The double burden of malnutrition in Indonesia
Cut Novianti Rachmi MD, MIPH
A thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy The University of Sydney 2018
Discipline of Child and Adolescent Health Sydney Medical School The University of Sydney
PAGE | 1
Authorship Statement
This thesis consists of three inter-related studies, which are presented in Chapters 2 to 9. The results of each study were initially prepared as manuscripts that have been either published or submitted for publication in a peer-reviewed journal at the time of thesis submission. I was the primary author of all seven of these manuscripts. Under my supervisors’ guidance, I was responsible for the following in each study:
Study 1: Systematic reviews (Chapter 2) I took the key role in all aspects of the study, including identification of study objectives, screening of eligible studies, identification of tools to critically appraise the eligible studies, the critical appraisal process, data extraction, analysis and interpretation, and drafting of the manuscript.
Part of Chapter 2 of this thesis is published as: Rachmi CN, Li M, Baur LA. Overweight and obesity in Indonesia: prevalence and risk factors – a literature review. Public Health. 2017; 147: 20-29.
Part of Chapter 2 of this thesis is submitted for publication: Rachmi CN, Li M, Baur LA. The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature. Submitted to International Journal of Public Health (under review).
Study 2: Quantitative studies (Chapters 4,5, and 6) I contributed to identifying the study questions, and had primary responsibility for cleaning the primary dataset for the secondary data analysis, conducting the statistical analyses, interpreting the findings, and drafting of the manuscripts.
Chapter 4 of this thesis is published as: Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight and overweight in children aged 2·0-4·9 years in Indonesia: prevalence trends and associated risk factors. PLoSOne. 2016; 11(5):e0154756.
Chapter 5 of this thesis is published as:
PAGE | 5
Table of contents
Title……………………………………………………………………………………….. 1 Statement of originality…………………………………………………………………… 2 Supervisor’s statement……………………………………………………………………. 3 Candidate’s statement…………………………………………………………………….. 4 Authorship statement……………………………………………………………………... 5 Table of contents………………………………………………………………………….. 8 Acknowledgements……………………………………………………………………….. 13 Publications arising from this thesis, awards and publications arising during candidature. 16 Abstract…………………………………………………………………………………… 20 List of abbreviations……………………………………………………………………… 22 Ethics clearance…………………………………………………………………………... 24 List of tables……………………………………………………………………………… 25 List of figures……………………………………………………………………………... 29 List of appendices………………………………………………………………………… 32 Chapter 1: Thesis overview and aims………………………………………………….. 33 1.1 Introduction…………………………………………………………………… 34 1.2 Research aims and objectives………………………………………………… 34 1.3 Overview of thesis……………………………………………………………. 35 Chapter 2: Review of the literature…………………………………………………….. 37 2.1 Definition, causes, and consequences of the double burden of malnutrition…. 38 2.1.1 Undernutrition………………………………………………………. 39 2.1.1.1 Stunting…………………………………………………… 39 2.1.1.2 Underweight………………………………………………. 42 2.1.2 Overnutrition………………………………………………………... 44 2.1.2.1 Overweight/ obesity………………………………………. 44 2.1.3 The double burden of malnutrition…………………………………. 46 2.2 Interventions/ policies on malnutrition……………………………………….. 47 2.3 References…………………………………………………………………….. 52 The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature…………………………. 56 Introduction……………………………………………………………………….. 58 Methods…………………………………………………………………………… 58
PAGE | 8 Inclusion and exclusion criteria…………………………………………... 58 Literature search process………………………………..………………… 59 Critical appraisal to ensure quality……………………………………….. 60 Definitions and cut off points for malnutrition…………………………… 60 Results…………………………………………………………………………….. 60 Characteristics of included studies……………………………………….. 60 Level of double burden of malnutrition…………………………………... 61 The double burden of malnutrition in each country………………….…… 63 Discussion………………………………………………………………………… 78 References………………………………………………………………………… 82 Overweight/obesity in Indonesia: prevalence and risk factors – a literature review……... 91 Introduction………………………………………………………..……………… 93 Methods…………………………………………………………………………… 93 Inclusion and exclusion criteria…………………………………………... 93 Literature search strategy…………………………………………………. 93 Screening of documents…………………………………………………... 93 Critical appraisal process…………………………………………………. 93 Definitions of overweight/ obesity………………………………………... 93 Results…………………………………………………………………………….. 93 Characteristics of included studies………………………………………... 93 Prevalence………………………………………………………………… 95 Risk factors……………………………………………………………….. 97 Discussion………………………………………………………………………… 99 References………………………………………………………………………… 100 Chapter 3: Quantitative studies………………………………………………………… 103 3.1 Overview and aims…………………………………………………………… 104 3.2 Methods……………………………………………………………………….. 105 3.2.1 Indonesian Family Life Survey (IFLS).……………………………. 105 3.2.2 Funding bodies and ethics approval for IFLS………………………. 106 3.2.3 Information collected and participants……………………………… 106 3.2.4 Data analysis approaches…………………………………………… 108 3.3 References……………………………………………..……………………… 110 Chapter 4: Stunting, underweight and overweight in children aged 2.0-4.9 years in Indonesia: prevalence trends and associated risk factors…………………………….. 111 Introduction……………………………………………………..………………… 113
PAGE | 9 Methods…………………………………………………………………………… 113 Indonesian Family Life Survey…………………………………………… 113 Outcome variables (anthropometric calculations) ……………………….. 114 Potential risk factors……………………………………………………… 114 Statistical analysis………………………………………………………… 115 Results…………………………………………………………………………….. 116 Characteristics of participants…………………………………………….. 116 Prevalence of stunting, underweight and ‘at risk of overweight’ and overweight/ obesity……………………………………………………….. 116 Associated risk factors……………………………………………………. 118 Discussion………………………………………………………………………… 120 Statement of principal findings…………………………………………… 120 Strengths and limitations………………………………………………….. 124 Comparison with other studies……………………………………………. 124 Implications for research, policy and practice……………………………………. 125 References………………………………………………………………………… 125 Chapter 5: Stunting coexisting with overweight in 2.0-4.9-year-old Indonesian children: prevalence, trends and associated risk factors from repeated cross- sectional surveys…………………………………………………………………………. 130 Introduction……………………………………………………………………….. 131 Methods…………………………………………………………………………… 132 Indonesian Family Life Survey…………………………………………… 132 Potentially associated risk factors………………………………………… 132 Statistical analysis………………………………………………………… 133 Results…………………………………………………………………………….. 133 Characteristics of participants…………………………………………….. 133 Prevalence of and risk factors for concurrent stunting and overweight…... 133 Odds of stunted children being overweight………………………………. 135 Discussion………………………………………………………………………… 135 Conclusions……………………………………………………………………….. 138 References………………………………………………………………………… 139 Chapter 6: Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? ……………………………………. 141 Introduction……………………………………………………………………….. 142 Methods…………………………………………………………………………… 143
PAGE | 10 Indonesian Family Life Survey…………………………………………… 143 Anthropometry and blood pressure calculations………………………….. 144 Statistical analysis………………………………………………………… 145 Results…………………………………………………………………………….. 145 Characteristics of participants…………………………………………….. 145 Stunted children and overweight/obesity in adolescence: 14-year cohort analysis……………………………………………………………………. 145 Stunted children and thinness in adolescence: 14-year cohort analysis...... 145 Stunted children and overweight/obesity in adolescence: 7-year cohort analyses…………………………………………………………………… 145 Prevalence ratio of stunted children being overweight/obese…………….. 145 Stunted children and high blood pressure………………………………… 147 Discussion………………………………………………………………………… 149 References………………………………………………………………………… 150 Chapter 7: Qualitative studies………………………………………………………….. 155 7.1 Overview and aims…………………………………………………………… 156 7.2 Methods……………………………………………………………………..… 156 7.2.1 Location and number of focus group discussion…………………… 156 7.2.2 Ethics approval……………………………………………………… 157 7.2.3 Participants………………………………………………………….. 157 7.2.4 Data analysis approaches…………………………………………… 157 7.3 References…………………………………………………………………….. 158 Chapter 8: Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study………………………………………………………………………… 159 Background……………………………………………………………………….. 161 Methods…………………………………………………………………………… 161 Study design and location………………………………………………… 161 Participants and recruitment process……………………………………... 161 Research team, data collection and approach to analysis………………… 161 Credibility and trustworthiness…………………………………………… 162 Results…………………………………………………………………………….. 164 Characteristics of participants…………………………………………….. 164 Emerging categories………………………………………………………. 164 Concept of overweight……………………………………………………. 164
PAGE | 11 Factors contributing to overweight……………………………………….. 165 Awareness and feelings towards overweight in children…………………. 168 Discussion………………………………………………………………………… 170 Conclusions……………………………………………………………………….. 171 Recommendations………………………………………………………………… 171 References………………………………………………………………………… 172 Chapter 9: Food choices made by primary carers (mothers/ grandmothers) in West Java, Indonesia…………………………………………………………………………... 174 Introduction……………………………………………………………………….. 176 Participants, material and methods…….………………………………………… 176 Study design, location, and participants………………………………….. 176 Research team, data collection and approach to analysis………………… 178 Triangulation and quality assurance……………………………………… 179 Results…………………………………………………………………………….. 179 Characteristics of participants…………………………………………….. 179 Emerging themes…………………………………………………………. 180 Discussion………………………………………………………………………… 191 Conclusion………………………………………………………………………... 194 References………………………………………………………………………… 196 Chapter 10: Conclusions and recommendations……………………………………… 199 10.1 Key findings…………………………………………………………………. 200 10.2 Future directions…………………………………………………………….. 201 10.2.1 Future directions based on findings presented in this thesis………. 202 10.2.1.1 Policies/ implementation………………………………… 203 10.2.1.2 Future research…………………………………………... 206 10.2.2 Additional recommendations to be considered by all levels of government in Indonesia………………………………………………….. 207 10.3 References…………………………………………………………………… 208 Appendix………………………………………………………………………………… 209
PAGE | 12
Acknowledgements
Thank you Allah, SWT. I know that You know how thankful I am for everything.
I undertook my medical school training in Indonesia and then my Masters and PhD degrees here at the University of Sydney. Let me tell you one thing – the PhD is the hardest to finish, yet the most enjoyable. I can proudly say that I have done much hard work; however, there are many people who eased my burden along the way. Let me use this space to acknowledge them.
My scholarship provider: Lembaga Pengelola Dana Pendidikan (LPDP), Ministry of Finance, The Republic of Indonesia. Thank you for funding me to achieve one of my goals in life.
My wonderful supervisors: Professor Louise Baur, Professor Mu Li, Dr Kingsley Agho and Dr Cynthia Hunter. There is absolutely no way to describe how I really feel about these teachers, but I hope these few words can describe how much I respect and treasure them. They are the ones that I give my highest respect to.
Professor Louise Baur, the one who always teaches by example. You have been a great role model, an excellent teacher, a terrific supervisor, and an amazing friend. You showed me how character, strength, and perseverance are just as important as the power of knowledge. I can assure you that I will always have a “fire in my belly”, Prof Baur.
Professor Mu Li, the one who taught me about tenacity and how to balance my role as a wife, a mother, and a scholar. I have admired her since I did my Masters degree back in 2008 and today, my admiration has only grown stronger. I am so lucky to have you as my second supervisor, no student could ask for a better combination.
Dr Kingsley Agho, the one who would not stop torturing me in my first few months of learning about STATA, only to ensure I have learnt everything I needed to know to do it on my own. I will forever value the friendship and I can beat you any day in the Chilli Challenge, Sir.
PAGE | 13 Dr Cynthia Hunter, the one that literally honed my qualitative research skills and ensured I did the best I could in each step of the qualitative study. I have learnt so much during the last two years, and I thank you for that.
Denise Yuille and Diane Hanlon, thank you for making my life as a student much easier. What would I do without the both of you?
Friends that started the journey with me and ended it together: Korri, Leonny, Cokorda, and Andri. I am going to miss our lunches, discussions, and ‘laughing at ourselves’ moments. I shall see you guys on the top of the stairs years from now.
Those I only met for a short time in my early years of the PhD journey yet held an important role in my study: Dr Sarah Garnett and Dr Mandy Ho. Thank you.
My parents (Erwin Djailani and Tuti Nuriah), parents in law (Moenawar and Andjar Pandarmaningsih), sisters and brothers, thank you for all the support and prayers. The youngest one is now officially a Doctor.
People come in to your life for a reason, a season, or a lifetime. These people came at the right time to save me from my most depressing moments. Hafizah Jusril, Arlene Junita, Rinaldi Ridwan, and Agus Lim, I cannot thank you guys enough for giving me the support I needed every single time. I love you guys. I hope you guys came to stay for a lifetime.
Dasapta Erwin Irawan, my husband, my partner, my colleague, my teacher, my best friend, and the love of my life. When I started this journey he said one sentence that he truly held until the moment I submitted this thesis: “Go ahead and get that PhD, I’ve got your back”. I love you. Thank you for being you and holding on to me in my lowest moment(s). Even at my worst, I’m best with you!
So what if I'm a woman. I do not want to be a slave of ignorance. – OST of Kartini movie, written by Melly Goeslaw (a famous Indonesian singer)
PAGE | 14 My daughter Khaira Salsabila Irawan, please hold on to those words until the end of time, as being a woman is a privilege. We can achieve anything we want in life. Anything! I love you so much.
My son Hasbi Kautsar Irawan, the PhD baby, without you this thesis might have been finished much earlier, but it would not be as much fun. Thank you for being part of the degree, and being such a nice baby both inside and outside my belly. I love you to the moon and back.
One of the reasons for me to undertake this PhD was to set an example for my daughter (back then there was only her) that nothing can hold you back from achieving your dreams: being a Muslim, being a woman, being a wife, being a mother, being older than other PhD students, coming from a poor family, or even coming from a developing country. As long as you are determined to achieve something, you can and you will.
My children, I hope I can read your thesis acknowledgement one day that says: “my mother, the one that led by example”. Go ahead and get your dreams, we’ve got your back.
If I have left anyone out, I am sure this page is not the only way I can thank him or her.
As a closing remark, I hope you enjoy reading my thesis, as much as I enjoyed writing it. This thesis is a way to let people know what I have been doing for the past 3 years, 10 months, 1 week and 5 days. But who is counting, right?
Best regards, Cut
PAGE | 15 Publications arising from this thesis, awards and publications arising during candidature
Publications arising from this thesis Papers in peer-reviewed journals Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight and overweight in children aged 2·0-4·9 years in Indonesia: prevalence trends and associated risk factors. PLoSOne. 2016; 11(5):e0154756.
Rachmi CN, Agho KE, Li M, Baur LA. Stunting coexisting with overweight in 2.0-4.9 year old Indonesian children: prevalence, trends and associated risk factors from four cross sectional surveys. Public Health Nutrition. 2016; 19 (15): 2698-707.
Rachmi CN, Agho KE, Li M, Baur LA. Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? International Journal of Public Health. 2016; 62:153-162.
Rachmi CN, Li M, Baur LA. Overweight and obesity in Indonesia: prevalence and risk factors – a literature review. Public Health. 2017; 147: 20-29.
Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. International Journal of Behavioral Nutrition and Physical Activity. 2017; 14(101); doi.org/10.1186/s12966-017-0556-1.
Papers submitted for publication/ undergoing review Rachmi CN, Hunter CL, Li M, Baur LA. Food choices made by primary carers (mothers/ grandmothers) in West Java, Indonesia. Submitted to Appetite (under review).
Rachmi CN, Li M, Baur LA. The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature. Submitted to International Journal of Public Health (under review).
PAGE | 16 Published abstracts or conference proceedings Rachmi CN, Agho KE, Li M, Baur LA. The double burden of malnutrition: a longitudinal, nationally representative study of Indonesian children. Australia and New Zealand Obesity Society (ANZOS) Conference, Sydney, Australia, 2014.
Rachmi CN, Agho KE, Li M, Baur LA. Concurrent stunting and obesity in children aged 2.0- 4.9 years in Indonesia: the double burden of malnutrition. European Congress on Obesity (ECO), Prague, Czech Republic, 2015.
Rachmi CN, Agho KE, Li M, Baur LA. Concurrent stunting and obesity in children aged 2.0 to 4.9 years in Indonesia: the double burden of malnutrition. Asian Congress on Nutrition (ACN), Tokyo, Japan, 2015.
Rachmi CN, Agho KE, Li M, Baur LA. Overweight, stunting and underweight in Indonesian children: what are the associated risk factors? Australia and New Zealand Obesity Society (ANZOS) Conference, Melbourne, Australia, 2015.
Rachmi CN, Agho KE, Li M, Baur LA. Concurrent stunting and overweight in 2.0-4.9 year old Indonesian children: data from four repeated cross-sectional surveys. International Congress on Obesity, Vancouver, Canada, 2016.
Rachmi CN, Agho KE, Li M, Baur LA. Childhood stunting and the risk of overweight/ obesity, thinness, and high blood pressure in adolescence: evidence from Indonesia. Asia Pacific Rim Universities (APRU) Global Health Conference, Sydney, Australia, 2016.
Rachmi CN, Agho KE, Li M, Baur LA. Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? The Obesity Society Annual Scientific Meeting, New Orleans, United States of America, 2016.
Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. IAAH – 11th World Congress on Adolescent Health, New Delhi, India, 2017.
PAGE | 17 Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. The Obesity Society Annual Scientific Meeting, Washington DC, United States of America, 2017.
Awards arising during candidature Sydney South East Asia Centre (SSEAC) Three Minute Thesis Competition 2016 – finalist.
Asia Pacific Rim Universities (APRU) Global Health Conference Sydney 2016 – Second place in Postgraduate Student Poster Competition.
Publications arising during candidature but not included in the thesis Peer-reviewed article Maulina T, Rachmi CN, Akhter R. The association between self-report of orofacial pain symptoms with age, gender, interference in activities, and socioeconomic factors surveyed in Indonesian community health centers. Asian Pacific Journal of Dentistry: APJD. 2014; 14(2):23-34.
Irawan DE, Priyambodho A, Rachmi CN, Wibowo D, Fahmi A. Bibliometric study to assist research topic selection: a case from research design on Jakarta’s groundwater (part 1). Research Ideas and Outcomes. 2016; 2: e9841; https://doi.org/10.3897/rio.2.e9841.
Irawan DE, Rachmi CN, Irawan H, Abraham J, Kusno K, Multazam MT, Rosada KK, Nugroho SH, Kusumah G, Holidin D, Aziz NA. Penerapan Open Science di Indonesia agar riset lebih terbuka, mudah diakses, dan meningkatkan dampak saintifik. Jurnal Berkala Ilmu Perpustakaan UGM. 2017; https://doi.org/10.22146/bip.17054.
Irawan DE, Priyambodho A, Rachmi CN, Wibowo DM. Looking at groundwater research landscape of Jakarta Basin for better water management. Journal of Physics: Conference Series. 2017; https://doi.org/10.1088/1742-6596/877/1/012033.
Preprints Irawan DE, Rachmi CN, Pratama A, Tulak GP, Rochman AD. Groundwater and river water interaction to solve water shortage: a case from Tasikmalaya, Indonesia. PeerJ PrePrints. 2014; 2:e720v1.
PAGE | 18 Book Irawan DE, Rachmi, CN. Menulis (Ilmiah) itu Menyenangkan (Book). 2016. ITB press; Bandung, Indonesia.
PAGE | 19
Abstract
Introduction The double burden of malnutrition (DBM) is characterised by overlapping conditions of under- and over-nutrition within one population, household, or even individual. Indonesia faces this situation. The overall aim of this thesis was to provide a better understanding of DBM in Indonesian children.
Methods This thesis used a mixed methods approach, combining a secondary data analysis from the Indonesian Family Life Survey (IFLS) and a qualitative study. Prevalence data for DBM, from Indonesian children aged 2.0-4.9 years at four different time points -1993, 1997, 2000, and 2007 - were calculated. In addition I undertook a longitudinal analysis of children aged 2.0 to 4.9 years at baseline (1993 and 2000) and their weight and blood pressure status 7 (2000 and 2007) and 14 years (2007) later. In the qualitative study I explored the perceptions of Indonesian mothers and grandmothers of under-five and elementary school-aged children towards child obesity and food choices. I conducted twelve focus group discussions in West Java with 94 carers from three different socioeconomic status groups.
Results DBM occurs in Indonesian children, with decreasing secular trends in stunting and underweight; and increasing trends in overweight/obesity. I found an inconsistent trend in the prevalence of concurrent stunting and overweight across 1993 to 2007. Although the cross sectional data showed that stunted children are more likely to be overweight/ obese, my longitudinal evidence showed the opposite. I found no association between stunting in early childhood and thinness, overweight/ obesity, or high blood pressure in the later years.
In the qualitative study, primary carers still define overweight subjectively. The “chubbier is healthier” concept may be a contributing factor, and even though some mothers realize that their children are overweight/ obese, denial of the effect of overweight/ obesity on their children’s physical ability is still apparent. There are complex reasons behind primary carers’ decision-making processes around their children’s food choices that may additionally contribute to the increased prevalence of overweight/ obesity.
PAGE | 20
Conclusion These results provide a better understanding of DBM in Indonesia. Such information is likely to reveal potential strategies to tackle this issue both in Indonesia, and potentially in other low- and middle-income countries.
PAGE | 21 List of abbreviations
ASEAN Association of South East Asian Nations BMI Body mass index BMIZ BMI-for-age z score CDC United States Centers for Disease Control and Prevention CI Confidence intervals CHW Children’s Hospital at Westmead DBM Double burden of malnutrition DBP Diastolic blood pressure FGD Focus group discussions GLLAMM Generalised linear latent and mixed models HAZ Height-for-age z score HDI Human development index IFLS Indonesian Family Life Survey HBP High blood pressure HREC Human Research Ethics Committee IOTF International Obesity Task Force KMS Kartu menuju sehat (health card) LMIC Low and middle-income countries NCD Non communicable disease NCHS/WHO National Center for Health Statistics/World Health Organization NTB Nusa Tenggara Barat OR Odds ratio PE Physical education Posyandu Pos pelayanan terpadu (community health posts) PKK Pembinaan kesejahteraan keluarga (association of family activities) PR Prevalence ratio Puskesmas Pusat kesehatan masyarakat (community health centers) PUGS/ PGS Pedoman umum gizi seimbang/ pedoman gizi seimbang (balanced nutrition guidelines) Riskesdas Riset kesehatan dasar (basic national health survey) SBP Systolic blood pressure SD Standard deviation
PAGE | 22
SEA South East Asia SES Socio economic status SUSENAS Survei sosioekonomi nasional (national socioeconomic survey) SUN Scaling Up Nutrition Svy Survey commands UNICEF United Nations Children’s Fund WAZ Weight-for-age z score WHO World Health Organization
PAGE | 23
Ethics clearance
The quantitative studies in chapters 4, 5, and 6 (secondary data analysis of the Indonesian Family Life Survey [IFLS] dataset) did not require further ethics clearance. The original ethics clearance for the IFLS was obtained from Institutional Review Boards in the USA (at Rand Corporation, Santa Monica, California) and in Indonesia (at Universitas Indonesia for waves 1 and 2 and at Universitas Gadjah Mada for subsequent waves).
The Human Research Ethics Committee from the Faculty of Medicine, Universitas Padjadjaran, in Bandung, Indonesia, approved the qualitative study in chapters 8 and 9 (473/UN6.C1.3.2/KEPK/PN/2016).
PAGE | 24 List of tables
Table Title Page
Chapter 2 – Literature review
Table 2.1 Definition and cut off points of undernutrition used throughout this 39 thesis according to the WHO Child Growth Standards for 0 to 5 years
Table 2.2 Definition and cut off points of overnutrition used throughout this 44 thesis according to the WHO Child Growth Standards for 0 to 5 years
The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature
Table 1 The double burden of malnutrition in Cambodia and Indonesia; 66 Country order alphabetically, study order by year of publication
Table 2 The double burden of malnutrition in Lao PDR, Malaysia, 71 Myanmar, and Philippines; Country order alphabetically, study order by year of publication
Table 3 The double burden of malnutrition in Thailand, and Vietnam; 75 Country order alphabetically, study order by year of publication
Supplementary Guidelines for critically appraising studies of prevalence or 86 Table 1 incidence of a health problem by Loney at al
Supplementary Criteria and cut-off points used in this literature review 87 Table 2
Supplementary Critical appraisal process using Loney et al.; Country order 88 Table 3 alphabetically, study order by year of publication
Overweight/obesity in Indonesia: prevalence and risk factors – a literature review
Table 1 Definitions of overweight/obesity, order by year of publication and 94 alphabetically within the same year
Table 2 Summary of critical appraisal of studies included in the review, 95 order by year of publication and alphabetically within the same year
Table 3 Summary of results from included studies; order by year of 96 publication and alphabetically within the same year
PAGE | 25 Table 4 Risk factors associated with overweight/obesity; order by year of 98 publication and alphabetically within the same year
Supplementary List of sources of the literature searching process 102 Table 1
Supplementary Guidelines for critically appraising studies of prevalence or 102 Table 2 incidence of a health problem by Loney at al
Chapter 3 – Quantitative studies
Table 3.1 List of modules collected in the household questionnaires of the 107 Indonesian Family Life Survey in 1993
Chapter 4 – Stunting, underweight and overweight in children aged 2.0-4.9 years in Indonesia: prevalence trends and associated risk factors
Table 1 Characteristics of children and parents in each wave of the 117 Indonesian Family Life Survey, n (%) or mean (standard deviation)
Table 2 Comparison of the prevalence of stunting, underweight and ‘at risk’ 119 and overweight/obesity within different variables (%, 95% CIs) (n=4,101)
Table 3 Unadjusted and adjusted odds ratios (OR) (95%CI) of potential risk 122 factor for stunting, underweight and ‘at risk’ and overweight/obesity (n=4,101)
Supplementary Adjusted odds ratios (95% confidence intervals) for the complete 129 Table 1 data, 5 and 10 imputation data sets (M=5 and M=10)
Chapter 5 - Stunting coexisting with overweight in 2.0-4.9-year- old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys
Table 1 Characteristics of children and parents in each wave (wave 1, 1993; 134 wave 2, 1997; wave 3, 2000; wave 4, 2007) of the Indonesian Family Life Survey
Table 2 Prevalence of concurrent stunting and overweight among children 136 aged 2.0-4.9 years (n 4101), Indonesian Family Life Survey
Chapter 6 – Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence?
Table 1 Characteristics of children aged 2.0-4.9 years in waves 1 (1993) and 146 followed up to wave 4 (2007): Prevalence (n [%]) and mean (95% Confidence intervals [CI])
PAGE | 26 Table 2 14-year cohort analysis: 765 children aged 2.0-4.9 years in 1993 and 147 their weight status (overweight/obesity and thinness) in 2007, according to height status in 1993
Table 3 Two cohort analyses of children age 2.0-4.9 years in baseline and 148 their overweight status 7 years later: 1,083 children in 1993 and 1,589 children in 2000
Electronic Comparison of baseline demographics between children aged 2.0- 152 supplemental 4.9 years who completed 4 waves of the survey 1993-2007 (n=765) material Table and those who only completed the first wave in 1993 (n=620); Data 1 from Indonesian Family Life Survey, Indonesia, 1993-2007
Electronic Characteristics of 1083 children aged 2·0-4·9 years in wave 1 153 supplemental (1993) followed up to wave 3 (2000) [7 years cohort-1]; Prevalence material Table (n [%]) and mean (95% Confidence Intervals [CI]). Data from 2 Indonesian Family Life Survey, Indonesia, 1993-2000
Electronic Characteristics of 1589 children aged 2·0-4·9 years in wave 3 154 supplemental (2000) followed up to wave 4 (2007) [7 years cohort-2] of the material Table Indonesian Family Life Survey, Indonesia, 2000-2007; Prevalence 3 (n [%]) and mean (95% Confidence Intervals [CI]). Data from Indonesian Family Life Survey, Indonesia, 2000-2007
Chapter 8 - Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school- aged children in Bandung, Indonesia: a qualitative study
Table 1 Questions used in focus group discussions in Greater Bandung 163 Area, Indonesia
Table 2 Characteristics of participants from 12 focus group discussions in 165 Greater Bandung Area, Indonesia.
Table 3 Category and subcategories on the perception of overweight carers 165 in Greater Bandung Area, Indonesia
Table 4 Quotes on concept of overweight according to carers from Bandung, 166 Indonesia
Table 5 Quotes on factors contributing to weight according to carers from 167 Bandung, Indonesia
Table 6 Quotes on awareness and feelings towards overweight in children 169 according to carers from Bandung, Indonesia
Chapter 9 – Food choices made by primary carers (mothers/ grandmothers) in West Java, Indonesia
Table 1 Characteristics of participants of 12 focus group discussions in 180 Greater Bandung Area, Indonesia
PAGE | 27 Table 2 Quotes on decision on food served at home and attitudes toward 183 instant noodles consumption according to carers in West Java, Indonesia
Table 3 Quotes on decisions on food/ snacks and drinks consumptions and 187 feelings about food/ snacks and drinks consumptions according to carers in West Java, Indonesia
Table 4 Quotes on considerations of money issues and Halal issues 190 according to carers in West Java, Indonesia
PAGE | 28 List of figures
Figure Title Page
Chapter 1 - Thesis overview and aims
Figure 1 Overview of studies undertaken in this thesis 36
Chapter 2 - Literature review
Figure 2.1 The concept and different levels of the double burden of malnutrition 38
Figure 2.2 World Health Organization’s The Double Burden of Malnutrition 39 Policy Brief
Figure 2.3 Prevalence of stunting in under 5 year children based on joint estimates 40 by UNICEF, WHO, and World Bank Group in 2017
Figure 2.4 Three Indonesian children aged 9 years 42
Figure 2.5 Determinants of child undernutrition according to UNICEF 43
Figure 2.6 Prevalence of overweight in under 5 year children based on joint 45 estimates by UNICEF, WHO, and World Bank Group in 2017
Figure 2.7 Causes of the double burden of malnutrition 47
Figure 2.8 Four platforms to overcome stunting 48
Figure 2.9 Five platforms to stop the increasing prevalence of childhood 49 overweight
Figure 2.10 World Health Organization: importance of tackling the double burden 50 of malnutrition – The Double Burden of Malnutrition Policy Brief
Figure 2.11 World Health Organization: Double-duty Actions – Double-duty 51 Actions for Nutrition Policy Brief
The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature
Figure 1 A flow diagram showing the flow of citations reviewed in this study 61
Figure 2 Different levels of the double burden of malnutrition in 8 ASEAN 63 countries
PAGE | 29 Overweight/obesity in Indonesia: prevalence and risk factors – a literature review
Figure 1 Literature searching process on the prevalence and risk factors of 94 overweight/obesity in Indonesia
Chapter 3 – Quantitative studies
Figure 3.1 Studies conducted using the Indonesian Family Life Survey waves 1 104 (1993), 2 (1997), 3 (2000), and 4 (2007)
Figure 3.2 Conceptual framework for analysis of risk factors related to 109 malnutrition in Indonesian children aged 2.0 to 4.9 years; Data from the Indonesian Family Life Survey
Chapter 4 – Stunting, underweight and overweight in children aged 2.0-4.9 years in Indonesia: prevalence trends and associated risk factors
Figure 1 Distribution of BMI z-score and its mean (vertical line) of 2.0-4.9 years 121 children of the Indonesian Family Life Survey (IFLS) wave 1 (1993), wave 2 (1997), wave 3 (2000), and wave 4 (2007)
Chapter 5 - Stunting coexisting with overweight in 2.0-4.9-year-old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys
Figure 1 Odds Ratios, with their 95% confidence intervals represented by 137 horizontal bars, of stunted children aged 2.0-4.9 years being overweight; wave 1 (1993), 2 (1997), 3 (2000), and 4 (2007) of the Indonesian Family Life Survey
Chapter 6 – Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence?
Figure 1 The number of participants in each original cohort and the final number 144 of eligible participants in the analysis; Data from Indonesian Family Life Survey, Indonesia wave 1 (1993), 2 (1997), wave 3 (2000 and wave 4 (2007)
Figure 2 Adjusted prevalence ratio (PR), 95% confidence intervals (CI), and P 149 values of stunted children aged 2.0-4.9 years at baseline to being overweight/obese 14 or 7 years later; Data from the Indonesian Family Life Survey, Indonesia, 1993-2007
Figure 3 Height status of 765 Indonesian children age 2.0-4.9 years in wave 1 149 (1993) and the prevalence (95% confidence intervals) of high systolic and diastolic blood pressure status in wave 4 (2007); Data from Indonesian Family Life Survey, Indonesia, 1993 and 2007
PAGE | 30 Chapter 8 - Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study
Figure 1 Numbers and participants of focus group discussions in Bandung, 162 Indonesia
Additional Kartu Menuju Sehat (KMS- Health Card) for boys aged 0–24 months 173 file 1
Chapter 9 – Food choices made by primary carers (mothers/ grandmothers) in West Java, Indonesia
Figure 1 Numbers and participants of focus group discussions in Bandung 177 Greater Area, West Java, Indonesia
Chapter 10 – Conclusions and recommendations
Figure 10.1 Future directions and recommendations for Indonesia 202
Figure 10.2 How can government support healthy food preferences? 204
Figure 10.3 Balanced Nutrition Guidelines 206
PAGE | 31 List of appendices
Appendix Title Page Appendix 1 Ethics approval from the Ethics Committee at Fakultas Kedokteran 209 Universitas Padjadjaran, Bandung
PAGE | 32
Chapter 1: Thesis overview and aims
1.1 INTRODUCTION 34 1.2 RESEARCH AIMS AND OBJECTIVES 34 1.3 OVERVIEW OF THESIS 35
PAGE | 33
Chapter 1: Thesis overview and aims
1.1 INTRODUCTION Indonesia has been dealing with two main types of undernutrition, namely stunting and underweight, for decades. The Indonesian government has implemented many policies on intervention and treatment of undernutrition, such as Raskin (free rice for the poor), Kadarzi (nutrition-conscious family) and Pemberian Makanan Tambahan (free supplementary feeding) for children. However, in the past decade, Indonesian government has begun to acknowledge the other end of the spectrum of malnutrition, overnutrition. Indonesian government policies and program currently deal with these two types of malnutrition as different conditions.
Although several studies of obesity prevalence in Indonesia have been published, there are many evidence gaps in monitoring prevalence trends of overweight/ obesity, especially in children. Additional issues include the lack of information about concurrent under- and over- nutrition occurring within the same individuals, as well as the effect of early under nutrition on later nutritional status. Finally, there is a lack of information about community understanding of obesity and the double burden of malnutrition in children in Indonesia. Such information would be a crucial step to understanding the best strategies to overcome this issue.
A narrative review of the existing literature on the double burden of malnutrition in South East Asia, along with literature on overweight/ obesity in the Indonesian context, is available in chapter 2.
1.2 RESEARCH AIMS AND OBJECTIVES This thesis aims to provide further information about the double burden of malnutrition in Indonesian children, with a specific focus to a combination of stunting and overweight/ obesity. The specific objectives of this thesis are to: 1. Document the existing literature on the double burden of malnutrition in South East Asian countries. 2. Document the existing literature on overweight/ obesity in Indonesia. 3. Provide information on the prevalence trends and associated risk factors for stunting, underweight, and overweight/ obesity in Indonesian children aged 2.0-4.9 years at four different time points: 1993, 1997, 2000, and 2007.
PAGE | 34
4. Provide information on the prevalence trends and associated risk factors for concurrent stunting and overweight/ obesity in Indonesian children aged 2.0-4.9 years at four different time points: 1993, 1997, 2000, and 2007. 5. Investigate the likelihood of stunted children of being overweight or obese compared to their healthy height peers. 6. Investigate whether previously stunted children in Indonesia are more likely to become overweight, thin, or have high blood pressure in later years. 7. Explore the perceptions of Indonesian mothers and grandmothers towards child obesity and related issues. 8. Explore the reasons behind mothers’/ grandmothers’ decision-making around food choices in West Java, Indonesia.
1.3 OVERVIEW OF THESIS A mixed-methods approach was conducted to answer the study objectives: 1. I conducted systematic reviews to answer objectives 1 and 2. Both of the systematic reviews looked at papers published from 1990 to 2016 and hence pick up on my published manuscripts. 2. I conducted cross sectional analyses to answer study objectives 3, 4 and 5. These cross sectional studies are secondary data analysis of the Indonesian Family Life Survey (IFLS) from wave 1 (1993) to wave 4 (2007). 3. I conducted a longitudinal analysis to answer study objective 6. This is also a secondary data analysis of the IFLS dataset, following children in separate 7- and 14- year cohorts. 4. I conducted a qualitative study to answer study objective 7 and 8. Twelve focus group discussions were conducted in three different districts in Greater Bandung Area, West Java, Indonesia.
The methods used in both of the systematic reviews are elaborated in Chapter 2. Chapter 3 provides detailed information on the quantitative studies reported in Chapters 4 to 6. Chapter 7 provides information on the qualitative study whose results are presented in Chapters 8 and 9. Chapter 10 provides a final summary of the results presented in the previous chapters along with a discussion on the implications and future research questions that might arise based on the findings. An overview of all the studies undertaken in this PhD thesis is available in Figure 1.1.
PAGE | 35 CHAPTER 2 LITERATURE REVIEW
The double burden of malnutrition in ASEAN countries - a Overweight and obesity in Indonesia: prevalence and risk Definition, causes, and consequences of malnutrition comprehensive review of the literature factors ? a literature review
Interventions/ policies on malnutrition Aim: to review the DBM in Association of South East Asian Nations Aim: to review published data on the prevalence of overweight and (ASEAN) countries, including levels (population [including country, obesity in Indonesian adults and children and adolescents, and the city, or any community], household, or individual), types (the type of reported associated risk factors. undernutrition and overweight), and prevalence.
CHAPTER 3 METHODS OF QUANTITATIVE STUDIES CHAPTER 7 METHODS OF QUALITATIVE STUDIES
CROSS-SECTIONAL STUDIES LONGITUDINAL STUDIES
CHAPTER 4 CHAPTER 5 CHAPTER 6 CHAPTER 8 CHAPTER 9 Stunting, underweight and Stunting coexisting with Are stunted young Indonesian Perceptions of overweight by Food choices made by primary overweight in children aged overweight in 2·0?4·9-year-old children more likely to be primary carers (mothers/ carers (mothers/ grandmothers) 2.0?4.9 years in Indonesia: Indonesian children: prevalence, overweight, thin, or have high grandmothers) of under five and West Java, Indonesia. prevalence trends and associated trends and associated risk blood pressure in adolescence? elementry school-aged children risk factors factors from repeated i n B a ndung, Indonesia: a Aim: c r o s s - s e ctional surveys Aims: qualitative study To explore the reasons behind Aims: T o d e t e r mine whether stunted m o t h e r s ?/ grandmothers? 1. To determine the temporal trends Aims: young children are at greater risk, in Aim: decision-making around food in the prevalence of underweight, 1. To describe the prevalence and adolescence, of To explore the perceptions of choices in West Java, Indonesia. stunting, and at risk of overweight/ trends of concurrent stunting and 1. Overweight/ obesity Indonesian mothers and overweight or obesity in Indonesian overweight in young Indonesian 2. Thinness grandmothers from different children aged 2.0-4.9 years at four children between 1993 to 2007 3. High blood pressure. socio-economic groups towards different time points: 1993, 1997, 2. To identify potential risk factors child overweight/ obesity and 2000, and 2007. associated with the phenomenon related issues. 2. To examine associated risk 3. To determine whether stunted factors. children are at greater risk of being overweight or obese compared to their healthy height peers.
CHAPTER 10 CONCLUSIONS AND RECOMMENDATIONS
Figure 1.1 Overview of studies undertaken in this thesis
PAGE | 36 Chapter 2: Review of the literature
2.1 DEFINITION, CAUSES, AND CONSEQUENCES OF THE DOUBLE 38 BURDEN OF MALNUTRITION 2.1.1 Undernutrition 39 2.1.1.1 Stunting 39 2.1.1.2 Underweight 42 2.1.2 Overnutrition 44 2.1.2.1 Overweight/ obesity 44 2.1.3 The double burden of malnutrition 46 2.2 INTERVENTIONS/ POLICIES ON MALNUTRITION 47 2.3 REFERENCES 52
PAGE| 37
Chapter 2: Review of the literature
2.1 DEFINITION, CAUSES, AND CONSEQUENCES OF THE DOUBLE BURDEN OF MALNUTRITION The double burden of malnutrition (DBM), a concept first introduced in 1992 at the International Conference on Nutrition1, is defined as the coexistence of any form of undernutrition with overnutrition (Figure 2.1)1-4. Much research has documented that this condition occurs at both the population level (in the same country, city, or community) and the household level (combination of undernutrition and overnutrition within members of a family)3-13. There is less published data that this condition may occur within the same individual, either simultaneously or at different stages in life3,8,13-15. It has also been proposed that undernutrition in early childhood may increase the risk of overnutrition in later life1,2.
POPULATION LEVEL HOUSEHOLD LEVEL INDIVIDUAL LEVEL
Stunting
Wasting Undernutrition Underweight
Micronutrient The double deficiencies burden of malnutrition At risk of overweight/ obesity
Overnutrition Overweight
Obesity
Figure 2.1 The concept and different levels of the double burden of malnutrition
The WHO has recently introduced the following infographic that further expands on this definition as shown in Figure 2.2.
PAGE | 38
Figure 2.2 World Health Organization’s The Double Burden of Malnutrition Policy Brief3. Available from: http://www.who.int/nutrition/publications/doubleburdenmalnutrition- policybrief/en/
2.1.1 Undernutrition There are several types of undernutrition, including stunting, wasting, underweight, or having any micronutrient deficiencies. Throughout this thesis, I will focus on two specific types of undernutrition - stunting and underweight.
Stunting is defined as having low length/ height for age. It is a form of chronic malnutrition that is usually fully established by the age of two years; therefore adult height is essentially determined at age two years16-18. Throughout this thesis, a child is considered to be stunted when his/ her height for age z score (HAZ) is less than -2 using a classic definition linked to the WHO Child Growth Standard19 as stated in Table 2.1.
Table 2.1 Definition and cut off points of undernutrition used throughout this thesis according to the WHO Child Growth Standards for 0 to 5 years (http://www.who.int/childgrowth/en/)
Category Type Cut off points Undernutrition Stunting Height for age z score <-2 Underweight Weight for age z score <-2
Underweight is defined as having low weight according to the child’s age. This thesis uses the WHO Growth Standard definition of underweight as having weight for age z score (WAZ)<- 2 (Table 2.1)19.
2.1.1.1 Stunting According to the 2014 World Health Organization (WHO) Stunting Policy Brief, more than one half (56%) of stunted children in the world live in Asia20. A joint estimates by UNICEF, WHO, and World Bank Group, published in 2017, showed that the prevalence of stunting in
PAGE | 39 under 5 year children was 22.9% globally and 25.8% in Southeast Asia21 (Figure 2.3). From 11 countries in the WHO South East Asia Region, the top four countries with the highest prevalence of stunting in under five children were Timor Leste, India, Nepal and Indonesia22.
Figure 2.3 Prevalence of stunting in under 5 year children based on joint estimates by UNICEF, WHO, and World Bank Group in 201721. Available from http://www.who.int/nutgrowthdb/estimates2016/en/
According to the World Health Organization conceptual framework on childhood stunting, there are four major issues that contribute to the development of stunting: household and family factors, inadequate complementary feeding, breastfeeding and infections23.
The household and family factors consist of two issues, maternal factors and the home environment. Maternal factors consist of a range of issues such as the health status (including the nutritional status) of the mother before conception, during pregnancy, and during lactation. Other issues are short stature in mothers, adolescent pregnancy, poor mental health and short birth spacing between two children. The health status of the mother influences the health status of the baby, as underweight mothers are more likely to have babies who experienced intrauterine growth restriction which in turn is a risk factor for stunting in early childhood23-25. The home environment includes but is not limited to issues such as food insecurity within the household, intra-household food allocation, and inadequate water and sanitation supply20,23.
PAGE | 40 Inadequate complimentary feeding includes three main issues: poor food quality, inadequate feeding practices, and food and water safety. Poor food quality is associated with a lack of dietary diversity as well as low intake of animal-source protein. This is also worsened by foods that are reduced in micronutrient and energy content. Inadequate feeding practices include reduced frequency and quantity of food, especially during and after an episode of infection/ illness. Food and water safety issues consist of a range of factors such as poor hygiene practices, contaminated water and food and also poor food preparation and storage23.
The third issue that contributes to stunting is inadequate breastfeeding practices, e.g. delayed onset, non-exclusive breastfeeding and early cessation of breastfeeding20,23. The last issue relates to clinical and subclinical infections, especially enteric, diarrhoeal, helminthic, and respiratory infections20,23. The severity, duration, and recurrence of infections are all important20. Recurrent infections and stunting are often described as a vicious circle, as recurrent infections might lead to stunted development and being stunted also increases the risk of a child acquiring an infectious disease (due to decreasing immune functions), mainly diarrhoea and pneumonia26-29.
Stunting is associated with a range of health, developmental and economic consequences, both in the short-term, and long-term23,30. The short-term consequences include higher mortality and morbidity, lower resistance to infectious diseases, and under-developed cognitive, motor, and language development23. Long-term consequences include poorer school progression and fewer completed grades of schooling, an increased risk for later degenerative diseases, and a greater risk for later obstetric complications in females due to the smaller pelvis1,17,23,24,31. Stunted children in early childhood who have rapid weight gain or who are exposed to high fat diets after the age of 2 years have a higher risk of becoming overweight/ obese in later life32-34. This has been linked to the later development of non- communicable diseases such as diabetes, stroke, hypertension and heart disease2,24. Stunting can occur as part of another vicious circle, as stunted children will grow to be stunted adults which in turn is a risk factor in women for having a stunted child 17,24.
To reduce the risk of stunting, several interventions have been taken into consideration as outline further below in Interventions/ Policies on Malnutrition section.
In conclusion, stunting is considered a major health problem for several reasons: 1) stunting affects many children in the world, 2) stunting has many health and functional consequences, 3) it is defined by one robust, internationally recognised standard, 4) it can be modified PAGE | 41 through an optimal intervention in the first 1000 days of life (from a mother’s pregnancy to a child’s second birthday), and 5) it requires multi sectoral interventions30. Figure 2.4 shows three Indonesian children aged 9 years where one of them is stunted.
Figure 2.4 Three Indonesian children aged 9 years. The child in the red t-shirt has a normal height for a 9-year child and the one in the black t-shirt is stunted.
2.1.1.2 Underweight In 2016, the WHO showed that the prevalence of underweight in under fives had declined by10.8% (from 24.8% in 1990 to 14.0% in 2016) globally and by 22.3% (from 49.2% in 1990 to 26.9% in 2016) in the Southeast Asia region. Although the prevalence is declining, the prevalence is still high, highlighting how major an issue underweight is for Southeast Asian countries in particular35.
In contrast, for children aged 5-19 years, there has been relatively little change in the prevalence of moderate and severe underweight over the past few decades. A 2017 paper showed there has only been a decrease in prevalence of 0.8% (from 9.2% to 8.4%) and 2.4% (from 14.8% to 12.4%) for girls and boys, respectively, over 41 years, from year 1975 to 201636. This paper also showed that the prevalence of moderate and severe underweight in children and adolescents is still higher than those who are affected by overweight/ obesity. Despite this slow pace of decline, the prevalence of child and adolescent overweight/ obesity is projected to surpass the prevalence of underweight by 202236.
PAGE | 42 According to UNICEF, the causes of maternal and child undernutrition can be identified into three layers of causes: immediate causes, underlying causes, and basic causes (Figure 2.5)37. The immediate causes include a combination of inadequate dietary intake and recurrent infections/ diseases. These two issues affect one another. The underlying causes include household food insecurity, inadequate feeding practices, and a combination of unhealthy household environment and deficiency in health services. The basic causes consist of three issues: 1) sociocultural, economic and political, 2) financial, human and social capital, and 3) household access to a range of access for example education, income, etc37.
Figure 2.5 Determinants of child undernutrition according to UNICEF37
Undernutrition is a crucial issue because when it happens in early life it will produce lifelong consequences38. Underweight in childhood is strongly correlated with the presence of infectious diseases. A child who is severely underweight27 has a 9.5 times higher risk of dying from diarrhoeal disease than a child who is not underweight. In adolescent pregnancy, underweight is related to higher maternal mortality rate, complications during pregnancy and delivery, and intrauterine growth retardation24,39.
PAGE | 43 Several interventions have been proven to reduce the prevalence of underweight and have been promoted by the World Food Program and UNICEF’s Ending Child Hunger and Undernutrition Initiative40. Examples include a hygiene education and promotion program, maternal nutrition education program, promotion of exclusive breastfeeding, promotion of recommended complimentary feeding practices, safety net and funds transfer, hand washing with soap program, and many more40.
2.1.2 Overnutrition 2.1.2.1 Overweight/ obesity Childhood overweight (0-5 years) is defined as having a body mass index (BMI) z score >+2 and ≤+3 while obesity is defined as having a BMI z score >+3, based on the WHO Growth Standard19. In this thesis, I also include the ‘at risk of overweight/ obesity’ category, defined as having a BMI z score ≥+1 and ≤+2 (Table 2.2)41.
Table 2.2 Definition and cut off points of overnutrition used throughout this thesis according to the WHO Child Growth Standards for 0 to 5 years (http://www.who.int/childgrowth/en/)
Category Type Cut off points Overnutrition At risk Body Mass Index z score ≥+1 and ≤+2 Overweight Body Mass Index z score >+2 and ≤+3 Obesity Body Mass Index z score >+3
In most low and middle income countries, the escalation of overweight/ obesity prevalence in children and adults is much higher than the reduction of underweight prevalence36,42. In a 2017 joint estimates report developed by UNICEF, WHO, and the World Bank Group, the prevalence of overweight in under 5 year children is 6.0% globally and 7.2% for Southeast Asia21 (Figure 2.6).
In a separate 2017 paper, the NCD Risk Factor Collaboration team (which collated 2,416 population-based studies from numerous countries) showed there was a plateau in the previously rising trends in BMI among children and adolescents aged 5-19 years36. Nevertheless, there are still high level prevalence rates in many high income countries, and prevalence rates are accelerating in South, East, and Southeast Asia36. Globally, there was a 4.9% and 6.9% increase in the prevalence of overweight/ obesity in girls and boys, respectively, aged 5-19 years, from 1975 to 201636.
PAGE | 44
Figure 2.6 Prevalence of overweight in under 5 year children based on joint estimates by UNICEF, WHO, and World Bank Group in 201721. Available from http://www.who.int/nutgrowthdb/estimates2016/en/
According to Shrimpton and Rokx, the causes of the obesity epidemic in low- and middle- income countries (LMICs) can be categorised into four interrelated major themes: (1) the health/ biological environment, (2) the economic/ food environment, (3) the physical/ built environment, and (4) the socio/ cultural environment1. The Health/ Biological Environment highlights the vast improvements in health that have resulted in many “surviving” populations with constrained early growth that are then exposed to an “obesogenic” environment, which puts them at a higher risk of accumulating body fat. The Economic/ Food Environment emphasises how better socio economic conditions influence food purchasing power in many countries. In relation to this, the increased penetration of frozen foods and other westernised consumption patterns in to low- and middle-income countries has contributed to the increasing prevalence of overweight/ obesity in the world. The Physical/ Built Environment focuses on physical activity patterns and the diminishing spaces for the community to be physically active, together with the rise of passive entertainment/ abundant screens. The Socio/ Cultural Environment highlights the influence made by media, culture and peer pressure on individuals’ choices around food and activities1.
Children with overweight/ obesity have a higher risk of overweight/ obesity in adulthood43. Overweight/ obesity is a well-known risk factor for many chronic disorders including non- communicable diseases (NCDs) such as diabetes, hypertension, stroke and ischemic heart
PAGE | 45 diseases1,44-47. Obesity in children has also been linked to poorer psychosocial function and educational accomplishment46-49.
2.1.3 The double burden of malnutrition The DBM has occurred globally, negating the previously held opinion that overweight/ obesity only affects people in developed countries whereas undernutrition affects those in the less developed and developing countries1,2. As the consequences of each end of the malnutrition spectrum are well known, the effect of having both ends of the spectrum present in a country are potentially very significant.
A report of the 2002 Bellagio Conference on Nutrition Transition classified the causes of the double burden of malnutrition into three particular transitions, as described in Figure 2.750. They were the nutrition transition, typified by changes in dietary patterns, consumptions, and energy expenditure linked to the development in a country’s economy. This transition can be seen in the shift from a high prevalence of undernutrition towards a higher prevalence of overnutrition. The epidemiological transition is typified by changes in population disease burden and also associated with economic development. This transition results in NCDs substituting for infectious diseases as the primary cause of mortality in a population. The demographic transition, typified by changes in increasing lifespan and changes in the make up of a population, can be seen in those countries which have higher proportions of older people compared to young people50.
PAGE | 46 NUTRITION EPIDEMIOLOGICAL DEMOGRAPHIC TRANSITION TRANSITION TRANSITION
Changes in Changes in Changes in
1. Dietary patterns O verall disease burden Population structure 2. Consumptions i n the community and lengthening of 3. Energy expenditure associated with lifespans associated with economic development economic development
Text Text Text Text A shift from high fertility and A shift from mortality (high A shift from high prevalence of proportions of young high prevalence of infectious diseases people) undernutrition and those related to to to undernutrition reduced mortality rising prevalence of to (increased proportions overnutrition non communicable of older people/ diseases (NCDs) changing age structure)
Figure 2.7 Causes of the double burden of malnutrition. Adapted from Popkin et al.50 and The WHO’s Double Burden of Malnutrition Policy Brief3.
2.2 INTERVENTIONS/ POLICIES ON MALNUTRITION Globally, there are many well-established interventions for each of the elements of the double burden of malnutrition, contained in documents such as the WHO Stunting Policy Brief20 and the WHO Overweight Policy Brief45.
The WHO Childhood Stunting Policy Brief highlights the importance of focusing on the first 1000 days of life, from pregnancy to a child’s second birthday, through several platforms20 (Figure 2.8). The first platform is to ensure an early initiation of breastfeeding and exclusive breastfeeding in the first six months without any additional solid food. Breastfeeding should ideally be continued until the child is two years of age. The second platform involves improving the quality of the diet in complementary feeding period. The quality refers to diversity in dietary intake as well as intake of food from animal sources. The third platform requires collaboration between the previously mentioned nutrition-specific interventions (those that deal with immediate causes of malnutrition) and nutrition-sensitive ones (those that address underlying causes of malnutrition by collaborating actions from different sectors), for example household, environmental, socioeconomic, and cultural factors. The last platform, which is the at the program level, should consider political commitment along with multi sectoral collaborations, involvement of community and an integrated service delivery20.
PAGE | 47
2. Improving the quality of 1. Optimal breastfeeding diet in complementary practices. feeding period.
Framework for action to overcome stunting - focusing on 1000-day period
4. Political commitment 3. Collaboration between and nutrition-specific and multi-sectoral nutrition-sensitive collaborations. interventions.
Figure 2.8 Four platforms to overcome stunting. Data sourced from The WHO Childhood Stunting Policy Brief20
The Ending Child Hunger and Undernutrition Initiative of the World Food Program and UNICEF proposes four pillars in efforts against undernutrition40. The first is to achieve an increased awareness about hunger and undernutrition as well as an understanding of prospective solutions. The second pillar is to strengthen policies and programs around hunger and nutrition. The third is to achieve increased capacities for community-based actions around these issues, and the last one is an increased efficiency and accountability through monitoring and evaluation40.
The WHO Childhood Overweight Policy Brief proposes that interventions/ policies should be tailored to the profile in each country or even more specifically, each area45. They recommend five platforms to reduce the risk of overweight/ obesity (Figure 2.9). The first platform is improving nutritional status and growth patterns through early life exposures, e.g. pre-pregnancy and during pregnancy counseling; exclusive breastfeeding and continuation of breastfeeding until age 2 years, with adequate complimentary feeding practices; regulating the marketing of breast milk substitutes; implementing baby friendly hospital initiative, and many more. The second platform is to improve community understanding and social norms, including through government-endorsed dietary guidelines. The third platform consists of interventions that limit the exposures of children towards marketing of unhealthy foods. The fourth is to influence the food system and food environment, such as ensuring access to safe
PAGE | 48 drinking water and supporting small shops selling healthy food in less-haves communities. The last intervention is to improve nutrition in neighbourhoods, including limiting the number of fast-food outlets around schools45.
1. Improving nutritional status and growth patterns
Framework for action to 2. Improving the 5. Improve nutrition community in neighbourhoods stop increasing prevalence understanding and of childhood overweight social norms
4. Influence the food 3. Interventions that system and food limit the marketing environment of unhealthy foods
Figure 2.9 Five platforms to stop the increasing prevalence of childhood overweight. Data sourced from The WHO Childhood Overweight Policy Brief45
Shrimpton and Rokx have argued that there is no framework that coordinates all of these fragmented solutions1. This issue is made more challenging by the limitations in capacity, especially in LMICs, to deliver the proposed interventions1.
However, in 2017 the WHO issued the Double Burden of Malnutrition Policy Brief3 and Double-duty Actions for Nutrition Policy Brief4, highlighting the importance of tackling both ends of the malnutrition spectrum. These policy briefs emphasise the importance of eliminating the double burden of malnutrition in order to achieve the United Nations Sustainable Development Goals as well as to foster improved health and economic development (Figure 2.10).
PAGE | 49
Figure 2.10 World Health Organization: importance of tackling the double burden of malnutrition – The Double Burden of Malnutrition Policy Brief3. Available from: http://www.who.int/nutrition/publications/doubleburdenmalnutrition-policybrief/en/
The policy briefs also introduce the concept of ‘double-duty actions’ referring to interventions, programs, and policies aimed at simultaneously reducing the prevalence of under- and overnutrition. Such actions include but are not limited to promoting breastfeeding, food marketing regulations, maternal nutrition and antenatal programs, healthy diet initiatives, nutrition education and access to healthy foods (Figure 2.11)3,4. Breastfeeding is a potential double-duty action because breast milk provides essential nutrients that protect against childhood stunting, wasting, and overweight/ obesity. Breastfeeding protects children from diarrhoeal diseases and other childhood infections and may protect mothers against obesity and NCDs in later life51-53.
Healthy diet initiatives range from programs/ interventions to optimise appropriate infant complementary feeding to a variety of interventions that support the adoption of healthy diets in schools and workplaces3,4. Other double-duty actions can be delivered through maternal nutrition and antenatal program. For example, an iron and folic acid supplementation program will protect the mother from having micronutrient deficiencies and help the development of the baby, simultaneously. Furthermore, during antenatal visits, nutrition education to mothers can help prevent excessive weight gain and thus gestational diabetes, and potentially protect the baby from overweight/ obesity in later life54,55. PAGE | 50
Figure 2.11 World Health Organization: Double-duty Actions – Double-duty Actions for Nutrition Policy Brief4. Available from: http://www.who.int/nutrition/publications/double- duty-actions-nutrition-policybrief/en/
PAGE | 51 2.3 REFERENCES
1. Shrimpton R, Rokx C. The double burden of malnutrition: a review of global evidence. Washington DC: World Bank; 2012. Report No.: 79525. 2. Shrimpton R, Rokx C. The Double Burden of Malnutrition in Indonesia. Jakarta, Indonesia: World Bank Jakarta; 2013. 3. World Health Organization. The double burden of malnutrition. Policy brief. Geneva: World Health Organization; 2017. 4. World Health Organization. Double-duty actions for nutrition. Policy brief. Geneva: World Health Organization; 2017. 5. Doak CM, Adair LS, Monteiro C, Popkin BM. Overweight and underweight coexist within households in Brazil, China and Russia. The Journal of Nutrition 2000;130:2965-71. 6. Roemling C, Qaim M. Dual burden households and intra-household nutritional inequality in Indonesia. Economics and Human Biology 2013;11:563-73. 7. Oddo VM, Rah JH, Semba RD, et al. Predictors of maternal and child double burden of malnutrition in rural Indonesia and Bangladesh. Am J Clin Nutr 2012;95:951-8. 8. Kroker-Lobos MF, Pedroza-Tobias A, Pedraza LS, Rivera JA. The double burden of undernutrition and excess body weight in Mexico. Am J Clin Nutr 2014;100:1652S-8S. 9. Freire WB, Silva-Jaramillo KM, Ramirez-Luzuriaga MJ, Belmont P, Waters WF. The double burden of undernutrition and excess body weight in Ecuador. Am J Clin Nutr 2014;100:1636S-43S. 10. Mendez M, Monteiro C, Popkin BM. Overweight exceeds underweight among women in most developing countries. Am J Clin Nutr 2005;81:714-21. 11. Jaacks LM, Slining MM, Popkin BM. Recent trends in the prevalence of under- and overweight among adolescent girls in low- and middle-income countries. Pediatr Obes 2015;10:428-35. 12. Doak CM, Adair LS, Bentley M, Monteiro C, Popkin BM. The dual burden household and the nutrition transition paradox. Int J Obes (Lond) 2005;29:129-36. 13. Abdullah A. The Double Burden of undernutrition and overnutrition in developing countries: an update. Curr Obes Rep 2015;4:337-49. 14. Bove I, Miranda T, Campoy C, Uauy R, Napol M. Stunting, overweight and child development impairment go hand in hand as key problems of early infancy: Uruguayan case. Early Hum Dev 2012;88:747-51. 15. Popkin BM, Richards MK, Montiero CA. Stunting is associated with overweight in children of four nations that are undergoing the nutrition transition. J Nutr 1996;126:3009-16. 16. Victora CG, de Onis M, Hallal PC, Blossner M, Shrimpton R. Worldwide timing of growth faltering: revisiting implications for interventions. Pediatrics 2010;125:e473-80. 17. Mani S. Is there Complete, Partial, or No Recovery from Childhood Malnutrition? – Empirical Evidence from Indonesia. Oxford Bulletin of Economics and Statistics 2012;74:691-715. 18. Cole TJ. Secular trends in growth. Proc Nut Soc 2000;59:317-24. 19. The WHO Child Growth Standards. World Health Organization, 2006. (Accessed 16 March 2014, at http://www.who.int/childgrowth/standards/en/.) 20. World Health Organization. WHO Global Nutrition Target: Stunting Policy Brief. WHO; 2014.
PAGE | 52
21. UNICEF/WHO/World Bank Group. Levels and trends in child malnutrition: Joint Child Malnutrition Estimates- Key findings of the 2017 edition: Unicef/WHO/World Bank Group 2017. 22. World Health Organization. World Health Statistics data visualizations dashboard - Child stunting: Prevalence of stunting in children under 5. 03-05-2017 ed: World Health Organization; 2017. 23. Stewart CP, Iannotti L, Dewey KG, Michaelsen KF, Onyango AW. Contextualising complementary feeding in a broader framework for stunting prevention. Maternal and Child Nutrition 2013;9:27-45. 24. Black RE, Victora CG, Walker SP, et al. Maternal and child undernutrition and oveweight in low-income and middle-income countries. Lancet 2013;382.9890:427-51. 25. Ozaltin E, Hill K, Subramanian S. Association of maternal stature with offspring mortality, underweight, and stunting in low- to middle-income countries. JAMA 2010;303:1507-15. 26. Kossmann J, Nestel P, Herrera M, El-Amin A, Fawzi W. Undernutrition and childhood infections: a prospective study of childhood infections in relation to growth in the Sudan. Acta Paediatrica 2000;89:1122-8. 27. Black R, Allen L, Bhutta Z, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 2008;371:243-60. 28. Olofin I, McDonald CM, Ezzati M, et al. Associations of Suboptimal Growth with All-Cause and Cause-Specific Mortality in Children under Five Years: A Pooled Analysis of Ten Prospective Studies. PloS one 2013:e64636. 29. Solomons N. Malnutrition and infection: an update. British Journal of Nutrition 2007;98:S5-S10. 30. Onis Md, Branca F. Childhood stunting: a global perspective. Maternal and Child Nutrition 2016;12 (Suppl. 1):12-26. 31. UNICEF. The State of the World's Children 2016: A fair chance for every child. USA: United Nations Children's Fund; 2016. 32. Martins P, Hoffman D, Fernandes M, et al. Stunted children gain less lean body mass and more fat mass than their non-stunted counterparts: a prospective study. British Journal of Nutrition 2004;92:819-25. 33. Sawaya A, Grillo L, Verreschi I, da Silva A, Roberts S. Mild stunting is associated with higher susceptibility to the effects of high fat diets: studies in a shantytown population in São Paulo, Brazil. The Journal of Nutrition 1998;128.2 supplement:415-20. 34. Sawaya A, Roberts S. Stunting and future risk of obesity: principal physiological mechanisms. Cad Saude Publica 2003;19 (Sup. 1):S21-S8. 35. World Health Organization. Global and regional trends by WHO Regions, 1990-2016: Underweight. Geneva: World Health Organization; 2017. 36. NCD Risk Factor Collaboration (NCD-RisC)*. Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128·9 million children, adolescents, and adults. Lancet 2017. 37. UNICEF. Improving child nutrition: The achievable imperative for global progress. New York, USA: United Nations Children's Fund (UNICEF); 2013. 38. Barker D. Mothers, babies, and disease in later life. Edinburgh: Churchill Livingstone; 1998. PAGE | 53
39. Han Z, Mulla S, Beyene J, Liao G, McDonald S. Maternal underweight and the risk of preterm birth and low birth weight: a systematic review and meta-analyses. International journal of epidemiology 2011;40. 40. World Food Programme and United Nations Children's Fund. Global Framework for Action: Ending Child Hunger and Undernutrition Initiative: World Food Programme and United Nations Children's Fund; 2006. 41. de Onis M, Blossner M, Borghi E. Global prevalence and trends of overweight and obesity among preschool children. Am J Clin Nutr 2010;92:1257-64. 42. Popkin BM. The nutrition transition and obesity in the developing world. J Nutr 2001;131:871S-3S. 43. Whitaker RC, Wright JA, Pepe MS, Seide KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. The New England journal of medicine 1997;337:869-73. 44. World Health Organization. Diet, Nutrition and the Prevention of Chronic Diseases. Geneva: WHO; 2003. 45. World Health Organization. Global nutrition targets 2025, Childhood overweight policy brief. Geneva: WHO; 2014. 46. Lobstein T, Baur L, Uauy R, TaskForce IIO. Obesity in children and young people: a crisis in public health. Obesity reviews : an official journal of the International Association for the Study of Obesity 2004;5 Suppl 1:4-104. 47. World Health Organization. Consideration of the evidence on childhood obesity for the Commission on Ending Childhood Obesity: report of the Ad hoc Working Group on Science and Evidence for Ending Childhood Obesity. Geneva: World Health Organization; 2016. 48. Quek Y, Tam W, Zhang M, Ho R. Exploring the association between childhood and adolescent obesity and depression: a meta-analysis. Obesity Reviews 2017;18:742-54. 49. Caird J, Kavanagh J, O’Mara-Eves A, et al. Does being overweight impede academic attainment? A systematic review. Health Education Journal 2014;73:497-521. 50. Popkin BM. An overview on the nutrition transition and its health implications: the Bellagio meeting. Public Health Nutr 2002;5:93-103. 51. World Health Organization. Exclusive breastfeeding to reduce the risk of childhood overweight and obesity. Biological, behavioural and contextual rationale. Geneva: World Health Organization; 2014. 52. World Health Organization. Exclusive breastfeeding for optimal growth, development and health of infants. World Health Organization. 53. Rollins N, Bhandari N, Hajeebhoy N, et al. Why invest, and what it will take to improve breastfeeding practices? The Lancet 2016;387:491-504. 54. World Health Organization. Nutrition counselling during pregnancy. Geneva: World Health Organization; 2017. 55. World Health Organization. WHO recommendations on antenatal care for a positive pregnancy experience. Geneva: World Health Organization; 2016.
PAGE | 54
______The next two sections of the literature review were initially conducted in the early stages of my PhD candidature and helped provide the rationale for conducting the quantitative analyses. However, it was not until the later stages of the PhD that I decided to publish some of the literature review. For this reason, I decided to update the literature searching process. As a result I captured some of the quantitative work that I had already published. ______
PAGE | 55 The double burden of malnutrition in Association of South East Asian Nations (ASEAN) countries – a comprehensive review of the literature
INTRODUCTION 58 METHODS Inclusion and exclusion criteria 58 Literature searching process 59 Document screening process 59 Critical appraisal to ensure quality 60 Definitions and cut off points for malnutrition 60 RESULTS Characteristics of included studies 60 Level of double burden of malnutrition 61 Population level 61 Household level 62 Individual level 62 The double burden of malnutrition in each country 63 Cambodia 64 Indonesia 64 Lao PDR 70 Malaysia 70 Myanmar 70 Philippines 70 Thailand 74 Vietnam 74 DISCUSSION 78 REFERENCES 82
This chapter has been submitted to International Journal of Public Health (under review).
PAGE | 56
ABSTRACT Objectives The double burden of malnutrition (DBM) – a combination of undernutrition and overnutrition – is a problem faced by many countries. This study aimed to comprehensively review the DBM in ASEAN countries, including levels (population, household, or individual), the type, and prevalence.
Methods We searched four electronic databases, Medline via OvidSP, Scopus, Global Health via OvidSP and Web of Science, from January 1990-May 2017. We included studies that reported prevalence, were published in English language peer-reviewed journals and were available in full text. Studies were formally assessed against a critical appraisal tool.
Results We included 48 studies from eight countries. Most studies were from Indonesia and there was only one study from Lao PDR. A range of criteria for anthropometry were used in these studies, hence, comparison is impossible. DBM happened at community, household, and individual levels, with different types of undernutrition (stunting/underweight/wasting/thinness) in combination with overweight/obesity.
Conclusions ASEAN countries need to strengthen surveillance using WHO standards to improve the comparability of data, further develop regional strategies to address DBM and investigate potential “double-duty actions” as suggested by the WHO.
Keywords: double burden malnutrition, ASEAN, undernutrition, overnutrition.
PAGE | 57
INTRODUCTION The double burden of malnutrition (DBM) occurs when there is a combination of overweight/ obesity with a form of undernutrition, such as underweight, wasting, stunting, or any micronutrient deficiencies1-4. Previous studies have found that DBM can occur at the level of the country, city, household or individual5-15. Many country-level representative surveys also provide solid evidence of this phenomenon of the combination of underweight children and overweight mothers5, 7, 8, 10, 16. The 2014 report from the International Food Policy Research Institute has called DBM “the new normal” for less-developed and developing countries3.
DBM also occurs in countries in southeast Asia. Southeast Asian countries can be classified in a number of ways. One such classification is the Association of South East Asian Nations (ASEAN), a political and economic alliance aimed at promoting regional peace, stability and economic growth17. In this paper, we use the ASEAN grouping because this comprises countries with similar characteristics who may thus share similar issues around nutrition. The ten ASEAN countries are Brunei Darussalam, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, Philippines, Singapore, Thailand, and Vietnam.
In this paper we aim to provide a comprehensive review of the prevalence of DBM in ASEAN countries in all age groups, in order to facilitate comparison between countries and to identify gaps in current knowledge so as to inform future research. Specifically, we review DBM at different levels (population [including country-, city-, or any community-level studies], household, or even individual), the types of DBM (the type of undernutrition associated with overweight), and the prevalence by age, housing area (urban/ rural) or education of the participants. As the currently available information is scattered, this review will provide a consolidated overview of DBM in these countries, including identifying similarities and differences in patterns and highlighting potential strategies/ interventions.
METHODS Inclusion and exclusion criteria The inclusion criteria of studies were: 1) studies that reported prevalence data of the double burden of malnutrition (overweight/ obesity with either stunting, or underweight, or wasting) in one or more of the ASEAN countries, including multi-country studies that included any ASEAN country; 2) studies published in peer-reviewed journals; 3) studies available in full text written in English; and 4) sample size >300. The exclusion criteria were editorials or
PAGE | 58 commentaries, qualitative studies, studies focusing on clinical features and intervention studies.
We did not include micronutrient deficiency studies in this review, because initial searches only identified one study that included micronutrient deficiency and overweight/ obesity and met the other inclusion criteria.
Literature search process We conducted a literature search using four electronic databases for English language papers: Medline via OvidSP, Scopus, Global Health via OvidSP and Web of Science, from January 1990 until May 2017. The search in Medline used the following key words/ MeSH terms: 1) overnutrition (overnutrition OR overweight OR obes* OR weight OR body weight OR fat), 2) undernutrition (undernutrition OR stunt* OR wasting OR thin* OR underweight), 3) malnutrition, 4) double burden (dual OR double OR burden), 5) prevalence (prevalence OR number* OR case*, incidence OR survey), 6) adults (adult* OR mother OR father OR parent* OR caretaker OR elder*), 7) children and adolescents (child* OR adolescen* OR school-age OR preschool OR under five ), 8) countries (ASEAN OR South East Asia OR Asia OR LMIC* OR developing countr* OR Brunei Darussalam OR Cambodia OR Indonesia OR Lao PDR OR Malaysia OR Myanmar OR Philippines OR Singapore OR Thailand OR Vietnam). We combined the results from 1) to 7) with AND and then combined the results with each of the countries item in number 8) with AND. We also limited the search to the English language and journal articles. Searches in three other electronic databases used similar search strategies and key words.
Furthermore, we identified papers through hand searching of the reference lists of journals and contacted several South East Asian public health researchers to seek full publications of conference abstracts.
Document screening process We screened publications identified from the literature and hand search process. The first step was to remove duplicates and ensure the relevance of the articles. We excluded non-English language articles. Titles and abstracts were then screened by CNR. After further exclusion of records according to our criteria, relevant full texts were then ready to be assessed to ensure quality.
PAGE | 59
Critical appraisal to ensure quality For the purposes of this review, we assessed full text articles against a tool established by Loney et al18, which was specifically developed for critically appraising prevalence or incidence studies. This tool includes three quality assessment criteria: 1) validity of the study methods (6 questions); 2) interpretation of results (1 question); and 3) applicability of results (1 question). The critical appraisal tool is available in Supplementary Table 1. In this review we included articles that scored 5 or more and fulfilled at least two of three assessment criteria. Afterwards, we grouped the data in each country based on the level of DBM: 1) population level (includes DBM that happened within the same country, city, or smaller population), 2) household level, or 3) individual level.
Supplementary Table 1 about here
Definitions and cut off points for malnutrition There were many different criteria and cut-off points for malnutrition used in the included studies, which are summarised in Supplementary Table 2. Several studies used a combination/ modification of criteria, as listed in the table.
Supplementary Table 2 about here
RESULTS Characteristics of included studies Figure 1 shows the results of the literature search. Of the 45 identified studies from our literature search identification and screening process, 44 scored 5 or more on the critical appraisal score (6 scored 8, 29 scored 7, and 9 scored 6); these 44 studies are included in the review. The results of the critical appraisal process are available in Supplementary Table 3. Four of the included studies were multi-country studies, and 40 were single country studies. The country with the most studies of DBM was Indonesia and that with the fewest was Lao PDR. We found no study on DBM from Singapore or Brunei Darussalam, resulting in studies from 8 countries to be further analysed. The earliest study was from 2000 and latest ones were from 2017.
PAGE | 60
Potentially relevant records identified Additional records identified through other n
o through four database searching (n= 2,157) sources (hand search, personal i t communication): n=12 a c
i f i
t n
e
d Records after duplicates removed (n=1427) I
Reasons records excluded: 1. Full text not available Title and abstract screened (n=1427) 2. Studies not including both under- and overnutrition 3. Not conducted in ASEAN g
n 4. Editorial i 5. Commentary n
e 6. Conference abstract e r c
S Full text articles excluded with Full-text articles assessed for eligibility (n=89) reasons: 1. Qualitative study 2. Focus on clinical features 3. Intervention study y
t Critical appraisal of studies using the criteria of i l i Loney et al (n=45) b Excluded because score less than 5 i g i l E
Studies included (n=44) d
e 1. Brunei Darussalam: 0 6. Myanmar: 2 d
u 2. Cambodia: 5 7. Philippines: 3 l
c 3. Indonesia: 15 8. Singapore: 0 n I 4. Lao PDR: 1 9. Thailand: 4 5. Malaysia: 10 10. Vietnam: 14
Figure 1. A flow diagram showing the flow of citations reviewed in this study
Supplementary Table 3 about here
Level of double burden of malnutrition We analysed the studies at three levels: studies that were conducted at the population level (country-, city-, or any community-level), those that were conducted at the household level, and those conducted to investigate whether DBM occurred within the same individual. From this point onwards, we calculated studies based on the country.
Population level All eight countries had studies that described the presence of DBM at the population level, with the number of studies from each country as follows: Vietnam (13 studies), Indonesia
PAGE | 61
(nine studies) and Malaysia (nine studies), Cambodia (four studies), Thailand (four studies), Myanmar (two studies), Philippines (two studies) and Lao PDR (one study). From 44 studies at the population level, 27 were nationally representative samples while the other 17 studies were conducted in a limited number of provinces, cities or communities. The 27 national studies were: 4/4 studies from Cambodia, 6/9 studies from Indonesia, 4/9 studies from Malaysia, 1/2 studies from Myanmar, 1/1 study from Lao PDR, 1/2 studies from Philippines, 2/4 studies from Thailand, and 8/13 studies from Vietnam.
Household level Studies looking at the occurrence of DBM within the same household, most of which were the combination of overweight mother and underweight child, were from 4 countries. Most studies were from Indonesia (five studies), followed by Malaysia, Philippines, and Vietnam (each with one study). Five of these 8 studies (4 from Indonesia and one from Vietnam) were nationally representative.
Individual Two studies, one each from Cambodia and Indonesia, reported DBM in the same individual using nationally representative data. Both studies were conducted in children and found concurrent stunting and overweight/ obesity. Both studies were published in 2016, with one conducted in under five children, and one in children aged 2.0-4.9 years. The complete list of studies according to the three levels of DBM is available in Figure 2.
PAGE | 62
Cambodia: 1. Arimond and Ruel (2004) POPULATION LEVEL 2. Corsi et al. (2011) 3. Jaacks et al. (2015) 4. Greffeuille et al. (2016) Malaysia: Thailand: 1. Lim et al. (2000) 1. Firestone et al. (2011) 2. E-Siong et al. (2002) 2. Rojroongwasinkul et al. (2013) 3. Ming Moy et al. (2004) 3. Sandjaja et al. (b) (2013) 4. Zaini et al. (2005) 4. Pongcharoen et al. (2017) Indonesia: 5. Zalilah et al. (2006) 1. Winkvist et al. (2000) 6. Soo et al. (2008) 2. Soekirman et al. (2002) 7. Moore et al. (2010) 3. Julia et al. (2004) 8. Sandajaja et al. (b) (2013) 4. Roemling and Qaim (2012) Vietnam: 9. Poh et al. (2013) 5. Sandjaja et al.(a) (2013) 1. Tuan et al. (2007) 6. Sandjaja et al. (b) (2013) 2. Cuong et al. (2007) 7. Hanandita and Tampubolon (2015) 3. Dieu et al. (2007) 4. Walls et al. (2009) 8. Jaacks et al. (2015) Myanmar: 9. Rachmi et al.(a) (2016) 1. Moore at al. (2010) 5. Kim et al. (2010) 6. Moore et al. (2010) 2. Soe at al. (2016) 7. Ha et al. (2011) 8. Laillou et al. (2012) 9. Ly et al. (2013) 10. Nguyen et al. (2013) Lao PDR: Philippines: 11. Sandjadja et al. (b) (2013) 1. Moore et al. (2010) 1. Florentino et al. (2002) 12. Laillou et al. (2014) 2. Moore et al. (2010) 13. Jaacks et al. (2015)
Indonesia: 1. Doak et al. (2005) HOUSEHOLD LEVEL 2. Oddo et al. (2012) 3. Roemling and Qaim (2013) 4. Vaezghasemi et al. (2014) Malaysia: Philippines: Vietnam: 5. Wibowo et al. (2015) 1. Wong et al. (2015) 1. Angeles-Agdeppa et al. 1. Doak et al. (2005) (2003)
INDIVIDUAL LEVEL
Cambodia: Indonesia: 1. Tzioumis et al. (2016) 1. Rachmi et al. (b) (2016)
Figure 2. Different levels of the double burden of malnutrition in 8 ASEAN countries; Country order alphabetically, study order by year of publication
The double burden of malnutrition in each country The included studies used different criteria and cut off points, which makes direct comparisons impossible. However, we highlight the characteristics of studies in each country in more detail. Details of the included studies are shown in Tables 1-3 in alphabetical order.
PAGE | 63
Cambodia Five studies were from Cambodia, four of which were conducted at the population level and one at the individual level (Table 1). Of the population level studies, two were in adults and adolescent to adult women, one in female adolescents, and the other one in children. The two studies in women19, 20 showed a decreased prevalence of underweight and increased prevalence of overweight/ obesity from 2000 to 200519 and 2000 to 201420. The study in female adolescents showed an increase of both underweight and overweight/ obesity prevalence over a five-year period (2000 to 2005)21. The other study involving children aged 6-23 months22 showed that stunting, wasting and overweight/ obesity were present at the population level, with a high prevalence of stunting and wasting. The individual level study in children aged 0-5 years showed an increased prevalence of concurrent stunting and overweight between 2000 to 201023.
Indonesia Of the 15 studies from Indonesia, 9 were at the population level, 5 at the household level and one at the individual level (Table 1). Six out of 9 studies at the population level studies were conducted in children and adolescents, one involving adolescent and adult participants and two involving adult participants. One country-level study of repeated cross-sectional analyses showed that from 1993 to 2007 the prevalence of underweight in adults decreased while the prevalence of overweight/ obesity increased24. In children and adolescent similar trends were found in two country-level studies (both repeated cross-sectional analyses)21, 25.
At the household level, one study found that underweight and overweight were present in different individuals (children or adults) in 22.0% of households in 19937 while a similar study found this occurred in 19.0% of households in 200726, both using the Indonesian Family Life Survey (IFLS) dataset. One study that performed repeated cross-sectional analyses in different years within the same household, also using IFLS dataset, showed an increasing prevalence of DBM (one overweight and one underweight household member) between 1993 (11.1%) and 2007 (16.1%)27. Another study using the Basic Health Survey (Riskesdas) dataset reported that the prevalence of one type (or any combination) of undernutrition (stunting, underweight, or wasting) in a child aged 2-5 years with an overweight mother was 29.8%28 of DBM in the households in 2015. A further study conducted in eight provinces documented the prevalence of overweight mother and stunted child aged 6-59 months to be 5.8% in 201210.
PAGE | 64
A study conducted at the individual level with children aged 2.0-4.9 years in four different years (1993, 1997, 2000 and 2007) showed an increased prevalence of concurrent stunting and overweight over the 14 year time period29.
PAGE | 65 Table 1. The double burden of malnutrition in Cambodia and Indonesia; Country order alphabetically, study order by year of publication No Authors (year Participants Criteria and cut-off points used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Wasting [W], Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Wasting [W], Obesity [O/OB] [W], Thinness Thinness [T] [T] Overweight [O] CAMBODIA Population level 1 Arimond and 2,049 Both sexes; Adult: WHO 1995 [O] BMI≥25 [S] and [U] in N/A [O/OB]: 4.0% [S]: 36.0% Ruel (2004)22 6-23 m Children: WHO 1995 children, and [O] in [W]: 19.0% [S] HAZ<-2 [W] WHZ<-2 mothers 2 Corsi et al. 2000: 5,292 Female; WHO 2000 [U] and [O] in adult N/A [O/OB] [U] 2000: 19.1% (2011)19 2005: 6,147 20-49 y [U] BMI<18.5 [O] BMI≥25 2000: 7.3% 2005: 17.8% 2005: 11.3% 3 Jaacks et al. 2005: 1,447 Female; IOTF 2000 [U] and [O] in N/A [O/OB] 2005 [U] 2005 (2015)21 2010: 1,582 15-18.9 y [U] BMI<18.5 [O] BMI≥25 adolescent Urban: 4.0% Urban: 24.0% Rural: 1.0% Rural: 21.0% 2010 2010 Urban: 4.0% Urban: 25.0% Rural: 2.0% Rural: 21.0% 4 Greffeuille et 2000: 15,351 Female; WHO 2000 [U] and [O] in adult N/A [O/OB] [U] al. (2016)20 2005: 16,823 15-49 y [U] (BMI<18.5) [O] BMI≥25 2000: 6.4% 2000: 20.7% 2010: 18,754 2005: 9.5% 2005: 20.3% 2014: 17,577 2010: 10.9% 2010: 19.1% 2014: 18.3% 2014: 13.9% Individual level 5 Tzioumis et al 2000: 1,753 Both sexes; WHO Growth Standards 2006 [S] and [O] 2000: 2.7% N/A N/A (2016)23 2005: 1,557 0-5 y [S] HAZ<-2 [O] WHZ>+2 concurrently in a 2005: 0.8% 2010: 1,449 child 2010: 1.2% INDONESIA Population level 1 Winkvist et al 4,132 Female; WHO 2000 [U] and [O] in N/A [O/OB] [U] 1996: 16.2% (2000)55 15-49 y [U] BMI<18.5 [O] BMI ≥25 adults 1996: 11.6% 1997: 14.4% 1997: 14.3%
PAGE | 66 No Authors (year Participants Criteria and cut-off points used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Wasting [W], Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Wasting [W], Obesity [O/OB] [W], Thinness Thinness [T] [T] Overweight [O] 2 Soekirman et al 1,367 Both sexes; WHO 1995 [S], [U], and [O] in N/A [O/OB] [S] (2002)56 8-10 y [S] HAZ<-2 [U] WAZ<-2 children Male: 17.76% Male: 11.79% [O] BMI>85th percentile Female: 15.29% Female: 8.97% [OB] BMI>95th percentile [U] Male: 12.95% Female: 7.35% 3 Julia et al 3,010 Female; WHO 1995 [S], [W], and [O] in N/A [O/OB] 2.7% [S]: 19.3% (2004)57 6-7.9 y [S] HAZ< -2 [W] WHZ<-2 children [W]: 5% Male; 6-8.9y IOTF 2000 [O] BMI≥25 4 Roemling and 1993: 10,227 Both sexes; WHO 2004 and Gurrici 1998 [U] and [O] in N/A [O/OB] [U] Qaim (2012)24 2000:17,041 20 - 75 y [U] BMI<18.5 [O] BMI≥23 adults 1993 1993 2007: 20,475 Male: 20.84% Male: 16.71% Female: 32.0% Female: 17.29% 2000 2000 Male: 24.56% Male: 16.66% Female: 39.62% Female: 14.47% 2007 2007 Male: 31.38% Male: 13.67% Female: 48.75% Female: 10.95% 5 Sandjaja et ala 7,211 Both sexes; WHO Growth Standard (2006) [S], [U], [W], and N/A [O] [S] (2013)47 0.5-12 y and Reference (2007) [O] in children Urban: 5.6% Urban: 25.1% [S] HAZ<-2 [U] WAZ<-2 Rural: 3.2% Rural: 39.2% [W] WHZ<-2 [OB] [U] [O] BMIZ>+2 (less than 5 years) Urban: 5.1% Urban: 19.2% and >+1 (≥5 years) Rural: 1.8% Rural: 28.9% [W] Urban: 5.8% Rural: 6.0% 6 Sandjaja et alb 2,396 Both sexes; WHO Growth Reference 2007 [S], [U], [T], and N/A [O]: 5.9% [S] : 29.0% (2013)58 6-12 y [S] HAZ<-2 [U] WAZ<-2 [O] in children [OB]: 5.6% [U]: 25.2% [T] BMIZ<-2 [O] BMIZ>+1 [T]: 9.7%
PAGE | 67 No Authors (year Participants Criteria and cut-off points used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Wasting [W], Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Wasting [W], Obesity [O/OB] [W], Thinness Thinness [T] [T] Overweight [O] 7 Hanandita and 645,032 Both sexes; WHO 2004 and Gurrici 1998 [U] and [O] in N/A [O/OB]: 17.9% [U]: 14.4% Tampubolon ≥15 y [U] BMI<18.5 [O] BMI≥25 adolescent and (2015) 59 adults 8 Jaacks et al 2000: 1,336 Female; IOTF 2000 [U] and [O] in N/A [O/OB] [U] (2015)21 2007: 1,069 15-18.9 y [U] BMI<18.5 [O] BMI≥25 adolescent 2000 2000 Urban: 7.0% Urban: 23.0% Rural: 6.0% Rural: 28.0% 2007 2007 Urban: 10.0% Urban: 28.0% Rural: 10.0% Rural: 25.0% 9 Rachmi et ala 4,101 Both sexes; WHO Growth Standards 2006 [S], [U], and [O] in N/A [O/OB] [S] 1993: 50.8% (2016)25 2.0-4.9 y [S] HAZ<-2 children 1993: 10.3% 1997: 48.6% [U] WAZ<-2 1997: 10.6 % 2000: 44.8% [O] BMIZ>+1 2000: 11.7% 2007: 36.7% 2007: 16.5% [U] 1993: 34.5% 1997: 34.6% 2000: 27.1% 2007: 21.4% Household level 10 Doak et al 6,461 Both sexes; Adults WHO 2000 [U] and [O] in 22% [O/OB] [U] (2005)7 households ≥2 y [U] BMI<18.5 [O] BMI≥25 children and adults Children: 5.1% Children: 32.2% Children IOTF 2000 Adults: 14.6% Adults: 19.2% [U] BMI<18.5 [O] BMI≥25 11 Oddo et al 247,126 Female; Adults WHO 2004 [O] BMI≥23 [S] in children and 5.80% [O/OB]: 32% [S]: 37% (2012)10 households mean 28y Children WHO Growth Standards [O] in mother Both sexes; 2006 [S] HAZ<-2 6-59 m
PAGE | 68 No Authors (year Participants Criteria and cut-off points used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Wasting [W], Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Wasting [W], Obesity [O/OB] [W], Thinness Thinness [T] [T] Overweight [O] 12 Roemling and 27,237 Both sexes; Adult: WHO 2004 Dual burden 1993: 11.1% N/A N/A Qaim (2013)27 households ≥2 y [U] BMI<18.5 [O] BMI≥25 household: one [O] 1997: 16.3% Children: WHO Growth Standard and one [U] 2000: 16.8% (2006) and Reference (2007) member. 2007: 16.1% [U] BMIZ<-2 [O] BMIZ>+1 13 Vaezghasemi et 9,743 Both sexes; Adults WHO 2000 [U] and [O] in 19% [O/OB] [U] al. (2014)26 households ≥2 y [U] BMI<18.5 [O] BMI≥25 children and adults Urban: 25% Urban: 18% (38,237 Children IOTF 2000 Rural: 17% Rural: 20% individuals) [U] BMI<18.5 [O] BMI≥25 14 Wibowo et al 1,899 Female; 18- Adults WHO 2000 [O] BMI ≥25 [S], [U], [W] and 29.80% N/A N/A (2015)28 households 40 y Children WHO Growth Standards any combination in Both sexes; 2006 the child and [O] 2-5 y [S] HAZ<-2 [U] WAZ<-2 mother [W] WHZ<-2 Individual level 15 Rachmi et alb 4,101 Both sexes; WHO Growth Standards 2006 [S] and [O] 1993: 6.4% N/A N/A (2016)29 2.0-4.9 y [S] HAZ<-2 [O] BMIZ>+1 concurrently in a 1997: 6.8% child 2000: 5.2% 2007: 7.2%
PAGE | 69 Lao PDR The nationally representative multi-country study of DBM in adults >18 years that included Lao PDR as one of the sites showed that DBM was present in adults 18 years and older in 2010 (Table 2). There was a higher prevalence of overweight/ obesity in females compared to males, but a higher prevalence of underweight in males compared to females30.
Malaysia There were 10 studies of DBM in Malaysia, nine at the population level (seven involving children and adolescents, two involving adults) and one at the household level (Table 2). The two country-level studies in adults30, 31 used the same criteria and cut off points for both underweight and overweight. Both showed an increased prevalence of overweight/ obesity and a decreased prevalence of underweight between 200031 and 201030, although Lim et al involved adults aged 20 years and older31 while Moore at al involved those aged 18 years and older30. Those studies in children and adolescents involved participants of different ages and used different criteria and cut off points so we were not able to make a comparison or identify prevalence trends. The one household level study showed a 19.4% prevalence of a combination of a stunted child and overweight mother and a 12.5% prevalence of a combination of underweight child and overweight mother32.
Myanmar We found two studies of DBM in Myanmar (Table 2). One, a multi-country study conducted in adults in 2010, showed a higher prevalence of underweight in male compared to female, but a higher prevalence of overweight/ obesity in female compared to male30. The other study, conducted at the city-level involving young children aged 6-24 months, showed a high prevalence of underweight (28.4%) and a low prevalence (1.5%) of overweight/ obesity33.
Philippines There were two population level studies conducted in the Philippines (Table 2). One city-level study in children aged 8-10 years showed a significantly higher (more than 4 times) prevalence of stunting and underweight in children from public schools compared to those from private schools34. In contrast, there was a significantly higher prevalence of overweight/ obesity in children from private schools compared to those in public schools34. The only study that was conducted at the household level revealed that 8.2% of households had an overweight mother and a underweight child35.
PAGE | 70 Table 2. The double burden of malnutrition in Lao PDR, Malaysia, Myanmar, and Philippines; Country order alphabetically, study order by year of publication No Authors (year Participants Criteria used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Thinness Thinness [T] Obesity [O/OB] [T] Overweight [O] LAO PDR Population level 1 Moore et al 3,536 Both sexes; WHO 2000 [U] and [O] in N/A [O] Male: 9.1% [U] Male: 9.3% (2010)30 ≥18 y [U] BMI<18.5 [O] BMI≥25 adults Female: 11.9% Female: 9.0% [OB] BMI≥30 [OB] Male:0.7% Female:1.7% MALAYSIA Population level 1 Lim et al 28,737 Both sexes; WHO 1995 [U] and [O] in N/A [O/OB]: 26.5% [U]: 12.7% (2000)31 ≥20 y [U] BMI<18.5 [O] BMI>25 adults 2 E-Siong et al 1,208 Both sexes; WHO 1995 [S], [U], and [O] in N/A [O/OB]: 17.8% [S]: 6.7% (2002)60 8-10 y [S] HAZ<-2 [U] WAZ<-2 children Male: 19.2% [U]: 7.1% [O] BMI>85th percentile Female: 16.5% [OB] BMI>95th percentile 3 Ming Moy et al 3,620 Both sexes; WHO 1995 [U] and [O] in N/A [O/OB]: 7.3% [U]: 14.8% (2004)61 Primary 5 [U] BMI<5th percentile children (11 y) [O] BMI>95th percentile Secondary 2 (14 y) and Secondary 4 (16 y) 4 Zaini et al 1,405 Both sexes; WHO 1995 [U] and [O] in N/A [O]: 16.3% [U] : 1.2% (2005)62 9-10 y [U] BMI<3rd percentile children [OB]: 6.3% [O] BMI>85th percentile [OB] BMI>95th percentile 5 Zalilah et al 6,555 Both sexes; WHO 1995 [U] and [O] in N/A [O/OB] [U] (2006)63 11-15 y [U] BMI<5th percentile adolescent Male:19.5% Male: 14.9% [O] BMI>85th percentile Female: 16.7% Female: 7.8%
PAGE | 71 No Authors (year Participants Criteria used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Thinness Thinness [T] Obesity [O/OB] [T] Overweight [O] 6 Soo et al 489 Both sexes; WHO 1995 [U] and (at risk of) N/A At risk [O]: 9.8% [U]: 3.1% (2008)64 15-17 y [U] BMI<5th percentile [O] in adolescent [O]: 8.6% At risk [O] BMI>85th percentile [O] BMI>95th percentile 7 Moore et al 4,546 Both sexes; WHO 2000 [U] and [O] in N/A [O] [U] (2010)30 ≥18 y [U] BMI<18.5 [O] BMI≥25 adults Male: 21.8 % Male: 10.1% [OB] BMI≥30 Female: 25.0% Female: 6.4% [OB] Male: 8.4% Female: 9.2% 8 Sandjaja et alb 2,262 Both sexes; WHO Growth Reference 2007 [S], [U], [T] and N/A [O]: 14.0% [S] : 5.8% (2013)58 6-12 y [S] HAZ<-2 [U] WAZ<-2 [O/OB] in children [OB]: 21.8% [U]: 9.9% [T] BMIZ<-2 [O] BMIZ>+1 [T]: 7.0% [OB] BMIZ>+2 9 Poh et al 3,542 Both sexes; WHO Growth Standards 2006 [S], [T] and [O] in N/A [O] Urban: 9.7% [S] Urban: 8.3% (2013)45 individuals 0.5-12.9 y and WHO Growth Reference children Rural: 9.9% Rural: 8.8% 2007 [OB] [T] Urban: 5.2% [S] HAZ<-2 [T] BMIZ<-2 Urban: 12.7% Rural: 6.0% [O] BMIZ>+1 (>5 y) and ≥+2 (0- Rural: 8.2% 5 y) [OB] BMIZ>+2 (>5 y) and ≥+3 (0-5 y) Household level 10 Wong et al 438 Female;15- Adult WHO 2000 [O] BMI ≥ 25 [O] mother and Stunted Child [O/OB] Children (2015)32 households 55 y Children WHO Growth Standards malnourished Overweight Adult: 26.0% [S] : 64.2% (931 mothers Both sexes; 2006 [S] HAZ<-2 children Mother: 19.4% Children: 5.2% [U]: 48.7% and 304 3-59 m [U] WAZ<-2 [O] BMIZ>+2 Underweight children) Child Overweight Mother:12.5% MYANMAR Population level
PAGE | 72 No Authors (year Participants Criteria used Combination of Prevalence of publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Thinness Thinness [T] Obesity [O/OB] [T] Overweight [O] 1 Moore et al 5,539 Both sexes; WHO 2000 [U] and [O] in N/A [O] Male: 5.1% [U] (2010)30 ≥18 y [U] BMI<18.5 [O] BMI≥25 adults Female: 8.8% Male: 14.6% [OB] BMI≥30 [OB] Male: 0.7% Female: 13.3% Female: 1.1% 2 Soe et al 385 Both sexes; WHO Child Growth Standards [U] and [O] in N/A [O/OB]: 1.5% [U]: 28.4% (2016)33 6-24 m 2006 children [U] WAZ<-2 SD [O] WAZ>+1.5 SD PHILIPPINES Population level 1 Florentino et al 1,208 Both sexes; WHO 1995 [S], [U], and [O] in N/A [O] [S] (2002)34 individuals 8-10 y [S] HAZ<-2 [U] WAZ<-2 children Public school: Public school: [O] BMI>85th percentile 5.8% 26.5% [OB] BMI>95th percentile Private school: Private school: 24.9% 6.6% [OB] [U] Public school: Public school: 3.3% 18.2% Private school: Private school: 12.0% 4.3% 2 Moore et al 7,557 Both sexes; WHO 2000 [U] and [O] in N/A [O] [U] (2010)30 individuals ≥18 y [U] BMI<18.5 [O] BMI≥25 adults Male: 11.2% Male: 16.5% [OB] BMI≥30 Female: 12.4% Female: 9.2% [OB] Male: 2.3% Female: 4.6% Household level 3 Angeles- 378 child- Female; Adults WHO 2000 [O] BMI≥ 25 [U] child and [O] 8.20% N/A N/A Agdeppa et al mother pairs >20y Children [U] WAZ <-2 mother (2003)35 Both sexes; 33-83 m
PAGE | 73 Thailand There were four studies of DBM in Thai children (Table 3). We are not able to show trends in prevalence because of differences in criteria and cut off points as well as the age of participants. One study showed a higher prevalence of stunting, underweight, and thinness in rural areas compared to urban areas whereas a higher prevalence of overweight/ obesity in urban areas54.
Vietnam There were 14 studies of DBM in Vietnam, 13 at the population level (seven in adults, four in adolescents, two in children through to adults) and only one at the household level (Table 3). Of the adult studies, three city- and country-level studies using similar criteria found a higher prevalence of overweight/ obesity in females and a higher prevalence of underweight in males57, 59, 38. One country-level study showing decreasing trends of prevalence of underweight from 2000 to 2005 and increasing trends of overweight prevalence in adults61. One study conducted at the household level showed that 5% of 4,600 households had a combination of underweight/ overweight in children and adults7.
PAGE | 74 Table 3. The double burden of malnutrition in Thailand, and Vietnam; Country order alphabetically, study order by year of publication No Authors (year of Participants Criteria used Combination of Prevalence publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Thinness [T] Obesity [O/OB] [W], Thinness Overweight [O], [T] Obesity [OB] THAILAND Population level 1 Firestone et al 4,610 Both sexes; CDC 2000 [S], [U] and [O] in N/A [O/OB]: 8.26% [S]: 19.85% (2011)66 2-10 y [S] <-2 SD Height for age children [U]: 27.79% [U] BMI for Age <5th percentile [O] BMI for Age>95th percentile 2 Rojroongwasinkul 3,119 Both sexes; WHO Growth Standards 2006 [S], [U], [T], [O] N/A [O] Urban: 6.9% [S] Urban: 4.1% et al (2013)46 0.5-12.9 y and WHO Growth Reference and [OB] in Rural: 8.0% Rural: 8.4% 2007 children [OB] [U] Urban: 6.4% [S] HAZ<-2 [U] WAZ<-2 Urban: 11.8% Rural: 10.2% [T] BMIZ<-2 Rural: 5.9% [T] Urban: 6.2% [O] BMIZ≥+1 and ≤+2 Rural: 6.4% [OB] BMIZ>+2 3 Sandjaja et alb 1,580 Both sexes; WHO Growth Reference 2007 [S], [U], [T], and N/A [O]: 9.7% [S]: 7.0% (2013)58 individuals 6-12 y [S] HAZ<-2 [U] WAZ<-2 [O] in children [OB]: 14.0% [U]: 13.2% [T] BMIZ<-2 [O] BMIZ>+1 [T]: 8.1% [OB] BMIZ>+2 4 Pongcharoen et al 845 Both sexes; WHO Growth Standards 2006 [S], [T] and [O] in N/A [O/OB]: 14.0% [S]: 5.0% (2017)67 individuals 3-5 y [S] HAZ<-2 [T] BMIZ<-2 children [T]: 3.0% [O] BMIZ>+2 VIETNAM Population level 1 Tuan et al 1992: 24,068 Both sexes; Adults: WHO 1995 [U] and [O] in N/A [O] Children [U] Children (2007)68 2002:158,019 2-17 y and 18- [U] BMI<18.5 [O] BMI≥25 children and adult 1992: 1.4% 1992: 32.1% 65 y Children: CDC 2000 2002: 1.8% 2002: 33.5% [U] BMI<5th percentile [O] Adults [U] Adults [O] BMI>85th percentile 1992: 2.0% 1992: 32.6% 2002: 5.2% 2002: 24.8%
PAGE | 75 No Authors (year of Participants Criteria used Combination of Prevalence publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Thinness [T] Obesity [O/OB] [W], Thinness Overweight [O], [T] Obesity [OB] 2 Cuong et al 1,488 Both sexes; WHO 2004 [U] and [O] in N/A [O/OB] [U] (2007)69 20-60 y [U] BMI<18.5 [O] BM≥23 adults Male: 31.6% Male: 22.6% Female: 33.6% Female: 18.9% 3 Dieu et al 670 Both sexes; CDC 2000 [S], [U], [W] and N/A [O]: 20.5% [S]: 2.3% (2007)70 48-65 m [S] Height for age<5th percentile [O] in children [OB]: 16.3% [U]: 2.7% [U] Weight for age<5th [W]: 2.6% percentile [W] BMI for age<5th percentile IOTF 2000 [O] BMI≥25 [OB] BMI≥30 4 Walls et al 978 Both sexes; WHO 200 0 and WHO 2004 [U] and [O] in N/A [O/OB] BMI≥25 [U] (2009)71 25-74 y [U] BMI<18.5 adults Male: 14.0% Male: 14.5% [O] BMI≥25 Female: 12.7% Female: 12.6% [O] BMI≥23 – Asian cut off [O/OB] BMI≥23 Male: 29.7% Female: 31.5% 5 Kim et al (2010)72 497 Both sexes; WHO 2004 [U] and [O] in N/A [O/OB] [U] individuals 19-60 y [U] BMI<18.5 [O] BMI≥23 adults Male: 21.6% Male: 14.2% Female: 20.4% Female: 18.8% 6 Moore et al 2,995 Both sexes; WHO 2000 [U] and [O] in N/A [O] Male: 2.7 % [U] (2010)30 individuals ≥18 y [U] BMI<18.5 [O] BMI≥25 adults Female: 7.2% Male: 22.8 % [OB] BMI≥30 [OB]: Female: 12.1% Male: - Female: 0.6% 7 Ha et al (2011)73 2000: 14,452 Both sexes; WHO 2000 and WHO 2004 [U] and [O] in N/A [O/OB] BMI≥25 [U] 2000: 25.0% 2005: 17,213 25-64 y [U] BMI<18.5 adults 2000: 3.5% 2005: 20.9% [O] BMI≥25 2005: 6.6% [O] BMI≥23 – Asian cut off [O/OB] BMI≥23 2000: 16.3% 2005: 11.7%
PAGE | 76 No Authors (year of Participants Criteria used Combination of Prevalence publication) Total Sex and age Stunting [S], Underweight [U], DBM DBM Overweight [O], Stunting [S], range Thinness [T] Stunting [S], Obesity [OB], or Underweight Overweight [O], Obesity [OB] Underweight [U], overweight/ [U], Wasting Thinness [T] Obesity [O/OB] [W], Thinness Overweight [O], [T] Obesity [OB] 8 Laillou et al 2,112 (1,526 Female; >18y Adult WHO 2000 [U] and [O] in N/A Adult Children (2012)74 women and Both sexes; [U] BMI<18.5 [O] BMI≥25 children and adult [O/OB]: 8.0% [S]: 23.2% 586 children) 0.5-6 y Children WHO Growth [U]: 18.1% Standards 2006 [S] HAZ<-2 [W]: 6.3% [U] WAZ<-2 [W] WHZ<-1 Adult [U]: 20.5% 9 Ly et al (2013) 75 1,621 Both sexes; WHO 2000 and WHO 2004 [U] and [O] in N/A [O/OB] [U]: 12.4% ≥35 y [U] BMI<18.5 adults BMI≥25: 16.0% [O] BMI≥25 [O/OB] [O] BMI≥23 – Asian cut off BMI≥23: 33.7% 10 Nguyen et al 952 Both sexes; WHO Growth Standards 2006 [S], [U], and [O] in N/A [O/OB]: 6.5% [S]: 14% (2013)76 0.5-4.9 y [S] HAZ<-2 [U] WAZ<-2 children [U]: 8.6% [O] BMIZ>+2 11 Sandjaja et alb 1,920 Both sexes; WHO Growth Reference 2007 [S], [U], [T], and N/A [O]: 7.0% [S]: 15.8% (2013)58 6-12 y [S] HAZ<-2 [U] WAZ<-2 [O] in children [U]: 22.0% [T] BMIZ<-2 [O] BMIZ>+1 [T]: 12.7% 12 Laillou et al 1,526 Female; 15-49 WHO 2004 [U] and [O] in [O/OB]: 20% [U]: 20.5% (2014)77 y [U] BMI<18.5 [O] BMI≥23 women
13 Jaacks et al 6,581 Female; 15- IOTF 2000 [U] and [O] in N/A [O/OB] [U] (2015)21 18.9 y [U] BMI<18.5 [O] BMI≥25 adolescents Urban: 1.0% Urban: 36.0% Rural: 1.0% Rural: 33.0% Household level 14 Doak et al (2005)7 4,600 Both sexes; Adults WHO 2000 [U] and [O] in 5.00% [O/OB]: [U]: households ≥2 y [U] BMI<18.5 [O] BMI≥25 children and adults Children 1.2% Children 43.7% Children IOTF 2000 Adults 2.0% Adults 34.4% [U] BMI<18.5 [O] BMI≥25
PAGE | 77 DISCUSSION This study is the first comprehensive review of high-quality publications on the double burden of malnutrition (a combination of undernutrition and overweight/obesity) in ASEAN countries. Four multi-country studies were included. Indonesia had the most studies on this issue, with only one study that met the quality criteria from Lao PDR, and no study from Singapore or Brunei Darussalam. Eight of the ASEAN countries reported 35 prevalence studies at the population level, seven at the household level and only two at the individual level. Our review also highlighted the range of criteria and cut-off points for anthropometry used in studies between and within countries, making it difficult to compare the prevalence rates; hence we depend on those studies with repeated cross-sectional analyses to provide data on prevalence trends.
At the population level, we found an increasing prevalence of overweight/ obesity and decreasing prevalence of underweight in Cambodia, Indonesia, and Vietnam. Synthesising our findings from all countries, there was a higher prevalence of overweight/ obesity in urban areas and in females, while underweight was more prevalent in rural areas and in males. At the household level, the prevalence of DBM ranged from 5.0% in Vietnam to as high as 29.8% in Indonesia with varying combinations of DBM. The two studies at the individual level showed an increased prevalence of concurrent stunting and overweight in under five children between 1993 to 2007 and 2000 to 2010.
At the population-level we found a higher prevalence of overweight compared to underweight in adults from six out of seven countries. This finding is in line with the changing face of malnutrition affecting all regions in the world. For example, in 2014, 1.9 billion adults aged 18 years or older were overweight and 462 million adults underweight36, 37.
Most of the studies in this review focussed on DBM at the population level, while recent studies from Latin American countries (e.g. Mexico38, Guatemala39, Colombia40 and Ecuador6) have had a greater focus on the household- and individual level. These studies have investigated the combination of stunted children and overweight mother at the household level and concurrent stunting and overweight/ obesity in children at the individual level6, 38-40.
The most common household-level combination of DBM found in this review was an underweight child with an overweight mother. Given that the largest number of children living with stunting and underweight is in Asia2, 41, it is clear that ASEAN countries should
PAGE | 78 have a major focus of their policies, program implementation and monitoring and evaluation systems on the combination of underweight/stunting with overweight/ obesity, both at the household, as well as the individual level.
One of the possible reasons behind the varying prevalence rates of DBM in different ASEAN countries is that the nutrition transition is at different stages in these countries. Other studies have suggested possible explanations behind the occurrence of DBM, especially at the household level. One study in Guatemala found that the socioeconomic status of dual burden households (i.e. stunted children with overweight mothers) was lower than those non-stunted households. The associated risk factors for dual burden households were short statured mothers, a higher number of births, mother currently working, and being indigenous42. Another study that analysed demographic health survey data from 18 low- and middle-income countries stated that factors associated with dual burden households (overweight mother and stunted/ underweight children) were higher maternal age (>30 years), lower level of maternal education, having more than two siblings under the age of 5 years living in the same house and living in urban areas43. However, they also concluded that the paradox in dual burden households could be best defined as a result of a “rapid secular increase” in the weight of the mothers43.
Several studies demonstrated a social patterning to DBM. For example, one study from the Philippines found a much higher prevalence of undernutrition (stunting and underweight) in children from public schools compared to those from private schools, but a higher prevalence of overweight/ obesity in children in private schools34. Such findings are similar to those in a study from Burkina Faso where overweight was more prevalent in children from private schools and stunting more prevalent in children from public schools44. Several studies found the ‘classic’ phenomenon of higher undernutrition in rural areas and higher overweight/ obesity in urban areas in Thailand, Malaysia, and Indonesia26, 45-47. This is in keeping with observations that undernutrition in low and middle income countries occurs more commonly in the poor, the urban poor and those living in rural areas, while overweight/ obesity affects people in higher socioeconomic status groups15, 48, 49. This phenomenon might be shifting as overweight prevalence can be found in both urban and rural areas21 as well as in the poorest to richest quintiles in many low and middle income countries8.
We did not find any study from Singapore and Brunei Darussalam. According to the 2014 and 2015 Global Nutrition Reports, there is no available information on under five stunting,
PAGE | 79 wasting, or overweight, although 6.2% of adults were documented as having obesity in Singapore3,4. The prevalence of under five stunting, wasting, and overweight in 2009 in Brunei Darussalam was 19.7%, 2.9%, and 8.3%, respectively, however, Brunei Darussalam has been listed as one of the countries with insufficient data to make a formal assessment4.
The main strength of the review is that we conducted a comprehensive literature search of studies published in English on DBM from the ASEAN countries. Secondly, we only included papers of high quality, supported by the use of a valid critical appraisal tool which also reduced subjectivity and reporting bias. Thirdly, this review provides the first collated summary of DBM in ASEAN countries. One limitation is there might be incompletely identified studies in other national languages.
In this review of DBM, the use of different criteria and cut-off points of both under- and overnutrition within and between different countries highlights the need for harmonisation of definitions to make comparisons possible50-54. While many country-level surveys are available, the need to facilitate the documentation of trends in prevalence is of great importance.
The World Health Organization recommends several strategies to address DBM. One of them is to create supportive environments for nutrition at all ages, including home, school and workplace settings, as well as the city context36. Regular monitoring of weight and height as well as supporting diet quality are crucial and should not be limited only to children and adolescents, as good maternal nutrition can prevent both under- and overnutrition in their offspring36, 37.
The information on the prevalence of DBM should be useful to health care practitioners and policy makers, providing country-level guidance for development and implementation of specific preventions/ interventions across the lifespan in the respective country. For example, Cambodia is experiencing a decreased prevalence of underweight and an increased prevalence of overweight in adult women, while there is an increased prevalence of both underweight and overweight in adolescent females19-21. Such findings highlight the need for specific policies to address the environmental and social determinants of malnutrition within schools, communities and work contexts36. Different approaches might be needed to decrease the prevalence of both under- and overnutrition and prevent future malnutrition.
PAGE | 80 Future research should have a strong focus at the individual level of DBM, the only work at this level in the ASEAN region having been conducted in Cambodia and Indonesia23, 29. Research should also aim to better understand the shared drivers of both under- and overnutrition, including biology (poor adolescent and maternal nutrition, early-life nutrition), environmental factors (food, health, social, living and working environments), and socioeconomic influences (poverty, inequality, food insecurity)37. By identifying these shared drivers, governments may be able to create shared platforms to address the different types of malnutrition in one action, for example through modification of national dietary guidelines or other national-level policies37.
To date most countries have had separate policies to address under- and overnutrition, but there has been no framework to address both conditions simultaneously2. The World Health Organization released two policy briefs in 2017 to set out the potential for “double-duty actions” to eliminate both ends of the malnutrition spectrum36, 37. These actions are not necessarily new actions,. However, actions previously aimed to overcome one form of malnutrition, may have potential to overcome several forms concurrently37. These include exclusive breastfeeding initiatives, optimising early nutrition, programs involving maternal nutrition and antenatal care, school food policies and regulations on food marketing37. ASEAN countries will benefit from revisiting existing actions currently in place to address under- or overnutrition issues and search for potential “double-duty actions”.
PAGE | 81 REFERENCES
1. Shrimpton R, Rokx C. The Double Burden of Malnutrition in Indonesia. Jakarta, Indonesia: World Bank Jakarta, 2013 Contract No.: Report 76192-ID. 2. Shrimpton R, Rokx C. The double burden of malnutrition: a review of global evidence. Washington DC: World Bank, 2012 79525. 3. International Food Policy Research Institute. Global Nutrition Report 2014: Actions and accountability to accelerate the World's Progress on Nutritioin. washington, DC: International Food Policy Research Institute, 2014. 4. International Food Policy Research Institute. Global Nutrition Report 2015: Actions and Accountability to Advance Nutrition and Sustainable Development Washington, DC: International Food Policy Research Institute, 2015. 5. Doak CM, Adair LS, Monteiro C, Popkin BM. Overweight and underweight coexist within households in Brazil, China and Russia. The Journal of Nutrition. 2000;130:2965-71. 6. Freire WB, Silva-Jaramillo KM, Ramirez-Luzuriaga MJ, Belmont P, Waters WF. The double burden of undernutrition and excess body weight in Ecuador. Am J Clin Nutr. 2014;100(6):1636S- 43S. Epub 2014/11/21. doi: 10.3945/ajcn.114.083766. PubMed PMID: 25411306. 7. Doak CM, Adair LS, Bentley M, Monteiro C, Popkin BM. The dual burden household and the nutrition transition paradox. Int J Obes (Lond). 2005;29(1):129-36. Epub 2004/10/27. doi: 10.1038/sj.ijo.0802824. PubMed PMID: 15505634. 8. Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, de Onis M, et al. Maternal and child undernutrition and oveweight in low-income and middle-income countries. Lancet. 2013;382.9890:427-51. Epub August 3, 2013. 9. Haddad L, Cameron L, Barnett I. The double burden of malnutrition in SE Asia and the Pacific: priorities, policies and politics. Health policy and planning. 2014. Epub 2014/10/18. doi: 10.1093/heapol/czu110. PubMed PMID: 25324529. 10. Oddo VM, Rah JH, Semba RD, Sun K, Akhter N, Sari M, et al. Predictors of maternal and child double burden of malnutrition in rural Indonesia and Bangladesh. Am J Clin Nutr. 2012;95(4):951-8. doi: http://dx.doi.org/10.3945/ajcn.111.026070. PubMed PMID: 22357721. 11. Fernald L, Neufeld L. Overweight with concurrent stunting in very young children from rural Mexico: prevalence and associated factors. European Journal of Clinical Nutrition. 2007;61:623-32. 12. Piernas C, Wang D, Du S, Zhang B, Wang Z, Su C, et al. The double burden of under- and overnutrition and nutrient adequacy among Chinese preschool and school-aged children in 2009-2011. Eur J Clin Nutr. 2015. Epub 2015/07/02. doi: 10.1038/ejcn.2015.106. PubMed PMID: 26130296. 13. Bove I, Miranda T, Campoy C, Uauy R, Napol M. Stunting, overweight and child development impairment go hand in hand as key problems of early infancy: Uruguayan case. Early Hum Dev. 2012;88(9):747-51. Epub 2012/05/09. doi: 10.1016/j.earlhumdev.2012.04.002. PubMed PMID: 22560815. 14. Monteiro C, Conde W, Popkin BM. is obesity replacing or adding to undernutrition? Evidence from different social classes in Brazil. Public Health Nutrition. 2002;5(1A):105-12. 15. Abdullah A. The Double Burden of undernutrition and overnutrition in developing countries: an update. Curr Obes Rep. 2015;4:337-49. 16. Frankenberg E, Thomas D. The Indonesia Family Life Survey (IFLS): Study Design and Results from Waves 1 and 2. Santa Monica, California: RAND, 2000 Contract No.: DRU-2238/1- NIA/NICHD. 17. ASEAN. Association of Southeast Asian Nations Jakarta: The ASEAN Secretariat; [cited 2017 20th June]. Available from: http://asean.org/. 18. Loney PL, Chambers LW, Bennett KJ, Roberts JG, Stratford PW. Critical appraisal of the health research literature: prevalence or incidence of a health problem. Chronic diseases in Canada. 1998;19(4):170-6. Epub 1999/02/24. PubMed PMID: 10029513. 19. Corsi DJ, Finlay JE, Subramanian SV. Global Burden of Double Malnutrition: Has Anyone Seen It? PloS one. 2011;6(9):e25120. doi: doi:10.1371/ journal.pone.0025120. 20. Greffeuille Vr, Sophonneary P, Laillou A, Gauthier L, Hong R, Hong R, et al. Inequalities in Nutrition between Cambodian Women over the Last 15 Years (2000–2014). Nutrients. 2016;8(224):1- 15. doi: doi:10.3390/nu8040224. 21. Jaacks LM, Slining MM, Popkin BM. Recent trends in the prevalence of under- and overweight among adolescent girls in low- and middle-income countries. Pediatr Obes.
PAGE | 82 2015;10(6):428-35. doi: 10.1111/ijpo.12000. PubMed PMID: 25558987; PubMed Central PMCID: PMCPMC4492920. 22. Arimond M, Ruel MT. Dietary Diversity Is Associated with Child Nutritional Status: Evidence from 11 Demographic and Health Surveys. J Nutr. 2004;134:2579–85. 23. Tzioumis E, Kay MC, Bentley ME, Adair LS. Prevalence and trends in the childhood dual burden of malnutrition in low- and middle-income countries, 1990–2012. Public Health Nutrition. 2016;19(8):1375–88. 24. Roemling C, Qaim M. Obesity trends and determinants in Indonesia. Appetite. 2012;58(3):1005-13. doi: http://dx.doi.org/10.1016/j.appet.2012.02.053. PubMed PMID: 22402303. 25. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, Underweight and Overweight in Children Aged 2.0-4.9 Years in Indonesia: Prevalence Trends and Associated Risk Factors. PloS one. 2016;11(5):e0154756. Epub 2016/05/12. doi: 10.1371/journal.pone.0154756. PubMed PMID: 27167973; PubMed Central PMCID: PMC4864317. 26. Vaezghasemi M, Ohman A, Eriksson M, Hakimi M, Weinehall L, Kusnanto H, et al. The effect of gender and social capital on the dual burden of malnutrition: a multilevel study in Indonesia. PloS one. 2014;9(8):e103849. doi: 10.1371/journal.pone.0103849. PubMed PMID: 25153321; PubMed Central PMCID: PMCPMC4143167. 27. Roemling C, Qaim M. Dual burden households and intra-household nutritional inequality in Indonesia. Economics and Human Biology. 2013;11(4):563-73. doi: 10.1016/j.ehb.2013.07.001. 28. Wibowo Y, Sutrisna B, Hardinsyah H, Djuwita R, M MK, Syafiq A, et al. Relationship between intra-household food distribution and coexistence of dual forms of malnutrition. Nutrition Research and Practice. 2015;9(2):174-9. 29. Rachmi CN, Agho KE, Li M, Baur LA. Stunting coexisting with overweight in 2.0-4.9-year- old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys. Public Health Nutr. 2016:1-10. Epub 2016/04/29. doi: 10.1017/S1368980016000926. PubMed PMID: 27120968. 30. Moore S, Hall JN, Harper S, Lynch JW. Global and national socioeconomic disparities in obesity, overweight, and underweight status. J Obes. 2010;2010. doi: 10.1155/2010/514674. PubMed PMID: 20721361; PubMed Central PMCID: PMCPMC2915815. 31. Lim T, Ding L, Zaki M, Suleiman A, Fatimah S, Siti S, et al. Distribution of Body weight, height and body mass index in a national sample of malaysian adults. The medical journal of Malaysia. 2000;55:108-28. 32. Wong CY, Zalilah MS, Chua EY, Norhasmah S, Chin YS, Nur’Asyura AS. Double-burden of malnutrition among the indigenous peoples (Orang Asli) of Peninsular Malaysia. BMC Public Health. 2015;15(680):1-9. doi: DOI 10.1186/s12889-015-2058-x. 33. Soe ZY, Chompikul J, Hong SA. Prevalence and risk factors of underweight status among children aged between six to twenty four months in Hlaing Tharyar Township of Myanmar. Journal of Public Health and Development 2016;14(3):29-44. 34. Florentino RF, Villavieja GM, Laña RD. Regional study of nutritional status of urban primary schoolchildren. 1. Manila, Philippines. Food and Nutrition Bulletin. 2002;23(1). 35. Angeles-Agdeppa I, Lana RD, Barba CV. A case study on dual forms of malnutrition among selected households in District 1, Tondo, Manila. Asia Pacific J Clin Nutr. 2003;12(4):438-46. 36. World Health Organization. The double burden of malnutrition. Policy brief.2017. Available from: http://apps.who.int/iris/bitstream/10665/255413/1/WHO-NMH-NHD-17.3-eng.pdf?ua=1. 37. World Health Organization. Double-duty actions for nutrition. Policy brief.2017. Available from: http://apps.who.int/iris/bitstream/10665/255414/1/WHO-NMH-NHD-17.2-eng.pdf?ua=1. 38. Kroker-Lobos MF, Pedroza-Tobias A, Pedraza LS, Rivera JA. The double burden of undernutrition and excess body weight in Mexico. Am J Clin Nutr. 2014;100(6):1652S-8S. Epub 2014/11/21. doi: 10.3945/ajcn.114.083832. PubMed PMID: 25411308. 39. Ramirez-Zea M, Kroker-Lobos MF, Close-Fernandez R, Kanter R. The double burden of malnutrition in indigenous and nonindigenous Guatemalan populations. Am J Clin Nutr. 2014;100(6):1644S-51S. Epub 2014/11/21. doi: 10.3945/ajcn.114.083857. PubMed PMID: 25411307. 40. Sarmiento OL, Parra DC, Gonzalez SA, Gonzalez-Casanova I, Forero AY, Garcia J. The dual burden of malnutrition in Colombia. Am J Clin Nutr. 2014;100(6):1628S-35S. Epub 2014/11/21. doi: 10.3945/ajcn.114.083816. PubMed PMID: 25411305. 41. Neufeld L, Osendarp S. Global, regional and country trends in underweight and stunting as indicators of nutrition and health of populations. Nestle Nutr Inst Workshop Ser. 2014;78:11-9.
PAGE | 83 42. Lee J, Houser RF, Must A, Fulladolsa PPd, Bermudez OI. Disentangling nutritional factors and household characteristics related to child stunting and maternal overweight in Guatemala. Economics and Human Biology. 2010;8:188-96. 43. Jehn M, Brewis A. Paradoxical malnutrition in mother–child pairs: Untangling the phenomenon of over- and under-nutrition in underdeveloped economies. Economics and Human Biology. 2009;7:28-35. 44. Daboné C, Delisle HF, Receveur O. Poor nutritional status of schoolchildren in urban and peri-urban areas of Ouagadougou (Burkina Faso). Nutr J. 2011;10(34):1-8. 45. Poh BK, Ng BK, Siti Haslinda MD, Nik Shanita S, Wong JE, Budin SB, et al. Nutritional status and dietary intakes of children aged 6 months to 12 years: findings of the Nutrition Survey of Malaysian Children (SEANUTS Malaysia). The British journal of nutrition. 2013;110 Suppl 3:S21-35. Epub 2013/09/18. doi: 10.1017/S0007114513002092. PubMed PMID: 24016764. 46. Rojroongwasinkul N, Kijboonchoo K, Wimonpeerapattana W, Purttiponthanee S, Yamborisut U, Boonpraderm A, et al. SEANUTS: the nutritional status and dietary intakes of 0.5–12-year-old Thai children. British Journal of Nutrition 2013;110:S36–S44. 47. Sandjaja S, Budiman B, Harahap H, Ernawati F, Soekatri M, Widodo Y, et al. Food consumption and nutritional and biochemical status of 05-12-year-old Indonesian children: the SEANUTS study. British Journal of Nutrition. 2013;110 Suppl 3:S11-20. PubMed PMID: 24016762. 48. Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. International journal of epidemiology. 2002;31(2):285-93. 49. Tzioumis E, Adair LS. Childhood dual burden of under- and over-nutrition in low- and middle-income countries: a critical review. Food Nutr Bull. 2014;35(2):230-43. 50. Mendez M, Monteiro C, Popkin BM. Overweight exceeds underweight among women in most developing countries. Am J Clin Nutr. 2005;81:714-21. 51. Abrantes MMo, Lamounier JA, Colosimo EAn. Recommendations for the use of body mass index for the classification of overweight and obese children and adolescents. Food and Nutrition Bulletin. 2002;23(3):262-6. 52. Bibiloni MdM, Pons A, Tur JA. Prevalence of Overweight and Obesity in Adolescents: A Systematic Review. ISRN Obesity. 2013;Article ID 392747:1-14. doi: http://dx.doi.org/10.1155/2013/392747. 53. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(May):1-6. 54. World Health Organization. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. The Lancet. 2004;363(January):157-63. 55. Winkvist A, Nurdiati DS, Stenlund H, Hakimi M. Predicting under- and overnutrition among women of reproductive age: a population-based study in central Java, Indonesia. Public Health Nutr. 2000;3(2):193-200. PubMed PMID: 10948386. 56. Soekirman, Hardinsyah, Jus'at I, Jahari AB. Regional study of nutritional status of urban primary schoolchildren. 2. West Jakarta and Bogor, Indonesia. Food & Nutrition Bulletin. 2002;23(1):31-40. PubMed PMID: 11975367. 57. Julia M, van Weissenbruch MM, de Waal HAD-v, Surjono A. Influence of socioeconomic status on the prevalence of stunted growth and obesity in prepubertal Indonesian children. Food & Nutrition Bulletin. 2004;25(4):354-60. PubMed PMID: 15646313. 58. Sandjaja, Poh BK, Rojroonwasinkul N, Le Nyugen BK, Budiman B, Ng LO, et al. Relationship between anthropometric indicators and cognitive performance in Southeast Asian school- aged children. The British journal of nutrition. 2013;110 Suppl 3:S57-64. Epub 2013/09/18. doi: 10.1017/S0007114513002079. PubMed PMID: 24016767. 59. Hanandita W, Tampubolon G. The double burden of malnutrition in Indonesia: Social determinants and geographical variations. SSM-Population Health. 2015;1:16-25. doi: http://dx.doi.org/10.1016/j.ssmph.2015.10.002. 60. E-Siong T, Khor S-C, Ooi H-E, Young S-I, Zakiyah O, Zulkafli H. Regional study of nutritional status of urban primary schoolchildren. 3. Kuala Lumpur, Malaysia. Food and Nutrition Bulletin. 2002;23(1):41-7. 61. Moy FM, Gan CY, Zaleha MKS. Body mass status of school children and adolescents in Kuala Lumpur, Malaysia. Asia Pac J Clin Nutr. 2004;13(4):324-9.
PAGE | 84 62. Zaini MA, Lim C, Low W, Harun F. Factor affecting nutritional status of Malaysian primary school children. Asia-Pacific journal of public health / Asia-Pacific Academic Consortium for Public Health. 2005;17(2):71-80. 63. Zalilah MS, K Mirnalini, Khor GL, Merlin A, Bahaman AS, Norinlah K. Estimates and Distribution of Body Mass Index in a Sample of Malaysian Adolescents. Med J Malaysia. 2006;61(1):48-54. 64. Soo KL, Shariff ZM, Nasir M, Taib M. Eating behaviour, body image and self-esteem of adolescent girls in Malaysia. Perceptual and motor skills. 2008;106:833-44. 65. Muhammad NA, Omar K, Shah SA, Muthupalaniappen LAP, Arshad F. Parental perception of their children’s weight status, and its association with their nutrition and obesity knowledge. Asia Pac J Clin Nutr. 2008;17(4):597-602. 66. Firestone R, Punpuing S, Peterson KE, Acevedo-Garcia D, Gortmaker SL. Child overweight and undernutrition in Thailand: Is there an urban effect? Social Science & Medicine. 2011;72:1420-8. 67. Pongcharoen T, Judprasong K, Jitngarmkusol S, Kriengsinyos W, Winichagoon P. Body mass index is associated with fat mass in normal, overweight/obese, and stunted preschool children in central Thailand. Asia Pac J Clin Nutr. 2017;26(4). doi: doi: 10.6133/apjcn.052016.02. 68. Tuan N, Tuong P, Popkin B. Body mass index (BMI) dynamics in vietnam. European Journal of Clinical Nutrition. 2008;62:78-86. 69. Cuong T, Dibley M, Bowe S, Hanh T, Loan T. Obesity in adults: an emerging problem in urban areas of Ho Chi Minh City, Vietnam. European Journal of Clinical Nutrition. 2007;61:673-81. 70. Dieu HT, Dibley MJ, Sibbritt D, Hanh TT. Prevalence of overweight and obesity in preschool children and associated socio-demographic factors in Ho Chi Minh City, Vietnam. International journal of pediatric obesity : IJPO : an official journal of the International Association for the Study of Obesity. 2007;2(1):40-50. Epub 2007/09/01. doi: 10.1080/17477160601103922. PubMed PMID: 17763009. 71. Walls HL, Peeters A, Son PT, Quang NN, Hoai NTT, Loi DD, et al. Prevalence of underweight, overweight and obesity in urban Hanoi, Vietnam. Asia Pac J Clin Nutr. 2009;18(2):234- 9. 72. Kim SH, Hwang J-Y, Kim MK, Chung HW, Nguyet TTP, Kim WY. Dietary factors related to body weight in adult Vietnamese in the rural area of Haiphong, Vietnam: the Korean Genome and Epidemiology Study (KoGES). Nutrition Research and Practice (Nutr Res Pract). 2010;4(3):235-42. doi: DOI: 10.4162/nrp.2010.4.3.235. 73. Ha DT, Feskens EJ, Deurenberg P, Mai LB, Khan NC, Kok FJ. Nationwide shifts in the double burden of overweight and underweight in Vietnamese adults in 2000 and 2005: two national nutrition surveys. BMC Public Health. 2011;11(62):1-9. 74. Laillou A, Pham TV, Tran NT, Le HT, Wieringa F, Rohner F, et al. Micronutrient Deficits Are Still Public Health Issues among Women and Young Children in Vietnam. PloS one. 2012;7(4):e34906. doi: doi:10.1371/journal.pone.0034906. 75. Ly KA, Ton TG, Ngo QV, Vo TT, Fitzpatrick AL. Double burden: a cross-sectional survey assessing factors associated with underweight and overweight status in Danang, Vietnam. BMC Public Health. 2013;13(35):1-10. 76. Nguyen BKL, Thi HL, Do VAN, Thuy NT, Huu CN, Do TT, et al. Double burden of undernutrition and overnutrition in Vietnam in 2011: results of the SEANUTS study in 0·5–11-year- old children. British Journal of Nutrition. 2013;110:S45–S56. 77. Laillou A, Yakes E, Le TH, Wieringa FT, Le BM, Moench-Pfanner R, et al. Intra-Individual Double Burden of Overweight and Micronutrient Deficiencies among Vietnamese Women. PloS one. 2014;9(10):e110499. doi: doi:10.1371/journal.pone.0110499. 78. Rachmi CN, Li M, Baur LA. Overweight and obesity in Indonesia: prevalence and risk factors – a literature review. Public Health. 2017; 147: 20-29.
PAGE | 85 Supplementary Tables
Supplementary Table 1. Guidelines for critically appraising studies of prevalence or incidence of a health problem by Loney at al18 – previously published in Rachmi et al78
Criteria Scoring system A. ARE THE STUDY METHODS VALID? . Are the study designs and sampling method appropriate for the research question? 1 point . Is the sampling frame appropriate? 1 point . Is the sample size adequate? 1 point . Are objectives, suitable and standard criteria used for measurement of the health 1 point outcome? . Is the health outcome measured in an unbiased fashion? 1 point 6. Is the response rate adequate? Are the refusers described? 1 point
B. WHAT IS THE INTERPRETATION OF THE RESULTS? 7. Are the estimates of prevalence or incidence given with confidence intervals and in 1 point detail by subgroup, if appropriate?
C. WHAT IS THE APPLICABILITY OF THE RESULTS? 8. Are the study subjects and the setting described in detail and similar to those of 1 point interest to you?
Total 8 points
PAGE | 86 Supplementary Table 2. Criteria and cut-off points used in this literature review*
No Name of criteria (abbreviations) Age in Cut off points years Overweight [O]/ Obesity [OB] Stunting [S] Underweight [U]/ wasting [W]/ thinness [T] Children and adolescents 1 World Health Organization (WHO 1995) 6-19 [O]: BMI≥ 85% percentile HAZ<-2 or <3rd percentile [W]: WHZ<-2 or <3rd [OB]: BMI ≥ 95% percentile percentile (<10 years) BMI for age<5th percentile (≥10 years) 2 International Obesity Task Force (IOTF 2000) 2-18 [O]: BMI-for-age-and sex ≥BMI of [T]: BMI-for-age-and sex 25 at age 18
PAGE | 87 Supplementary Table 3. Critical appraisal process using Loney et al.18; Country order alphabetically, study order by year of publication
Authors and year of publication Random Unbiased Adequate Measures Outcome Adequate Confidence Study Score sample or sampling sample were the measured response intervals, subjects whole frame size standards in an rate, subgroup described population unbiased refusers analysis fashion described
CAMBODIA 1 Arimond and Ruel (2004)22 Y Y Y Y Y N/A Y Y 7 2 Corsi et al (2011)19 Y Y Y Y Y N/A Y Y 7 3 Jaacks et al (2015)21 Y Y Y Y Y N/A Y Y 7 4 Tzioumis et al (2016)23 Y Y Y Y Y N/A Y Y 7 5 Greffeuille et al (2016)20 Y Y Y Y Y N/A Y Y 7 INDONESIA 1 Winkvist et al (2000)55 Y Y Y Y Y Y N Y 7 2 Soekirman et al (2002)56 Y Y Y Y Y N/A Y Y 7 3 Julia et al (2004)57 Y N Y Y Y Y Y Y 7 4 Doak et al (2005)7 Y Y Y Y N/A N/A Y Y 6 5 Roemling and Qaim (2012)24 Y Y Y Y Y N/A Y Y 7 6 Oddo et al (2012)10 Y Y Y Y Y N/A N Y 6 7 Sandjaja et ala (2013)47 Y Y Y Y Y N/A Y Y 7 8 Sandjaja et alb (2013)58 Y Y Y Y Y N/A Y Y 7 9 Roemling and Qaim (2013)27 Y Y Y Y Y N/A Y Y 7 10 Vaezghasemi et al. (2014) 26 Y Y Y Y Y N/A Y Y 7 11 Wibowo et al (2015)28 Y Y Y Y Y N/A Y Y 7 12 Hanandita and Tampubolon (2015)59 Y Y Y Y Y N/A N Y 6 13 Jaacks et al (2015)21 Y Y Y Y Y N/A Y Y 7 14 Rachmi et ala (2016)25 Y Y Y Y Y N/A Y Y 7 15 Rachmi et alb (2016)29 Y Y Y Y Y N/A Y Y 7 LAO PDR
PAGE | 88 Authors and year of publication Random Unbiased Adequate Measures Outcome Adequate Confidence Study Score sample or sampling sample were the measured response intervals, subjects whole frame size standards in an rate, subgroup described population unbiased refusers analysis fashion described 1 Moore et al (2010)30 Y Y Y Y N N/A Y Y 6 MALAYSIA 1 Lim et al (2000)31 Y Y Y Y Y Y Y Y 8 2 E Siong et al (2002)60 Y Y Y Y Y N/A Y Y 7 3 Ming Moy et al (2004)61 Y Y Y Y Y Y N Y 7 4 Zaini et al (2005)62 Y Y Y Y Y N/A Y Y 7 5 Zalilah et al (2006)63 Y Y Y Y Y N/A Y Y 7 6 Soo et al (2008)64 Y Y Y Y Y Y N Y 7 7 Muhammad et al (2008)65 N N Y Y Y N/A N Y 4 8 Moore et al (2010)30 Y Y Y Y N N/A Y Y 6 9 Sandjaja et alb (2013)58 Y Y Y Y Y N/A Y Y 7 10 Poh et al (2013)45 Y Y Y Y Y N/A Y Y 7 11 Wong et al (2015)32 Y Y Y Y Y Y Y Y 8 MYANMAR 1 Moore et al (2010)30 Y Y Y Y N N/A Y Y 6 2 Soe et al (2016)33 Y Y Y Y Y N/A N Y 6 PHILIPPINES 1 Florentino et al (2002)34 Y Y Y Y Y N/A Y Y 7 2 Angeles-Agdeppa et al (2003)35 Y Y Y Y Y Y N Y 7 3 Moore et al (2010)30 Y Y Y Y N N/A Y Y 6 THAILAND 1 Firestone et al (2011)66 Y Y Y Y Y N/A N Y 6 2 Rojroongwasinkul et al (2013)46 Y Y Y Y Y N/A Y Y 7 3 Sandjaja et alb (2013)58 Y Y Y Y Y N/A Y Y 7 4 Pongcharoen et al (2017)67 Y Y Y Y Y N/A N Y 6
PAGE | 89 Authors and year of publication Random Unbiased Adequate Measures Outcome Adequate Confidence Study Score sample or sampling sample were the measured response intervals, subjects whole frame size standards in an rate, subgroup described population unbiased refusers analysis fashion described VIETNAM 1 Doak et al (2005)7 Y Y Y Y N/A N/A Y Y 6 2 Tuan et al (2007)68 Y Y Y Y Y Y Y Y 8 3 Cuong et al (2007)69 Y Y Y Y Y N/A Y Y 7 4 Dieu et al (2007)70 Y Y Y Y Y N/A Y Y 7 5 Walls et al (2009) 71 Y Y Y Y Y Y Y Y 8 6 Kim et al (2010)72 Y Y Y Y Y N/A Y Y 7 7 Moore et al (2010)30 Y Y Y Y N N/A Y Y 6 8 Ha et al. (2011)73 Y Y Y Y Y N/A Y Y 7 9 Laillou et al (2012)74 Y Y Y Y Y Y Y Y 8 10 Ly et al. (2013)75 Y Y Y Y Y Y N Y 7 11 Nguyen et al (2013)76 Y Y Y Y Y N/A N Y 6 12 Sandjaja et alb (2013)58 Y Y Y Y Y N/A Y Y 7 13 Laillou et al (2014)77 Y Y Y Y Y Y Y Y 8 14 Jaacks et al (2015) 21 Y Y Y Y Y N/A Y Y 7
PAGE | 90 Overweight/obesity in Indonesia: prevalence and risk factors – a literature review
INTRODUCTION 93 METHODS Inclusion and exclusion criteria 93 Literature search strategy 93 Screening of documents 93 Critical appraisal process 93 Definitions of overweight/ obesity 93 RESULTS Characteristics of included studies 93 Prevalence 95 Prevalence in children and adolescents 95 Prevalence in adults 97 Risk factors 97 Demographic and socioeconomic factors 97 Location: rural vs urban 97 Lifestyle factors 97 Nutritional factors 99 DISCUSSION 99 REFERENCES 100
This part of the chapter has been previously published as: Rachmi CN, Li M, Baur LA. Overweight and obesity in Indonesia: prevalence and risk factors – a literature review. Public Health. 2017; 147: 20-29.
PAGE | 91 public health 147 (2017) 20e29
Available online at www.sciencedirect.com
Public Health
journal homepage: www.elsevier.com/puhe
Review Paper Overweight and obesity in Indonesia: prevalence and risk factorsda literature review
C.N. Rachmi a,b,*, M. Li c, L. Alison Baur a,c a Discipline of Child & Adolescent Health, The Children's Hospital at Westmead, University of Sydney Clinical School, Sydney, Australia b Fakultas Kedokteran Universitas Padjadjaran, Bandung, Indonesia c Sydney School of Public Health, The University of Sydney, Sydney, Australia article info abstract
Article history: Objectives: Overweight/obesity is a problem faced by both high- and low- and middle- Received 15 November 2016 income countries. This review aimed to report published data on the prevalence of over- Received in revised form weight and obesity in Indonesian children, adolescents, and adults, along with the asso- 29 January 2017 ciated risk factors. Accepted 1 February 2017 Study design: Literature review. Methods: We conducted a literature search for articles published in English (through Medline via OvidSP, Scopus, Global Health via OvidSP and Web of Science electronic da- Keywords: tabases) and Indonesian languages (several websites, direct contact with Indonesian public Obesity health researchers, practitioners and Ministry of Health staff) from earliest to March 2016. Children We screened the results and ensured the quality of included studies with Loney's tools for Adolescent critically appraising prevalence or incidence studies. Adults Results: We included 17 papers on the topic which were available in full text and passed the Prevalence critical appraisal process. The prevalence of overweight/obesity has increased over the past Indonesia two decades in Indonesian children, adolescents and adults. Prevalence rates are higher in boys than girls among children, but higher in females in the adolescent and adult age groups. The prevalence of overweight/obesity is also higher in those living in urban areas and with higher income or education. Conclusions: Overweight/obesity is a serious public health problem in Indonesia with a continuing increase in its prevalence. Interventions at the household level and beyond are needed to successfully lower the prevalence of overweight/obesity in the country. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Discipline of Child & Adolescent Health, The Children's Hospital at Westmead, University of Sydney Clinical School, Locked Bag 4001, Westmead, NSW 2145, Australia. Tel.: 61 82121786786, 61 450531511. þ þ E-mail addresses: [email protected], [email protected] (C.N. Rachmi). http://dx.doi.org/10.1016/j.puhe.2017.02.002 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved. PAGE | 92 public health 147 (2017) 20e29 21
Asia OR LMIC* OR developing country*). We combined the Introduction results from 1) to 6) with AND. Searches in other databases used similar search strategies and key words. The prevalence of overweight/obesity has escalated in many We also identified papers written in the Indonesian lan- 1e3 developing countries which face the ‘nutrition transition’. guage through searches of: 1) the National Institute of Health Within these countries, the double burden of malnu- Research and Development website (http://ejournal.litbang. d trition whereby both undernutrition and overweight co- depkes.go.id/); 2) Indonesian journals; and 3) Indonesian d 1,3e7 exist within the population is a common finding. Ministry of Health websites (www.depkes.go.id). We con- South East Asia is one of the regions facing the double tacted several Indonesian public health researchers and burden of malnutrition. With a long history of undernutrition practitioners, and Ministry of Health staff, seeking full publi- in these countries, the prevalence of overweight and over- cations of conference abstracts. The complete list of sources is weight/obesity is now increasing, ranging from 8% to 30% in available in Supplementary Table 1. adult males and 8e52% in adult females within the World Health Organization's South East Asia Region member coun- Screening of documents tries.7,8 Compared to other South East Asian countries, Indonesia currently has the highest prevalence rate of over- The English and Indonesian language publications we identi- 9 weight/obesity in under five children. fied were screened for relevance and quality. Indonesian In Indonesia the prevalence of childhood stunting remains language documents usually provided an English abstract. high, whereas the prevalence of overweight/obesity in all age The full text of Indonesian-language papers was read by CNR 10,11 groups is rising. Evidence regarding the magnitude and and HJ, who are native Indonesian-language speakers. Three distribution of overweight/obesity in Indonesia is still scarce, authors assessed titles and abstracts to ensure topics were with no single report documenting the results from various relevant to the review. The full text of the relevant documents studies in the country. A better understanding of the preva- was then assessed for quality to be included in the review. lence of overweight/obesity, along with its determinants, will aid decision-making about the best way to implement and/or Critical appraisal process evaluate a range of policies and strategies. This article reviews published data on the prevalence of There are few well-evaluated tools for critically appraising overweight and obesity in Indonesian adults, and children and prevalence or incidence studies. We used that created by Loney adolescents, and the reported associated risk factors. et al.,12 which was developed for the purpose of appraising prevalence or incidence studies. The quality of the documents was assessed against three quality assessment criteria: 1) val- Methods idity of the study methods (6 points); 2) interpretation of results (1 point); and 3) applicability of results (1 point; details available Inclusion and exclusion criteria in Supplementary Table 2).12 For this review, we included pa- pers that scored 5 or more, out of a total score of 8, and covered We included studies that: 1) reported data for people living in at least two of three criteria. The critical appraisal guideline is Indonesia (including multi-country studies with Indonesia as available in Supplementary Table 2. one of the countries); 2) investigated the prevalence of, and/or Subsequently, data were grouped into the following cate- risk factors for, overweight/obesity; 3) published in a peer- gories and subcategories: 1) prevalence, (a) children and ado- reviewed journal; and 4) were available in full text (not lescents, (b) adults; and 2) risk factors, (a) demographic and editorial, commentary, or abstract for conferences). Docu- socioeconomic (b) location, (c) lifestyle, and (d) nutritional ments were excluded if they were: not presented in either factors. English or Indonesian; qualitative studies; or focussing on clinical features of overweight/obesity, interventions, treat- Definitions of overweight/obesity ment of overweight/obesity and/or the prevalence/treatment of obesity-related disorders. There are many definitions of overweight/obesity in children and adults. We have summarised the different definitions in Literature search strategy Table 1. However, some of the research included in this review used a combination of more than one definition and/or cut-off We conducted a literature search in four electronic databases points. for papers published in English language: Medline via OvidSP, Scopus, Global Health via OvidSP and Web of Science, from earliest until March 2016. The search in Medline used the Results following key words/MeSH terms: 1) overweight and obesity (overweight OR obese* OR weight OR body weight OR fat), 2) Characteristics of included studies prevalence (prevalence OR number* OR case*, incidence OR survey), 3) risk factor (risk factor* OR cause OR aetiology), 4) Fig. 1 shows the results of the literature search, screening and adults (adult* OR mother OR father OR parent* OR caretaker OR eligibility assessment. elder*), 5) children and adolescents (child* OR adolescent* OR Of the 17 included papers on the prevalence of overweight school-age), 6) Indonesia (Indonesia OR South East Asia OR and obesity in Indonesia, ten scored 7 out of 8 with the critical PAGE | 93 22 public health 147 (2017) 20e29
Table 1 e Definitions of overweight/obesity, order by year of publication and alphabetically within the same year. No Criteria Abbreviated Age (years) Overweight Obesity
Children and adolescents 1 National Center for Health Statistics/ NCHS/WHO 1977 2e18 Weight-for-age z Weight-for-age z score WHO International Growth Reference score (WAZ) > 2 (WAZ) > 3 þ þ 2 World Health Organization WHO 1995 6e19 Body mass index BMI 95% percentile (BMI) 85% percentile 3 International Obesity TaskForce IOTF 2000 2e18 BMI-for-age-and sex > BMI-for-age-and sex > BMI of 25 at age 18 BMI of 30 at age 18 4 United States Centers for Disease CDC 2000 2e20 BMI 85% percentile BMI 95% percentile Control and Prevention (CDC) growth charts 5 World Health Organization Child WHO growth 0e5 BMI-for-age z score BMIZ 3 SD þ Growth Standards 0e5 years standards 2006 (BMIZ) 2 standard þ deviation (SD) 6 World Health Organization child WHO growth 5e19 BMIZ 1 SD BMIZ 2 SD þ þ growth reference data for 5e19 years reference 2007 Adults 1 World Health Organization WHO 1995 18 Body Mass Index (BMI) > 25 BMI 30 2 Gurrici cut-off values for obesity Gurrici 1998 18 N/A BMI >27 for Indonesians 3 World Health Organization; The WHO 2000 18 BMI 25 BMI 30 International Classification of adult underweight, overweight and obesity according to BMI 4 World Health Organization; appropriate WHO 2004 18 BMI 23 BMI 25 body mass index for Asian populations
Fig. 1 e Literature searching process on the prevalence and risk factors of overweight/obesity in Indonesia. PAGE | 94 public health 147 (2017) 20e29 23
appraisal tool, six scored 6 and one scored 5. Table 2 sum- marises the critical appraisal analysis.
Among the 17 studies, nine included children and adoles- Score cents as participants, one included both children and adults (2 years and older), and seven included only adults. Two of these were multiple country studies in South East Asia that included Indonesia.13,14 Twelve studies were national-level stud- ies4,11,14e23 and five were regional studies.13,24e27 Ten studies described Study subjects also reported related risk factor: five of these studies in chil- dren and adolescents and five in adults.
Prevalence
A summary of the included studies is available in Table 3, with analysis Confidence studies presented in order of the year of publication, and then alphabetically for studies in the same year. intervals, subgroup
Prevalence in children and adolescents
The studies used a range of definitions for overweight and
obesity, making it impossible to compare the results across Adequate studies. response rate, A secondary data analysis4 using the Indonesian Family refusers described Life Survey (IFLS)28 wave 1 (1993), data found the prevalence of overweight/obesity in children aged 2e18 years was 5.1% (In- ternational Obesity TaskForce criteria). In another study conducted in 1999 in two areas in Central Java province,26 the children (boys aged 6.0e8.9 years and girls aged 6.0e7.9 years) Outcomes were divided into three groups: non-poor urban, rural and measured in
poor urban group. Overweight (International Obesity Task- an unbiased fashion Force criteria) was approximately five times higher in non- poor urban children (4.9%) compared with their rural coun- terparts (1.0%) and seven times higher compared with the
26 were the standard poor urban children (0.7%). Measures Using the World Health Organization (WHO) 1995 criteria, a 2001 survey in two cities, Bogor and West Jakarta, revealed the prevalence of overweight/obesity in private school boys aged size sample
8e10 years (32.7%) was almost three times higher than that for Adequate state school boys (11.0%). The prevalence of overweight in private school girls aged 8e10 years (21.2%) was nearly twice 13 that of their state school counterparts (12.5%). frame sampling One 2002 study compared three communities in Indonesia, Unbiased Jakarta, Jogjakarta and Kuta. The prevalence of obesity in stu- dents aged 13e15 years old was 8.0% (US Centers for Disease Control and Prevention [CDC] 2000 criteria), with a significantly YYNYYY YYY YYY Y Y7 YY YY YY Y YY YY Y NY Y Y YY Y YY Y Y YY Y Y YY Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y N/A Y Y Y N Y Y N Y Y N Y Y Y Y Y N N/A N/A Y N/A Y Y Y Y N/A Y N/A Y N/A Y Y N/A N/A Y Y N/A Y Y N/A N/A N N/A Y N/A Y N/A Y N/A Y Y N/A Y Y N/A Y Y Y Y Y 7 Y Y 6 Y Y Y Y 5 Y 6 Y Y 6 Y Y 6 Y 6 Y 7 Y 7 Y 7 Y 7 Y 7 7 6 7 7 higher prevalence being associated with an urban setting
(Jakarta), private schools, higher income families and children population 27 who had higher computer or play station usage (>3 h). sample or whole A secondary data analysis of the 2005 Nutritional Status
15 15 23
Survey, found 6.0% of children and adolescents aged 6e18 18 22 21 years in Jakarta were overweight, with a higher prevalence in 13 25 14 16 11 27 boys compared with girls (CDC 2000 criteria). There was an 24 20 4 17 inverse association between age and the prevalence of over- 26 weight and obesity, where the highest prevalence of obesity 19 was found in younger school children aged 6e9 years (10.9%),
with lower prevalence rates in older age groups i.e. 9.1e12.0, Summary of critical appraisal of studies included in the review, order by year of publication and alphabetically within the same year.
12.1e15.0 and 15.1e18 years (7.7%, 5.1% and 3.7%, e Soekirman et al., 2002 Julia et al., 2004 Doak et al., 2005 Prihantini and Jahari, 2007 Collins et al., 2008 Azhari et al., 2009 Razzaque et al., 2009 Susilowati, 2011 Roemling and Qaim, 2012 Sari and Mansyur, 2012 Diana et al., 2013 Sandjaja et al., 2013 Sandjaja et al., 2013 Popkin and Slining, 2013 Sari and Amaliah, 2014 Sohn, 2014 Rachmi et al., 2016 respectively).15 A 2007 study among 3108 high-school students (16e18 Table 2 1 Abbreviations: Y, yes; N, no; N/A, information not available in the article. 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 years) in Aceh24 found that 2.7% had obesity (CDC 2000 No Authors, year of publication Random PAGE | 95 24
Table 3 e Summary of results from included studies; order by year of publication and alphabetically within the same year. No Authors, year of publication Year of Level of study Number of participants Criteria used Prevalence of overweight/obesity study [O/O] or obesity 1 Soekirman et al., 200213 2001 Regional 1367 children (8e10 years) WHO 1995 [O/O]dboys: 17.8%; girls: 15.3% 2 Julia et al., 200426 1999 Regional 1738 boys (6.0e8.9 years) and 1272 IOTF 2000 [O/O]: 3.5% girls (6.0e7.9 years) WHOeNCHS 1995 [O/O]: 2.3% 3 Doak et al., 20054 1993 National 6461 households (children and IOTF 2000 and WHO 2000 [O/O]dchildren and adolescents: adolescents 2e18 years; adults) 5.1%; adults: 14.6% 15
4 Prihantini and Jahari, 2007 2005 National 7195 children and adolescents CDC 2000 [O/O]: 6.0% 20 (2017) 147 health public (6e18 years) 5 Collins et al., 200827 2002 Regional 1758 adolescents (12e15 years) CDC 2000 Obesity: 8.0% 6 Azhari et al., 200924 2007 Regional 3108 adolescents (16e18 years) CDC 2000 Obesity: 2.7%; boys: 2.2%; girls: 3.2% 7 Razzaque et al., 200925 2005 Regional 2000 adults (25e64 years) WHO 2000 [O/O] e male: 10.4%; female: 27.9% 8 Susilowati, 201117 2010 National 44,604 adolescents (10e19 years) WHO 2004 [O/O]: 9.8%; boys: 8.7%; girls: 10.8% 9 Roemling and Qaim, 201223 1993, National 47,743 adults (20e75 years) WHO 2004 and Gurrici [O/O] e 1993 male: 20.8%; female: 32.0% 2000, 2007 Indonesia-specific 1998 [O/O] e 2000 male: 24.6%; female: 39.6% [O/O] e 2007 male: 31.2%; female: 48.8% 10 Sari and Mansyur, 201222 2010 National 16,142 adults ( 60 years) WHO 2000 [O/O]: 19.7% 11 Diana et al., 201320 2010 National 57,167 women (19e55 years) WHO 2000 [O/O]: 29.4%; rural 25.9%; urban 32.4% 12 Popkin and Slining, 201318 1993, 2007 National 3873 and 10, 285 women (19e49 WHO 2004 [O/O] e 1993: 17.7%; 2007: 31.2% years) in 1993 and 2007, respectively 14 e 13 Sandjaja et al., 2013 2011 National 7211 children (0.5e12 years) WHO 2006 for 0e5 years; [O/O]: 15.8%; rural: 5.1%; urban: 10.7% 29 WHO 2007 for 5e19 years 14 Sandjaja et al., 201316 2011 National 6746 children (6e12 years) WHO 2007 for 5e19 years [O/O]: 11.5% 15 Sohn, 201419 2007 National 4846 men and 5345 women ( 40 years) WHO 2000 and WHO 2004 Obesity (BMI 25)dmale: 20.1%; female: 36.2% Obesity (BMI 27)dmale: 10.1%; female: 21.8% 16 Sari and Amaliah, 201421 2007, 2010 National 408,351 adults (2007) and 125,563 BMI 25 [O/O]d2007: 19.8%; male: 14.8%; Female: 24.6% adults (2010) (19e55 years) [O/O]d2010: 23.0%; male: 17.0%; female: 28.7% 17 Rachmi et al., 201611 1993, 1997, National 4101 children (2.0e4.9 years) WHO 2006 for 0e5 years At risk/[O/O]d1993: 10.3% 2000, 2007 (at risk BAZ > 1) At risk/[O/O]d1997: 10.6% þ At risk/[O/O]d2000: 11.7% At risk/[O/O]d2007: 16.5%
Abbreviations: WHO: World Health Organization; IOTF: International Obesity Task Force; NCHA: National Center for Health Statistics; CDC: Centers for Disease Control and Prevention; BMI: body mass index; BAZ: BMI for age Z score.
PAGE | 96 public health 147 (2017) 20e29 25
criteria), with a higher prevalence in girls (3.2%) compared focussing on older participants (>60 years) of the 2010 Ris- with boys (2.2%). Another study17 analysed part of the Ris- kesdas found that the prevalence of overweight (WHO 2000 kesdas (Basic National Health Survey) 2010 data to determine criteria) was higher in women (22.9%) compared with men the prevalence of overweight (WHO 2004 criteria) in adoles- (16.0%).22 cents aged 10e19 years. The prevalence of overweight and In summary, in adults, the prevalence of overweight has obesity was higher among girls (10.8%) than boys (8.7%). increased since the early 1990s,21,23 with higher prevalence Another study provided a comparison of prevalence rates rates in women compared with men.19,21e23,25 (WHO Growth Standard 2006) in four different years (1993, 1997, 2000 and 2007) using the IFLS28 data set in children aged Risk factors 2.0e4.9 years. The prevalence of those who were ‘at risk’ (> 1 þ SD)/overweight/obese in those years was 10.3%, 10.6%, 11.7% A summary of the included studies on risk factors is shown in and 16.5%, respectively, with a significant (P < 0.05) difference Table 4. between boys (14.4% mean prevalence overall) and girls (10.9%).11 Demographic and socioeconomic factors A 2011 study conducted throughout Indonesia14 found the prevalence of overweight/obesity in children aged 0.5e12 In children and adolescents, those in the youngest age group years in urban areas was more than double that of children in were more likely to be overweight/obese compared with older rural areas (10.7% and 5.1%, respectively; WHO Growth Stan- age groups.11,15,17 Boys had significantly higher prevalence of dard 2006, WHO Growth Reference 2007). In another 2011 overweight than girls.11,15,27 In 2010, girls aged 10e19 years study,16 the prevalence of obesity in children aged 6e12 years who were married had a 1.87 fold greater chance of being over all Indonesia was 11.5% (WHO Growth Reference 2007). overweight (95% CI: 1.48e2.36) compared with those who were In summary, among child participants, the prevalence of not married or who were divorced.17 overweight was higher in boys compared with girls,11,13e15,24 A higher prevalence of obesity was found in children of in urban areas,26,27 in those attending private schools13,27 working mothers compared with stay-at-home mothers and in more recent years.11 In adolescents, however, the (5.4%), especially mothers who worked in the private sector prevalence of overweight was higher in girls than in boys.17,24 (26.9%). Other significant factors associated with higher prevalence rates of overweight/obesity in children included a Prevalence in adults higher family income,27 being in the higher economic quintile (a sum of the household's food and non-food expenditures),17 Four studies conducted secondary data analyses of the IFLS.28 having, mother/father who were overweight/obese,11 and The first one used data from the years 1993, 2000 and 2007,23 higher parental education (father/mother went to and showed a substantial increasing prevalence of over- university).11,15 weight/obesity at each measurement period in both sexes. For adults, women had a higher chance of being over- Within the three waves, the rates of overweight (WHO 2004 weight compared to men.21,22,25 Other factors related to criteria) and obesity (Gurrici 1998 criteria) were higher in overweight included being married (for women),20,21,23 having women compared with men. The second study4 used 1993 a high living standard/expenditure (indicating higher in- data and found the prevalence of overweight/obesity (WHO come),20,23 or being in the highest socioeconomic groups.21 2000 criteria) in adults was 14.6% (body mass index [BMI] 25 kg/m2). The third study used the 1993 and 2007 data18 and Location: rural vs urban found that the prevalence of overweight (WHO 2000 criteria) in women aged 19e49 years nearly doubled over the 14 years. One study found that adolescent girls living in urban areas The fourth study used the 2007 IFLS28 data but in different age had a 1.26 fold greater chance of being overweight (95% CI: groups (>40 years).19 Weight status was calculated against the 1.09e1.45) compared with their counterparts living in rural WHO 2000 and the WHO 2004 criteria. Regardless of the cut-off areas.17 Two other studies identified that the chance of being points used, the prevalence of obesity in Indonesia was higher overweight/obese in adults aged 19e55 years was significantly in women compared with men. higher in those living in urban areas.20,21 A prevalence survey in Purworejo in 2005 found the prev- alence of overweight (WHO 2000 criteria) was much higher in Lifestyle factors women than men aged 25e64 years (23.9% and 9.5%, respectively).25 Undertaking regular household chores (e.g. sweeping and One secondary data analysis of the National Health Survey mopping the floor of the house, doing laundry) was associated 15 (Riskesdas) in years 2007 and 2010,21 found the combined with low risk of overweight in children. One cross-sectional prevalence of overweight/obesity (BMI 25 kg/m2) in adults study showed a strong association between obesity and the aged 19e55 years increased from 19.8% in 2007 to 23.0% in level of physical activity. Students aged 16e18 years who were 2010, with a higher prevalence in women (24.6% and 28.7%, inactive, assessed using the International Physical Activity respectively) compared with men (14.8% and 17.0%, respec- Questionnaire, had a 2.58 higher risk (95% CI: 1.37e4.85) of 24 tively) in both years. being obese compared with their active counterparts. 20,23 Another study using the 2010 Riskesdas data set of women Two studies revealed that adults who had lighter levels aged 19e55 years found the prevalence of overweight/obesity of daily physical activity, such as having sedentary or house- (WHO 2000 criteria) in women to be 29.4%.20 A 2012 analysis keeping jobs, were also at greater risk of being overweight. PAGE | 97 26
Table 4 e Risk factors associated with overweight/obesity; order by year of publication and alphabetically within the same year. No Authors, year of Year of study Demographic and Location (rural Lifestyle Nutritional publication Socioeconomic vs urban)
1 Prihantini and 2005 Being in the youngest Rarely helping in household chores Often eats fried food Jahari, 200715 age group (6e9 years) including oily/fried snacks Being male (>4 /week) Â Higher parental education Habits of having supplement (university) tablets/drinks Parental occupation 2 Collins et al., 200827 2002 Being male Higher family income (>2 million IDR) 3 Azhari et al., 200924 2007 Physical activity (less active) Physical fitness (less fit) 4 Razzaque et al., 200925 2005 Being female ulchat 4 21)20 (2017) 147 health public Age (increasing from 25 to 34, 35e44, 45e54 age group and then decreasing at 55e64 age group) 5 Susilowati, 201117 2010 Being in the youngest Living in urban area age group (10e12 years) Being married Being in the higher quintile economic status 6 Roemling and 1993, 2000, 2007 Being married Lighter physical activity at work Higher food expenditure Qaim, 201223 Having a high living standard Having television (leisure related) (individual) (equals to higher expenditure) Higher meat and dairy share (individual) e 29 7 Sari and Mansyur, 201222 2010 Being female Living in urban area Presence of caregiver 8 Diana et al., 201320 2010 Being married Living in urban area Has sedentary physical activity Higher income Protective factor: higher education 9 Sari and Amaliah, 201421 2007, 2010 Being female Living in urban area Being married Higher socioeconomic status Work as civil servant/employee/military/ police 10 Rachmi et al., 201611 1993, 1997, Being in the youngest age 2000, 2007 group (2.0e2.9 years) Being male Having parents who were overweight/ obese, Having fathers with university education
PAGE | 98 public health 147 (2017) 20e29 27
Having a television in the house (as a proxy to estimate how Being married was a significant risk factor for over- people spent their leisure time) was positively associated with weight/obesity, both in adults and adolescents.17,20,21,23 This BMI.23 In older adults, having a caregiver at home was asso- is an important culturally related risk factor to be high- ciated with a 1.26 (95%CI: 1.14e1.40) fold chance of being lighted in this review, because marriage at a young age is overweight.22 still a relatively common practice in Indonesia. In 2016, 14% of Indonesian adolescents (both sexes) were married by the Nutritional factors age of 18 years.39 Strengths of this review include that it is the first A study which investigated dietary intake of vegetables, fast comprehensive report of studies on the prevalence of over- food, meat, fried foods and soft drinks by food frequency weight obesity in Indonesian population published in Indo- questionnaire on children and adolescents found that the nesian or English languages. Furthermore, we applied a prevalence of overweight was higher in those who ate fried systematic approach to the literature search, including local foods including oily/fried snacks >4 times per week compared Indonesian journal websites. The use of the quality assess- with their counterparts who did not (7.2%e4.7%, respec- ment criteria12 enabled a standardised approach to judgement tively).15 In adults, overweight was associated with a greater and reduced the risk of subjectivity. A limitation is that there total food expenditure as well as an increased intake of meat were a number of identified studies not available in full text and dairy products.23 which therefore could not be included in the review. It has been suggested that developing countries should focus on conducting periodic nutrition surveillance to monitor Discussion the rate of overweight/obesity.9 In Indonesia a variety of sur- veys have been conducted with different definitions of over- This is the first review bringing together high-quality published weight and obesity being used, making it difficult to compare data on the prevalence of overweight and obesity in Indone- the prevalence rates within Indonesia as well as other parts of sian adults and children, and reported associated risk factors. the world. The prevalence of overweight/obesity has increased over the Our review highlights the need for involvement of a range past two decades in all the age groups. There are clear sex of stakeholders to lower the prevalence of overweight/obesity differences, with prevalence rates being higher in boys than in the country. Based on the results of this prevalence review, girls, although a higher prevalence in females in the adolescent the government of Indonesia will need to promote standard and adult age groups. At all ages, rates are higher in those living surveillance methods and cut-offs as well as periodically in urban areas. The findings also highlight the challenges with monitor the results of nutrition-related surveys and develop the use of different definitions and cut-off points for over- policies to deal with overweight/obesity. Based on the risk weight/obesity in the included studies; making it is impossible factors review, the government needs to fully implement their to compare prevalence rates between most of them. current policies on the promotion of healthy eating, support of The prevalence of overweight/obesity in adult Indonesians public transport and provision of public green space and increased rapidly over the 14 years from 1993 to 2007, by sidewalks.40,41 The government may also need to develop around 11% points (20.8%e31.2%) in men23 and 13e16% policies to promote physical activity42 and healthy eating (17.7%e31.2% and 32.0%e48.8%) in women.18,23 In contrast, environment in school settings.43 Other relevant strategies the prevalence of overweight/obesity globally rose by about include the regulation of food marketing and consideration of 8.1% in men and 8.2% in women over the 33 years from 1980 to specific taxes on unhealthy beverages.44 2013.2 Thus, within less than half the time frame, the preva- lence of overweight/obesity in Indonesian women nearly doubled compared with that of women globally. Our results Author statements show that the prevalence of overweight/obesity in Indonesian children is higher in boys compared with girls, comparable to Acknowledgements findings in children from South Asia29 and in Vietnam, China and Thailand.30 This might be related to cultural issues in This literature review is part of Cut Rachmi's PhD studies, Asian countries where families may still favour boys over which is supported by a scholarship from Lembaga Pengelola girls. Dana Pendidikan (LPDP), the Republic of Indonesia. We found that living in urban areas was strongly associ- The authors would like to thank Ms Hafiza Jusril for her ated with overweight/obesity, at any age,17,20e22 similar to help in identifying the available sources in the literature what has been described in several South Asian countries29 searching process. and other developing countries.31 We also showed that higher economic status was associated with overweight/ Ethical approval obesity, both in children and adults,17,21,23,27 a pattern also 29,31e37 seen in other low- and middle-income countries. None sought. Furthermore, the results showed that overweight/obesity was associated with low levels of physical activity (less active, rarely involved in household chores)15,24 and higher con- Funding sumption of fried foods.15 Such findings are in keeping with None declared. well-documented risk factors for obesity globally.38 PAGE | 99 28 public health 147 (2017) 20e29
Competing interests 13. Soekirman Hardinsyah, Jus'at I, Jahari AB. Regional study of nutritional status of urban primary schoolchildren. 2. West None declared. Jakarta and Bogor, Indonesia. Food Nutr Bull 2002;23(1):31e40. PubMed PMID: 11975367. 14. Sandjaja S, Budiman B, Harahap H, Ernawati F, Soekatri M, Authors' contributions Widodo Y, et al. Food consumption and nutritional and biochemical status of 05-12-year-old Indonesian children: the All authors contributed in designing the study; analysing and SEANUTS study. Br J Nutr 2013;110(Suppl. 3):S11e20. PubMed interpreting the data; writing, revising and reviewing the draft PMID: 24016762. 15. Prihantini S, Jahari AB. Risk factors of obesity in school of the manuscript. children age 6e18 years in DKI Jakarta. Penelit Dizi Dan Makanan (PGM) 2007;30(1):32e40. 16. Sandjaja Poh BK, Rojroonwasinkul N, Le Nyugen BK, references Budiman B, Ng LO, et al. Relationship between anthropometric indicators and cognitive performance in Southeast Asian school-aged children. Br J Nutr 1. Popkin BM, Adair LS, Ng SW. Now and then: the global 2013;110(Suppl. 3):S57e64. http://dx.doi.org/10.1017/ nutrition transition: the pandemic of obesity in developing S0007114513002079. Epub 2013/09/18. PubMed PMID: countries. Nutr Rev 2012;70(1):3e21. http://dx.doi.org/10.1111/ 24016767. j.1753-4887.2011.00456.x. 17. Susilowati D. The relationship between overweight and socio 2. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, demographic status among adolescent girls in Indonesia. Bul Margono C, et al. Global, regional, and national prevalence of Penelit Sist Kesehat 2011;14(1):1e6. Epub January 2011. overweight and obesity in children and adults during 18. Popkin BM, Slining MM. New dynamics in global obesity 1980e2013: a systematic analysis for the Global Burden of facing low- and middle-income countries. Obes Rev Disease Study 2013. Lancet 2014;384(9945):766e81. http:// 2013;14(Suppl. 2):11e20. http://dx.doi.org/10.1111/obr.12102. dx.doi.org/10.1016/S0140-6736(14)60460-8. PubMed PMID: Epub November 2013. 24880830; PubMed Central PMCID: PMCPMC4624264. 19. Sohn K. Sufficiently good measures of obesity: the case of a 3. World Health Organization. Report of the first meeting of the ad developing country. J Biosoc Sci 2014:1e21. http://dx.doi.org/ hoc working group on science and evidence for ending childhood 10.1017/S0021932013000692. Epub 15 January 2014. obesity. Geneva, Switzerland: World Health Organization; 20. Diana R, Yuliana I, Yasmin G, Hardinsyah. Risk factors of 2014. overweight among Indonesian women. J Gizi Dan Pangan 4. Doak CM, Adair LS, Bentley M, Monteiro C, Popkin BM. The 2013;8(1):1e8. dual burden household and the nutrition transition paradox. 21. Sari K, Amaliah N. Sociodemographic factors and obesity on Int J Obes (London) 2005;29(1):129e36. http://dx.doi.org/10.1038/ adults in Indonesia, year 2007 and 2010 (data analysis of Basic sj.ijo.0802824. Epub 2004/10/27. PubMed PMID: 15505634. Health Survey 2007 and 2010). J Ekol Kesehat 2014;13(4):328e39. 5. Vorster HH, Bourne LT, Venter CS, Oosthuizen W. 22. Sari K, Mansyur M. Female, live in urban, and the existence of Contribution of nutrition to the health transition in a caregiver increased risk over-nutrition in elderly: an developing countries: a framework for research and Indonesian national study 2010. Health Sci Indones intervention. Nutr Rev 1999;57(11):341e9. Epub 2000/01/11. 2012;3(1):9e14. Epub June 2012. PubMed PMID: 10628185. 23. Roemling C, Qaim M. Obesity trends and determinants in 6. Worsley A. Food habits and beliefs in transitional societies. Indonesia. Appetite 2012;58(3):1005e13. http://dx.doi.org/ Asia Pac J Clin Nutr 1998;7(3/4):287e92. Epub 1998/12/01. 10.1016/j.appet.2012.02.053. PubMed PMID: 22402303. PubMed PMID: 24393685. 24. Azhari Ishak S, Wilujeng LK. Relationship between physical 7. International Food Policy Research Institute Global Nutrition activity, fitness and body image with obesity among senior Report. Actions and accountability to advance nutrition and high school students at Banda Aceh Municipality. Bul Penelit sustainable development. Washington, DC: International Food Sist Kesehat 2009;12(3):319e29. Epub Juli 2009. Policy Research Institute, 2015; 2015. 25. Razzaque A, Nahar L, Van Minh H, Ng N, Juvekar S, Ashraf A, 8. World Health Organization Regional Office for South East Asia et al. Social factors and overweight: evidence from nine Asian Region. Noncommunicable diseases in the South-East Asia region: INDEPTH Network sites. Glob Health Action 2009;2. http:// situation and response 2011. India: World Health Organization, dx.doi.org/10.3402/gha.v2i0.1991. Epub 2009/12/23. PubMed Regional Office for South East Asia Region; 2011. PMID: 20027257; PubMed Central PMCID: PMC2785100. 9. ASEAN/UNICEF/WHO. Regional report on nutrition security in 26. Julia M, van Weissenbruch MM, de Waal HA, Surjono A. ASEAN, vol. 2. Bangkok: UNICEF; 2016. Influence of socioeconomic status on the prevalence of 10. Badan Penelitian dan Pengembangan Kesehatan Kementrian stunted growth and obesity in prepubertal Indonesian Kesehatan Republik Indonesia. Status Gizi. In: Departemen children. Food Nutr Bull 2004;25(4):354e60. PubMed PMID: Kesehatan Indonesia. Laporan Hasil Riset Kesehatan Dasar 15646313. Indonesia tahun 2013, Riskesdas Dalam Angka. Jakarta, 27. Collins AE, Pakiz B, Rock CL. Factors associated with obesity Indonesia: CV Kiat Nusa; 2014. p. 386e415. in Indonesian adolescents. Int J Pediatr Obes 2008;3:58e64. 11. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight http://dx.doi.org/10.1080/17477160701520132. and overweight in children aged 2.0-4.9 years in Indonesia: 28. Rand Corporation. The Indonesian Family Life Survey (IFLS) Santa prevalence trends and associated risk factors. PLoS One Monica. Califonia: Rand Corporation; 2014 [cited 2014 22 2016;11(5):e0154756. http://dx.doi.org/10.1371/ February]. Available from: http://www.rand.org/labor/FLS/ journal.pone.0154756. Epub 2016/05/12. PubMed PMID: IFLS.html. 27167973; PubMed Central PMCID: PMC4864317. 29. Mistry SK, Puthussery S. Risk factors of overweight and 12. Loney PL, Chambers LW, Bennett KJ, Roberts JG, Stratford PW. obesity in childhood and adolescence in South Asian Critical appraisal of the health research literature: prevalence countries: a systematic review of the evidence. Public Health or incidence of a health problem. Chronic Dis Can 2015;129(3):200e9. http://dx.doi.org/10.1016/ 1998;19(4):170e6. Epub 1999/02/24. PubMed PMID: 10029513. j.puhe.2014.12.004. PubMed PMID: 25746156. PAGE | 100 public health 147 (2017) 20e29 29
30. Li M, Dibley MJ. Child and adolescent obesity in Asia. In: 2015;9(3):220e7. http://dx.doi.org/10.1016/j.orcp.2014.10.219. Baur LA, Twigg SM, Magnusson RS, editors. A modern epidemic, PubMed PMID: 25434691. experts perspectives on obesity and diabetes. Sydney: Sydney 38. Malik VS, Willet WC, Hu FB. Global obesity: trends, risk University Press; 2012. p. 171e90. factors and policy implications. Nat Rev Endocrinol 2013;9(1). 31. Gupta N, Goel K, Shah P, Misra A. Childhood obesity in 39. UNICEF. The State of the World's Children 2016: a fair chance for developing countries: epidemiology, determinants, and every child. USA: United Nations Children's Fund; 2016. prevention [Review] Endocr Rev 2012;33(1):48e70. http:// 40. World Health Organization. Obesity: preventing and managing dx.doi.org/10.1210/er.2010-0028. the global epidemic; report of a WHO consultation. Geneva, 32. Mo-suwan L, Geater AF. Risk factors for childhood obesity in a Switzerland: World Health Organization; 1999. Available transitional society in Thailand. Int J Obes Related Metab Disord from: www.who.int/iris/handle/10665/42330. 1996;20(8):697e703. Epub 1996/08/01. PubMed PMID: 8856390. 41. World Health Organization. Global strategy on diet, physical activity 33. Mo-suwan L, Tongkumchum P, Puetpaiboon A. Determinants and health. Geneva, Switzerland: World Health Organization; of overweight tracking from childhood to adolescence: a 5 y 2004. Available from: www.who.int/dietphysicalactivity/ follow-up study of Hat Yai schoolchildren. Int J Obes Related strategy/eb11344/strategy_english_web.pdf. Metab Disord 2000;24(12):1642e7. Epub 2000/12/29. PubMed 42. World Health Organization. Global recommendations on physical PMID: 11126218. activity for health. Geneva, Switzerland: World Health 34. Sakamoto N, Wansorn S, Tontisirin K, Marui E. A social Organization; 2010. Available from: https://www.ncbi.nlm. epidemiologic study of obesity among preschool children in nih.gov/books/NBK305057/. Thailand. Int J Obes Related Metab Disord 2001;25(3):389e94. 43. World Health Organization. School policy framework: http://dx.doi.org/10.1038/sj.ijo.0801477. Epub 2001/04/25. implementation on the WHO global strategy on diet, physical PubMed PMID: 11319637. activity and health. Geneva, Switzerland: World Health 35. Li M, Dibley MJ, Sibbritt D, Yan H. Factors associated with Organization; 2008. Available from: www.who.int/ adolescents' overweight and obesity at community, school dietphysicalactivity/SPF-en-2008.pdf. and household levels in Xi'an City, China: results of 44. World Health Organization. Fiscal policies for diet and prevention hierarchical analysis. Eur J Clin Nutr 2008;62(5):635e43. http:// of noncommunicable diseases: technical meeting report, 5e6 May dx.doi.org/10.1038/sj.ejcn.1602757. Epub 2007/04/19. PubMed 2015. Geneva, Switzerland: World Health Organization; 2016. PMID: 17440524. Available from: apps.who.int/iris/bitstream/10665/250131/1/ 36. Dieu HT, Dibley MJ, Sibbritt D, Hanh TT. Prevalence of 9789241511247-eng.pdf. overweight and obesity in preschool children and associated socio-demographic factors in Ho Chi Minh City, Vietnam. Int J Pediatr Obes 2007;2(1):40e50. http://dx.doi.org/10.1080/ Appendix A. Supplementary data 17477160601103922. Epub 2007/09/01. PubMed PMID: 17763009. 37. Koirala M, Khatri RB, Khanal V, Amatya A. Prevalence and Supplementary data related to this article can be found at factors associated with childhood overweight/obesity of http://dx.doi.org/10.1016/j.puhe.2017.02.002. private school children in Nepal. Obes Res Clin Pract
PAGE | 101
Electronic Supplementary File Supplementary Table 1. List of sources of the literature searching process English Languange Indonesian Language 1. Medline via Ovid 1. Jurnal Kesehatan Badan Penelitian dan Pengembangan Kesehatan (National Institute of Health Research and Development) which contains: a) Buletin Penelitian Kesehatan b) Media Penelitian dan Pengembangan Kesehatan c) Health Science Journal of Indonesia d) Penelitian Gizi Makanan e) Buletin Penelitian Sistem Kesehatan f) Media Gizi Mikro Indonesia g) Journal for Social Determinants of Health 2. Scopus 2. Other Indonesian Journals, for example: Bandung Medical Journal, Journal of the Indonesian Medical Association, The Journal of the Indonesian Medical Students’ Association, Media Komunikasi Kementrian Kesehatan RI, Pharmamedika Health Journal, Medical Journal of Indonesia, Medicinus, Kesmas The National Journal of Public Health); 3. Global Health via Ovid SP 3. Indonesian Ministry of Health websites (www.depkes.go.id) 4. Web of Science 4. Personal communications with Indonesian public health researchers and practitioners, and Ministry of Health staff
Supplementary Table 2. Guidelines for critically appraising studies of prevalence or incidence of a health problem by Loney at al [14] Criteria Scoring system A. ARE THE STUDY METHODS VALID? 1. Are the study designs and sampling method appropriate for the research question? 1 point 2. Is the sampling frame appropriate? 1 point 3. Is the sample size adequate? 1 point 4. Are objectives, suitable and standard criteria used for measurement of the health 1 point outcome? 5. Is the health outcome measured in an unbiased fashion? 1 point 6. Is the response rate adequate? Are the refusers described? 1 point
B. WHAT IS THE INTERPRETATION OF THE RESULTS? 7. Are the estimates of prevalence or incidence given with confidence intervals and in 1 point detail by subgroup, if appropriate?
C. WHAT IS THE APPLICABILITY OF THE RESULTS? 8. Are the study subjects and the setting described in detail and similar to those of 1 point interest to you?
Total 8 points
PAGE | 102
Chapter 3: Quantitative studies
3.1 OVERVIEW AND AIMS 104 3.2 METHODS 3.2.1 Indonesian Family Life Survey (IFLS) 105 3.2.2 Funding bodies and ethics approval for IFLS 106 3.2.3 Information collected and participants 106 3.2.4 Data analysis approaches 108 3.3 REFERENCES 110
PAGE| 103
Chapter 3: Quantitative studies
3.1 OVERVIEW AND AIMS This chapter provides an introduction to the quantitative studies reported in the next three chapters. Chapters 4 and 5 present cross-sectional analyses of the Indonesian Family Life Survey dataset of children aged 2.0-4.9 years at four different time-points: 1993 (wave 1), 1997 (wave 2), 2000 (wave 3), and 2007 (wave 4). Chapter 6 presents the analysis of the same dataset longitudinally, following up children aged 2.0-4.9 years at baseline, 7 and 14 years later. A more detailed explanation of the analyses performed is illustrated in Figure 3.1.
1993 1997 2000 2007 Year of Survey (wave 1) (wave 2) (wave 3) (wave 4)
Cross-sectional Children aged 2.0 - 4.9 years in each wave of the studies Indonesian Family Life Survey
Longitudinal studies s e i d
u Start End
t 2.0 - 4.9 years at 1993
s 14 years
f
o follow-up
e
p Start End y 2.0 - 4.9 years at 1993
T 7 years follow-up
2.0 - 4.9 years at 2000 Start End 7 years follow-up
Figure 3.1 Studies conducted using the Indonesian Family Life Survey waves 1 (1993), 2 (1997), 3 (2000), and 4 (2007)
As detailed in chapter 2.3, there is gap in understanding of prevalence trends in malnutrition in Indonesia, especially amongst children. Thus, the first aim of the cross-sectional studies was to determine the prevalence trends and associated risk factors for three forms of malnutrition: stunting, underweight, and overweight/ obesity in Indonesian children aged 2.0- 4.9 years at four different time points: 1993, 1997, 2000, and 2007. The methods of analysis, results, and discussion are presented in Chapter 4 of this thesis.
PAGE| 104
The results from chapter 4 raised the question of whether a specific form of the double burden of malnutrition – concurrent stunting and overweight/ obesity within the same individual – occurs in Indonesian children. We aimed to determine the prevalence trends and associated risk factors for concurrent stunting and overweight/ obesity in Indonesian children aged 2.0- 4.9 years at four different time points: 1993, 1997, 2000, and 2007. Analysis was also performed to determine whether stunted children are at greater risk of being overweight or obese compared to their healthy height peers. The methods, results, and discussion are presented in Chapter 5 of this thesis.
The outcomes from chapters 4 and 5 raised a more specific question of whether the previously stunted children in the earlier waves (1993 and 2000) were more likely to become overweight, thin, or have high blood pressure in later years, namely in 2000 and 2007. The complete methods and results of the longitudinal studies are presented in Chapter 6 of this thesis.
3.2 METHODS 3.2.1 Indonesian Family Life Survey (IFLS) The Indonesian Family Life Survey (IFLS) is one of the largest on-going longitudinal surveys in Indonesia and the only one that targets the same participants originally included in the first wave (1993), and their family members and descendants1. The survey has been undertaken in 5 consecutive waves: 1993, 1997, 2000, 2007, and 2014. For the studies presented in this thesis, we only used the dataset from waves 1 (1993) to 4 (2007) of the survey. The IFLS has made substantial efforts in tracking down participants from the first wave (1993). In cases where participants have moved to another province in Indonesia, if that province is included in one of the survey areas, the IFLS team tracked down the family to be included in the survey. Thus, their re-contact rate is high when compared to longitudinal studies in many other countries1-4. In year 1993 (first wave), 13 out of 27 provinces in Indonesia were included in the survey. The selection of these provinces was based on the degree of heterogeneity of the population in each province, with the result that the survey was conducted in four out of the five main islands (Sumatra, Java, Kalimantan and Sulawesi).
The nature of this dataset, which tracks down “original participants” and their offspring at different time points, allows us to analyse the dataset both cross-sectionally and longitudinally. More detailed information about this survey has been published in several field reports2-4. In summary, in each province, the enumeration areas were determined according to the 1993 SUSENAS (National Socioeconomic Survey) sampling frame, from the
PAGE| 105
1990 census. The process resulted in 321 enumeration areas from those 13 provinces, and then randomly selected households from urban and rural enumeration areas1-3.
3.2.2 Funding bodies and ethics approval for IFLS The main funding bodies for this survey across all waves were the US National Institute for Child Health and Human Development and the US National Institute of Aging. Other funding bodies for this survey were the United States Agency for International Development, the World Health Organization, the Ford Foundation, John Snow (OMNI project), the Hewlett Foundation, the Futures Group (the POLICY project) and the International Food Policy Research Institute. The IFLS survey and its procedures were reviewed and approved by Institutional Review Boards in the USA (at Rand Corporation) and in Indonesia (at Universitas Indonesia for waves 1 and 2 and at Universitas Gadjah Mada for subsequent waves)1-3.
3.2.3 Information collected and participants The Indonesian Family Life Survey consists of community-facility and household questionnaires. The community-facility questionnaire includes information about village statistical records, the village head, association of family activities (PKK), government health centres, community health posts, physicians/private clinics, traditional practitioners and schools (elementary, junior, and senior high). This thesis focuses more on the household questionnaire books, which includes questions about household roster, household economy, adult information, ever-married woman information, child information, and anthropometry. A control book was used by the IFLS teams to document the sampling/enumeration record, re- contact information, and tracking forms. The IFLS involves participants of all ages, categorised as children (0-14 years), (ever- and never-married) adults (15-49 years), and all older adults (>50 years)1-3.
Informed consent was obtained from all individual participants included in the study. Parents (mother or father) gave informed consent for their children. For the control, household roster, and household economy books, the interviewer administered questions to the household head or spouse of the head. For the next three books (adult, ever-married woman, and child information), the interviewer collected individual level-data, with children being helped by their parent. Trained nurses measured height, weight, and other measurements for interviewed adults and children and recorded the results in the anthropometry book2-3. The complete list of modules collected during the survey is available in Table 3.1.
PAGE| 106
Table 3.1 List of modules collected in the household questionnaires of the Indonesian Family Life Survey in 19932-3
BOOK MODULE CONTENTS CA Anthropometry Height and weight K Control Book Sampling/ enumeration record Re-contact information Questionnaire tracking form I Household Roster Household member roster Household characteristics Consumption Outpatient care provider knowledge II Household Economy Farm business Non-farm business Labor and non-labor income Household assets Household economy shocks Health insurance III Adult Information Education history Employment history Marital history Birth summary (women 50+) Migration history Tobacco smoking Health condition Acute morbidity Self-treatment Outpatient utilization Inpatient utilization Non-co resident family roster and transfers Other transfers Individual assets, non-labor income IV Ever-married Woman Information Marital history Pregnancy summary Pregnancy and infant feeding history Contraceptive knowledge/use Contraceptive calendar V Child Information Child education history Child acute morbidity Child self-treatment Child outpatient utilization Child inpatient utilization
In this thesis, for the cross-sectional studies, I included children aged 2.0 to 4.9 years in each of the four waves (1993, 1997, 2000, and 2007) of the Indonesian Family Life Surveys. For the longitudinal studies, I included data from children who were aged 2.0 to 4.9 years at baseline (either in 1993 [Wave 1] or 2000 [Wave 3]) and their follow up data seven (1993 to 2000 and 2000 to 2007) and fourteen (1993 to 2007) years later.
PAGE| 107
3.2.4 Data analysis approaches The conceptual framework for the analyses in the next three chapters was adapted from the Davison and Birch Ecological Model for Childhood Overweight, to fit not only with the available variables in the Indonesian Family Life Survey dataset, but also to fit with stunting and underweight as other forms of malnutrition (see Figure 3.2)5.
All data analysis was conducted using STATA Data Analysis and Statistical Software version 13 (STATACorp, College Station, TX)6. The dataset of the survey is available at the RAND corporation website (http://www.rand.org/labor/FLS/IFLS.html) with prior registration and the requirement of reporting the papers published using this dataset.
The next process was data cleaning, which took nearly 5 months. Afterwards, I combined information from seven separate questionnaire books in each wave for the cross sectional studies. Although IFLS is a longitudinal study in nature, the survey recorded different children aged 2.0-4.9 years in each wave because the shortest length of time between each survey was 3 years. For example, children aged 2.0 years at the time of the first wave (1993) were 6.0 years at the next wave (1997). For the longitudinal studies, there was the additional step of linking the information from all child participants aged 2.0-4.9 years at baseline to the next waves (7 or 14 years later). Under the supervision of one of my supervisors, Dr. Agho, I created specific ID linking information for every child in earlier waves to the next waves.
During this process, I approached and maintained close communication with a senior research associate at Rand Organization, Ms Christine Peterson, to clarify issues I encountered along the way, including ethics approval. Every step of the analysis was checked and approved by Dr Kingsley Agho and all analyses based on the results were discussed with and approved by my other supervisors, Prof. Louise Baur and Prof. Mu Li.
PAGE| 108
FACTORS ASSOCIATED WITH MALNUTRITION IN INDONESIA
COMMUNITY LEVEL FACTORS 1. Region (province) 2. Housing area (urban/rural)
PARENTAL AND HOUSEHOLD LEVEL FACTORS 1. Parent's age 2. Parent's marital status 3. Parent's height/weight 4. Ever check up during pregnancy (mother) 5. Parent's education 6. Household Wealth Index
CHILD LEVEL FACTORS 1. Age 2. Sex 3. Birth weight 4. Ever breastfed 5. Age of weaning 6. Age starting complementary food 7. Child's height/weight
UNDERNUTRITION HEALTHY OVERNUTRITION 1. Stunting HEIGHT AND 1. At risk of overweight 2. Underweight WEIGHT 2. Overweight 3. Obese
Figure 3.2 Conceptual framework for analysis of risk factors related to malnutrition in Indonesian children aged 2.0 to 4.9 years; Data from the Indonesian Family Life Survey.
PAGE| 109
3.3 REFERENCES
1. Rand Corporation. The Indonesian Family Life Survey (IFLS) Santa Monica, Califonia: Rand Corporation; 2014 [cited 2014 22 February]. Available from: http://www.rand.org/labor/FLS/IFLS.html. 2. Frankenberg E, Karoly LA, Gertler P, Peterson CE, Wesley D. The 1993 Indonesian Family Life Survey: Overview and Field Report. Santa Monica, California: RAND, 1995. 3. Serrato C, Melnick G. The Indonesian Family Life Survey Overview and Descriptive Analysis. Santa Monica, California: RAND, 1995. Contract No.: DRU-1191-AID. 4. Frankenberg E, Thomas D. The Indonesia Family Life Survey (IFLS): Study Design and Results from Waves 1 and 2. Santa Monica, California: RAND, 2000. Contract No.: DRU-2238/1-NIA/NICHD. 5. Davison KK, Birch LL. Childhood overweight: a contextual model and recommendations for future research. Obesity reviews : an official journal of the International Association for the Study of Obesity. 2001;2(3):159-71. Epub 2002/07/18. PubMed PMID: 12120101; PubMed Central PMCID: PMC2530932. 6. Stata Corp. STATA statistical software: Release 13. College station, TX: Stata Corp LP; 2014.
PAGE| 110
Chapter 4: Stunting, underweight and overweight in children aged 2.0-4.9 years in Indonesia: prevalence trends and associated risk factors
INTRODUCTION 113 METHODS Indonesian Family Life Survey 113 The survey and data collection 113 Participants 114 Ethics 114 Outcome variables (anthropometric calculations) 114 Potential risk factors 114 Child factors 115 Parental and household factors 115 Community factors 115 Statistical analysis 115 RESULTS Characteristics of participants 116 Prevalence of stunting, underweight and ‘at risk of overweight’ and overweight/ 116 obesity Associated risk factors 118 DISCUSSION Statement of principal findings 120 Strengths and limitations 124 Comparison with other studies 124 IMPLICATIONS FOR RESEARCH, POLICY AND PRACTICE 125 REFERENCES 125
This chapter has been previously published as: Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight and overweight in children aged 2·0 -4·9 years in Indonesia: prevalence trends and associated risk factors. PLoSOne. 2016; 11(5):e0154756.
PAGE| 111 RESEARCH ARTICLE Stunting, Underweight and Overweight in Children Aged 2.0–4.9 Years in Indonesia: Prevalence Trends and Associated Risk Factors
Cut Novianti Rachmi1*, Kingsley E. Agho2, Mu Li3, Louise Alison Baur1,3
1 Discipline of Paediatrics and Child Health, The Children’s Hospital at Westmead (University of Sydney Clinical School), Sydney, NSW, Australia, 2 School of Science and Health, Western Sydney University- Campbelltown Campus, Sydney, NSW, Australia, 3 Sydney School of Public Health, The University of a11111 Sydney, Sydney, NSW, Australia * [email protected]; [email protected]
Abstract
OPEN ACCESS Objective Citation: Rachmi CN, Agho KE, Li M, Baur LA (2016) Stunting, Underweight and Overweight in Children The double burden of malnutrition affects many low and middle-income countries. This Aged 2.0–4.9 Years in Indonesia: Prevalence Trends study aimed to: a) determine temporal trends in the prevalence of underweight, stunting, and Associated Risk Factors. PLoS ONE 11(5): and at risk of overweight/ overweight or obesity in Indonesian children aged 2.0–4.9 years; e0154756. doi:10.1371/journal.pone.0154756 and b) examine associated risk factors. Editor: Yanqiao Zhang, Northeast Ohio Medical University, UNITED STATES Design Received: December 17, 2015 A repeated cross-sectional survey. This is a secondary data analysis of waves 1, 2, 3, and 4 Accepted: April 19, 2016 (1993, 1997, 2000, and 2007) of the Indonesian Family Life Survey, which includes 13 out Published: May 11, 2016 of 27 provinces in Indonesia. Height, weight and BMI were expressed as z-scores (2006 Copyright: © 2016 Rachmi et al. This is an open WHO Child Growth Standards). Weight-for-age-z-score <-2 was categorised as under- access article distributed under the terms of the weight, height-for-age-z-score <-2 as stunted, and BMI-z-score >+1, >+2, >+3 as at-risk, Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any overweight and obese, respectively. medium, provided the original author and source are credited. Results Data Availability Statement: All relevant data are There are 938, 913, 939, and 1311 separate children in the 4 waves, respectively. The prev- within the paper. The raw dataset are available from http://www.rand.org/labor/FLS/IFLS/access.html, for alence of stunting decreased significantly from waves 1 to 4 (from 50.8% to 36.7%), as did researchers who meet the criteria for access. the prevalence of underweight (from 34.5% to 21.4%). The prevalence of ‘at-risk’/over-
Funding: This research was performed as part of weight/obesity increased from 10.3% to 16.5% (all P<0.01). Stunting and underweight were Cut Novianti Rachmi’s PhD studies, for which she related to lower birth weight, being breastfed for 6 months or more, having parents who received a scholarship from Lembaga Pengelola were underweight or had short stature, and mothers who never attended formal education. Dana Pendidikan (LPDP), the Republic of Indonesia. Stunting was also higher in rural areas. Being at-risk, or overweight/obese were closely LPDP had no role in study design, data collection and related to being in the youngest age group (2–2 9 years) or male, having parents who were analysis, decision to publish, or preparation of the Á manuscript. overweight/obese or having fathers with university education.
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 112 1 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Competing Interests: The authors have declared Conclusions that no competing interests exist. The double burden of malnutrition occurs in Indonesian children. Development of policy to combine the management of chronic under-nutrition and over-nutrition is required.
Introduction Although once considered a problem only of developed countries, the prevalence of obesity has risen during the past 30 years in low and middle-income countries (LMICs).[1–4] At the same time, many LMICs are still dealing with the prevalent public health issue of under nutrition, a situation often described as the “double burden of malnutrition”. This “double burden” may occur in the same country, city or household (mother/ child pair), or in the same individual at different stages of his or her life.[5–14] In a 2013 Indonesian national survey, Riset Kesehatan Dasar (Riskesdas), the prevalence of underweight in children under 5 years was reported as 19.6%, that of stunting as 37.2%, and that of combined overweight and obesity as 11.9%. That survey also documented the preva- lence of combined stunting and overweight/obesity in this age group as 6.8%, a higher preva- lence than that of children who were of healthy height but who were overweight or obese (5.1%).[15] While such data are undoubtedly important, the inclusion of children aged less than two years, the period of life when stunting is developing,[16] may make interpretation of the results more difficult. The 2006 WHO Growth Standard also emphasises the use of BMI- for-age as the index of weight relative to height starting at 2 years of age.[17] To our knowledge, there has been no Indonesian report that has reported the trends in prevalence of underweight, stunting and overweight in children aged 2.0–4.9 years, along with the associated risk factors. A better understanding of the double burden of malnutrition in Indonesia, especially in early childhood, would aid decision-making about potential strategies to tackle the problem. Thus, the primary aim of this study was to determine the temporal trends in the prevalence of underweight, stunting, and at risk of overweight/ overweight or obesity in Indonesian chil- dren aged 2.0–4.9 years at four different time points: 1993, 1997, 2000, and 2007. The second- ary aim was to examine associated risk factors.
Methods Indonesian Family Life Survey The survey and data collection. We performed a secondary analysis of data collected in 1993 (wave 1), 1997 (wave 2), 2000 (wave 3), and 2007 (wave 4) from the Indonesian Family Life Survey (IFLS). The IFLS is a longitudinal survey of a stratified random sample of house- holds involving both questionnaires and anthropometric measurements, that is representative of 83% of the Indonesian population, and which was collected under the supervision of the Rand Corporation.[18] In the first wave (1993), the survey included 13 out of 27 provinces in Indonesia. These were selected based on the heterogeneity of the communities in these prov- inces, and included four of the five main islands in Indonesia (Sumatra, Java, Kalimantan and Sulawesi). Enumeration areas were further determined from these provinces based on the 1993 SUSENAS (National Socioeconomic Survey) sampling frame, from the 1990 census.[18–20] Details of the complete sampling scheme and survey methods have been described in field reports.[19, 20] In brief, IFLS randomly selected 321 enumeration areas within 13 provinces, then 20 households from each urban enumeration area, and 30 households from each of the rural enumeration areas. The first survey in 1993 included over 7,000 households and more
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 113 2 / 17 The DBM in 2.0–4.9 Year Indonesian Children
than 22,000 individuals. The second, third and fourth surveys aimed to contact all the partici- pants from the first survey and their offspring and split-off households, with a re-contact rate of more than 90% in each subsequent wave of the households in the first wave. To date, IFLS is the only survey that provides both cross-sectional and longitudinal data in Indonesia. In this paper, each wave was treated as a separate cross sectional survey. Data from the IFLS surveys are publicly available from the Rand Corporation website [18]. Data collections for wave 1 (1993) were performed by a total of 21 field teams. Two trained team members visited every household and collected questionnaire and anthropometric data (length/ height and weight) of each household member. The length/height of children was measured using Shorr measuring boards Model 420, and weight was measured using Seca Model 770 scales (SECA, Los Angeles, CA, USA). Children who were unable to be measured alone were weighed with their parent after their parent’s weight was ascertained. Standing height was used for every child older than two years.[19] Subsequent waves applied similar methods to that of the first one. In the IFLS survey, the information gathered from interviewees is recorded in seven separate questionnaire books.[20] We combined the household, household economy, child information, adult information and anthropometry from the four waves of datasets, resulting in 4,101 chil- dren with complete matching data for child (apart from birth weight), household and parental, and community level factors. Participants. For the purposes of this study, children aged 2.0–4.9 years within each wave who had complete records for height, weight, age, and sex were selected. Even though the sur- vey visited the same household in each wave, because waves were three to seven years apart, this resulted in a different group of 2.0–4.9 year old children in each of the waves. This is also the reason for the different number of children in each wave. For the potential risk factors anal- ysis, we combined data from all waves. Ethics. From personal communication with the IFLS research team, the IFLS survey and its procedures were reviewed and approved by Institutional Review Boards in the USA (at Rand Corporation, Santa Monica, California) and in Indonesia (Ethics Committees of Univer- sitas Gadjah Mada, Yogyakarta, and earlier at Universitas Indonesia, Jakarta). Written informed consent was obtained from all participants. Written consent was also obtained from the next of kin, caretakers, or guardians on behalf of the children enrolled in the survey.
Outcome variables (anthropometric calculations) 2 2 We calculated child body mass index (BMI; weight/height ; kg/m ) and expressed weight, height and BMI as z-scores using age and sex specific references from the LMS Growth Pro- gram.[21] Weight, height and BMI z-scores were then calculated against the 2006 WHO Child Growth Standards for children <5 years.[17] Children with weight-for-age-z-score <-2 were categorised as underweight, those with height-for-age-z-score <-2 as stunted, and those with BMI-z-score >+1, >+2, >+3 as at risk, overweight and obese, respectively.[17, 22] The ‘at-risk’ category (BMI-z-score >+1 and +2SD) was introduced by De Onis et al in 2010 for children less than five years.[22] Biologically implausible values were identified and discarded using cut off points from the WHO Anthro software (version 3.2.2, January 2011) for the Child Growth Standards (igrowup).[23]
Potential risk factors The conceptual framework we used was modified from the ecological model of childhood obe- sity, to include the potential risk factors associated with childhood malnutrition.[24] The
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 114 3 / 17 The DBM in 2.0–4.9 Year Indonesian Children
potential risk factors for childhood malnutrition were categorised into child, parental/ house- hold and community level factors. Child factors. These consisted of the child’s age (2.0–2.9, 3.0–3.9, and 4.0–4.9 years), sex, anthropometry (birth weight [low, healthy, and high birth weight], current weight and height), and nutrition history (ever breastfed, age of weaning [full cessation of breastfeeding], and age of starting complementary foods [less or equal to/more than 6 months]). Parental and household factors. These included parents’ age, marital status, anthropome- try (weight and height), and maternal antenatal care history (ever/never had check-up during pregnancy). The household level factors included the parents’ education (never attended any formal education, attended primary school, middle school, and university or higher), and the household wealth index that measures the economic status of a household. The household wealth index was constructed by assigning weights to eleven household assets, including the house the family lived in, another house/building, farmland, live stock/poultry/fishpond, vehi- cles (cars, boats, bicycles, motorbikes), household appliances (radio, tape recorder, TV, fridge, sewing or washing machine), savings or deposits or stocks, jewellery, receivables and other assets (household furniture and utensils) using the survey data and principle components anal- ysis method. The household wealth index was then calculated as the sum of the weighted scores for each item. The wealth index was used to rank all households across the four surveys. The household wealth index variable was categorised into five quintiles (poorest, poorer, middle, richer and richest) but for analyses in this study this index was divided into three categories. The bottom 40% of households was classified as poor households, the next 40% as the middle households and the top 20% as rich households. The complete formula and calculation of determining household wealth index have been described and used in several publications. [25–27] For the parents, BMI was categorised using the World Health Organization International Classification of underweight, overweight, and obesity.[28] Those with heights below -2 stan- dard deviations (SD) on the WHO 2007 Standard Growth Reference for School-aged Children and Adolescents (using the cut off points at age 19 years for male and female) were classed as having short stature.[29] Community factors. Community level factors included the housing area (urban/rural) and region. The latter was classified into four, primarily based on the main Indonesian islands: Sumatra, Java, Bali and Nusa Tenggara Barat (NTB), and Kalimantan and Sulawesi.
Statistical analysis Data were then analysed using STATA Data Analysis and Statistical Software version 13 (STA- TACorp, College Station, TX).[30] The Survey (‘Svy’) command was used to adjust for cluster- ing (enumeration areas) and sampling weights. The prevalence of underweight, stunting, at risk and overweight/obesity for each of the potential risk factors was calculated, and presented as percentage with 95% confidence intervals (CIs). Children’s weight, height and BMI within each wave were described using means and standard deviations. To determine the associations between the potential risk factors in child, parental and com- munity level and stunting, underweight, “at risk of overweight” and overweight/obesity in chil- dren, we combined the data from all waves. The GLLAMM (Generalised Linear Latent and Mixed Models) package with the logit link and binomial family in STATA was used to adjust for clustering and sampling weights. Univariate and multivariate binary logistic regression analysis was performed.[31] All dependent variables were categorised as dichotomous variables and odds ratios calculated. In the multivariable model, a staged modelling technique was employed.
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 115 4 / 17 The DBM in 2.0–4.9 Year Indonesian Children
In the first modelling stage, community level factors were first entered into the model to assess their associations with the study outcomes. A manually executed backward elimination method was conducted to select factors significantly associated with the outcomes. In the sec- ond model, the significant factors in the first stage were added to parental and household fac- tors and this was followed by backward elimination procedure. A similar approach was used for the child factors in the third stages. A staged stepwise regression was performed because it a) avoids the degree of correlation between the important predictors; b) produces better models and a better understanding of the data; c) produces the best model (estimates) for our study; and d) avoids reporting redundant predictors. The staged stepwise regression went from the most distal set of factors (community) to the most proximal (child), because the child is a sub- set of the community and any future population health intervention would start from the com- munity (more general) level and work towards the individual (more specific) level. To avoid any statistical bias, we tested and reported any collinearity in the final model. The odds ratios with 95% CIs were calculated in order to assess the adjusted risk of independent variables, and those with P < 0.05 were retained in the final model.
Results Characteristics of participants Table 1 shows the percentage of sample characteristics of the participants in each wave, with a total of 4,101 children aged 2.0–4.9 years. There were no significant differences in child’s age and sex across the four waves, although mean values for weight, height, and BMI increased over time. Of the 2,420 children who had a recorded birth weight, most (82.2%) had birth weight between 2.5–4.0 kg. Across all four waves, the vast majorities of children (>93%) were ever breastfed, and most (>87%) ceased being breastfed after the age of 6 months, but were given complementary food before the age of 6 months (>64%). At the parent level, most of the mothers and fathers were aged over 30 years at the time of the surveys, were currently married, and had a normal BMI. In the first three waves, just over one-half of mothers (53.3–55.5%) were classified as having short stature, whereas in wave 4, this had decreased to 45.5%. The prevalence of short stature in the fathers was higher across all four waves (57.1–59.8%). In all waves, a greater proportion of mothers ever had a check-up during their pregnancy and the majority of mothers and fathers had attended primary school or had a higher level of education. At the community level, in all four waves there was a similar number of children living in urban and rural areas.
Prevalence of stunting, underweight and ‘at risk of overweight’ and overweight/obesity Table 2 shows the trends in prevalence for stunting, underweight, ‘at risk of overweight’ and overweight/obesity across the four waves, as well as the prevalence for each of the potential risk factors. The prevalence of stunting decreased significantly over 14 years–from 50.8% in 1993 (wave 1) to 36.7% in 2007 (wave 4). The same phenomenon occurred with the prevalence of underweight, which decreased significantly from 34.5% in wave 1 to 21.4% in wave 4. In con- trast, the prevalence of combined at risk of overweight, overweight and obesity increased signif- icantly over the time period, from 10.3% to 16.5%. Fig 1 shows the distribution of BMI-z-score for each of the four survey waves. From wave 1 to 4, there was a successive shift to the right in both the mean BMI-z-score and BMI distribution.
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 116 5 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Table 1. Characteristics of children and parents in each wave of the Indonesian Family Life Survey, n (%) or mean (standard deviation).
Characteristics Wave 1 (n = 938) Wave 2 (n = 913) Wave 3 (n = 939) Wave 4 (n = 1311) CHILD LEVEL FACTORS Age 2.0–2.9 313 (33.4%) 274 (30.0%) 299 (31.8%) 425 (32.4%) 3.0–3.9 326 (34.8%) 257 (28.2%) 294 (31.3%) 454 (34.6%) 4.0–4.9 299 (31.8%) 382 (41.8%) 346 (36.9%) 432 (33.0%) Sex Male 501 (53.4%) 462 (50.6%) 482 (51.3%) 633 (48.3%) Female 437 (46.6%) 451 (49.4%) 457 (48.7%) 678 (51.7%) Birth weight (n = 2420) <2.5 kg 46 (7.8%) 27 (6.1%) 19 (7.7%) 75 (6.6%) 2.5-<4.0 kg 473 (80.6%) 357 (81.1%) 198 (79.8%) 963 (84.1%) 4.0 kg 68 (11.6% 56 (12.8%) 31 (12.5%) 107 (9.3%) Ever breastfed Yes 938 (100%) 850 (93.1%) 880 (93.7%) 1267 (96.6%) No 0 (0%) 63 (6.9%) 59 (6.3%) 41 (3.4%) Age of weaning <6 mo 41 (4.4%) 62 (6.8%) 80 (8.5%) 168 (12.8%) 6 mo 897 (95.6%) 851 (93.2%) 859 (91.5%) 1143 (87.2%) Age starting complementary food <6 mo 699 (74.5%) 670 (73.4%) 720 (76.7%) 846 (64.5%) 6 mo 239 (25.5%) 243 (26.6%) 219 (23.3%) 465 (36.5%) Child weight and height Weight, mean (SD) 12.5 (0.06) 12.8 (0.06) 12.9 (0.06) 13.3 (0.05) Height, mean (SD) 91.1 (0.21) 91.8 (0.21) 91.8 (0.18) 92.8 (0.16) BMIa, mean (SD) 15.1 (0.04) 15.2 (0.05) 15.2 (0.04) 15.4 (0.04) PARENTAL AND HOUSEHOLD LEVEL FACTORS Mother's age <30 years 467 (49.8%) 368 (40.3%) 339 (36.1%) 624 (47.6%) 30 years 471 (50.2%) 545 (59.7%) 600 (63.9%) 687 (52.4%) Father's age <30 years 217 (23.1%) 156 (17.2%) 145 (15.4%) 181 (13.8%) 30 years 721 (76.9%) 757 (82.8%) 794 (84.6%) 1130 (86.2%) Parents' marital status Currently married 938 (100%) 913 (100%) 939 (100%) 1303 (99.4%) Formerly married 0 (0%) 0 (0%) 0 (0%) 8 (0.6%) Mother's BMIb Underweight 111 (11.8%) 95 (10.4%) 75 (8.0%) 121 (9.2%) Normal weight 671 (71.5%) 636 (69.7%) 606 (64.5%) 759 (57.9%) Overweight/obese 156 (16.7%) 182 (19.9%) 258 (27.5%) 431 (32.9%) Father's BMIb Underweight 109 (11.6%) 105 (11.5%) 113 (12.0%) 140 (10.7%) Normal weight 714 (76.1%) 715 (78.3%) 665 (70.8%) 905 (69.0%) Overweight/obese 115 (12.3%) 93 (10.2%) 161 (17.2%) 266 (20.3%) Mother's heightc Normal height 417 (44.5%) 417 (45.7%) 438 (46.7%) 714 (54.5%) Short stature 521 (55.5%) 496 (54.3%) 501 (53.3%) 597 (45.5%) Father's heightc (Continued)
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 117 6 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Table 1. (Continued)
Characteristics Wave 1 (n = 938) Wave 2 (n = 913) Wave 3 (n = 939) Wave 4 (n = 1311) Normal height 377 (40.2%) 376 (41.2%) 403 (42.9%) 545 (41.6%) Short stature 561 (59.8%) 537 (58.8%) 536 (57.1%) 766 (58.4%) Ever had check up during pregnancy Yes 821 (87.5%) 545 (59.7%) 653 (69.5%) 1247 (95.1%) No 117 (12.5%) 368 (40.3%) 286 (30.5%) 64 (4.9%) Mother's education No education 96 (10.2%) 58 (6.4%) 56 (6.0%) 38 (2.9%) Primary school 508 (54.2%) 516 (56.5%) 448 (47.7%) 395 (30.1%) Junior and high school 297 (31.7%) 285 (31.2%) 331 (35.3%) 577 (44.0%) University or more 37 (3.9%) 54 (5.9%) 104 (11.0%) 301 (23.0%) Father's education No education 70 (7.5%) 41 (4.5%) 39 (4.2%) 19 (1.5%) Primary school 462 (49.3%) 459 (50.3%) 411 (43.8%) 262 (20.0%) Middle school 337 (35.9%) 342 (37.5%) 357 (38.0%) 282 (21.5%) University or more 69 (7.3%) 71 (7.7%) 132 (14.0%) 748 (57.0%) Household's wealth index Poor 446 (47.6%) 801 (87.7%) 404 (43.0%) 582 (44.4%) Middle 141 (15.0%) 69 (7.6%) 192 (20.5%) 269 (20.5%) Rich 351 (37.4%) 43 (4.7%) 343 (36.5%) 460 (35.1%) COMMUNITY LEVEL FACTORS Housing area Urban 435 (46.4%) 413 (45.3%) 429 (45.7%) 696 (53.1%) Rural 503 (53.6%) 500 (54.7%) 510 (54.3%) 615 (46.9%) Region Sumatra 248 (26.4%) 209 (22.9%) 206 (21.9%) 322 (24.6%) Java 466 (49.7%) 502 (55.0%) 519 (55.3%) 647 (49.4%) Bali & Nusa Tenggara Barat 130 (13.9%) 115 (12.6%) 103 (11.0%) 204 (15.6%) Kalimantan & Sulawesi 94 (10.0%) 87 (9.5%) 111 (11.8%) 138 (10.4%)
SD, Standard Deviation aBased upon the 2006 WHO Child Growth Standards for children <5 years [17] bBased upon the WHO BMI International Classification cut-off points [28] cHeight below -2 Standard Deviation according to WHO Standard Growth Reference for School-aged Children and Adolescents [29] doi:10.1371/journal.pone.0154756.t001
Associated risk factors Table 3 shows the unadjusted and adjusted odds ratios of all identified potential risk factors for stunting, underweight, and at risk of and overweight/obese. The multivariate analysis yielded similar results for the two forms of under nutrition: stunting and underweight. In the final model, the factors related to the higher probability of being stunted and underweight were lower birth weight (<2.5 kg), being breastfed for 6 months or more, having a mother or father who was underweight or had short stature, and mothers who never attended formal education. The only difference was that the probability of being stunted was also higher when a child lived in a rural area. Children had a greater probability of being at risk, or overweight/obese when they were in the youngest age group (2–2 9 years), were male, or had a mother and father who were over- Á weight/obese or had fathers who had attended university.
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 118 7 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Table 2. Comparison of the prevalence of stunting, underweight and ‘at risk’ and overweight/obesity within different variables (%, 95% CIs) (n = 4,101).
Variables Height for Age Z score Weight for Age Z score <-2 BMI Z score >+1 (At risk of overweight and <-2 (Stunted) (Underweight) overweight/obese) PREVALENCE IN EACH WAVE Wave 1 (1993) 50.8 (47.5–53.9) a 34.5 (31.6–37.7) a 10.3 (8.5–12.5) a Wave 2 (1997) 48.6 (45.4–51.8) 34.6 (31.6–37.8) 10.6 (8.8–12.8) Wave 3 (2000) 44.8 (41.7–48.0) 27.1 (24.3–30.0) 11.7 (9.8–13.9) Wave 4 (2007) 36.7 (34.1–39.3) 21.4 (19.3–23.7) 16.5 (14.6–18.6) CHILD LEVEL FACTORS Age 2.0–2.9 48.0 (45.3–50.7) a 30.4 (28.0–33.0) 16.9 (14.9–19.0) a 3.0–3.9 42.5 (39.8–45.1) 26.5 (24.1–28.9) 12.5 (10.8–14.4) 4.0–4.9 43.0 (40.5–45.6) 29.1 (26.8–31.5) 9.1 (7.7–10.7) Sex Male 43.7 (41.6–45.8) 27.7 (25.8–29.7) 14.4 (12.9–15.9) a Female 45.2 (43.0–47.4) 29.6 (27.7–31.6) 10.9 (9.6–12.4) Birth weight (n = 2420) <2.5 kg 57.5 (49.9–64.8) a 38.9 (31.8–46.5) a 13.2 (8.8–19.2) a 2.5-<4.0 kg 38.5 (36.4–40.6) 24.4 (22.5–26.3) 13.4 (12.0–15.0) 4.0 kg 33.6 (28.1–39.5) 14.9 (11.1–19.7) 21.0 (16.5–26.4) Ever breastfed Yes 44.7 (42.9–46.4) 28.8 (27.3–30.4) a 13.5 (12.3–14.7) No 30.0 (17.9–45.7) 7.5 (2.4–20.8) 22.5 (12.1–37.9) Age of weaning <6 mo 24.0 (19.0–29.9) a 11.2 (7.7–15.9) a 17.2 (12.9–22.6) 6 mo 45.3 (43.6–47.3) 29.3 (27.5–31.1) 12.9 (11.6–14.2) Age starting complementary food <6 mo 43.9 (41.8–46.0) 28.3 (26.4–30.2) 13.6 (12.2–15.1) 6 mo 46.6 (43.4–49.8) 29.8 (26.9–32.8) 13.3 (11.3–15.7) PARENTAL AND HOUSEHOLD LEVEL FACTORS Mother's age <30 years 44.1 (41.8–46.4) 29.5 (27.4–31.6) 12.9 (11.4–14.5) 30 years 44.7 (42.7–46.7) 28.0 (26.2–29.9) 12.6 (11.3–13.9) Father's age <30 years 47.1 (43.2–51.0) 30.6 (27.1–34.3) 12.4 (10.5–14.6) 30 years 45.8 (44.0–47.6) 29.8 (28.1–31.5) 11.9 (10.7–13.1) Parents' marital status Currently married 44.4 (42.7–46.0) 28.8 (27.6–30.1) 12.7 (11.7–13.7) Formerly married 37.5 (12.5–71.5) 37.4 (12.5–71.5) 12.5 (1.7–53.7) Mother's BMI b Underweight 51.6 (46.4–56.1) a 41.5 (36.8–46.4) a 8.5 (6.1–11.6) a Normal 46.7 (44.8–48.6) 29.9 (28.2–31.7) 11.6 (10.4–12.8) Overweight/obese 35.9 (33.1–38.9) 20.3 (17.9–22.8) 17.2 (15.0–19.7) Father's BMI b Underweight 54.4 (49.5–59.3) a 41.8 (37.1–46.7) a 7.6 (5.3–10.6) a Normal 48.1 (46.1–49.0) 30.5 (28.7–32.3) 12.4 (11.2–13.7) Overweight/obese 30.5 (26.6–34.7) 18.5 (15.4–22.2) 13.2 (10.7–16.4) Mother's height c (Continued)
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 119 8 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Table 2. (Continued)
Variables Height for Age Z score Weight for Age Z score <-2 BMI Z score >+1 (At risk of overweight and <-2 (Stunted) (Underweight) overweight/obese) Normal height 32.9 (30.9–35.1) a 22.0 (20.2–23.9) a 11.9 (10.7–13.3) Short stature 55.2 (53.1–57.3) 34.9 (32.9–36.9) 13.4 (12.0–14.9) Father's height c Normal height 35.3 (32.9–37.9) a 24.4 (22.2–26.7) a 11.6 (10.1–13.4) Short stature 54.1 (51.9–56.3) 34.1 (32.1–36.2) 12.1 (10.7–13.6) Ever had check up during pregnancy Yes 43.2 (41.4–44.03) a 27.6 (26.0–28.2) a 13.7 (12.5–15.0) a No 47.3 (44.5–50.2) 31.2 (28.6–33.9) 10.2 (8.6–12.1) Mother's education No education 60.1 (53.9–66.0) a 42.7 (36.7–49.0) a 11.3 (7.9–15.8) a Primary school 51.9 (49.6–54.1) 34.8 (32.6–37.0) 11.9 (10.5–13.4) Middle school 37.4 (35.0–39.9) 23.0 (20.9–25.2) 12.3 (10.7–14.1) University or more 29.8 (26.0–34.0) 15.7 (12.8–19.2) 17.5 (14.4–21.1) Father's education No education 50.3 (42.8–57.8) a 40.2 (33.1–47.8) a 10.1 (6.3–15.6) a Primary school 55.8 (53.4–58.3) 35.3 (32.9–37.6) 12.5 (11.0–14.2) Junior and high school 38.9 (36.3–41.5) 26.4 (24.1–28.9) 10.6 (9.1–12.4) University or more 32.8 (30.0–35.8) 19.3 (17.0–21.9) 16.1 (14.0–18.5) Household's wealth index Poor 46.3 (44.3–48.4) a 29.3 (27.5–31.2) 12.6 (11.4–14.1) Middle 41.2 (37.4–44.2) 28.5 (25.0–32.2) 12.4 (10.1–15.3) Rich 42.3 (39.5–44.1) 27.5 (25.0–30.1) 12.9 (11.1–14.9) COMMUNITY LEVEL FACTORS Housing area Urban 34.9 (32.9–37.0) a 23.4 (21.6–25.3) a 13.0 (11.6–14.6) Rural 53.3 (51.2–55.4) 33.5 (31.5–35.5) 12.4 (11.0–13.8) Region Sumatra 49.3 (46.2–52.4) a 29.5 (26.8–32.5) a 11.5 (9.6–13.6) a Java 39.8 (37.3–41.8) 25.7 (23.9–27.6) 14.4 (13.0–15.9) Bali & Nusa Tenggara Barat 52.0 (47.8–56.1) 35.5 (31.6–39.6) 10.5 (8.2–13.4) Kalimantan & Sulawesi 46.5 (41.8–51.2) 32.6 (28.3–37.1) 9.5 (7.1–12.7) aP value <0.05 (significant difference between each group within each risk factor) bBased upon the WHO BMI International Classification cut-off points [28] cHeight below -2 Standard Deviation according to WHO Standard Growth Reference for School-aged Children and Adolescents [29] doi:10.1371/journal.pone.0154756.t002
Discussion Statement of principal findings Our study, a secondary analysis of a series of representative surveys in Indonesian children aged 2.0–4.9 years, showed that, between 1993 and 2007, there was a significant decrease in the prevalence of both stunting and underweight. Over the same period there was a significant increase in the prevalence of ‘at risk of overweight’ and overweight/obesity in this age group. The issues of underweight and stunting are points of focus for Indonesian health authorities. Strategies to tackle them have included a variety of nutrition programs[32–34] as well as major improvements in the delivery of health care services, hygiene, and sanitation, including access
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 120 9 / 17 The DBM in 2.0–4.9 Year Indonesian Children
Fig 1. Distribution of BMI z-score and its mean (vertical line) of 2.0–4.9 years children of the Indonesian Family Life Survey (IFLS) wave 1 (1993), wave 2 (1997), wave 3 (2000), and wave 4 (2007). IFLS1: dotted line; IFLS2: dashed line; IFLS3: solid line; IFLS4: thick solid line. doi:10.1371/journal.pone.0154756.g001
to clean water.[33] These different approaches help explain the decrease in the prevalence of both underweight (from 34.5% to 21.4%) and stunting (from 50.8% to 36.7%) in children aged 2.0–4.9 years and over the 14 year period of the survey. The exact reason behind the rising prevalence of at risk/ overweight/ obesity in young chil- dren in Indonesia is not known. However, de Onis et al suggested that the rising global preva- lence of overweight/obesity in early childhood resulted from changes in both nutrition and physical activity patterns.[22] For example, improvements in the economy have resulted in changes in active transport and incidental activity[35] as well as a shift in dietary intake. [33, 36] Like many other developing countries, education is an important issue for Indonesia. Across the four waves, the level of education improved for both sexes, with increasing numbers of mothers and fathers experiencing formal education. The near doubling of the percentage of university-graduate parents is seen in wave 2 to 3, and from wave 3 to 4, the percentage of mothers with university education more than doubled. In families with limited income, the culture in many Asian countries still influences parents to choose their boys over the girls to go to university, because they will become the breadwinner for the family. This is supported by the higher numbers of men, compared to women, who graduated from university in all four waves. The percentage of poor households in wave 2 was much higher than in the other waves (87.7% vs. 47.6% in wave 1, 43.0%, in wave 3, 44.4% in wave 4). This might be an effect of the Asian financial crisis, which happened in 1997 (around wave 2 of the survey, conducted from late 1997 to 1998). Indonesia was one of the countries most affected by the crisis. At that time, many families had a dramatic decrease in family income as the inflation rate was 80% in that year.[37] Our study indicates that the “double burden of malnutrition” is present in Indonesian chil- dren aged 2.0–4.9 years. We also identified that both stunting and underweight were associated with a lower birth weight, being breastfed for more than 6 months, indices of parental under nutrition, and lack of maternal formal education. In contrast, being “at risk of overweight/obe- sity” or overweight or obese was associated with being in the youngest age group (2.0–2.9 years), being male, parental over nutrition, and high paternal formal education.
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 121 10 / 17 LSOE|DI1.31junlpn.145 a 1 2016 11, May DOI:10.1371/journal.pone.0154756 | ONE PLOS Table 3. Unadjusted and adjusted odds ratios (OR) (95%CI) of potential risk factor for stunting, underweight and ‘at risk’ and overweight/obesity (n = 4,101).
Variables Height for Age Z score <-2 (Stunted) Weight for Age Z score <-2 (Underweight) BMI Z score >+1 (At risk of overweight and overweight/ obese)
Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted
OR CI P value OR CI P value OR CI P value OR CI P value OR CI P value OR CI P value
PREVALENCE IN EACH WAVE Wave 1 (1993) Ref Ref Ref Wave 2 (1997) 0.89 0.77–1.03 0.131 0.95 0.76–1.13 0.490 1.10 0.82–1.48 0.511 Wave 3 (2000) 0.64 0.56–0.79 <0.001 0.62 0.53–0.72 <0.001 1.17 0.87–1.56 0.288 Wave 4 (2007) 0.49 0.41–0.56 <0.001 0.45 0.37–0.55 <0.001 1.86 1.43–2.41 <0.001 CHILD LEVEL FACTORS Age 2–2.9 Ref Ref Ref Ref e 3–3.9 0.81 0.69–0.95 0.010 0.85 0.75–0.96 0.075 0.71 0.56–0.87 <0.001 0.59 0.46–0.75 <0.001 4–4.9 0.82 0.70–0.96 0.010 0.93 0.90–1.15 0.431 0.53 0.39–0.62 <0.001 0.42 0.33–0.55 <0.001 Sex Male Ref Ref Ref Ref Female 1.03 0.91–1.17 0.650 1.01 0.86–1.14 0.933 0.75 0.62–0.90 0.002 0.76 0.62–0.94 0.010 Birth weight <2.5 kg Ref Ref c Ref Ref d Ref 2.5-<4 kg 0.54 0.38–0.73 <0.001 0.62 0.39–0.98 0.040 0.56 0.40–0.79 <0.001 0.69 0.43–1.10 0.125 0.99 0.62–1.58 0.993 4 kg 0.43 0.29–0.66 <0.001 0.49 0.28–0.87 0.010 0.33 0.20–0.52 <0.001 0.39 0.21–0.74 0.004 1.47 0.85–2.54 0.158 Ever breastfed Yes Ref Ref Ref No 0.69 0.35–1.39 0.309 0.21 0.06–0.71 0.010 2.11 0.99–4.49 0.056 Age of weaning <6 mo Ref Ref Ref Ref Ref 6 mo 2.73 1.96–3.14 <0.001 3.16 1.91–5.23 <0.001 2.73 2.54–6.57 <0.001 3.89 1.92–7.87 <0.001 0.78 0.53–1.15 0.217 Age starting complementary food <6 mo Ref Ref Ref 6 mo 1.05 0.90–1.24 0.475 1.04 0.89–1.24 0.588 0.99 0.81–1.27 0.864 PARENTAL AND HOUSEHOLD LEVEL FACTORS Mother's age <30 years Ref Ref Ref 30 years 1.06 0.95–1.19 0.386 0.96 0.83–1.10 0.598 1.00 0.84–1.22 0.832 Father's age 2.0 in DBM The <30 years Ref Ref Ref 30 years 1.01 0.84–1.19 0.950 0.99 0.85–1.24 0.753 0.97 0.75–1.26 0.846 Parents' marital status Currently married Ref Ref Ref
– – – –
Formerly married 0.47 0.71 3.12 0.436 1.95 0.34 11.07 0.447 0.93 0.06 12.67 0.958 Children Indonesian Year 4.9 Mother's BMIa c d e PAGE | 122 PAGE Underweight Ref Ref Ref Ref Ref Ref Normal 0.79 0.64–0.97 0.020 0.71 0.48–1.00 0.050 0.59 0.48–0.74 <0.001 0.56 0.38–0.81 0.002 1.23 0.86–1.72 0.246 1.21 0.82–1.80 0.328 Overweight/obese 0.48 0.39–0.62 <0.001 0.55 0.36–0.83 0.004 0.35 0.27–0.45 <0.001 0.39 0.25–0.61 <0.001 1.85 1.23–2.67 0.001 1.88 1.24–2.87 0.003 Father's BMIIa Underweight Ref Ref c Ref Ref d Ref Ref e Normal 0.79 0.64–0.98 0.030 0.90 0.62–1.31 0.593 0.64 0.51–0.80 <0.001 0.87 0.58–1.29 0.496 1.66 1.05–2.62 0.005 1.65 1.12–2.42 0.010 1/17 / 11 Overweight/obese 0.33 0.25–0.44 <0.001 0.45 0.28–0.72 0.001 0.31 0.22–0.43 <0.001 0.53 0.32–0.90 0.020 1.72 1.18–2.51 0.020 1.49 0.09–2.38 0.092 (Continued) LSOE|DI1.31junlpn.145 a 1 2016 11, May DOI:10.1371/journal.pone.0154756 | ONE PLOS Table 3. (Continued)
Variables Height for Age Z score <-2 (Stunted) Weight for Age Z score <-2 (Underweight) BMI Z score >+1 (At risk of overweight and overweight/ obese)
Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted
OR CI P value OR CI P value OR CI P value OR CI P value OR CI P value OR CI P value
Mother's heightb Normal height Ref Ref Ref Ref Ref Short stature 2.46 2.16–2.80 <0.001 2.21 1.76–2.78 <0.001 1.77 1.54–2.04 <0.001 1.30 1.01–1.68 0.040 1.12 0.93–1.35 0.199 Father's heightIa Normal height Ref Ref Ref Ref Ref Short stature 2.17 1.88–2.51 <0.001 1.91 1.51–2.41 <0.001 1.63 1.39–1.91 <0.001 1.49 1.15–1.94 0.002 1.07 0.87–1.33 0.472 Ever had check up during pregnancy Yes Ref Ref Ref No 1.10 0.96–1.26 0.167 1.17 1.01–1.36 0.030 0.72 0.58–1.12 0.004 Mother's education No education Ref Ref c Ref Ref d Ref Primary school 0.77 0.58–1.00 0.050 0.92 0.49–1.75 0.822 0.69 0.52–0.90 0.008 0.76 0.40–1.44 0.411 0.92 0.62–1.37 0.688 Middle school 0.37 0.28–0.49 <0.001 0.47 0.24–0.91 0.020 0.34 0.26–0.46 <0.001 0.36 0.19–0.70 0.003 0.99 0.66–1.49 0.985 University or more 0.31 0.22–0.42 <0.001 0.59 0.29–1.19 0.144 0.23 0.16–0.33 <0.001 0.28 0.13–0.59 0.001 1.60 1.03–2.48 0.030 Father's education No education Ref Ref Ref Ref e Primary school 1.32 0.97–1.80 0.070 0.73 0.53–1.00 0.050 1.68 0.98–2.90 0.058 1.78 1.01–3.17 0.040 Junior and high school 0.60 0.44–0.82 0.002 0.45 0.32–0.62 <0.001 1.33 0.76–2.32 0.310 1.33 0.74–2.41 0.332 University or more 0.47 0.34–0.65 <0.001 0.31 0.22–0.44 <0.001 2.22 1.28–3.86 0.004 2.19 1.17–4.11 0.010 Household's wealth index Poor Ref Ref Ref Middle 0.82 0.68–0.95 0.040 0.98 0.81–1.21 0.966 0.94 0.72–1.23 0.683 Rich 0.79 0.69–0.92 0.003 0.91 0.78–1.07 0.295 1.01 0.80–1.23 1.000 COMMUNITY LEVEL FACTORS Housing area Urban Ref Ref Ref Ref Rural 2.23 1.95–2.55 <0.001 1.55 1.22–1.97 <0.001 1.77 1.52–2.05 <0.001 0.91 0.75–1.10 0.338 Region Sumatra Ref Ref Ref Java 0.69 0.53–0.90 0.007 0.75 0.54–1.04 0.089 1.27 1.06–1.50 0.049 h B n2.0 in DBM The Bali & Nusa Tenggara Barat 1.06 0.65–1.71 0.804 1.12 0.70–1.80 0.617 0.80 0.50–1.26 0.343 Kalimantan & Sulawesi 0.88 0.52–1.49 0.650 1.08 0.66–1.77 0.750 0.73 0.45–1.18 0.203
aBased upon the WHO BMI International Classification cut-off points [28] bHeight below -2 Standard Deviation according to WHO Standard Growth Reference for School-aged Children and Adolescents [29] – c P-value for the whole category: (Birth weight, P-value = 0.047); (Mother’s BMI, P-value = 0.017); (Father’s BMI, P-value <0.001); (Mother’s education, P-value <0.001) Children Indonesian Year 4.9
PAGE | 123 PAGE d P-value for the whole category: (Birth weight, P-value = 0.013); (Mother’s BMI, P-value <0.001); (Father’s BMI, P-value = 0.035); (Mother’s education, P-value <0.001) e P-value for the whole category: (Age, P-value <0.001); (Mother’s BMI, P-value <0.001); (Father’s BMI, P-value = 0.034); (Father’s education, P-value = 0.005) doi:10.1371/journal.pone.0154756.t003 2/17 / 12 The DBM in 2.0–4.9 Year Indonesian Children
Strengths and limitations This study is the first to elaborate on the temporal trends in, and potential risk factors for, the three major forms of malnutrition in Indonesian children aged 2.0–4.9 years: stunting, under- weight, and overweight/obesity. Another strength includes the use of the category of ‘at risk of overweight’ i.e. children with BMI-z-score >+1SD and +2.[22] In addition, analyses are based on data derived from a representative sample of the Indonesian population with the use of sam- pling weights in the analysis to reduce bias, and measurements were performed by trained profes- sionals. To handle missing values in the birth weight variable in our dataset, we performed multiple imputations and the results show no differences between the complete data and the 5 and 10 imputation data sets (S1 Table). One limitation is the use of repeated cross sectional sur- veys, which does not allow us to infer causality. In addition, the risk factor analyses were limited: we were not able to investigate all potential risk factors, such as parental occupation or health knowledge, due to insufficient data or the question not being addressed in the questionnaire.
Comparison with other studies Several studies using different data sources in Indonesia have yielded comparable prevalence results to our study, although all were performed in children aged 0–5 years and usually focused on one type of malnutrition.[15, 38–40] None has previously identified the presence of the double burden of malnutrition in this age group (2.0–4.9 years). We opted to use the age of 2.0 years as the lower age limit in this study to ensure the process of stunting was fully devel- oped in these children, making this study different from any previously published studies. However, both under nutrition (stunting or underweight) and concurrent overweight/obesity have been documented in populations of under-five children in several Asian and Latin Amer- ica countries such as Malaysia, Vietnam, China, Nepal, Ecuador, Mexico, Guatemala, and Colombia.[36, 41–46] Furthermore, studies of temporal trends in such countries as Mexico and Colombia show the same phenomenon of decreasing prevalence of stunting and an increasing prevalence in overweight/obesity.[43, 46] The prevalence of stunting in 2.0–4.9 Indonesian children decreased over the period of the four surveys, although it was still considered to be high, at 36.7% in the most recent wave of the study in 2007. A 2013 review showed wide-ranging prevalence rates of stunting in under five children, such as 23.3% in Vietnam, 32.3% in the Philippines, 35.1% in Myanmar, 44.2% in Ethiopia, and 59.3 in Afghanistan.[47] A Malaysian study showed a 14.2% prevalence of stunt- ing in children aged 1.0–3.9 years in 2013.[44] The prevalence of underweight in this age group in Indonesia also decreased over the period of the survey, to 21.4% at wave 4. A recent review of the prevalence of moderate to severe under- weight in children under five in Asian countries showed marked variations in prevalence, ranging from 8.0% in China, 11.0% in Malaysia, 18.0% in Thailand, 28.0% in the Philippines, to 47.0% in India and 48.0% in both Nepal and Bangladesh.[36] The prevalence of underweight in under-five children in Mexico, Guatemala, and Colombia are 2.8%, 1.0%, and 3.4%, respectively.[43, 45, 46] The prevalence of at risk/overweight/obesity in wave 4 was 16.5%, lower than in Vietnam 2005, where 36.8% of 4.0–5.0 year old children were overweight/obese.[41] The prevalence of at risk/ overweight/ obesity in under five children in Ecuador in 2012 was 30.2%[42] and in under five children in Mexico and Guatemala was 9.0% and 4.9%, respectively.[43] However, prevalence rates are age specific, and the definition/cut offs used for underweight and overweight/obesity in different studies might be different, highlighting the need for caution in making direct comparisons between studies. Our findings are consistent with other research showing strong associations between stunt- ing and a lower birth weight,[38, 48, 49] a longer duration of breastfeeding,[49] short-statured
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 124 13 / 17 The DBM in 2.0–4.9 Year Indonesian Children
mothers,[11, 38, 48] underweight mothers,[48] less educated mothers,[11, 38, 48–50] and liv- ing in rural areas.[38] Our study also emphasised the association between underweight chil- dren and mothers’ and fathers’ education levels.[40] Likewise, we found an association between overweight/obesity and male sex,[11] and maternal overweight/obesity.[11, 48]
Implications for research, policy and practice Our study highlights the emerging issue of the double burden of malnutrition in young Indo- nesian children. Many interventions and strategies that are already in place for the manage- ment of under nutrition in Indonesia, e.g. the Healthy and Fit due to Balance Nutrition (Sehat dan Bugar berkat Gizi Seimbang) or the Scaling Up Nutrition movement,[32, 34] may need to be modified to respond to the problem by balancing the risk factors associated to each condi- tion. For example, interventions that aim to correct under nutrition in early life need to empha- sise the importance of both linear growth and appropriate weight.[33, 51–53] Fortification of complementary foods needs to be balanced so as not to make previously ‘healthy weight’ chil- dren become overweight/obese. Further interventions in school age children need to balance the promotion of healthy diets as well as physical activity. Another important point is to ensure that interventions start as early as possible. For exam- ple, improvements in diets should start with adolescent girls and young women in their pre- pregnancy state, in order to prevent having underweight mothers, which is in turn a risk factor for having stunted and/ underweight children. The fact that breastfeeding decreases the preva- lence of obesity in later life is irrefutable;[54] however, prolonged breastfeeding, in association with poor feeding practices may be associated with stunting.[49, 55] Therefore, parental educa- tion, especially of mothers, regarding the importance of breastfeeding combined with healthy feeding practices is important. Currently, these different conditions of malnutrition are treated as separate issues.[33] There should be a policy that combines the management of concurrent under and over nutri- tion. Future studies should also aim at exploring whether children who are stunted in early life are more likely to be overweight/obese in later life.
Supporting Information S1 Table. Adjusted odds ratios (95% confidence intervals) for the complete data, 5 and 10 imputation data sets (M = 5 and M = 10, respectively). (DOCX)
Acknowledgments We would like to thank Christine Peterson from Rand Corporation for her assistance regarding the IFLS datasets and ethics information.
Author Contributions Conceived and designed the experiments: CNR KA ML LAB. Performed the experiments: CNR KA ML LAB. Analyzed the data: CNR KA ML LAB. Contributed reagents/materials/anal- ysis tools: CNR KA ML LAB. Wrote the paper: CNR KA ML LAB.
References 1. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014; 384(9945):766–81. doi: 10.1016/S0140- 6736(14)60460-8 PMID: 24880830
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 125 14 / 17 The DBM in 2.0–4.9 Year Indonesian Children
2. World Health Organization. Report of the First Meeting of the Ad hoc Working Group on Science and Evidence for Ending Childhood Obesity. Geneva, Switzerland: World Health Organization, 2014. 3. Popkin BM, Adair LS, Ng SW. Now and then: The Global Nutrition Transition: the pandemic of obesity in developing countries. Nutrition Reviews. 2012; 70(1):3–21. doi: 10.1111/j.1753-4887.2011.00456.x PMID: 22221213 4. Stanojevic S, Kain J, Uauy R. Secular and seasonal tends in obesity in Chilean preschool children, 1996–2004. Journal of Pediatric Gastroenterology and Nutrition. 2008; 47:339–43. doi: 10.1097/MPG. 0b013e31816eb448 PMID: 18728531 5. Doak CM, Adair LS, Bentley M, Monteiro C, Popkin BM. The dual burden household and the nutrition transition paradox. Int J Obes (Lond). 2005; 29(1):129–36. 6. Roemling C, Qaim M. Dual burden households and intra-household nutritional inequality in Indonesia. Economics and Human Biology. 2013; 11(4):563–73. doi: 10.1016/j.ehb.2013.07.001 PMID: 23890671 7. Oddo VM, Rah JH, Semba RD, Sun K, Akhter N, Sari M, et al. Predictors of maternal and child double burden of malnutrition in rural Indonesia and Bangladesh. Am J Clin Nutr. 2012; 95(4):951–8. doi: 10. 3945/ajcn.111.026070 PMID: 22357721 8. Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, de Onis M, et al. Maternal and child undernu- trition and oveweight in low-income and middle-income countries. Lancet. 2013; 382.9890:427–51. doi: 10.1016/S0140-6736(13)60937-X PMID: 23746772 9. Bove I, Miranda T, Campoy C, Uauy R, Napol M. Stunting, overweight and child development impairment go hand in hand as key problems of early infancy: Uruguayan case. Early Human Develop- ment. 2012; 88:747–51. doi: 10.1016/j.earlhumdev.2012.04.002 PMID: 22560815 10. Albala C, Vio F, Kain J, Uauy R. Nutrition transition in Latin America: the case of Chile. Nutrition Reviews. 2001; 59(6):170–76. PMID: 11444593 11. Fernald L, Neufeld L. Overweight with concurrent stunting in very young children from rural Mexico: prevalence and associated factors. European Journal of Clinical Nutrition. 2007; 61:623–32. PMID: 17136036 12. Sawaya A, Grillo L, Verreschi I, da Silva A, Roberts S. Mild stunting is associated with higher suscepti- bility to the effects of high fat diets: studies in a shantytown population in São Paulo, Brazil. The Journal of Nutrition. 1998; 128.2 supplement:415–20. 13. Monteiro C, Conde W, Popkin BM. is obesity replacing or adding to undernutrition? Evidence from dif- ferent social classes in Brazil. Public Health Nutrition. 2002; 5(1A):105–12. PMID: 12027272 14. Doak CM, Adair LS, Monteiro C, Popkin BM. Overweight and underweight coexist within households in Brazil, China and Russia. The Journal of Nutrition. 2000; 130:2965–71. PMID: 11110855 15. Badan Penelitian dan Pengembangan Kesehatan KKRI. Laporan Hasil Riset Kesehatan Dasar (Ris- kesdas) Indonesia tahun 2013. In: Indonesia DK, editor. Jakarta, Indonesia: CV Kiat Nusa; 2014. 16. Victora CG, de Onis M, Hallal PC, Blossner M, Shrimpton R. Worldwide timing of growth faltering: revis- iting implications for interventions. Pediatrics. 2010; 125(3):e473–80. doi: 10.1542/peds.2009-1519 PMID: 20156903 17. World Health Organization. The WHO Child Growth Standards: World Health Organization; 2006 [cited 2014 16 March]. Available from: http://www.who.int/childgrowth/standards/en/. 18. Corporation R. The Indonesian Family Life Survey (IFLS) Santa Monica, Califonia: Rand Corporation; 2014 [cited 2014 22 February]. Available from: http://www.rand.org/labor/FLS/IFLS.html. 19. Frankenberg E, Karoly LA, Gertler P, Peterson CE, Wesley D. The 1993 Indonesian Family Life Survey: Overview and Field Report. Santa Monica, California: RAND, 1995. 20. Serrato C, Melnick G. The Indonesian Family Life Survey Overview and Descriptive Analysis. Santa Monica, California: RAND, 1995 Contract No.: DRU-1191-AID. 21. LMS Growth Program 2010 [Internet]. 2010 [cited June 2014]. Available from: http://www. healthforallchildren.com/?product=lmsgrowth. 22. de Onis M, Blossner M, Borghi E. Global prevalence and trends of overweight and obesity among pre- school children. Am J Clin Nutr. 2010; 92:1257–64. doi: 10.3945/ajcn.2010.29786 PMID: 20861173 23. World Health Organization. WHO Anthro version 3.2.2 and macros: World Health Organization; 2011 [cited 2014 15 September]. Available from: http://www.who.int/childgrowth/software/en/. 24. Davison KK, Birch LL. Childhood overweight: a contextual model and recommendations for future research. Obesity reviews: an official journal of the International Association for the Study of Obesity. 2001; 2(3):159–71. 25. Filmer D, Pritchett LH. Estimating wealth effects without expenditure data—or tears: an application to educational enrollments in states of India. Demography. 2001; 38(1):115–32. PMID: 11227840
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 126 15 / 17 The DBM in 2.0–4.9 Year Indonesian Children
26. Unicef. Childhood under threat: the state of the world's children 2005. In Childhood under threat: the state of the world's children 2005. New York, USA: Unicef, 2004 Contract No.: ISBN 92-806-3817-3. 27. World Health Organization. The World Health Report: 2005: make every mother and child count. Geneva Switzerland World Health Organization, 2005 Contract No.: ISBN 92 4 156290 0 28. World Health Organization. World Health Organization International Classification of adult underweight, overweight, and obesity Geneva: WHO; 2014 [cited 2015 March 2015]. Available from: http://www. who.int/bmi. 29. World Health Organization. WHO Growth Reference for School-aged Children and Adolescents: WHO; 2007 [cited 2015 March 2015]. Available from: http://www.who.int/growthref/en/. 30. Corp S. STATA statistical software: Release 13. College station, TX: Stata Corp LP; 2014. 31. Rabe-Hesketh S, Skrondal A. Multilevel and longitudinal modelling using Stata. Third ed. Texas: Stata Press; 2012. 32. Nutrition SU. Scaling Up Nutrition Movement: Indonesia [cited 2014 November]. Available from: http:// scalingupnutrition.org/sun-countries/indonesia. 33. Shrimpton R, Rokx C. The Double Burden of Malnutrition in Indonesia. Jakarta, Indonesia: World Bank Jakarta, 2013 Contract No.: Report 76192-ID. 34. Soekirman, Hadi H, Thaha R, H H, I Ja, E A, et al. Sehat dan Bugar berkat Gizi Seimbang. Jakarta: Gramedia-Kompas Publishing; 2011. 35. Shrimpton R, Rokx C. The double burden of malnutrition: a review of global evidence. Washington DC: World Bank, 2012 79525. 36. Khor GL. Food-based approaches to combat the double burden among the poor: challenges in the Asian context. Asia Pac J Clin Nutr. 2008; 17 Suppl 1:111–5. PMID: 18296315 37. Nasution A. The meltdown of the Indonesian economy in 1997–1998: Causes and responses. Seoul Journal of Economics. 1998; 11(4):447. 38. Semba RD, de Pee S, Sun K, Sari M, Akhter N, Bloem MW. Effect of parental formal education on risk of child stunting in Indonesia and Bangladesh: a cross-sectional study. Lancet. 2008; 371(9609):322– 8. doi: 10.1016/S0140-6736(08)60169-5 PMID: 18294999 39. Bardosono S, Sastroamidjojo S, Lukito W. Determinants of child malnutrition during the 1999 economic crisis in selected poor areas of Indonesia. Asia Pac J Clin Nutr. 2007; 16(3):512–26. PMID: 17704034 40. Waters H, Saadah F, Surbakti S, Heywood P. Weight-for-age malnutrition in Indonesian children, 1992–1999. International journal of epidemiology. 2004; 33(3):589–95. PMID: 15155707 41. Dieu HT, Dibley MJ, Sibbritt D, Hanh TT. Trends in overweight and obesity in pre-school children in urban areas of Ho Chi Minh City, Vietnam, from 2002 to 2005. Public Health Nutrition. 2008; 12(5):702– 9. doi: 10.1017/S1368980008003017 PMID: 18664311 42. Freire WB, Silva-Jaramillo KM, Ramirez-Luzuriaga MJ, Belmont P, Waters WF. The double burden of undernutrition and excess body weight in Ecuador. Am J Clin Nutr. 2014; 100(6):1636S–43S. doi: 10. 3945/ajcn.114.083766 PMID: 25411306 43. Kroker-Lobos MF, Pedroza-Tobias A, Pedraza LS, Rivera JA. The double burden of undernutrition and excess body weight in Mexico. Am J Clin Nutr. 2014; 100(6):1652S–8S. doi: 10.3945/ajcn.114.083832 PMID: 25411308 44. Poh BK, Ng BK, Siti Haslinda MD, Nik Shanita S, Wong JE, Budin SB, et al. Nutritional status and die- tary intakes of children aged 6 months to 12 years: findings of the Nutrition Survey of Malaysian Chil- dren (SEANUTS Malaysia). The British journal of nutrition. 2013; 110 Suppl 3:S21–35. doi: 10.1017/ S0007114513002092 PMID: 24016764 45. Ramirez-Zea M, Kroker-Lobos MF, Close-Fernandez R, Kanter R. The double burden of malnutrition in indigenous and nonindigenous Guatemalan populations. Am J Clin Nutr. 2014; 100(6):1644S–51S. doi: 10.3945/ajcn.114.083857 PMID: 25411307 46. Sarmiento OL, Parra DC, Gonzalez SA, Gonzalez-Casanova I, Forero AY, Garcia J. The dual burden of malnutrition in Colombia. Am J Clin Nutr. 2014; 100(6):1628S–35S. doi: 10.3945/ajcn.114.083816 PMID: 25411305 47. de Onis M, Dewey K, Borghi E, Onyango A, Blössner M, Daelmans B, et al. The World Health Organiza- tion's global target for reducing childhood stunting by 2025: rationale and proposed actions. Maternal and Child Nutrition. 2013; 9(supplement 2):6–26. 48. Bove I, Miranda T, Campoy C, Uauy R, Napol M. Stunting, overweight and child development impairment go hand in hand as key problems of early infancy: Uruguayan case. Early Hum Dev. 2012; 88(9):747–51. doi: 10.1016/j.earlhumdev.2012.04.002 PMID: 22560815
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 127 16 / 17 The DBM in 2.0–4.9 Year Indonesian Children
49. Tiwari R, Ausman LM, Agho KE. Determinants of stunting and severe stunting among under-fives: evi- dence from the 2011 Nepal Demographic and Health Survey. BMC Pediatr. 2014; 14:239. doi: 10. 1186/1471-2431-14-239 PMID: 25262003 50. Monteiro CA, Benicio MH, Conde WL, Konno S, Lovadino AL, Barros AJ, et al. Narrowing socioeco- nomic inequality in child stunting: the Brazilian experience, 1974–2007. Bull World Health Organ. 2010; 88(4):305–11. doi: 10.2471/BLT.09.069195 PMID: 20431795 51. Uauy R, Kain J, Mericq V, Rojas J, Corvalan C. Nutrition, child growth, and chronic disease prevention. Annals of Medicine. 2008; 40:11–20. doi: 10.1080/07853890701704683 PMID: 18246473 52. World Health Organization. WHO Global Nutrition Target: Stunting Policy Brief2014. Available from: http://www.who.int/nutrition/topics/globaltargets_stunting_policybrief.pdf. 53. Bloem MW, de Pee S, Hop le T, Khan NC, Laillou A, Minarto, et al. Key strategies to further reduce stunting in Southeast Asia: lessons from the ASEAN countries workshop. Food Nutr Bull. 2013; 34(2 Suppl):S8–16. PMID: 24049992 54. Long-term effects of breastfeeding: a systematic review [Internet]. World Health Organization. 2013. Available from: http://apps.who.int/iris/bitstream/10665/79198/1/9789241505307_eng.pdf. 55. Marquis GS, Habicht JP, Lanata CF, Black RE, Rasmussen KM. Association of breastfeeding and stunting in Peruvian toddlers: an example of reverse causality. International journal of epidemiology. 1997; 26(2):349–56. PMID: 9169170
PLOS ONE | DOI:10.1371/journal.pone.0154756 May 11, 2016 PAGE | 128 17 / 17
Supplementary Table 1. Adjusted odds ratios (95% confidence intervals) for the complete data, 5 and 10 imputation data sets (M=5 and M=10) Variables Complete data M=5 M=10 AOR [95% CI] P>z AOR [95% CI] P>t AOR [95% CI] P>t Housing area 1.62 1.28 2.06 0.000 1.71 1.38 2.12 0.000 1.70 1.38 2.11 0.000 Mother’s BMI 0.77 0.64 0.94 0.010 0.77 0.63 0.92 0.005 0.77 0.63 0.92 0.005 Fathers’s BMI 0.65 0.51 0.81 0.000 0.62 0.50 0.78 0.000 0.62 0.50 0.78 0.000 Mother’s height 2.23 1.78 2.80 0.000 2.14 1.73 2.64 0.000 2.14 1.73 2.64 0.000 Father’s height 1.95 1.54 2.45 0.000 1.89 1.53 2.34 0.000 1.89 1.53 2.35 0.000 Mother’s education 0.71 0.60 0.83 0.000 0.71 0.61 0.82 0.000 0.71 0.61 0.82 0.000 Birth weight 0.72 0.55 0.96 0.023 0.79 0.60 1.03 0.082 0.81 0.62 1.06 0.117 Age of weaning 3.07 1.87 5.05 0.000 2.69 1.73 4.19 0.000 2.69 1.72 4.18 0.000
PAGE| 129
Chapter 5: Stunting coexisting with overweight in 2.0-4.9-year-old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys
INTRODUCTION 131 METHODS Indonesian Family Life Survey 132 Data collection 132 Inclusion criteria 132 Ethics 132 Anthropometric indices calculations 132 Potentially associated risk factors 132 Child-level factors 132 Parental- and household- level factors 132 Community-level factors 133 Statistical analysis 133 RESULTS 133 Characteristics of participants 133 Prevalence of and risk factors for concurrent stunting and overweight 133 Odds of stunted children being overweight 135 DISCUSSION 135 CONCLUSIONS 138 REFERENCES 139
This chapter has been previously published as: Rachmi CN, Agho KE, Li M, Baur LA. Stunting coexisting with overweight in 2.0-4.9 year old Indonesian children: prevalence, trends and associated risk factors from four cross sectional surveys. Public Health Nutrition. 2016 Oct;19(15):2698-707.
PAGE| 130 Public Health Nutrition: page 1 of 10 doi:10.1017/S1368980016000926
Stunting coexisting with overweight in 2·0–4·9-year-old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys
Cut Novianti Rachmi1,*, Kingsley Emwinyore Agho2, Mu Li3 and Louise Alison Baur1,3 1Discipline of Paediatrics and Child Health, The Children’s Hospital at Westmead (University of Sydney Clinical School), Locked Bag 4001, Westmead, NSW 2145, Australia: 2School of Science and Health, University of Western Sydney, Penrith, NSW, Australia: 3Sydney School of Public Health, The University of Sydney, Sydney, NSW, Australia
Submitted 22 September 2015: Final revision received 20 March 2016: Accepted 30 March 2016
Abstract Objective: The persistence of undernutrition, along with overweight and obesity, constitute the double burden of malnutrition. The present study aimed to: (i) describe the prevalence and trends of concurrent stunting and overweight in Indonesian children; (ii) identify potentially associated risk factors; and (iii) determine whether stunted children are at greater risk of overweight compared with those of healthy height. Design: A secondary data analysis of children aged 2·0–4·9 years in four cross-sectional studies of the Indonesian Family Life Survey. Children’s height and BMI Z-scores were calculated based on the WHO Child Growth Standards (2006). We defined ‘concurrent stunting and overweight’ as height-for-age Z-score <−2 and BMI Z-score >+1. Multivariate generalised linear latent and mixed models were used to determine associated risk factors. Setting: Thirteen out of twenty-seven provinces in Indonesia. Subjects: Children (n 4101) from four waves of the Indonesian Family Life Survey (1993–2007). Results: There were inconsistent trends in the prevalence of concurrent stunting and overweight from waves 1 to 4. Children were more likely to be stunted and
Public Health Nutrition overweight when they were in the youngest age group (2·0–2·9 years), were weaned after the age of 6 months, had short-statured mothers or lived in rural areas. Stunted children were significantly more likely to be overweight than fi Keywords healthy-height children (OR > 1) but did not differ signi cantly different across Stunting each wave (OR = 1·34–2·01). Overweight/obesity Conclusions: Concurrent stunting and overweight occurs in Indonesian children Indonesia aged 2·0–4·9 years. Current policies and programmes need to be tailored for the Children management of this phenomenon. Double burden
The ‘double burden of malnutrition’–the persistence of and Mexico – have reported the phenomenon in children undernutrition, along with a rapid increase in overweight aged less than 5 years(4,13–16). However, to our and obesity – is now recognised as ‘the new normal’(1,2). knowledge, no study has reported on concurrent stunting Many countries face the double burden of malnutrition, and overweight prevalence, or the associated risk factors, although it is a more common phenomenon in countries in any paediatric age group in South-East Asia. A better where stunting rates are high(1). Several studies and understanding of the risk factors related to concurrent reviews have also shown that this phenomenon can be stunting and overweight would improve prevention and found at the level of both the family (mother and child management approaches aimed at overcoming this double burden)(3–8) and the individual(4,9,10). problem. The co-occurrence of stunting and overweight has been The present paper is the second of a series of secondary described in children from such countries as Mexico, data analyses on the double burden of malnutrition in China, Russia, South Africa, Brazil and the USA(9–12). A few Indonesia. In our first paper we showed that, over a studies – from Uruguay, Ecuador, Guatemala, South Africa 14-year time frame (1993 to 2007), the prevalence of
*Corresponding author: Email [email protected]; [email protected] © The Authors 2016 PAGE | 131 2 CN Rachmi et al. stunting in Indonesian children aged 2·0–4·9 years at 2 years because the process of stunting is more decreased by 14·1 % from 50·8 % to 36·7 %, while the prominent before the age of 2(21). prevalence of overweight increased by 6·2 % from 10·3% to 16·5 % (all P < 0·01). We also identified that associated Ethics risk factors for a higher probability of being stunted or Ethics approval was granted from the Institutional Review underweight included lower birth weight (<2·5 kg), being Board at Rand Corporation (USA) and from the Ethics breast-fed for 6 months or more, having a mother or father Committee at Universitas Indonesia (Indonesia) for the who was underweight or short-statured, and mothers with first wave and the Ethics Committee at Universitas Gadjah no formal education. The likelihood of being stunted Mada (Indonesia) for the next three waves. was also higher when a child lived in a rural area (all P < 0·05). Children were more likely to be at risk or Anthropometric indices calculations overweight/obese if they were in the youngest age group BMI was calculated as weight/height2 (kg/m2). Using Pan (2·0–2·9 years), male, had parents who were overweight/ and Cole’s LMS Growth Program(22), the children’s height obese and fathers with high formal education (university and BMI Z-scores were calculated based upon the WHO (17) or more; all P < 0·05) . Child Growth Standards (2006)(23). Stunting was defined as Herein we elaborate further on the co-occurrence of height-for-age Z-score <−2. Children with BMI Z-score stunting and at risk of or overweight/obesity in the same >+1, >+2 and >+3 were categorised as being at risk of individual – we refer to this as ‘concurrent stunting and overweight, overweight and obese, respectively(23,24). For overweight’. The aims of the current paper were to: the purposes of the present study, we defined ‘concurrent (i) describe the prevalence and trends of concurrent stunting stunting and overweight’ as children with the combination and overweight in young Indonesian children between 1993 of height-for-age Z-score <−2 and BMI Z-score >+1. and 2007; (ii) identify potential risk factors associated with the phenomenon; and (iii) determine whether stunted children are at greater risk of being overweight or obese Potentially associated risk factors compared with their healthy-height peers. The conceptual framework for the current analysis was modified from the ecological model of childhood obesity of Davison and Birch(25) to include not only the available Methods variables in the IFLS data set, but also stunting as a form of malnutrition. The potential risk factors were divided into Indonesian Family Life Survey three categories: child-, parental- and household-, and Data collection community-level factors. Data were from the first four waves of the Indonesian Family Life Survey (IFLS) in the years 1993, 1997, 2000 and Child-level factors (18) Public Health Nutrition 2007 . Details of the IFLS have been described in our These consisted of the child’s age, sex, birth weight first paper and several previously published field (for whom this had been recorded), whether they were fi reports(17,19,20). In brief, IFLS is a longitudinal, nationally ever breast-fed, age of weaning (de ned as full cessation representative survey of a stratified random sample of of breast-feeding), age of starting complementary foods, households involving both questionnaires and anthropo- and their current weight and height. Age was divided into metric measurements. The first wave (1993) recruited three groups: 2·0–2·9, 3·0–3·9 and 4·0–4·9 years. Birth participants from thirteen of the twenty-seven Indonesian weight was categorised as low birth weight (<2·5 kg), provinces. The next three surveys followed the same healthy birth weight (2·5– < 4·0 kg) and high birth weight survey and measurement methods as the first one, and (≥4·0 kg). Both age of weaning and age of starting had a very high re-contact rate (>90 %). Trained complementary foods were divided into two groups: professionals collected the data. The original design of the <6 months and ≥6 months. survey was longitudinal, targeting the same families and the children of the families from the first wave of the Parental- and household-level factors survey(18–20). In the present paper we use cross-sectional Parental-level factors included maternal and paternal study design. We analysed data from children aged factors. Maternal factors included mothers’ age, BMI, 2·0–4·9 years in each wave, resulting in different groups of height and maternal history of check-up during preg- children in each of the four waves. nancy. Maternal age was categorised as <30 or ≥30 years and maternal BMI was categorised based on the WHO Inclusion criteria BMI International Classification cut-off points of ≥25 and Inclusion criteria were children aged 2·0–4·9 years who ≥30 kg/m2 for overweight and obesity, respectively(26). had complete records for child information (height, Because of the lack of consistent definitions of stunting for weight, age and sex) and matching parental-, household- men and women in the literature, for the purposes of the and community-level data. The minimum age was chosen current analysis we categorised height as short stature
PAGE | 132 Stunted and overweight Indonesian children 3 (height-for-age Z-score <−2) or healthy height (height-for- avoid statistical error in our analyses, we repeated the age Z-score ≥− 2), based upon a standard age of 19 years backward elimination process using a different approach. and the WHO Standard Growth Reference for School- First, only variables among community-, child-, parental- Aged Children and Adolescents(27). Maternal history of and household-level variables with P < 0·20 identified in check-up during pregnancy was categorised as ever or the univariate analysis were entered for the backward never had check-up (yes/no variable). Paternal factors elimination process. Second, we double-checked the included fathers’ age, BMI and height using the same backward elimination by including all community, child, cut-off points as mothers, and parental marital status. parental and household variables, and only the variables The household-level factors included mothers’ and with P < 0·05 were retained in the final model (i.e. child’s fathers’ education (divided into four groups: never age group, age of weaning, maternal height and housing attended any formal education, attended primary school, area). Third, we tested and reported any collinearity in the middle school, and university or higher) and the house- final model. The odd ratios and 95 % confidence intervals hold’s wealth index, assessed by calculation of a score were calculated for each variable and were used to mea- involving the ownership of eleven household assets sure the impact of the adjusted estimates on the study by using weights. We ranked the households into five outcome. The significant Bonferroni-adjusted P values are quintiles: poorest, poorer, average, richer and richest. For reported. the analysis, households in the bottom two quintiles were The odds ratio of becoming overweight for those who categorised as poor, those in the middle two quintiles as were stunted was calculated by dividing the probability of average, and those in the highest quintile as rich being overweight in stunted children by the probability of households(28). being overweight in the healthy-height children.
Community-level factors Results The community factors included the housing area (rural and urban) and region. Four regions were included in the Characteristics of participants study: Sumatra, Java, Bali and Nusa Tenggara Barat, and The sociodemographic characteristics of the participants and Kalimantan and Sulawesi. their parents are shown in Table 1. There were a total of 4101 children aged 2·0–4·9 years in all four waves, with a Statistical analysis similar percentage of children in each age band and sex. In the IFLS, each household completed several separate In all four waves most children were born in the healthy questionnaires, each with different types of information weight range (2·5–4·0kg) and were breast-fed until (e.g. anthropometry, household economy, child informa- 6 months. Throughout all four waves, a little more than half tion, adult information). These different files were merged of the mothers were aged ≥ 30 years or were classified as in order to build the data set for analysis. We used having short stature, except in wave 4 where 54·5% of
Public Health Nutrition sampling weights in the analysis to reduce bias; however, mothers were of healthy height. As many as 83 % of all we did not adjust sampling weight for children aged fathers were aged ≥30 years during the data collection and 2·0–4·9 years because sub-samples are mutually exclusive. just over half (54 %) of fathers were of short stature. The Frequency tabulations were first conducted to describe the prevalence of underweight in both mothers and fathers distributions of data used in the study, followed by remained relatively constant throughout the four waves, prevalence estimates using the Taylor-series linearisation while the prevalence of overweight in both mothers and method to examine the impact of all potential predictors fathers increased over time. At the household level, there using χ2 tests and multiple testing with the Bonferroni were more educated fathers than mothers. From years 1993 correction was carried out by dividing the 5 % significance to 2000, there were more people living in rural areas, but by level by the number of χ2 tests performed. year 2007 more families were living in urban areas. The unadjusted odd ratios for factors associated with stunting and overweight were examined using GLLAMM Prevalence of and risk factors for concurrent (generalised linear latent and mixed models)(29). This was stunting and overweight followed by multivariable analyses after controlling for Table 2 shows the prevalence of concurrent stunting and community-, child-, parental- and household-level factors. overweight as well as the associated risk factors. The All statistical analyses were conducted using the statistical prevalence indicates that children aged 2·0–2·9 years were software package STATA/MP version 13.1 (2014) and significantly more likely to be stunted and overweight than multilevel models were fitted using STATA commands to those children aged 4·0–4·9 years. Children whose fathers adjust for the variability of clustering. had a healthy BMI, whose mothers and fathers were of In the multivariable analysis models, a manual stepwise short stature, whose mothers had a check-up during backward elimination process was used to identify factors pregnancy, who were breast-fed for ≥6 months or who that were significantly associated with the study outcome lived in rural areas had a higher prevalence of concurrent using a 5 % significance level. In order to minimise or stunting and overweight.
PAGE | 133 Public Health Nutrition Characteristic hl level Child Survey Life Family 1 Table 4 aetllevel Parental oshl level Household Age Maternal vrbreast-fed Ever Parents g fweaning of Age Mother Paternal g fsatn opeetr foods complementary starting of Age Sex hl egtadheight and weight Child it egt( weight Birth Father 2 Mother uiradhg col3735 337 school high and Junior M (kg/m BMI and mean (cm), Height e 3 0 5 93 850 100 938 Yes urnl are 3 0 1 0 3 0 3399 1303 100 939 100 913 100 938 married Currently 3 hc pdrn pregnancy during up Check o006 6 63 0 0 No omrymrid00000080 8 0 0 0 0 10 96 0 0 education No married Formerly nvriyo oe6 7 69 more or University Father Mother 4 rmr col5854 508 school Primary ae5153 501 Male Father egt(g,ma and mean (kg), Weight uiradhg col2731 297 school high and Junior eae4746 437 Female Mother nvriyo oe3 3 37 more or University 2 oeuain7 7 70 education No Father rmr col4249 462 school Primary ≥ < ≥ < ≥ < · · · · 0 0 0 5 4 ots4 4 41 months 6 ots8795 897 months 6 ots6974 699 months 6 ots2925 239 months 6 2 elh egt3740 377 height Healthy hr ttr 2 55 521 stature Short hr ttr 6 59 561 stature Short e 2 87 821 Yes o1712 117 No newih 1 11 111 Underweight elh egt6171 671 weight Healthy vregtoee1616 156 Overweight/obese newih 0 11 109 Underweight elh egt7476 714 weight Healthy elh egt4744 417 height Healthy vregtoee1512 115 Overweight/obese < ≥ < ≥ – – – – · · g6 11 68 kg 0 g4 7 46 kg 5 hrceitc fcide n aet nec ae(ae1 93 ae2 97 ae3 00 ae4 07 fteIndonesian the of 2007) 4, wave 2000; 3, wave 1997; 2, wave 1993; 1, (wave wave each in parents and children of Characteristics ’ 0yas4749 467 years 30 0yas4150 471 years 30 0yas2723 217 years 30 0yas7176 721 years 30 2 3 4 education s ’ < education s ’ · · · er 1 33 313 years 9 er 2 34 326 years 9 er 9 31 299 years 9 4 ’ ’ ’ aia status marital height§ s age s BMI s ’ ’ ’ age s BMI s height§ s · g4380 473 kg 0 2 n ) † ‡ 2420) ‡ enand mean , SD SD SD n ae1( 1 Wave rMa or % Mean or 15 91 12 · · · 70 0 0 07 06 53 n 3)Wv ( 2 Wave 938) · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 415 91 04 12 21 06 5 81 357 6 7 41 376 2 7 30 274 4 9 54 496 5 6 40 368 8 612 56 6 3 58 537 8 4 37 342 9 4 59 545 5 4 59 545 2 5 28 257 8 6 40 368 5 26 62 4 86 58 2 17 71 3 5 17 156 1 510 95 8 5 93 851 6 8 41 382 8 5 82 757 9 3 69 636 5 7 73 670 5 1 56 516 2 6 50 462 4 8 19 182 7 4 26 243 5 0 11 105 6 8 31 285 7 76 49 27 8 451 6 1 78 715 1 1 45 417 5 310 93 3 45 54 9 14 41 5 5 50 459 3 SD n rMa or % Mean or · · · 90 79 50 0 15 77 n 1)Wv ( 3 Wave 913) · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 1 43 411 3 612 06 515 91 05 21 9 79 198 1 0 42 403 2 9 31 299 0 0 53 501 3 3 36 339 3 8 93 12 880 31 1 8 3 57 536 8 5 38 357 5 5 69 653 7 0 63 600 7 96 59 9 9 31 294 2 8 30 286 3 08 80 8 66 56 4 3 14 132 7 4 15 145 2 58 75 4 5 91 859 2 4 36 346 8 9 84 794 8 0 64 606 7 2 76 720 4 4 47 448 5 8 51 482 6 5 27 258 9 1 23 219 6 1 12 113 5 3 35 331 2 97 48 19 1 457 4 6 70 665 3 3 46 438 7 6 17 161 2 0 11 104 9 94 39 5 SD n rMa or % Mean or · · · 90 89 40 0 24 76 n 3)Wv ( 4 Wave 939) · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 6 20 262 8 613 06 415 92 04 18 6 84 963 8 4 41 545 9 2 32 425 8 9 45 597 3 2 47 624 1 2796 9 1267 107 7 5 6 58 766 1 8 21 282 0 2795 1247 5 8 52 687 9 13 41 3 5 34 454 3 44 64 5 6 12 168 5 82 38 0 4 57 748 0 8 13 181 4 2 9 121 0 1387 1143 5 3 33 432 9 1086 1130 6 5 57 759 5 4 64 846 7 9 30 395 7 3 48 633 3 3 32 431 5 6 36 465 3 4 10 140 0 7 44 577 3 56 51 75 678 7 7 0 69 905 8 1 54 714 7 6 20 266 2 0 23 301 0 91 19 2 SD n rMa or % Mean or · · · 20 32 20 0 42 79 n 31 l ae ( waves All 1311) · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 5438 1594 0 512 05 415 91 04 16 9182 1991 1 8045 1870 6 3132 1311 4 1551 2115 5 7843 1798 6 9595 10 3935 262 6 3 0399 54 4093 4 2231 4 3832 1318 5 8970 2899 1 3356 2303 4 6 4 166 4 680 3132 1331 6 2229 1202 9 5 8 351 8 4 6 248 9 0024 1020 0 9 17 699 8 0 9 402 2 7091 3750 2 4935 1459 0 4283 3402 2 6265 2672 9 9571 2935 5 8745 1867 1 0850 2078 3 0725 1027 9 1628 1166 5 6 11 467 7 4036 1490 0 6 6 49 167 6 2023 7 9973 2999 0 9648 1986 5 3 15 635 3 9 12 496 0 6 4 169 5 SD n NRachmi CN PAGE | 134 PAGE rMa or % Mean or · · · 80 88 20 0 22 81 n 4101) tal et · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 9 27 13 04 2 6 0 6 8 9 9 8 4 1 7 2 1 2 5 3 6 1 9 0 8 4 5 0 2 6 5 7 0 4 4 3 9 3 1 4 5 1 1 SD . Stunted and overweight Indonesian children 5
Table 1 Continued
Wave 1 (n 938) Wave 2 (n 913) Wave 3 (n 939) Wave 4 (n 1311) All waves (n 4101)
Characteristic n or Mean % or SD n or Mean % or SD n or Mean % or SD n or Mean % or SD n or Mean % or SD Household’s wealth index Poor 446 47·6 801 87·7 404 43·0 582 44·4 2233 54·5 Average 141 15·0 69 7·6 192 20·5 269 20·5 671 16·4 Rich 351 37·4 43 4·7 343 36·5 460 35·1 1197 29·1 Community level Housing area Urban 435 46·4 413 45·3 429 45·7 696 53·1 1973 48·1 Rural 503 53·6 500 54·7 510 54·3 615 46·9 2128 51·9 Region Sumatra 248 26·4 209 22·9 206 21·9 322 24·6 985 24·0 Java 466 49·7 502 55·0 519 55·3 647 49·4 2134 52·0 Bali and Nusa Tenggara Barat 130 13·9 115 12·6 103 11·0 204 15·6 552 13·5 Kalimantan and Sulawesi 94 10·0 87 9·5 111 11·8 138 10·4 430 10·5
†Based upon the 2006 WHO Child Growth Standards for children <5 years(23). ‡Based upon the WHO BMI International Classification using general cut-off points(26). §Height-for-age Z-score <−2(27).
Univariate analysis indicated that, compared with 1993, height. This is the first study to show the trends in the the odds of being stunted and overweight increased by 29 % prevalence of concurrent stunting and overweight along in 2007. Children aged 3·0–3·9 and 4·0–4·9 years, and those with the associated risk factors in early childhood in a with overweight/obese fathers were significantly less likely South-East Asian population of children. to be stunted and overweight. Children who were breast-fed The finding that children in the youngest age group after the age of 6 months were 3·49 times more likely to be (2·0–2·9 years) were more likely to experience concurrent stunted and overweight than children who were breast-fed stunting and overweight compared with the older children for less than 6 months. Children whose fathers and mothers highlights the importance of interventions starting as early were of short stature were significantly more likely to be as possible. The WHO, in its policy briefs for stunting(30) stunted and overweight. and also for overweight(31), emphasizes the need for After adjusting for potential confounders, the risk factors multisectoral approaches as well as interventions needing for stunted and overweight were: youngest age group to occur prior to, during and beyond pregnancy. (2·0–2·9 years), breast-fed after the age of 6 months, born As a country undergoing transition, an investment in to mothers who were classified as having short stature and education is definitely on the agenda in Indonesia. Over
Public Health Nutrition living in rural areas. There was no collinearity found in the the duration of the study, the education level of both final model. mothers and fathers improved, as shown by the decline in the percentage of mothers and fathers who had no formal Odds of stunted children being overweight education. Between waves 2 and 3, the number of parents Figure 1 shows the odds of those who were stunted being who went to university nearly doubled. The same overweight, compared with their healthy-height peers, for phenomenon happened between waves 3 and 4 for each wave of data collection. At all time points, stunted women. In many Asian cultures, when it comes to edu- children were significantly more likely to be overweight cation, parents often prefer to send their boys to school, than children who were not stunted (OR > 1). because they will become the head of the family. Within each wave in our study, there was a higher proportion of men with university education compared with women. Discussion The present analysis showed no association between parental education and concurrent stunting and over- The current study presents a series of cross-sectional weight. However, in our first paper(17) we found that surveys from four different time points (1993, 1997, 2000 stunting itself was associated with mothers not having and 2007) over 14 years in Indonesia. We show that formal education. Our first paper also detailed the asso- concurrent stunting and overweight occurs in the 2·0–4·9 ciation between child overweight/obesity status and year age group in Indonesian children, and is more likely fathers who attended university(17),afinding in keeping in the 2·0–2·9 year age group, in children who were with other studies that have shown a positive association breast-fed for longer than 6 months, who lived in rural between socio-economic position and child obesity in areas or whose mothers had short stature. In all four low- and middle-income countries(32,33). waves, stunted children were significantly more likely to Although several reports have shown that the double be overweight/obese compared with children of healthy burden of malnutrition may occur in the same family or
PAGE | 135 Public Health Nutrition aetllevel Parental aibe%9 IO 5%CI % 95 OR CI % 95 % wave each in Prevalence Variable 2 Table 6 oshl level Household hl level Child Maternal ae1(93 6 (1993) 1 Wave ae2(97 6 (1997) 2 Wave Paternal vrbreast-fed Ever Mother Parents ae3(00 5 (2000) 3 Wave Sex ae4(07 7 (2007) 4 Wave g fweaning of Age Age it egt( weight Birth g fsatn opeetr foods complementary starting of Age Father Mother hc-pdrn pregnancy during Check-up 4 Father e 7 Yes Mother oeuain7 6 education No married Currently uiradhg col3 school high and Junior o12 No omrymrid0n/a 7 0 school Primary married Formerly ae7 Male eae6 Female nvriyo oe6 more or University Father uiradhg col5 school high and Junior 2 3 2 Mother nvriyo oe6 more or University oeuain7 education No Father rmr col8 school Primary ≥ < ≥ < < ≥ · · · · 0 0 0 5 4 ots2 months 6 ots7 months 6 2 ots6 months 6 ots8 months 6 e 7 Yes omlhih 4 height Normal o5 No hr ttr 7 stature Short newih 6 Underweight oml6 Normal vregtoee6 Overweight/obese newih 5 Underweight oml7 Normal omlhih 4 height Normal vregtoee2 Overweight/obese hr ttr 8 stature Short < ≥ < ≥ – – – – · · g9 kg 0 g7 kg 5 ’ 0yas6 years 30 0yas6 years 30 rvlneo ocretsutn n vregtaogcide gd2 aged children among overweight and stunting concurrent of Prevalence 0yas7 years 30 0yas6 years 30 4 2 3 education s ’ < education s ’ · · · er 3 years 9 er 10 years 9 er 5 years 9 4 ’ ’ ’ aia status marital height§ s age s BMI s ’ ’ ’ age s BMI s height§ s · g6 kg 0 n ‡ 2420) ‡ tnigadoewih ndutdAdjusted Unadjusted overweight and Stunting · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 85 *6 0* 35 *3 9* 45 87 24 56 56 85 82 45 16 24 74 55 24 93 25 55 75 26 06 26 *0 1* 05 14 14 *3 3* 06 44 *8 2* 46 24 84 05 *3 5* 41 47 54 75 16 14 · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ,8 7, ,8 4, ,7 3, ,6 9, ,8 0, ,10 5, ,6 1, ,13 5, ,8 4, ,8 3, ,4 9, ,9 6, ,8 3, ,9 2, ,11 9, 7 8, ,6 0, ,5 0, ,7 4, ,26 3, ,7 8, ,8 1, ,8 0, ,8 0, ,5 9, 7 0, ,7 8, ,7 8, ,8 5, ,8 1, ,6 3, ,11 7, ,8 4, ,12 1, ,7 7, ,7 0, ,5 7, ,4 4, ,9 3, ,9 6, ,7 7, ,10 5, ,12 1, · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · Ref. 0 Ref. 0 30 Ref. 1 Ref. 2 10 71 61 90 Ref. 7 Ref. 1 Ref. 0 Ref. 3 80 10 20 71 50 Ref. 2 81 e.Ref. Ref. 1 10 70 70 Ref. 0 13 60 51 20 10 e.Ref. Ref. 5 20 61 00 Ref. 9 · · · · · · · · 31 71 Ref. 7 72 e.Ref. Ref. 9 Ref. 2 01 Ref. 1 · · · · · · · · · · · · · · · · · · · · · · · · · · 80 88 80 58 10 71 90 39 81 38 90 09 40 34 10 81 70 70 77 51 45 60 06 70 97 90 29 50 85 40 94 70 97 91 49 20 72 10 31 10 51 30 93 10 41 41 84 40 84 20 22 · · · · · · · · · · · · · · · · · · · · · · · · · · 0 1 50, 0 39, 9 1 69, 5 2 85, 3 0 53, 7 2 67, 5 1 05, 6 1 76, 5 0 25, 5 1 55, 6 1 36, 7 1 57, 0 5 00, 3 1 63, 9 1 59, 3 1 93, 7 1 67, 4 1 54, 5 2 55, 2 8 42, 2 1 42, 5 2 85, 1 0 21, 2 2 42, 6 2 66, 1 1 91, · 0 – 4 · · · · · · · · · · · · · · · · · · · · · · · · · · · er ( years 9 30 63 40 74 68 20 12 60 96 40 94 90 89 30 83 90 59 47 00 20 70 37 60 06 90 99 10 41 00 60 00 80 80 08 30 33 40 04 10 61 20 22 20 02 50 85 38 70 07 < < < n 0 0 0 11,Idnsa aiyLf Survey Life Family Indonesian 4101), P · · · · · · · · · · · · · · · · · · · · · · · · · · 167 817 0 001 303 152 010 382 020 609 0 0 001 309 298 120 050 779 917 124 186 475 868 0 2 007 229 219 016 0 1 001 584 OR R9 CI % 95 OR · · · · 61 66 70 57 40 44 81 98 · · · · 2 2 22, 1 0 41, 0 0 30, 0 7 20, NRachmi CN PAGE | 136 PAGE · · · · 80 58 10 81 65 10 41 † < 0 tal et P · · · · 001 001 001 010 . Stunted and overweight Indonesian children 7
Table 2 Continued
OR
Stunting and overweight Unadjusted Adjusted†
Variable % 95 % CI OR 95 % CI P OR 95 % CI P Household’s wealth index Poor 7·06·0, 8·1 Ref Average 5·64·0, 7·70·81 0·56, 1·16 0·253 Rich 6·04·7, 7·50·86 0·64, 1·14 0·290 Community level Housing area Urban 4·7* 3·9, 5·7 Ref. Ref. Rural 8·17·0, 9·41·79 1·38, 2·35 <0·001 1·66 1·19, 2·32 0·003 Region Sumatra 6·95·5, 8·7 Ref. Java 6·75·7, 7·80·98 0·71, 1·47 0·923 Bali and Nusa Tenggara Barat 6·34·6, 8·70·89 0·49, 1·63 0·711 Kalimantan and Sulawesi 4·93·2, 7·40·68 0·35, 1·32 0·256
Ref., reference category; n/a, not applicable. *P value <0·003 (Bonferroni adjusted). †Independent variables adjusted for are child-, parental- and household-, and community-level factors. ‡Based upon the WHO BMI International Classification using general cut-off points(26). §Height-for-age Z-score <−2(27).
OR 95 % CI P value
1993 1.66 1.07, 2.55 0.0218
1997 2.01 1.30, 3.11 0.0017
2000 1.34 1.05, 1.73 0.0205 Public Health Nutrition
2007 1.44 1.10, 1.91 0.0015
0.0 1.01.52.02.53.0 3.5 OR (95 % CI) Fig. 1 Odds ratios, with their 95 % confidence intervals represented by horizontal bars, of stunted children aged 2·0–4·9 years being overweight; wave 1 (1993), wave 2 (1997), wave 3 (2000) and wave 4 (2007) of the Indonesian Family Life Survey
even individual in Indonesia(1,7,34–36),nonehasspecifically report(16) which showed a prevalence of 19 %, albeit in a looked at the prevalence of this phenomenon in very young small sample of 162 children aged 3 years. Another impor- children or explored the associated risk factors. Our findings tant factor is that different definitions and cut-off points are show the prevalence of concurrent stunting and overweight used to determine the prevalence of the double burden of in Indonesia to be more than 5 % in all four waves, with an malnutrition. The lack of one standardised definition and set overall rate increase of 0·06 % per year over the 14-year of cut-offs makes interpretation of different studies more period. Although this prevalence is still relatively low, difficult. Future research in this area would benefitfroma studies from other countries have generally shown a lower consensus on standard definitions and cut-off points. Other prevalence of combined stunting and overweight in both research recommendations include more focus on eating school-aged and pre-school-aged children(4,10,12,14,15,37,38). behaviours and physical activity. There is also a need for The two exceptions are a report from Uruguay, where a prospectively designed studies. concurrent prevalence of 13·2 % was documented in There are few reports of risk factors associated with 2046 children aged 0–59 months(13), and a South African concurrent stunting and overweight in childhood,
PAGE | 137 8 CN Rachmi et al. particularly in nationally representative samples(9,39). In a both in Indonesia and globally(35,47).Thus,thecon- study of 7555 children aged 2–6 years from rural areas in sequences of concurrent stunting and overweight, although Mexico, Fernald and Neufeld found similar results to our never having been addressed previously, are of great study, whereby concurrent stunting and overweight was importance. Furthermore, as has been highlighted in reports higher in children whose mothers had short stature(9),a from the World Bank and other reviews(1,8,34), there are finding also documented by Keino et al. in their review on internationally recognised policies and strategies for com- determinants of this phenomenon in children from sub- bating stunting(1,8,30,48,49) and overweight/obesity(1,8,31); Saharan Africa(39). Other factors found to be related to although these have usually been addressed separately. concurrent stunting from these studies were lower socio- A few very recent policies have specifically addressed the economic status, lower maternal education, large house- double burden of malnutrition, such as with the South hold size and younger mothers(9,39). One of the findings in African food-based dietary guidelines(50). our study is the apparently counterintuitive observation The strengths of our study include the large sample size that a longer duration of breast-feeding is associated with of participants and the use of trained observers to under- concurrent stunting and overweight. Several studies have take anthropometric measurements. The representative shown that prolonged breast-feeding duration protects nature of sampling and the use of similar methods against obesity in childhood(40,41) and also stunting(42,43). throughout the four waves allow comparison of results Therefore, our finding must be interpreted with caution, between waves. We also included sampling weights keeping in mind that many other factors influence a child’s during the analyses to reduce potential bias. nutritional status, including the age of commencing com- One limitation of the study is the cross-sectional design, plementary feeding, the type of complementary feeding, limiting the ability to explore causation. Another is the the use of formula milk and the family hygiene and unavailability, or limited amount, of data on children’s and sanitation conditions. adults’ physical activity. In addition, we were not able to We compared the four risk factors associated with assess other aspects of the child’s eating behaviours concurrent stunting and overweight found in the current (including their breast-feeding details: exclusively, analysis with those we identified for stunting or for over- combined with formula milk, or predominantly formula weight as individual phenomena in our previous paper(17). milk), nor the family hygiene and sanitation conditions, all One – being in the youngest age group (2·0–2·9 years) – of which influence children’s nutritional status. Even was a similar risk factor for being overweight alone. The though 95 % of these children were ever breast-fed, more other associated risk factors – being breast-fed for more than 70 % were started on complementary foods at less than 6 months, having a short-statured mother and living than 6 months of age. The quality of these complementary in a rural area – were similar to the risk factors for stunting foods will also have an impact on weight status and linear alone. No new risk factors associated with concurrent growth. stunting and overweight were identified. However, as we
Public Health Nutrition have earlier stated in the ‘Methods’ section, we can only Conclusion report on the potential risk factors available from the data sets. In conclusion, our paper demonstrates that the double We also investigated whether stunted children are at burden of malnutrition occurs at the individual level in greater risk of being overweight compared with their Indonesian children aged 2·0–4·9 years. Such data should healthy-height peers. Bove et al. had similar results to our serve as the catalyst for developing policy and pro- findings, where stunted children aged 0–4·9 years were grammes in dealing with concurrent stunting and over- 2·7 (95 % CI 1·8, 4·1) times more likely to be overweight/ weight. It is important that both policy makers and health obese (BMI Z-score >+2) than children of healthy practitioners work together in addressing this public health height(13). A study by Popkin et al. also showed in children problem. aged 3 to 9 years from four different countries (Russia, China, South Africa and Brazil) that there was a significant association between stunting and overweight/obesity (OR of Acknowledgements 1·7to7·7)(11). Interestingly, in a 2004 study from Indonesia – among 3010 prepubertal school-aged children, randomly Acknowledgements: The authors would like to thank sampled from one urban (Yogyakarta) and one rural area Dr Christine Peterson from Rand Corporation for her (Gunung Kidul) – the stunted children were less likely to be assistance regarding the IFLS data sets and ethics infor- overweight compared with their non-stunted peers(36). mation. Financial support: This research was performed Stunting is closely related to lower cognitive perfor- as part of C.N.R.’s PhD studies, for which she received a mance(1,44,45),poorermotordevelopment(1,45) and lower scholarship from Lembaga Pengelola Dana Pendidikan immune function(1). Overweight is an important underlying (LPDP), the Republic of Indonesia. LPDP had no role in cause for the occurrence of many non-communicable the design, analysis or writing of this article. Conflict of diseases(46), one of the main causes of death and disability interest: None. Authorship: All authors formulated the
PAGE | 138 Stunted and overweight Indonesian children 9
research questions, carried out the study, analysed and 16. Mamabolo RL, Alberts M, Steyn NP et al. (2005) Prevalence interpreted the data, and wrote, revised and reviewed the and determinants of stunting and overweight in 3-year-old draft of the article. Ethics of human subject participation: black South African children residing in the Central Region of Limpopo Province, South Africa. Public Health Nutr 8, This study was conducted according to the guidelines laid 501–508. down in the Declaration of Helsinki and all procedures 17. Rachmi CN, Agho KE, Li M et al. (2016) Stunting, under- involving human subjects/patients were approved by the weight and overweight in children aged 2·0–4·9 years in by Institutional Review Boards in the USA (at Rand Indonesia: prevalence trends and associated risk factors. PLOS One (In the Press). Corporation) and Indonesia (at Universitas Gadjah Mada 18. Rand Corporation (2014) The Indonesian Family Life Survey and earlier at Universitas Indonesia). Written informed (IFLS). http://www.rand.org/labor/FLS/IFLS.html (accessed consent was obtained from all subjects/patients. May 2014). 19. Frankenberg E, Karoly LA, Gertler P et al. (1995) The 1993 Indonesian Family Life Survey: Overview and Field Report. Santa Monica, CA: Rand Corporation; available at http:// References www.rand.org/pubs/drafts/DRU1195z1.html 20. Serrato C & Melnick G (1995) The Indonesian Family Life 1. Shrimpton R & Rokx C (2012) The Double Burden of Survey: Overview and Descriptive Analysis of the Population, Malnutrition: A Review of Global Evidence. Washington, Health, and Education Data. Santa Monica, CA: Rand DC: World Bank. Corporation; available at http://www.rand.org/pubs/drafts/ 2. International Food Policy Research Institute (2014) Global DRU1191.html Nutrition Report 2014: Actions and Accountability to 21. Victora CG, de Onis M, Hallal PC et al. (2010) Worldwide Accelerate the World’s Progress on Nutrition. Washington, timing of growth faltering: revisiting implications for inter- DC: IFPR. ventions. Pediatrics 125, e473–e480. 3. Doak CM, Adair LS, Monteiro C et al. (2000) Overweight 22. Pan H & Cole TJ (2010) LMSGROWTH, a Microsoft Excel and underweight coexist within households in Brazil, China add-in to access growth references based on the LMS method and Russia. J Nutr 130, 2965–2971. (Version 2.68). http://www.healthforallchildren.com/shop- 4. Freire WB, Silva-Jaramillo KM, Ramirez-Luzuriaga MJ et al. base/shop/software/lmsgrowth/ (accessed April 2014). (2014) The double burden of undernutrition and excess 23. World Health Organization (2006) The WHO Child Growth body weight in Ecuador. Am J Clin Nutr 100, issue 6, Standards. http://www.who.int/childgrowth/standards/en/ 1636S–1643S. (accessed May 2014). 5. Doak CM, Adair LS, Bentley M et al. (2005) The dual burden 24. de Onis M, Blossner M & Borghi E (2010) Global prevalence household and the nutrition transition paradox. Int J Obes and trends of overweight and obesity among preschool (Lond) 29, 129–136. children. Am J Clin Nutr 92, 1257–1264. 6. Black RE, Victora CG, Walker SP et al. (2013) Maternal and 25. Davison KK & Birch LL (2001) Childhood overweight: a child undernutrition and oveweight in low-income and contextual model and recommendations for future research. middle-income countries. Lancet 382, 427–451. Obes Rev 2, 159–171. 7. Oddo VM, Rah JH, Semba RD et al. (2012) Predictors of 26. World Health Organization (2006) World Health Organization maternal and child double burden of malnutrition in rural Global Database on Body Mass Index. http://apps.who.int/ Indonesia and Bangladesh. Am J Clin Nutr 95, 951–958. bmi/index.jsp?introPage=intro_3.html (accessed May 2014). 8. Haddad L, Cameron L & Barnett I (2015) The double burden 27. World Health Organization (2007) WHO Reference 2007. fi Growth reference data for 5–19 years. http://www.who.int/
Public Health Nutrition of malnutrition in SE Asia and the Paci c: priorities, policies and politics. Health Policy Plan 30, 1193–1206. growthref/en/ (accessed May 2014). 9. Fernald L & Neufeld L (2007) Overweight with concurrent 28. Filmer D & Pritchett LH (2001) Estimating wealth effects stunting in very young children from rural Mexico: pre- without expenditure data – or tears: an application to educa- valence and associated factors. Eur J Clin Nutr 61, 623–632. tional enrollments in states of India. Demography 38,115–132. 10. Piernas C, Wang D, Du S et al. (2015) The double burden of 29. Rabe-Hesketh S & Skrondal A (2012) Multilevel and under- and overnutrition and nutrient adequacy among Longitudinal Modelling Using Stata, 3rd ed. College Chinese preschool and school-aged children in 2009–2011. Station, TX: Stata Press. Eur J Clin Nutr 69, 1323–1329. 30. World Health Organization (2014) WHO Global Nutrition 11. Popkin BM, Richards MK & Montiero CA (1996) Stunting is Targets 2025: Stunting Policy Brief. http://www.who. associated with overweight in children of four nations int/nutrition/topics/globaltargets_stunting_policybrief.pdf that are undergoing the nutrition transition. J Nutr 126, (accessed November 2014). 3009–3016. 31. World Health Organization (2014) Global Nutrition Targets 12. Iriart C, Boursaw B, Rodrigues GP et al. (2013) Obesity and 2025: Childhood Overweight Policy Brief. http://www. malnutrition among Hispanic children in the United States: who.int/nutrition/globaltargets_overweight_policybrief.pdf double burden on health inequities. Rev Panam Salud (accessed November 2014). Publica 34, 235–243. 32. Gupta N, Goel K, Shah P et al. (2012) Childhood obesity in 13. Bove I, Miranda T, Campoy C et al. (2012) Stunting, over- developing countries: epidemiology, determinants, and weight and child development impairment go hand in hand prevention. Endocr Rev 33, 48–70. as key problems of early infancy: Uruguayan case. Early 33. Wang Y & Lim H (2012) The global childhood obesity Hum Dev 88, 747–751. epidemic and the association between socio-economic status 14. Kroker-Lobos MF, Pedroza-Tobias A, Pedraza LS et al. and childhood obesity. Int Rev Psychiatry 24,176–188. (2014) The double burden of undernutrition and excess 34. Shrimpton R & Rokx C (2013) The Double Burden of body weight in Mexico. Am J Clin Nutr 100, issue 6, Malnutrition in Indonesia. Jakarta: World Bank Jakarta. 1652S–1658S. 35. Badan Penelitian dan Pengembangan Kesehatan KKRI 15. Ramirez-Zea M, Kroker-Lobos MF, Close-Fernandez R et al. (2014) Status Gizi. In Laporan Hasil Riset Kesehatan Dasar (2014) The double burden of malnutrition in indigenous Indonesia tahun 2013, Riskesdas Dalam Angka, 1st ed., and nonindigenous Guatemalan populations. Am J Clin pp. 386–415 [Departemen Kesehatan, editor]. Jakarta: CV Nutr 100, issue 6, 1644S–1651S. Kiat Nusa.
PAGE | 139 10 CN Rachmi et al.
36. Julia M, van Weissenbruch MM, de Waal HA et al. (2004) 44. Sandjaja, Poh BK, Rojroonwasinkul N et al. (2013) Influence of socioeconomic status on the prevalence of Relationship between anthropometric indicators and stunted growth and obesity in prepubertal Indonesian cognitive performance in Southeast Asian school-aged children. Food Nutr Bull 25, 354–360. children. Br J Nutr 110, Suppl. 3, S57–S64. 37. Severi C & Moratorio X (2014) Double burden of 45. Sudfeld CR, McCoy DC, Danaei G et al. (2015) Linear undernutrition and obesity in Uruguay. Am J Clin Nutr 100, growth and child development in low- and middle-income issue 6, 1659S–1662S. countries: a meta-analysis. Pediatrics 135, e1266–e1275. 38. Li Y, Wedick NM, Lai J et al. (2011) Lack of dietary diversity 46. World Health Organization (2003) Diet, Nutrition and the and dyslipidaemia among stunted overweight children: the Prevention of Chronic Diseases. Report of a Joint WHO/FAO 2002 China National Nutrition and Health Survey. Public Expert Consultation. WHO Technical Report Series no. 916. Health Nutr 14, 896–903. Geneva: WHO. 39. Keino S, Plasqui G, Ettyang G et al. (2014) Determinants of 47. Lozano R, Naghavi M, Foreman K et al. (2012) Global and stunting and overweight among young children and ado- regional mortality from 235 causes of death for 20 age lescents in sub-Saharan Africa. Food Nutr Bull 35, 167–178. groups in 1990 and 2010: a systematic analysis for the 40. Armstrong J & Reilly JJ; Child Health Information Team Global Burden of Disease Study 2010. Lancet 380, (2002) Breastfeeding and lowering the risk of childhood 2095–2128. obesity. Lancet 359, 2003–2004. 48. Bloem MW, de Pee S, Hop le T et al. (2013) Key strategies to 41. Zheng JS, Liu H, Li J et al. (2014) Exclusive breastfeeding is further reduce stunting in Southeast Asia: lessons from the inversely associated with risk of childhood overweight in a ASEAN countries workshop. Food Nutr Bull 34, 2 Suppl., large Chinese cohort. J Nutr 144, 1454–1459. S8–S16. 42. Marquis GS, Habicht JP, Lanata CF et al. (1997) Association 49. de Onis M, Dewey K, Borghi E et al. (2013) The World of breastfeeding and stunting in Peruvian toddlers: an Health Organization’s global target for reducing childhood example of reverse causality. Int J Epidemiol 26, 349–356. stunting by 2025: rationale and proposed actions. Matern 43. Simondon KB, Simondon F, Costes R et al. (2001) Breast- Child Nutr 9, Suppl. 2, 6–26. feeding is associated with improved growth in length, but 50. Vorster HH, Badham JB & Venter CS (2013) An introduction not weight, in rural Senegalese toddlers. Am J Clin Nutr 73, to the revised food-based dietary guidelines for 959–967. South Africa. S Afr J Clin Nutr 26, 3 Suppl., S5–S12. Public Health Nutrition
PAGE | 140
Chapter 6: Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence?
INTRODUCTION 142 METHODS Indonesian Family Life Survey 143 Data collection 143 Study participants 143 Anthropometry and blood pressure calculations 144 Statistical analysis 145 RESULTS Characteristics of participants 145 Stunted children and overweight/obesity in adolescence: 14-year cohort analysis 145 Stunted children and thinness in adolescence: 14-year cohort analysis 145 Stunted children and overweight/obesity in adolescence: 7-year cohort analyses 145 Prevalence ratio of stunted children being overweight/obese 145 Stunted children and high blood pressure 147 DISCUSSION 149 REFERENCES 150
This chapter has been previously published as: Rachmi CN, Agho KE, Li M, Baur LA. Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? International Journal of Public Health. 2017; 62(1):153-162.
PAGE| 141 Int J Public Health (2017) 62:153–162 DOI 10.1007/s00038-016-0905-x
ORIGINAL ARTICLE
Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence?
Cut Novianti Rachmi . Kingsley Emwinyore Agho . Mu Li . Louise Alison Baur
Received: 21 April 2016 / Revised: 20 September 2016 / Accepted: 24 September 2016 / Published online: 4 October 2016 Ó Swiss School of Public Health (SSPH+) 2016
Abstract stunting and thinness at adolescence or in the odds/likeli- Objectives To determine whether stunted young children hood of having high systolic or diastolic blood pressure. are at greater risk of (1) overweight/obesity or thinness, and Conclusions We found no association between early life (2) high blood pressure (HBP) in adolescence. stunting and overweight/obesity, thinness and HBP in Methods A secondary data analysis using the Indonesian adolescence. Family Life Survey waves 1 (1993) to 4 (2007). We gen- erated a 14-year follow-up cohort (1993–2007) and two Keywords Stunting Overweight/obesity 7-year cohorts (1993–2000 and 2000–2007) of children aged Double burden CohortsÁ Indonesia ChildrenÁ Á Á Á 2.0–4.9 years. Stunting (HAZ \ -2), thinness (BMIZ \ -2), and overweight/obesity (BMIZ [ ?1) were deter- mined based upon the WHO Child Growth Standards. HBP Introduction ([90th percentile) was interpreted using the 4th Report on the Diagnosis of HBP in Children and Adolescents. For many low- and middle-income countries, the rising Results 765, 1083, and 1589 children were included in the prevalence of childhood obesity is in addition to the 14-year cohort, and the two 7-year cohort analyses, existing problem of stunting, leading to one of the forms of respectively. In the 7-year cohorts, early life stunting was the double burden of malnutrition (IFPRI 2015). Stunting, inversely associated with overweight/obesity (prevalence defined as low length/height-for age, is a form of chronic ratio 0.32 and 0.38, respectively; P \ 0.05), but no sig- malnutrition that has been a major nutritional problem nificant association was found with the 14-year cohort. mainly in developing countries. In 2011, there were There was no significant association between childhood approximately 25.7 % under five children globally who were stunted (length or height-for age z score \-2), Electronic supplementary material The online version of this 28.0 % in developing countries, and 7.2 % in developed article (doi:10.1007/s00038-016-0905-x) contains supplementary countries (de Onis et al. 2013). material, which is available to authorized users. Several cross-sectional studies show that stunted chil- dren are more likely to be overweight/obese, compared to C. N. Rachmi (&) L. A. Baur Discipline of ChildÁ and Adolescent Health, The Children’s those with healthy heights (Bove et al. 2012; Popkin et al. Hospital at Westmead (University of Sydney Clinical School), 1996; Rachmi et al. 2016a). However, whether stunting is a Sydney, Australia risk factor for the later development of overweight/obesity e-mail: [email protected]; [email protected] is still contested. A 22-month longitudinal case–control K. E. Agho study of Brazilian girls aged 7–11 years at baseline found School of Science and Health, Western Sydney University, an association between mild stunting and later obesity Sydney, Australia (Sawaya et al. 1998). Others have found no prospective relation between stunting in children less than six years and M. Li L. A. Baur SydneyÁ School of Public Health, The University of Sydney, subsequent overweight/obesity in adolescence or adulthood Sydney, Australia (Benefice et al. 2001; Cameron et al. 2005; Schroeder et al.
123 PAGE | 142 154 C. N. Rachmi et al.
1999; Timaeus 2012; Walker et al. 2001, 2007). Indeed, families, and their individual members, from wave 1 in Gigante et al. (2007) found that stunting in early life was 1993 to the most recent survey. IFLS details have been associated with a reduced fat mass index and body mass described in several reports, including our first paper index in adolescence. Some have argued that previous (Frankenberg et al. 1995; IFLS 2014; Serrato and Melnick cross-sectional studies have been hampered by errors in 1995; Rachmi et al. 2016b). In summary, there are four height measurement (Timaeus 2012). There are limited waves of the survey for which data have been publicly longitudinal studies in countries with high rates of stunting, released, i.e., 1993, 1997, 2000, and 2007. The survey and none has been conducted in Asia (Schroeder et al. sampled families from 13 out of Indonesia’s 27 provinces 1999; Benefice et al. 2001; Cameron et al. 2005; Gigante in 1993 using stratified random sampling. It comprised et al. 2007; Walker et al. 2007; Timaeus 2012). questionnaires followed directly by anthropometric mea- Children who are stunted in early life have also been surements, with each subsequent survey using similar found to have a higher mean systolic blood pressure in later methods to the first wave. The data collection included childhood, irrespective of current size (Gaskin et al. 2000). direct interviews for adults and proxy interviews for chil- A separate study showed that stunted adolescents who are dren and infants. Questionnaires covered a broad range of currently overweight have a higher mean systolic blood issues including relationships between members of the pressure compared to their healthy height and overweight household, economic, and non-economic measures of counterparts (Clemente et al. 2012). Both studies reported household wealth, individual data for adults and children, mean blood pressure with neither using cut-off points for and location of the house. These data were recorded in hypertension. The use of hypertension cut-off points is seven different books. Trained nurses also weighed and important in determining whether stunting in early child- measured all adults and children in each household. In this hood poses a specific risk for hypertension in adolescence. study, we combined information from several books and This paper is the third in a series of secondary data divided them into three clusters: the child, household, and analyses on the double burden of malnutrition in Indonesia community level factors. This longitudinal survey has a using data from the Indonesian Family Life Survey (IFLS). high recontact rate ([90 % of households and individuals) The previous papers focused on cross-sectional analyses of in every wave as compared to wave 1 (Frankenberg et al. children aged 2.0–4.9 years. This paper focuses on stunted 1995; Frankenberg and Thomas 2000; IFLS 2014; Serrato children aged 2.0–4.9 years in 1993, and follows them and Melnick 1995). longitudinally through to 2007, i.e., 14 years later. We also performed analyses on two separate 7-year cohorts of Study participants children with the same baseline age group (one from 1993 to 2000, and the other from 2000 to 2007) to determine The survey’s design allowed us to follow children aged whether there were secular trends in any associations. We 2.0–4.9 years at baseline in year 1993 to age aimed to determine whether stunted young children are at 15.0–18.9 years in 2007. In addition, we also followed greater risk, in adolescence, of (1) overweight/obesity or children aged 2.0–4.9 years in either 1993 or 2000 in two thinness, and (2) high blood pressure. 7-year cohorts, i.e., to 2000 and to 2007, respectively. In 2013, one in every three Indonesian children was For the purposes of this study, we only included children stunted (BPPK 2014). This study provides valuable infor- with complete records of child level factors [birth dates, sex, mation about the long-term effects of stunting. While many height, weight, and blood pressure (only in wave 4)], strategies are currently in place in Indonesia and other household level factors (mothers’ education and household developing countries to reduce the prevalence of stunting, wealth index), and community level factors (housing area such information about early life stunting and later life and region). The flowchart in Fig. 1 shows the total number outcomes would help with the development of new policies/ of eligible participants in each original cohort and the final strategies to deal with individuals who are already stunted. number of children included in the analysis. From 1385 children available in wave 1 (1993), only 765 children had complete records in all four waves. The Electronic Supple- Methods mentary Materials Table 1 shows there were no differences in mean values for the baseline variables between those who Indonesian family life survey completed all four waves and those who did not apart from a slight decrease in the prevalence of underweight (33.3 vs Data collection 34.3 %, P = 0.001) This suggests that the results of the analysis will not be affected by censored data. The Indonesian Family Life Survey is the only Indonesian Mothers’ education was categorized as: never attended survey that provides longitudinal data tracing the same any formal education, attended primary school, and
123 PAGE | 143 Are stunted young Indonesian children more likely to be overweight, thin, or have high blood… 155
Fig. 1 The number of participants in each original cohort and the final number of eligible participants in the analysis; Data from Indonesian Family Life Survey, Indonesia wave 1 (1993), 2 (1997), wave 3 (2000), and wave 4 (2007). FU followed-up, LFU lost to follow up, Implausible values using cut-off points from the WHO Anthro software (WHO 2011) for child growth standards (igrowup)
attended middle school or higher. We calculated the weight/height2), height, and weight were expressed as Household Wealth Index by giving a score involving the z scores using the LMS Growth program (LMS 2010). possession of eleven householdassets,andrankedthem Z scores were calculated against the 2006 WHO Child in five quintiles from poorest to richest using principal Growth Standard (WHO 2006) for children under five components analysis method. Households in the bottom years in the first wave and the 2007 WHO Growth Refer- two quintiles were categorized as poor, those in the ence for School-aged Children and Adolescents for the middle two quintiles as middle and the highest quintile as three subsequent waves (WHO 2007). We identified and rich households (Filmer and Pritchett 2001). Housing then discarded biologically implausible values using cut- area was classified as urban or rural areas and the regions off points from the WHO Anthro software (WHO 2011) for were divided into four regions: Sumatra, Java, Bali and child growth standards (igrowup) as the following: weight- Nusa Tenggara Barat (NTB), and Kalimantan and for-age z score (WAZ) \-6 and WAZ [5, height-for-age Sulawesi. z score (HAZ) \-6 and HAZ [6, BMI z score (BMIZ) \-5 and BMIZ [5. We defined stunting as an Anthropometry and blood pressure calculations HAZ \-2, thinness as a BMIZ \-2, and overweight/ obesity as a BMIZ [?1 according to the 2006 WHO Child Trained nurses measured children’s height using Shorr Growth Standards for Children \5 years (WHO 2006) and measuring boards Model 420. Children’s weight was the 2007 WHO Standard Growth Reference for School- measured using Seca Floor Model 770 scales (SECA, Los aged Children and Adolescents (WHO 2007). In children Angeles, CA, USA), with digital read out (Frankenberg under five, we combined the ‘at risk of overweight/obesity, et al. 1995). The same methods were used in the next three overweight, and obesity’ categories, defined as waves. Blood pressure was measured twice in the same BMIZ [?1, as suggested by de Onis et al. (2010). visit using an Omron self-inflating meter that produced a Systolic and diastolic blood pressures (SBP and DBP) digital read out (Frankenberg and Thomas 2000). We used were interpreted using the percentile table in the Fourth the mean of two measurements in the analysis. Blood Report on the Diagnosis, Evaluation, and Treatment of pressure measurement started in wave 2 of the survey, and High Blood Pressure in Children and Adolescents was only collected in children aged 15 years or older. (NHBPEP 2004). High systolic or diastolic blood pressure Thus, we only analyzed blood pressure in wave 4 in this was defined as a blood pressure [90th percentile, unless study, because the children were [15 years by then. the 90th percentile was [120 mmHg for systolic We calculated age based on the birth dates and dates of or [80 mmHg for diastolic, in which case the latter cut-off anthropometric measurements. Body mass index (BMI; points were used (NHBPEP 2004).
123 PAGE | 144 156 C. N. Rachmi et al.
Statistical analysis wave 2 (5.0 %), before rising again to wave 4 (7.6 %). The prevalence of thinness increased from wave 1 (6.3 %) to We merged data files from several separate questionnaires wave 3 (12.9 %), and then decreased in wave 4 (8.9 %). In and anthropometry files, and analyzed the data using wave 4, 35.3 % of adolescents had high SBP and 29.0 % STATA Data Analysis and Statistical Software version 13 had high DBP. The characteristics of the two 7-year (Stata Corp 2014). We generated two data sets. The first cohorts are available in Electronic Supplementary Materi- consisted of children aged 2.0–4.9 years in 1993 who were als (ESM Tables 2, 3). followed over 14 years to 2007. The second consisted of two sets of 7-year follow-up of children in the same age Stunted children and overweight/obesity bracket at baseline in 1993 and 2000, who were followed to in adolescence: 14-year cohort analysis years 2000 and 2007, respectively. We conducted fre- quency tabulations to describe distributions. Results are Of the children who were either stunted or of healthy presented as numbers and prevalence or means and stan- height in 1993, 8.6 % (95 % CI 6.2–11.9) and 6.5 % (95 % dard deviations. The next step was performed using the CI 4.5–9.5), respectively, were overweight/obese in 2007 Survey (‘Svy’) commands that adjust for cluster sampling (Table 2). design and weights. The process comprised calculating 95 % confidence intervals around prevalence estimates. Stunted children and thinness in adolescence: 14-year The prevalence of stunted children in 1993 who become cohort analysis overweight/obese 14 years later (2007), as well as the prevalence of stunted children in 1993 and 2000 who Table 2 shows the prevalence of previously stunted chil- become overweight/obese seven years later (in 2000 and dren in wave 1 who were thin in wave 4. The prevalence of 2007), is presented with 95 % confidence intervals (CIs). thinness within previously stunted children was 8.4 % Likewise, we computed the prevalence of stunted children (6.0–11.6), which did not differ significantly from the in 1993 who fell in the thinness category fourteen years prevalence of thinness in those who were previously of later in 2007. healthy height [9.4 % (6.9–12.8)]. The prevalence ratio of The prevalence ratio of stunted children being over- stunted children becoming thin 14 years later, compared to weight/obese in later childhood or adolescence were those of healthy height, was not significant at 0.9 (95 % CI computed by dividing the likelihood of being overweight/ 0.6–1.4); P = 0.497. obese in previously stunted children to the likelihood in children with healthy heights. The same calculations were Stunted children and overweight/obesity also used to compute the likelihood of thinness (14 years in adolescence: 7-year cohort analyses later) in previously stunted children. The prevalence of high blood pressure in adolescents who were either stunted Table 3 shows that the prevalence of overweight/obesity at or of healthy height in early childhood is presented with 7 years follow-up (i.e., 2000 or 2007) in children who were 95 % CIs, along with the odds ratio. Comparisons were stunted at baseline (i.e., 1993 and 2000) differed signifi- undertaken using the Chi-squared test. cantly, [1.7 % (0.9–3.1) and 5.0 % (3.4–7.3), respectively], as did the prevalence in those who were of healthy height at baseline [7.7 (5.6–10.4) and 16.5 (13.6–20.0), Results respectively].
Characteristics of participants Prevalence ratio of stunted children being overweight/ obese 765 children were included in the 14-year, and 1083 and 1589 children were included in the two 7-year cohort Figure 2 shows the prevalence ratio of those who were analyses, respectively. Characteristics of the children stunted at baseline being overweight/obese either seven or included in the 14-year follow-up, by wave, are shown in 14 years later, compared to their healthy height peers. The Table 1. In the 14-year cohort, there were a slightly higher 14-year analysis showed that the prevalence ratio of stun- percentage of boys compared to girls in the survey, with a ted young children becoming overweight/obese 14 years similar mean age for boys and girls within each wave. The later was not significant at 1.26 (95 % CI 0.76–2.07). In prevalence of stunting decreased from wave 1 (49.0 %; both of the 7-year analyses, stunting had a significant aged 2.0–4.9 years) to wave 4 (30.5 %; aged inverse association toward later overweight/obesity (PR 16.0–18.9 years). The prevalence of at risk/overweight or 0.32; 95 % CI 0.17–0.60 and PR 0.38; 95 % CI 0.27–0.53, obesity was highest in wave 1 (10.0 %), then decreased to respectively).
123 PAGE | 145 Are stunted young Indonesian children more likely to be overweight, thin, or have high blood… 157
Table 1 Characteristics of children aged 2.0–4.9 years in waves 1 (1993) and followed-up to wave 4 (2007): prevalence (n [%]) and mean [95 % confidence intervals (CI)] Wave 1 (1993) Wave 2 (1997) Wave 3 (2000) Wave 4 (2007)
Age (years) [mean (95 % CI)] Girls 3.4 (3.3–3.5) 7.4 (7.3–7.5) 10.2 (10.1–10.3) 17.5 (17.4–17.6) Boys 3.4 (3.4–3.5) 7.5 (7.4–7.5) 10.2 (10.1–10.3) 17.6 (17.5–17.6) Sex Girls 349 (45.6 %) Boys 416 (54.4 %) Weight for age z scorea Underweight 255 (33.3 %) 296 (38.7 %) N/Ab N/Ab Not underweight 510 (66.7 %) 469 (61.3 %) N/Ab N/Ab Height for age z scorec Stunted 382 (49.9 %) 315 (41.2 %) 315 (41.2 %) 233 (30.5 %) Healthy height 383 (50.1 %) 450 (58.8 %) 450 (58.8 %) 532 (69.5 %) Body mass index z scored Thinness 48 (6.3 %) 77 (10.0 %) 99 (12.9 %) 68 (8.9 %) Healthy weight 640 (83.7 %) 650 (85.0 %) 622 (81.3 %) 639 (83.5 %) (At risk and) overweight/obese 77 (10.0 %) 38 (5.0 %) 44 (5.8 %) 58 (7.6 %) Mother’s education Middle school or more 200 (26.1 %) Primary school 510 (66.7 %) No education 55 (7.2 %) Household’s wealth index Poor 380 (49.7 %) 748 (97.8 %) 323 (42.2 %) 347 (45.4 %) Middle 101 (13.2 %) 11 (1.4 %) 161 (21.1 %) 156 (20.4 %) Rich 284 (37.1 %) 6 (0.8 %) 281 (36.7 %) 262 (34.2 %) Housing area Rural 402 (52.5 %) Urban 363 (47.5 %) Region Sumatra 167 (21.8 %) Java 429 (56.1 %) Bali and Nusa Tenggara Barat 107 (14.0 %) Kalimantan and Sulawesi 62 (8.1 %) Systolic blood pressure (mmHg) Normal N/A N/A N/A 495 (64.7 %) High N/A N/A N/A 270 (35.3 %) Diastolic blood pressure (mmHg) Normal N/A N/A N/A 543 (71.0 %) High N/A N/A N/A 222 (29.0 %) Data from Indonesian Family Life Survey, Indonesia, 1993–2007 a Based upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2, 3, 4 (WHO 2007) b N/A not available because weight for age z score calculations only available until age 10 years c Based upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2, 3, 4 (WHO 2007) d Based upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2, 3, 4 (WHO 2007)
123 PAGE | 146 158 C. N. Rachmi et al.
Table 2 14-year cohort analysis: 765 children aged 2.0–4.9 years in 1993 and their weight status (overweight/obesity and thinness) in 2007, according to height status in 1993 Stunteda in 1993 (wave 1) Overweight/obesity (n=382) 2007 (wave 4) Healthy heighta in 1993 2007 (wave 4) Overweight/obeseb (wave 1) (n = 383) Overweight/obeseb [% (95 % CI)] [% (95 % CI)]
Overall 8.6 (6.2–11.9) Overall 6.5 (4.5–9.5) Sex Sex Girls 5.4 (2.8–10.0) Girls 4.4 (2.2–8.6) Boys 11.2 (7.6–16.2) Boys 8.4 (5.3–13.1) Mothers’ education Mothers’ education Middle school or more 2.6 (0.6–9.8) Middle school or more 8.2 (4.4–14.6) Primary school 10.9 (7.6–15.2) Primary school 5.4 (3.1–9.0) No education 5.4 (1.3–19.9) No education 11.1 (2.7–36.3) Household wealth index Household wealth index Poor 8.2 (4.7–13.9) Poor 6.3 (3.4–11.3) Middle-class 8.0 (4.6–13.6) Middle-class 7.1 (3.9–12.3) Rich 10.6 (5.6–19.2) Rich 5.9 (2.2–14.8) Housing area Housing area Rural 10.4 (7.1–15.1) Rural 5.8 (3.1–10.5) Urban 5.9 (3.1–11.0) Urban 7.1 (4.3–11.5)
Stunteda in 1993 (wave 1) Thinness (n = 382) 2007 (wave 4) Healthy heighta in 1993 2007 (wave 4) Thinnessb (wave 1) (n = 383) Thinnessb [% (95 % CI)] [% (95 % CI)]
Overall 8.4 (6.0–11.6) Overall 9.4 (6.9–12.8) Sex Sex Girls 7.7 (4.5–12.9) Girls 9.4 (5.9–14.6) Boys 8.9 (5.7–13.5) Boys 9.4 (6.1–14.3) Mothers’ education Mothers’ education Middle school or more 9.0 (4.3–17.6) Middle school or more 9.0 (5.0–15.6) Primary school 8.6 (5.8–12.6) Primary school 9.1 (6.0–13.3) No education 5.4 (1.3–19.9) No education 16.7 (5.3–41.8) Household wealth index Household wealth index Poor 8.1 (4.6–14.0) Poor 8.1 (4.7–13.5) Middle-class 8.6 (5.1–14.3) Middle-class 8.4 (4.9–14.0) Rich 8.2 (3.9–16.4) Rich 14.7 (8.0–25.3) Housing area Housing area Rural 10.9 (7.4–15.6) Rural 8.7 (5.3–14.0) Urban 4.6 (2.2–9.4) Urban 10.0 (6.6–14.8) Data from Indonesian family life survey, Indonesia, 1993–2007 The prevalence [95 % confidence intervals (CI)] presented was unadjusted a Height for age z score below -2 standard deviation according to the WHO Growth Standards (WHO 2006) b BMI for age z score below -2 standard deviation according to the WHO Growth Reference (WHO 2007)
Stunted children and high blood pressure stunted or of healthy height at baseline. The odds ratio of stunted children having high systolic or diastolic Figure 3 shows there was no statistically significant blood pressure was not significant (OR 0.80; 95 % CI difference in the prevalence of high systolic/diastolic 0.59–1.07 and OR 0.84; 95 % CI 0.62–1.15, blood pressure in adolescence between those who were respectively).
123 PAGE | 147 Are stunted young Indonesian children more likely to be overweight, thin, or have high blood… 159
Table 3 Two cohort analyses of children age 2.0–4.9 years in baseline and their overweight status 7 years later: 1,083 children in 1993 and 1589 children in 2000 Stunteda in 1993 (wave 1) 7-years cohort 1 (n = 544) 2000 (wave 3) Healthy heighta in 1993 2000 (wave 3) Overweight/obeseb (wave 1) (n = 539) Overweight/obeseb [% (95 % CI)] [% (95 % CI)]
Overall stunted 1.7 (0.9–3.1) Overall 7.7 (5.6–10.4) Sex Sex Girls 1.0 (0.3–2.7) Girls 7.9 (5.0–12.4) Boys 2.3 (1.0–4.9) Boys 7.5 (4.9–11.3) Mothers’ education Mothers’ education Middle school or more 6.1 (2.2–15.9) Middle school or more 14.6 (9.7–21.3) Primary school 0.9 (0.4–2.2) Primary school 5.9 (3.7–9.2) No education 1.8 (0.2–12.1) No education 0.0 Household wealth index Household wealth index Poor 1.5 (0.6–3.7) Poor 9.9 (6.2–15.4) Middle-class 1.6 (0.6–4.2) Middle-class 5.4 (3.1–9.4) Rich 2.1 (0.5–8.4) Rich 8.4 (4.4–15.6) Housing area Housing area Rural 1.3 (0.5–3.1) Rural 3.7 (1.8–7.4) Urban 2.8 (1.2–6.6) Urban 13.7 (9.9–18.6) Stunteda in 2000 (wave 3) 7-years cohort 2 (n = 709) 2007 (wave 4) Healthy heighta in 2000 2007 (wave 4) Overweight/obeseb (wave 1) (n = 880) Overweight/obeseb [% (95 % CI)] [% (95 % CI)]
Overall Stunted 5.0 (3.4–7.3) Overall 16.5 (13.6–20.0) Sex Sex Girls 4.2 (2.5–6.9) Girls 10.9 (7.6–15.4) Boys 5.8 (3.4–9.7) Boys 21.8 (17.3–27.2) Mothers’ education Mothers’ education Middle school or more 2.9 (1.1–7.5) Middle school or more 21.5 (15.7–28.7) Primary school 5.5 (3.6–8.4) Primary school 14.9 (11.5–19.1) No education 6.2 (1.4–23.2) No education 2.3 (0.3–16.0) Household wealth index Household wealth index Poor 3.6 (2.0–6.7) Poor 15.7 (11.4–21.1) Middle-class 5.5 (3.0–9.8) Middle-class 20.2 (14.9–26.7) Rich 7.0 (3.2–14.7) Rich 11.8 (7.2–18.7) Housing area Housing area Rural 4.7 (2.9–7.6) Rural 14.3 (10.5–19.1) Urban 5.8 (3.1–10.3) Urban 19.7 (15.3–24.9)
Data from Indonesian Family Life Survey, Indonesia, 1993, 2000, and 2007 The prevalence [95 % confidence intervals (CI)] presented was unadjusted P \ 0.05 are presented in italic a Height for age z score below -2 Standard Deviation according to the WHO Growth Standards (WHO 2006) b BMI for age z score below -2 Standard Deviation according to the WHO Growth Reference (WHO 2007)
123 PAGE | 148 160 C. N. Rachmi et al.
The prevalence of overweight/obesity in stunted chil- dren, while relatively low, was significantly higher in the more recent 7-year cohort [5.0 % (3.4–7.3) compared to 1.7 % (0.9–3.1)], and even higher in the 14-year cohort [8.6 % (6.2–11.9)]. We explored the probability of stunted young children becoming overweight/obese seven and 14 years later. In contrast with our previous findings in four-repeated cross-sectional surveys within four different years, where stunted children were significantly more Fig. 2 Adjusted prevalence ratio (PR), 95 % confidence intervals likely to be overweight compared to their healthy heights (CI), and P values of stunted children aged 2.0–4.9 years at baseline to being overweight/obese 14 or 7 years later; data from the counterparts (Rachmi et al. 2016a), in this study, we found Indonesian family life survey, Indonesia, 1993–2007. The PRs, CIs, no significant prospective association between stunting in and P values are presented in the following order: wave 1 (1993) to 4 early childhood and overweight/obesity 14 years later. This (2007) [14 years], wave 1 (1993) to 3 (2000) [7 years], and wave 3 finding supports those of previous longitudinal studies in (2000) to 4 (2007) [7 years]. Independent variables adjusted for are sex, mother’s education, housing area (rural/urban), and household’s adolescents in South Africa (Cameron et al. 2005; Timaeus wealth index 2012), Senegal (Benefice et al. 2001), and Jamaica (Walker et al. 2001). We also found a significant inverse association Discussion of stunting in early childhood towards overweight/obesity in both 7-year cohorts. This is similar to the results of In the 14-year follow-up, just 8.6 % of young children who Gigante et al. (2007), although their study consisted only of were stunted were overweight/obese in adolescence. The adolescent males. One prospective cohort study found that prevalence of thinness in previously stunted children was also non-stunted girls (aged 9–24 months) in Jamaica were relatively low at 8.4 %. Stunting in young children was significantly more likely to be overweight in adolescence inversely associated with overweight/obesity 7 years later in (17–18 years), compared to their stunted counterparts, with both of the 7-year cohort analyses (P \ 0.05), but there was no no significant difference being found in the boys (Walker significant association in the 14-year follow-up. We did not et al. 2007). find any significant association between early childhood The association between stunting in early childhood and stunting and later thinness in adolescence. While a third of the later overweight/obesity was different between the two adolescents had high systolic and diastolic blood pressure, no 7-year cohorts and the 14-year cohort. We are uncertain of difference in prevalence was found between those who in the explanation for this, although a ‘puberty effect’ might early life were stunted, versus those who were of healthy be in operation. One longitudinal study in Senegalese girls height. There was no significant association found for stunted aged 11.4 years at baseline who were followed up to age children with high systolic or high diastolic blood pressure. 15.5 years showed that stunted girls continued to have a
Fig. 3 Height status of 765 Indonesian children age 2.0–4.9 years in wave 1 (1993) and the prevalence (95 % confidence intervals) of high systolic and diastolic blood pressure status in wave 4 (2007); data from Indonesian family life survey, Indonesia, 1993 and 2007. Height status was determined based upon the 2006 WHO Child Growth Standards (WHO 2006) for children \5 years. High systolic/diastolic blood pressure was determined based upon the 4th Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents (NHBPEP 2004)
123 PAGE | 149 Are stunted young Indonesian children more likely to be overweight, thin, or have high blood… 161 low height-for-age throughout puberty but had a catch-up difference in most of the baseline characteristics, apart in body mass (Benefice et al. 2001). Another study pro- from the prevalence of underweight in those who were in posed a ‘catch-up period’ in stunted girls during puberty, four waves of the study compared to those who were not. since stunted girls who were thinner compared to their Some authors have suggested that stunted young chil- healthy height counterpart in baseline have shown no sig- dren are at greater risk of overweight or obesity in later nificant difference in body mass by the age of 17–18 years years (Martins et al. 2004; Sawaya et al. 1998; Sawaya and (Walker et al. 2007). Roberts 2003). If this was the case, many low- and middle- About one-third of our main cohorts had high systolic income countries that currently have high rates of stunting and diastolic blood pressure in adolescence. This is com- would face very high prevalence rates of adult obesity, as parable to a study in 10,453 Chinese students aged well as obesity-related diseases, in future years. For 15-17 years, where the prevalence rates of relative high example, stunting prevalence is very high in countries, blood pressure (defined as systolic and/or diastolic blood such as Myanmar (35.1 %), Ethiopia (40.4 %), Pakistan pressure C95 % percentile, specific for age and sex) for the (45.0 %), Papua New Guinea (49.5 %), and Timor Leste 15-, 16-, and 17-year age groups were 29.4, 30.7, and (57.7 %) (IFPRI 2016). However, using BMI as an indi- 30.3 %, respectively (Ying-Xiu et al. 2015). Similarly, a cator of overweight/obesity, we found that, among 382 study involving 1203 Taiwanese junior-high school stu- children who were stunted at age 2.0–4.9 years, 83.0 % dents aged 12–14 years showed the prevalence of had a healthy BMI in adolescence, with 8.6 % of them hypertension (systolic blood pressure or diastolic blood being overweight/obese and 8.4 % being thin. While BMI pressure C95th percentile, specific for age, sex, and height) z scores is a valid measure of overweight/obesity, future was 29.7 % in boys and 20.7 % in girls (Lin et al. 2012). In research of overweight and obesity in early childhood our study, we found no difference in the prevalence of high should consider using other body composition parameters systolic/diastolic blood pressure in adolescence between to detect the possibility of central obesity, with follow-up those who were stunted and those who were of healthy into young and mid-adulthood. height as young children. We did not find any association Chronic undernutrition and overnutrition in Indonesia between stunting and hypertension in adolescents. There is and other Asian countries are of great importance. In those little evidence regarding the prospective association countries where undernutrition remains a major health between childhood stunting and high blood pressure in concern in early childhood, the development of policies adolescence, irrespective of weight status. One longitudinal and programs should combine the prevention and man- study in Jamaica found that stunting at 9–24 months was agement of both under and overnutrition, and implement significantly associated with higher systolic blood pressure close monitoring of growth starting as early as possible. at age 7–8 years, although they compared mean blood pressure and did not use cut-off points for high blood Acknowledgments CNR would like to thank Lembaga Pengelola pressure, as in our analysis (Gaskin et al. 2000). Dana Pendidikan (LPDP), the Republic of Indonesia for her Ph.D. scholarship. This is the first Asian study to show the prospective associations between stunting in early life and later over- Compliance with ethical standards weight/obesity, thinness, and high blood pressure. To ensure we identified stunted young children, we started at Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of age two years, by which age the process of stunting is the institutional and/or national research committee and with the 1964 generally already established (Victora et al. 2010). Helsinki declaration and its later amendments or comparable ethical Strengths of the study include the relatively long-term standards. Ethics approval for IFLS was provided in the United States follow-up (14 years) of the main cohort; the high recontact of America (Institutional Review Board at Rand Corporation, Santa Monica, California) and Indonesia (Ethics Committees of the rate and the fact that trained professionals collected both Universitas Indonesia and the Universitas Gadjah Mada). questionnaire and anthropometric data, reducing potential measurement errors. The nature of the study (i.e., sec- Informed consent Informed consent was obtained from all individ- ondary data analysis) meant we were limited to performing ual participants included in the study. analyses on the existing data. Other potential factors that might indicate an association between early childhood stunting and adolescent weight status—such as fat mass, References lean mass, waist circumference—were not available in the data set. A further potential limitation is the reduced Badan Penelitian dan Pengembangan Kesehatan (BPPK) Kementrian Kesehatan Republik Indonesia (2014) Status Gizi. In: Indonesia number of participants who completed all four waves of the DK (ed) Laporan Hasil Riset Kesehatan Dasar Indonesia tahun survey; however, comparisons between those who com- 2013, Riskesdas Dalam Angka. CV Kiat Nusa, Jakarta, pleted the survey and those who did not showed no pp 386–415
123 PAGE | 150 162 C. N. Rachmi et al.
Benefice E, Garnier D, Simondon KB, Malina RM (2001) Relation- pressure in children and adolescents. Pediatrics 114(2 Suppl 4th ship between stunting in infancy and growth and fat distribution Report):555–576 during adolescence in Senegalese girls. Eur J Clin Nutr Popkin BM, Richards MK, Montiero CA (1996) Stunting is associ- 55(1):50–58 ated with overweight in children of four nations that are Bove I, Miranda T, Campoy C, Uauy R, Napol M (2012) Stunting, undergoing the nutrition transition. J Nutr 126(12):3009–3016 overweight and child development impairment go hand in hand Rachmi CN, Agho KE, Li M, Baur LA (2016a) Stunting, coexisting as key problems of early infancy: Uruguayan case. Early Hum with overweight in 2.0–4.9 year old Indonesian children: Dev 88(9):747–751 prevalence, trends and associated risk factors from four cross Cameron N, Wright MM, Griffiths PL, Norris SA, Pettifor JM (2005) sectional surveys. Public Health Nutr. Available on CJO2016. Stunting at 2 years in relation to body composition at 9 years in doi:10.1017/S1368980016000926 African urban children. Obesity Res 13(1):131–136 Rachmi CN, Agho KE, Li M, Baur LA (2016b) Stunting, underweight Clemente APG, Santos CD, Silva AAB, Martins VJ, Marchesano AC and overweight in children aged 2.0–4.9 years in Indonesia: et al (2012) Mild stunting is associated with higher blood prevalence trends and associated risk factors. PLoSOne pressure in overweight adolescents. Arq Bras Cardiol 98:6–12 11(5):e0154756. doi:10.1371/journal.pone.0154756 de Onis M, Blossner M, Borghi E (2010) Global prevalence and Sawaya A, Roberts S (2003) Stunting and future risk of obesity: trends of overweight and obesity among preschool children. Am principal physiological mechanisms. Cad Saude Publ 19(Sup. J Clin Nutr 92:1257–1264 1):S21–S28 de Onis M, Dewey K, Borghi E, Onyango A, Blo¨ssner M et al (2013) Sawaya A, Grillo L, Verreschi I, da Silva A, Roberts S (1998) Mild The World Health Organization’s global target for reducing stunting is associated with higher susceptibility to the effects of childhood stunting by 2025: rationale and proposed actions. high fat diets: studies in a shantytown population in Sa˜o Paulo, Matern Child Nutr 9:6–26 Brazil. J Nutr 128(2 supplement):415–420 Filmer D, Pritchett LH (2001) Estimating wealth effects without Schroeder DG, Martorell R, Flores R (1999) Infant and child growth expenditure data—or tears: an application to educational enroll- and fatness and fat distribution in Guatemalan adults. Am J ments in states of India. Demography 38(1):115–132 Epidemiol 149(2):177–185 Frankenberg E, Thomas D (2000) The Indonesia family life survey Serrato C, Melnick G (1995) The Indonesian family life survey (IFLS): study design and results from waves 1 and 2. RAND, overview and descriptive analysis. RAND, Santa Monica Santa Monica Stata Corp (2014) STATA statistical software: Release 13. Stata Corp Frankenberg E, Karoly LA, Gertler P, Peterson CE, Wesley D (1995) LP, College station The 1993 Indonesian family life survey: overview and field Timaeus IM (2012) Stunting and obesity in childhood: a reassessment report. RAND, Santa Monica using longitudinal data from South Africa. Int J Epidemiol Gaskin P, Walker SP, Forrester TE, Grantham-McGregor S (2000) 41(3):764–772. doi:10.1093/ije/dys026 Early linear growth retardation and later blood pressure. Eur J Victora CG, de Onis M, Hallal PC, Blossner M, Shrimpton R (2010) Clin Nutr 54:563–567 Worldwide timing of growth faltering: revisiting implications for Gigante DP, Victora CG, Horta BL, Lima RC (2007) Undernutrition interventions. Pediatrics 125(3):e473–e480. doi:10.1542/peds. in early life and body composition of adolescent males from a 2009-1519 birth cohort study. Br J Nutr 97(5):949–954 Walker SP, Gaskin P, Powell CA, Bennett FI (2001) The effects of Indonesia Family Life Survey (IFLS). http://www.rand.org/labor/ birth weight and postnatal linear growth retardation on body FLS/IFLS.html. Accessed May 2014 mass index, fatness and fat distribution in mid and late International Food Policy Research Institute (2015) Global nutrition childhood. Public Health Nutr 5(3):391–396 report 2015: actions and accountability to advance nutrition and Walker SP, Chang SM, Powell CA (2007) The association between sustainable development. Washington, DC early childhood stunting and weight status in late adolescence. International Food Policy Research Institute (2016) Global nutrition Int J Obes (Lond) 31:347–352 report 2016: from promise to impact: ending malnutrition by World Health Organization (2006) The WHO child growth standards. 2030. Washington, DC In: World Health Organization. http://www.who.int/childgrowth/ Lin FH, Chu NF, Hsieh AT (2012) The trend of hypertension and its standards/en/. Accessed 16 March 2014 relationship to the weight status among Taiwanese young World Health Organization (2007) WHO growth reference for school- adolescents. J Hum Hypertens 26(1):48–55 aged children and adolescents. In: WHO. http://www.who.int/ LMS Growth Program 2010 (2010) http://www.healthforallchildren. growthref/en/. Accessed March 2015 com/?product=lmsgrowth. Accessed June 2014 World Health Organization (2011) WHO Anthro version 3.2.2 and Martins P, Hoffman D, Fernandes M, Nascimento C, Roberts S et al macros. In: World Health Organization. http://www.who.int/ (2004) Stunted children gain less lean body mass and more fat childgrowth/software/en/. Accessed 15 Sept 2014 mass than their non-stunted counterparts: a prospective study. Br Ying-Xiu Z, Gui-Zhi S, Jin-Shan Z, Ming L, Zun-Hua C (2015) J Nutr 92:819–825 Monitoring of blood pressure among children and adolescents in National High Blood Pressure Education Program (NHBPEP) Work- a coastal province in China: results of a 2010 survey. Asia- ing Group on Children and Adolescents (2004) The fourth Pacific J Public Health/Asia-Pacific Acad Cons Public Health report on the diagnosis, evaluation, and treatment of high blood 27(2):NP1529–36. doi:10.1177/1010539512444777
123 PAGE | 151
ELECTRONIC SUPPLEMENTAL MATERIAL International Journal of Public Health Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? Cut Novianti Rachmi, Kingsley Emwinyore Agho, Mu Li, Louise Alison Baur
Electronic supplemental material Table 1. Comparison of baseline demographics between children aged 2.0-4.9 years who completed 4 waves of the survey 1993-2007 (n=765) and those who only completed the first wave in 1993 (n=620); Data from Indonesian Family Life Survey, Indonesia, 1993-2007 Completed 4 Available only P value waves (765) in wave 1 (620) Age (years) [mean (95% CI)] Girls 3·4 (3·3-3·5) 3·5 (3·4-3·6) 1.000 Boys 3·4 (3·4-3·5) 3·5 (3·4-3·6) Sex Girls 349 (45·6%) 301 (48.5%) 0.280 Boys 416 (54·4%) 319 (51.5%) Weight for Age Z-scorea Underweight 255 (33·3%) 296 (34.3%) 0.001 Not underweight 510 (66·7%) 407 (65·7%) Height for Age Z-scoreb Stunted 382 (49·9%) 312 (50·3%) 0.914 Healthy height 383 (50·1%) 308 (49·7%) Body Mass Index Z-scorec Thinness 48 (6·3%) 51 (8·2%) 0.297 Healthy weight 640 (83·7%) 501 (80·8%) (At risk &) overweight/obese 77 (10·0%) 68 (11·0%) Mother's education Middle school or more 200 (26·1%) 146 (23.5%) 0.073 Primary school 510 (66·7%) 409 (66.0%) No education 55 (7·2%) 65 (10.5%) Household's wealth index Poor 380 (49·7%) 302 (48.7%) 0.890 Middle 101 (13·2%) 80 (13.3%) Rich 284 (37·1%) 238 (38.4%) Housing Area Rural 402 (52·5%) 323 (52.1%) 0.871 Urban 363 (47·5%) 297 (47.9%) Region Sumatra 167 (21·8%) 140 (22.6%) 0.565 Java 429 (56·1%) 355 (57.3%) Bali & Nusa Tenggara Barat 107 (14·0%) 71 (11.4%) Kalimantan & Sulawesi 62 (8·1%) 54 (8.7%) aBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) bN/A: Not Available because Weight for Age Z-score calculations only available until age 10 years cBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) dBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) P value calculated with t-test and chi square test
PAGE| 152
Electronic supplemental material Table 2. Characteristics of 1083 children aged 2·0-4·9 years in wave 1 (1993) followed up to wave 3 (2000) [7 years cohort-1]; Prevalence (n [%]) and mean (95% Confidence Intervals [CI]). Data from Indonesian Family Life Survey, Indonesia, 1993-2000 Wave 1 (1993) Wave 3 (2000) Age (years) [mean (95% CI)] Girls 3·4 (3·3-3·5) 10·2 (10·1-10·3) Boys 3·4 (3·4-3·5) 10·2 (10·1-10·3) Sex Girls 492 (45·4%) Boys 591 (54·6%) Weight for Age Z-scorea Underweight 360 (33·2%) N/Ab Not underweight 723 (66·8%) N/Ab Height for Age Z-scorec Stunted 544 (50·2%) 452 (41·7%) Healthy height 539 (49·8%) 631 (58·3%) Body Mass Index Z-scored Thinness 72 (6·6%) 135 (12·5%) Healthy weight 900 (83·1%) 887 (81·9%) (At risk &) overweight/obese 111 (10·3%) 61 (5·6%) Mother's education Middle school or more 263 (24·3%) Primary school 749 (69·2%) No education 71 (6·5%) Household's wealth index Poor 546 (50·4%) 451 (41·6%) Middle 142 (13·1%) 229 (21·1%) Rich 395 (36·5%) 403 (37·3%) Housing Area Rural 587 (54·2%) Urban 496 (45·8%) Region Sumatra 270 (24·9%) Java 590 (54·5%) Bali & Nusa Tenggara Barat 135 (12·5%) Kalimantan & Sulawesi 88 (8·1%) aBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) bN/A: Not Available because Weight for Age Z-score calculations only available until age 10 years cBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) dBased upon the WHO Growth Standards for wave 1 ( WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007)
PAGE| 153
Electronic supplemental material Table 3. Characteristics of 1589 children aged 2·0-4·9 years in wave 3 (2000) followed up to wave 4 (2007) [7 years cohort-2] of the Indonesian Family Life Survey, Indonesia, 2000-2007; Prevalence (n [%]) and mean (95% Confidence Intervals [CI]). Data from Indonesian Family Life Survey, Indonesia, 2000-2007 Wave 3 (2000) Wave 4 (2007) Age (years) [mean (95% CI)] Girls 3·5 (3·4-3·6) 10·9 (10·9-11·0) Boys 3·5 (3·4-3·5) 10·9 (10·8-11·0) Sex Girls 792 (49·8%) Boys 797 (50·2%) Weight for Age Z-scorea Underweight 422 (26·6%) N/Ab Not underweight 1167 (73·4%) N/Ab Height for Age Z-scorec Stunted 709 (44·6%) 555 (34·9%) Healthy height 880 (55·4%) 1034 (65·1%) Body Mass Index Z-scored Thinness 105 (6·6%) 210 (13·2%) Healthy weight 1299 (81·8%) 1201 (75·6%) (At risk &) overweight/obese 185 (11·6%) 178 (11·2%) Mother's education Middle school or more 356 (22·4%) Primary school 1180 (74·3%) No education 53 (3·3%) Household's wealth index Poor 718 (45·2%) 706 (44·4%) Middle 313 (19·7%) 317 (20·0%) Rich 558 (35·1%) 566 (35·6%) Housing Area Rural 905 (57·0%) Urban 684 (43·0%) Region Sumatra 344 (21·7%) Java 895 (56·3%) Bali & Nusa Tenggara Barat 189 (11·9%) Kalimantan & Sulawesi 161 (10·1%) aBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) bN/A: Not Available because Weight for Age Z-score calculations only available until age 10 years cBased upon the WHO Growth Standards for wave 1 (WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007) dBased upon the WHO Growth Standards for wave 1 ( WHO 2006) and WHO Growth Reference for waves 2,3,4 (WHO 2007)
PAGE| 154
Chapter 7: Qualitative studies
7.1 OVERVIEW AND AIMS 156 7.2 METHODS 7.2.1 Location and number of the focus group discussion 156 7.2.2 Ethics approval 157 7.2.3 Participants 157 7.2.4 Data analysis approaches 157 7.3 REFERENCES 158
PAGE| 155 Chapter 7: Qualitative studies
7.1 OVERVIEW AND AIMS This chapter provides an introduction to the qualitative studies reported in the following two chapters. The results of the quantitative analysis results involving children aged 2.0-4.9 years have shown a significant increase in the prevalence of combined at risk of overweight, overweight and obesity: from 10.3% in 1993 to 16.5% in 2007 (see Chapter 4)1. The risk factors associated with overweight/obesity were being in the youngest age group (2.0-2.9 years), being male, having parents who were overweight/obese and having a father with university education1.
These results raise the question of what are the most effective strategies to tackle the issue of childhood obesity. In order to address this question and to gain a better understanding of the role families, particularly the primary carers, play, I undertook a qualitative study in Bandung, Indonesia of mothers/grandmothers of preschool- and primary school-aged children. The qualitative study aimed to explore the perceptions of Indonesian mothers and grandmothers towards child obesity and related issues. Such information would provide valuable insights into how obesity is perceived by mothers/grandmothers of young children, and would also reveal potential strategies for tackling childhood obesity in Indonesia. The complete results of the qualitative studies are presented in Chapters 8 and 9 of this thesis.
7.2 METHODS
7.2.1 Location and number of focus group discussion We randomly selected 3 areas from the Greater Bandung Area (Bandung City, Bandung Regency, and Bandung Barat Regency) for this study. A number was assigned to each city/ regency and we chose the city according to the last digit in a table of random numbers. These three areas have different Human Development Index (HDI) scores of high, medium, and low2. The HDI scores acted as our compass towards different levels of socio-economic status (SES). Cisarua, Margahayu, and Sukasari districts represent the low, intermediate, and high SES areas, respectively. We conducted four focus group discussions (FGDs; two with mothers of pre- school aged children and two with mothers of primary-school-aged children) in each area, resulting in a total of 12 FGDs.
PAGE| 156 7.2.2 Ethics approval Ethics approval was obtained from the Ethics Committee at Fakultas Kedokteran Universitas Padjadjaran, Bandung (Appendix 1). According to the new guidelines from the University of Sydney Human Research Ethics Committee (HREC), once the in-country ethics clearance was obtained we were only required to send a copy of the ethics clearance approval letter to the University of Sydney HREC without further application.
7.2.3 Participants We conducted 12 FGDs with 6-10 participants in each group. There were 94 participants in total. The inclusion criteria for the participants were mothers/grandmothers of children under five (balita) or children aged 7-12 years old (primary school aged children) who hold or share the primary role in decision making for the care of the child or who are responsible, in part or in whole, for feeding the child, willing to spare 120 minutes of their time for the study (30 minutes for introduction session and 90 minutes for discussion), and willing to give consent to the researcher for recording the discussion, with special attention given to anonymity in the transcript. The exclusion criteria were mothers/grandmothers who only able to attend one session, either introduction or discussion. Prior to the FGD, the researcher and research assistant had a 30-minute meeting with the participants, to build rapport and explain the general aim of the research and what we will do during the FGD. The complete process of recruitment of participants is available in each published or submitted paper (Chapters 8 and 9).
7.2.4 Data analysis approaches The process of transcription started immediately after the first FGD to ensure that we could do corrections in the transcripts and explore more issues in subsequent FGDs in a timely manner. We anticipated conducting 6-9 FGDs. However, in reality data saturation was not reached by then. After discussion within the research team we continued to 12 groups. All data were analysed using NVIVO version 113.
We used the grounded theory approach as the basis of our analyses4-7. This approach is characterised by a data-driven process. In summary, concepts that emerged from our data were combined to create theories. These data were obtained through our data collection processes, observations, and interactions between our participants4-7.
A step by step approach to our data analysis is also available in each of the two chapters.
PAGE| 157 7.3 REFERENCES
1. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, Underweight and Overweight in Children Aged 2.0-4.9 Years in Indonesia: Prevalence Trends and Associated Risk Factors. PloS one. 2016; 11:e0154756. 2. Indeks Pembangunan Manusia [https://jabar.bps.go.id/Subjek/view/id/26 - subjekViewTab3|accordion-daftar-subjek1] 3. NVivo qualitative data analysis Software. vol. Version 11: QSR International Pty Ltd.; 2014. 4. Strauss A: Qualitative analysis for social scientists. Cambridge: Cambridge University Press; 1987. 5. Strauss A, Corbin J: Basics of qualitative research: grounded theory procedures and techniques. Newbury Park, CA: Sage; 1990. 6. Glaser B, Strauss A: The discovery of grounded theory: strategies for qualitative research. Chicago: Aldine; 1967. 7. Green J, Thorogood N: Qualitative Methods for Health Research. Second edition. Great Britain: SAGE Publications Ltd; 2009.
PAGE| 158
Chapter 8: Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study
BACKGROUND 161 METHODS Study design and location 161 Participants and recruitment process 161 Research team, data collection and approach to analysis 161 Credibility and trustworthiness 162 RESULTS Characteristics of participants 164 Emerging categories 164 Concept of overweight 164 How carers define overweight 164 Normal birthweight range 164 “Chubbier is healthier” 164 Ideal body 164 Factors contributing to overweight 165 Dietary factors 165 Activity levels and sedentary behaviours in children 166 Heredity 167 Awareness and feelings towards overweight in children 168 Characteristics of overweight 168 Food portions in overweight children 168 Eating frequency in overweight children 168 Feelings about overweight children 168 Sensitivity in carers of overweight children 168 Children’s current weight and height 168 DISCUSSION 170 CONCLUSIONS 171 RECOMMENDATIONS 171 REFERENCES 172
This chapter has been previously published: Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. International Journal of Behavioral Nutrition and Physical Activity. 2017; 14(101); DOI 10.1186/s12966-017-0556-1.
PAGE| 159 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 DOI 10.1186/s12966-017-0556-1
RESEARCH Open Access Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study Cut Novianti Rachmi1* , Cynthia Louise Hunter2,3, Mu Li3 and Louise Alison Baur1,3
Abstract Background: The prevalence of childhood overweight has increased in the past two decades in Indonesia. Even though prevalence is escalating, there is a lack of qualitative evidence to assist in the design and implementation of strategies to tackle this issue. This study aimed to explore the view of primary carers (mothers and grandmothers) from different socio-economic-status groups, on childhood overweight in the Greater Bandung Area, Indonesia. Methods: We conducted 12 focus groups discussions with a total of 94 carers of under-five and 7–12 years children, from June to October 2016. We used the grounded theory approach in our analysis. Results: Three main categories emerged: the concept of overweight, factors contributing to overweight, and awareness and feelings towards overweight children. Most carers from all SES groups defined overweight subjectively, while a few from the low SES group defined it objectively. Most carers from low and high SES groups agreed with the concept “chubbier is healthier”. All carers had some knowledge of the main factors that contribute to childhood overweight: dietary factors, activity levels and sedentary behavior, and hereditary factors. Carers from all SES groups described similar characteristics of overweight; carers from low and intermediate SES groups had mixed feelings while all high SES carers have negative feelings about overweight children, mostly related to stigma. However, carers who identified their own children as being overweight expressed sensitivity about this weight status, especially their physical abilities. Almost all carers knew their children’s current weight while less than two thirds knew their children’s height. Conclusions: There are several policy implications. Firstly, health-related knowledge of the primary carers is of great importance and needs augmenting. To increase that knowledge, there is a role for front-line health practitioners (doctors/midwives/nurses) to be more active in educating the community. Secondly, simpler and more effective ways to disseminate healthy lifestyle messages to carers is required. Thirdly, by placing more emphasis on carers monitoring their children’s growth may encourage carers to take steps to keep their children in the healthy weight and height ranges. Fourthly, the Department of Education may need to improve the quality and quantity of physical activity in schools. Keywords: Overweight, Children, Perceptions,Mothers,Indonesia,Qualitative
* Correspondence: [email protected]; [email protected] 1Discipline of Child & Adolescent Health, The Children’s Hospital at Westmead, University of Sydney Clinical School, Locked Bag 4001, Westmead, NSW 2145, Australia Full list of author information is available at the end of the article
© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. PAGE | 160 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 2 of 13
Background intermediate, and high socio-economic status areas, Overweight/ obesity in both children and adults, while respectively. initially an issue mainly in western countries, is now a The rationale for selecting areas of different SES was major emerging health problem in countries undergoing to obtain a diversity of views of mothers/grandmothers economic transition – including Indonesia [1, 2]. towards children’s growth, development and overweight Indonesia, the world’s fourth most populous country, and to observe whether SES is a contributing factor to has been dealing with the problem of children’s un- overweight. Hereafter, we will refer to mothers/grand- dernutrition, including stunting and underweight for mothers as carers. decades, as a consequence of which many nutrition po- licies and programs were focused on addressing these is- Participants and recruitment process sues [3]. Overweight, on the other hand, has only been Twelve focus group discussions (FGDs) were conducted, recognised as an Indonesian health problem for little six with carers of under five children and six with carers more than a decade [2, 3]. of elementary-school-aged children (7–12 years),1 from The prevalence of childhood overweight in Indonesia June to October 2016 (Fig. 1). Inclusion criteria were has significantly increased in the last 20 years, with carers of children who hold or share the primary role in higher prevalence rates found among boys in general decision-making for the care of the child or were re- and boys and girls living in urban areas [2]. To our sponsible, in part or in whole, for child feeding and ea- knowledge, qualitative evidence about overweight in ting, including food shopping and budgeting. Indonesia is scarce, with the available evidence mainly To recruit participants from low and intermediate focused on mothers with overweight children, [4] and SES areas, we contacted Community Health Centres none considering the views of people from different (Puskesmas) in the chosen areas to reach carers of socio-economic groups. under five children. We collaborated with local mid- Parents, particularly mothers, hold an important role wives and health kader (mothers who are trained and in child feeding, role modelling for healthy eating and work for the Puskesmas and Community Health Posts active living, and the prevention of overweight [5–11]. [Posyandu] for a small incentive). These kader have Many potential strategies in tackling overweight will in- knowledge of the people living in the communities [13]. volve working with families, especially the primary carers We provided a verbal explanation of the project, and of young children. These strategies will benefit from an disseminated the participant information statement to understanding of the knowledge, attitudes, and beha- invite participation. To recruit carers of 7–12 years viours of the primary carers. Thus, this study aimed to children, we asked for assistance from class coordina- explore the perceptions of Indonesian mothers and tors in public elementary schools (a mother from each grandmothers from different socio-economic groups to- grade in elementary school who acts as the information wards child overweight/ obesity and related issues. In source for fellow mothers). We also placed project infor- the Indonesian language, people refer to overweight/ mation including contact details in students’ communi- obesity with the same word, kegemukan, except those in cation books (book that teachers use to communicate the health sector/ academics. Hereafter, we will refer to with mothers of individual children). overweight/ obesity as overweight. Recruitment of participants from high SES groups was by invitation to local private schools and paediatricians’ offices. In Indonesia, children from socially advantaged Methods families do not go to Puskesmas, but to private medical Study design and location practices or specialists, and they attend private schools. We undertook a qualitative study in Bandung, West Java, We used the same approach in each school (i.e. contac- Indonesia. West Java Province consists of 27 cities/regen- ting the class coordinator), and asked the nurse in the cies and we selected three cities/regencies (Bandung City, paediatrician’s office to assist us in providing information Bandung Regency, and Bandung Barat Regency) based on about and contact details for the study to the carers. high population numbers and mixed ethnic groups, to Interested carers then directly contacted the research explore the situation in Indonesia [12]. Every district in team who gave further explanation about the discussion each city/regency was numbered and we chose the se- process and took details. Carers who gave verbal consent lected areas based on the last digit from a table of random to participate were later approached to set the date and numbers. These three areas have different Human Devel- time for the FGD. opment Index scores (high, medium, and low); these were used as a measure of different levels of socioeconomic Research team, data collection and approach to analysis status [12]. The locations of the selected areas (Cisarua, The field research team consisted of the first author Margahayu, and Sukasari districts) represent the low, (CNR) and two research assistants (DMP and TMF). They
PAGE | 161 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 3 of 13
Fig. 1 Numbers and participants of focus group discussions in Bandung, Indonesia. SES: socio economic status; Low SES: FGDs 5, 6, 7, and 8; Intermediate SES: FGDs 3, 4, 7, and 8; High SES: FGDs 1, 2, 11, and 12 underwent a three-day training program in qualitative re- terms that have no exact English translation were kept search, concentrating on FGDs, with CLH in Bandung. and explained in English inside parentheses next to Question guidelines were developed in English and fina- those terms. lised during the training. Afterwards, the guidelines were We used the grounded theory approach in our analysis translated into the Indonesian and Sundanese languages [14–16]. All transcripts were analysed using NVivo for and a pilot FGD was conducted to test the questions Mac version 11.0.0 [17]. The research assistants were guidelines (Table 1). This process included a discussion of not involved in the analysis phase. The analysis was con- potential issues and possible resolutions that might be ducted following the steps: 1) CNR and CLH read all encountered during the FGDs. transcripts several times to familiarise themselves with CNR acted as a moderator in all of the FGDs, with two the data. 2) CNR undertook the open coding and com- research assistants (DMP and TMF) observing and ma- pared the coding within and between transcripts. 3) naging the recordings and time keeping. Prior to each CNR and CLH identified categories and sub-categories FGD, the study team explained the FGD process to the of ideas/concepts. 4) Once the final version of categories participants and asked them to fill in the basic demo- and subcategories was agreed upon, a final combination graphic information form and provide written informed of subcategories took place. 5) CNR and CLH identified consent. Participants were offered the language of their similarities and differences between and within the dif- choice to be used during the FGDs, either Indonesian or ferent SES groups. 6) The quotations that best repre- Sundanese (the language of West Java), in which all three sented each category and subcategory were identified. moderators were fluent. Each FGD began with a 30-min As part of the analysis, CNR quantified comments ice-breaker, to build rapport between participants and re- throughout the transcripts looking at responses from searchers. We highlighted how much everyone’s opinion individual carers. Quotations representing each SES was valued and encouraged the participants to speak grouping were selected from multiple FGD groups. 7) freely. Similarly, and to ensure confidentiality, we did not All authors (CNR, CLH, ML, and LAB) agreed upon the videotape the discussion. To ensure consistency of the dis- final version of the categories and subcategories. LAB cussion the moderator used the same guideline, including and ML, experts in childhood obesity and childhood nu- a set of prompt questions. All discussions were digitally trition, were significantly involved in the design phase of recorded, and lasted 50–90 min. Group size ranged from the research and writing up of the manuscript. 6 to 9 people. Participants each received 11 USD for participation. Credibility and trustworthiness Transcription and translation started immediately after Credibility and trustworthiness were addressed in several the first FGD, giving the team some time to reflect on ways. One researcher (CNR) was the moderator of all 12 the discussion results. Transcription was done only in FGDs and could, therefore, describe the situation and Indonesian by DMP and TMF and checked by CNR to convey the opinions stated directly (words) or indirectly confirm that the transcripts were verbatim to the re- (intonation of voice or facial expressions) consistently. cordings. The research assistants then translated the We reduced interpretive bias by involving another ana- transcripts into English; both Indonesian and English lyst (CLH) who was not present during the data col- transcripts were checked by CLH and CNR to ensure lection. CNR also conducted triangulation with the two the quality of the translation. Sundanese or Indonesian research assistants and used the notes made during the
PAGE | 162 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 4 of 13
Table 1 Questions used in focus group discussions in Greater Bandung Area, Indonesia No Questions (English) Indonesian Sundanese 1 There is quite a lot of talk around these days Akhir-akhir ini banyak pembicaraan tentang Jaman ayeuna seuer pisan pacariosan tentang about being obese or being overweight in terlalu gemuk pada anak dan dewasa, bayi, barudak sareng dewasa anu parawakanna both adults and in children and babies - can termasuk bayi, menurut ibu, apa arti terlalu ageung teuing, menurut ibu, kumaha artina ari you tell me what you think this means? When I gemuk? Ketika saya mengatakan terlalu gemuk, parawakanna ageung teuing teh? Upami abdi say overweight/obesity, what is the first thing apa hal pertama yang ibu pikirkan? Menurut nyarios parawakan anu ageung teuing, naon that comes in your mind? In your opinion, ibu, artis yang berat badannnya ideal siapa? hal pertama anu ibu pikirkeun? Menurut ibu, which celebrities have the ideal body? saha artis anu parawakanna ideal? 2 In your opinion, when can you say a baby/ Menurut ibu, kapan seorang bayi/anak bisa Menurut ibu, iraha bayi/barudak teh disebat child is overweight/obese? dibilang terlalu gemuk? parawakanna ageung teuing? Prompts: e.g. they have multiple chins, they Bantuan: Bantuan: wear clothes several sizes bigger than their • Mukanya (dagu berlipat) • Rarayna (dagu berlipat) age, they have trouble learning how to walk/ • Ukuran baju • Ukuran bajuna run, they eat adult size portion of food • Terlambat belajar berjalan/lari (bayi) • Telat diajar jalan/lumpat(bayi) • Makan banyak seperti orang dewasa • Tuangna siga jalmi dewasa 3 What do you think of the saying ‘chubbier is Apa pendapat ibu tentang ungkapan “lebih Kumaha pendapat ibu tentang paribasa healthier’? What do you think about the gemuk lebih sehat”? Bagaimana pendapat ibu “langkung ageung langkung sehat”? Kumaha relationship between body weight and health? tentang hubungan berat badan dengan pendapat ibu tentang hubungan abot badan kesehatan? sareng kasihatan? 4 What factors do you think contribute/ Menurut ibu, hal apa saja yang mempengaruhi Menurut ibu, naon wae anu ngaruh kana berat influence to a baby/ child’s weight? berat badan bayi/anak? badan bayi/anak? Prompts: Is it dietary? Bantuan: Apakah makanan? Bantuan: Apakah makanan? Mention the following only if they are not Sebut hal berikut hanya bila hal-hal ini tidak Sebut hal berikut hanya bila hal-hal ini tidak mentioned: genetic, family eating patterns, disinggung: turunan/ bawaan (genetik), pola disinggung: turunan/ bawaan (genetik), pola physical activity, screen time (length of time makan keluarga, kegiatan fisik, waktu layar makan keluarga, kegiatan fisik, waktu layar spends in front of a screen – TV, computers, (waktu yang dihabiskan di depan layar - TV, (waktu yang dihabiskan di depan layar - TV, iPad, video games, reading, phone/games, komputer, iPad, permainan video, membaca, komputer, iPad, permainan video, membaca, playing cards, board games, etc), sleep, other telefon/ permainan, bermain kartu, permainan telefon/ permainan, bermain kartu, permainan activities (soccer, kite flying, hide and seek, hop lainnya), tidur, kegiatan lainnya (main bola, lainnya), tidur, kegiatan lainnya (main bola, scotch, jump rope, other outside activities: layangan, petak umpet, engkle, lompat tali, layangan, petak umpet, engkle, lompat tali, buying something at store)? kegiatan lain di luar rumah: belanja ke kegiatan lain di luar rumah: belanja ke In your knowledge, what do your children warung)? warung)? usually have as snacks? Sepengetahuan ibu, apa yang dimakan anak Sapengetahuan ibu, naon wae cemilan anu Prompts: ibu sebagai cemilan? dituang ku anak ibu? • Soda Bantuan: Bantuan: • Instant noodles • Soda • Soda How do you feel about your children’s snacks • Mie instan • Mie instan choices? Bagaimana perasaan ibu tentang pilihan Kumaha perasaan ibu tentang pilihan cemilan cemilan anak ibu? anak ibu? 5 How much do you expect your baby to weigh Menurut ibu berapa berat badan bayi yang Menurut ibu, sabaraha abot bayi anu at birth? About how much do you think a diharapkan saat lahir? Menurut ibu berapa diharapkeun pas lahir? Menurut ibu, sabaraha healthy baby should weigh by 6 months and seharusnya berat badan bayi saat usia 6 bulan kedahna abot bayi pas usia 6 sasih sareng by age 12 months? dan 1 tahun? sataun? 6 Can you tell me how much your child weighs Apakah ibu tahu berapa berat badan dan Kinten-kintena ibu terang teu sabaraha abot and his/her height now? tinggi badan anak ibu sekarang? sareng tinggi anak ibu ayeuna? 7 Are you satisfied with your child’s growth up Apakah ibu puas dengan pertumbuhan anak Apa ibu puas sareng pertumbuhan anak ibu to this point? If there is anything you would ibu sampai saat ini? Kalau ada hal yang ingin dugika ayeuna? Upaya aya hal anu bade like to change about your child’s weight, what diubah mengenai berat badan anak ibu, diubah soal berat badan anak ibu, naon anu would it be? kira-kira hal apa? bade diubah? 8 Who do you think plays an important role in Ibu, kalau di rumah siapa yang berperan dalam Ibu, upami di bumi saha anu berperan dina influencing your child’s weight in the house? pertumbuhan anak ibu? pertumbuhan anak ibu? Prompts: Bantuan: Bantuan: • Who is responsible in determining the kind of • Siapa yang menentukan jenis makanan anak? • Saha anu nangtoskeun jenis makanan anak? food the child consume? • Siapa yang memberikan makanan anak? • Saha anu masihan makanan anak? • Who feed the child? 9 Have you ever had any discussions regarding Apakah ibu pernah berdiskusi tentang Apa ibu pernah diskusi tentang pertumbuhan your baby/ child’s growth? pertumbuhan anak ibu? anak ibu? If yes, what is the discussion about? Mengenai apa? Diskusina tentang apa? Prompts: For example if you think he/ she is Bantuan: Misalnya apakah menurut ibu anak Bantuan: Misalnya apa menurut ibu anak ibu the tallest/ biggest one in the class/ ibu yang paling tinggi/ besar di kelas/ nu pang tinggi/ besarna di kelas/ lingkunganna. neighbourhood? lingkungan. Sareng saha? With who do you usually have the discussion Dengan siapa? Bantuan: with? Bantuan: • Suami Prompts: • Suami • Mertua • Husband • Mertua • Kader
PAGE | 163 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 5 of 13
Table 1 Questions used in focus group discussions in Greater Bandung Area, Indonesia (Continued) • In laws • Kader posyandu • Dokter • Posyandu (health post) kader • Dokter • Tatanggi • Doctor • Tetangga • Neighbours 10 What do you think when you see children who Apa yang ibu pikirkan ketika melihat anak yang Naon anu ibu pikirkeun upami ningali anak anu appear overweight/obese? Can you elaborate menurut ibu terlalu gemuk? Tolong jelaskan! menurut ibu parawakanna ageung teuing? Tiasa more on that? dijelaskeun bu.!
FGDs to ensure the nuances captured during the discus- In contrast, although the same subjective pattern was sions were agreed upon by three facilitators. seen in the low SES groups (32/35), three carers defined overweight objectively, based on Kartu Menuju Sehat Results (KMS – Health Card). KMS is a card used since the Characteristics of participants 1970s in Indonesia (Additional file 1). In 2010, the Indo- A total of 94 carers gave their consent and participated nesian Ministry of Health released a regulation to use in the FGDs. We had similar numbers of carers in the this card for children aged under 5. The purpose of the under 5 years and elementary groups. Most carers from KMS is to monitor children’s growth, to record chil- the low and intermediate SES groups had graduated ei- dren’s immunisation history, health services usage, and a ther from elementary or middle and high school, but tool to educate the community [18]. none from university. In contrast, most carers from the All carers (32/32) from the intermediate SES groups high SES groups were university graduates with only a defined overweight in a subjective way. A few (3/32) de- few having middle and high school qualifications alone. scribed overweight differently. We had more mothers compared to grandmothers in all three SES groups (Table 2). We had carers from 11 dif- Normal birthweight range ferent ethnicities; however, we did not find that ethnicity Carers from different SES groups had similar expecta- played a part in the differences in the discussions. tions for a baby’s birthweight. Carers from high SES groups expected their children to be heavier at age Emerging categories 6 months and 1 year, compared to the other two groups. Emerging categories are shown in Table 3. There were several marked differences between the three SES groups involved. Although we expected differences would arise “chubbier is healthier” from carers of children under 5 years old and carers of One third (12/35) of carers from the low SES groups elementary school children within the same SES group, agreed and two thirds (23/35) disagreed with the term such differences were not found. “chubbier is healthier”. They mentioned the concept of healthy overweight as an overweight baby/child who is Concepts of overweight still very active and is developmentally normal. This broad category portrays carers’ opinions around the All carers (32/32) from the intermediate SES groups concepts of overweight in children, “chubbier is health- were against the concept of “chubbier is healthier”. They ier” and the ideal adult body, including their expecta- specifically differentiated between cuteness and health tions of normal weight ranges at certain ages. Relevant status in overweight children. However, they also in- excerpts from the transcripts are presented in Table 4. cluded the concept of healthy overweight. One fourth (7/27) of the high SES carers agreed with “chubbier is healthier”, while the rest (20/27) disagreed, How carers define overweight highlighting past experiences of dealing with the over- Most (23 out of 27) carers in the high SES category de- weight status of their children. fined overweight subjectively, for example through visual interpretation of body size. The researchers did not pro- vide any particular definition of overweight nor differen- Ideal body tiate between overweight and obesity at the beginning of Carers from all three SES groups referred to Indonesian our discussions, since most Indonesians use the same media celebrities they considered had ideal bodies. word, kegemukan. One carer stated that in her opinion The ideal body was described as tall, slim, skinny, or the two terms are the same. Few (4/27) high SES carers model-like in women and a well-muscled body in defined it in a more objective way (i.e. based on specific men. However, it was difficult to name child celebri- measurements). ties they felt had an ideal body.
PAGE | 164 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 6 of 13
Table 2 Characteristics of participants from 12 focus group discussions in Greater Bandung Area, Indonesia Low SES (groups 5, 6, 7, 8) Intermediate SES (groups 3, 4, 9, 10) High SES (groups 1, 2, 11, 12) Number of participants Total 35 32 27 Under five 17 (48.6%) 16 (50.0%) 13 (48.1%) Elementary 18 (51.4%) 16 (50.0%) 14 (51.9%) Education Elementary or less 16 (45.7%) 11 (34.4%) 0 Middle & High School 16 (45.7%) 19 (59.4%) 3 (11.1%) Diploma 3 (8.6%) 2 (6.2%) 6 (22.2%) University or higher 0 0 18 (66.7%) Distribution of participants Mothers 31 23 (71.9%) 25 (92.6%) Grandmothers 4 9 (28.1%) 2 (7.4%) Number of children at home Two or less 24 (68.6%) 16 (50.0%) 17 (63.0%) More than two 11 (31.4%) 16 (50.0%) 10 (37.0%) Ethnicities 5 (Sunda, Jawa, Padang, 5 (Sunda, Jawa, Betawi, 7 (Sunda, Jawa, Riau, Dayak, Batak, Bali) Minahasa, Aceh) Batak, Betawi, Bugis) SES socio economic status
Factors contributing to overweight Dietary factors This category represents carers’ views on what they con- Carers from all SES groups stated that there was a range sidered were the three main factors contributing to over- of dietary factors contributing to overweight such as ex- weight: dietary factors, physical activity and sedentary cessive sugar consumption, unhealthy snacks, certain behaviours, and heredity (Table 5). types of food, big portions, frequent eating, and parental eating/snacking habits. Intermediate and high SES groups commented upon Table 3 Category and subcategories on the perception of vitamin supplements, milk, cooking oil, and mealtimes overweight in carers in Greater Bandung Area, Indonesia (especially dinner). Diets high in carbohydrate (rice or Category Subcategory noodles) were also a factor. 1. Concept of overweight 1.1 How carers define overweight Rice is Indonesia’s staple food. All SES groups consi- 1.2 Normal weight range dered rice as a “must have”, everyday food. One mother (high SES group) admitted that it is part of their identity 1.3 Chubbier is healthier as an Indonesian. 1.4 Ideal body shape It seems that what the family will cook/eat depended 2. Factors contributing to 2.1 Dietary factors heavily on what the children requested, because if the overweight 2.2 Activity levels and sedentary carers did not fulfill the children’s wishes the children’s behaviour in children behaviour may become difficult, for example by being 2.3 Heredity cranky or do not want to eat anything through out the 3. Awareness and feelings towards 3.1 Characteristics of overweight day. Most carers (25/32) from intermediate SES groups overweight in children 3.2 Food portion size in admitted that their children were heavily influenced by overweight children TV food commercials. 3.3 Eating frequency in Low and intermediate SES carers in particular mentioned overweight children that the ideal menu should consist of empat sehat lima 3.4 Feelings about overweight sempurna (Four Healthy, Five Perfect – an Indonesian children ideal meal incorporating carbohydrates, proteins, vegeta- 3.5 Sensitivity in carers of own bles, fruits, and milk). This concept, introduced by the overweight children Indonesian government in 1952, is now considered dated 3.6 Children’s current weight and because it only gives information about variety of food height without considering quantity [19]. In 1995 the government
PAGE | 165 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 7 of 13
Table 4 Quotes on concept of overweight according to carers from Bandung, Indonesia Low SES (FGD 5, 6, 7, 8) Intermediate SES (FGD 3, 4, 7, 8) High SES (FGD 1, 2, 11, 12) How carers • P1: Too much fat. • P2: Excessive fat. • P1: Weight and height not balanced/ define • P3: Big body. • P5: Body weighs more than it proportional. overweight • P4: Posture not appropriate with height and should be. • P4: (Someone who has a) big body. age range. • P7: Around here it means you • P8: Excessive weight. • P8: Weight exceeds normal range based on are healthy. • P7: Obesity is the same with overweight. age, based on the chart in KMS-Kartu Menuju • P6: In here it means you are happy. • P8: Weight is more than standard; I Sehat (Health Card) know there’s a way to measure it. Normal • Birth: 2.5–3.5 kg • Birth: 2.8–3.5 kg • Birth: 2.5–3.5 kg weight • 6 months: 5–9 kg • 6 months: 6–8 kg • 6 months: 8–10 kg range • 1 year: 9–12 kg • 1 year: 7–11 kg • 1 year: 8–13 kg Chubbier • P5: I personally don’t know what measurements • P4: I think it’s cute, but definitely • P3: Umm… When she was 6 months old, is healthier would count as obese, what obesity is. But not healthier. the doctor said she was obese. So yeah, according to my knowledge... I agree with “the • P8: (They have) lots of diseases. she was fat, chubby. At the beginning I chubbier, the healthier”. Being chubby is healthy. • P3: If they’re too fat, they are often thought she was cute. But when the doctor That’s what I think. sick and that’s not good. said she was obese, it frightened me. • P4: But, if it’s me, Doc, I want to have a fat child. • P1: Well, there are fat people who • P2: My child often got an asthma attack Who doesn’t want a fat child, right? Fat and are healthy and unhealthy. when she was fat. healthy, they’re so nice to look at. • P7: I think I agree with chubbier is • P6: It’s (chubbier is healthier) not true. • 1P2: If you mean a healthy fat, well... then that healthier, if it’s in toddlers. And also • P7: The child can be big, but actually, the is okay, maybe. But... if it’s unhealthy fat, it means in babies, I think child is not healthy. they have too much. • P5: Cholesterols. • P7: I want a child who is fat and healthy.. it would be better.. uh, I want it to be that way. Ideal body • P6: Luna Maya (famous model and movie star in Indonesia). shape • P7: Krisdayanti (famous singer in Indonesia), she still has a nice body even though she has children. • P8: Ideally, approximately 40 kg. • P3: Being ideal is about having nice height, too. P2:Steven William (half Indonesian, half English movie star). • P6: Like a model, hahaha. introduced new guidelines Pedoman Umum Gizi Seimbang playing outside although they also had screen time (PUGS - Balanced Nutrition Guidelines) that contain (watching TV at night). These families usually had information on both food variety and portion size along “physical activity time” once a week such as hiking the with the importance of exercise and water consumption mountains, running, and family sports time. In many [19, 20]. None of the carers mentioned PUGS. low and intermediate SES areas in Indonesia children must spend time (from 5 pm–6.30 pm) in Quran recita- Activity levels and sedentary behaviours in children tion class every day at the mosque. All elementary schools in Indonesia have a mandatory In contrast, children from high SES groups spent most 2 h physical education class per week. This was consi- of their after school time indoors at home. In addition to dered not enough by all low and intermediate SES the school physical education class, there were extracur- carers, who said children should engage in more physical ricular activities such as dancing, martial arts, and soc- activity. Carers stated that boys spent their time in PE cer. Eight (8/27) carers thought that, ideally, children class by playing soccer and girls by having aerobic acti- should have more (maybe around 3 h) physical activities vities. They also mentioned the 2 h includes time to outside school every day because the activities at school change clothes and sometimes time to walk to the are not enough. Half (13/27) of the carers stated chil- nearby sports field (low and intermediate SES), leaving dren do not need extra activity outside school because only 45–60 min for playing sports. All children in the they are moving around the house anyway - they have low SES groups went to public schools (free of any bigger houses. All children in these groups went to pri- charge), while children in intermediate SES groups went vate schools. Classes in private schools start at 7 am and to public schools or Islamic schools (madrasah; some fee finish at 11.30 am for kindergarten, 1.30 pm for 1st and payment required but not as expensive as private 2nd grade of elementary school, and 3.30 pm for 3rd-6th schools). Public kindergartens start at 7.30 am and end graders. Most children had other extracurricular activ- at 9.30 am, 1st-3rd grades of elementary schools start at ities e.g. piano class, math class, swimming class. These 7.00 am and end at 10.00 am, while 4th–6th grades start arrangements resulted in children arriving home at 6 pm at 7.00 am and end at 12.00 pm. Children from low and or even 6.30 pm, when they usually had dinner, watched intermediate SES groups spent their time after school TV or played with their tablets and then went to sleep.
PAGE | 166 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 8 of 13
Table 5 Quotes on factors contributing to weight according to carers from Bandung, Indonesia Low SES Intermediate SES High SES Dietary factors • P1: It’s mainly caused by one’s dietary • P2: They eat too much • P1: We are Indonesians, we have to habit that is too... where one does not P5: It has to be empat sehat lima eat rice. have appropriate portions. Also, frequent sempurna (Four Healthy, Five Perfect – • P7: She has a big appetite, drinks a lot of eating can cause obesity. an Indonesian ideal meal incorporating milk, and her father is also fat. Could that • P4: It could be that, Ma’am, aa, obesity carbohydrates, proteins, vegetables, be a factor? comes from eating too many sugary fruits, and milk). • P2: I rarely consume noodles now, when foods, excessive sugar consumption. Or it • P2: Like when you sleep right after you I do, I only put half of the MSG. Hahaha. could be from the snacks, Ma’am; they’re eat, lack of exercise, eating snacks… • P6: So you add salt to it? not very healthy. these things make you bigger. • P2: Aa, nope, I add eggs and mix them • P7: Even if he has eaten two packs of • P9: Aa, they eat a bowl of noodle and a with butter, so it has the right amount of noodles, if he has not eaten rice, he’ll plateful of rice. saltiness from the butter. consider it as if he hasn’t eaten yet. M: So, noodles and rice. How often do • P6: But that’s where the fat come from they eat that, Ma’am? (the butter). P9: Aa, sometimes they eat those twice • P4: It can be the quantity of the milk.P5: a day. It can be how much the children drink P5: Noodle can also cause obesity. the milk. • P7: Initially, A (name) drank too many • P3: Maybe it’s because they drink a lot sugary drinks like... something like Teh of milk. Gelas (sugary drinks brand). So he • P4: For example, he drinks brand A, became obese. although he only drinks a little, but the composition is different. Maybe one of the compositions stimulates the appetite. Activity levels • P1: After they eat, they’re just sitting • P6: Lifestyle. • P4: Video games. and sedentary down, watching TV. So that, aa, there is M: What kind of lifestyle? • P5: This (imitates someone playing video behaviours in no chance for them to burn the calories. P6: One that lacks exercise. games) isn’t being active. children That might have caused (the person • P8: My child can’t stand staying at home, • All: Hahaha to be). so he plays outside. • P4: Because, the children can, umm, even • P7: Fat. • M: What do you think about the physical if they don’t go outside, they can do the • P2: He is out somewhere to play (right education class at school? activities at home.P3: They can play after he came home from school). • P6: It’s not long enough. around the house, moving here and Changes his uniform, then goes straight • P4: Right, it’s not enough. there, right? outside to play. • P3: They mostly play outside, of P4: Yes. They can go up and down. • P3: After school my daughter usually course.P5: Yeah, outdoors. P5: On the stairs, right? have her lunch, then….. Then after lunch, • P8: He reads Qur’an in the evening. • P2: For my son of course it’s inside. that’s in... in the afternoon, aa, she is • P5: He reads the Qur’an, ma’am. • P3: Heeh, he arrived in the afternoon, joining Qur’an (Moslem’s holy book) • P3: My child reads the Qur’an in the and then go straight to his math course. recitation. evening and afterwards he plays for a And then it’s Maghrib (praying times- • P7: Maybe, being fat is caused by lack while M: Outside or inside the house? 6 pm) when he came home; he only of...maybe lack of activities? • P3: He plays outside. spends some time outside the house on • P8: They fly kites. Saturdays and Sundays, and that doesn’t • P5: They just run around. happen a lot either. • P8: They ride bicycles, Doc. • P2: Yeah, same here. Even on Saturdays • P7: Right, bicycles. and Sundays he doesn’t spend much time outside. Heredity • P1: Those who have inherited it from.... • P8: There’s the heredity factor. • P2: It’s inherited. I am one of the big • P6: From the parents. • P3: Yes, it’s inherited. families. Yeah, my husband and I, we’re • P3: Being large is in my family’s blood • P5: That’s why the children become fat. big and tall people. • P8: But he inherits it from his father. His • P7: If the parents were fat, the children • P7: Several diseases are heredity like father is also fat. would also be fat. cancer, asthma, but not overweight. • P2: That’swhenanexcessivedietis • P1: Heredity. • P8: The shape and size of the body is not balanced by exercise, which was • P2: It can be inherited. probably not affected by heredity. mentioned earlier. Heredity is a factor • P4: Fatness can be inherited. • P6: It might be both. Maybe, the fat that can play into this, too, right? Like people already has a heredity factor and genes? Right? then… • P9: People in the village always think • P5: It’s worsened by wrong dietary habit. like this, it’s normal if the children are • P6: The heredity, maybe, heredity factor? fat because the father is also fat, just • P4: Do you mean genetic factor? like that. • P6: So, their bones are naturally big. • P5: Yes, since they were little, it’s already big.
Heredity from high SES groups elaborated further on the concept of Carers from low and intermediate SES groups had a similar heredity based on their personal observation of their chil- understanding of heredity i.e. overweight is simply what has dren/nephews. A few (5/27) from high SES groups thought been going on in their family for several generations. Carers that heredity is not afactorthatcontributestooverweight.
PAGE | 167 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 9 of 13
Awareness and feelings towards overweight in children All carers from the high SES groups shared the Relevant excerpts for this category are available in same negative feelings towards overweight children. Table 6. Carers from all SES groups shared their concern about overweight-related health problems such as Characteristics of overweight heart disease, diabetes, difficulty in breathing, choles- All SES groups mentioned similar characteristics: phy- terol problems, or high blood pressure. sical appearance (“bulging belly”, head circumference, “stomach folds”, plump cheeks, “hulk like”, “big butt”, no Sensitivity in carers of overweight children neck), size of their clothes, difficulty in movements, ti- All carers who mentioned that their children were ring easily, panting breath, and having a lower IQ level. overweight expressed similar sensitive attitudes when Intermediate and high SES carers added that overweight it came to their own children, irrespective of SES. For children get sick more often compared to their healthy example, they placed more attention on their chil- weight counterparts. dren’s capabilities in physical activities rather than their weight status. Food portions in overweight children All carers stated that overweight children eat bigger por- Children’s current weight and height tions compared to healthy weight children, admitting Most carers of the low and intermediate SES groups that sometimes their children eat the same portion of knew their children’scurrentweight(29/35and25/ food as their parents. One mother (low SES group) 32, respectively), but only some (21/35 and 15/32, stated that as one of her efforts to reduce her child’s respectively) knew their children’s height. These fam- weight, she gave her smaller portions, different from the ilies had their children’s height and weight measured recommended serving size on the box of instant food. regularly at the Posyandu until they were 5 years old. From the carers’ facial expressions and the way they Eating frequency in overweight children stated their experience of going to Posyandu to have the Carers from the low SES groups were aware that their child measured, it is indicative that carers from low and overweight children eat more than three times a day. intermediate SES groups may consider this as just an One carer stated that her obese child eats regularly three obligation to tick the checklist instead of an opportun- times a day at home, but every time he visits a relative ity to monitor their children’s development. In public or a neighbour, he accepts their invitation to eat. It is elementary schools, weight and height are measured customary in Indonesia to offer people (children or once a year. adults) food whenever they visit. It is considered polite All high SES group carers agreed that their indicator and respectful for the host to do so, and the guest of health is their children’s weight, with 22/27 knowing should accept, in order to show respect. their children’s current weight, while only a little more Carers in the intermediate and high SES groups also than half (15/27) knew their height. Five carers (5/13) raised the issue of snacks, as their children would eat admitted they were only concerned about the children’s three or four times a day and snack in between. weight and not height; one even stated she never paid attention to her children’s heights. One mother stated Feelings about overweight children that the only opinion that matters for her was her Carers from low SES groups had mixed feelings about trusted doctor’s. Under 5 children are weighed during overweight babies/children, one fourth (9/35) having visits to the paediatrician whereas 7–12 year old children positive feelings such as happiness and an eagerness to are weighed every 2 months at (private) school. Even pinch the child to express their feelings. Pinching chil- with this scheme, many carers did not pay much atten- dren on their cheeks to express their feelings is common tion to their children’s current weight or height. in Indonesia, for example when people think the child is One third (35/94) of all carers from all SES groups cute. However, more carers (26/35) had negative feelings were satisfied with their children’s growth; (including all such as fear that their children will become overweight, carers who stated that their children were healthy feeling sorry for the overweight child because they might weight/overweight), while 59/94 carers were not satis- be bullied at school or because of their inability to par- fied. Carers admitted that they compared their children’s ticipate in some of the schools’ physical activities, or weight or height to that of their peers at school, in the feeling that overweight children are “just lazy”. playground or other public places. One mother from the Similar to carers from the low SES groups, carers from intermediate SES groups stated that she compared her the intermediate SES groups also had mixed feelings about son’s height with his friends because she thinks that he overweight babies/children. Several (6/32) expressed posi- is short and felt relieved when she saw that his peers tive feelings but most (26/32) portrayed negative feelings. were mostly ‘as short as he is’.
PAGE | 168 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 10 of 13
Table 6 Quotes on awareness and feelings towards overweight in children according to carers from Bandung, Indonesia Low SES Intermediate SES High SES Characteristics • P5: If it’s in babies, Ma’am, from here • P8: Uh... the belly is bloated and it looks • P4: For example, babies his age are of overweight (pointing at her arms), we can see it from like the person is pregnant. supposed to be in the standing and their arms. So, in Sundanese, gegeretan • P2: When you bring them out to walk, started to walking stage (but the fat (Sundanese term- babies with fat folds; it they get tired easily. baby cannot do that just yet). looks as if their flesh have been sliced). • P6: They run for a little bit and they are P6: Yeah, it’s like, the children seem to • P7: The size should be the plus one; like (makes panting noises). experience difficulties to move. It looks otherwise it must be sewn. Hahaha • P7: They are slow; they walk slowly. difficult for them to walk. I see that a lot, • P3: ... for example after they walk for a • P5: So, for example, when the baby is 4 and I use to think, “That is overweight”. little while, they go hah hah hah months old, the other child can lie flat on • P5: It’s probably because she has tough (laboured breathing noises). the stomach, but the fat baby can’t do it. bones, bigger bones and however much • P5: Sundanese people said that the way weight she loses she will never be as they walk is ngajegag (walking with the thin as other kids. The problem is, leg open, like frog legs), right? currently she moves slower than her younger sibling and she has breathing problems, too • P7: If the person is overweight, he’s prone to diseases and becomes sick easily. Food portion • P9: So when I gave her the food, I gave • P3: My son (elementary school) eats the • P6: She only eats zuppa soup as main size in her less, I don’t follow the serving size same portion as mine. meal.P2: How many bowls?P6: Only overweight guidelines from the box. • P7: Like earlier this morning, before I three bowls. children • P1: For normal children, it’s just the came here, we have breakfast with fried • P5: Because we think the children are still standard serving...for the overweight rice. Well, after fried rice, he asked more, on the growth period (so we let the children, they will have as much as he asked for noodles. I gave him one children eat a lot). adults. pack of noodle, and he ate it until there’s • P7: Parents are also afraid if the children nothing left. get sick, because there are so many • P4: In a day? Well, actually, he eats three diseases right now. times a day, but the portion is more than • P6: So sometimes they make their me. For me, for example, I eat one full children eat a lot to prevent them plate. My child can eat up to three full from getting sick. portion. Eating • P4: My niece, 5th grade, 58 kg, she eats • P7: He’s always eating. He never stops • P5: Yeah, people do that you know. My frequency in at 6 pm, then she eats again at 8 pm. eating. son, my eldest one weighs almost 70 k. overweight Mostly, she eats meatballs, she rarely eats • P5: He eats four times a day, and he also He eats one meal at 10 am, and then children any vegetables. eats snacks like bread, cakes, sweets. another one at 12. In the afternoon, • P2: A child should eat three meals a day • P2: He eats…he eats at noon. The others when he finally gets home, he eats right? This one eats more than three only eat in the morning and after Isya another meal. He just wanders around times and she eats noodles, sometimes (the last Moslem’s prayer time in a day, in the kitchen looking for food, and two packages aren’t enough for her, only Indonesia, it’s at about seven o’clock suddenly he has a plate full of rice. four is. p.m). When the others eat after Isya, my • P4: For my child, if he eats regularly like • P9: My (obese) child eats regularly, three son also joins them. After that, he three times a day, his weight is stable. times at home, but even when he has watches TV until 11 o’clock. Sometimes, Three meals: morning, noon, and eaten and our neighbour offers him at ten o’clock, he eats again. That means afternoon around six pm. Stable, but something to eat, he will eat it. When he he doesn’t make a move (after his last once there are snacks. Oh, his weight comes to his aunt’s house and she offers meal), then he directly goes to sleep. goes skyrocketing. him food, he will eat it. Feelings about • P5: Who doesn’t want a fat child? • P7: When they’re little, it’skindofcute • P8: When we look at the fat children, overweight • P7: They are cute. I want to pinch them. (to be fat). we’ll think they are cute…But actually it’s children • P2: I feel gereget (kind of loving • P3: Yes, cute, like on TV. a pity, fat children are having difficulties something cute mixed with desire to • P6: It scares me. to move. pinch, in Sundanese), Ma’am. • P1: If the child is too fat, sometimes I feel • P6: I’m afraid my child will become • P8: Happy to see a fat child, but pity him pity for him/her, just like the one on TV. that fat. as well • P8: I feel sorry for him, Doctor. • P2: The problem is that they are vulnerable • P6: Because it’sapity.It’s difficult. Yes, to diseases, I’m afraid of that. Underweight the fat children have difficulties to children are vulnerable to diseases, but so move. They become uh, they’re lazy, are the big ones, hahaha. they sleep a lot. • P7: Out of fear of becoming like A (name), so we limit our child’s diet... • P9: Horrified. Sensitivity in • P6: This doesn’t happen to my child, • P8: I don’t think a fat person gets tired • P5: My child, who is fat, is actually very carers of own he is fat but healthy. easily because mine is fat (and he does active and can run quickly. overweight • P9: Well, when he runs around, he not get tired easily). • P8: When he’s racing with his thinner children doesn’t get tired easily. friends, he runs faster. • P4: Fat people are not always slower.
PAGE | 169 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 11 of 13
Table 6 Quotes on awareness and feelings towards overweight in children according to carers from Bandung, Indonesia (Continued) Children’s • P2: I just want it [my child’s weight] to be • P3: I have too many under 5 children, • P7: Importantly, it must be according to current weight just normal, ascending in a straight line everytime I go to Posyandu (community the chart, or the doctor’s opinion. If the and height [on the KMS- Kartu Menuju Sehat (Health health post) to weigh them; I bring 3 or doctor says okay, I’ll be thankful. Card)]. 4 children. • P6: I’m aware that I am short, so my • P4: I thought that my child was the only children won’t be tall. short one, but it turns out that there are a lot of kids who are as short as he is.
Discussion when other “subject matter” teachers were not available This is the first study in Indonesia that investigates the [24]. Another US study, using pedometers to count knowledge, attitudes and beliefs surrounding childhood steps, showed that typical PE classes only account for overweight of mothers/grandmothers from different SES 8–11% of the total steps made throughout the day groups. We identified 3 categories and 13 subcategories [25]. A 3-year intervention study of primary school as stated in Table 3. However, due to word limitations, children in Sweden showed that to successfully reduce we can only discuss several subcategories: normal weight weight gain, children needed to be physically active range, ideal body shape, characteristics of overweight, for more than 40 min in each physical education set- activity levels and sedentary behaviour, heredity, how ting every day [26]. Carers seemed less worried about carers define overweight, children’s current weight and physical education at school because children from height, and chubbier is healthier. low and intermediate SES groups spend most of their All three SES groups had similar expectations about after school hours playing outside the house, which an appropriate weight at birth, 6 and 12 months. This add to their physical activity hours; while the high SES might have resulted from the participation of mothers in children engaged more in extracurricular activities. the Posyandu, [21] community counselling usually done Many carers assigned heredity or genetic predispo- in Puskesmas, or consultations with health practitioner sition as the ‘expected and accepted’ cause of childhood or paediatricians. All participants had similar ideas about overweight, a finding also documented in Australia [27] the ideal adult body type, demonstrating the influence of and the United States [23, 28, 29]. the media in their lives. It is very likely children of pa- Most carers knew the current weight of their children rent participants watched television with their parents but less than 60% of these carers knew about their chil- and were influenced by what is considered the ideal dren’s height. Most carers defined overweight subjec- adult body type. Thus, children who absorbed large tively from physical appearance. This finding is in amounts of screen time could relate to the ideal for keeping with studies from Germany, [30] the United adults but not have a connection with the young child’s Kingdom, [22] the United States, [23, 31] and in bodyweight because there is no such ideal for children Vietnam [32]. While some carers knew how to define either in the media or in everyday reality. Linking this to overweight objectively, this knowledge is not very useful inadequate PE in school curricula seems to suggest that when they do not know their child’s current weight or a child’s ideal body type is not considered as important. height. Knowing how to measure overweight and then All carers had similar opinions about the characteristics knowing a child’s weight and height status may be impor- of overweight children: overweight children had bigger tant first steps in recognising whether children are under- portion sizes and ate more frequently compared to their weight, of a healthy weight or overweight [22, 31–33]. healthy weight counterparts. How carers characterised Some carers agreed with the ‘chubbier is healthier’ overweight is similar to studies in the United Kingdom concept that is also found in other cultures. For example in [22] and the United States [23] where parents were Turkey (plump kids are healthy), [34] Vietnam (chubbiness found to rely on visual appearance, exercise capability in is good), [32] the United States (“baby fat” is cute and relation to peers, and clothing sizes. healthy [31]; a bigger infant is a better infant [28]) and the The carers recognised that dietary factors, activity United Kingdom (a big baby is a healthy baby) [35]. The levels and sedentary behaviours, and heredity all contri- concept of overweight symbolising family wellbeing is also buted to the development of overweight. Most carers found in China [36]. agreed that the 2-h physical education (PE) class at Carers from the high SES groups expressed their opi- school is not enough, especially as this includes chan- nions freely and in a more confident manner compared ging clothes and getting to the sportsfield. One qualita- to the other two groups. This is probably related to their tive study with US school principals, PE teachers and educational background, with more carers from the high students highlighted the need for improvement in the SES groups being university graduates, compared to in-school physical activity; one student referred to PE none from the other groups. Within the low SES groups, class as a “dumping ground”, where children were sent we found that the women from villages with an active
PAGE | 170 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 12 of 13
midwife (in terms of educating the carers) were involved differences on the perceptions of overweight not only be- more actively in the discussion. This may highlight the tween different SES groups but also within the same SES significant role played by front line staff in the health groups. system, e.g. midwives and general practitioners. Two studies in Indonesia support this issue, stating that the Recommendations availability and higher participation in Posyandu in Irrespective of these differences, one of the policy impli- Indonesia have significantly improved children’s nutri- cations around the perception of carers is the impor- tion status and had a protective effect on the develop- tance of educating mothers and other primary carers of ment of overweight in children [21, 37]. children around healthier lifestyles. This study highlights One aspect to highlight in our study is the interaction the need for improvement within Indonesia in the dis- between participants within each focus group, an advan- semination of current policies such as PUGS and the use tage of conducting FGDs [38]. Some carers supported of KMS, for example through the media (newspaper and each other and some questioned another’s opinion, television). The use of the KMS card needs to be imple- bringing many insights to the topic and richness to our mented with all levels of health practitioners, making it data. Comparison between the three SES groups re- the national reference in Indonesia to regularly monitor vealed more similarities in low and intermediate SES children’s growth. Consideration should be given to rou- groups in the way they perceived childhood overweight tine anthropometric measurements, currently under- and related topics. We also found that in a few subca- taken in private elementary schools, being implemented tegories (how carers define overweight; chubbier is in every school, thus providing feedback to parents healthier), carers from low SES groups shared similar about their children’s weight and height status. There is knowledge with their counterparts from the higher SES room for review and revision of the quality and quantity groups. For some issues, the low SES carers even of physical education class in elementary schools. The showed better knowledge compared to their counter- Department of Education may need to increase the times parts from higher SES, for example in the knowledge to for children’s physical activities during school hours, e.g. define overweight objectively and knowledge about the during recess or before and after school physical acti- current weight and height of their children. vities. Future research should aim to explore the food A strength of this study is that it is the first in Indonesia choices made by carers and the involvement of other to investigate the perceptions of overweight in carers from people in the household (father, grandparents) in making three different SES groups. Secondly, in contrast to the diet-related decisions. other Indonesian study where only mothers of overweight children were involved, our study recruited carers with Endnotes children of diverse weight status, providing a variety of 1 perspectives. Thirdly, we had the same number of groups We included carers of under five and elementary in each status category, allowing us to compare the results school children because of the practicalities in recruiting between the three groups involved. Fourthly, we suc- these mothers through Community Health Centers, pri- cessfully attracted participation from carers of low vate practices and schools. Carers of 5 and 6-year-olds SES groups, even though low SES participants are (kindergarten age) are much harder to recruit because considered by some to be difficult to recruit [31]. One kindergarten is not mandatory in Indonesia. of our study’slimitationsisthatbecausethenumber of mother participants outweighed the number of Additional file grandmother participants, comparison of these two groups’ perspectives was unavailable. Although our Additional file 1: Kartu Menuju Sehat (KMS- Health Card) for boys aged 0–24 months. Source: http://www.depkes.go.id/resources/download/info- qualitative research findings may lack generalisability, terkini/Kartu%20Menuju%20Sehat%20KMS.pdf. (PDF 2475 kb) we believe our findings are supportive of findings from other countries while producing new replicable fin- Acknowledgements dings that will represent carers’ views if done in other CNR would like to thank Lembaga Pengelola Dana Pendidikan (LPDP), the regions of Indonesia. Republic of Indonesia for her PhD scholarship.
Funding Conclusions This study was funded by the Children’s Hospital at Westmead (CHW), Australia. This study sought to explore the perceptions of carers CHW did not have any role in the design of the study and collection, analysis, towards childhood overweight. Although there were simi- and interpretation of data and in writing the manuscript. larities in the way carers from different SES groups per- Availability of data and materials ceived overweight, the differences were of great interest. The complete datasets used and/or analysed during the current study are In discussion of health-related issues, there were available from the corresponding author on reasonable request.
PAGE | 171 Rachmi et al. International Journal of Behavioral Nutrition and Physical Activity (2017) 14:101 Page 13 of 13
Authors’ contributions 15. Strauss A, Corbin J. Basics of qualitative research: grounded theory All authors contributed equally in designing the study. CNR analysed and procedures and techniques. Newbury Park, CA: Sage; 1990. interpreted the data under the supervision of CLH, ML and LB. All authors 16. Glaser B, Strauss A. The discovery of grounded theory: strategies for contributed in writing the manuscript and have read, revised and approved qualitative research. Chicago: Aldine; 1967. the final draft of the manuscript. 17. NVivo qualitative data analysis Software. vol. Version 11: QSR International Pty Ltd.; 2014. http://www.qsrinternational.com/support/faqs/how-do-i-cite- Ethics approval and consent to participate nvivo-10-nvivo-9-or-nvivo-8-in-my-wo. We obtained ethics approval from the Ethics Committee at Fakultas Kedokteran 18. Kartu Menuju Sehat [http://www.depkes.go.id/resources/download/info- Universitas Padjadjaran,Bandung(no.473/UN6.C1.3.2/KEPK/PN/2016)dated16th of terkini/Kartu Menuju Sehat KMS.pdf]. June 2016. Further permits were obtained from three health departments, 19. Inilah Perbedaan ‘4 Sehat 5 Sempurna’ Dengan ‘Gizi Seimbang’ [http:// Puskesmas,andschoolsofthreeareas.PriortoeachFGD,theparticipants www.depkes.go.id/article/view/16051300001/inilah-perbedaan-4-sehat-5- provided written informed consent. sempurna-dengan-gizi-seimbang-.html]. 20. Departemen Kesehatan RI. Direktorat Jenderal Bina Kesehatan Masyarakat Consent for publication DKR: Pedoman Umum Gizi Seimbang (PUGS) - panduan untuk petugas. Not applicable. Indonesia DKR ed. Jakarta, Indonesia: Depkes RI; 2003. 21. Anwar F, Khomsan A, Sukandar D, Riyadi H, Mudjajanto ES. High Competing interests participation in the Posyandu nutrition program improved children The authors declare that they have no competing interest. nutritional status. Nutr Res Pract. 2010;4:208–14. 22. Jones AR, Parkinson KN, Drewett RF, Hyland RM, Pearce MS, Adamson AJ. Gateshead millennium study Core T: parental perceptions of weight status Publisher’s Note in children: the Gateshead millennium study. Int J Obes. 2011;35:953–62. Springer Nature remains neutral with regard to jurisdictional claims in published 23. Goodell LS, Pierce MB, Bravo CM, Ferris AM. Parental perceptions of maps and institutional affiliations. overweight during early childhood. Qual Health Res. 2008;18:1548–55. 24. Gamble A, Chatfield SL, Cormack ML Jr, Hallam JS. Not enough time in the Author details 1 day: a qualitative assessment of in-school physical activity policy as viewed Discipline of Child & Adolescent Health, The Children’s Hospital at by administrators, teachers, and students. J Sch Health. 2017;87:21–8. Westmead, University of Sydney Clinical School, Locked Bag 4001, 2 25. Tudor-Locke C, Lee SM, Morgan CF, Beighle A, Pangrazi RP. Children's Westmead, NSW 2145, Australia. School of Social and Political Sciences, The pedometer-determined physical activity during the segmented school day. University of Sydney, RC Mills Building (A26), Sydney, NSW 2006, Australia. 3 Med Sci Sports Exerc. 2006;38:1732–8. Sydney School of Public Health, The University of Sydney, Edward Ford 26. Sollerhed AC, Ejlertsson G. Physical benefits of expanded physical education Building (A27), Fisher Rd, Sydney, NSW 2006, Australia. in primary school: findings from a 3-year intervention study in Sweden. Scand J Med Sci Sports. 2008;18:102–7. Received: 26 April 2017 Accepted: 19 July 2017 27. Jackson D, McDonald G, Mannix J, Faga P, Firtko A. Mothers’ perceptions of overweight and obesity in their children. Aust J Adv Nurs. 2005;23:8–13. 28. Baughcum AE, Burklow KA, Deeks CM, Powers SW, Whitaker RC. Maternal References feeding practices and childhood obesity: a focus group study of low- 1. International Food Policy Research Institute. Global nutrition report 2015: income mothers. Arch Pediatr Adolesc Med. 1998;152:1010–4. actions and accountability to advance nutrition and sustainable development. 29. Jain A, Sherman SN, Chamberlin LA, Carter Y, Powers SW, Whitaker RC. Why Washington, DC: International Food Policy Research Institute; 2015. don't low-income mothers worry about their preschoolers being 2. Rachmi CN, Li M, Baur L. Overweight and obesity in Indonesia: prevalence overweight? Pediatrics. 2001;107:1138–46. and risk factors–a literature review. Public Health. 2017; 30. Sikorski C, Riedel C, Luppa M, Schulze B, Werner P, Konig HH, Riedel-Heller 3. Shrimpton R, Rokx C. The double burden of malnutrition in Indonesia. SG. Perception of overweight and obesity from different angles: a Jakarta, Indonesia: World Bank Jakarta; 2013. qualitative study. Scand J Public Health. 2012;40:271–7. 4. Rizona F, Susetyowati S, Lusmilasari L. Mother's feeding behaviours on 31. Eli K, Howell K, Fisher PA, Nowicka P. “a little on the heavy side”:a overweight children. Int J Community Med Public Health. 2016;3:831–6. qualitative analysis of parents’ and grandparents’ perceptions of 5. Tucker P, Irwin JD, He M, Bouck LM, Pollett G. Preschoolers’ dietary preschoolers’ body weights. BMJ Open. 2014;4:e006609. behaviours: parents’ perspectives. Can J Diet Pract Res. 2006;67:67–71. 32. Do LM, Larsson V, Tran TK, Nguyen HT, Eriksson B, Ascher H. Vietnamese 6. Towns N, D'Auria J. Parental perceptions of their child's overweight: an mother's conceptions of childhood overweight: findings from a qualitative integrative review of the literature. J Pediatr Nurs. 2009;24:115–30. study. Glob Health Action. 2016;9:30215. 7. Scaglioni S, Salvioni M, Galimberti C. Influence of parental attitudes in the 33. He M, Evans A. Are parents aware that their children are overweight or development of children eating behaviour. Br J Nutr. 2008;99(Suppl 1):S22–5. obese? Do they care? Can Fam Physician. 2007;53:1493–9. 8. Oude Luttikhuis H, Baur L, Jansen H, Shrewsbury VA, O'Malley C, Stolk RP, 34. Esenay FI, Yigit R, Erdogan S. Turkish Mothers’ perceptions of their Children’s Summerbell CD. Interventions for treating obesity in children. Cochrane weight. J Spec Pediatric Nurs. 2010;15:144–53. Database Syst Rev. 2009. doi:10.1002/14651858.CD001872.pub2. 35. Southwell O, Fox JR. Maternal perceptions of overweight and obesity in 9. Golley RK, Hendrie GA, Slater A, Corsini N. Interventions that involve parents children: a grounded theory study. Br J Health Psychol. 2011;16:626–41. to improve children's weight-related nutrition intake and activity patterns - 36. Li J, Lei J, Wen S, Zhou L. Sex disparity and perception of obesity/ what nutrition and activity targets and behaviour change techniques are overweight by parents and grandparents. Paediatr Child Health. associated with intervention effectiveness? Obes Rev. 2011;12:114–30. 2014;19:e113–6. 10. Niemeier BS, Hektner JM, Enger KB. Parent participation in weight-related 37. Andriani H, Liao CY, Kuo HW. Association of Maternal and Child Health health interventions for children and adolescents: a systematic review and Center (Posyandu) availability with child weight status in Indonesia: a meta-analysis. Prev Med. 2012;55:3–13. National Study. Int J Environ Res Public Health. 2016;13 11. Gibson EL, Kreichauf S, Wildgruber A, Vogele C, Summerbell CD, Nixon C, 38. Green J, Thorogood N. Qualitative methods for Health Research. Second Moore H, Douthwaite W, Manios Y, ToyBox-Study G. A narrative review of edition edn. Great Britain: SAGE Publications Ltd; 2009. psychological and educational strategies applied to young children's eating behaviours aimed at reducing obesity risk. Obes Rev. 2012;13(Suppl 1):85–95. 12. Indeks Pembangunan Manusia [https://jabar.bps.go.id/Subjek/view/id/26- subjekViewTab3|accordion-daftar-subjek1]. 13. Operational Working Group Ministry of Health: Pedoman Umum Pengelolaan Posyandu. (Health Mo ed. pp. 6–7. Jakarta: Departemen Kesehatan Republik Indonesia; 2011:6–7. 14. Strauss A. Qualitative analysis for social scientists. Cambridge: Cambridge University Press; 1987.
PAGE | 172
Additional File 1. Kartu Menuju Sehat (KMS- Health Card) for boys aged 0–24 months. Source: http://www.depkes.go.id/resources/download/info- terkini/Kartu%20Menuju%20Sehat%20KMS.pdf.
PAGE| 173 Chapter 9: Food choices made by primary carers (mothers/ grandmothers) in West Java, Indonesia
INTRODUCTION 176 PARTICIPANTS, MATERIAL AND METHODS Study design, location, and participants 176 Research team, data collection and approach to analysis 178 Triangulation and quality assurance 179 RESULTS Characteristics of participants 179 Emerging themes 180 Decisions on food served at home 180 Attitudes toward instant noodles consumption 181 Decisions on food/ snacks and drinks consumptions 185 Feelings about food/ snacks and drink choices 186 Consideration of money issues 189 Consideration of Halal issues 189 DISCUSSION 191 CONCLUSIONS 194 REFERENCES 196
This chapter has been submitted to Appetite (under review).
PAGE| 174 ABSTRACT Background Indonesian children currently face a decreasing prevalence of underweight and an increasing prevalence of overweight. For children aged 0 to 12 years, mothers/grandmothers hold the primary role in feeding practices. This study specifically aimed to explore mothers’/ grandmothers’ decision-making around food choices in West Java, Indonesia.
Methods Twelve focus groups discussions were conducted between June and October 2016. Ninety- four carers of children aged under-five and 7-12 years from three socioeconomic groups participated. Data analysis followed a grounded theory approach.
Results We developed six main themes: 1) Decisions on food served at home, which were mainly influenced by socioeconomic status and family member requests. 2) Attitudes toward instant noodles consumption. The low and intermediate SES children consumed noodles on a daily basis, even up to five times daily, while the high SES children only consumed noodles as ‘emergency’ food. 3) Decisions on food/ snacks and drinks consumptions. Low and intermediate SES children bought snacks/ sugary drinks from street vendors while the high SES children bought them at supermarket. 4) Feelings about food/ snacks and drinks choices. All carers shared similar concerns and uneasiness about their children’s choices. 5) Consideration of money issues, which were prominent among the low and intermediate SES carers. 6) Consideration of Halal issues, which were only mentioned by high SES carers. Most similarities were found between the low and intermediate SES groups, while most differences arose between these two groups compared to the high SES.
Conclusions The six interrelated categories affected one another around participants’ decision-making around the family’s food consumption. Developing countries could improve their multi- sectoral collaboration in ensuring better availability and affordability of healthier food. Education targeted at behavior modification toward healthier lifestyles should be implemented countrywide.
Keywords: food choices; mothers; grandmothers; children; Indonesia; qualitative 1
PAGE | 175 INTRODUCTION The double burden of malnutrition is a condition where two types of malnutrition, undernutrition (underweight, stunting, wasting, thinness, or any form of micronutrient deficiencies) and overnutrition (overweight/ obesity), are present1,2. According to the Global Nutrition Report 2016, Indonesia is ranked 105 out of 128 countries (from lowest to highest prevalence) for overweight in children under five years of age, and 108 out of 132 countries for stunting, a form of chronic undernutrition (prevalence 36.4%)3. Thus, Indonesia is dealing with the double burden of malnutrition2,4,5. In under five children in Indonesia the prevalence of undernutrition (namely stunting and underweight) has decreased in the past two decades6,7 while the prevalence of overnutrition (overweight/ obesity) has increased6-8.
For pre-adolescent children in particular, mothers or carers hold the primary role in food choices, food purchases, food preparation and the serving of food to children9-12. Therefore, mothers and other carers are an important target group with whom to explore food choices and consumption. Better knowledge of factors influencing food choices and consumption in childhood is of great importance for everyone to understand the issue of childhood overweight, as well as other aspects of child health. Of the few studies of childhood obesity in Indonesia, most have used quantitative methods to estimate the prevalence and associated risk factors of obesity4,8,13,14. To our knowledge, no study has included a diverse range of socioeconomic groups and explored the reasons behind the food choices made by parents and grandparents who are the main carers of children in Indonesia.
This study is part of a research project around childhood overweight/ obesity in Indonesia. Our previous analyses have explored carers’ knowledge around the concept of overweight, factors related to overweight, and awareness and feelings towards children who are overweight15. The present study aimed to explore the reasons behind mothers’/ grandmothers’ decision-making around children’s food choices and consumption in West Java, Indonesia. Such information can provide new insights to governments and policy implementers to tackle the issue of childhood overweight in Indonesia and elsewhere.
PARTICIPANTS, MATERIAL AND METHODS Study design, location, and participants This qualitative study used focus group discussions (FGDs) which were conducted from June to October 2016. We selected three out of 27 provinces in West Java: Bandung City, Bandung Regency, and Bandung Barat Regency. These three provinces comprise the Bandung Greater
PAGE | 176 Area16 and they have different Human Development Index scores, which act as a measure of different socioeconomic status (SES). The three districts were: (1) Cisarua district from Bandung Barat Regency, a low SES area, (2) Margahayu district from Bandung Regency, an intermediate SES area, and (3) Sukasari district from Bandung city, a high SES area16. Cisarua and Sukasari districts are 18 km apart, while Margahayu district is 30 and 40 km away from Sukasari and Cisarua, respectively.
We conducted four FGDs in each district, representing the three levels of SES, totaling twelve FGDs. Within each district, we selected two groups of carers of under five children and two groups of carers of 7-12 years children (Figure 1).
Cisarua#District#2 2#FGDs#with#carers#of#pre2school#children# Low#SES## (4#FGDs)# 2#FGDs#with#carers#of#primary#school#children#
12#Focus# Margahayu#District# 2#FGDs#with#carers#of#pre2school#children# Group# Medium#SES# Discussions# (4#FGDs)# 2#FGDs#with#carers#of#primary#school#children#
Sukasari#District#2# 2#FGDs#with#carers#of#pre2school#children# High#SES## (4#FGDs)# 2#FGDs#with#carers#of#primary#school#children#
Figure 1. Numbers and participants of Focus Group Discussions in Bandung Greater Area, West Java, Indonesia.
To be included in the study, participants were required to be the primary carer of the children, meaning that they held or took a major part in taking care of the children or were responsible for the feeding practices of their children (including preparing, cooking, and feeding the children). We aimed to gain a range of different perspectives of childhood overweight by selecting mothers and grandmothers (both of whom will be referred to as carers hereafter) from different SES areas.
The recruitment process was similar for low and intermediate SES areas, where we collaborated with the Community Health Centres (Puskesmas) and public elementary schools in the districts. In Puskesmas, we obtained assistance from local midwives and health kader. Health kader are mothers who have knowledge of the people in their communities and train to work for the Puskesmas and Community Health Posts (Posyandu) for a small incentive17. We PAGE | 177 explained the project to midwives and health kader and they helped to disseminate the participant information statement. In public elementary schools we received assistance from the teachers and class coordinators (one of the mothers from each grade who holds the responsibility as the connector to disseminate information from teachers to fellow mothers). We also disseminated the participant information sheets through the students’ communication books.
For high SES carers, we sought assistance from nurses from private schools in the area and paediatricians’ offices, as children from this SES group do not go to Puskesmas for medical advice. We contacted the class coordinators in the private schools and asked the school nurse to disseminate the participant information sheets to potential carers.
Those who were interested in the study then contacted the research team who gave additional information as needed. When verbal consent was obtained, we took the recruit’s details and set the date and time for the focus group discussion.
Research team, data collection and approach to analysis A team of researchers (first author and two medical practitioners as research assistants) received research training delivered by CLH in Bandung, Indonesia in order to ensure a deeper understanding of qualitative research in general and FGDs in particular. The semi- structured/ structured moderator guide was finalized and translated (into the Indonesian and Sundanese languages) during the training. Afterwards, we conducted a pilot FGD and discussed and attempted resolution of potential issues that might arise during the FGDs. The results of this pilot FGD were not included in our analysis.
The first author took the role of moderator, one research assistant as an observer, and the other as recording manager/ timekeeper. Before starting the FGDs, research assistants asked the participants to fill in basic demographic information and to sign the consent form. The moderator then explained the process of the FGDs and offered to conduct the discussion in either the Indonesian or Sundanese language. Sundanese is the language of West Java and all research team members were proficient speakers. We had a 30-minute ice-breaker, during which we built rapport with all carers and emphasized the importance of having everyone sharing their opinion and experience. To ensure consistency, the moderator always used the moderator guide, which contained a set of questions and prompts. We had 6-9 carers in each
PAGE | 178 FGD and everyone received $11 USD for participating in the study. The discussions lasted for 50-90 minutes and were recorded using a digital/ mobile phone recorder.
After the first FGD, we began the transcription and translation, where transcriptions were firstly written in the Indonesian language and checked by CNR. This approach gave us the advantage of being able to reflect on the discussion. Afterwards, we translated the transcripts to English. Both CNR and CLH checked the Indonesian and English transcripts to ensure accuracy of transcript and translation. Some of the Sundanese and Indonesian terms that are untranslatable to English were left in their original forms with an explanation of the terms in English.
We chose the grounded theory approach in our analysis18-20. NVivo for Mac version 11.0.0 was used to assist in the data coding and analysis processes21. CNR, under full supervision from CLH, used the following approach: (1) read the transcripts several times; (2) coded the transcripts openly and compared the results within and between transcripts; (3) identified categories of concepts; (4) reviewed and finalized the categories; (5) identified similarities and differences within and between SES groups; (5) quantified comments from individual carers in each FGD; and (6) chose which excerpts best represented each category. Excerpts representing each category were selected from different FGD groups. Finally, CNR, CLH, LB, and ML checked and agreed upon the final version of categories arising from the transcripts.
Triangulation and quality assurance After each FGD, the research team performed triangulation by discussing the observational notes each one made during the process, including the voice intonation or facial expressions that a particular participant made, and the flow and nuances of the discussion that we captured. Another researcher (CLH) helped with the analysis to reduce interpretive bias, because she was not present during the FGDs. CLH also ensured quality of the analysis by checking that all categories originated from the transcripts.
RESULTS Characteristics of participants There were 94 carers who participated in the FGDs, with relatively comparable numbers of participants from the under five and the 7-12 years groups. Two thirds of carers from the high SES groups were university graduates while most carers from the low and intermediate SES
PAGE | 179 were middle school graduates. We had 15 grandmothers who participated in all 12 FGDs while the rest (79 participants) were mothers. We had a total of 11 different ethnicities in the 12 groups (Table 1).
Table 1. Characteristics of participants of 12 focus group discussions in Greater Bandung Area, Indonesia. Low SES Intermediate SES High SES (FGDs 5, 6, 7, 8) (FGDs 3, 4, 9, 10) (FGDs 1, 2, 11, Total number of participants 35 32 12)27 Carers of under fives 17 16 13 Carers of elementary 18 16 14 Educationschoolchildren Elementary schoolchidlren or less 15 11 0 Middle School (Junior or 16 19 3 DiplomaHigh) 3 2 6 University or higher 0 0 18 Distribution of participants Mothers 31 23 25 Grandmothers 4 9 2 Number of children at home Two or less 24 16 17 More than two 11 16 10 Ethnicities 5 (Sunda, Jawa, 5 (Sunda, Jawa, 7 (Sunda, Jawa, Padang, Batak, Betawi, Minahasa, Riau, Dayak, Bali) Aceh) Batak, Betawi, Bugis) SES: socio economic status FGDs: focus group discussions
Emerging themes We found six emerging themes: 1) Decisions on food served at home. 2) Attitudes toward instant noodles consumption. 3) Decision on food/ snacks drinks consumption. 4) Feelings about food/ drinks consumption. 5) Consideration of money issues. 6) Consideration of Halal issues.
Decisions on food served at home In Indonesia, public primary school starts at 7 am and runs to 12 noon and private schools start at 7 am and run to 1.30 pm. Thus, children of public schools do not eat any lunch in school; they usually eat breakfast, lunch and dinner at home. For private school students, they usually have lunch in schools, with a meal that is prepared from home or based on the catering menu provided by the school.
PAGE | 180 Amongst the low SES carers, decisions about food served at the table were based on the carers’ decisions and their children’s requests. Whether or not the children’s requests were accommodated depended largely on the family’s financial situation. We observed that with the low SES carers, although they wanted to ask their children about food choices that they would cook/ eat for the day, money concerns dictated how far they could accommodate their children’s wishes. This was more of an issue with low SES carers compared with the intermediate SES carers. Intermediate SES carers stated that they made most of the decisions about the food served, although sometimes they asked other family members such as their husband or the children. Intermediate SES carers had more freedom – in terms of purchasing power – to provide the food according to their children/ husbands' requests, however several carers still had money concerns. For detail excerpts please see Table 2.
In contrast, carers from high SES groups stated that everyone in the family had a say about what they would cook/ eat for the whole day. For example, the mother chose the kind of protein (beef, chicken, or fish), but the rest of family had a say about how they wanted it cooked (roasted, baked, fried). Similar to carers from intermediate SES, the main reason for asking their children’s opinion was to entice and ensure they ate the food served at home.
Attitudes toward instant noodles consumption In many Asian countries, noodles are considered a staple food. Indonesia has the second highest consumption of instant noodles per country per year globally22. Instant noodles were first available in Indonesia in 1968 as the brand SuperMi©, which was then followed by IndoMie© in 1972. These are the most famous noodle brands in Indonesia. Many people refer to instant noodles as IndoMie©/ SuperMi© even though they eat different brands. Indonesia is currently the biggest producer of instant noodles in the world. A single packet of instant noodles costs 0.15 USD while one kilogram of rice costs 1.04 USD. One kilogram of rice will typically feed 6 people, but other food needs to be consumed with the rice, which will be an extra cost. It takes around 3 minutes to cook instant noodles and 30-40 minutes to cook rice.
Six carers from low SES groups said they served noodles as snacks or a side dish, which meant the family ate it with rice or had another meal after their noodles consumption. But most (29 out of 35) agreed that when they did not have money to prepare a proper meal, noodles become the substitute meal. Children from low SES groups consumed noodles every day, even up to five times a day. The difference between the low SES group and two thirds (23/32) of the carers from the intermediate SES groups whose children had a similar patterns
PAGE | 181 of noodle consumption pattern was that the other one third (9/32) of intermediate SES carers limited the amount of noodles their children consumed. These intermediate SES carers also started taking precautionary action to reduce the potential negative health effects of noodles, e.g. straining the noodles out of the cooking water and placing them in fresh hot water. From what mothers discussed during the FGDs, there is a belief that consuming too many noodles could cause overweight and have other detrimental effects on health. Please refer to Table 2 for detail excerpts.
Compared with the low SES carers whose children consumed instant noodles (IndoMie© or SuperMi©) on a daily basis, the high SES carers utilised noodles as ‘emergency/ occasional food’ when they had no other choice, when their children had a poor appetite because they were sick, or when they forgot to cook the rice at home. These high SES carers (27/27) limited their children’s consumption of noodles for health reasons, such as the detrimental effect of monosodium glutamate and other preservatives. Whenever they can, they prefer to make home-cooked noodles to reduce any of the detrimental effects of instant noodles.
PAGE | 182 Table 2. Quotes on decision on food served at home and attitudes toward instant noodles consumption according to carers in West Java, Indonesia Categories Low SES (FGDs 5, 6, 7, 8) Intermediate SES (FGDs 3, 4, 9, 10) High SES (FGDs 1, 2, 11, 12) Decision on P3: They will say they want grilled fish. But I just P2: I cook whatever I want, Doc. P6: If my children want to eat something, I give food served at give him fried noodles. If there is no fish, just eat them what they want home noodles, okay? P1: Sometimes my grandchild asks for certain food. I will cook it to make sure that he will eat the food. P7: It’s better than having extra food at the table. I P6: I don’t always give them what they want, but if just ask them, what do you want to eat today? I have the materials I will make it. If I don’t have it P7: I ask my child first because if I don’t cook what then I just make what I can with the ingredients I she wants, she won’t eat it. P9: If she was having one of those difficult times (to have. eat), yes, I follow her request, because there’s no P8: I just ask my husband and children what they other way, it’s better than if she doesn’t eat P1: I, uh, have a younger sibling who cares for his want to eat and I follow it. anything. child according to... well, he did go to a Doctor for counseling. So, the Doctor said that a child at this P4: When I’m the one who decides, my child and P3: Our family discusses the menu together. My age, if I’m not mistaken, at about this age, they husband won’t eat and the food is wasted. child would say, “Mom, let’s have this for dinner.” shouldn’t be introduced to salt... or sugar, We cook it together. something like that. So, he was too strict, he followed the theory. I observed that his child has P2: It’s also the mother. I mean, we already buy the grown very little, Doc. The child is so small, and stock, for example I bought the chicken, but I ask the food that his child eats is purely homemade, the children what kind of chicken that they want without sugar or sweetener. Instead of salt, he uses (fried, soup)? broth and made it by himself. Then, when he started working, uh, he moved (to other city) with P5: Oh… Yes, if it’s choosing food, sometimes I his wife and the child was left with their decide it by myself. Because their mother really grandmother. The grandmother was less strict in can’t… cannot make a decision about her child… feeding the child...... But the child’s growth Automatically, I do it by myself. I make some actually became better... variations, now… I cannot just force the child, for example, that they must eat vegetables. If the child does not want to eat vegetables, we still cannot force him. But, the day after tomorrow, I will try it again. Attitudes P8: Noodles should be forbidden, right? Aa. but the P1: I give it to him twice a week. P4: I forbid her from eating instant noodles. toward grandmother gives him noodle and egg everyday. P6: But it’s better than (having) nothing (to eat). instant So his body is big. It is still okay if he only eats P4: I add the rice so he’ll be full. P9: Yes , it’s better than nothing. noodles noodles once a day, right? But he eats noodles in P6: Yes, indeed. When my nephew was sick, I gave consumption the morning, afternoon, and night. He always eats P7: When I cook them, I throw away the water. him noodles. It was better than nothing. noodle. Not to consider the extra rice. P5: I do, too.
PAGE | 183 P6: You throw away the water after you boil it. P7: You can’t feel it now, but in the future, you P5: A child should eat three meals a day right? This P7: And replace it with fresh water. will. Yesterday, I went to the hospital to see my one eats more than three times and she eats noodles, friend’s child after he was hospitalized because he sometimes two packages aren’t enough for her, only P8: I give my children noodles everyday. ate a lot of noodles. Then, I decided to give less four is. noodles to my children. P3: For me, I limit the noodle consumption, Doctor. P7: Even if he has eaten two packs of noodles, if he M: What do you mean by limit it? P1: For me it’s once a month. has not eaten rice, he'll consider it as if he hasn't P3: So, in a week, I only give the children one pack P2: Right, rarely, maybe not even once a month. eaten yet. of noodle. P5: Noodles are only for emergency, when there’s no other food available.
P8: We have noodles every time I forgot to cook the rice. SES: socio economic status FGDs: focus group discussions
PAGE | 184 Decisions on food/ snacks and drinks consumptions Street vending is part of the daily life of Indonesians. Street vendors sell snacks and drinks. The types of snacks sold at street vendors include gorengan (Indonesian fried food consists of vegetables, fruit, or tofu covered by flour), meatballs, porridge, cilok (snack made of steamed flour), and batagor (Indonesian fried dumplings usually served with peanut sauce). The types of drinks sold at street vendors include sweetened tea, ice tea, ice syrup, fruity flavoured ice, and es campur (mixed fruit with ice and condensed milk). In Indonesia, tea, juices and cordials (sweet fruit-flavored drinks) are referred to as sugary drinks.
Carers from low SES groups stated that they and their children chose to buy snacks from street vendors. Even when the carers made their own snacks at home; it was usually fried food (e.g. tofu, bananas, fried vegetables cooked with flour). Most low SES carers (32/35) preferred to buy snacks for reasons varying from affordability to practicality. Please see Table 3 for relevant excerpts. Mothers stated that soda drinks were not widely consumed compared to sugary drinks because they are more expensive.
According to the carers, children from intermediate SES groups preferred to buy snacks from street vendors, although some of the children already knew about hygiene and preferred cleaner vendors who used gloves when handling the food. Carers also preferred to buy the snacks from street vendors, although they made their own snacks during Ramadan (fasting month for Muslims) because it is considered a special time of year. Even so, carers preferred either to buy snacks from “more hygienic” street vendors, especially food that they can see while being prepared, or to buy the snacks from the store (brands that are well known). Most of the carers (26/32) bought the snacks for different reasons, for example time, availability and practicality: saving their energy for other activities. Soda drinks were rarely consumed, such as for the Idul Fitri celebrations at the end of Ramadan. Carers from low and intermediate SES groups described how their children consumed sugary drinks on a daily basis.
Most of the carers (22/27) from high SES groups preferred to buy snacks at the supermarket as part of the monthly grocery shopping list, however a few carers preferred to make their own at home. The two main reasons for buying the snacks was because carers were too busy or not in the mood for snack making. Children of high SES groups consumed soda drinks, some of them drank soda on a daily basis while the rest drank them when dining out with the family (approximately twice a week). This was in contrast with children of low SES groups.
PAGE | 185 High SES carers also stated that they preferred their children to consume snacks like cereal or chocolates, even though they knew the snacks were high in sugar.
Feelings about food/ snacks and drinks consumptions All carers from all three SES groups had negative feelings (worried, annoyed, broken-hearted) towards their children’s food/ drinks consumptions (Table 3). They expressed a strong feeling of helplessness or impotence. They could articulate the concept of unhealthy food, yet they felt they had no other choice but to let the children consume unhealthy food, compared to being faced with the children’s anger or the fact that they would not have anything to eat. These negative feelings were generated by the self-discovered knowledge of the effect of unhealthy food. There were other sources of knowledge, for example TV investigative shows on street food or social media.
PAGE | 186 Table 3. Quotes on decisions on food/ snacks and drinks consumptions and feelings about food/ snacks and drinks consumptions according to carers in West Java, Indonesia Categories Low SES (FGDs 5, 6, 7, 8) Intermediate SES (FGDs 3, 4, 9, 10) High SES (FGDs 1, 2, 11, 12) Decisions on P6: When it is really difficult for him to eat, I make P4: My child told me that at her school, the batagor P9: I put a limit on shopping; I go to the store once food/ snacks him milk puddings so that he will eat that as a vendor uses their hands directly to pick the food. So a week and buy lots of food so my children will not and drinks snack. But it is still difficult, Ma’am. Now, small she doesn't want to buy it anymore. buy snacks outside. I don’t tell them about buying consumptions shops sell something like jellies... The children snacks. If they want to buy something, I have to prefer to buy them from there... The children P3: Fried bananas, bala- bala, especially during know what kind of snack will be bought. She never nowadays prefer to buy (the snacks). Ramadan . I make cireng (fried flattened tapioca eats snacks from the street vendors outside of her flour batter), too elementary school either. P8: Who drinks the Good Day Mochaccino? Often times, it’s for both of us (mother and child). P8: Yes, well, I pick the most visible one. P8: I give my children soda drinks. P7: The clean ones. P6: They drink it, but only a quarter of a cup. P2: Yeah, Teh Gelas (sugary drinks). But Teh P8: And the hygienic ones. I must be able to see it P7: I’ve given it to them. Gelas (sugary drinks) is not that bad, Ma’am. when it is being made. P2: Only when we go out to eat. Unlike those with lots of colors, Ma’am, like Neslo (sugary drinks). P5: I give soda drinks to my child. Especially, on P4: Me? I cook my children’s snacks when the P1: Right, Neslo (sugary drinks), the one that costs Idul Fitri, it’s usually available Hilal showed (Arabic term for significance of the five hundred rupiahs (4 cents USD). P2: Right, it (soda drinks) is a gift; it comes with New Moon in the Islamic lunar calendar. In THR (the money or package given by companies to Indonesia, this means once in a blue moon). P9: My child drinks Teh Gelas everyday, especially the employees on Idul Fitri) when they are at school. P5: If I feel like it, then I make my children’s P8: Same here, mam. When he comes home, he P1: Just the same. I give my child Teh Gelas snacks. already has Teh Gelas on his hands. (sugary drinks) everyday because it’s the cheapest. P3: If I’m in the mood...(then I make the snacks). P2: My children like street food better; they don’t P2: When I don’t have much time, I’ll buy the eat the snacks I cook for them. snacks at the store. P2: I have other things to do. P6: I’m usually busy. P3: I prefer to buy it because it’s affordable. P4: For me, I prefer them to buy something from P2: But the main factor is malas (lack of will/ the store, the snacks that have a brand name on it. unmotivated). P6: Sometimes there’s no money hahaha. The snacks we see on TV, I trust the commercials. P3: Malas . P1: It’s expensive. Even I’m not one hundred percent sure. If my P5: Yes, malas . children buy the snacks that don’t have well-known P2: Even though the ingredients are there, I’m just P7: If I had more money, then I’ll buy ingredients brand, well... (I don’t let my children eat those). too malas to do it. to make my own snacks. P3: I’m malas . P2: Because (buying) it is more practical. P4: Too malas to move. P6: Simple and cheap, snacks cost only one P6: I don’t want to feel tired. thousand rupiahs (8 cents USD). P8: I can save my energy and everything else. P2: I prefer them to eat Koko Krunch (brand of cereal). P9: (Making your own snacks) is inefficient and it P7: Right, and I’m afraid when I cook, my younger P7: Right, something like Koko Krunch and requires extra funds. children will play with the stove (because I have 5 chocolates, even though it contains more sugar.
PAGE | 187 children). P2: Right. Feelings about P1: Yes, it makes me feel like a failure of a mother P3: Worry, I’m actually worried. P9: Well, for me, sometimes I feel worried. food/ snacks (when I see what my children eat), it breaks my P8: Yes, worried. I feel like some of his food and drinks heart. P2: I’d rather he didn’t drink Ale-ale choices are worrying. consumptions P1: It contains preservatives. P9: We know that those chips that have flavors P3: I’m afraid it contains some weird ingredients. P2: On TV, there was a case, uh… like that. sometimes contain something that’s not healthy. They used textile dyes. It contains MSG. P6: On TV, street snacks are shown to have P4: Textile dyes. preservatives and food coloring. They also use P1: In (other sugary) drinks, too. P4: So I feel sad seeing him eating something that ingredients like rotten fish, so I forbid my children contains MSG. I prefer him to eat chocolates. from buying it, but they keep on buying it. P5: The drinks sold at schools are different; they’re mixed with all kinds of things. It makes me scared. P5: I don’t like it (his choices of foods) . P2: I’m worried they will eat that a lot. I see on TV P4: It’s horrifying, isn’t it… that there are lots of unhealthy snacks, we don’t P6: I’m afraid the vendor might use chemicals for P1: Annoyed, I feel so annoyed. make the snacks by ourselves, and moreover if they the drinks. buy it at the market or on the street, sometimes it P3: Yes, chemicals. P8: I’m afraid because the beverages contain a lot has expired. This happened to me when my son P6: And artificial sweeteners. of preservatives. bought Popmie (Indonesian cup noodle brand) and P6: Yes. it didn’t smell right. When I checked it, it had P4: Sad. P3: The preservatives. already expired. He bought it from K (name) and P8: Worried. P8: Yes. she often sells expired snacks, so I always need to P6: I’m afraid he might get addicted. check them and after that, I never bought anything P5: I’m worried and afraid. from her again. P2: His favorite is noodles. But if he eats it P8: I’m afraid of the preservatives. everyday, I feel that it’s not okay. But when I don’t P5: Sometimes they (the street vendors) use plastic P2: For example, the sugar that they use. I mean follow his request, he’ll get mad. He won’t eat. when they fry the snacks. that it’s not real sugar. P8: It’s artificial sweetener. P4: But sometimes in the morning, the child will P4: My child is a picky eater. If... if my child wants P2: The sugar is not a real sugar. ask for fries. But when he saw there’s nothing in to eat something, there’s no other way, Ma’am, it’s P3: It’s saccharine. the refrigerator, and we make something else, he better than having him cry or get mad at school, so prefers to go to school without (having any I have to say yes. P7: If I don’t cook noodles for him, he will whine breakfast). all day. P1: Actually, I think everyone is probably worried (about our children food choices). Everyone is P3: But now, my youngest child, who is four years worried, Doc. But maybe, there is nothing that we old, is a very picky eater. Yesterday, when I didn’t can do about it, I think. give him noodles, he didn’t eat anything. He only ate three spoonful of rice. He didn’t want to eat P3: It’s better than getting them upset. anything. But if there’s noodle, he definitely wants to eat. SES: socio economic status FGDs: focus group discussions
PAGE | 188 Consideration of money issues Both low and intermediate SES carers shared their experiences of how monetary issues interfered with their desires of providing healthy food for the family. One low SES mother (Table 4 column 1) stated that her family ate noodles as main meals every day especially when they were short of money. Another low SES mother shared a story of when her son asked her to have healthier food in the house. She stated that she also wanted to have healthier foods like other families but they did not have the money to do that. If she used the money to buy vegetables then there would not be enough money for the children to go to school, an issue with which her son agreed. A mother from the intermediate SES group stated a similar view.
Two intermediate SES mothers were involved in an argument; one said that you could provide cheap vegetables to the children, and another mother answered that her children would not consume spinach soup without adding any sausage in it, otherwise they would refer to it as “leaves with water”. She explained that all she could afford was the spinach and adding sausages to it would be too expensive and therefore she chose not to provide spinach soup at home.
No mothers from the high SES made any comments around money issues. Instead, they commented about Halal food being important.
Consideration of Halal issues Halal foods are those which are permissible, and which have been processed according to the Islamic law, from the slaughtering of animals until the preparation of the food23. The only group that discussed this issue was the high SES carers. Five carers (out of 27) stated that they had more concerns about the halal status of the food/ snacks than the nutritional values of the food that their children consume. One mother stated that she thought most mothers would agree that they considered Halalness as the most important thing when considering to buy snacks. She added that they knew that snacks would cause overweight/ obesity, but as long as it was Halal then it was acceptable. One mother had a similar view as she said: “It’s okay to be fat as long as it’s a Halal fatness (God allows it).” Relevant excerpts are available in Table 4. We did not find any comments regarding Halal issues from the low and intermediate SES mothers.
PAGE | 189
Table 4. Quotes on considerations of money issues and Halal issues according to carers in West Java, Indonesia Categories Low SES (FGDs 5, 6, 7, 8) Intermediate SES (FGDs 3, 4, 9, 10) High SES (FGDs 1, 2, 11, 12) Consideration P3: Right. I said I also want to do that, eat healthy P5: But I think it’s cheaper if you make some No one mentioned about this issue of money food like the others. But, we don’t have any money vegetables for them? Like spinach soup? issues for that. If I buy vegetables, how about the money P7: P5, If I make spinach soup, the spinach alone for the transportation to school? So he will reply costs only two thousand rupiahs for a bundle. If I “You’re right, Mom”. buy two bundles, then it will costs four thousand rupiahs. That’s not including sausage (so it’s P1: I consider noodles as the primary food, Ma’am. expensive). If I don’t add sausage, the children For us, when we do not have money, we will just would say, “What is this? Is it just leaves with eat noodles. water?”
P8: Just cook from what we have. There’s no choice P4: That’s the reality. I want to change the dietary but what we have. habit (with healthy food) but I don’t have money for that. Hahaha.
P6: If we buy it, I can just eat it right away. If I want to cook it by myself, it’s not enough to buy all the ingredients with just six thousands rupiahs. Consideration No one mentioned about this issue No one mentioned about this issue P2: For me as long as it’s Halal, I don’t really care of Halal if it makes us fat or not. issues P9: Well, for me, every time I buy my children’s snacks, I only search for the Halal logo. That’s it. P4: I do, too. P6: It’s okay to be fat as long as it’s a Halal fatness (God allows it).
P8: I think we all do the same thing, we think the most important thing is whether the snacks are Halal or not. If it’s about fatness, once you consume snacks, whether it’s only a little or in abundance, you will eventually get fat. So as long as the snacks are Halal, it’s fine by me. SES: socio economic status FGDs: focus group discussions
PAGE | 190 DISCUSSION This study aimed to explore the reasons behind Indonesian carers’ decision-making in food consumptions for their pre-adolescent children based on the perspectives from three different socio economic groups. Six themes emerged: 1) Decisions on food served at home. 2) Attitudes toward instant noodles consumption. 3) Decision on food/ snacks drinks consumptions. 4) Feelings about food/ drinks consumptions. 5) Consideration of money issues. 6) Consideration of Halal issues. Although there were similarities and differences within and between the three SES groups, we found more similarities between the low and intermediate SES carers, and more differences between these groups compared to the high SES carers.
Carers from the low and intermediate SES groups made most of the decisions about the food served at home and only sometimes involved their children in the decision-making. This was particularly due to monetary considerations, but they considered family member’s request when they had money. High SES carers stated that the decisions always involved all family members, so that children would eat the food being served. One qualitative study conducted in Australia with parents of children aged 3-5 years also described the significant influence that children made on the food served at home24. Another qualitative study in Canada emphasized the importance of considering all family members’ “likes and dislikes” in taste, stating that mothers preferred to cook food that the family will eat, even if it lacks nutritional value25. A United Kingdom study identified mothers who concentrated on pleasing their children, where one mother stated that she would not spend money on food that her children would not eat26.
Instant noodles are high in carbohydrate, fat and salt, yet low in fibre, vitamins, and minerals22,27. One study among over 10,000 Korean adults found an association between instant noodles consumption and a higher prevalence of metabolic syndrome in women28. In our study, children from low and intermediate SES groups consumed noodles on a daily basis while carers from the high SES group referred to noodles as the occasional food or as their last option. A study with primary school children in Vietnam reported that 50% of these children consumed at least one packet of noodles weekly, with a mean of 2.9 portions weekly29. Several researchers have acknowledged that the increasing trend of noodles consumption in Asia provides a window of opportunity to improve micronutrient intake through fortification of wheat flour – the main ingredient in instant noodles – or through the seasonings inside the noodle packet22,30. In our study, some low and intermediate SES group
PAGE | 191 carers stated that their children consumed instant noodles with rice, or carers added rice to their children’s noodles to make them ‘full”. This practice would result in an excessive intake of carbohydrate in these children. As documented in a study of 2,728 Singaporean adults, a high consumption of noodles combined with rice may result in hyperglycaemia31. In our study, carers from intermediate and high SES groups were trying to limit their children’s noodles consumption because they considered noodles to be unhealthy food. Similar evidence was found in a study undertaken in three European cities (Maastricht, Liege, Aachen) where higher and middle class carers were more likely to limit their children’s consumption of unhealthy food compared to the lower SES carers in the same study32.
Carers from high SES groups stated that the factors that prevented them from making their children’s food/ snacks were time limitations, not being in “the mood” or “not feeling like it”, or simply “malas (unmotivated/ lack of will)”. Even though all the participants were housewives or stay at home carers, there were other factors that might play a role. For example the number of children in a family may be important, as one mother stated that she was afraid if she was busy in the kitchen making food/ snacks, then no one could watch her children and one of her five children might play with the stove in the kitchen. In addition, none of our low and intermediate SES mothers had household help, so they were also responsible for household cleaning and washing and ironing the clothes. Several qualitative studies in a review by Pocock et al. had similar findings33. Mothers stated that being “tired or unmotivated” prevented them from preparing healthy menus. In a study of the perspectives of American children (aged 11-14 years) on this topic children reported that one of the main reasons for eating out or ordering take-away food was because “mom does not feel like cooking”34. Other evidence of time limitation being a factor in providing healthy food comes from two qualitative studies and two reviews of qualitative studies25,33,35,36.
In our study, most carers from low and intermediate SES groups reported that their children consumed large quantities of sugary drinks and only children from the high SES consumed soda drinks. Similar findings from one Australian study and one North American study showed parents consider sugary drinks as “everyday” drinks24,37 and soda drinks as a “sometimes” drink24. Two other North American studies showed that sweetened beverages and soda drinks were major contributors to higher daily intakes of energy and sugars38,39. This may be one of the contributors to the increasing prevalence of childhood overweight in Indonesia.
PAGE | 192 All carers, regardless of SES, shared the same concerns and negative feelings about their children’s choices and consumptions of food/ snacks and drinks. They admitted that they were worried about these choices and consumption level, but felt they cannot do anything about it since the unintended consequences would be the children not eating anything throughout the day or being upset/ crying at school. One systematic review documented that 9 studies showed carers’ food decisions were largely influenced by their feelings regarding their children’s health35.
Money appeared to be an important issue in the decision-making to provide healthier food in the low and intermediate SES mothers. This is comparable to other studies where carers were concerned that they would not have enough money to buy healthy food/ ingredients24,26,40,41. The need to put quantity over quality when money is limited, is a finding also documented elsewhere41. One qualitative study with 11-14 year old U.S. children participants in a Northeastern suburb highlighted that monetary issues affects their parents’ ability to buy food34. One possible consideration for the Indonesian government is ways to improve their citizens’ accessibility to healthy food. Strategies that have been proven effective in Canada include collaborations with supermarket chains to provide discount coupons to buy fruits and vegetables, or the introduction of farmers’ markets in low- and middle-income areas to ensure healthier food availability and affordability42,43.
Several high SES carers discussed Halal issues. One study in the United Kingdom reported how Muslim families checked the ‘Halalness’ of ingredients on the food labels or of meat and other meat products as a routine activity. These families could also check the dietary information from the food label, although this was not seen as a priority44. This might provide an opportunity for the Indonesia government to educate the community to look at the dietary information when they are checking on the Halal logo.
In contrast, we found no comment about Halal issues in the low and intermediate SES groups. This may be due to several reasons: 1) Halal-ness is a given condition in Indonesia, 2) It is related to a hierarchy of choices, and 3) the vast majority of food is considered Halal in Indonesia (it is very rare to find non-Halal food in traditional markets, unless in cities where the majority of the population are non-Muslim).
Although we have discussed these six categories independently, we realize that all six were interrelated and affect each other in the decision-making process. For example money issues
PAGE | 193 have an impact on decisions on food served at the table and intersect with noodles consumption. Another example is how every SES group bought snacks/ drinks from street vendors, but their feelings about snack choices and money issues also influenced these decisions. For example, children from intermediate SES groups preferred “more hygienic’ street vendors that might be more expensive. We did not find any noteworthy differences between carers of under five children and those of elementary school children. One of the reasons could be that some mothers had under five and school age children in the same household.
A strength of our study on food choices is that we accessed different views from mothers/ carers of low, intermediate, and high socio economic status, despite some researchers considering it a challenge to include participation from low SES participants26,45. Another strength is that this is the first study in Asia to document that some mothers give more attention to the Halal status, instead of the nutritional value, of the food/ snacks that their children consume. This finding is likely to be found in other Muslim countries. From the methodological perspective, interaction between participants allowed us to capture different ideas from these carers. We quantified participants’ comments to make our findings clearer and we ensured the quality of our analysis by performing triangulation and reducing interpretive bias. One limitation of our study is that we did not explore the nutrition knowledge of carers.
CONCLUSIONS In summary, this study presents evidence that food choices are influenced by many factors, particularly socioeconomic status and the requests of children and/ or other family members. Differences were found in the factors that influence those decisions. We found unhealthy food consumption (e.g. instant noodles) in the low SES families was strongly associated with carers’ monetary issues. On the other hand, high SES carers placed more emphasis on “Halal- ness” over nutritional values in selecting food/ snacks for their children. Future policies should incorporate pathways to reduce the price of healthier food ingredients, thus reducing the impact of economic constraints especially for the low and middle-income families.
It will take multi sectoral collaboration (e.g Departments of Health, Farming, Education, and Urban Planning) to implement policies aimed at improving the availability of healthier food. The Indonesian government might need to build collaborations with universities, as well as colleges and schools, to augment the education of health professionals to reach families and
PAGE | 194 educate them about age-specific nutrition-effective strategies. Other ways to improve carers’ and health professionals’ knowledge around overweight/ obesity can be conducted through professional training, conferences or seminars for parents. Education programs around healthy eating could also be included in school-curricula throughout Indonesia. Another aspect to consider is how to implement behavior modification in feeding practices that will inform healthy family dietary habits. Given that most Indonesians are Muslim, the role of religion may need to be considered in the development and implementation of food policies. Future studies should consider other aspects of food consumption, such as the nutrition knowledge of carers and carers’ consideration of children’s growth in deciding food consumption. Research with adolescents might raise other factors related to food choices.
ETHICAL CLEARANCE We obtained ethics approval from the Ethics Committee at Fakultas Kedokteran Universitas Padjadjaran, Bandung (no. 473/UN6.C1.3.2/KEPK/PN/2016) dated 16th of June 2016. We also obtained further permits from three health departments in each area to start the study in the Puskesmas and schools.
ACKNOWLEDGEMENTS CNR would like to thank Lembaga Pengelola Dana Pendidikan (LPDP), the Republic of Indonesia for her PhD scholarship.
FUNDING This study was funded by The Children’s Hospital at Westmead (CHW), Australia. CHW did not have any role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
PAGE | 195 REFERENCES
1. World Health Organization. The double burden of malnutrition. Policy brief. Geneva: World Health Organization; 2017. 2. Shrimpton R, Rokx C. The double burden of malnutrition: a review of global evidence. Washington DC: World Bank; 2012. Report No.: 79525. 3. International Food Policy Research Institute. From promise to impact: ending malnutrition by 2030. Washington, DC: International Food Policy Research Institute; 2016. 4. Shrimpton R, Rokx C. The Double Burden of Malnutrition in Indonesia. Jakarta, Indonesia: World Bank Jakarta; 2013. 5. International Food Policy Research Institute. Global Nutrition Report 2014: Actions and accountability to accelerate the World's Progress on Nutrition. Washington, DC: International Food Policy Research Institute; 2014. 6. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, Underweight and Overweight in Children Aged 2.0-4.9 Years in Indonesia: Prevalence Trends and Associated Risk Factors. PloS One 2016;11:e0154756. 7. Badan Penelitian dan Pengembangan Kesehatan KKRI. Status Gizi. In: Indonesia DK, ed. Laporan Hasil Riset Kesehatan Dasar Indonesia tahun 2013, Riskesdas Dalam Angka. Jakarta, Indonesia: CV Kiat Nusa; 2014:386 - 415. 8. Rachmi CN, Li M, Baur L. Overweight and obesity in Indonesia: prevalence and risk factors–a literature review. Public Health 2017; 147: 20-29; doi: 10.1016/j.puhe.2017.02.002. 9. Crombie IK, Kiezebrink K, Irvine L, et al. What maternal factors influence the diet of 2-year-old children living in deprived areas? A cross-sectional survey. Public Health Nutrition 2009;12:1254-60. 10. Gross SM, Pollock ED, Braun B. Family influence: key to fruit and vegetable consumption among fourth- and fifth-grade students. Journal of Nutrition Education and Behaviour 2010;42:235-41. 11. Pliner P. Cognitive schemas: how can we use them to improve children's acceptance of diverse and unfamiliar foods? The British Journal of Nutrition 2008;99 Suppl 1:S2-6. 12. Wansink B. Profiling nutritional gatekeepers: Three methods for differentiating influential cooks. Food Quality and Preference 2003;14:289-97. 13. Usfar AA, Lebenthal E, Atmarita, Achadi E, Soekirman, Hadi H. Obesity as a poverty-related emerging nutrition problems: the case of Indonesia. Obesity Reviews 2010;11:924-8. 14. Julia M, van Weissenbruch MM, de Waal HAD-v, Surjono A. Influence of socioeconomic status on the prevalence of stunted growth and obesity in prepubertal Indonesian children. Food & Nutrition Bulletin 2004;25:354-60. 15. Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. International Journal of Behavioral Nutrition and Physical Activity 2017;14:1-13. 16. Indeks Pembangunan Manusia. Statistics of Jawa Barat, 2016. (Accessed March, 2016, at https://jabar.bps.go.id/Subjek/view/id/26 - subjekViewTab3|accordion-daftar- subjek1.) 17. Operational Working Group Ministry of Health Republic Indonesia. Pedoman Umum Pengelolaan Posyandu. In: Health Mo, ed. Jakarta, Indonesia2011:6-7. PAGE | 196 18. Strauss A. Qualitative analysis for social scientist. Cambridge: Cambridge University Press; 1987. 19. Strauss A, Corbin J. Basics of qualitative research: grounded theory procedures and techniques. Newbury Park, CA: Sage; 1990. 20. Glaser B, Strauss A. The discovery of grounded theory: strategies for qualitative research. Chicago: Aldine; 1967. 21. NVivo qualitative data analysis Software. QSR International Pty Ltd.; 2014. 22. Gulia N, Dhaka V, Khatkar BS. Instant noodles: processing, quality, and nutritional aspects. Critical Reviews in Food Science and Nutrition 2014;54:1386-99. 23. Khalek AA, Ismail SHS. Why are we eating Halal - using the theory of planned behavior in predicting halal food consumption among generation Y in Malaysia. International Journal of Science and Humanity 2015;5:608-12. 24. Petrunoff NA, Wilkenfeld RL, King LA, Flood VM. 'Treats', 'sometimes foods', 'junk': a qualitative study exploring 'extra foods' with parents of young children. Public Health Nutrition 2014;17:979-86. 25. Slater J, Sevenhuysen G, Edginton B, O'Neil J. 'Trying to make it all come together': structuration and employed mothers' experience of family food provisioning in Canada. Health Promotion International 2011;27:405-15. 26. Hardcastle SJ, Blake N. Influences underlying family food choices in mothers from an economically disadvantaged community. Eating Behaviors 2016;20:1-8. 27. Park J, Lee JS, Jang YA, Chung HR, Kim J. A comparison of food and nutrient intake between instant noodle consumers and non-instant noodle consumers in Korean adults. Nutrition Research and Practice 2011;5:443-9. 28. Shin HJ, Cho E, Lee HJ, et al. Instant noodle intake and dietary patterns are associated with distinct cardiometabolic risk factors in Korea. J Nutr 2014;144:1247-55. 29. Huong TL, Brouwer I, Nguyen K, Burema J, Kok F. The effect of iron fortification and de-worming on anaemia and iron status of Vietnamese schoolchildren. British Journal of Nutrition 2007;97:955-62. 30. Bronder KL, Zimmerman SL, van den Wijngaart A, Codling K, Johns KA, Pachon H. Instant noodles made with fortified wheat flour to improve micronutrient intake in Asia: a review of simulation, nutrient retention and sensory studies. Asia Pacific Journal of Clinical Nutrition 2017;26:191-201. 31. Zuniga YL, Rebello SA, Oi PL, et al. Rice and noodle consumption is associated with insulin resistance and hyperglycaemia in an Asian population. The British Journal of Nutrition 2014;111:1118-28. 32. Hupkens C, Knibbe R, Drop M. Social class differences in food consumption: The explanatory value of permissiveness and health and cost implications. European Journal of Public Health 2000;10:108-13. 33. Pocock M, Trivedi D, Wills W, Bunn F, Magnusson J. Parental perceptions regarding healthy behaviours for preventing overweight and obesity in young children: a systematic review of qualitative studies. Obesity reviews : an official journal of the International Association for the Study of Obesity 2010;11:338-53. 34. Holsten JE, Deatrick JA, Kumanyika S, Pinto-Martin J, Compher CW. Children's food choice process in the home environment. A qualitative descriptive study. Appetite 2012;58:64-73.
PAGE | 197 35. Paes VM, Ong KK, Lakshman R. Factors influencing obesogenic dietary intake in young children (0–6 years): systematic review of qualitative evidence. BMJ Open 2015;5:e007396. 36. Tucker P, Irwin JD, He M, Bouck LM, Pollett G. Preschoolers' dietary behaviours: parents' perspectives. Canadian Journal of Dietetic Practice and Research 2006;67:67-71. 37. Neumark-Sztainer D, MacLehose R, Loth K, Fulkerson JA, Eisenberg ME, Berge J. What's for dinner? Types of food served at family dinner differ across parent and family characteristics. Public Health Nutrition 2012;17:145-55. 38. Nicklas TA, O'neil CE, Fulgoni VL. Relationship between snacking patterns, diet quality, and risk of overweight and abdominal obesity in children. International Journal of Child Health and Nutrition 2013;2:189-200. 39. Piernas C, Popkin BM. Trends in snacking among U.S. children. Health Affairs (Project Hope) 2010;29:398-404. 40. Styles JL, Meier A, Sutherland LA, Campbell MK. Parents' and caregivers' concerns about obesity in young children: a qualitative study. Family and Community Health 2007;30:279-95. 41. Dammann KW, Smith C. Factors affecting low-income women's food choices and the perceived impact of dietary intake and socioeconomic status on their health and weight. Journal of Nutrition Education and Behaviour 2009;41:242-53. 42. Mah CL, Thang H. Cultivating Food Connections: The Toronto Food Strategy and Municipal Deliberation on Food. International Planning Studies 2013;18:96-110. 43. Mah CL, Cook B, Rideout K, Minaker LM. Policy options for healthier retail food environments in city-regions. Canadian Journal of Public Health 2016 2016;107(Suppl. 1):eS64–eS7. 44. Rawlins E, Baker G, Maynard M, harding S. Perceptions of healthy eating and physical activity in an ethnically diverse sample of young children and their parents: the DEAL prevention of obesity study. Journal of Human Nutrition and Dietetics 2012;26:132- 44. 45. Eli K, Howell K, Fisher PA, Nowicka P. "A little on the heavy side": a qualitative analysis of parents' and grandparents' perceptions of preschoolers' body weights. BMJ Open 2014;4:e006609.
PAGE | 198
Chapter 10: Conclusions and recommendations
10.1 KEY FINDINGS 200 10.2 FUTURE DIRECTIONS 201 10.2.1 Future directions based on findings presented in this thesis 202 10.2.1.1 Policies/ implementation 203 10.2.1.2 Future research 206 10.2.2 Additional recommendations to be considered by all levels of 207 government in Indonesia 10.3 REFERENCES 208
PAGE| 199
Chapter 10: Conclusions and recommendations
The double burden of malnutrition (DBM) is happening in Indonesia as well as many other low- and middle-income countries. To overcome the DBM, an understanding of current trends, risk factors and community perceptions is crucial. The overall aim of this thesis was to provide a better understanding of the double burden of malnutrition and especially stunting and overweight/ obesity in children in Indonesia.
10.1 KEY FINDINGS 1. We documented that the double burden of malnutrition exists in Indonesian children aged 2.0 to 4.9 years at the population level, with a decreasing prevalence of stunting (from 50.8% to 36.7%) and underweight (from 34.5% to 21.4%) and an increasing prevalence of overweight/ obesity (from 10.3% to 16.5%) over the 14-year period from 1993 to 2007. There was a shift to the right in the distribution of BMI-z-score at the four time points of 1993, 1997, 2000, and 2007, indicating that Indonesian children became heavier over the years. Similar risk factors for being stunted or underweight were having a low birth weight (<2.5 kg), a longer breastfeeding period (equal to or more than 6 months), having one or both parents who were stunted or underweight, and mothers with no formal education. A child was also at higher risk of being stunted when he/ she lived in a rural area. The prevalence of overweight/ obesity was higher in younger children (2.0 to 2.9 years), males, children with one or both parents with overweight/ obesity, and children whose fathers had completed university level of education1.
2. The double burden of malnutrition also occurred in Indonesian children at the individual level, where there were inconsistent trends in the prevalence of concurrent stunting and overweight in children aged 2.0 to 4.9 years at the four different time points of 1993, 1997, 2000, and 2007. Associated risk factors were being in the youngest age group (2.0 to 2.9 years), being breastfed for 6 months or longer, living in a rural area, or having a mother who was stunted. The cross-sectional studies showed that at four different time points, stunted children were more likely to have overweight/ obesity compared to the healthy height children, with an Odds Ratio >12.
PAGE| 200
3. The longitudinal studies showed contrasting evidence in that stunted children were not more likely to have overweight/ obesity compared to their healthy height peers. There were also no associations between stunting in early childhood (2.0 to 4.9 years) and being thin or having high blood pressure in the adolescent years (i.e. 7 or 14 years later)3.
4. To understand the reasons behind the increasing prevalence of overweight/ obesity, focus group discussions with primary carers (mothers/ grandmothers) of under-five and elementary school aged children were conducted to elicit their perceptions around childhood overweight. These carers came from three different socioeconomic status areas. Three main themes emerged from the discussions: the concept of overweight, factors contributing to overweight, and awareness and feelings towards overweight in children4.
5. Using the focus groups discussions, we also investigated the reasons underlying the decision-making processes of mothers/ grandmothers in determining their children’s food choices. Six themes emerged: decisions on food served at home, attitudes toward instant noodles consumption, decisions on food/ snacks and drinks consumptions, feelings about food/ drinks choices, consideration of money issues, and consideration of Halal issues.
10.2 FUTURE DIRECTIONS There are several World Health Organization and UNICEF documents aimed at tackling malnutrition5-9. As previously elaborated in Chapter 2, the WHO stunting policy brief proposed that interventions be conducted through four main platforms, with a focus on the first 1000 days of life5. These platforms are breastfeeding, complementary feeding, collaboration between nutrition-specific and nutrition-sensitive interventions, and multi sectoral collaborations on the program level5.
WFP and UNICEF’s Ending Child Hunger and Undernutrition Initiative propose four main pillars for their global framework for action, especially by supporting regional, national and local efforts9. These pillars are increasing awareness around hunger and undernutrition, strengthening national policies, increasing capacities in communities especially those that were affected with child hunger and undernutrition, and increasing the efficiency and accountability within different stakeholders and partners9.
The WHO childhood obesity policy brief suggests the best solution to address the rising prevalence of childhood obesity should be tailored to the condition of each country, and more
PAGE| 201 specifically, to different parts of the country6. For example, in areas where many schools do not provide lunch services it is less important to develop school food policies. Likewise, it is less important to target changes through TV advertising in areas where TV watching is not a common practice6.
The double burden of malnutrition and double-duty action for nutrition policy briefs highlight how governments and communities should find actions/ interventions that are already in place that will have an effect in tackling both under- and overnutrition at the same time7,8.
10.2.1 Future directions based on findings presented in this thesis Figure 10.1 shows a summary of possible future directions for all levels of Indonesian governments in the areas of policies and research. They are based on the evidence from the quantitative and qualitative studies presented in this thesis5-8,10.
FUTURE DIRECTIONS
POLICIES/ IMPLEMENTATION RESEARCH
1. Ensure availability, accessibility and Cross-sectional research: affordability of healthy food 1. dietary intake 2. screen time 3. physical activities 2. Interventions aimed at double-duty 4. environmental factors influencing dietary actions intake, screen time and physical activities
3. National dietary guidelines that are Longitudinal research: easy to comprehend 1. previously stunted children and their chance of being overweight/ obese in adulthood 2. whether socioeconomic status influences a 4. Improve the surveillance system to child's body mass index/ risk of being monitor malnutrition overweight/ obese
6. Make surveillance data readily Qualitative research: accessible to monitor trends and adjust 1. surveys at city/ regency level to inform tailored interventions interventions 2. understanding of current national guidelines on balanced nutrition 3. views of existing interventions/ strategies
Figure 10.1 Future directions and recommendations for Indonesia
PAGE| 202 10.2.1.1 Policies/ implementation In terms of policies/ implementation, then it will be important for several different ministries at all levels of government within Indonesia to develop policies to ensure the availability, accessibility and affordability of healthy food, especially in low- and middle-income neighbourhoods. In our qualitative work, the issue of the accessibility of healthy food and drinks was raised. For example, more socially disadvantaged families could not afford good quality food and chose instant noodles and soda drinks. The availability, accessibility and affordability of healthy food does not depend solely on the policies developed by governments; it takes multiple parties to make this work, for example food producers, food industries, and schools11.
One example provided by a paper in the Lancet describes how an individual’s food choices are influenced by a complex array of factors including different levels of government, food producers, schools as institutions, the type and regulation of marketing of food especially those aimed at children, as well as food retailers (Figure 10.2)11.
The second recommendation is for all levels of government to identify interventions, both existing and new, that have the potential to include ‘double-duty actions’, i.e. they are aimed at reducing the prevalence of under- and overnutrition simultaneously7-8. One of the most obvious actions is support for the early initiation of breastfeeding and for exclusive breastfeeding for the first six months of a child’s life. Breastfeeding should be continued until a child is two years of age, with appropriate complementary feeding practices7-8.
PAGE| 203
Figure 10.2 How can government support healthy food preferences?11 Available from: http://www.thelancet.com/infographics/obesity-food-policy
The next recommendation is for the Indonesian government to investigate the current national dietary guidelines and generate a simpler version for the community. Based on my qualitative study findings, most carers refer to the empat sehat lima sempurna (four healthy five perfect) guideline, which was developed in 1952 and has since been superseded. This guideline emphasises the consumption of rice, side dishes (meat or other source of protein), vegetables, fruit and look at milk as an item that will make the diet “perfect”. The newest dietary
PAGE| 204 guideline, developed in 1995, is called Pedoman Gizi Seimbang (Balanced Nutrition Guidelines – PGS, previously called PUGS). This guideline proposes a daily food intake that contains nutrients in the type and amount that suits the needs of the body (Figure 10.3). In addition, PGS gives attention to four main principles, namely: become familiar with eating a variet of foods, maintain a clean lifestyle, the importance of active lifestyles and sports, and monitor body weight. However, PGS has 13 messages that might be harder for carers to remember: 1. Eat a wide variety of foods. 2. Eat food to meet energy needs. 3. Eat a carbohydrate source of half of your energy needs. 4. Limit your consumption of fats and oils to a quarter of your energy needs. 5. Use the iodized salt. 6. Eat food sources of iron. 7. Give only breast milk to infants until the age of 4 months. 8. Get used to having breakfast. 9. Drink plenty of clean and safe water. 10. Do physical activity and exercise regularly. 11. Avoid drinking alcoholic beverages. 12. Eat foods that are safe for health. 13. Read the packaged food label12-14. The Indonesian government might also consider ways to to improve the community’s awareness around nutrition and healthy lifestyle recommendations.
The fourth recommendation is for all levels of Indonesian government to improve the surveillance system to monitor malnutrition. This would include educating mothers and kader (voluntary health workers in the community) about the importance of regular measurements of the children’s length/ height and weight, for initial screening purposes. Any child health examination should include measurements of linear growth and an explanation to parents of their child’s current growth and development. These efforts will help early detection of, and therefore early treatment for, malnutrition.
The last recommendation is to ensure that surveillance data are readily accessible for all stakeholders. This will allow academics to do their part to help all levels of government within Indonesia to monitor trends and allow the various levels of government to adjust the double- duty intervention actions accordingly.
PAGE| 205
Figure 10.3 Balanced Nutrition Guidelines12. Available from: http://gizi.depkes.go.id/download/Pedoman%20Gizi/PGS%20Ok.pdf
10.2.1.2 Future research The results from my studies have highlighted specific recommendations for main areas of research: cross sectional studies, longitudinal studies, and qualitative studies. Cross-sectional research should cover other risk factors related to childhood obesity in Indonesia, e.g. more aspects of dietary intake, screen time, physical activities, and the physical activity environment (access to playgrounds, parks, sidewalks for pedestrians, bike lanes, etc.). Cross sectional research should also combine information from existing data repositories to complement each other.
Future longitudinal research should explore outcomes of previously stunted children and their chance of being overweight/ obese in adulthood. Other areas to explore include understanding
PAGE| 206 the influence of socioeconomic status on a child's BMI/ chance of overweight/ obesity in the future.
For future qualitative research, it would be interesting to conduct stakeholder interviews at the city/ regency level to inform the development of tailored interventions relevant to the areas’ needs and capabilities. Other research should be conducted to identify community members’ awareness of the current balanced nutrition guidelines (Pedoman Gizi Seimbang) and related attitudes and practices around it. Academics should consider conducting qualitative research around existing intervention/ strategies to see what works and what does not work, and the associated barriers and enablers. This will provide important feedback to all levels of Indonesian government to support decision-making.
10.2.2 Additional recommendations to be considered by all levels of government in Indonesia All levels of government of Indonesia should consider the following actions: 1. Conduct qualitative research with policy-makers within all levels of government to better understand the challenges in implementation of existing programs/ policies, 2. Conduct policy mapping around the issue of the double burden of malnutrition. This will provide all levels of government with an overall view of available strategies and potential opportunities to combine one or more interventions to support implementation. It may provide opportunities for collaboration between provinces or cities in the same province. 3. Re-invigorate the program of family planning. This program was proven to be a successful approach in previous years with their “two children is enough and two years between each child” slogan. This way, families are able to ensure each child can develop optimally and they can afford to raise their children. 4. Focus interventions on adolescent girls, especially in areas where early marriage is commonly practiced. As a supporting act, the government should also improve the socialisation of health and social consequences of early marriages to prevent having teenage brides. 5. Improve public spaces to support healthy lifestyles, for example bike lanes, pedestrian lanes, parks and outdoor facilities. 6. Improve all interventions related to hygiene and sanitation, as well as educating the community about the importance of working together to improve all community-based programs. These efforts might include a more prominent campaign around hand-washing using soap, safe water practices, defecation education, and many other interventions.
PAGE| 207 10.3 REFERENCES
1. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, Underweight and Overweight in Children Aged 2.0-4.9 Years in Indonesia: Prevalence Trends and Associated Risk Factors. PloS one 2016;11:e0154756. 2. Rachmi CN, Agho KE, Li M, Baur LA. Stunting coexisting with overweight in 2.0- 4.9-year-old Indonesian children: prevalence, trends and associated risk factors from repeated cross-sectional surveys. Public Health Nutr 2016:1-10. 3. Rachmi CN, Agho KE, Li M, Baur LA. Are stunted young Indonesian children more likely to be overweight, thin, or have high blood pressure in adolescence? Int J Public Health 2016. 4. Rachmi CN, Hunter CL, Li M, Baur LA. Perceptions of overweight by primary carers (mothers/grandmothers) of under five and elementary school-aged children in Bandung, Indonesia: a qualitative study. International Journal of Behavioral Nutrition and Physical Activity 2017;14:1-13. 5. World Health Organization. WHO Global Nutrition Target: Stunting Policy Brief. Geneva:WHO; 2014. 6. World Health Organization. Global Nutrition Target: Childhood Overweight Policy Brief. Geneva: WHO; 2014. 7. World Health Organization. The Double Burden of Malnutrition Policy Brief. Geneva: World Health Organization; 2017. 8. World Health Organization. Double-duty Actions for Nutrition Policy Brief. Geneva: World Health Organization; 2017. 9. World Food Programme and United Nations Children's Fund. Global Framework for Action: Ending Child Hunger and Undernutrition Initiative: World Food Programme and United Nations Children's Fund; 2006. 10. Shrimpton R, Rokx C. The Double Burden of Malnutrition in Indonesia. Jakarta, Indonesia: World Bank Jakarta; 2013. 11. Hawkes C, Smith TG, Jewell J, et al. Smart food policies for obesity prevention. Lancet 2015;385:2410-21. 12. Ministry of Health Republic of Indonesia. Pedoman Gizi Seimbang. Jakarta, Indonesia: Ministry of Health Republic of Indonesia; 2014. 13. Direktorat Jenderal Bina Kesehatan Masyarakat, Departemen Kesehatan Republik Indonesia. Pedoman Umum Gizi Seimbang (PUGS) - Panduan untuk Petugas. In: Departemen Kesehatan Republik Indonesia, ed. Jakarta, Indonesia: Depkes RI; 2003. 14. Inilah Perbedaan '4 Sehat 5 Sempurna' Dengan 'Gizi Seimbang' Kementrian Kesehatan Republik Indonesia, 2016. (Accessed July, 2016, at http://www.depkes.go.id/article/view/16051300001/inilah-perbedaan-4-sehat-5-sempurna- dengan-gizi-seimbang-.html.)
PAGE| 208
Appendix
Appendix - Ethics approval from the Ethics Committee at Fakultas Kedokteran Universitas Padjadjaran, Bandung
PAGE| 209