The association between parental and before and the development of offspring overweight and obesity in childhood, adolescence and young adulthood

Nurzalinda Zalbahar

BSc, MSc

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2016 School of Public Health

Abstract

Background and aims: Parental obesity is one of the important factors that is strongly linked with the development of offspring obesity. When both parents are obese, the risk of offspring being obese is greater than when one parent is obese. Irrespective of the exposure time of parental overweight and obesity, studies have consistently found that parental overweight and obesity is positively associated with offspring overweight and obesity. Few longitudinal birth-cohort studies have examined pre-pregnancy parental overweight or obesity and whether it is positively associated with offspring overweight and obesity in the long term. Additionally, there are limited studies reporting on the association of prenatal parental obesity with offspring body mass index (BMI) change, and even less is known about the association between long-term post-partum parental BMI change and the risk of offspring obesity. The aim of the current study is to investigate the association between parental obesity and offspring obesity over the life course, using data from birth-cohort studies. This study examines the longitudinal association between prenatal parental BMI and offspring BMI and adiposity in infancy, childhood, adolescence and adulthood. This study also investigates the association between prenatal parental BMI and offspring BMI change in childhood, adolescence and adulthood. Finally, this study adds to research examining the association between 21 years maternal post-partum weight change (PPWC) and the risk of adulthood obesity in offspring.

Methods: Data are from two cohorts-Universiti Sains Malaysia Pregnancy Cohort which consists of 145 respondents who provided details of pre-pregnancy parental weight and height and weight and length of offspring at the 12-month follow-up; and the Mater-University of Queensland Study of Pregnancy (MUSP) which consists of 1494 participants with complete weight and height for pre- pregnancy paternal–maternal and offspring weight and height at 5, 14 and 21 years. A series of models were used to test the association between prenatal parental BMI and offspring outcomes to adjust the potential covariates obtained from either cohorts (e.g. maternal factors: gestational weight gain; offspring factors: birth weight, TV watching hours at 14 years). Multiple linear regression, logistic regression and multinomial regression were used (continuous or categorical predictors) to examine the association between pre-pregnancy parental BMI and its categories, and offspring BMI and its categories. Multiple imputation was used to address the issue of missing values in covariates. Generalized estimating equation models were used to examine the longitudinal parental– offspring BMI association in infants and were again used to determine the differential effect size of prenatal parental BMI on childhood, adolescence and adulthood. Finally, we analysed the ii

association between 21 years maternal postpartum BMI change and offspring BMI and waist circumference (WC at 21 years using multiple linear regression and multinomial logistic regression.

Results: Major findings are as follows: 1) A significant association was found between maternal BMI and child’s weight for age z-score (WAZ), weight for height z-score (WHZ) and BMI for age z-score (BAZ) at 12 months; although it was attenuated in the fully adjusted model. 2) After adjustment for confounders, for each unit increase in paternal and maternal BMI, the BMI of young adult offspring increased by 0.33 kg/m2 and 0.35 kg/m2, and the waist circumference (WC) increased by 0.76 cm and 0.62 cm, respectively. Offspring at 21 years were at six times more at risk of being overweight/obese (OW/OB) when both parents were OW/OB, compared to offspring of normal-weight parents. 3) A total of 14.7% of the offspring experienced BMI change from normal at 5 years to OW/OB at 14 years, 15.3% from normal at 14 years to OW/OB at 21 years. Overall, the strength of the association of parental BMI with offspring BMI was stronger as offspring became older. When both parents were OW/OB, these associations were stronger than when one parent was OW/OB. 4) In two decades, mothers who had highest weight change were associated with adult offspring’s risk of OW/OB and abdominal obesity with odds of 1.47 (95%CI: 1.08–2.00) and 1.50 (95%CI: 1.10–2.02), respectively.

Conclusions: These findings suggest that parental BMI before pregnancy has a long-term association with offspring BMI from early childhood to adulthood. We found mother–offspring and father–offspring associations were of similar strength and direction. There was no offspring male– female differential association. Our study further demonstrated that parental pre-pregnancy nutritional status is an important predictor in early life that influences offspring overweight and obesity throughout life. Intervening to reduce the prevalence of overweight or obesity in early life by targeting parents may have potential benefits. Encouraging adults to maintain a healthy lifestyle in this challenging obesogenic environment may benefit both themselves and subsequent generations. Studies are warranted to further explore parental lifestyle (dietary and physical activities) before pregnancy and the impact of intervention to improve parental lifestyle on offspring outcomes in the long term.

iii

Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

iv

Publications during candidature

Peer-reviewed papers

Published or accepted 1. Zalbahar N, Mohamed Jan HJ, Loy SL, Najman J, McIntyre HD, Mamun A. Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia. Journal of Obesity Research and Clinical Practice. 2015; In Press 2. Zalbahar N, Najman J, McIntyre HD, Mamun A. Parental pre-pregnancy BMI influences on offspring BMI and waist circumference at 21 years. Australian New Zealand Journal of Public Health. (accepted 24th May 2016)

Book Chapters

No publications

Conference abstracts

1. 8th World Congress on Developmental Origin of Health and Disease 2013 conference, Singapore. 17th – 20th Nov 2013. Title: Impact of pre-pregnancy parental overweight and obesity on offspring obesity: systematic analysis and meta-analysis. (Oral)(Abstract published in Annals of Nutrition and , Sept. 2013) 2. Abstract accepted by 12th Asian Congress of Nutrition 2015 conference, 14th – 18th May 2015. Yokohama, Japan. Title: Maternal-paternal pre-pregnancy obesity associated with offspring obesity in long-term. (Poster)

v

Publications included in this thesis

Publication citation 1 – incorporated as Chapter 4. Publication 1 Zalbahar N, Najman J, McIntyre HD, Loy SL, Hamid Jan JM, Mamun A. Citation 1 Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia. Obesity Research and Clinical Practice. 2015; doi: 10.1016/j.orcp.2015.08.002 Author Development of research question: Zalbahar N (90%), Mamun A (10%) contribution Data management and statistical analysis: Zalbahar N (85%), Loy SL (10%), Mamun A (5%) Interpretation of results: Zalbahar N (75%), Mamun A (15%), %), McIntyre HD (5%), Najman J (5%) Wrote the manuscript: Zalbahar N (100%) Edited the manuscript: Zalbahar N (70%), Mamun A (10%), Loy SL (10%), McIntyre HD (5%), Najman J (5%)

Publication citation 2 – incorporated as Chapter 5. Publication 2 Zalbahar N, Najman J, McIntyre HD, Mamun A. Prenatal parental overweight Citation 2 and obesity and the risk of offspring overweight and obesity at 21 years. Australian New Zealand Journal of Public Health. (Accepted, 24th May, 2016) Author Development of research question: Zalbahar N (80%), Mamun A (10%), contribution McIntyre HD (5%), Najman J (5%) Data management and statistical analysis: Zalbahar N (90%), Mamun A (10%) Interpretation of results: Zalbahar N (75%), Mamun A (15%), %), McIntyre HD (5%), Najman J (5%) Wrote the manuscript: Zalbahar N (100%) Edited the manuscript: Zalbahar N (80%), Mamun A (10%), McIntyre HD (5%), Najman J (5%)

vi

Contributions by others to the thesis

All contributions have been listed in statement for jointly-authored articles above.

Statement of parts of the thesis submitted to qualify for the award of another degree

None.

vii

Acknowledgements

This PhD journey and the doctoral research was full of challenges—from the research proposal stage to the submission process. No words can describe what I went through, and it was difficult to know what was achievable in the timeframe after a series of tough times. ‘Verily, with every hardship comes ease!’ I boundlessly sought Allah for strength and held to the amazing people surrounding me with infinite support. All praise to Allah, the Almighty. For only with His blessings I had the strength to finally finish the thesis. Most of all, I would like to express my sincere appreciation to my supervisors, Associate Professor Abdullah Al Mamun, Professor Jake Najman and Professor David H. McIntyre for their intellectual assistance throughout my doctoral studies journey.

Thank you Mamun for guiding me with your immense knowledge to make this thesis possible. Also, many thanks for being a gentle critic, helping me to improve my research, analysis and writing skills. Thank you for always re-assuring me with, ‘You are nearly there’ when finishing the thesis seemed impossible. Also, thank you for being very patient with my frantic panic, especially towards the end of the journey. Thank you Jake and David for the hard questions which I could never answer precisely and for challenging my ideas. Your attention to detail drove me to finally learn to become a better researcher. Thank you for shaping me to be critical and to widen my research to include various perspectives. Thank you so much for unfailingly dedicating your time to read through my writing.

I would like to acknowledge the post graduate administration officers Mary Roset and Alison Bath at the School of Public Health, The University of Queensland, for being extremely helpful, always being there to assist in the administrative work and lending your ears as great listeners. I want to thank MUSP data manager Greg Shuttlewood, dedicated in assisting and providing the data from MUSP for the analysis; and to thank the investigators and all participants in the MUSP study. I also acknowledge all the participants and investigators of the Universiti Sains Malaysia Pregnancy Cohort Study for their contribution to this study. The study would not have been accomplished without these great commitments. I thank Dr Hamid Jan and Dr. Loy See Ling for exclusively supporting the research and giving advice.

viii

I would like to thank my sponsors, the Ministry of Education Malaysia for the Skim Latihan Akademik Bumiputera scholarship and the Universiti Putra Malaysia for approving my study leave which provided me the opportunity to pursue a PhD at The University of Queensland, Australia. I am also thankful to the School of Public Health for the top-up scholarship for fee waiver.

I thank my fellow colleagues and friends, especially Duratul Ain and June Joseph for their true friendship, for sharing common stories we faced as students, and for all the fun we have had in the last four years. I am grateful to Dr. Syamhanin and Dr. Norashidah for their generous help in easing my journey when things got tough, and for inspiring me with their achievements and motivation. I also want to thank Tanvir, Munim, and Dr. Sumon for always helping with statistics and giving ideas for writing. I thank Nargess, Dr. Evelyn, Gwendaline, Sweatha, Ana Maria, and Tania for their support, motivation and friendship. I thank all of them for keeping a sense of humour when I had lost mine.

I will always be very grateful to my parents, Haji Zalbahar Adam and Hajah Patimah Othman who deserve a special mention for their love, and also to my brother and sister-in-law for supporting me spiritually and in many other ways to persevere in this difficult journey.

ix

Keywords

Pre-pregnancy, parental, paternal, maternal, overweight, obesity, weight change, BMI change, cohort study

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 111706, Epidemiology, 50 % ANZSRC code: 111103, Nutrition Physiology, 15 % ANZSRC code: 010402, Biostatistics, 35 %

Fields of Research (FoR) Classification

FoR code: 1111, Nutrition and Dietetics, 50% FoR code: 1117, Public Health and Health Services, 50%

x

Table of contents

Abstract ...... ii

Declaration by author ...... iv

Publications during candidature ...... v Peer-reviewed papers ...... v Book Chapters ...... v Conference abstracts ...... v

Publications included in this thesis ...... vi

Contributions by others to the thesis ...... vii Statement of parts of the thesis submitted to qualify for the award of another degree ...... vii

Acknowledgements ...... viii

Keywords ...... x

Australian and New Zealand Standard Research Classifications (ANZSRC) ...... x

Fields of Research (FoR) Classification ...... x

Table of contents ...... xi

List of Figures ...... xiv

List of Abbreviations ...... xvii

Chapter 1: General background ...... 1 1.1 Introduction ...... 1 1.2 Research aim ...... 6 1.3 Research objectives and hypotheses ...... 6 1.4 Organisation of the study ...... 7

Chapter 2: Literature review ...... 16 2.1 Prevalence and burden of overweight and obesity...... 16 2.2 Systematic review of parental – offspring BMI and adiposity association ...... 18 Prenatal parental – offspring BMI and obesity association ...... 22 After birth parental – offspring BMI and obesity association ...... 34 Parental BMI and obesity in relation to the trajectories, tracking and change in offspring BMI and overweight/obesity ...... 42

xi

2.3 Other early life factors related to overweight and obesity ...... 47 Gestational weight gain ...... 47 Birth weight and risk of offspring overweight and obesity ...... 48 Maternal smoking during pregnancy and the risk of offspring obesity ...... 49 Infant breastfeeding protective against overweight and obesity ...... 49 Rapid weight gain and catch-up growth ...... 50 2.4 Pathways and mechanisms of parental-offspring obesity ...... 51 2.5: Summary of the literature review ...... 53

Chapter 3: Methodology ...... 75 3.1 Overall study design and population ...... 75 USM Pregnancy Cohort ...... 75 MUSP ...... 78 3.2 Measurement of exposure ...... 80 USM Pregnancy Cohort ...... 80 MUSP ...... 80 3.3 Measurement of outcomes ...... 80 USM Pregnancy Cohort ...... 80 MUSP ...... 81 3.4 Measurement of covariates and other independent variables ...... 82 USM Pregnancy Cohort ...... 82 MUSP ...... 83 3.6: Statistical analysis ...... 84 Descriptive analysis ...... 85 Multivariable analysis ...... 85 Chapter 4: Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in

Malaysia ...... 89 4.1 Context ...... 89 4.2: Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia...... 90 Chapter 5: Prenatal parental overweight and obesity and the risk of offspring overweight and obesity at 21 years...... 112

xii

5.1: Context ...... 112 5.2: Prenatal parental overweight and obesity and the risk of offspring overweight and obesity at 21 years...... 113 Chapter 6: Parental pre-pregnancy obesity and the risk of offspring weight and body mass index change from childhood to adulthood...... 134 6.1 Context ...... 134 6.2: Parental pre-pregnancy obesity and the risk of offspring weight and body mass index change from childhood to adulthood ...... 135 Chapter 7: Impact of maternal anthropometric change from pre-pregnancy to

21 years postpartum on adult offspring anthropometry and central obesity. .. 156 7.1 Context ...... 156 7.2 Impact of maternal anthropometric change from pre-pregnancy to 21 years postpartum on adult offspring anthropometry and central obesity...... 157

Chapter 8: Discussion and conclusions ...... 176 8.1 Introduction ...... 176 8.2 Main findings ...... 176 8.3 Biological and environmental mechanisms which may explain the associations ...... 180 8.4 Implications ...... 182 8.5 Strengths and limitations ...... 183 Strengths ...... 183 Limitations ...... 184 8.6 Conclusion ...... 185 8.7 Future research and recommendations ...... 185

Appendix A ...... 192

Appendix B ...... 194

xiii

List of Figures

Figure 1.1 Parental early-life determinant pathways of association towards offspring OW/OB across life ...... 4 Figure 2.1 Keyword combinations used for systematic review ...... 20 Figure 2.2 Flow diagram of the selection process for articles ...... 21 Figure 2.3 Summary of maternal and offspring BMI association followed up in months and years (childhood, adolescence and adulthood) ...... 26 Figure 2.4 Summary of maternal and offspring BMI association in adult males and females ...... 26 Figure 2.5 Summary of maternal and offspring overweight/obese (BMI categories) in childhood and adolescence ...... 33 Figure 2.6 Developmental programming of obesity altered by maternal nutrition (source: [203])52 Figure 3.1 Subject flow diagram of USM Pregnancy Cohort Study [2]...... 77 Figure 3.2 Flow diagram of participation in MUSP study phases [3] ...... 79 Figure 4.1 Infant growth status showing mean for weight, weight for age z-score (WAZ), length for age z-score (LAZ), weight for length z-score (WLZ) and body mass index for age z-score (BAZ) at 0, 2, 6 and 12 month according to parental a) paternal and b) maternal BMI categories (<25 kg/m2 and 25 kg/m2) ...... 99 Figure 5.1 Distribution of mean offspring BMI at 21 years by prenatal parental BMI (n = 2229)125 Figure 6.1Distribution of offspring BMI change pattern in categories: (A) 5 and 14 years and (B) 14 and 21 years by prenatal parental BMI category in combination: Both NW parents; NW mother and OW/OB father; OW/OB mother and NW father; both OW/OB parents...... 142 Figure 6.2 Adjusted regression coefficient of overall parental–offspring BMI and BMI-z score association and the separated effect at three developmental stages (5, 14 and 21 years). (A) Paternal–offspring association; (B) Maternal–offspring association. Adjusted for maternal age at birth, baseline family income, maternal smoking habit, maternal gestational weight gained, offspring’s birth weight, offspring’s sex, and offspring’s lifestyle at 14 years (family meals, TV watching and sports participation)...... 147 Figure 7.1(Supplement) Overall, male and female mean, 95% CI of offspring BMI, weight and WC at 21years by maternal absolute BMI change from pre-pregnancy to 21 years post partum in quantiles ...... 169 Figure 7.2 (supplement) Overall, male and female mean, 95% CI of offspring BMI, weight and WC at 21years by maternal relative BMI change from pre-pregnancy to 21 years post partum in quantiles ...... 170 xiv

List of Tables Table 2.1 Inclusion and exclusion criteria for the literature search ...... 20 Table 2.2 Summary of parental (paternal and maternal) BMI (kg/m2 in continuous scale) associations with offspring BMI in childhood and adolescence ...... 24 Table 2.3 Summary of maternal and offspring adiposity association in childhood ...... 28 Table 2.4 Summary of maternal and offspring BMI and obesity association childhood and adulthood ...... 29 Table 2.5 Summary of parental and offspring overweight/obese (BMI categories) in childhood, adolescence and adulthood ...... 31 Table 2.6 Summary of (continuous) parental and offspring BMI associations from childhood, adolescence and adulthood ...... 35 Table 2.7 Summary of parental BMI and offspring adiposity in childhood and adolescence ...... 36 Table 2.8a Summary of parental and offspring overweight/obese association in childhood and adulthood ...... 40 Table 2.9 Summary of parental and offspring overweight/obese association in childhood and adulthood ...... 41 Table 2.10 Summary of studies on parental BMI and obesity on offspring BMI trajectories and weight change ...... 45 Table 3.1 Birth cohorts used in this study ...... 75 Table 3.2 Definitions of growth indicators in terms of z-scores ...... 81 Table 3.3 Formula for the computation of physical activity scores...... 83 Table 3.4 Summary of overall predictors, outcome and other independent variables ...... 84 Table 3.5 Statistical method based on objectives ...... 85 Table 4.1 Distribution (mean and standard deviation) of infant WAZ, LAZ, WLZ and BAZ at 12 months of age by selected background factors ...... 100 Table 4.2 Distribution (mean and standard deviation) of infant skinfold thickness at 12 months of age by parental BMI in two categories...... 103 Table 4.3 Association of parental pre-pregnancy BMI and offspring growth factors throughout 12 months ...... 104 Table 5.1 Comparison of the characteristics of participants who provided information on pre- pregnancy parental BMI, offspring BMI and WC at 21 years versus the participants who were excluded from the analyses ...... 120 Table 5.2: Parental and offspring factors by offspring BMI and WC at 21 years (n = 222) ...... 121

xv

Table 5.3 Unadjusted and adjusted regression coefficient and odds ratios between paternal, maternal BMI and parental weight status with offspring outcomes at 21 years (n = 2229) ...... 123 Table 5.4: Associations of parental BMI and offspring BMI and WC by sex at 21 years, male (n =1114) and female (n = 1115) ...... 124 Table 6.1 Comparison of the characteristics of participants who provided complete information and those who did not provide complete information...... 143 Table 6.2 Prevalence and change in overweight or obesity status between ages 5, 14 and 21 years...... 144 Table 6.3 Unadjusted and adjusted ORs with 95% CI of offspring’s changing patterns of BMI categories from stage to stage by parental BMI categories ...... 145 Table 6.4 Adjusted associations of parental BMI and offspring’s relative BMI change (continuous scale and tertile) from 5 to 14 years, 14 to 21 years and 5 to 21 years ...... 146 Table 7.1 Complete and incomplete information on maternal weight pre-pregnancy and at 21 years postpartum, and on offspring weight and WC at 21 years...... 162 Table 7.2 Maternal (pre-pregnancy, first clinical visit, 21 years postpartum and changes at 21 years postpartum) and offspring (weight, BMI, WC and its categories) measurements ...... 163 Table 7.3 Distribution of the mean (SD) of offspring weight, WC and BMI at 21years by maternal relative weight and BMI change at 21 years postpartum ...... 164 Table 7.4 Distribution n (%) of offspring adulthood overweight/obesity and abdominal obesity by maternal weight and BMI change in quantiles ...... 165 Table 7.5 The OR of OW/OB (BMI) and abdominal obesity of offspring at 21 years by maternal postpartum relative weight and BMI change in quantiles at 21 years ...... 166 Table 7.6 (Supplement): Association between maternal 21-year postpartum weight and BMI change with offspring weight, BMI and waist circumference at 21 years ...... 167 Table 7.7(Supplement) Distribution of the mean (SD) of offspring weight, waist circumference and BMI at 21years by maternal absolute weight and BMI change ...... 168

xvi

List of Abbreviations

BMI Body mass index WC Waist circumference USM Universiti Sains Malaysia MUSP Mater-University of Queensland Study of Pregnancy FFQ Food frequency questionnaire TWG Total gestational weight gain GWG Gestational weight gain Kcal Kilocalories; standard unit of energy intake HR High risk NR No risk OR Odds ratio RR Risk ratio p Significance level GEE General estimation equation PPWC Postpartum weight change PPBMIC Postpartum body mass index change OW Overweight OB Obesity OW/OB Overweight and obese UW/NW Underweight and Normal weight

xvii

Chapter 1: General background

1.1 Introduction

The global prevalence of overweight and obesity (OW/OB), reported by the World Health Organization (WHO) has doubled since 1980 [1]. Some 35% of adults aged 20 years and above were overweight in 2008 and 11% of them were obese. The WHO data also reveals that globally in 2010, 42 million children under the age of five were overweight. The alarming increasing rate of obesity worldwide has been a great concern due to its adverse public health consequences. It is well known that obesity is significantly associated with increasing risk of non-communicable diseases (NCD) such as type 2 mellitus (T2DM) and cardiovascular diseases (CVD). Increasing evidence also suggests that obesity is a major driver of metabolic morbidities, certain types of cancers and mental illnesses [2].

There is a high degree of continuity of OW/OB across the life course. A systematic review of the literature reported that OW/OB in childhood persists into adulthood [3]. Studies have also reported that individuals who were not OW/OB by childhood might become OW/OB during adulthood [4-6]. Gaining weight is a result of an energy imbalance due to excessive intake of food compared to energy expenditure. Multiple potential factors could lead to the occurrence of these increases in weight, from childhood to adulthood. Socio-cultural and familial environment factors are reported to be associated with offspring OW/OB throughout life [7-13]. It is necessary to consider labour saving technology and advances in transportation, such as washing machines, and gadgets, and different forms of transportation. Because of these, lifestyles are more sedentary; there is less movement and walking/cycling to school may not be an option for children, all factors that lead to a reduction in energy expenditure. Easy access to unhealthy food from school tuck shops and fast food restaurants, less food preparation at home, and fewer family meals may also influence the risk of obesity.

Moreover, the role of genetics and epigenetics (the expression of genes without changes in DNA sequences) contributes to the development of obesity [14-16]. A common form of genes that was identified initially was FTO ( mass and obesity associated gene) and recently, research in genome wide association studies has identified other genetic markers such as LEP, LEPR, POMC, PCSK1, MC4R and BDNF [17, 18]. Recent studies have also proposed another potential influence known as the ‘intrauterine over-nutrition hypothesis’ underlying the maternal-offspring obesity association 1

[19]. This hypothesis is based on the concept of excess maternal adiposity and during pregnancy which leads to metabolic alterations related primarily to excess fetal nutrition [20, 21]. These metabolic changes eventually predispose the offspring towards metabolic dysfunction and the risk of OW/OB later in life [22].

The earliest studies suggested that maternal undernutrition and food insecurity was associated with an increased risk of coronary health disease among their children in adulthood (known as Barker’s Theory) [23-25]. These studies later led to further investigation on early life factors as determinants of adverse diseases in later life and develop a foundation on developmental origin of health and diseases [22]. Years later, the ‘developmental origin of health and disease’ (DOHad) have been established and a large number of studies have demonstrated the association of fetal experiences to the risk of later adult chronic diseases such as cancers, CVD, and obesity [26, 27]. It began with associating birth weight as one of the early determinants of the risk of adulthood obesity. Later it evolved to multiple factors such as babies subjected to under and overnutrition in utero, parental genes and lifestyles contributing to the risk of being obese later in life. Thus, parental obesity is an important determinant in early life associated with developmental of obesity in their offspring.

It is generally agreed that prevention of obesity should start in early life. Targeting young children, in school-based intervention programs and other initiatives that involved parents in family-based interventions have been tested. It is also generally agreed that parents are role models and part of a broader prevention strategy [28, 29]. However, our understanding of child nutrition and the circumstances in which parental BMI contributes to the development of offspring OW/OB remains very limited [30, 31].

Parental OW/OB are one of the most important parts of genetic and familial-environmental factors that are strongly linked with the development of offspring overweight and obesity. Most studies, including cross-sectional [32-41], longitudinal [42-46], birth-cohort [10, 47-60], and case-control designs [61, 62], have consistently found that parental OW/OB are strongly associated with offspring OW/OB. When both parents are OW/OB, the risk of OW/OB offspring is greater than when one parent is OW/OB [10, 48, 54, 63]. While some studies have reported parental prenatal BMI and obesity are linked with offspring OW/OB in childhood [50, 64, 65], others have reported parental postnatal BMI and obesity are associated with offspring OW/OB [42, 66]. There is no

2

previous study which has examined whether prenatal and postnatal exposure to parental OW/OB has differential effects on the development of offspring OW/OB. If prenatal exposure has a stronger effect compared to the postnatal exposure to parental OW/OB this might support the over-nutrition hypothesis. If the reverse is the case (i.e. if post-natal exposure has a stronger effect than prenatal exposure), it suggests the importance of the postnatal environment. If there is no difference for offspring OW/OB between pre- and post-natal exposure to parental OW/OB it might suggest there is no differential effect of parental exposure of OW/OB on offspring OW/OB. No study has examined these possibilities.

Irrespective of the time of exposure to parental OW/OB, studies have consistently found that parental OW/OB are associated with offspring OW/OB. The association of parental and offspring OW/OB is characterised by different strengths of association between childhood, adolescence and adulthood [46, 52, 67, 68]. It is unknown whether the relationship between parental and offspring OW/OB differs across the life course of the offspring.

There are multiple pathways which link parental OW/OB to offspring OW/OB. A simple conceptual framework (Figure 1.1) has been developed based on literature for the parental and offspring obesity association [69, 70]. The pathways of this association are mainly determined by genetic factors, intrauterine programming and shared environmental factors. These factors could include the contribution and influence of parental factors on child OW/OB in the short and long term. The contributing factors of parental prenatal and postnatal OW/OB may have two different pathways; both parents who are OW/OB may contribute to the outcome or, father or mother may contribute differentially towards offspring OW/OB.

3

Genetic

Intrauterine programming

Environmental factors

Figure 1.1 Parental early-life determinant pathways of association towards offspring OW/OB across life

As shown in Figure 1.1, there are multiple layers of interaction between biological and environmental factors involving parents that potentially influence the development of offspring OW/OB. The first inner layer, the parental–offspring OW/OB link may reflect a simple genetic predisposition. Adiposity from parents may be transferred by ‘fat genes’ to the child and associated with an increase in the child’s BMI later in life [71]. Others have reported that an insulin-like growth factor is imprinted and expressed from the paternal copy of the gene, and regulates foetal growth and weight gain during conception or later after birth [72]. Recently, altered genetic expression (without change to underlying DNA sequences) due to interactions with the environment (referred to as ‘epigenetic’ effects) have also been identified as potential contributors to offspring OW/OB [73].

Intrauterine development refers to the influence of maternal factors which occur during pregnancy. Previous findings have suggested that maternal factors, such as weight status (pre-pregnancy BMI and gestational weight gain during pregnancy) and lifestyle (smoking habits, alcohol consumption, dietary intake and physical activity), have an impact on offspring obesity in infancy through to adulthood [74]. Offspring exposed to under-nutrition and over-nutrition in utero may later develop obesity across the life course [20, 75]. Mothers who were undernourished may have infants who are more likely to gain weight in the period following the birth and have higher chances of increased BMI and waist circumference (WC) in later life [25, 76]. Conversely, the over-nutrition hypothesis 4

may explain why mothers who are OW/OB before and during pregnancy may directly influence their offspring to increase weight [77]. The association could be due to high fat stores accumulated by the child in the womb and resulting in larger infant size [78]. Moreover, the consumption of high fat and energy-dense foods during pregnancy may deliver excess glucose, free fatty acids and amino acids which may influence the development of the offspring [79].

Finally, parental–offspring OW/OB associations may be explained through a range of shared environmental factors. This may influence offspring size and adiposity due to offspring sharing a similar environment and lifestyle to their parents. Parents’ unhealthy eating habits, such as high- energy and high-fat food consumption, as well as a sedentary way of life, may influence offspring behaviour in early childhood which may persist into adulthood [80]. Such a pattern of excess consumption of energy-dense foods may influence childhood and later obesity—independently of other genetic and intrauterine factors. Unhealthy dietary intake later in life has been observed in animal models [81]. The exposure to the unhealthy lifestyle and behaviour of parents who are OW/OB may create unhealthy habits which persist throughout the life of the offspring. This may further lead the child to gain weight later in life.

The aim of this thesis is to evaluate the association between parental obesity and the development of offspring obesity in childhood, adolescence and adulthood. In this thesis, two birth cohorts were used partly to assess the research aim. First, we used data from the Universiti Sains Malaysia (USM) Pregnancy Cohort Study. This birth cohort dataset consists of parental information at early pregnancy periods, second trimester, and third trimester and after birth. It includes socio- demographic background, anthropometric data, body composition and adiposity measurements. Maternal lifestyle factors—dietary pattern and physical activity during pregnancy—were also included in the dataset associated with this cohort. However, the USM only studied short-term effects with detailed information collected from the prenatal period to one year after birth. The second dataset, the Mater University Study of Pregnancy (MUSP), is a long-term dataset which includes data on critical developmental points (i.e. childhood, adolescence and adulthood). MUSP is an ongoing prospective longitudinal study, initially established in the early 1980s and it involves multiple stages of data collection since 1981–1983. The diversity and uniqueness of data obtained from the MUSP cohort means it is possible to examine the long-term links between parental and offspring BMI and adiposity.

5

1.2 Research aim

The overall aim of this study is to investigate the association of parental pre-pregnancy, pregnancy and after pregnancy BMI and obesity on the development of offspring BMI, waist circumference (WC) and obesity throughout the life course, including childhood, adolescence and adulthood.

1.3 Research objectives and hypotheses

This study has the following specific objectives and hypotheses:

1. To critically review the literatures on the associations of parental pre-pregnancy, pregnancy and after pregnancy BMI and obesity on the development of offspring BMI, waist circumference (WC) and obesity throughout the life stages. Hypothesis: There is an association of parental OW/OB with the risk of offspring obesity across life stages. The association between parental before pregnancy OW/OB is stronger compared to parental OW/OB after birth on the risk of offspring obesity. 2. To investigate the association between pre-pregnancy parental BMI and its categories (normal weight, overweight and obesity) and body composition and offspring growth development and body composition in first year of life. Hypothesis: There is an association between pre-pregnancy parental BMI and its categories and body composition and offspring growth development and body composition in the first year of life. 3. To examine the prospective association between pre-pregnancy parental BMI and its categories and offspring BMI and WC at 21 years. Hypothesis: There is an association between pre-pregnancy parental BMI and its categories and offspring BMI and WC at 21 years. 4. To examine whether pre-pregnancy parental overweight and obesity predict change of offspring BMI from 5 years to 14 years and 14 years to 21 years. Hypothesis: Pre-pregnancy parental overweight and obesity predict change in offspring BMI from childhood to adolescence and adolescence to adulthood. 5. To determine whether maternal 21-years postpartum weight and BMI change predict the risk of offspring overweight and obesity in adulthood. Hypothesis: Maternal 21-years postpartum weight and BMI change predict offspring overweight and obesity in adulthood.

6

1.4 Organisation of the study

This thesis consists of eight chapters, beginning with the introduction and concluding in Chapter 8. Following the introduction (Chapter 1), the literature review in Chapter 2 provides an overview of the thesis, commencing with an overview of the prevalence and burden of OW/OB. The next section is the comprehensive systematic search of the prior literature in relation to parental BMI and obesity and their impact on the development of offspring BMI and obesity. This chapter also includes a brief review of parental BMI and obesity in relation to trajectories, tracking and changes in offspring BMI and obesity, and other early life factors associated with the risk of offspring OW/OB. The pathways explaining parental–offspring obesity associations are discussed as well. In Chapter 3 (methodology), a brief methodology is presented regarding participants’, measurements, and analysis used in order to determine the association between parental BMI and obesity and offspring BMI and obesity. Results are then presented in Chapters 4 to 7. These include one published, one accepted, and two submitted manuscripts. Chapter 4 consists of the manuscript on the short-term impact of prenatal parental BMI and adiposity on offspring body composition and adiposity. Chapter 5 concerns the association between parental BMI and obesity before pregnancy, with offspring BMI and WC in adulthood. Then, Chapter 6 describes the influence of prenatal parental BMI and obesity on offspring BMI change from childhood to adulthood, and the differential effect size of prenatal parental BMI on offspring BMI in childhood, adolescence and adulthood. Chapter 7 consists of documenting and analysing postpartum maternal BMI change over 20 years following the index pregnancy, with the risk of adulthood offspring obesity and abdominal obesity. Finally, all of the results are summarised in Chapter 8, followed by a discussion of implications and limitations of the thesis, recommendations for future work and conclusions.

References [1] WHO. (2013). Global database on body mass index. Available: http://www.who.int/bmi/index.jsp?introPage=intro_3.html. [2] J. J. Reilly, E. Methven, Z. C. McDowell, B. Hacking, D. Alexander, L. Stewart, et al., "Health consequences of obesity," Archives of Disease in Childhood, vol. 88, pp. 748-752, September 1, 2003 2003. [3] A. S. Singh, C. Mulder, J. W. R. Twisk, W. Van Mechelen, and M. J. M. Chinapaw, "Tracking of childhood overweight into adulthood: a systematic review of the literature," Obesity Reviews, vol. 9, pp. 474-488, 2008.

7

[4] S. S. Guo, W. Wu, W. C. Chumlea, and A. F. Roche, "Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence," The American Journal of Clinical Nutrition, vol. 76, pp. 653-658, September 1, 2002 2002. [5] C. Y. Li, M. I. Goran, H. Kaur, N. Nollen, and J. S. Ahluwalia, "Developmental trajectories of overweight during childhood: Role of early life factors," Obesity, vol. 15, pp. 760-771, Mar 2007. [6] V. Burke, Beilin, L.J., Dunbar, D., "Family lifestyle and parental body mass index as predictors of body mass index in Australian children: A longitudinal study," International Journal of Obesity, vol. 25, 2001. [7] M. Kanciruk, J. J. Andrews, and T. Donnon, "Family history of obesity and risk of childhood overweight and obesity: A meta-analysis," International Journal of Medicine Health Pharmacology Biomedical Engineering, vol. 8, pp. 244-56, 2014. [8] J. Laitinen, C. Power, and M. Järvelin, "Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity," American Journal of Clinical Nutrition, vol. 74, pp. 287-294, 2001. [9] M. R. Lowe, L. M. Shank, R. Mikorski, and M. L. Butryn, "Personal History of Dieting and Family History of Obesity are Unrelated: Implications for Understanding Weight Gain Proneness," Eating Behaviors, vol. 17, 2015. [10] A. A. Mamun, D. A. Lawlor, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Family and early life factors associated with changes in overweight status between ages 5 and 14 years: findings from the Mater University Study of Pregnancy and its outcomes," International Journal of Obesity and Related Metabolic Disorders, vol. 29, pp. 475-482, 2005. [11] A. A. Mamun, M. J. O'Callaghan, G. Williams, and J. M. Najman, "Television watching from adolescence to adulthood and its association with BMI, waist circumference, waist-to- hip ratio and obesity: a longitudinal study," Public Health Nutrition, vol. 16, pp. 54-64, 2013. [12] P. M. Anderson and K. F. Butcher, "Childhood obesity: trends and potential causes," The Future of children, vol. 16, pp. 19-45, 2006. [13] K. Harrison, K. K. Bost, B. A. McBride, S. M. Donovan, D. S. Grigsby-Toussaint, J. Kim, et al., "Toward a Developmental Conceptualization of Contributors to Overweight and Obesity in Childhood: The Six-Cs Model," Child Development Perspectives, vol. 5, pp. 50- 58, 2011.

8

[14] R. J. F. Loos, D. I. W. Phillips, R. Fagard, G. Beunen, C. Derom, C. Mathieu, et al., "The influence of maternal BMI and age in twin on in the offspring," Diabetes Care, vol. 25, pp. 2191-2196, Dec 2002. [15] K. E. Rhee, S. Phelan, and J. McCaffery, "Early determinants of obesity: genetic, epigenetic, and in utero influences," International Journal of Pediatric, vol. 2012, p. 463850, 2012. [16] J. Wardle, S. Carnell, C. Haworth, and R. Plomin, "Evidence for a strong genetic influence on childhood adiposity despite the force of the obesogenic environment," American Journal of Clinical Nutrition, vol. 87, pp. 398 - 404, 2008. [17] T. M. Frayling, N. J. Timpson, M. N. Weedon, E. Zeggini, R. M. Freathy, C. M. Lindgren, et al., "A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity," Science, vol. 316, pp. 889-94, May 11 2007. [18] D. Albuquerque, E. Stice, R. Rodriguez-Lopez, L. Manco, and C. Nobrega, "Current review of genetics of human obesity: from molecular mechanisms to an evolutionary perspective," Molecular Genetics and Genomics, vol. 290, pp. 1191-221, Aug 2015. [19] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater- university study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb 15 2007. [20] E. Oken and M. W. Gillman, "Fetal origins of obesity," Obesity Research, vol. 11, pp. 496- 506, 2003. [21] R. C. Whitaker and W. H. Dietz, "Role of the prenatal environment in the development of obesity," Journal of Pediatrics, vol. 132, pp. 768-776, 5// 1998. [22] P. D. Taylor and L. Poston, "Developmental programming of obesity in mammals," Experimental Physiology, vol. 92, pp. 287-298, March 1, 2007 2007. [23] D. J. P. Barker and C. Osmond, "Infant mortality, childhood nutrition, and ischaemic heart disease in england and wales," The Lancet, vol. 327, pp. 1077-1081, 1986/05/10 1986. [24] D. J. P. Barker, "Maternal nutrition, fetal nutrition, and disease in later life," Nutrition, vol. 13, pp. 807-813, 1997/09/01 1997. [25] A. C. J. Ravelli, Jhp, C. Osmond, D. J. P. Barker, and O. P. Bleker, "Obesity at the age of 50 y in men and women exposed to famine prenatally," American Journal of Clinical Nutrition, vol. 70, pp. 811-816, 1999.

9

[26] M. A. Hanson and P. D. Gluckman, "Developmental origins of health and disease – Global public health implications," Best Practice & Research Clinical Obstetrics & Gynaecology, vol. 29, pp. 24-31, 1// 2015. [27] R. Robinson, "The fetal origins of adult disease," BMJ, vol. 322, p. 375, 2001. [28] P. Sung-Chan, Y. W. Sung, X. Zhao, and R. C. Brownson, "Family-based models for childhood-obesity intervention: a systematic review of randomized controlled trials," Obesity Reviews, vol. 14, pp. 265-278, 2013. [29] H. M. Yavuz, M. H. van Ijzendoorn, J. Mesman, and S. van der Veek, "Interventions aimed at reducing obesity in early childhood: a meta-analysis of programs that involve parents," Journal of Child Psychology and Psychiatry, pp. n/a-n/a, 2014. [30] J. Dodd, R. Grivell, C. Crowther, and J. Robinson, "Antenatal interventions for overweight or obese pregnant women: a systematic review of randomised trials," British Journal of Obstatric and Gyneacology, vol. 117, pp. 1316 - 1326, 2010. [31] J. M. Dodd, C. A. Crowther, and J. S. Robinson, "Dietary and lifestyle interventions to limit weight gain during pregnancy for obese or overweight women: A systematic review," Acta Obstetricia et Gynecologica Scandinavica, vol. 87, pp. 702-706, 2008. [32] M. Birbilis, Moschonis, G., Mougios, V., Manios, Y., "Obesity in adolescence is associated with perinatal risk factors, parental BMI and sociodemographic characteristics," European Journal of Clinical Nutrition, vol. 67, 2013. [33] L. Gomez-Lopez, Van Hulst, A., Barnett, T.A., Roy-Cagnon, M.-H., Tremblay, A., O'Loughlin, J., Lambert, M., "Does parental body mass index status modify the associations among birth weight, early growth and childhood adiposity?," Paediatrics and Child Health (Canada), vol. 18, 2013. [34] M. H. Jiang, Y. Yang, X. F. Guo, and Y. X. Sun, "Association between child and adolescent obesity and parental weight status: A cross-sectional study from rural north China," Journal of International Medical Research, vol. 41, pp. 1326-1332, Aug 2013. [35] H. S. Lee, K. J. Duffey, C. I. Kim, and B. M. Popkin, "The relationship between family and child weight status by household structure in South Korea: 2007-2010," Nutrition & Diabetes, vol. 3, Jun 2013. [36] K. McCarthy, Y. L. Ye, S. Yuan, and Q. Q. He, "Parental Weight Status and Offspring Cardiovascular Disease Risks: a Cross-Sectional Study of Chinese Children," Preventing Chronic Disease, vol. 12, Jan 2015.

10

[37] A. M. Oliveira, A. C. Oliveira, M. S. Almeida, N. Oliveira, and L. Adan, "Influence of the family nucleus on obesity in children from northeastern Brazil: a cross-sectional study," Bmc Public Health, vol. 7, Sep 2007. [38] S. Plachta-Danielzik, B. Landsberg, M. Johannsen, D. Lange, and M. J. Muller, "Determinants of the prevalence and incidence of overweight in children and adolescents," Public Health Nutrition, vol. 13, pp. 1870-1881, Nov 2010. [39] S. Santiago, I. Zazpe, M. Cuervo, and J. A. Martinez, "Perinatal and parental determinants of childhood overweight in 6-12 years old children," Nutricion Hospitalaria, vol. 27, pp. 599-605, Mar-Apr 2012. [40] R. C. Whitaker, C. M. Deeks, A. E. Baughcum, and B. L. Specker, "The relationship of childhood adiposity to parent body mass index and eating behavior," Obesity Research, vol. 8, pp. 234-240, 2000. [41] L. Xu, L. Dubois, D. Burnier, M. Girard, and D. Prud'homme, "Parental overweight/obesity, social factors, and child overweight/obesity at 7 years of age," Pediatrics International, vol. 53, pp. 826-31, Dec 2011. [42] L. Dubois and M. Girard, "Early determinants of overweight at 4.5 years in a population- based longitudinal study," International Journal of Obesity, vol. 30, pp. 610-617, Apr 2006. [43] Y. Li, K. Brock, R. Cant, L. Ke, and S. Morrell, "Parental obesity as a predictor of childhood overweight/obesity in Australian migrant children," Obesity Research & Clinical Practice, vol. 2, pp. 179-187, Sep 2008. [44] A. M. Magarey, L. A. Daniels, T. J. Boulton, and R. A. Cockington, "Predicting obesity in early adulthood from childhood and parental obesity," International journal of obesity, vol. 27, pp. 505-513, 2003. [45] W. Stewart Agras, Hammer, L.D., McNicholas, F., Kraemer, H.C., "Risk factors for childhood overweight: A prospective study from birth to 9.5 years," Journal of Pediatrics, vol. 145, 2004. [46] R. C. Whitaker, Wright, J.A., Pepe, M.S., Seidel, K.D., Dietz, W.H., "Predicting obesity in young adulthood from childhood and parental obesity," New England Journal of Medicine, vol. 337, pp. 869 - 873, 1997. [47] P. Chivers, H. Parker, M. Bulsara, L. Beilin, and B. Hands, "Parental and early childhood influences on adolescent obesity: a longitudinal study," Early Child Development and Care, vol. 182, pp. 1071-1087, 2013/04/21 2012.

11

[48] B. Durmus, L. R. Arends, L. Ay, A. C. Hokken-Koelega, H. Raat, A. Hofman, et al., "Parental anthropometrics, early growth and the risk of overweight in pre-school children: the Generation R Study," Pediatric Obesity, vol. 8, pp. 339-50, Oct 2013. [49] L. Fairley, G. Santorelli, D. A. Lawlor, M. Bryant, R. Bhopal, E. S. Petherick, et al., "The relationship between early life modifiable risk factors for childhood obesity, ethnicity and body mass index at age 3 years: findings from the Born in Bradford birth cohort study," BMC Obesity, vol. 2, p. 9, 2015. [50] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child Cohort Study: A Family-based Approach to Studying the Role of the Intrauterine Environment in Childhood Adiposity," American Journal of Epidemiology, vol. 176, pp. 83- 92 July 15, 2012 2012. [51] D. H. Heppe, J. C. Kiefte-de Jong, B. Durmus, H. A. Moll, H. Raat, A. Hofman, et al., "Parental, fetal, and infant risk factors for preschool overweight: the Generation R Study," Pediatric Research, vol. 73, pp. 120-7, Jan 2013. [52] J. K. Lake, C. Power, and T. J. Cole, "Child to adult body mass index in the 1958 British birth cohort: associations with parental obesity," Archives of Disease in Childhood, vol. 77, pp. 376-381, Nov 1997. [53] M. Mostazir, Jeffery, A., Voss, L., Wilkin, T.J., "Gender-assortative waist circumference in mother-daughter and father-son pairs, and its implications. An 11-year longitudinal study in children (EarlyBird 59)," Pediatric Obesity, vol. 9, 2014. [54] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357-1359, Jun 2005. [55] N. Stettler, A. M. Tershakovec, B. S. Zemel, M. B. Leonard, R. C. Boston, S. H. Katz, et al., "Early risk factors for increased adiposity: a cohort study of African American subjects followed from birth to young adulthood," American Journal of Clinical Nutrition, vol. 72, pp. 378-383, Aug 2000. [56] V. Svensson, J. A. Jacobsson, R. Fredriksson, P. Danielsson, T. Sobko, H. B. Schiöth, et al., "Associations between severity of obesity in childhood and adolescence, obesity onset and parental BMI: A longitudinal cohort study," International Journal of Obesity, vol. 35, pp. 46-52, 2011.

12

[57] S. R. Veena, G. V. Krishnaveni, S. C. Karat, C. Osmond, and C. H. Fall, "Testing the fetal overnutrition hypothesis; the relationship of maternal and paternal adiposity to adiposity, insulin resistance and cardiovascular risk factors in Indian children," Public Health Nutrition, vol. 16, pp. 1656-66, Sep 2013. [58] V. Vogels, D. L. A. Prosthumus, E. C. M. Mariman, F. Bouwman, A. D. M. Kester, P. Rump, et al., "Determinants of overweight in a cohort of Dutch children," American Journal of Clinical Nutrition, vol. 84, pp. 717-724, 2006. [59] S. M. T. Williams, R. W; Taylor, B. J., "Secular changes in BMI and the associations between risk factors and BMI in children born 29 years apart," Pediatric Obesity, vol. 8, pp. 21-30, Feb 2013. [60] N. Zalbahar, H. J. Mohamed Jan, S. L. Loy, J. Najman, H. D. McIntyre, and A. Mamun, "Association of parental body mass index before pregnancy on infant growth and body composition: Evidence from a pregnancy cohort study in Malaysia," Obesity Research and Clinical Practice, 2015. [61] K. Bammann, J. Peplies, S. De Henauw, M. Hunsberger, D. Molnar, L. A. Moreno, et al., "Early Life Course Risk Factors for Childhood Obesity: The IDEFICS Case-Control Study," PLoS ONE, vol. 9, p. e86914, 2014. [62] K. M. Rathnayake, A. Satchithananthan, S. Mahamithawa, and R. Jayawardena, "Early life predictors of preschool overweight and obesity: a case-control study in Sri Lanka," BMC Public Health, vol. 13, Oct 2013. [63] E. Freeman, R. Fletcher, C. E. Collins, P. J. Morgan, T. Burrows, and R. Callister, "Preventing and treating childhood obesity: time to target fathers," International Journal of Obesity, vol. 36, pp. 12-15, Jan 2012. [64] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: a 16-year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct 2011. [65] C. M. Murrin, G. E. Kelly, R. E. Tremblay, and C. C. Kelleher, "Body mass index and height over three generations: evidence from the Lifeways cross-generational cohort study," BMC Public Health, vol. 12, p. 81, 2012. [66] C. Power, Pouliou, T., Li, L., Cooper, R., HyppöÃnen, E., "Parental and offspring adiposity associations: Insights from the 1958 British birth cohort," Annals of Human Biology, vol. 38, 2011.

13

[67] M. Kivimäki, D. A. Lawlor, G. D. Smith, M. Elovainio, M. Jokela, L. Keltikangas-Järvinen, et al., "Substantial intergenerational increases in body mass index are not explained by the fetal overnutrition hypothesis: The Cardiovascular Risk in Young Finns Study," American Journal of Clinical Nutrition, vol. 86, pp. 1509-1514, 2007. [68] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166-1175, 2011. [69] R. Cooper, E. Hyppönen, D. Berry, and C. Power, "Associations between parental and offspring adiposity up to midlife: the contribution of adult lifestyle factors in the 1958 British Birth Cohort Study," American Journal of Clinical Nutrition, vol. 92, pp. 946-953, 2010. [70] Q. Wu and M. Suzuki, "Parental obesity and overweight affect the body-fat accumulation in the offspring: the possible effect of a high-fat diet through epigenetic inheritance," Obesity Reviews, vol. 7, pp. 201-208, 2006. [71] S. N. Yazbek, S. H. Spiezio, J. H. Nadeau, and D. A. Buchner, "Ancestral paternal genotype controls body weight and food intake for multiple generations," Human Molecular Genetics, vol. 19, pp. 4134-4144, November 1, 2010 2010. [72] L. M. E. McCowan, R. A. North, E. M. Kho, M. A. Black, E. H. Y. Chan, G. A. Dekker, et al., "Paternal Contribution to Small for Gestational Age Babies: A Multicenter Prospective Study," Obesity, vol. 19, pp. 1035-1039, May 2011. [73] E. R. Kyung, P. Suzanne, and M. Jeanne, "Early Determinants of Obesity: Genetic, Epigenetic, and In Utero Influences," International Journal of Pediatrics, vol. 2012, p. 9, 2012. [74] P. M. Catalano and H. M. Ehrenberg, "Review article: The short- and long-term implications of maternal obesity on the mother and her offspring," BJOG: An International Journal of Obstetrics & Gynaecology, vol. 113, pp. 1126-1133, 2006. [75] S. D. Parlee and O. A. MacDougald, "Maternal nutrition and risk of obesity in offspring: The Trojan horse of developmental plasticity," Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, vol. 1842, pp. 495-506, 3// 2014. [76] R. C. Painter, T. J. Roseboom, and O. P. Bleker, "Prenatal exposure to the Dutch famine and disease in later life: an overview," Reproductive Toxicology, vol. 20, pp. 345-52, Sep-Oct 2005.

14

[77] J. S. Dyer and C. R. Rosenfeld, "Metabolic Imprinting by Prenatal, Perinatal, and Postnatal Overnutrition: A Review," Seminars in Reproductive Medicine, vol. 29, pp. 266-276, May 2011. [78] K. Shankar, A. Harrell, X. Liu, J. M. Gilchrist, M. J. J. Ronis, and T. M. Badger, "Maternal obesity at conception programs obesity in the offspring," American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, vol. 294, pp. R528-R538, February 2008 2008. [79] M. Alfaradhi and S. Ozanne, "Developmental programming in response to maternal over- nutrition," Frontiers in Genetics, vol. 2, 2011-June-3 2011. [80] A. A. Mamun, M. J. O'Callaghan, S. M. Cramb, J. M. Najman, G. M. Williams, and W. Bor, "Childhood behavioral problems predict young adults' BMI and obesity: evidence from a birth cohort study," Obesity (Silver Spring), vol. 17, pp. 761-6, Apr 2009. [81] S. L. Kirk, A.-M. Samuelsson, M. Argenton, H. Dhonye, T. Kalamatianos, L. Poston, et al., "Maternal Obesity Induced by Diet in Rats Permanently Influences Central Processes Regulating Food Intake in Offspring," PLoS ONE, vol. 4, p. e5870, 2009.

15

Chapter 2: Literature review

In this chapter reviews the overall context of this thesis, which begins with the prevalence and burden of overweight and obesity. Then, because parental–offspring obesity associations have been widely researched, the focus of the literature review narrows further, highlighting prospective and retrospective studies that deal with the associations between prenatal and postnatal parental BMI and adiposity and offspring BMI and adiposity throughout life. In this section a systematic review, fulfilling Objective 1, which was outlined previously is conducted.

The next section consists of a brief discussion regarding the influence of parental BMI and obesity on offspring BMI trajectories across life. Subsequently, other early life factors that are associated with offspring overweight and obesity are briefly reviewed. Finally, the main pathways and mechanisms that relate to parental–offspring BMI and obesity associations is summarised.

2.1 Prevalence and burden of overweight and obesity

Globally, the prevalence of obesity has doubled over the last three decades and it has become one of the major public health concerns. Although the speed of the increase has now slowed down in some developed countries, obesity is still rapidly increasing in most middle-income and developing countries [1, 2]. Low, et al. [3] reported in a review that the increasing trend of obesity prevalence among adults will reach a peak in the 40–60 year age group. In the United States, 35.5% of adult men and 35.8% of adult women have been reported to be obese [4]. Recently the prevalence of obesity among adults in Australia was found to be 28.0% [5] and in the United Kingdom it was 33.9% [6]. A review of the Asia Pacific Cohort Studies [7] reported that 20–50% of adults were overweight and obese in developing countries. The combined prevalence of overweight and obesity has increased by 414% in China, from 3.7% in 1982 to 19.0% in 2002. Overweight and obesity prevalence among the mid-adulthood population (aged 40–56 years) in Malaysia has increased from 20% in 1996 to 43% in 2006, according to Khambalia and Seen [8]. In a later publication in 2012, the age-adjusted prevalence of overweight and obesity among adults over the age of 18 years in Malaysia was reported as 38.3% and 34.0%, respectively [9].

Rapid changes, such as an increasingly sedentary lifestyle and high-calorie diets, have contributed to several developing countries facing the double burden of under-nutrition and over-nutrition. There are other regions where under-nutrition still persists. For example, in Indonesia the

16

prevalence of under-nutrition is still high. However, the prevalence of overweight and obesity in Indonesia has also increased from 1993 to 2007: around 21% to 31% among males and 30% to 49% among females [10]. The prevalence of obesity in women of childbearing age has been closely monitored due to the association with adverse pregnancy outcomes [11, 12, 13 14]. Furthermore, infants born to mothers who gained excess weight during pregnancy suffered impaired health status in both the short and long term [14, 15]. In the United Kingdom, maternal obesity has increased considerably over time, more than doubling from 7.6% to 15.6% over 19 years among mothers attending their first antenatal check-up [16]. In Australia between 1997 and 2002, one in three pregnant women were found to be overweight or obese, regardless of the trimester [11]. Other population-based studies in Asia or developing countries are rare; but, a study in an obstetric population in Hong Kong reported that 31.7% of women at the first trimester visit were overweight and obese [17].

Although the prevalence of children being overweight and obese has stabilised in some developed countries, these rates remain high and may increase [18]. In 2011, approximately 40 million children under five [1] were overweight and obese. A recent study by Ogden, et al. [19] reported that since the 1980s, the prevalence of childhood obesity in the United States has doubled, and has tripled in adolescents. In Australia between 1985 and 2008 there was a slight increase in overweight and obesity prevalence among children aged 2–18 years (from 21% overweight to 25%, and from 5% obese to 6%) [20]. The estimated prevalence of childhood overweight and obesity in Africa in 2010 was 8.5% and is expected to reach 12.7% by 2020. The prevalence of overweight and obesity is lower in Asia than in Africa (4.9% in 2010), but the absolute number of affected children (18 million) is higher in Asia [21]. Moreover, results from the National Health Morbidity Survey III in Malaysia showed that one out-of-five children aged 7–12 years were overweight and obese [22]. In countries which are facing the challenge of the double burden of under- and over-nutrition, for example, Indonesia, it has been shown that among children under five years who were undernourished, 12% of them were overweight and obese [23]. The prevalence of childhood overweight and obesity remains high in many parts of the world, and childhood overweight and obesity is likely to have a significant health impact later in life.

The growing prevalence of overweight and obesity across different age groups has a significant negative health impact. The strong associations of overweight and obesity with co-morbidities have been reviewed systematically by Guh, et al. [24]. Notably, being overweight and obese has been

17

documented as increasing the risk of cardiovascular diseases (CVDs), diabetes and cancers [25]. Being overweight and obese in childhood was also associated with increased incidence of metabolic syndromes, CVD and T2DM in adulthood [26]. Multiple health consequences of being overweight and obese have also amplified the burden on healthcare providers and the economy of a country. Furthermore, obesity was estimated to affect total healthcare expenditures from 0.7% in France to 2.8% in the United States, with higher medical costs associated with obese individuals compared with their normal-weight peers [27]. As such, Wang et al. 2011 reported that the continuing rise of obesity in the United Kingdom and the United States has led to an increase in health and economic costs which was projected to add to the health-care budget by US$48–66 billion a year in the United States and by £1·9–2 billion a year by 2030 in the United Kingdom [28]. In 2005, the total annual direct cost of overweight and obesity in Australia was approximately A$21 billion [29]. The overall cost of obesity to Australian society and government was estimated at A$58.2 billion in 2008, including the cost of disability, loss of well-being and premature death resulting from obesity and its impacts, productivity costs, health system costs and carer costs [30].

Therefore, with the current burden of increasing overweight and obesity within various populations, multiple intervention programs have been actively developed worldwide to address the emerging prevalence of overweight and obesity [31]. However, Hesketh and Campbell [32] reported that few interventions, which were intended to reduce obesity during early childhood, focused on childhood anthropometry and physical activities. Furthermore, there was also a wide variety of interventions for obesity, including community-based, family-based and school-based interventions. However, more research and consistent methods are needed to understand the comparative effectiveness of obesity prevention programs in these differing settings [33]. So far, most of the intervention strategies have only shown a short term and limited positive impact.

2.2 Systematic review of parental – offspring BMI and adiposity association

Parental BMI and obesity have been reported to be one of the strongest early-life factors associated with the development of offspring overweight and obesity [34-37]. Most of the cross-sectional studies have demonstrated this association—primarily in childhood [38-53], while some have reported this association during adolescence [43, 54, 55]. For example, a large population-based cross-sectional study by Whitaker et al. reported that the risk of childhood obesity was predicted by having obese parents with odds of 22.3 (95%CI: 10.3–48.4) compared to children who had normal- weight parents [56]. They also reported that maternal BMI (beta-coefficient: 0.27) had a stronger 18

association than paternal BMI (beta-coefficient: 0.23) with childhood BMI, after adjusting for offspring age, sex, family socio-economic factors and clustering by family. A meta-analysis of observational studies concluded that a higher risk of childhood overweight and/or obesity was associated with any family members who were obese [35]. This study also reported that children were at two times the risk of being overweight/obese when either father or mother were obese as compared to children with no parental history of obesity. However, the analyses included a variety of observational study designs (case control, cross-sectional and prospective) and were primarily cross-sectional studies. As with all cross-sectional designs, there are major limitations in determining whether reported relationships between paternal–maternal BMI and offspring BMI are cause and effect. Apart from the study design differences between cross-sectional and longitudinal studies, there are differences in terms of when the parents’ measurements were obtained. Two distinct timings of parental-weight status measurement employed in previous studies may be divided into ‘prenatal’ and ‘after birth’. Also, varying relationships of paternal and maternal weight status with offspring weight status have been observed, depending on whether the associations are presented as effect size (beta-coefficient, odd ratios or risk ratios). Moreover, different periods of parental BMI and obesity exposure may explain differences in mechanisms which contribute to the association with offspring BMI and obesity.

BMI is a common indicator used to categorise a person’s weight status (i.e. underweight, normal weight, overweight or obese). One major reference standard has been used widely to define overweight and obesity in adults using the WHO’s BMI cut-off [57] (Although some Asia-Pacific and Asian countries have introduced their own reference standards, modified to suit their populations). As for children and adolescents, several reference standards have been used to determine overweight and obesity. Major globally-known growth standards commonly used for children and adolescence are the Centre of Disease Control (CDC), International Obesity Task Force (IOTF) and WHO. For some of these, criteria for BMI are age and sex standardised [58-61].

A large number of prospective studies have examined associations between parental BMI and/or obesity and offspring BMI and/or obesity. Consequently, a systematic review was undertaken and the search of the literature was narrowed to prospective/retrospective study designs, dealing with parental BMI and other proxy measurements of adiposity as the exposure, in relation to the development of offspring BMI and other measurements of adiposity. This systematic review is based on PRISMA (Preferred reporting items for systematic review and meta-analysis) guidelines.

19

A combination of keywords related to the target population and exposures were used, as shown in Figure 2.1. Comprehensive searches from online databases were conducted and reviewed (Figure 2.2).

A combination of four groups of keywords [(title/abstract and MeSH) terms] related to our target population and exposure:  obesity OR obese OR "body mass" OR "body fat" OR "body " OR "body fatness" OR overweight OR "body composition" OR "waist circumference"

 parent* OR mother* OR father* OR maternal OR paternal  "offspring" OR offspring OR son* OR daughter* OR child* OR adolescence OR adult*

 Longitudinal OR prospective OR retrospective OR cohort* OR panel

Figure 2.1 Keyword combinations used for systematic review

Titles and abstracts of studies found in the database search were screened, based on the inclusion criteria shown in Table 2.1, and assessed for quality. The search was restricted to the last 30 years. Studies were required to have at least one follow-up, with measurement of offspring BMI and overweight/obesity or any measure of adiposity. Other adiposity measurement in this search may capture studies measuring parental (maternal or/and paternal) BMI using waist circumference, skinfold thickness, BIA (Bioelectrical impedance analysis), DEXA (Dual-energy X-ray absorptiometry, total body water or others). Singleton-birth offspring aged no less than 6 months or more than 60 years at follow-up were included. Short-listed articles in full-text were screened and reviewed for inclusion in the systematic review.

Table 2.1 Inclusion and exclusion criteria for the literature search

Inclusion Exclusion Study design Prospective/retrospective Cross-sectional, case control, and intervention Participants Human studies Animal studies Follow-up At least one time - Exposure Parental, maternal and/or paternal Family, stepmother and stepfather Outcomes Singleton-birth offspring aged Twin, siblings between 6 months and 60 years old (infant, preschool, childhood, adolescence and adulthood) Measurements BMI (continuous or categorical), other Body image/ type scale used adiposity measurements (using waist circumference and skinfold thickness, DEXA, BIA, and total body water) 20

Records identified through Additional records identified through database searching other sources PubMed, CINAHL, Web of (n = 4) Science, Scopus and Embase (n = 15772)

Records after screened for duplicates (n = 8418) Full-text articles excluded  Cross-sectional and case control study design (n = 37)  GWG as main exposure (n = 5)  Overall Metabolic Syndrome risk factor as outcome (n = 1)  Non-English (n = 6)

 Offspring secular trend as Records after abstracts screened for outcome (n = 1) unrelated topics  Intergenerational (n = 13)

 Offspring outcome as risk (n = 223) categories (n = 1)  Offspring aged less than 6 months and more than 60 years (n = 1)  Exposure missing value (n = 2)  Combination of parental factors with other factors (n = 2) Full-text articles assessed for eligibility  Severe obesity (n = 1)  Twin outcome (n = 2) (n = 154)  Similar cohorts outcome and measure (n = 2)

Secondary outcome Studies included in systematic review  Offspring weight change as Primary outcome outcome (n = 3) (n = 63)  Offspring growth and BMI in trend and trajectories as outcome (n = 15)

Figure 2.2 Flow diagram of the selection process for articles

21

The primary outcome was divided into two main sections. The first section included studies using parental weight status measured before and/or during pregnancy. The second section included studies which reported parental weight status measured after the birth of the child (at least one month after the birth of the index child). This time period may include major influences from shared familial environmental factors such as behavioural lifestyle and habits.

The secondary outcome of the systematic review included studies which considered the associations of parental BMI and obesity on the trajectories, tracking, and changes of offspring BMI and obesity throughout childhood.

Prenatal parental – offspring BMI and obesity association

Several studies from the literature search reported prenatal parental BMI and overweight/obesity in relation to the development of offspring BMI and overweight/obesity. These findings were classed as ‘prenatal’ because measurements of the father, mother or both parents were taken before or during pregnancy. This classification was used regardless of whether the measurements were self- reported, or taken by the researcher. Initially, a total of 42 studies were found [37, 62-99]. Two studies were excluded due to multiple reports about the same cohorts [67, 100, 101]. Other studies were identified as potentially using similar cohorts, but reported different outcome measures; for example, age at time of outcomes, effect size determination, and gender assortative findings. Fifteen of these studies reported associations between parental and offspring BMI as regression coefficients after adjustment for potential confounding [66, 69, 72, 74, 80, 82, 83, 85, 88, 94, 96, 97, 102, 103]. Despite this, each study used different selections and numbers of confounding factors for adjustment in their models. Of these, as shown in Table 2.2, four had examined paternal and maternal BMI associations with offspring BMI in childhood and adolescence [66, 72, 82, 83]. Another study reported the association between parental adiposity, using sum of skinfold thickness and offspring BMI [97]. These studies reported that there was no difference between maternal and paternal BMI on offspring BMI in children under 10 years [72, 83, 97]. Although, Lawlor et al. reported that among teenage offspring (14 years) in the MUSP (Mater – University of Queensland Study of Pregnancy) cohort, the association between maternal-offspring BMI (β = 0.353; 95% CI: 0.304–0.401) were slightly higher compared to paternal-offspring BMI association (β = 0.251; 95% CI: 0.199–0.304) [82]. Similarly, Chivers et al. reported that a unit increase of maternal BMI was associated with a 0.037 kg/m2 (p <0.001) increase in 14 year old offspring BMI, and a unit increase in adjusted paternal BMI was associated with a 0.023 kg/m2 (p = 0.005) increase [66]. From these 22

five studies, an increase in paternal and maternal BMI before pregnancy was found to be associated with an increase in offspring BMI in childhood and adolescence. Also, there appeared to be a slightly higher effect size of maternal BMI compared to paternal BMI as the age of their offspring increased.

23

Table 2.2 Summary of parental (paternal and maternal) BMI (kg/m2 in continuous scale) associations with offspring BMI in childhood and adolescence Author and Cohort profile/ Population, Exposure Outcome Outcome Results year population country measured age (yrs) Fleten et al., Norwegian Mother 29,216 parents During BMI 3 In final model, on childhood BMI z score, a 1 SD increase was 2012 [72] and Child Cohort and child trios, pregnancy 0.149 (0.130–0.168) for maternal BMI and 0.122 (0.103–0.140) Study (MoBa) Norway for paternal BMI. On offspring BMI, an increase of 1 kg/m2 was 0.036 (0.031–0.040) for parental BMI and 0.037 (0.032–0.043) for paternal BMI Leary et al., Avon Longitudinal 4654 parents During BMI 7.5 A unit increase in maternal BMI and paternal BMI was associated 2010 [83] Study of parents and child trios, pregnancy with 0.218 (0.183–0.254) and 0.197 (0.161–0.233) BMI increase and children UK in male offspring (age adjusted using standardised regression) (ALSPAC) while in female offspring BMI increased 0.288 (0.249–0.328) and 0.197 (0.158–0.235) Veena et al., The Mysore 504 parents and During BMI 9.5 In adjusted model, 1 SD increase in parental SS was associated 2012 [97] Parthenon study child trios, India pregnancy with 0.09 (0.06–0.12) maternal SS and 0.06 (0.03–0.08) paternal SS. BMI: 0.95 (0.57–1.33) for maternal BMI and 0.61 (0.27– 0.95) for paternal BMI Chivers et The Western 1403 parents During BMI 14 In final adjusted model, increase in 1kg/m2 maternal BMI and al., 2012 [66] Australian and child trios, pregnancy paternal BMI increased offspring BMI 0.037kg/m2 (p <0.001) Pregnancy Cohort Australia and 0.023 kg/m2 (p <0.005) (Raine Study) Lawlor et al., The Mater 3340 parents During BMI 14 In final adjusted model for standardised offspring BMI, a 1 SD 2007 [82] University of and child trios, pregnancy increase in maternal BMI was 0.362 SD (0.323–0.402) compared Queensland Study Australia to paternal BMI 0.239 SD (0.197–0.282); (p <0.0001) of Pregnancy (MUSP) Abbreviations: SD = standard deviation; SS = adiposity Note: Confidence intervals in brackets are at 95%

24

Nine studies reported predominantly maternal–offspring BMI associations from as early as 6 months old to 40 years old [69, 74, 80, 81, 85, 88, 94, 96, 102]. Across these nine studies, there were increases in the effect size of associations between mother and offspring with the increase of age (childhood to adulthood) as shown in Figure 2.3. However, only four studies reported the differential effect of maternal BMI on male and female offspring BMI in adulthood, depicted in Figure 2.4. An early childhood study by Deierlein et al. found that in an adjusted model, an increase in maternal BMI was associated with an increase of 0.02 (95% CI: 0.003–0.050) weight for length z-score (WLZ) in their 6 month old offspring [69]. Similarly, weak effect sizes were reported by Mesman et al. in 14- month old offspring—0.031 kg/m2 (95% CI: 0.019–0.043) BMI increase with each unit increment in maternal BMI [85]. Two studies have reported outcomes for offspring in adolescence [74, 80]. Gaillard and colleagues found that a unit increase of maternal BMI was associated with an increase of 0.32 kg/m2 (95% CI: 0.26–0.37) BMI among 17 year old offspring [74]. Similarly, in Europe, each unit increment of maternal BMI was associated with an 0.39 kg/m2 (95% CI: 0.30–0.47) BMI increase in 18 year old offspring, but this study included only male offspring [80]. Amongst adult offspring, the effect sizes were larger than at earlier ages in other studies. Reynolds et al. reported that each unit increase in maternal BMI was found to be associated with an 0.35 kg/m2 increase in offspring BMI at 30 years [88]. There was a slightly higher effect size among those in a 32 year old Jerusalem cohort: Hochner et al. found that a unit increment of maternal BMI was associated with an offspring BMI increase of 0.481 kg/m2 (95% CI: 0.377– 0.585) [102]. In a Brazilian young adult cohort, each unit of maternal BMI increase was associated with an 0.65 kg/m2 (95% CI: 0.41–0.88) BMI increase in male adults and an 0.94 kg/m2 (95% CI: 0.66–1.22) increase in female adults [94]. Findings from the Northern Finland Birth cohort show that each unit increase of maternal BMI was associated with an increase in adult offspring BMI of 0.11 kg/m2 (95% CI: 0.07–0.16) for males and 0.14 kg/m2 (95% CI: 0.09–0.18) for females [103]. A study among two different age groups from a similar cohort has also been reported by Terry and colleagues [96]. This study found that there was a significant association between maternal BMI and adult offspring BMI. They reported that each unit increase in maternal BMI was associated with a 0.30 kg/m2 (standard error [SE]: 0.08) BMI increase in offspring at age 20 years and 0.38kg/m2 (SE: 0.13) at age 40 years.

25

Hochner (2012) 32 years (P <0.001) Reynolds (2010) 32 years Gaillard (2015) 17 years

Mesman (2009) 14 months ■ Linear regression with (p-value)  Linear regression and confidence interval (CI) Deierlein (2010) 6 months

-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Beta coefficients, 95% CI or P-value

Figure 2.3 Summary of maternal and offspring BMI association followed up in months and years (childhood, adolescence and adulthood)

Female ▲ Female (regression and CI)) → Female (regression and SE) Terry (2007) ■ Male (Regression and CI) 40 years (SE: 0.13) Laitinen (2001) 31 years Tequenes (2009) 23 years Terry (2007) 20 years (SE: 0.08) Male

Laitinen (2001) 31 years Tequenes (2009) 23 years Koupil (2008) 18 years

-0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Beta Coefficients, 95% CI or SE

Figure 2.4 Summary of maternal and offspring BMI association in adult males and females

26

Table 2.3 shows six studies that have examined the parental BMI and offspring adiposity association (sum of skinfold thickness, fat mass index or body fat percentage) [62, 64, 65, 75, 84]. One of these studies reported that, after adjustment for other early life factors, increased maternal and parental BMI were not significantly associated with offspring skinfold thickness at 2 years [62]. However, among a Singapore birth cohort at the two-year follow-up, significant associations were found between maternal BMI and offspring skinfold thickness [84]. They reported that each unit increase of maternal BMI was associated with a 0.033 cm (95% CI: 0.014–0.052) increase in subscapular and 0.025 cm (95% CI: 0.006–0.044) increase in triceps skinfold thickness at age 2 years. Castillo et al. reported that a unit increase in adjusted maternal BMI was associated with an increase in 0.07 kg/m2 (95% CI: 0.02–0.11) fat mass index and 0.18% (95% CI: 0.03–0.34) increase in body fat percentage in childhood [64]. In a follow-up study, reported by Gale et al., among 9 year olds at Prince Anne Maternity Hospital, Southampton, it was noted that a unit increase of maternal BMI was associated with an increase in fat mass index of 0.26 kg/m2 (95% CI: 0.04–0.48) in male offspring and 0.42 kg/m2 (95% CI: 0.29–0.56) in female offspring [75]. Furthermore, a study of children aged 9 years who were in the highest tertile for percentage body fat showed higher maternal BMI [65]. Blair et al. found that overweight and obese mothers were associated with offspring change in percentage body fat, with an increase of 3.4 units (95% CI: 1.4–6.0) and 4.0 units (95% CI: 0.4–7.7), respectively, in the final model [63]. Table 2.4 presents another six studies that reported associations between higher maternal BMI and childhood offspring BMI, risk of overweight offspring, and offspring body fat percentage [68, 73, 87, 89, 92, 95]. Adjusted maternal BMI was associated with an increase of 0.10 kg/m2 (95% CI: 0.08–0.12) in BMI, 0.21% (95% CI: 0.13–0.29) in percentage body fat and 1.15 times (95% CI: 1.10–1.20) increased risk of overweight offspring at 5–6 years [73]. Terry et al. found that an increase in a unit of maternal BMI was associated with an 0.07 kg/m2 (SE: 0.004) BMI increase in their 1-year old offspring [95]. They also reported that higher maternal BMI is associated with a 1.10 times risk (95% CI: 1.07–1.12) of having overweight offspring. Similarly, among Chilean children higher maternal BMI predicted offspring overweight at 7 years, with unadjusted odds of 1.97 (95% CI: 1.38–2.80) [89]. Another study showed that higher maternal BMI was a predictor of overweight children at 2 years, with odds of 1.07 (95% CI: 1.04–1.10) [68]. The risk of being an obese child at 4 years was predicted by maternal BMI after adjustment for gestational age and parity with odds of 2.27 (p <0.001) [87]. A study done by Stettler et al. demonstrated that adiposity risk (sum of skinfold thickness) in young adult offspring was associated with a 1.15 increase (95% CI: 1.06–1.25) for each unit increase in maternal BMI [92].

27

Table 2.3 Summary of maternal and offspring adiposity association in childhood Author and Cohort profile/ Population, Exposure Outcome Outcome Results year population country measured age (yrs) Ay et al., Generation R Study 1039 mothers During Skinfold 2 Adjusted parental BMI not associated with 2008 [62] and child dyads, pregnancy offspring SFT at 24 months. But weight and height USA of parents independently associated.

Lin et al., Growing in 976 mothers and During Skinfold 2 Maternal pre-pregnancy BMI increase of 1 SD was 2015 [84] Singapore Towards a child dyads, pregnancy associated with skinfold thickness of 3.28% (1.43– Healthy Outcomes Singapore 5.17) and 2.46% ( 0.61–4.36), subscapular and (GUSTO) triceps at 2 yrs

Castillo et Perinatal Pelatos 3156 mothers During Body fat 6–7 Adjusted maternal pre-pregnancy BMI unit al., 2015 Study and child dyads, pregnancy % increase, associated with offspring 0.07 kg/m2 [64] Brazil (0.02–0.11) FMI and 0.18 % (0.03–0.34) body fat % at 6 yrs

Gale et al., Prince Anne Maternity 216 mothers and During FMI 9 In final model, 1 SDS maternal pre-pregnancy BMI 2007 [75] Hospital, child dyads, UK pregnancy increase, was associated with an 0.26 SDS (0.04– Southampton 0.48) for males and 0.42 SDS (0.29–0.56) of FMI

Catalano et Metro Health 63 mothers and During Body fat 9 Obese pre-pregnancy mothers predicts OR 4.03 al., 2009 Medical Centre child dyads, pregnancy % (1.23–13.22) in 3 tertiles BMI in the adjusted model [65] USA (FCV) on offspring Blair et al., The Auckland 591 mothers and Recalled Body fat 7 Adjusted maternal pre-pregnancy BMI of OW/OB 2007 [63] Birthweight child dyads, % associated with 3.7% (1.4–6.0) and 4.0% (95% CI Collaborative Study New Zealand 0.4–7.7) body fat of the child

Abbreviations: SFT = subcutaneous fat mass; FMI = fat mass index; FCV = first clinical visit; OW/OB = overweight or obese; SDS = standard deviation scores Note: Confidence intervals in brackets are at 95%

28

Table 2.4 Summary of maternal and offspring BMI and obesity association childhood and adulthood Author and Cohort profile/ Population, Exposure Out- Outcome Results year population country measured come age (yrs) Gademan et Amsterdam Born 1727 mothers During BMI 5–6 In final model, a maternal pre-pregnancy BMI al., 2013 [73] Children and their and child dyads, pregnancy and OB increase of 1 unit kg/m2, 0.10 (95% CI 0.08– Development Study Netherlands 0.12) in OW mothers predicted OR: 1.15 (ABCD) (1.10–1.20) risk of OW at 6 yrs

Terry et al., Collaborative Perinatal 20,523 mothers During BMI 7 Increase in maternal pre-pregnancy BMI of 1 2011 [95] Project, USA and child dyad, pregnancy and OB kg/m2, was 0.07 kg/m2 (p <0.005), OR 1.10 USA (1.08–1.12) for males, and 0.07 kg/m2 (p <0.005) and OR 1.09 1.07–1.11) for females de Hoog et al., Amsterdam Born 3156 mothers During OW 2 Maternal OW predicted OR: 1.09 (1.07–1.12) 2011 [68] Children and their and child dyads, pregnancy OW at 2 yrs Development Study Netherlands (ABCD) Olson et al., Bassett Mothers Health 321 mothers and During OW/OB 4 Adjusted maternal pre-pregnancy OW/OB 2010 [87] Project (BMHP) cohort child dyads, pregnancy predicted offspring OB, OR: 2.268 (p <0.001) USA at 4 yrs

Rios- Castillo 54 public kindergartens 652 mothers and Self- OW 7 Unadjusted OR of OW pre-pregnancy mothers et al., 2015 (National Association of child dyads, recalled 4 was 1.97 (1.38–2.80) compared to normal- [89] Kindergartens) Chile years after weight mothers of offspring OW at 7 yrs birth Stettler et al., The National 447 mothers and During skinfold 23 Unadjusted maternal pre-pregnancy BMI 2000 [92] Collaborative Perinatal child dyads, pregnancy predicted offspring adiposity OR 1.15 (1.06– Project (CPP) USA 1.25) in adulthood

Abbreviations: OW = overweight; OR = odds ratio; OW/OB = overweight or obese Note: Confidence intervals in brackets are at 95%

29

Fourteen studies found that there were significant associations between parental and offspring BMI in its categories (overweight/obese) in childhood, adolescence and adulthood [37, 63, 70, 71, 76-79, 86, 90, 91, 93, 98, 99]. These studies reported outcomes as odd ratios, while another three reported outcomes as relative risks [63, 71, 90]. Only four studies (Table 2.5) were found to have both parents overweight/obese measured as exposure [37, 70, 77, 86]. Another ten studies reported maternal overweight/obese and offspring overweight/obese associations (Figure 2.5).

Durmus et al. reported that when both parents were obese, the risk of their child being overweight was 6.52 (95% CI: 3.44–12.38) compared to a child with normal-weight parents [70]. From the ALSPAC cohort, Reilly et al. reported that the risk of offspring being obese at 7 years was 10.44 times higher (95% CI: 5.11–21.32) than offspring of normal-weight parents. They also reported separately the impact of obese mothers and obese fathers, which was associated with 4.25 times (95% CI: 2.86–6.32) and 2.54 times (95% CI: 1.72–3.75) the risk of having obese children compared to normal weight mothers and fathers, respectively [104]. Similarly, in the MUSP cohort, O’ Callaghan et al. reported that overweight and obese offspring were separately predicted by overweight and obese parents [86]. For example, they found that offspring of obese mothers had 3.9 times (95% CI: 2.3–6.4) the odds of being obese at 5 years compared to normal-weight mothers. However, the risk was not significant for maternal–offspring overweight (OR: 1.2, 95% CI: 0.8– 1.8). Jaaskelainen et al. found that parental-offspring overweight/obesity associations differed by gender in adolescent offspring [77]. As such, they found that the odds of male offspring of overweight mothers being overweight/obese at age 16 years were 2.03 times higher (95% CI: 1.46- 2.81) compared to male offspring of normal weight mothers. Daughters of overweight mothers were 2.73 times (95% CI: 1.97–3.77) more likely to be overweight/obese compared to daughters of normal-weight mothers. They found that overweight fathers predicted male offspring being overweight/obese at age 16 years with the odds of 1.95 (95% CI: 1.48–2.58) compared to sons of normal–weight fathers, while the female offspring of overweight fathers showed odds of 2.61 (95% CI: 1.94–3.50) for overweight compared to daughters of normal–weight fathers. There are no controversies, but more similarities and agreement that parental overweight and/or obesity predicts the risk of offspring overweight and obesity later in life; the risk was higher when both parents were overweight/obese compared to normal weight parents.

30

Table 2.5 Summary of parental and offspring overweight/obese (BMI categories) in childhood, adolescence and adulthood Author and Cohort Number of Exposure Outcome Outcome Results year profile/ population, measured age (yrs) population country Durmus et Generation 5227 parents During OW 4 OB parents in final al., 2012 R Study and child pregnancy model predicted OR: [70] trios, USA 6.52 (3.44–12.38) child OW at 4 yrs O' Callaghan The Mater 4062 parents During OW/OB 5 Adjusted pre-pregnancy et al., 1997 University and child pregnancy maternal OW predicted [86] of trios, (FCV) offspring OB OR: 2.4 Queensland Australia (1.6-3.8); and maternal Study of OB predicted child OB Pregnancy, OR: 3.9 (2.3–6.4). Australia Paternal OW predicted [1980– OR 2.8 (1.8–4.5) OW 1984] offspring; and paternal OB predicted OR: 2.0 (1.1–3.6) OB offspring; and OR: 2.1 (1.4.0–3.3) OW offspring at 5 yrs Reilly et al., Avon 5493 parents During OB 7 In final model, both OB 2005 [37] Longitudin and child pregnancy parents predicted OR: al Study of trios, UK 10.44 (5.11–21.32), parents and paternal OB OR: 2.54 children, (1.72–3.75) and UK late maternal OB OR: 4.25 1980s (2.86–6.32) offspring OB at 5 yrs Jaaskelainen Northern 4788 parents During OW/OB 16 Parental pre-pregnancy et al., 2011 Finland and child pregnancy OB in prediction of [77] Birth dyads, offspring OW: mother– Cohort, Finland son: OR: 4.36 (2.50– Finland 7.59); mother–daughter: [1986] OR 3.95 (2.34–6.68). while, father–son: OR 3.17 (1.70–5.92); father–daughter: OR 5.58 (3.09–10.07)

Abbreviations: OW = overweight; OB = obese; OW/OB = overweight or obese; OR = odds ratio Note: Confidence intervals in brackets are at 95%

Ten studies identified maternal overweight and/or obesity before pregnancy as an independent predictor of offspring overweight and/or obesity regardless of other prenatal and early childhood factors [63, 71, 76, 78, 79, 90, 91, 98, 99, 105]. Of these ten studies, three studies reported their results as relative risks [71, 78, 90]. Children of overweight and obese mothers carried a risk of being overweight and/or obese in childhood of 1.4 times and of 6.38 times for being overweight/or obese in adulthood (Figure 2.5). We categorised the outcome into under 5 years [78, 79, 98, 99], 5 31

to 10 years [76], and more than 10 years [90, 91, 93]. In Figure 2.5, the distribution of odds ratio and relative risk were based on offspring age, the least risk to the highest risk, and lower BMI categories (overweight) to the highest BMI categories (obesity) of studies found. There seems no difference between the risk of offspring being overweight or obese under 5 years and above 5 years, when their mothers were overweight, obese, or overweight/obese. Given the variability in outcome measures used (such as overweight, obese, overweight/obese, body fat percentage, combined female and male offspring and separated male and female offspring) pooled data analysis or meta- analysis were not possible.

32

AgeAge > 10> 10years years Steube (2009); 36 - 56y

Steube (2009); 36 - 56y

Rooney (2010); 19y

Steube (2009); 18y

Steube (2009); 18y

Salsberry (2007); 13y

Salsberry (2007); 13y Age >5 years <10 Age > 5 to < 10 years years p < 0.001 Hamilton (2010); 9y p < 0.001 Hamilton (2010); 9y Age <5 years Age < 5 years

Janjua (2012); 5y

Janjua (2012); 5y

Janjua (2012); 5y

Janjua (2012); 5y

Kitsantas (2010); 4y

Whitaker (2004); 4y

Whitaker (2004); 4y

Whitaker (2004); 4y

Fairley (2015); 3y

Fairley (2015); 3y

Whitaker (2004); 3y

Whitaker (2004); 3y

Whitaker (2004); 3y

Whitaker (2004); 2y

Whitaker (2004) ; 2y ■ Relative Risk (RR) Whitaker (2004) ; 2y  Odds Ratio (OR) Wen (2015); 2y

-2 0 2 4 6 8 10 12 Figure 2.5 Summary of maternal and offspring overweight/obese (BMI categories) in childhood and adolescence 33

After birth parental – offspring BMI and obesity association

Twenty-three studies were found to investigate postnatal parental–offspring BMI and obesity associations [34, 106-127]. Studies were designated as ‘postnatal’ when the degree of obesity of the parents (father, mother or both) was measured at least one month after the birth of the index child. Six studies reported a direct (continuous) association between parental and offspring BMI [107, 108, 113, 115, 116, 125]. Another fourteen studies reported results in terms of risk–of overweight and obesity in relation to parental-weight status. Another two studies reported parental BMI and offspring adiposity (body fat percentage and skinfold thickness) and one study reported parental and offspring waist circumference [106, 120, 122]. Comparison of all studies was difficult as the definitions of outcomes and the referent or control groups varied markedly.

Five of the six studies which presented continuous results (summarised in Table 2.6), found a positive association between parental BMI and offspring BMI [107, 108, 113, 115, 116, 125]. Two of these studies reported significant associations between parental and adolescent BMI [113, 125], while another three studies reported that parental BMI measured during childhood has significant associations with the adult offspring BMI [108, 113, 115, 116]. However, one study by Chen et al. (2007) reported a significant association only between maternal BMI and offspring BMI [107]. There were three studies as shown in Table 2.7 which reported postnatal parental BMI associations with offspring adiposity using measures such as body fat percentage, skinfold thickness and abdominal obesity [106, 122, 128]. Ajala et al. found a moderate (r = 0.30, p <0.001) association between parental BMI and offspring skinfold thickness at 8 years [106], while Salbe et al. reported that a unit increase in maternal BMI was associated with an 0.22 increase in childhood body fat percent. These findings, however, were not adjusted for other confounders [122]. Both female and male offspring were reported to have increased risk of adiposity when their parents had a high waist circumference. Having high risk (WC) parents was associated with an increase of 1.62 cm (95% CI; 0.09–3.16) for female offspring and 1.32 cm (95% CI: 0.09–2.55) for male offspring waist circumference, respectively [128].

34

Table 2.6 Summary of (continuous) parental and offspring BMI associations from childhood, adolescence and adulthood Author and Cohort profile/ Population, Age (yrs) Outcome Outcome Results year population country exposure age (yrs) measured Heude et al., The Fleurbix- 167 parents and 8 BMI 16 Adjusted for aged and standardised z- score, a unit 2005 [113] Laventine Ville child trios, increase in maternal BMI associated with 0.18 (p Sante Study France <0.01) BMI in offspring at 16 yrs—slightly higher (FLVS) than paternal BMI (0.09, p <0.13), but not significant

Trudeau et al., Trosi-Rivieres 134 parents and 10–12 BMI 12 Offspring BMI at 35 yrs associated with parental 2003 [125] semilongitudinal child trios, BMI, but only for mothers and daughters, study of growth Canada coefficient = 0.37 (p <0.001) and development Kivimaki et The 1918 parents 3–18 BMI 24–39 Adjusted BMI–BMI maternal–offspring 0.31 ( 0.29– al., 2007 [115] Cardiovascular and child trios, 0.36), paternal-offspring 0.29 (0.23–0.35) Risk in Young Finland Finns Study Kvaavik et al., The Oslo Youth 482 parents and 13 BMI 33 Adjusted BMI–BMI maternal–offspring 0.07 (p = 2003 [116] Study child trios, 0.12), paternal–offspring 0.12 (p <0.001) Finland Cooper et al., Perinatal 9346 parents 11 BMI 44–45 Adjusted BMI–BMI maternal–son 0.21 (0.18–0.25), 2010 [108] Mortality Survey and child trios, maternal–daughter 0.29 (0.25–0.34); Paternal–son USA 0.27 (0.23–0.31), paternal–daughter 0.22 (0.17– 0.28)

Chen et al., Elementary school 307 mother and 7–8 BMI 8–9 Adjusted higher BMI–BMI maternal–offspring 2007 [107] in Northern child dyads, 0.062 (p <0.02) Taiwan Taiwan Abbreviations: - Note: Confidence intervals in brackets are at 95%

35

Table 2.7 Summary of parental BMI and offspring adiposity in childhood and adolescence Author and year Cohort profile/ Population, Age (yrs) Outcome Outcome Results population country exposure age (yrs) measured

Ajala et al., 2011 Early Bird Study 224 parents and 5 Skinfold 8 Parental BMI associated with offspring SF [106] child trios, UK thickness (r=0.30, p <0.001). Daughter–mothers (r=0.39, p <0.001), daughter–fathers (r=0.26, p<0.01), sons–mother (r=0.22, p <0.01) and sons– fathers (r=0.06, p <0.5) Salbe et al., 2002 Pima Indian 138 parents and 5 Body fat % 10 A unit increase in maternal BMI was [122] children child trios, USA associated with 0.22 kg/m2 (p <0.05) of offspring Body fat %. Mostazir et al., The Early Bird 233 Parents and 5 WC 15 In adjusted model, the mean WC of daughters 2012 [128] Study child trios, UK of HR mothers was 1.62 cm (0.09–3.16) greater than NR. The mean WC of sons of HR fathers was 1.32 cm (0.09–2.55) higher than NR fathers Abbreviations: SF = skinfold; WC = waist circumference; HR = high risk; NR = No risk Note: Confidence intervals in brackets are at 95%

36

In Table 2.8 and 2.9, seven out of fourteen studies, which presented the results as categories, found an association between parental and offspring overweight/obesity in childhood when offspring were aged 4 to 12 years [110-112, 118, 121, 123, 124]. From these seven studies, three studies reported children, who were exposed to a combination of both parents either overweight or obese, were at risk of being overweight after three to eight years follow-up [111, 112, 123]. Another three of these studies separately reported mothers and fathers who were overweight or obese in relation to the risk of offspring overweight/obesity [118, 121, 124], while one study reported only maternal overweight associations with offspring overweight risk in childhood [121]. Another seven studies examined the impact of parental overweight/obesity on adult offspring overweight/obesity. Four studies reported the risk of overweight/obese offspring when a combination of both parents were overweight/obese [34, 117, 119, 126]. Another three studies reported separately on the differential effects of overweight and/or obese mothers and fathers on the risk of offspring being overweight and/or obese in adulthood [109, 114, 127].

There were slightly higher odds of children of overweight/obese parents being overweight/obese in adulthood compared to children of overweight/obese parents who were followed-up in short-term periods. The risk of having obese offspring later in life were higher when both parents were obese compared to offspring with normal or overweight parents. Freeman et al. reported that children of overweight parents during early childhood were 2.52 times (95% CI: 1.07–3.79) more likely to be overweight at 8–9 years of age than children of normal-weight parents [112]. They also reported that children of obese parents were 11.09 time (95% CI: 1.66–74.09) more likely to be obese compared to children of normal-weight parents. Dubois et al. found that at a 4-year follow-up, children of overweight parents were 3.2 times (95% CI: 1.7–5.8) more likely to be overweight compared to children of normal parents, while children with one overweight parent (either father or mother) were 2.1 times (95% CI: 1.3–3.6) more likely to be overweight compared to children with normal-weight parents (after adjustment for gestational age and the child’s birth weight) [111]. Another study found that children of overweight parents (measured at 6 months after birth) were at risk of becoming overweight at the 9-year follow-up(r = 0.38; p <0.001) [123].

Unadjusted associations were reported by De Coen et al. They found that children of overweight mothers were 2.99 times (95% CI: 1.80–4.97) more likely to become overweight at 5 years old compared to children of normal weight mothers [110]. They also reported however, that there was no significant increased risk of being overweight among children of an overweight father. Li et al. found that children of obese mothers and obese fathers were 2.9 times (95% CI: 1.8–4.8) and 2.6 37

times (95% CI: 1.7–4.1), respectively, more at risk of being overweight and obese at 9 years old compared to children with normal mothers and fathers [118]. They also found that children with obese mothers and obese father have a 4.8 times (95% CI: 2.5–8.6) and 4.4 times (95% CI: 2.5–7.8) higher risk of being obese at 9 years old compared to children of normal-weight fathers and mothers. Children with two obese parents were 4.9 times (95% CI: 3.0–8.0) more likely to be obese at the 3-year follow-up compared to children with two normal-weight parents. Suka et al. [124] found that male and female children of overweight/obese mothers were 2.43 times (95% CI: 1.60– 3.61) and 3.77 times (95% CI: 2.64–5.43) more likely respectively to be obese at age 9 years compared to children with normal weight mothers. They also found that male and female children of overweight/obese fathers were at risk of being obese at 2.12 (95% CI: 1.61–2.77) and 2.25 (95% CI: 1.72–2.95) times, respectively, compared to children with normal-weight fathers. These results were adjusted for obesity at baseline. Olvera et al. reported that children with overweight mothers at baseline (4 years old) were 3.3 times (95% CI: 0.63–17.63) more likely to be overweight after 8 years compared to children of normal–weight mothers. However this association was no longer significant after adjustment for other confounders [121].

Whitaker et al. reported that children with two obese parents were at 13.6 times (95% CI: 3.7–50.4) higher risk of being obese in adulthood compared to children with normal-weight parents [126]. They also reported that if any one of the parents was obese, the risk of the child being obese in adulthood was 3.2 times (95% CI: 1.8–5.7) compared to children with normal weight parents. Lake et al. [117] found that male and female children of overweight parents carried higher risks of being obese at 33 years: 3.41 times (95% CI: 2.36–4.93) and 2.65 times (95% CI: 1.95–3.62), respectively, compared to children of normal-weight parents. They also reported that male and female children of obese parents were 8.42 times (95% CI: 5.47–13.00) and 6.75 (95% CI: 4.57– 9.96) at risk of being obese, respectively, compared to children of normal parents. Another study done by Magarey et al [119] found that children of overweight parents had 4.3 times (95% CI: 2.1– 8.9) the risk of being overweight at 20 years compared to children of normal-weight parents. They also reported that if any one of the parents was overweight, the risk of the child being overweight was 2.1 times (95% CI: 1.0–4.5) higher compared to the children of normal-weight parents. Han et al. reported that when both parents were obese, the odds of their child being obese in adulthood was 11.77 times (95% CI: 7.70–18.00) higher compared to a child of normal-weight parents [34]. They also reported that when both parents were overweight, the odds of their child being obese as an adult were 10.40 times higher (95%CI: 4.65-13.29) compared to a child of normal–weight parents.

38

The odds of the child becoming an overweight or obese adult when only one of their parents was overweight or obese were 1.96 (95% CI: 1.54–2.48) for one overweight parent, and 2.14 (95%CI: 1.62-2.81) for one obese parent compared to the child of normal-weight parents.

Williams [127] reported the results as relative risks and found that children of obese mothers carried 1.9 times (1.2–2.5) the risk of becoming overweight adults compared to children of normal-weight mothers, while children of obese fathers carried 1.3 times (95% CI: 1.0–1.7) the risk of becoming overweight adults compared to children of normal-weight fathers. Crossman et al. reported that sons of an obese father were1.6 times more likely to become overweight adults compared to sons of normal-weight fathers while daughters of obese mothers were 1.39 times more likely to become overweight adults compared to daughters of normal mothers [109]. Juonala et al. found that children of mothers with higher BMI were 1.64 times (95% CI: 1.47–1.83) more likely to become obese adults compared to children of normal-weight mothers, while children of higher-BMI fathers were 1.57 times (95% CI: 1.39–1.77) more likely to become obese adults compared to child of normal- weight fathers [114].

39

Table 2.8a Summary of parental and offspring overweight/obese association in childhood and adulthood Author Cohort profile/ Population, Age Out- Outcome Results and year population country exposure come age (yrs) measured

Dubois and Quebec Longitudinal 1550 parents 5 mths OW 4.5 In adjusted OR, one parent OW/OB predicted 2.1 (1.3–3.6) of Girard., Study of Child and child offspring OW; and both OW/OB parents predicted OR: 3.2 (1.7–5.8), 2006 [111] Development trios, Canada compared to normal-weight parents (QLSCD) Stewart- Hospitals and health 150 parents 6 mths OW 9.5 Parental OW correlates with offspring OW at 9.5 years (0.38, p Agras et al., community in San and child <0.001) 2004 [123] Francisco Bay Area. trios, USA Freeman et Longitudinal Study 3285 parents 4–5 yrs OW/ 8–9 Adjusted model, parental OW predicted offspring OW/OB at follow- al., 2011 of Australian and child OB up, OR: 2.02 (1.07–3.79 p <0.05) and OR: 5.32 (1.14–24.74, p [112] Children trios, <0.05). Both OB parents predicted OB offspring, OR: 11.09 (1.66– Australia 74.09, p <0.05) Li et al., Sydney aircraft noise 586 parents 9 yrs OW/ 12 Parental OB predicted an offspring BMI increase of more than 2 2008 [118] and child blood and child OB kg/m2 at 12 years with RR: 1.5 (0.7–3.2) for maternal OB and RR: pressure study trios, 1.3 (0.7–2.4) for paternal OB. Although not significant Australia De Coen et Prevention of 473 parents 5 yrs OW 6.5 and 8 Adjusted baseline BMI. Parental OW/OB predicted offspring OW al., 2013 overweight among and child at18 months, OR: 2.62 (1.46–4.69) for mothers, and OR: 2.16 (1.21– [110] pre-school and trios, 3.86) for fathers. At 30 months, OR: 2.99 (1.80–4.97) for mothers, school children (POP Belgium and OR: 2.01 (0.97–4.16) for fathers project) Suka et al., Toyoma Prefecture, 6102 parents 11 yrs OB 9 Compared with normal-weight parents, paternal OB at baseline 2002 [124] Japan and child predicted offspring OW/OB at 9 years with odds of 2.12 (1.61-2.77) trios, Japan for males and 2.25 (1.72-2.95) for females while mothers OW/OB predicted 2.43 (1.60-3.61) for males and 3.77 (2.64-5.34) for females Olvera et Al Bienester del 69 Mother 4–8 yrs OW 7–12 Maternal OW/OB at baseline predicted offspring OB during al., 2007 Nino ( to the and child childhood OR: 1.2 (1.05-1.31), increasing in adolescence, OR: 2.7 [121] Wellbeing of the dyad, USA (CI: 0.61-11.81) and adulthood, OR: 3.3 (0.63-17.63), but not child) significant Abbreviations: Abbreviations: OW = overweight; OB = obese; OW/OB = overweight or obese; RR = relative risk; OR = odds ratio Note: Confidence intervals in brackets are at 95%

40

Table 2.9 Summary of parental and offspring overweight/obese association in childhood and adulthood Author Cohort profile/ Population, Age Out- Outcome Results and year population country Exposure come Age (yrs) measured

Han et al., The Midspan 2338 Parents and 4–8 yrs OB 33–59 In the adjusted model, one OB parent at baseline predicted 1.96 2015 [34] Family Study, child trios, (1.54–2.48) OW and 2.14 (1.62–2.81) OB in adult offspring. Two Scotland Scotland OB parents at baseline predicted 10.40 (4.65–13.29) OB and 11.77 (7.70–18.00) OW in adult offspring. Magarey et The Adelaide 155 parents and 8 yrs OW/ 20 Both parents OW/OB predicted offspring OW/OB at 11, 15 and 20 al., 2002 Nutrition Study child trios, OB yrs, RR: 26.7(3.6–196.4), RR: 6.6 (2.3–19.0), and RR: 4.3 (2.1– [119] (ANS) Australia 8.9). Maternal OW/OB predicted OW/OB at 11 and 15 yrs, RR: 2.2 (1.4–3.4) and 3.8 (2.0–6.9), paternal OW/OB predicted offspring OW/OB, RR: 2.0 (1.3–3.2) and RR: 1.1 (0.6–1.8) Lake et al., The British Birth 11407 Parents and 11 yrs OB 33 Parental OB predicted adult offspring OB with OR: 8.42 (5.47– 1997 [117] Cohort child trios, UK 13.00) for sons and OR: 6.75 (4.57–9.96) for daughters Whitaker et Group Health 854 parents and 2 yrs OB 21–29 Parental OB during childhood to adolescence predicted offspring al., 1997 Cooperative of child trios, USA OB in adulthood, OR: 13.6 (3.7-50.4), 15.3 (5.7–41.3), OR: 5.0 [126] Puget Sound (2.1–12.1), OR: 2.0 (0.8–5.2), OR: 5.6 (2.5–12.4) Williams., Queen Mary 482 parents and 11 yrs OW 20 In adjusted model, parental OB compared with normal-weight at 2001 [127] Hospital child trios, New baseline predicted offspring OW at 21yrs with RR: 1.9 (1.2–2.5) Zealand and RR: 1.3 (1.0–1.7)

Crossman The National 6378 Parents and 12–20 yrs OW/ 18–26 In final model, parental OB at baseline predicted daughter OB at 6 et al., 2006 Longitudinal child trios, USA OB yr follow-up, OR: 1.39 (p <0.05) for maternal and OR: 1.58 (p [109] study of <0.05) for paternal OB. While, predicted OB for son only was OR: Adolescence 1.60 ( p <0.05) Health Juonala et The 2119 parents and 3–18 yrs OB 24–39 Maternal BMI predicted offspring OB (age and sex adjusted) in al., 2011 Cardiovascular child, Finland adulthood with OR: 1.64 (1.47–1.83) and paternal OR: 1.57 (1.39– [114] Risk in Young 1.77). In final model, maternal BMI increased 1 kg/m2; family Finns Study income and childhood BMI associated with the increase in offspring BMI (AUC: 0.772, 0.745–0.800) Abbreviations: OW = overweight; OB = obese; OW/OB = overweight or obese; RR = relative risk OR = odds ratio Note: Confidence intervals in brackets are at 95%

41

Parental BMI and obesity in relation to the trajectories, tracking and change in offspring BMI and overweight/obesity

Several studies have reported that parental BMI and obesity impact on offspring BMI trajectories and change, across age periods (Table 2.9). One study has shown that maternal BMI and obesity are associated with offspring body composition from birth to childhood. This association was found to differ with the sex of the offspring. Andres et al. (2015) reported that maternal overweight and obesity were significantly associated with offspring body fat trajectories [129]. They found that boys born to obese mothers showed higher body mass index for age z-score (BAZ) and body fat and lower fat free mass (FFM) starting at 2 years and maintained until 6 years compared to boys born to normal-weight mothers. By contrast, girls born to obese mothers increased their body fat at age 5 to 6 years, while girls from normal-weight mothers showed a decline in body fat after age 3. However, there were no significant differences in the patterns of trajectories between maternal normal weight, overweight and obese mothers. Giles et al. (2015) reported that maternal obesity in early pregnancy was associated with the risk of accelerating growth trajectories during childhood [130]. They found four distinct growth trajectories from birth to 3.5 years in an Australian birth cohort. Maternal obesity was associated with an accelerating trajectory with odds of 3.72 (95% CI: 1.15–12.05) compared to the intermediate growth group. Haga et al. (2012) reported that in a Japanese cohort of children born in Koshu City there were six distinct BMI trajectories from birth to 11 years [131]. They reported that maternal BMI in early pregnancy, maternal smoking during pregnancy, and unfavourable behaviours during early pregnancy were associated with the ‘progressive overweight’ and ‘progressive obesity’ groups. Three trajectories of overweight among participants in the US National Longitudinal Survey of Youth (NLSY79) at 2 to 12 years were reported by Li et al., (2007). The trajectories were normal (83.9%), early onset (10.9%) and late onset (5.2%) [132]. Pre- pregnancy maternal overweight and obesity were suggested as one of the factors associated with offspring early onset and late onset of overweight. For example, the reported adjusted odds of children with obese mothers were 5.8 times (95% CI: 2.6–13.0) for late onset classes of childhood overweight compared to children of normal-weight mothers. Considering early childhood, Pryor et al. (2015) reported that three trajectories of overweight were found for children in Quebec from 5 months to 5 years: early onset (11.0%), late-onset (16.6%), and never overweight (72.5%) [133]. In their findings, parental overweight was reported to predict both early onset and late onset overweight groups. When both parents were overweight, the risk of a child being in the early and late onset groups was 7.29 times (95% CI: 4.08–13.02) and 5.01 (95% CI: 3.24–7.74) higher, respectively, compared to offspring of the never overweight group.

42

Fuiano et al. (2008) reported that parental BMI was associated with the risk of offspring being persistently overweight or obese in childhood (1 to 2 years follow-up) among Italian school children [134]. Overweight and obese parents were reported to predict persistently overweight and obese children at the two year follow-up. When both parents were obese at baseline, the risk of offspring being persistently overweight/obese throughout the two-year study was 3.90 times (95%CI: 2.28- 6.70) higher compared to offspring of normal-weight parents. Kristiansen et al. (2014) reported that among 1256 Norwegian birth-cohort children, parental overweight was associated with the risk of having a stable high-body-size at 7 years of age after adjustment of prenatal and offspring early life factors [135]. A stable high-body-size was defined as those in the highest tertile of the ponderal index at birth and overweight/obese at 7 years of age. Parental overweight/obesity was reported to predict persistent high-body-size of offspring at 7 years of age with odds of 2.16 (95% CI: 1.57– 2.97) for mothers and 1.40 (95% CI: 1.00–1.93) for fathers compared to normal-weight status parents. They found that the odds were significantly higher among female offspring compared to males. Recently, researchers using a Netherlands birth cohort found four overweight trajectory groups from birth to 11 years (never overweight, persistent overweight, overweight reduction, and late onset overweight) [136]. They reported that among the persistent overweight group, children were more likely to have overweight parents in childhood compared to the overweight reduction group children. The odds of children at 8 years being in the persistent trajectory group if they had overweight mothers before pregnancy and overweight fathers were 1.85 (95% CI: 1.37–2.49) and 1.75 (95 CI: 1.21–2.55), respectively. Ziyab et al. (2014) identified four overweight trajectories among a birth cohort from The Isle of Wight in the United Kingdom who were followed for 18 years. They defined the trajectories as normal, early persistent obesity, delayed overweight, and early transient overweight [137]. Maternal early-pregnancy overweight (OR: 3.16 95%CI: 1.52- 6.58) was reported as one of the strong independent predictors associated with membership of the early-persistent obesity group of offspring compared to normal-weight mothers.

Heude et al. (2005) reported that the cross-sectional relationship between parental–offspring anthropometric measurements from childhood to post-puberty and parental BMI and adiposity was associated with offspring BMI and adiposity [113]. They found that the association between maternal and offspring BMI and anthropometric measurements decreased in the first two years of life and became evident again after 2 years of age. However, the relationship between paternal and offspring BMI and anthropometric measurements that were persisting after 2 years of age differed between differing parameters of adiposity distribution across childhood (e.g. skinfold thickness,

43

waist and hip circumference). Further, the prospective associations with mothers’ BMI at baseline with offspring BMI change were significant, but this was not found for fathers’ BMI. Similarly, Linabery et al. (2012) reported that the Fels birth cohort study showed that parental weight-status categories (normal weight, overweight and obese) were associated with offspring growth (BMI) from infancy to 3 years of age [138]. They compared paternal–maternal overweight, obese and normal weight categories with offspring BMI trajectories and found that maternal obesity was more significantly associated with offspring early childhood growth than paternal obesity. However, they reported that maternal obesity was associated with offspring BMI at birth and 1.5 years onwards. A study done by Pryor et al. (2011) suggested three growth trajectories among children from 5 months to 8 years of age in Quebec [139]: low stable (54.5%), moderate (41.0%), and rising (4.5%). Maternal BMI and smoking during pregnancy were found to predict offspring in high-rising BMI trajectories. They found that children of overweight mothers were 2.38 times (95% CI: 1.38–4.54) more likely to have a high-rising BMI trajectory compared to children of normal-weight mothers, and the risk was 6.33 times (95% CI: 3.82–11.85) higher among children of obese mothers compared to children of normal–weight mothers after adjustment for child birth weight, gestational age, sex, maternal smoking during pregnancy, family income, and maternal education.

Only a few studies have examined the impact of parental BMI on offspring BMI and weight change [140-142], and they demonstrated that parental overweight and obesity predicted offspring weight status change from childhood to adolescence. They reported that the odds of children of overweight and obese parents (father and mother) being overweight/obese at 14 years from normal weight at 5 years were 16.7 times (95% CI: 3.97–9.59) more than children with normal-weight parents. Hesketh et al. (2009) reported that parental BMI was found to be associated with offspring BMI-z change in Australian primary school children. However this analysis was not adjusted for other early life factors [140]. Lawrence et al. (2014) reported that maternal pre-pregnancy BMI was associated with offspring BMI change from age 17 to 32 years. They found that a unit increase in pre-pregnancy BMI was associated with an 0.83 kg/m2 change in offspring BMI from adolescence to adulthood [141].

44

Table 2.10 Summary of studies on parental BMI and obesity on offspring BMI trajectories and weight change

Author and Cohort Population, Exposure Outcome Outcome year profile/ country measured age population Giles et al., The 556, Maternal BMI Association between 0–3.5 yrs 2014 [130] Generation 1 Australia in early maternal BMI and pregnancy offspring BMI growth trajectories Li et al., NLSY79 1739, USA Maternal BMI, Association between 2–12 yrs 2007 [132] recalled at maternal pre-pregnancy offspring aged BMI and offspring 14–22 yrs overweight trajectories Ziyab et al., The Isle of 1240, UK Maternal BMI, Association between 1–18 2014 [137] Wight birth antenatal visit maternal BMI and cohort offspring BMI trajectories Haga et al., Project Koshu, 1518, Japan Maternal BMI, Association between 0–12 2012 [131] Japan during maternal BMI and pregnancy: offspring BMI trajectories Maternal and Child Health Handbook Kristiansen Norwegian 3771, Parental BMI, Association between 0–7 yrs et al., 2014 Mother and Norway offspring aged parental BMI and [135] Child Cohort 6 mths offspring persistently Study (MoBa) being high-body-size van Rossem Prevention and 3550, Parental BMI, Association between 0–11yrs et al., 2014 Incidence of Netherlands offspring aged parental BMI and [136] Asthma and 3 mths offspring overweight Mite Allergy pattern (PIAMA) Andres et the Beginnings 325, USA Maternal BMI Association between 0–6 yrs al., 2014 study when offspring maternal BMI and [129] aged 3 mths longitudinal offspring BMI and body composition Mamun et The Mater 2934, Parental BMI, Association between 5 and al., 2005 University of Australia first clinical parental BMI and 14yrs [142] Queensland visit offspring weight change Study of Pregnancy (MUSP), Australia [1980-1984] Pryor et al., Quebec 1552, Maternal BMI Association between 6–12 yrs 2015 [133] Longitudinal Canada when offspring Maternal BMI (and other Study of Child aged 17 mths early factors) and Development trajectories of offspring (QLSCD) BMI mid childhood Pryor et al., Quebec 1950, Maternal BMI Association between 0.5–8 yrs 2011 [139] Longitudinal Canada when offspring maternal BMI (and other Study of Child aged 17 mths early factors) and Development offspring BMI trajectories (QLSCD) 45

Fuino et al., Southern Italy 623, Italy Parental BMI Association between From 2008 [134] (San Marco) at baseline parental BMI (other early baseline (offspring aged factors) and childhood to 2 yrs 3–8 yrs) BMI trajectories. follow- up Greenlund et CARDIA 5115, USA Parental body Association between From al., 1996 (development size at baseline parental fatness (body baseline [143] of coronary (offspring aged size) and adult offspring to 7 yrs artery disease 18–30 yrs) BMI change after 7 years follow- risk factors in up young adults) Hesketh et Health and 1234, Parental BMI Association between From al., 2009 Young Australia at baseline parental BMI and baseline [140] Victorians (offspring aged childhood BMI change to 3 yrs Study 5–10 yrs) after 3 years follow- up Heude et al, Fleurbaix- 124, France Parental BMI Association of parental From 2005 [113] Laventie Ville when child anthropometric birth to Sante Study aged 5–11yrs measurement and childhood (FLVS I and offspring adiposity II) throughout childhood Perez-pastor The Early Bird 226, UK Parental BMI Association between 5–8 yrs et al, 2009 when child parental BMI and [144] aged 5 yrs offspring BMI gain Mo-Suwan Primary and 2252, Parental BMI Association between From et al., 2000 secondary Thailand when child parental BMI and baseline [145] school, Hat aged 5 yrs overweight tracking from to 5 years Yai, Thailand childhood to adolescence follow up Linaberry et Fels 912, USA Parental BMI Association between 0–3.5 yrs al, 2013 Longitudinal when child parental BMI on infant [138] Study aged1yr growth Lorenco et Households in 225, Brazil Maternal BMI Association between 0.5–10 al, 2015 area of at baseline maternal BMI (and other yrs [146] Acrelandia, (children aged early life factors) on Western Brazil 0–36 mths) childhood BMI trajectories Lawrence et Jerusalem 556, Israel Maternal pre- Association between 17–35 yrs al., 2014 Perinatal Study pregnancy BMI maternal BMI and GWG [141] (JPS) from birth on offspring BMI change record from adolescence to adulthood

46

2.3 Other early life factors related to overweight and obesity

In addition to parental weight status and sociodemographic factors, other factors that predict changes in offspring BMI include high gestational weight gain, maternal smoking, maternal dietary habits and physical activitiy, birth weight, breastfeeding, rapid weight gain, and infant sleep duration [37, 147-149]

Gestational weight gain

Optimal maternal weight gain during pregnancy differs based on pre-pregnancy BMI; but it is generally classified according to the US Institute of Medicine recommendations [150] which are for normal-BMI women 11.5–16.0 kg, for overweight women 7.0–11.5 kg, and for obese women 5.0– 9.0 kg. Excessive weight gained during pregnancy is reported to be associated with the risk of offspring health later in life. Gaining excessive weight among mothers was associated with offsprings’ risk of overweight and obesity [93, 102, 151, 152]. Other studies have also found that excessive gestational weight gained during pregnancy is linked with offspring adiposity and abdominal obesity (Josefson, et al. [153], [154]). Josefson et al. reported in 2013 that neonates of mothers who gained excessive weight had 50% more fat mass and 3% greater body fat compared to infants born to mothers with optimal weight gain [153]. It was further concluded in a meta- analysis done by Nehring et al. in 2013, that excessive gestational weight gain contributes to 1.38 (95% CI: 1.21–1.57) pooled odd ratios for overweight in childhood [155]. In an extension of this finding, researchers using a bias-adjusted meta-analysis recently reported that excess gestational weight gain contributes to offspring obesity over the life course with a similar effect size [156]. However, some studies have reported that pre-pregnancy weight status appears to be a more important determinant of offspring overweight and obesity compared to excessive gestational weight gained [100, 157].

Nonetheless, genetic factors were also found to be associated with offspring obesity in a systematic review of twin and adoption studies by Gaillard et al. [158]. From this review, the association of environmental factors, such as home environment, with offspring obesity was strong in early childhood but disappeared at older ages among adoptees. Further, strong associations were found among twins, suggesting that genetic factors may have greater influence than environmental factors on offspring obesity. However, the outcomes of this review considered only childhood and adolescence. Long-term follow-up of twins and adopted children is ongoing. Most of the studies reviewed were conducted among caucasian cohorts and the socio-economic background of

47

participants in the adoptee studies was predominantly low income [158]. Of all these factors discussed above, parental weight status during pregnancy has been suggested by Fleten, et al. [72] as a very important early life determinant associated with offspring obesity. Moreover, maternal overweight or obesity per se has been suggested to influence childhood obesity through the intrauterine environment in the nine months of pregnancy [90, 159]. Although, there are multiple animal and epidemiological studies relating in-utero development to later obesity, it is still unclear whether the intrauterine environment may partly predict offspring obesity in a long term association [160].

Birth weight and risk of offspring overweight and obesity

Offspring birth weight has been associated with the risk of being overweight and obese later in life. Reilly et al. 2005 suggested that an extra 100g in birth weight is associated with the odds of 1.05 (95% CI: 1.03–1.07) being obese by age 7 years. Using categorised birth groups, Dubois and Girard (2006), found that the odds of overweight in childhood were 2.3 times (95% CI: 1.30–7.20) higher for infants born heavier than 4000 g compared to infants weighing 3000g to 4000g [111]. However, both of these studies were reported with unadjusted odds ratios. Other early life factors, such as maternal gestational weight gain, maternal age at birth and breastfeeding, may attenuate the link between birth weight and outcome. A recent three-year follow-up of a cohort of 55,925 children in China reported that any increase in birth weight was associated with the risk of overweight and obesity in childhood compared with infants born with a weight of 2500–2999 g. This outcome was robust after adjustment for maternal weight status, gestational weight gained and other offspring life style factors [161]. This large population study showed a short-term relationship. A systematic review and meta-analysis reported that higher birth weight (> 4000g) was associated with an increased risk of 2.07 times (95% CI; 1.91–2.24) of being obese in later life compared to those with a birthweight less than 4000 g [162]. However, this review was mainly restricted to Asian/Chinese populations and the possibilities of selection bias needs to be considered. Similar findings were reported in another review and meta-analysis including 643,902 participants in 26 countries. This review indicated that higher birth weight (> 4000g ) was associated with increased risk of overweight in the long term, with the overall odds being 1.60 times (95% CI: 1.45–1.77) more than those with a normal birth weight [163]. Although these reviews have found that higher birth weight impacts upon the development of offspring overweight and obesity throughout life, it has also been suggested this this effect may depend on maternal weight status.

48

Maternal smoking during pregnancy and the risk of offspring obesity

Several previous studies have reported that maternal smoking during pregnancy is associated with the risk of offspring overweight and obesity [107, 164-166]. Grzeskowiak et al demonstrated that maternal smoking in early or late pregnancy has a significant association with an increase in childhood BMI [167]. Researchers doing a meta-analysis of seven prospective studies reported that the odds of mothers who smoked during pregnancy of having overweight children were 1.47 (95% CI: 1.26–1.73) compared to non-smoking mothers [168]. Similar finding were observed by another meta-analysis of 14 observational studies done by Oken et al. [169]. Both reviews suggested that maternal smoking and offspring overweight and obesity associations were present in the short term, but that long-term impacts and associations of maternal smoking were less defined [170-172].

Infant breastfeeding protective against overweight and obesity

The association of breastfeeding with the risk of overweight and obesity has been widely studied. An inverse association between breastfeeding and offspring overweight has been reported in previous research. A review by Arenz et al. in 2004 of nine observational studies (69,000 participants) concluded that breastfeeding may have a continuous but small protective effect against offspring obesity in children [173]. When considered along with other early life factors, being breastfed for the first year of life was recently associated with decreased adjusted odds of 0.85 (95% CI: 0.74–0.99) for being overweight in childhood compared to those who were never breastfed [174]. Children who are exclusively breastfed for more than 6 months may have a greater advantage in terms of risk reduction of later overweight and obesity in childhood. The dose-response effects of breastfeeding may have different protective effects on the risk of overweight and obesity [175, 176]. Harder et al. (2005) reported that an increase of one month in breastfeeding duration was associated with a 4% decrease in the risk of overweight in children. Oddy et al. reported in 2014 that children who were breastfed less than 4 months showed a greater increase in BMI compared to those children who were breastfed more than 12 months [177]. However, the protective associations of breastfeeding may not persist in the long term, and long-term data remain inconclusive [178]. Although breastfeeding has been shown to protect against overweight and obesity later in life, other early life factors, such as maternal obesity, have been found to have a stronger impact on offspring risk of obesity [177, 179]. Michels et al. (2007) reported on 35,526 participants in the Nurses’ Health study who were followed up until young adulthood. They found that those who were

49

exclusively breastfed had lower rates of obesity in childhood but not in adolescence and adulthood [180].

Rapid weight gain and catch-up growth

Infants who are born underweight or preterm are potentially exposed to rapid weight gain and catch-up growth in their early life. Rapid weight gain is generally defined as abnormal acceleration of growth over a specified time period [181]. Ken and Ruth (2006) concluded that rapid infancy weight gain from birth up to 2 years of age was linked with increased obesity in childhood [182]. Monteiro et al (2005) concluded that rapid weight gain in the first year of life is associated with obesity later in the life course [183]. However, some potential concerns of these reviews were selection bias, a mixture of study designs, and differences in adjustment for confounding factors. There were also no homogenous measures of rapid weight gain between studies in order to objectively identify the exact period for intervention. Those who were born at term were also reported to experience rapid weight gain, subsequently showing an increase in the risk of overweight in later life. Hitze et al. found in 2010 that children in the adequate gestational age (AGA) group, who experienced rapid weight gain during the first two years of life, showed a higher prevalence of overweight compared to those who were not exposed to rapid weight gain [184]. This condition may differ between populations with different sociodemographic backgrounds. Goodell et al. found, in 2009 in a smaller sample size (n = 203) of children drawn from a low income population, that those who experienced rapid weight gain during their first year of life were 9.24 times more likely to be obese by the age of 3 years compared to those who had normal weight gain [181]. However, an association between rapid weight gain and the risk of subsequent obesity has also been found among high-income populations and different ethnicities. According to Stettler et al. those children from a multicentre cohort study of 5479 children aged 7 years who experienced a 100 g/month weight gain over four months were at increased risk of overweight with odds of 1.17 (95% CI: 1.11–1.24) after adjustment for birth weight and gestational age [185]. Also, in a relatively small multi-ethnic cohort, rapid infant weight gain was associated with increased risk of being overweight at 4 years of age, independently of breastfeeding history, maternal BMI and birth weight [186].

These early life factors discussed above are the influences most commonly identified as associated with the risk of overweight and obesity in offspring. These factors, however, also serve to confound or possibly mediate the associations between parental overweight and obesity and the risk of offspring overweight and obesity in later life. Combinations of several early life factors and having 50

more than one early life factor have also been found to increase the risk of overweight and obesity compared to having only one early life factor [187].

2.4 Pathways and mechanisms of parental-offspring obesity

Intergenerational studies looking at the trends in parental–offspring BMI during the current obesity epidemic have reported that there was a stable increase in parent and child association that may reflect the transmission of adiposity from parents to child [188-190]). These findings also reflect the resemblance of parents with their offspring due to genetic predisposition, shared familial environment, and intrauterine programming. Other recent reviews have described how parental- offspring BMI and obesity are driven by genetic- and epigenetic-intrauterine programming and shared familial environment [191-193]. However, no studies have objectively examined how these three mechanisms interact with each other.

Parents with adiposity genes may transmit obesity to offspring [158]. A meta-analysis of nine twin studies found that genetic factors had a strong effect on the variation of BMI from 1 to 18 years. Parental BMI showed moderate to high heritability from childhood to adolescence. Recently, researchers from the Quebec Family studies group used several adiposity phenotypes and BMI to estimate that heritability from parents to their offspring ranged from 35% to 60% [194]. The fat mass and obesity-associated (FTO) gene was the first of the adiposity-related genes that was discovered to be associated with common forms of obesity [195]. After a few waves of genome wide association studies (GWAS), a review done by Frayling et al. [196 ] recently reported that more than 35 loci of genes were found to be associated with the increase in human BMI. The search for underlying genotypes that cause obesity has been challenging, due to the complex interactions involved in the regulation of adiposity. Genetic architecture may differ across populations, social culture backgrounds and ethnicities. Therefore, some group are more likely to be predisposed to disease and obesity than other groups [196]. However, changes in genes might take time and strongly interact with environments. With the changes in energy balance through less energy expenditure and more energy intake, there may be epigenetic consequences. Biological programming of genetic factors will take time, probably years. However, epigenetic changes may be quite rapid

Intrauterine programming has been proposed as an explanation for the development of obesity [197, 198]. Initially, studies of individuals who had low birth weight (due to undernourished mothers) 51

were associated with the increase risk of being obese in later life [199, 200]. In recent years, it has been demonstrated that over-nutrition during pregnancy influences the risk of the development of obesity in offsprings in later life [201, 202]. Maternal obesity due to excess nutrient intake may be related to the high availability of plasma glucose, and amino acids [203, 204]. These characteristics may be transmitted to the child who is eventually predisposed to the risk of being overweight and obese [205, 206]. Also, previous studies have reported that maternal over-nutrition influences hormonal changes and appetite in offspring. Therefore, intrauterine programming has been proposed as an explanation for the association between low birth weight due to maternal malnutrition or high birth weight due to maternal over-nutrition [203].

Maternal diet Offspring Insuline resistance Weight gain obesity Inflammation Metabolites Hormones

Figure 2.6 Developmental programming of obesity altered by maternal nutrition (source: [203])

This concept of obesity beginning in utero and being dependent on maternal nutrition, and the hormonal and metabolic environment is shown in Figure 2.6. The probability that offspring organs, cellular composition, gene expression and epigenome may be modified as a consequence of maternal nutrition needs to be considered. When the child interacts with the postnatal environment and neonatal growth, the susceptibility towards risk of obesity may also increase (Desai et al., 2013). Nutrition provides fuel to accommodate the optimal needs of foetal growth and development, thus sub-optimal maternal nutrition may create a susceptibility to diseases in later life.

Finally, through familial shared-environmental factors, parents may influence the development of weight and BMI in their offspring. For example, living in an obesogenic family, children may acquire unhealthy dietary habits, preference for energy-dense food and an inactive lifestyle [191].

52

This impact on children’s eating increases the risk of obesity, and may also have a long-term effect [207, 208]. Davison et al. reported that children living with parents who had a high-energy, high-fat diet and were physically inactive (obesogenic) had a greater risk of a higher BMI, a higher percentage of body fat, a higher fat diet, and spending more hours watching TV than children from non-obesogenic families [208]. Moreover, children may develop a set of habits modelled by their parents’ behaviour and practice in relation to food preference, daily activities, and beliefs [209]. Paternal and maternal behaviour and habits are associated with children’s diet and physical lifestyle. Some mothers appear to simultaneously monitor their offspring’s screen time and dietary consumption. Fathers’ reinforcement has been associated with child physical activity as well [209]. Thus, parents are the enhancer of lifestyle intervention in preventing and treating obesity in children, through their influence on their children’s physical activity, eating behaviour, and screen time [210-212]. A trial for changes in the family-food environment reported that children experienced positive dietary change when their parents were practicing healthy food habits [213]. Although this study sample size was small and only focused on dietary fat intake, the positive family-food environment, high nutrition knowledge, and perceived responsibility reduced the child’s fat intake. In addition, Shloim et al. reviewed parenting style, feeding style and feeding practices, and weight status of children 4–12 years [214]. This study suggests that indulgent and uninvolved parenting and feeding styles were associated with a higher child BMI. The association between parental dietary intake and young adult obesity-related dietary behaviour was also reviewed [215]. In one review, parental dietary intake was associated with offsprings’ fruit and vegetable intake and energy and fat intake. In another study, parental feeding practices, and parental modelling of eating behaviour and amount of time spent watching TV were found to be likely to promote fatness in 5–6 year old children [216]. In a study by McCurdy et al., regardless of the family socio-economic background, parents who practiced healthy-food behaviour were found to have healthy-weight children [217]. All these findings may suggest that the importance of parental behaviour, perception and practices in relation to food intake and lifestyle partly influence offspring weight status in the short and long term.

2.5: Summary of the literature review

The accumulation of evidence from observational and other studies has shown that children who have parents who are overweight and obese in early life, have a higher chance of being overweight and obese in childhood, adolescence and adulthood. These studies have shown that there are two main links: 1) prenatal parental–offspring association and 2) postnatal parental–offspring 53

association. Irrespective of prenatal or postnatal parental exposure to obesity, studies have found a positive association between parents and their offspring. However, there is no clear distinction between these associations. There is no comparison as to whether prenatal or postnatal exposure to parental obesity has differential effects on offspring overweight and obesity. This comparison may provide us with new knowledge to distinguish whether the intrauterine environment has greater effect than the postnatal environment on the development of offspring overweight and obesity.

In addition, most studies that have examined the association between prenatal exposure to parental BMI and obesity and offspring BMI and obesity are based on birth-cohort studies. With a few exceptions, most studies have had a short follow-up time period. As a result, more studies have investigated the short-term link between parent and offspring BMI and less is known about the long-term association between prenatal-parental and offspring BMI and obesity. Additionally, studies examining offspring obesity in relation to parental obesity have been limited to BMI as the measure of obesity. Only a few studies have reported associations between prenatal parental BMI and obesity and offspring abdominal obesity and adiposity measures such as waist circumference, body fat percentage and skinfold thickness.

Some studies, which have examined parental BMI and obesity, have done so at only one time point (e.g. pre-pregnancy BMI); likewise, the prospective association with offspring BMI and obesity has been examined at only one time point. Much less is known about the association between parental BMI before birth and offspring BMI and obesity that persists through different stages of life, and whether the strength of this association differs at different critical phases of child development.

Few studies have compared paternal–offspring and maternal–offspring BMI, overweight and obesity. Although most studies have reported the maternal–offspring link was stronger than the paternal–offspring link, the evidence is still inconclusive. Studies which have examined the parental–offspring overweight and obesity association have considered some confounding factors in their multivariable analyses, including parental age, education, parity, maternal smoking during pregnancy, gestational age and birth weight. However, the majority of these studies did not have dietary information and details of the physical activities of parents or offspring. Therefore, it is unclear whether adjusting for diets and physical activity would attenuate the parental–offspring BMI and obesity association.

54

References [1] World Health Organization (WHO). (2015, 30 April). Obesity and overweight, Fact Sheet. Available: http://www.who.int/mediacentre/factsheets/fs311/en/ [2] M. Ezzati and E. Riboli, "Behavioral and Dietary Risk Factors for NoncommunicableDiseases," New England Journal of Medicine, vol. 369, pp. 954-964, 2013. [3] S. Low, M. C. Chin, and M. Deurenberg-Yap, "Review on Epidemic of Obesity," Ann Acad Med Singapore vol. 38, pp. 57-65, 2009. [4] K. M. Flegal, M. D. Carroll, B. K. Kit, and C. L. Ogden, "PRevalence of obesity and trends in the distribution of body mass index among us adults, 1999-2010," JAMA, vol. 307, pp. 491-497, 2012. [5] A. B. o. Statistic. (2013, 3/03/2016). Australian Health Survey. [6] WHO. (2013). Global database on body mass index. Available: http://www.who.int/bmi/index.jsp?introPage=intro_3.html. [7] C. Asia Pacific Cohort Studies, "The burden of overweight and obesity in the Asia–Pacific region," Obesity Reviews, vol. 8, pp. 191-196, 2007. [8] A. Z. Khambalia and L. S. Seen, "Trends in overweight and obese adults in Malaysia (1996– 2009): a systematic review," Obesity Reviews, vol. 11, pp. 403-412, 2010. [9] M. M. Yasin, A. M. Daher, N. M. Nasir, A. S. Ramli, M. Miskan, N. K. Keat, et al., "220 Overweight and Obesity in Malaysia: An Epidemiology Survey," Journal of , vol. 30, p. e67 10.1097/01.hjh.0000420044.04145.6c, 2012. [10] C. Roemling and M. Qaim, "Obesity trends and determinants in Indonesia," Appetite, vol. 58, pp. 1005-1013, 6// 2012. [11] L. K. Callaway, J. B. Prins, A. M. Chang, and H. D. McIntyre, "The prevalence and impact of overweight and obesity in an Australian obstetric population," Medical Journal of Australia, vol. 184, pp. 56-9, Jan. 2006. [12] P. M. Catalano and H. M. Ehrenberg, "Review article: The short- and long-term implications of maternal obesity on the mother and her offspring," BJOG: An International Journal of Obstetrics & Gynaecology, vol. 113, pp. 1126-1133, 2006. [13] D. Dabelea and T. Crume, "Maternal Environment and the Transgenerational Cycle of Obesity and Diabetes," Diabetes, vol. 60, pp. 1849-1855, Jul. 2011. [14] J. M. Crane, P. Murphy, L. Burrage, and D. Hutchens, "Maternal and perinatal outcomes of extreme obesity in pregnancy," Journal of obstetrics and gynaecology Canada : JOGC = Journal d'obstetrique et gynecologie du Canada : JOGC, vol. 35, pp. 606-611, 07/ 2013.

55

[15] J. R. O'Reilly and R. M. Reynolds, "The risk of maternal obesity to the long-term health of the offspring," Clinical Endocrinology, vol. 78, pp. 9-16, 2013. [16] N. Heslehurst, J. Rankin, J. R. Wilkinson, and C. D. Summerbell, "A nationally representative study of maternal obesity in England, UK: trends in incidence and demographic inequalities in 619 323 births, 1989-2007," International Journal of Obesity, vol. 34, pp. 420-428, Mar. 2010. [17] T. Y. Leung, T. N. Leung, D. S. Sahota, O. K. Chan, L. W. Chan, T. Y. Fung, et al., "Trends in maternal obesity and associated risks of adverse pregnancy outcomes in a population of Chinese women," BJOG, vol. 115, pp. 1529-37, Nov. 2008. [18] T. I. M. Olds, C. Maher, S. H. I. Zumin, S. PÉNeau, S. Lioret, K. Castetbon, et al., "Evidence that the prevalence of childhood overweight is plateauing: data from nine countries," International Journal of Pediatric Obesity, vol. 6, pp. 342-360, 2011. [19] C. L. Ogden, M. D. Carroll, B. K. Kit, and K. M. Flegal, "PRevalence of obesity and trends in body mass index among us children and adolescents, 1999-2010," JAMA, vol. 307, pp. 483-490, 2012. [20] T. S. Olds, G. R. Tomkinson, K. E. Ferrar, and C. A. Maher, "Trends in the prevalence of childhood overweight and obesity in Australia between 1985 and 2008," International Journal of Obesity, vol. 34, pp. 57-66, 10/13/online 2009. [21] M. de Onis, M. Blössner, and E. Borghi, "Global prevalence and trends of overweight and obesity among preschool children," The American Journal of Clinical Nutrition, vol. 92, pp. 1257- 1264, Nov. 2010. [22] B. M. Naidu, S. Z. Mahmud, R. Ambak, S. M. Sallehuddin, H. A. Mutalip, R. Saari, et al., "Overweight among primary school-age children in Malaysia," Asia Pacific Journal of Clinical Nutrition, vol. 22, p. 408, 2013. [23] A. A. Usfar, E. Lebenthal, Atmarita, E. Achadi, Soekirman, and H. Hadi, "Obesity as a poverty-related emerging nutrition problem: the case of Indonesia," Obesity Reviews, vol. 11, pp. 924-928, 2010. [24] D. Guh, W. Zhang, N. Bansback, Z. Amarsi, C. L. Birmingham, and A. Anis, "The incidence of co-morbidities related to obesity and overweight: a systematic review and meta- analysis," BMC Public Health, vol. 9, p. 88, 2009. [25] A. Abdullah, R. Wolfe, J. U. Stoelwinder, M. de Courten, C. Stevenson, H. L. Walls, et al., "The number of years lived with obesity and the risk of all-cause and cause-specific mortality," International Journal of Epidemiology, Feb. 2011.

56

[26] F. M. Biro and M. Wien, "Childhood obesity and adult morbidities," The American Journal of Clinical Nutrition, vol. 91, pp. 1499S-1505S, May 2010. [27] D. Withrow and D. A. Alter, "The economic burden of obesity worldwide: a systematic review of the direct costs of obesity," Obesity Reviews, vol. 12, pp. 131-141, 2011. [28] Y. C. Wang, K. McPherson, T. Marsh, S. L. Gortmaker, and M. Brown, "Health and economic burden of the projected obesity trends in the USA and the UK," The Lancet, vol. 378, pp. 815-825, //27 2011. [29] S. Colagiuri, C. Lee, R. Colagiuri, D. Magliano, J. E. Shaw, P. Z. Zimmet, et al., "The cost of overweight and obesity in Australia," Medical Journal of Australia, vol. 192, pp. 260-4, 2010. [30] Access Economics for Diabetes Australia, "The growing cost of obesity in 2008: three years on," 2008. Canberra: Diabetes Australia. [31] C. M. Doak, T. L. S. Visscher, C. M. Renders, and J. C. Seidell, "The prevention of overweight and obesity in children and adolescents: a review of interventions and programmes," Obesity Reviews, vol. 7, pp. 111-136, 2006. [32] K. D. Hesketh and K. J. Campbell, "Interventions to Prevent Obesity in 0–5 Year Olds: An Updated Systematic Review of the Literature," Obesity, vol. 18, pp. S27-S35, 2010. [33] S. N. Bleich, J. Segal, Y. Wu, R. Wilson, and Y. Wang, "Systematic Review of Community- Based Childhood Obesity Prevention Studies," Pediatrics, Jun. 2013. [34] T. S. Han, C. L. Hart, C. Haig, J. Logue, M. N. Upton, G. C. M. Watt, et al., "Contributions of maternal and paternal adiposity and smoking to adult offspring adiposity and cardiovascular risk: the Midspan Family Study," British Medical Journal Open, vol. 5, Nov. 2015. [35] M. Kanciruk, J. J. Andrews, and T. Donnon, "Family history of obesity and risk of childhood overweight and obesity: A meta-analysis," International Journal of Medical Health Pharmacy and Biomedical Engineering, vol. 8, pp. 244-56, 2014. [36] T. Parsons, C. Power, S. Logan, and C. Summerbell, "Childhood predictors of adult obesity: a systematic review," International Journal of Obesity Related Metabolic Disorders, vol. 23, pp. S1 - S107, 1999. [37] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357- 1359, Jun. 2005. [38] K. Bammann, J. Peplies, S. De Henauw, M. Hunsberger, D. Molnar, L. A. Moreno, et al., "Early Life Course Risk Factors for Childhood Obesity: The IDEFICS Case-Control Study," PLoS ONE, vol. 9, p. e86914, 2014.

57

[39] C. S. Barroso, Roncancio, A., Hinojosa, M.B., Reifsnider, E., "The association between early childhood overweight and maternal factors," Childhood Obesity, vol. 8, 2012. [40] I. Bralic, J. Vrdoljak, and V. Kovacic, "Associations between parental and child overweight and obesity," Collegium Antropologicum, vol. 29, pp. 481-486, Dec. 2005. [41] V. Burke, Beilin, L.J., Dunbar, D., "Family lifestyle and parental body mass index as predictors of body mass index in Australian children: A longitudinal study," International Journal of Obesity, vol. 25, 2001. [42] L. Y. Gibson, Byrne, S.M., Davis, E.A., Blair, E., Jacoby, P., Zubrick, S.R., "The role of family and maternal factors in childhood obesity," Medical Journal of Australia, vol. 186, 2007. [43] X. F. Guo, L. Q. Zheng, Y. Li, X. Y. Zhang, S. S. Yu, H. M. Yang, et al., "Prevalence and risk factors of being overweight or obese among children and adolescents in northeast China," Pediatric Research, vol. 74, pp. 443-449, Oct. 2013. [44] M. A. Hernandez-Valero, A. V. Wilkinson, M. R. Forman, C. J. Etzel, Y. M. Cao, C. H. Barcenas, et al., "Maternal BMI and country of birth as indicators of childhood obesity in children of Mexican origin," Obesity, vol. 15, pp. 2512-2519, Oct. 2007. [45] K. McCarthy, Y. L. Ye, S. Yuan, and Q. Q. He, "Parental Weight Status and Offspring Cardiovascular Disease Risks: a Cross-Sectional Study of Chinese Children," Preventing Chronic Disease, vol. 12, Jan. 2015. [46] L. N. Nobre, K. C. Silva, S. E. D. Ferreira, L. L. Moreira, A. D. Lessa, J. A. Lamounier, et al., "Early determinants of overweight and obesity at 5 years old in preschoolers from inner of Minas Gerais, Brazil," Nutricion Hospitalaria, vol. 28, pp. 764-771, May-Jun. 2013. [47] A. M. Oliveira, A. C. Oliveira, M. S. Almeida, N. Oliveira, and L. Adan, "Influence of the family nucleus on obesity in children from northeastern Brazil: a cross-sectional study," BMC Public Health, vol. 7, Sep. 2007. [48] R. Patel, R. M. Martin, M. S. Kramer, E. Oken, N. Bogdanovich, L. Matush, et al., "Familial associations of adiposity: findings from a cross-sectional study of 12,181 parental-offspring trios from Belarus," Plos One, vol. 6, Jan. 2011. [49] S. Santiago, I. Zazpe, M. Cuervo, and J. A. Martinez, "Perinatal and parental determinants of childhood overweight in 6-12 years old children," Nutricion Hospitalaria, vol. 27, pp. 599-605, Mar-Apr. 2012. [50] J. L. Santos, J. Kain, P. Dominguez-Vásquez, L. Lera, M. Galván, C. Corvalán, et al., "Maternal anthropometry and feeding behavior toward preschool children: association with

58

childhood body mass index in an observational study of Chilean families," International Journal of Behavioral Nutrition and Physical Activity, vol. 6, 2009. [51] M. Sekine, Yamagami, T., Hamanishi, S., Handa, K., Saito, T., Nanri, S., Kawaminami, K., Tokui, N., Yoshida, K., Kagamimori, S., "Parental obesity, lifestyle factors and obesity in preschool children: results of the Toyama Birth Cohort Study," Journal of Epidemiology, vol. 12, 2002. [52] R. C. Whitaker, C. M. Deeks, A. E. Baughcum, and B. L. Specker, "The relationship of childhood adiposity to parent body mass index and eating behavior," Obesity research, vol. 8, pp. 234-240, 2000. [53] L. Xu, L. Dubois, D. Burnier, M. Girard, and D. Prud'homme, "Parental overweight/obesity, social factors, and child overweight/obesity at 7 years of age," Pediatrics International, vol. 53, pp. 826-31, Dec. 2011. [54] S. Plachta-Danielzik, B. Landsberg, M. Johannsen, D. Lange, and M. J. Muller, "Determinants of the prevalence and incidence of overweight in children and adolescents," Public Health Nutrition, vol. 13, pp. 1870-1881, Nov. 2010. [55] L. M. Steffen, Dai, S., Fulton, J.E., Labarthe, D.R., "Overweight in children and adolescents associated with TV viewing and parental weight. Project HeartBeat!," American Journal of Preventive Medicine, vol. 37, 2009. [56] K. L. Whitaker, M. J. Jarvis, R. J. Beeken, D. Boniface, and J. Wardle, "Comparing maternal and paternal intergenerational transmission of obesity risk in a large population-based sample," American Journal of Clinical Nutrition, vol. 91, pp. 1560 - 1567, 2010. [57] World Health Organization (WHO), "Obesity: preventing and managing a global epidemic, Report of a WHO Consultant on Obesity. Geneva: World Health Organization.," 1998. [58] World Health Organization (WHO). (2008, 6 May). World Health Organization Child Growth Standards: Training course on child growth assessment - interpreting growth indicators. Available: http://www.who.int/childgrowth/training/en/ [59] P. C. D. Johnson, Logue, J., McConnachie, A., Abu-Rmeileh, N.M.E., Hart, C., Upton, M.N., Lean, M., Sattar, N., Watt, G., "Intergenerational change and familial aggregation of body mass index," European Journal of Epidemiology, vol. 27, pp. 53-61, 2012. [60] T. Cole, M. Bellizzi, K. Flegal, and W. Dietz, "Establishing a standard definition for child overweight and obesity worldwide: international survey," British Medical Journal, vol. 320, pp. 1240-1243, 2000. [61] R. J. Kuczmarski, C. L. Ogden, L. M. Grummer-Strawn, K. M. Flegal, S. S. Guo, R. Wei, et al., "CDC growth charts: United States," Adv Data, pp. 1-27, Jun. 8 2000.

59

[62] L. Ay, A. C. S. Hokken-Koelega, D. O. Mook-Kanamori, A. Hofman, H. A. Moll, J. P. Mackenbach, et al., "Tracking and determinants of subcutaneous fat mass in early childhood: The Generation R Study," International Journal of Obesity, vol. 32, pp. 1050-1059, 2008. [63] N. J. Blair, J. M. D. Thompson, P. N. Black, D. M. O. Becroft, P. M. Clark, D. Y. Han, et al., "Risk factors for obesity in 7-year-old European children: the Auckland Birthweight Collaborative Study," Archives of Disease in Childhood, vol. 92, pp. 866-871, Oct 2007. [64] H. Castillo, I. S. Santos, and A. Matijasevich, "Relationship between maternal pre- pregnancy body mass index, gestational weight gain and childhood fatness at 6-7 years by air displacement plethysmography," Maternal and Child Nutrition, vol. 11(4), pp. 606-17, 2015. [65] P. M. Catalano, Farrell, K., Thomas, A., Huston-Presley, L., Mencin, P., De Mouzon, S.H., Amini, S.B., "Perinatal risk factors for childhood obesity and metabolic dysregulation," American Journal of Clinical Nutrition, vol. 90, 2009. [66] P. Chivers, H. Parker, M. Bulsara, L. Beilin, and B. Hands, "Parental and early childhood influences on adolescent obesity: a longitudinal study," Early Child Development and Care, vol. 182, pp. 1071-1087, 2013/04/21 2012. [67] G. Davey Smith, C. Steer, S. Leary, and A. Ness, "Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC)," Archives of Disease in Childhood, vol. 92, pp. 876-880, 2007. [68] M. L. A. de Hoog, M. van Eijsden, K. Stronks, R. Gemke, and T. G. M. Vrijkotte, "Overweight at age two years in a multi-ethnic cohort (ABCD study): the role of prenatal factors, birth outcomes and postnatal factors," Bmc Public Health, vol. 11, Aug. 2011. [69] A. L. Deierlein, A. M. Siega-Riz, L. S. Adair, and A. H. Herring, "Effects of pre-pregnancy body mass index and gestational weight gain on infant anthropometric outcomes," Journal of Pediatrics, vol. 158, pp. 221-226, 2011. [70] B. Durmus, L. R. Arends, L. Ay, A. C. Hokken-Koelega, H. Raat, A. Hofman, et al., "Parental anthropometrics, early growth and the risk of overweight in pre-school children: the Generation R Study," Paediatric Obesity, vol. 8, pp. 339-50, Oct 2013. [71] L. Fairley, G. Santorelli, D. A. Lawlor, M. Bryant, R. Bhopal, E. S. Petherick, et al., "The relationship between early life modifiable risk factors for childhood obesity, ethnicity and body mass index at age 3 years: findings from the Born in Bradford birth cohort study," BMC Obesity, vol. 2, p. 9, 2015. [72] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child Cohort

60

Study: a family-based approach to studying the role of the intrauterine environment in childhood adiposity," American Journal of Epidemiology, vol. 176, pp. 83-92 July 15, 2012. [73] M. G. J. Gademan, M. Vermeulen, A. Oostvogels, T. J. Roseboom, T. L. S. Visscher, M. van Eijsden, et al., "Maternal prepregancy BMI and lipid profile during early pregnancy are independently associated with offspring's body composition at age 5-6 years: the ABCD Study," Plos One, vol. 9, Apr. 2014. [74] R. Gaillard, M. Welten, W. H. Oddy, L. J. Beilin, T. A. Mori, V. W. V. Jaddoe, et al., "Associations of maternal prepregnancy body mass index and gestational weight gain with cardio- metabolic risk factors in adolescent offspring: a prospective cohort study," BJOG: An International Journal of Obstetrics & Gynaecology, vol. 123, pp. 207-216, 2016. [75] C. R. Gale, M. K. Javaid, S. M. Robinson, C. M. Law, K. M. Godfrey, and C. Cooper, "Maternal size in pregnancy and body composition in children," Journal of Clinical Endocrinology & Metabolism, vol. 92, pp. 3904-3911, Oct. 2007. [76] E. R. Hamilton, J. O. Teitler, and N. E. Reichman, "Mexican American Birthweight and Child Overweight: Unraveling a Possible Early Life Course Health Transition," Journal of Health and Social Behavior, vol. 52, pp. 333-348, Sep. 2011. [77] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: a 16- year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct. 2011. [78] N. Z. Janjua, Mahmood, B., Islam, M.A., Goldenberg, R.L., "Maternal and early childhood risk factors for overweight and obesity among low-income predominantly black children at age five years: A prospective cohort study," Journal of Obesity, vol. ?, 2012. vol. no. required [79] P. Kitsantas, L. R. Pawloski, and K. F. Gaffney, "Maternal prepregnancy body mass index in relation to Hispanic preschooler overweight/obesity," European Journal of Pediatrics, vol. 169, pp. 1361-1368, Nov. 2010. [80] I. Koupil and P. Toivanen, "Social and early-life determinants of overweight and obesity in 18-year-old Swedish men," International Journal of Obesity, vol. 32, pp. 73-81, Jan. 2008. [81] J. Laitinen, Power, C., Järvelin, M.-R., "Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity," American Journal of Clinical Nutrition, vol. 74, 2001. [82] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater-university

61

study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb. 2007. [83] S. Leary, G. Davey Smith, and A. Ness, "No evidence of large differences in mother- daughter and father-son body mass index concordance in a large UK birth cohort," International Journal of Obesity, vol. 34, pp. 1191-1192, 2010. [84] X. Lin, I. M. Aris, M. T. Tint, S. E. Soh, K. M. Godfrey, G. S. Yeo, et al., "Ethnic differences in effects of maternal pre-pregnancy and pregnancy adiposity on offspring size and adiposity," Journal of Clinical Endocrinology Metabolism, p. jc20151728, Jul. 2015. [85] I. Mesman, T. J. Roseboom, G. J. Bonsel, R. J. Gemke, M. F. van der Wal, and T. G. M. Vrijkotte, "Maternal pre-pregnancy body mass index explains infant's weight and BMI at 14 months: results from a multi-ethnic birth cohort study," Archives of Disease in Childhood, vol. 94, pp. 587-595, Aug. 2009. [86] M. J. O'Callaghan, G. M. Williams, M. J. Andersen, W. Bor, and J. M. Najman, "Prediction of obesity in children at 5 years: a cohort study," Journal of Paediatric Child Health, vol. 33, pp. 311-6, Aug. 1997. [87] C. M. Olson, M. M. Demment, S. J. Carling, and M. S. Strawderman, "Associations between mothers' and their children's weights at 4 years of age," Childhood Obesity, vol. 6, pp. 201- 207, Aug. 2010. [88] R. M. Reynolds, C. Osmond, D. I. W. Phillips, and K. M. Godfrey, "Maternal BMI, Parity, and Pregnancy Weight Gain: Influences on Offspring Adiposity in Young Adulthood," Journal of Clinical Endocrinology & Metabolism, vol. 95, pp. 5365-5369, Dec. 2010. [89] I. Rios-Castillo, S. Cerezo, C. Corvalán, M. Martínez, and J. Kain, "Risk factors during the prenatal period and the first year of life associated with overweight in 7-year-old low-income Chilean children," Maternal and Child Nutrition, 2012. [90] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166- 1175, 2011. [91] P. J. Salsberry, Reagan, P.B., "Taking the long view: The prenatal environment and early adolescent overweight," Research in Nursing and Health, vol. 30, 2007. [92] N. Stettler, A. M. Tershakovec, B. S. Zemel, M. B. Leonard, R. C. Boston, S. H. Katz, et al., "Early risk factors for increased adiposity: a cohort study of African American subjects followed from birth to young adulthood," American Journal of Clinical Nutrition, vol. 72, pp. 378-383, Aug. 2000.

62

[93] A. M. F. M. R. M. K. B. Stuebe, "Maternal-recalled gestational weight gain, pre-pregnancy body mass index, and obesity in the daughter," International Journal of Obesity, 2009. [94] A. L. L. Tequeanes, D. P. Gigante, M. C. F. Assuncao, D. A. G. Chica, and B. L. Horta, "Maternal anthropometry is associated with the body mass index and waist:height ratio of offspring at 23 years of age," Journal of Nutrition, vol. 139, pp. 750-754, Apr. 2009. [95] M. B. Terry, Wei, Y., Esserman, D., McKeague, I.W., Susser, E., "Pre-and postnatal determinants of childhood body size: cohort and sibling analyses," Journal of Developmental Origins of Health and Disease, vol. 2, pp. 99-111, 2011. [96] M. B. Terry, Y. Wei, and D. Esserman, "Maternal, birth, and early-life influences on adult body size in women," American Journal of Epidemiology, vol. 166, pp. 5-13, Jul. 2007. [97] S. R. Veena, G. V. Krishnaveni, S. C. Karat, C. Osmond, and C. H. Fall, "Testing the fetal overnutrition hypothesis; the relationship of maternal and paternal adiposity to adiposity, insulin resistance and cardiovascular risk factors in Indian children," Public Health Nutrition, vol. 16, pp. 1656-66, Sep. 2013. [98] L. M. Wen, L. A. Baur, C. Rissel, H. L. Xu, and J. M. Simpson, "Correlates of body mass index and overweight and obesity of children aged 2 years: findings from the Healthy Beginnings Trial," Obesity, vol. 22, pp. 1723-1730, Jul. 2014. [99] R. C. Whitaker, "Predicting preschooler obesity at birth: the role of maternal obesity in early pregnancy," Pediatrics, vol. 114, pp. e29-36, Jul. 2004. [100] J. Laitinen, Jääskeläinen, A., Hartikainen, A.-L., Sovio, U., Vääräsmäki, M., Pouta, A., Kaakinen, M., Järvelin, M.-R., "Maternal weight gain during the first half of pregnancy and offspring obesity at 16 years: a prospective cohort study," BJOG: An International Journal of Obstetrics and Gynaecology, vol. 119, 2012. [101] U. M. Stamnes Køpp, Dahl-Jørgensen, K., Stigum, H., Frost Andersen, L., Næss, O., Nystad, W., "The associations between maternal pre-pregnancy body mass index or gestational weight change during pregnancy and body mass index of the child at 3 years of age," International journal of obesity (2005), vol. 36, 2012. [102] H. Hochner, Friedlander, Y., Calderon-Margalit, R., Meiner, V., Sagy, Y., Avgil-Tsadok, M., Burger, A., Savitsky, B., Siscovick, D.S., Manor, O., "Associations of maternal prepregnancy body mass index and gestational weight gain with adult offspring cardiometabolic risk factors: the Jerusalem perinatal family follow-up study," Circulation, vol. 125, pp. 1381-9, 2012.

63

[103] J. Laitinen, C. Power, and M. Järvelin, "Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity," American Journal of Clinical Nutrition, vol. 74, pp. 287-294, 2001. [104] J. J. Reilly, E. Methven, Z. C. McDowell, B. Hacking, D. Alexander, L. Stewart, et al., "Health consequences of obesity," Archives of Disease in Childhood, vol. 88, pp. 748-752, Sep. 2003. [105] A. M. Stuebe, H. Lyon, A. H. Herring, J. Ghosh, A. Wise, K. E. North, et al., "Obesity and diabetes genetic variants associated with gestational weight gain," American Journal of Obstetrics and Gynecology, vol. 203, Sep. 2010. [106] O. Ajala, A. E. Fr Meaux, J. Hosking, B. S. Metcalf, A. N. Jeffery, L. D. Voss, et al., "The relationship of height and body fat to gender-assortative weight gain in children. A longitudinal cohort study (EarlyBird 44)," International Journal of Pediatric Obesity, vol. 6, pp. 223-8, Aug. 2011. [107] J. L. Chen, D. Wall, C. Kennedy, V. Unnithan, and C. H. Yeh, "Predictors of increased body mass index in Chinese children," Progress in Cardiovascular Nursing, vol. 22, pp. 138-144, 2007. [108] R. Cooper, E. Hyppönen, D. Berry, and C. Power, "Associations between parental and offspring adiposity up to midlife: the contribution of adult lifestyle factors in the 1958 British Birth Cohort Study," American Journal of Clinical Nutrition, vol. 92, pp. 946-953, 2010. [109] A. Crossman, Anne Sullivan, D., Benin, M., "The family environment and American adolescents' risk of obesity as young adults," Social Science and Medicine, vol. 63, 2006. [110] V. De Coen, I. De Bourdeaudhuij, V. Verbestel, L. Maes, and C. Vereecken, "Risk factors for childhood overweight: a 30-month longitudinal study of 3- to 6-year-old children," Public Health Nutrition, vol. 17, pp. 1993-2000, Sep. 2014. [111] L. Dubois and M. Girard, "Early determinants of overweight at 4.5 years in a population- based longitudinal study," International Journal of Obesity, vol. 30, pp. 610-617, Apr. 2006. [112] E. Freeman, R. Fletcher, C. E. Collins, P. J. Morgan, T. Burrows, and R. Callister, "Preventing and treating childhood obesity: time to target fathers," International Journal of Obesity, vol. 36, pp. 12-15, Jan. 2012. [113] B. Heude, A. Kettaneh, R. Rakotovao, J. L. Bresson, J. M. Borys, P. Ducimetiere, et al., "Anthropometric relationships between parents and children throughout childhood: the Fleurbaix- Laventie Ville Sante Study," International Journal of Obesity (London), vol. 29, pp. 1222-9, Oct. 2005.

64

[114] M. Juonala, J. Juhola, C. G. Magnussen, P. Wurtz, J. S. A. Viikari, R. Thomson, et al., "Childhood environmental and genetic predictors of adulthood obesity: The Cardiovascular Risk in Young Finns Study," Journal of Clinical Endocrinology & Metabolism, vol. 96, pp. E1542-E1549, Sep 2011. [115] M. Kivimäki, D. A. Lawlor, G. D. Smith, M. Elovainio, M. Jokela, L. Keltikangas-Järvinen, et al., "Substantial intergenerational increases in body mass index are not explained by the fetal overnutrition hypothesis: The Cardiovascular Risk in Young Finns Study," American Journal of Clinical Nutrition, vol. 86, pp. 1509-1514, 2007. [116] E. Kvaavik, G. S. Tell, and K. I. Klepp, "Predictors and tracking of body mass index from adolescence into adulthood - follow-up of 18 to 20 years in the Oslo youth study," Archives of Pediatrics & Adolescent Medicine, vol. 157, pp. 1212-1218, Dec. 2003. [117] J. K. Lake, C. Power, and T. J. Cole, "Child to adult body mass index in the 1958 British birth cohort: associations with parental obesity," Archives of Disease in Childhood, vol. 77, pp. 376- 381, Nov. 1997. [118] Y. Li, K. Brock, R. Cant, L. Ke, and S. Morrell, "Parental obesity as a predictor of childhood overweight/obesity in Australian migrant children," Obesity Research & Clinical Practice, vol. 2, pp. 179-187, Sep. 2008. [119] A. M. Magarey, L. A. Daniels, T. J. Boulton, and R. A. Cockington, "Predicting obesity in early adulthood from childhood and parental obesity," International Journal of Obesity, vol. 27, pp. 505-513, 2003. [120] M. Mostazir, Jeffery, A., Voss, L., Wilkin, T.J., "Gender-assortative waist circumference in mother-daughter and father-son pairs, and its implications. an 11-year longitudinal study in children (EarlyBird 59)," Pediatric Obesity, vol. 9, 2014. [121] N. Olvera, Sharma, S., Suminski, R., Rodríguez, A., Power, T., "BMI tracking in Mexican American children in relation to maternal BMI," Ethnicity and Disease, vol. 17, 2007. [122] A. D. Salbe, Weyer, C., Lindsay, R.S., Ravussin, E., Antonio Tataranni, P., "Assessing risk factors for obesity between childhood and adolescence: I. Birth weight, childhood adiposity, parental obesity, insulin, and leptin," Pediatrics, vol. 110, 2002. [123] W. Stewart Agras, Hammer, L.D., McNicholas, F., Kraemer, H.C., "Risk factors for childhood overweight: a prospective study from birth to 9.5 years," Journal of Pediatrics, vol. 145, 2004.

65

[124] M. Suka, Sugimori, H., Yoshida, K., Sekine, M., Yamagami, T., Kagamimori, S., "Parental influence on the development of obesity in 9-year-old Japanese children: The Toyama Birth Cohort Study," Environmental Health and Preventive Medicine, vol. 7, 2002. [125] F. Trudeau, R. J. Shephard, S. Bouchard, and L. Laurencelle, "BMI in the Trois-Rivières study: child-adult and child-parent relationships," American Journal of Human Biology, vol. 15, pp. 187-191, 2003. [126] R. C. Whitaker, Wright, J.A., Pepe, M.S., Seidel, K.D., Dietz, W.H., "Predicting obesity in young adulthood from childhood and parental obesity," New England Journal of Medicine, vol. 337, pp. 869 - 873, 1997. [127] S. Williams, "Overweight at age 21: the association with body mass index in childhood and adolescence and parents' body mass index. A cohort study of New Zealanders born in 1972-1973," International Journal of Obesity and Related Metabolic Disorders, vol. 25, pp. 158-63, Feb. 2001. [128] M. B. Mostazir, B. S. Metcalf, J. Hosking, A. J. Streeter, A. N. Jeffery, L. D. Voss, et al., "Gender-assortative waist gain and its metabolic impact in mother-daughter and father-son pairs: A 10-year longitudinal study of trios," Diabetes, vol. 61, p. A320, 2012. [129] A. Andres, Hull, H.R., Shankar, K., Casey, P.H., Cleves, M.A., Badger, T.M., "Longitudinal body composition of children born to mothers with normal weight, overweight, and obesity," Obesity, vol. 23, 2015. [130] L. C. Giles, M. J. Whitrow, M. J. Davies, C. E. Davies, A. R. Rumbold, and V. M. Moore, "Growth trajectories in early childhood, their relationship with antenatal and postnatal factors, and development of obesity by age 9 years: results from an Australian birth cohort study," International Journal of Obesity, vol. 39, pp. 1049-1056, Jul. 2015. [131] C. Haga, Kondo, N., Suzuki, K., Sato, M., Ando, D., Yokomichi, H., Tanaka, T., Yamagata, Z., "Developmental trajectories of body mass index among Japanese children and impact of maternal factors during pregnancy," PLoS ONE, vol. 7, 2012. [132] C. Y. Li, M. I. Goran, H. Kaur, N. Nollen, and J. S. Ahluwalia, "Developmental trajectories of overweight during childhood: role of early life factors," Obesity, vol. 15, pp. 760-771, Mar. 2007. [133] L. E. Pryor, M. Brendgen, R. E. Tremblay, J. B. Pingault, X. C. Liu, L. Dubois, et al., "Early risk factors of overweight developmental trajectories during middle childhood," Plos One, vol. 10, Jun. 2015. [134] N. Fuiano, A. Rapa, A. Monzani, A. Pietrobelli, G. Diddi, A. Limosani, et al., "Prevalence and risk factors for overweight and obesity in a population of Italian schoolchildren: a longitudinal study," Journal of Endocrinological Investigation, vol. 31, pp. 979-984, Nov. 2008.

66

[135] A. L. Kristiansen, M. Bjelland, A. L. Brantsaeter, M. Haugen, H. M. Meltzer, W. Nystad, et al., "Tracking of body size from birth to 7 years of age and factors associated with maintenance of a high body size from birth to 7 years of age--the Norwegian Mother and Child Cohort study (MoBa)," Public Health Nutrition, vol. 18, pp. 1746-55, Jul. 2015. [136] L. van Rossem, A. H. Wijga, B. Brunekreef, J. C. de Jongste, M. Kerkhof, D. S. Postma, et al., "Overweight in infancy: which pre- and perinatal factors determine overweight persistence or reduction? A birth cohort followed for 11 years," Annals of Nutrition and Metabolism, vol. 65, pp. 211-219, 2014. [137] A. H. Ziyab, W. Karmaus, R. J. Kurukulaaratchy, H. M. Zhang, and S. H. Arshad, "Developmental trajectories of Body Mass Index from infancy to 18 years of age: prenatal determinants and health consequences," Journal of Epidemiology and Community Health, vol. 68, pp. 934-941, Oct. 2014. [138] A. M. Linabery, R. W. Nahhas, W. Johnson, A. C. Choh, B. Towne, A. O. Odegaard, et al., "Stronger influence of maternal than paternal obesity on infant and early childhood body mass index: the Fels Longitudinal Study," Pediatric Obesity, vol. 8, pp. 159-169, Jun. 2013. [139] L. E. Pryor, Tremblay, R.E., Boivin, M., Touchette, E., Dubois, L., Genolini, C., Liu, X., Falissard, B., Côté, S.M., "Developmental trajectories of body mass index in early childhood and their risk factors: an 8-year longitudinal study," Archives of Pediatrics and Adolescent Medicine, vol. 165, 2011. [140] K. Hesketh, J. Carlin, M. Wake, and D. Crawford, "Predictors of body mass index change in Australian primary school children," International Journal of Pediatric Obesity, vol. 4, pp. 45-53, 2009. [141] G. M. Lawrence, Shulman, S., Friedlander, Y., Sitlani, C.M., Burger, A., Savitsky, B., Granot-Hershkovitz, E., Lumley, T., Kwok, P.-Y., Hesselson, S., Enquobahrie, D., Wander, P.L., Manor, O., Siscovick, D.S., Hochner, H., "Associations of maternal pre-pregnancy and gestational body size with offspring longitudinal change in BMI," Obesity, vol. 22, 2014. [142] A. A. Mamun, D. A. Lawlor, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Family and early life factors associated with changes in overweight status between ages 5 and 14 years: findings from the Mater University Study of Pregnancy and its outcomes," International Journal of Obesity and Related Metabolic Disorders, vol. 29, pp. 475-482, 2005. [143] K. J. Greenlund, Liu, K., Dyer, A.R., Kiefe, C.I., Burke, G.L., Yunis, C., "Body mass index in young adults: associations with parental body size and education in the CARDIA study," American Journal of Public Health, vol. 86, 1996.

67

[144] E. M. Perez-Pastor, B. S. Metcalf, J. Hosking, A. N. Jeffery, L. D. Voss, and T. J. Wilkin, "Assortative weight gain in mother-daughter and father-son pairs: an emerging source of childhood obesity. Longitudinal study of trios (EarlyBird 43)," International Journal of Obesity, vol. 33, pp. 727-735, Jul. 2009. [145] L. Mo-Suwan, Tongkumchum, P., Puetpaiboon, A., "Determinants of overweight tracking from childhood to adolescence: a 5 y follow-up study of Hat Yai schoolchildren," International Journal of Obesity, vol. 24, 2000. [146] B. H. Lourenço, E. Villamor, R. A. Augusto, and M. A. Cardoso, "Influence of early life factors on body mass index trajectory during childhood: a population-based longitudinal analysis in the Western Brazilian Amazon," Maternal and Child Nutrition, vol. 11, pp. 240-252, 2015. [147] M. B. Terry, Y. Wei, D. Esserman, I. W. McKeague, and E. Susser, "Pre- and postnatal determinants of childhood body size: cohort and sibling analyses," Journal of Developmental Origins of Health and Disease, vol. 2, pp. 99-111, Apr. 2011. [148] T. D. Brisbois, A. P. Farmer, and L. J. McCargar, "Early markers of adult obesity: a review," Obesity Reviews, vol. 13, pp. 347-367, Apr. 2012. [149] L. Monasta, G. D. Batty, A. Cattaneo, V. Lutje, L. Ronfani, F. J. Van Lenthe, et al., "Early- life determinants of overweight and obesity: a review of systematic reviews," Obesity Reviews, vol. 11, pp. 695-708, 2010. [150] I. O. M. (IOM). Institute of Medicine. Weight Gain During Pregnancy: Reexamining the Guidelines. : Washington, DC: The National Academies Press, 2009. [151] A. Beyerlein, I. Nehring, P. Rzehak, J. Heinrich, M. J. Muller, S. Plachta-Danielzik, et al., "Gestational weight gain and body mass index in children: results from three german cohort studies," Plos One, vol. 7, p. e33205, 2012. [152] Oken, K. P. Kleinman, M. B. Belfort, J. K. Hammitt, and M. W. Gillman, "Associations of gestational weight gain with short- and longer-term maternal and child health outcomes," American Journal of Epidemiology, vol. 170, pp. 173-180, Jul. 2009. [153] J. L. Josefson, J. A. Hoffmann, and B. E. Metzger, "Excessive weight gain in women with a normal pre-pregnancy BMI is associated with increased neonatal adiposity," Pediatric Obesity, vol. 8, pp. e33-e36, 2013. [154] J. L. Kaar, T. Crume, J. T. Brinton, K. J. Bischoff, R. McDuffie, and D. Dabelea, "Maternal obesity, gestational weight gain, and offspring adiposity: the Exploring Perinatal Outcomes among Children Study," Journal of Pediatrics, vol. 165, pp. 509-515, Sep. 2014.

68

[155] I. Nehring, S. Lehmann, and R. von Kries, "Gestational weight gain in accordance to the IOM/NRC criteria and the risk for childhood overweight: a meta-analysis," Pediatric Obesity, vol. 8, pp. 218-224, Jun. 2013. [156] A. A. Mamun, M. Mannan, and S. A. R. Doi, "Gestational weight gain in relation to offspring obesity over the life course: a systematic review and bias-adjusted meta-analysis," Obesity Reviews, vol. 15, pp. 338-347, 2014. [157] R. Gaillard, B. Durmuş, A. Hofman, J. P. Mackenbach, E. A. P. Steegers, and V. W. V. Jaddoe, "Risk factors and outcomes of maternal obesity and excessive weight gain during pregnancy," Obesity, vol. 21, pp. 1046-1055, 2013. [158] K. Silventoinen, B. Rokholm, Kaprio, J., and T. I. A. Sorensen, "The genetic and environmental influences on childhood obesity: a systematic review of twin and adoption studies," International Journal of Obesity, vol. 34, pp. 29-40, 2009. [159] U. M. Stamnes Koepp, L. Frost Andersen, K. Dahl-Joergensen, H. Stigum, O. Nass, and W. Nystad, "Maternal pre-pregnant body mass index, maternal weight change and offspring birthweight," Acta Obstetricia et Gynecologica Scandinavica, vol. 91, pp. 243-9, Feb. 2012. [160] H. V. Mark, O. K. Stefan, and H. B. Bernhard, "Is Later Obesity Programmed In Utero?," Current Drug Targets, vol. 8, pp. 923-934, 2007. [161] N. Li, E. Liu, S. Sun, J. Guo, L. Pan, P. Wang, et al., "Birth weight and overweight or obesity risk in children under 3 years in China," American Journal of Human Biology, vol. 26, pp. 331-336, 2014. [162] Z. B. Yu, S. P. Han, G. Z. Zhu, C. Zhu, X. J. Wang, X. G. Cao, et al., "Birth weight and subsequent risk of obesity: a systematic review and meta-analysis," Obesity Reviews, vol. 12, pp. 525-542, 2011. [163] K. Schellong, S. Schulz, T. Harder, and A. Plagemann, "Birth Weight and Long-Term Overweight Risk: Systematic Review and a Meta-Analysis Including 643,902 Persons from 66 Studies and 26 Countries Globally," PLoS ONE, vol. 7, p. e47776, 2012. [164] J. Gravel, B. Potter, and L. Dubois, "Prenatal exposure to maternal cigarette smoke and offspring risk of excess weight is independent of both birth weight and catch-up growth," ISRN Epidemiology, vol. 2013, p. 8, 2013. [165] E. Oken, S. Y. Huh, E. M. Taveras, J. W. Rich-Edwards, and M. W. Gillman, "Associations of maternal prenatal smoking with child adiposity and blood pressure," Obesity Research, vol. 13, pp. 2021-8, Nov. 2005.

69

[166] L. Wang, H. M. Mamudu, and T. Wu, "The impact of maternal prenatal smoking on the development of childhood overweight in school-aged children," Pediatric Obesity, vol. 8, pp. 178- 188, 2013. [167] L. E. Grzeskowiak, N. A. Hodyl, M. J. Stark, J. L. Morrison, and V. L. Clifton, "Association of early and late maternal smoking during pregnancy with offspring body mass index at 4 to 5 years of age," Journal of Developmental Origins of Health and Disease, vol. 6, pp. 485-492, 2015. [168] S. F. Weng, S. A. Redsell, J. A. Swift, M. Yang, and C. P. Glazebrook, "Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy," Archives of Diseases in Childhood, vol. 97, pp. 1019-1026, Oct. 2012. [169] E. Oken, E. B. Levitan, and M. W. Gillman, "Maternal smoking during pregnancy and child overweight: systematic review and meta-analysis," International Journal of Obesity (London), vol. 32, pp. 201-10, Feb. 2008. [170] T. Ino, "Maternal smoking during pregnancy and offspring obesity: meta-analysis," Pediatrics International, vol. 52, pp. 94-99, 2010. [171] A. A. Mamun, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Maternal smoking during pregnancy predicts adult offspring cardiovascular risk factors – evidence from a community- based large birth cohort study," PLoS ONE, vol. 7, p. e41106, 2012. [172] C. Power and B. J. Jefferis, "Fetal environment and subsequent obesity: a study of maternal smoking," International Journal of Epidemiology, vol. 31, pp. 413-419, Apr. 2002. [173] S. Arenz, R. Ruckerl, B. Koletzko, and R. von Kries, "Breast-feeding and childhood obesity[mdash]a systematic review," International Journal of Obesity and Related Metabolic Disorders, vol. 28, pp. 1247-1256, 08/17/online 2004. [174] S. F. Weng, Redsell, S.A., Nathan, D., Swift, J.A., Yang, M., Glazebrook, C., "Estimating overweight risk in childhood from predictors during infancy," Pediatrics, vol. 132, 2013. [175] M. W. Gillman, S. L. Rifas-Shiman, Camargo, C. A. Jr, and et al., "RIsk of overweight among adolescents who were breastfed as infants," JAMA, vol. 285, pp. 2461-2467, 2001. [176] T. Harder, R. Bergmann, G. Kallischnigg, and A. Plagemann, "Duration of breastfeeding and risk of overweight: a meta-analysis," American Journal of Epidemiology, vol. 162, pp. 397-403, Sep. 2005. [177] W. H. Oddy, T. A. Mori, R.-C. Huang, J. A. Marsh, C. E. Pennell, P. T. Chivers, et al., "Early infant feeding and adiposity risk: from infancy to adulthood," Annals of Nutrition and Metabolism, vol. 64, pp. 262-270, 2014.

70

[178] T. J. Parsons, C. Power, and O. Manor, "Infant feeding and obesity through the lifecourse," Archives of Disease in Childhood, vol. 88, pp. 793-794, Sep. 2003. [179] C. G. Owen, R. M. Martin, P. H. Whincup, G. D. Smith, and D. G. Cook, "Effect of infant feeding on the risk of obesity across the life course: a quantitative review of published evidence," Pediatrics, vol. 115, pp. 1367-1377, 2005. [180] K. B. Michels, W. C. Willett, B. I. Graubard, R. L. Vaidya, M. M. Cantwell, L. B. Sansbury, et al., "A longitudinal study of infant feeding and obesity throughout life course," International Journal of Obesity, vol. 31, pp. 1078-1085, 04/24/online 2007. [181] L. S. Goodell, D. B. Wakefield, and A. M. Ferris, "Rapid weight gain during the first year of life predicts obesity in 2-3 year olds from a low-income, minority population," Journal of Community Health, vol. 34, pp. 370-5, Oct. 2009. [182] K. Ong and R. Loos, "Rapid infancy weight gain and subsequent obesity: Systematic reviews and hopeful suggestions," Acta Paediatrica, vol. 95, pp. 904-908, 2006. [183] P. O. A. Monteiro and C. G. Victora, "Rapid growth in infancy and childhood and obesity in later life – a systematic review," Obesity Reviews, vol. 6, pp. 143-154, 2005. [184] B. Hitze, A. Bosy-Westphal, S. Plachta-Danielzik, and F. Bielfeldt, "Long-term effects of rapid weight gain in children, adolescents and young adults with appropriate birth weight for gestational age: the kiel obesity prevention study," Acta pædiatrica (Oslo), vol. 99, pp. 256-262, 2010. [185] N. Stettler, B. S. Zemel, S. Kumanyika, and V. A. Stallings, "infant weight gain and childhood overweight status in a multicenter, cohort study," Pediatrics, vol. 109, pp. 194-199, 2002. [186] B. A. Dennison, A. D. Barbara, S. E. Lynn, H. S. Howard, and M. P. Robert, "Rapid infant weight gain predicts childhood overweight," Obesity (Silver Spring, Md.), vol. 14, pp. 491-499, 2006. [187] C. A. Robinson, A. K. Cohen, D. H. Rehkopf, J. Deardorff, L. Ritchie, R. T. Jayaweera, et al., "Pregnancy and post-delivery maternal weight changes and overweight in preschool children," Preventive Medicine, vol. 60, pp. 77-82, Mar. 2014. [188] T. A. Ajslev, L. Angquist, K. Silventoinen, J. L. Baker, and T. I. Sorensen, "Trends in parent-child correlations of childhood body mass index during the development of the obesity epidemic," PLoS One, vol. 9, p. e109932, 2014. [189] T. A. Ajslev, L. Angquist, K. Silventoinen, J. L. Baker, and T. I. A. Sorensen, "Stable intergenerational associations of childhood overweight during the development of the obesity epidemic," Obesity, vol. 23, pp. 1279-1287, Jun. 2015.

71

[190] D. Devakumar, C. Grijalva-Eternod, M. Cortina-Borja, J. Williams, M. Fewtrell, and J. Wells, "Disentangling the associations between parental BMI and offspring body composition using the four-component model," American Journal of Human Biology, pp. n/a-n/a, 2016. [191] S. L. Anzman, B. Y. Rollins, and L. L. Birch, "Parental influence on children's early eating environments and obesity risk: implications for prevention," International Journal of Obesity, vol. 34, pp. 1116-1124, 07//print 2010. [192] E. R. Kyung, P. Suzanne, and M. Jeanne, "Early determinants of obesity: genetic, epigenetic, and in utero influences," International Journal of Pediatrics, vol. 2012, p. 9, 2012. [193] Agatha A. van der Klaauw and I. S. Farooqi, "The hunger genes: pathways to obesity," Cell, vol. 161, pp. 119-132, 3/26/ 2015. [194] J.-P. Chaput, L. Pérusse, J.-P. Després, A. Tremblay, and C. Bouchard, "Findings from the Quebec Family Study on the Etiology of Obesity: genetics and environmental highlights," Current Obesity Reports, vol. 3, pp. 54-66, 2014/03/01 2014. [195] T. M. Frayling, N. J. Timpson, M. N. Weedon, E. Zeggini, R. M. Freathy, C. M. Lindgren, et al., "A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity," Science, vol. 316, pp. 889-94, May. 2007. [196] D. Albuquerque, E. Stice, R. Rodriguez-Lopez, L. Manco, and C. Nobrega, "Current review of genetics of human obesity: from molecular mechanisms to an evolutionary perspective," Molecular Genetics and Genomics, vol. 290, pp. 1191-221, Aug. 2015. [197] E. Oken and M. W. Gillman, "Fetal origins of obesity," Obesity Research, vol. 11, pp. 496- 506, 2003. [198] N. Farah, J. Hogan, V. O'Dwyer, B. Stuart, M. Kennelly, and M. J. Turner, "Influence of maternal glycemia on intrauterine fetal adiposity distribution after a normal oral glucose tolerance test at 28 weeks gestation," Experimental Diabetes Research, vol. 2011, p. 951203, 2011. [199] T. J. Parsons, Power, C., Manor, O., "Fetal and early life growth and body mass index from birth to early adulthood in 1958 British cohort: longitudinal," British Medical Journal, vol. 323, 2001. [200] R. C. Whitaker and W. H. Dietz, "Role of the prenatal environment in the development of obesity," Journal of Pediatrics, vol. 132, pp. 768-776, 5// 1998. [201] M. Alfaradhi and S. Ozanne, "Developmental programming in response to maternal over- nutrition," Frontiers in Genetics, vol. 2, Jun. 2011. [202] E. C. Cottrell and S. E. Ozanne, "Early life programming of obesity and metabolic disease," Physiology & Behavior, vol. 94, pp. 17-28, 4/22/ 2008.

72

[203] M. Desai, M. Beall, and M. Ross, "Developmental Origins of Obesity: Programmed Adipogenesis," Current Diabetes Reports, vol. 13, pp. 27-33, 2013/02/01 2013. [204] M. Li, D. Sloboda, and M. Vickers, "Maternal obesity and developmental programming of metabolic disorders in offspring: evidence from animal models," Experimental diabetes research, vol. 2011 , 2011. [205] A. J. Drake and R. M. Reynolds, "Impact of maternal obesity on offspring obesity and cardiometabolic disease risk," Reproduction, vol. 140, pp. 387-398, Sep. 2010. [206] S. P. Ford and N. Tuersunjiang, "Maternal obesity: how big an impact does it have on offspring prenatally and during postnatal life?," Expert Review of Endocrinology & Metabolism, vol. 8, pp. 261-273, 2013. [207] K. K. Davison, Birch, L.L., "Obesigenic families: Parents' physical activity and dietary intake patterns predict girls' risk of overweight," International Journal of Obesity, vol. 26, 2002. [208] K. K. Davison, L. A. Francis, and L. L. Birch, "Reexamining obesigenic families: Parents' obesity-related behaviors predict girls' change in BMI," Obesity Research, vol. 13, pp. 1980-1990, Nov. 2005. [209] A. B. Lloyd, D. R. Lubans, R. C. Plotnikoff, C. E. Collins, and P. J. Morgan, "Maternal and paternal parenting practices and their influence on children's adiposity, screen-time, diet and physical activity," Appetite, vol. 79C, pp. 149-157, Apr. 2014. [210] K. J. Gruber and L. A. Haldeman, "Using the family to combat childhood and adult obesity," Preventing Chronic Disease, vol. 6, p. A106, 2009. [211] R. Jago, K. K. Davison, R. Brockman, A. S. Page, J. L. Thompson, and K. R. Fox, "Parenting styles, parenting practices, and physical activity in 10- to 11-year olds," Preventive Medicine, vol. 52, pp. 44-47, 1// 2011. [212] H. Patrick and T. A. Nicklas, "A review of family and social determinants of children’s eating patterns and diet quality," Journal of the American College of Nutrition, vol. 24, pp. 83-92, 2005. [213] G. Hendrie, G. Sohonpal, K. Lange, and R. Golley, "Change in the family food environment is associated with positive dietary change in children," International Journal of Behavioral Nutrition and Physical Activity, vol. 10, p. 4, 2013. [214] N. Shloim, L. R. Edelson, N. Martin, and M. M. Hetherington, "Parenting Styles, Feeding Styles, Feeding Practices, and Weight Status in 4-12 Year-Old Children: A Systematic Review of the Literature," Frontiers in Psychology, vol. 6, p. 1849, 2015.

73

[215] K. van der Horst, A. Oenema, I. Ferreira, W. Wendel-Vos, K. Giskes, F. van Lenthe, et al., "A systematic review of environmental correlates of obesity-related dietary behaviors in youth," Health Education Research, vol. 22, pp. 203-226, Apr. 2007. [216] K. J. Campbell, D. A. Crawford, and K. Ball, "Family food environment and dietary behaviors likely to promote fatness in 5-6 year-old children," International Journal of Obesity (London), vol. 30, pp. 1272-80, Aug. 2006. [217] K. McCurdy, K. S. Gorman, T. Kisler, and E. Metallinos-Katsaras, "Associations between family food behaviors, maternal depression, and child weight among low-income children," Appetite, vol. 79C, pp. 97-105, Apr. 2014.

74

Chapter 3: Methodology

To achieve the study aim and objectives, information from two birth cohorts was used: Universiti Sains Malaysia (USM) Pregnancy cohort and MUSP (Mater – University of Queensland Study of Pregnancy) birth cohort. Data from the USM cohort was used to examine the short-term impact of parental pre-pregnancy BMI and adiposity on offspring BMI and adiposity; and data from the MUSP cohort was used to examine the long-term effect of parental pre-pregnancy BMI and adiposity on offspring BMI and adiposity. The MUSP cohort was also used to investigate the differential effect of pre-pregnancy parental BMI and obesity over the development of offspring BMI and obesity, across the life course. Also examined was the impact that exposure to parental overweight and obesity before pregnancy has on offspring BMI change, and whether the changes between maternal 20-year postpartum weight and BMI influence offspring weight, BMI and waist circumference in adulthood.

3.1 Overall study design and population

Table 3.1 presents an overview of the design and population for both of the birth cohorts. Table 3.1 Birth cohorts used in this study Cohort Design Population Universiti Sains Malaysia Longitudinal 214 parents and offspring (USM) Pregnancy Cohort Study

Mater – University of Longitudinal 7223 parents and offspring Queensland Study of Pregnancy (MUSP)

USM Pregnancy Cohort

The USM Pregnancy Cohort is used to investigate Objective 2 and relevant maternal lifestyle factors, such as dietary pattern and physical activity. However, this cohort study only looks at the short-term association of pre-pregnancy parental and offspring obesity and adiposity, that is, up to one year of age of offspring. The USM Pregnancy Cohort consisted of 188 maternal and offspring pairs. The birth cohort was initially developed to determine oxidative stress during pregnancy and it started in April 2010 and finished in December 2012. This study was conducted in the state of Kelantan, Peninsular Malaysia and consisted of 214 pregnant women interviewed at baseline. All of the subjects were recruited from the Kubang Kerian Health Clinic and Obstetrics and Gynaecology Clinic of Hospital USM. Mothers and offspring from this cohort were followed up at 18 weeks and

75

34 weeks of gestation; at birth; and post-birth at 2 months, 6 months and 12 months. The USM study was approved by the Research Ethics Committee (Human) HUSM and Medical Research Ethics committee (Ministry of Health Malaysia). Written informed consent was obtained from each participant. The sampling procedure used to obtain the cohort is shown in Figure 3.1

Data were collected using interview-administered questionnaires which consisted of three sections. The first section was for parental and infant socio-demographic background data, including parental age, educational level, occupation, monthly household income, marital status, obstetric history, medical history, smoking history, and birth outcomes. The second section was for parental and infant anthropometry data (paternal data at baseline only), such as weight, height, body fat distribution, skin-fold thickness measurements and waist and hip circumference (for fathers only). Finally, the third section was for maternal and infant dietary data. Maternal physical activity was added after the birth. Detailed study methods have been reported previously [1].

76

Eligible pregnant women

n = 408  194 refusals

Consent and initial interview (≤24 weeks of gestation)

n = 214

 26 lost to follow-up

Pregnancies

Second trimester (14-24 weeks of gestation)

Prenatal phase Prenatal n = 188  24 developed or hypertension Third trimester  3 lost to follow-up (≥32 weeks of gestation)

n = 161

 8 with preterm delivery  2 mothers lost to follow-up Babies delivered (≥37 weeks of gestation) Mothers, n = 151 Infants, n = 153

 4 withdrawals

Mother–Infant Visits 2 months visits

n = 149

 3 withdrawals Postnatal phase Postnatal 6 months visits

n = 146  1 withdrawal  7 pregnant 12 months visits Mothers, n = 138 Infants, n = 145

Figure 3.1 Subject flow diagram of USM Pregnancy Cohort Study [2]

77

MUSP

For the purpose of examining the third, fourth and fifth objectives, prospective data were used from MUSP. The MUSP is a prospective study of 7,223 women and their offspring. Mothers received antenatal care at a major public hospital in Brisbane between 1981 and 1983 and delivered a live singleton child who was not adopted before leaving the hospital. Mothers and children have been followed-up prospectively with maternal questionnaires being administered at the first antenatal clinic visit; 3–5 days post-delivery; then at 6 months; and 5, 14 and 21 years after birth. Ethical approval was obtained from the ethics committee at the University of Queensland and Mater Hospital, Queensland, Australia. Participants provided signed informed consent for their participation and that of their children. The response rate at the 6-month follow-up was 93%; at 5– year follow-up 73%; at 14–year follow-up 72%, and at 21–year follow-up 57%. In general, participants lost to follow-up were more likely to be males and of Asian and Aboriginal and/or Torres Strait Islander background (all P-values <0.001). Their mothers were more likely to be teenagers, less educated, single or cohabitating, have three or more children, used tobacco and alcohol during pregnancy, and be anxious and depressed at their first antenatal visit. Figure 3.2 provides details of the MUSP cohort using flow diagrams.

78

Figure 3.2 Flow diagram of participation in MUSP study phases [3]

79

3.2 Measurement of exposure

USM Pregnancy Cohort

Parental BMI was derived from weight and height obtained during the 18 weeks of gestation. Parental height were measured using SECA 206, microtoise tape to the nearest 0.1cm. Both maternal and paternal weight, fat percentage, fat mass and fat-free mass—were obtained using the TANITA SC-330 body composition analyser. Maternal pre-pregnancy weight during first clinical visit was self-reported. Paternal BMI was derived by weight in kilograms divided by the height in metre square. Maternal pre-pregnancy BMI was calculated using measured height in pregnancy and pre-pregnancy self-reported weight. Parental BMI was categorised based on the WHO, 1998 [4] cut-off: underweight (BMI <18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2) and obese (BMI >30.0 kg/m2). Overweight and obese categories were combined into an ‘overweight and obese’ (OW/OB) group for further analysis.

MUSP

The main exposures to determine the association between parental weight status or BMI and offspring weight status or BMI throughout the life course were parental BMI before or during pregnancy. Paternal weight and height were self-reported by the mother. Maternal pre-pregnancy weight was self-reported and height was measured at the baseline data collection. Both parental BMI were further calculated and divided into categories defined using WHO guidelines [4]. Underweight (BMI <18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0– 29.9 kg/m2) and obese (BMI >30.0 kg/m2). Overweight and obese categories were then combined into an OW/OB group for further analysis. To determine the association of maternal weight change with offspring BMI/weight change, maternal weight change will be calculated by subtracting maternal pre-pregnancy weight from maternal weight at 21-years post-partum.

3.3 Measurement of outcomes

USM Pregnancy Cohort

Child data and adiposity measurement including weight, length, MUAC, TSF and SSF were obtained and measured by the USM researchers. All these measurements when the offspring were 2, 6 and 12 months old were considered. Weight and length of the child were measured according to standard procedure (Mirnalini et al., 2007) by a researcher at each data collection follow-up during a clinical visit. Infant recumbent crown-heel length was measured to the nearest 0.1cm using a baby

80

weighing scale with attached measuring rod and head positioners (Seca 334, Hamburg, Germany). . Baby weighing scale (Seca 334, Hamburg, Germany) was used to measured weight and it was calibrated to the nearest 0.001 kg. Mother and infant were weighed together on the body composition analyser (Tanita SC-330ST, Tokyo, Japan) and maternal weight was subtracted from the total weight. This situation was limited to only infants who were restless. Child weights, length, weight for length and BMI at 12 months were computed using the WHO Anthro Software (Version 3.2.2). These growth indices were converted to sex and age specific z-scores using WHO child growth standards 2006 [5] Infant nutritional status was classified according to the z-scores as shown in Table 3.2.

Table 3.2 Definitions of growth indicators in terms of z-scores

Z-score Growth Indicators Length-for-age Weight-for-age Weight-for- BMI-for-age length >3 Note 2 Obese Obese >2 Overweight Overweight Note 3 >1 Possible risk of Possible risk of

overweight overweight 0 (medium) <–1 <–2 Stunted Underweight Wasted Wasted <–3 Severely Severely Severely wasted Severely wasted stunted underweight Notes: 1. Measurements in the shaded area indicate a normal range. 2. Very tall child, but it is rarely a problem unless it is so excessive which may indicate an endocrine disorder such as a growth-hormone-producing tumour. 3. A child may have a growth problem, but this is better assessed from weight-for-length or BMI-for-age. MUSP

Measured weight and height—as well as BMI of offspring at the age of 5, 14 and 21 years—were considered as main outcomes. At the 14- and 21-year follow-ups, offsprings’ height without shoes was measured to the nearest centimetre using a portable stadiometer. Their weight was measured in light clothing with a scale accurate to 0.2 kg. Two measures of weight were taken with a five- minute interval, and the mean of these two measures was used in all analyses. At 21 years, waist circumference (WC) was measured horizontally using a tape roughly in line with the respondent’s navel, directly against the skin without compressing the skin. The average of two measures was taken. Similarly, the hip measure was repeated twice and the average taken. Adolescents’ BMI was calculated based on measured height and weight at the 14-year follow-up, and BMI was categorised as normal or overweight according to standard definitions derived from international surveys by 81

Cole et al, 2000. At the 21-year follow-up, BMI was categorised into normal (<25 kg/m2), overweight (25–29 kg/m2) and obese (30 kg/m2) using the WHO classification of BMI cut-offs. WC (in centimetres) and waist-hip ratio (WHR) were measured at the same time. WC was categorised for males as: <94 cm normal, 94 to <102 cm abdominal overweight, and 102 cm abdominal obesity; and for females as: <80 cm normal, 80 to <88 cm abdominal overweight, and 88 cm abdominal obesity. WHR was categorised for males as: <0.90 cm normal, 0.90 to <1.0 cm overweight, and 1.0 cm obese; and for females as: <0.80 cm normal, 0.80 to <0.85 cm overweight and 0.85 cm obese. Measurements in the shaded area indicate a normal range. For further analyses, overweight and obese offspring were combined due to small numbers in the obese group.

3.4 Measurement of covariates and other independent variables

USM Pregnancy Cohort

Maternal gestational weight gain (GWG) was defined as the difference between self-reported pre- pregnancy weight and the last measured weight before delivery. Maternal BMI and GWG were further grouped according to the IOM (Institute of Medicine) [6] criteria: underweight (BMI <18.5 kg/m2) should gain 12.5–18.0 kg during pregnancy, normal weight (BMI 18.5–24.9 kg/m2) should gain 11.5–16.0 kg, overweight (BMI 25.0–29.9 kg/m2) should gain 7.0–11.5kg and obese (BMI >30.0 kg/m2) should gain 5.0–9.0 kg. Mothers who gained below and over the IOM recommendation were grouped as ‘inadequate’ and ‘excess’, respectively.

Household monthly income, maternal educational level, maternal age, parity, gestational age, maternal physical activity and maternal dietary intake as well as child birth weight were identified as potential confounders and were adjusted during the analysis. Household monthly income, maternal educational level, parity and gestational age were transformed into categorical variables, such as for monthly household income divided into three categories: low (RM5600), while maternal educational level was divided into primary plus secondary, diploma and tertiary level. Parity, which showed the number of previous children, was classified into two categories: 0 and 1 while gestational age was classified into preterm (<37 weeks) and term (>37 weeks). Maternal age was in years and maintained as continuous variables. Birth weight of offspring was recorded to the nearest gram.

82

Additionally, maternal physical activity measurement was determined using a short form of the International Physical Activity Questionnaire (IPAQ) and values given as metabolic equivalents (MET-minutes/week). Maternal physical activities over the period of the past one week were obtained in the postnatal phase via interview. There were three specific types of activity that were assessed by the IPAQ: walking, moderate-intensity activities, and vigorous-intensity activities. Total score of physical activity was computed from the summation of the duration (in minutes) and frequency (days) of these three types of activity, according to the IPAQ’s scoring protocol (Table 3.3); and classified as low, moderate and high levels of physical activity [7].

Table 3.3 Formula for the computation of physical activity scores

Physical activity Score of physical activity (MET-minutes/week) Walking 3.3 x walking minutes x walking days Moderate-intensity 4.0 x moderate-intensity activity minutes x moderate days activity Vigorous-intensity 8.0 x vigorous-intensity activity minutes x vigorous-intensity activity days Total physical activity the sum of walking + moderate + vigorous MET-minutes/week scores

Using 24 hours dietary recall, mothers were interviewed at 18 weeks of gestation to obtain maternal dietary intake. In the third trimester, maternal food consumption in past six months was collected using a validated semi-quantitative food-frequency questionnaire [8],. Dietary pattern scores were used as continuous variables.

MUSP

Possible confounding factors were considered for this cohort in the models. Some of the confounding factors were categorical variables and included parental educational attainment (classified into three categories: incomplete education, complete secondary education and completed further/higher education), household income ( low income: A$10400), maternal age (13–19, 20–35 and >35 years) maternal smoking habits (pre-pregnancy and/or in early pregnancy only, throughout pregnancy and never), maternal alcohol consumption during pregnancy (never, 1–2, and > 2), gestational age (preterm and term). Maternal total weight gain was considered as a continuous variable. GWG during pregnancy was calculated from maximum weight measured in pregnancy and the mother's self-reported pre-pregnancy weight.

83

For offspring, early life factors such as birth weight were used as continuous variables and adjusted in some models. Offspring gender (male, female) and offspring breastfeeding duration were used at the 6-month follow-up; and at the 14-year follow-up, offspring lifestyle factors, such as TV watching (3 hours or <3 hours during weekdays and weekends), sports participation (3 hours or <3 hours spent on sports/exercise per week), and family meal times (at least once a day, a few times a week, and less than that) were used.

Overall measurement of primary predictor and outcome variables, based on my study objectives, are shown in Table 3.4. Table 3.4 Summary of overall predictors, outcome and other independent variables

Other independent variables Objectives Primary predictors Outcome variables (covariates) USM cohort Household income, maternal Offspring weight education level, maternal age at Pre-pregnancy status and adiposity 12 birth, parity, gestational age, parental BMI and months. Considering gestational weight gain, maternal 2 adiposity during early weight and length at physical activity and dietary pregnancy (18 weeks birth, 2 months, and 6 intake during pregnancy, gestation) months offspring birth weight and sex.

MUSP cohort Pre-pregnancy Baseline: parental educational Offspring BMI and parental (maternal and attainment, annual household WC at 21 years paternal) BMI income Birth: maternal age, Offspring weight and maternal smoking during BMI change from 5 to pregnancy, gestational age, Pre-pregnancy 14years, 14 to 21years gestational weight gain, 3 parental BMI and its and 5 to 21years 6-month follow-up: offspring categories Offspring weight and birth weight and sex, BMI at 5, 14 and 21 breastfeeding duration 14-year years follow-up: offspring lifestyle (hours spent watching TV, sports Maternal weight and Offspring weight, BMI participations, eats together with 4 BMI change at 21 and WC at 21 years family, and fast food years postpartum consumption).

3.6: Statistical analysis

To examine the pre-pregnancy parental–offspring obesity/BMI association, descriptive and analytical analyses were initially conducted using SPSS version 16.0 for both the USM and MUSP cohorts. Statistically significant association, difference and precision were determined by 95% confidence intervals. All the variables were screened for multicollinearity using VIF (variance

84

inflation factor) before being considered in the regression analysis. VIF with the value less than 4.0 were used as rule of thumb. In Table 3.5 the statistical methods used for each of the four objectives are listed.

Table 3.5 Statistical method based on objectives Objectives Method 1 2 3 4 Descriptive statistic X X X Pearson correlation X X X Linear regression X X X Logistic regression X X X Chi-square test X X X T-Test independent samples X X X Generalised Estimating Equation (GEE) X X X Multiple Imputation X X

Descriptive analysis

Data exploration involving parental and offspring socio-demographic background was used. Missing data from MUSP were dealt with by screening for incomplete data sets for predictors (particularly parental weight and height) and main outcomes (offspring BMI at 5, 14 and 21 years). A total of 3519 persons at 5 years, 3317 persons at 14 years, and 3195 persons at 21 years provided complete weight and height data for analyses. For the USM cohort, from a total of 214 persons, data from 153 persons was used, after excluding those with a preterm infant. All continuous data for both cohorts were reported as mean and standard deviations while categorical data were reported as percentages. For statistical analyses, the F test was used for continuous data and chi-squared test was used for categorical data.

Multivariable analysis

USM Pregnancy Cohort A series of multiple linear regression were used to examine the association between parental pre- pregnancy BMI and offspring weight and growth status at one year. Parental BMI is the main predictor in this analysis and child weight for age z-score (WAZ), length for age z-score (LAZ), weight for length z-score (WLZ) and BMI z-score (BAZ) through to 12 months of age are the main outcomes. I used separate models for each of the outcomes. Firstly, simple linear regression was used where one of the child growth factors is the outcome and maternal and paternal BMI is the predictor (i.e. an unadjusted model). Next, in the second model, parental socio-demographic factors, 85

such as age, education and household income, were included. In the third model, in addition to the unadjusted model I have included offspring factors, such as birth weight and gestational age. In the fourth model, maternal physical activities and dietary pattern during pregnancy were included. Finally, all variables from Model 1 to Model 4 were included. Results of this model are interpreted as the rate of change in child weight/growth for every unit change of maternal or paternal BMI score and presented as regression coefficients with 95% confidence intervals.

Similarly, to examine the prospective association between parental body composition during pregnancy and child growth and body fat, we have used the multiple linear regression where outcomes are children’s skin-fold thickness and circumference at one year: child WAZ, LAZ, WLZ and BAZ; and predictor is parental adiposity distribution during pregnancy (i.e. total fat mass, fat free mass, fat percentage and visceral fat).

MUSP To determine the association between parental BMI and obesity and child BMI and obesity throughout life, a series of multiple linear regressions and logistic regressions were used. Additionally, generalised estimating equations (GEE) were used to determine the overall estimate and differential impact of prenatal parental BMI on offspring BMI across life. A series of models were used to determine predictors of outcomes. In the first model, father’s or mother’s BMI and offspring gender were adjusted. In the second model, parental factors such as education attainment during baseline, annual household income and maternal age at birth, maternal gestational weight gain, and maternal smoking habit during pregnancy were added. In the third model, offspring early life factors (birth weight, breastfeeding duration and gestational age) were added. Finally, all variables were included in the final models from Model 1 to Model 3, including offspring lifestyle factors at 14 years such as TV watching, sports involvement, fast food consumption and meal time with family.

To examine the links of pre-pregnancy parental BMI association with offspring weight and BMI change at a different stage of life, six categorical outcomes were considered: normal weight at 5 and 14 years, normal weight at 14 and 21 years, normal weight at 5 years and overweight and obese at 14 years, normal weight at 14 years and overweight and obese at 21 years, overweight and obese at 5 and 14 years, and overweight and obese at 14 and 21 years. Likewise, multinomial logistic regression was used to examine pre-pregnancy parental weight status in categories—both parents normal weight/ underweight (NW/UW); maternal NW/UW and paternal OW/OB; paternal NW/UW 86

and maternal OW/OB; and both parents OW/OB—with offspring BMI weight change in categories at 5, 14 and 21 years.

To explore the association between parental BMI and child BMI through the life course, parental BMI before pregnancy was the predictor, and the main outcome was offspring BMI at 5, 14 and 21 years. Multiple linear regression was used to examine the association with different series of models for each outcome (BMI at 5, 14 and 21 years). In additional analysis, considering the differences in range of BMIs between parents and offspring, BMI z scores were derived by adjusting the age for parent and adult offspring BMI; while for children aged 5 and 14 years, BMI z scores were standardised by adjusted age and sex, based on CDC growth references [9]. Maternal relative weight and BMI change were stratified into quintiles to further examine the impact of maternal change in 21 years on offspring outcomes.

All results were interpreted as regression coefficient with 95% confidence interval. Multiple logistic regression were used, adjusted for covariates (in final models) as stated above. Odd ratio and 95% confidence interval were used to interpret the results for categorical predictors and outcomes.

References [1] S. Loy and J. Hamid Jan, "The Universiti Sains Malaysia Pregnancy Cohort Study: Maternal-infant Adiposity Development until the First Year of Life," Health and the Environment Journal, vol. 5, pp. 50-64, 2014. [2] S. L. Loy, "A prospective study on maternal oxidative stress in pregnancy and postpartum and infant adiposity development during the first year of life.," 21, The Malaysian Journal of Medical Sciences : MJMS, Penerbit Universiti Sains Malaysia, Universiti Sains Malaysia, 2014. [3] J. M. Najman, R. Alati, W. Bor, A. Clavarino, A. Mamun, J. J. McGrath, et al., "Cohort Profile Update: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology, Dec. 2014. vol 44(1) & page 78-78f [4] WHO, "Obesity: Preventing and Managing a Global Epidemic, Report of a WHO Consultant on Obesity". Geneva: World Health Organization, 1998. [5] World Health Organization (WHO). (2008, May). World Health Organization Child Growth Standards: Training course on child growth assessment - interpreting growth indicators. Available: http://www.who.int/childgrowth/training/en/ [6] IOM, Institute of Medicine. Weight Gain During Pregnancy: Reexamining the Guidelines. : Washington, DC: The National Academies Press, 2009. 87

[7] IPAQ. (2005, Sep.). Guidelines for Data Processing and Analysis of The International Physical Activity Questionnaire (IPAQ)-Short and Long Forms, revised on November, 2005. . Available: http://www.ipaq.ki.se/scoring/pdf. [8] S. L. Loy and J. M. Hamid Jan, "Relative validity of dietary patterns during pregnancy assessed with a food frequency questionnaire," International Journal of Food Sciences and Nutrition, vol. 64, pp. 668-673, 2013. [9] T. Cole, M. Bellizzi, K. Flegal, and W. Dietz, "Establishing a standard definition for child overweight and obesity worldwide: international survey," BMJ, vol. 320, pp. 1240 - 1243, 2000.

88

Chapter 4: Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia

4.1 Context

In the critical literature reviews we have found that most of the studies examined the association of parental pre-pregnancy BMI and offspring BMI and obesity could not adjust for important confounders around pregnancy, such as valid measures of diets and physical activity. In this chapter, data from the Universiti Sains Malaysia Pregnancy Cohort were used due to the richness of data about parental lifestyle factors which were collected during pregnancy. This cohort data also provided other anthropometry measurements which were objectively measured during each data collection phase. Therefore, by using this cohort, we were able to observe the association between parental BMI and adiposity before pregnancy with offspring anthropometric measurements in early childhood.

From this short-term follow-up study, we identified that parental BMI influences offspring growth and anthropometric measurement in the first year of life. However, the association was not significant after adjusting for maternal dietary intake and physical activity during pregnancy. This may suggest that the short-term association between parental and offspring BMI and obesity may be confounded by dietary intake and energy expenditure of the parents. This chapter forms a manuscript which has been accepted for publication in the Journal of Obesity Research and Clinical Practice.

89

4.2: Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia.

Nurzalinda Zalbahar (MSc), Hamid Jan Mohamed Jan (PhD), Loy See Ling (PhD), Jake Najman (PhD), Harold David McIntyre (MD), Abdullah Mamun (PhD).

Abstract Background and objectives While the prevalence of obesity has peaked in many developed countries, it is rapidly increasing in most of the developing countries. Parental body mass index (BMI) is known as an important factor that is strongly linked with the development of offspring overweight and obesity. However, there are still limited studies focusing on the association of parental BMI before pregnancy on offspring growth and body composition in early life, particularly in developing countries. Methods Data from the Universiti Sains Malaysia (USM) Pregnancy Cohort, which consisted of 153 mother–offspring pairs, were used. Data were collected using interview-administered questionnaires and anthropometric measurements were obtained. Multiple linear regression and generalised equation estimation (GEE) were used to examine the direction and impact of the association between parental BMI and child growth and body composition (weight for age, height for age, body mass index for age, weight for height and fat mass at age 2 months, 6 months, and 12 months). Potential confounders, including validated measures of maternal diet and physical activity during pregnancy, were considered. Results Of 153 parents, one-quarter of the mothers and 42.2% of fathers before pregnancy were overweight or obese, respectively. Significant association was found between maternal BMI and child’s weight for age z-score (WAZ), weight for height z-score (WLZ) and body mass index for age z-score (BAZ), although it was attenuated in the fully adjusted model. No significant association was found between parental fat percentage and child’s skinfold thickness and body fat percentage at 12 months of age. Conclusions Having high pre-pregnancy BMI may increase BMI and WLZ of offspring in early life. Findings from this study may emphasise the importance among Malaysians of monitoring maternal weight status, particularly before and during pregnancy and during the early life of offspring.

Introduction 90

The global epidemic of overweight and obesity is a major public health concern. The prevalence of overweight and obesity is rising dramatically among children, adolescents and adults [1]. Increasing evidence consistently suggests that parental overweight and obesity is one of the most important familial and environmental factors that are strongly linked with the development of offspring weight status [2-4]. When both parents are OW/OB, the risk of offspring being OW/OB is stronger compared to only one parent being OW/OB [5]. Some studies also reported that the association between mother–offspring pairs is stronger than father–offspring pairs [6, 7], although others found no difference [8, 9].

The positive associations of maternal weight on infant growth and infant birth weight are well known [10, 11]. Recent meta-analysis suggested that maternal weight status not only increased maternal post-partum weight retention and health risk, but also influenced the adiposity and metabolic syndrome components among offspring [12]. Infants who have high- end distribution of weight status and higher size or growth will increase the risk of being obese later in life [13]. Additionally, offspring who are born to OW/OB mothers have a high risk of developing greater BMI and body fat in early childhood [14-16]. Although most studies reported a strong positive association between parental and offspring weight status, none of them controlled the association by using a validated measure of dietary patterns and physical activities of parents during pregnancy [2, 5]. Most studies relied on the self-reported maternal and paternal BMI before pregnancy, which might provide a biased estimate of parental weight status. Understanding the associations between parental and offspring anthropometrics and body compositions may further assist in the development of effective early obesity intervention, involving early-life environmental factors.

Studies related to parental–offspring body composition or weight status associations were mainly from the western and developed countries and remained under-studied in developing countries. Therefore, this study aimed to determine the association of parental body composition—including weight status and body fat distribution on offspring anthropometrics at 2 months, 6 months and 12 months of age—by using the prospective USM pregnancy cohort data.

Subjects and method

91

Data were derived from the USM pregnancy cohort study, which was established in 2010 and initially aimed to determine the causes of maternal and infant adiposity in Kelantan State, Peninsular Malaysia. As a baseline, the USM pregnancy cohort consisted of 214 pregnant women who were interviewed after providing written informed consent. Of these women, 153 mothers–infant pairs were finally included in the prenatal phase. A total of 61 women were lost to follow-up or excluded from the study because of developing gestational diabetes or pregnancy induced hypertension and with preterm labour. All women were recruited from the Kubang Kerian Health Clinic and Obstetrics and Gynaecology Clinic of Hospital USM. Mothers and offspring from this cohort were followed up at 18 weeks of gestation, 34 weeks of gestation, at birth, and post-birth at 2 months, 6 months and 12 months. This birth cohort study was approved by the Research Ethics Committee (Human) HUSM and Medical Research Ethics committee (Ministry of Health Malaysia).

Data were collected using interview-administered questionnaires which consisted of three sections. The first section was for parental and infant socio-demographic background information, including parental age, educational level, occupation, monthly household income, marital status, obstetric history, medical history, smoking history, and birth outcomes. The second section was for parental and infant anthropometry data (paternal data at baseline only and consisted of 122 respondents) such as weight, height, body fat distribution, skin-fold thickness measurements, waist and hip circumferences. Finally, the third section was for maternal and infant dietary data. Maternal physical activities were collected using the IPAQ (seven days recalled) during the second trimester at 18 weeks of pregnancy. Detailed methods of the study have been reported elsewhere [17].

Measurements of Exposure Parental weight and body composition including fat percentage, fat mass and fat free mass were obtained using the TANITA SC-330 body composition analyser. Both paternal and maternal height was measured using microtoise tape SECA 206 to the nearest 0.1 cm. Paternal BMI (weight in kilograms divided by the height in metre square) was calculated using measured height and weight. Maternal pre-pregnancy BMI was calculated using measured height in pregnancy and self-reported weight in pre-pregnancy. Parental BMI was further categorised based on the WHO [18] cut-offs: underweight (BMI < 18.5), normal weight (BMI 18.5 – 24.9), overweight (BMI 25.0 – 29.9) and obese (BMI > 30.0).

92

Overweight and obese categories were then combined into an OW/OB group for further analysis.

Measurements of outcome Children’s anthropometric measurements—including weight, length, TSF and SSF—were measured by the trained researchers. Weight and length of offspring were measured according to the standard procedure [19] at birth, and at 2, 6 and 12 months of age. Child recumbent crown-heel length was measured using a baby weighing scale with attached measuring rod and head positioners (Seca 334, Hamburg, Germany). The measurement was taken to the nearest 0.1 cm. Weight was measured to the nearest kilogram using a baby weighing scale (Seca 334, Hamburg, Germany) attenuated to the nearest 0.001 kg. For the child who was restless, mother and child were weighed together on the body composition analyser (Tanita SC-330ST, Tokyo, Japan) and maternal weight was subtracted from the total weight, but this situation was limited to only a few cases. Child weight, length, weight for length and BMI at 2, 6 and 12 months were converted to sex and age specific z-scores using WHO child growth standards 2004 [20]. These growth indices were computed using the WHO Anthro Software (Version 3.2.2). Infant nutritional status was defined according to the z-scores [21]. Children’s body fat percentage was derived from the child’s skinfolds based on Sen et al. (2010) [22] equations for boys and girls.

Measurements of covariates Household monthly income, maternal educational attainment, maternal age, parity, gestational age, total gestational weight gain (TGWG), maternal physical activity and maternal dietary intake at early pregnancy as well as child birth weight were identified as potential confounders. Household monthly income was divided into three categories: low (RM5600); Maternal educational level was divided into primary, secondary, and diploma/tertiary level. Parity was classified into two categories: no parity and at least one parity. Maternal age was in years and gestational age was in weeks, maintained as a continuous variable. Total gestational weight gain was calculated from the difference between self-reported maternal pre-pregnancy weight and the last recorded maternal pregnancy weight.

93

Maternal physical activity at 18 weeks of pregnancy was determined using the IPAQ and it was expressed as metabolic equivalents (MET-minutes/week) following the IPAQ guideline (IPAQ, 2005). Maternal physical activities over the past week were obtained in the prenatal phase via interview. There were three specific types of activity assessed by the IPAQ, including walking, moderate-intensity activities and vigorous-intensity activities. Overall, the total score of physical activity was computed from the summation of the duration (in minutes) and frequency (days) of these three types of activity, according to the IPAQ’s scoring protocol, and classified as low, moderate and high levels of physical activity. We have presented three specific types of activities as physical activity patterns (e.g. walking, moderate-intensity activities and vigorous-intensity activity) in pregnancy and physical activity level as classified by IPAQ.

Maternal dietary intake was obtained via interview through 24-hour dietary recall in the prenatal phase, at 18 weeks of pregnancy, to estimate energy and nutrient intakes. A four- stage, multiple-pass interviewing technique was used to conduct the dietary recall. At Stage 1, women reported on all foods and beverages eaten in the previous 24 hours, starting from the first food in the morning to the last food before going to bed. At Stage 2, a detailed description of food intake was recorded, such as cooking methods and brand names. The Atlas of Food Exchanged & Portion Sizes [23], containing coloured pictures of prepared foods and household measures, was used to aid in estimating portion sizes. At Stage 3, the dietary recall was reviewed to ensure all food items were recorded correctly, including supplement intake. At Stage 4, data on portion sizes were converted to grams and entered into the Nutritionist ProTM Diet Analysis software (Axxya Systems LLC., USA) to obtain energy and nutrient values, based on the Nutrient Composition of Malaysian Foods database and the US Department of Agriculture (USDA) Foods database. For local mixed dishes, which are not available in the database, at least two local recipes were referred for each dish to obtain the average energy and nutrient values, and entered into the software. Nutrient values for branded food items were entered according to food labels. Dietary intake was later categorised as adequate or inadequate intake. Dietary adequacy was assessed by comparing maternal energy and nutrient intake with Malaysian Recommended Nutrient Intakes (RNI) [24]. Adequacy for macronutrients was achieved if the distribution of the daily calorie intake was within these ranges: total carbohydrate 55–70%, total protein 10–15% and total fat 20– 30%.

94

In the third trimester, mothers were asked to recall their food intake during the past six months in pregnancy using a validated semi-quantitative food frequency questionnaire (FFQ) [25]. Frequency of intake for each food category was then reported as either daily, weekly, monthly or never, followed by estimating the amount of intake based on standard portion sizes. Two major dietary patterns—healthy pattern and less-healthy pattern—were identified from FFQ through the factor loading score. The healthy pattern was loaded by high intake of fish and other seafood, dairy products, fruits, vegetables, nuts and legumes. The less-healthy pattern was characterised by high intake of condiments, tea and coffee, meat and offal, confectioneries, oils and fats, and cereals [25].

Statistical analysis Statistical analyses were conducted using Stata version 13.0 and statistical significant association, difference and precision were determined by the 95% confidence interval. For USM Pregnancy Cohort, a total of 153 mother–infant pairs were used. All continuous data for both cohorts were reported as mean and standard deviation, while categorical data were reported as a percentage. For any statistical analysis, the F test was used for continuous data and the chi square test for categorical data.

To examine the association between parental pre-pregnancy BMI and offspring weight and growth status up to one year, a series of multiple linear regressions were used. In this analysis, the main exposure was parental BMI and the main outcomes were child WAZ, LAZ,WLZ, BAZ, and skinfold thickness including triceps, biceps, subscapular, suprailiac and percentage body fat. Separate models were used for each of the outcomes. Initially, simple linear regression was used, where one of the child growth factors was the outcome and maternal and paternal BMI was the exposure (i.e. an unadjusted model). In the second model, in addition to the unadjusted model, parental socio-demographic factors such as age, education and household income were included. In the third model, in addition to the unadjusted model, offspring factors were included such as birth weight and gestational age. In the fourth model, in addition to the unadjusted model, maternal physical activities as calories expenditure, dietary intake at 18 weeks of gestation as calories intake, and prenatal less-healthy pattern were included. In the final model all variables from Model 1 to Model 4 were included with the other parent’s BMI adjusted. We also used GEE to examine the

95

longitudinal association of parental BMI and offspring growth throughout 12 months. Results of this model were interpreted as rate of change in child weight/growth indicators for every unit change of maternal or paternal BMI score. Results were presented as a regression coefficient with a 95% confidence interval.

Results The mean (standard deviation) maternal pre-pregnancy BMI was 22.58 (4.10) kg/m2, and one-quarter of the mothers were overweight and obese. The mean (SD) paternal BMI during early pregnancy was 24.76 (5.80) kg/m2 and 42.2% of the fathers were overweight and obese. The mean (SD) of maternal GWG was 11.76 (5.46) kg, with 26.8% of women having gained excess weight during pregnancy. Maternal mean age was 29.32 (4.44) years. The average household family income for this cohort was RM 3,225.71 (2,499.43). The majority of parents were employed and had attained tertiary level education. All mothers were non- smokers while nearly half of the fathers were smokers. About a quarter of the women were nulliparous. The mean (SD) of the child birth weight was 3.11(0.42) kg and 45.1% offspring were males. The mean BMI for the children at 12 months of age was 16.65(1.42) kg/m2.

Distribution of selected baseline characteristics by infant growth at 12 months is shown in Table 4.1. Parental and infant characteristics of maternal pre-pregnancy BMI, TGWG, maternal education, family monthly household income and infant sex were associated with a significant difference in the means of at least one of the infant anthropometric outcomes. However, no difference was found between paternal characteristic on the means of the child anthropometric outcomes. Mean WAZ, LAZ, WLZ, and BAZ increased across categories of pre-pregnancy parental BMI categories, maternal educational level, and maternal METs_18 weeks. Women with more than one parity demonstrated the lowest of all child anthropometric outcomes. Means for all four anthropometric outcomes for children increased across categories of birth weight. Maternal physical activity pattern during 18 weeks of gestation showed that the majority of mothers participated less in moderate and vigorous activity. Those mothers who walked five or more days each week spent approximately 108 minutes per week on vigorous activity. 26% of mothers spent 30–40 minutes per day doing moderate activity. Overall energy expenditure level was low for most women. The mean difference of infant’s skinfold thickness by parental BMI categories (Table 4.2) was not statistically significant and the difference was modest. Figure 4.1 shows the mean of infant

96

weight and growth status (WAZ, LAZ, WLZ and BAZ) at birth, and at 2, 6 and 12 months of age, according to paternal and maternal BMI categories (<25 kg/m2 and >25kg/m2). The mean weight of infants increased throughout the months and showed no difference between parental BMI groups. As for the mean WAZ and LAZ, these declined as the age increased. Infant weight, WLZ and BAZ were shown to increase across the age of the infant and was slightly higher for those infants with parents who were in the OW/OB (>25kg/m2) group.

Table 4.3 shows the average growth rates of children for every unit increment of maternal and paternal BMI. We used GEE which averaged four repeated measures (birth, 2 months, 6 months and 12 months) of offspring growth, adjusting for the correlation structure of the data. We have considered exchangeable correlation structure. Except for WAZ and LAZ, other growth indicators were positively associated with the maternal pre-pregnancy BMI. These associations were robust and did not attenuate considerably, adjusting for maternal socio- economic factors. However, all these associations were not statistically significant after adjusting for maternal diets (maternal total calories intake and less-healthy dietary pattern), physical activities, and paternal BMI.

In the additional analyses, when we further examined the association of parental body composition with offspring growth and body composition, we found no statistically significant association (results are not shown).

97

a) Paternal BMI b) Maternal BMI 9

8

7

Kg 6

5 <25BMI kg/m2 < 25kg/m 2 4 >25BMI kg/m2 > 25kg/m 2 3 0 2 6 12 0 2 6 12 Child age (Month) Child age (Month)

Mean weight (kg) of infant with (CI)

a) Paternal BMI b) Maternal BMI BMI < 25kg/m2 0.5 <25 BMI>25 > 25kg/m2

0

Z-score -0.5

-1

-1.5 0 2 6 12 0 2 6 12 Child age (month) Child age (month) Mean weight for age z-score (WAZ) of infant with (CI)

a) Paternal BMI b) Maternal BMI 2 <25BMI < 25kg/m 0.5 >25BMI > 25kg/m2 0

-0.5 Z-score -1

-1.5

-2 0 2 6 12 0 2 6 12 Child age (month) Child age (month) Mean length for age z-score (LAZ) of infant with (CI)

98

a) Paternal BMI b) Maternal BMI

1.5

1

0.5

0 Z-score -0.5

-1 2 <25BMI < 25kg/m -1.5 >25BMI > 25kg/m2 -2 0 2 6 12 0 2 6 12 Child age (month) Child age (month) Mean weight for length z-score (WLZ) of infant with (CI)

a) Paternal BMI b) Maternal BMI 1

0.5

0 Z-score -0.5

<25BMI < 25kg/m2 -1 >25BMI > 25kg/m2

-1.5 0 2 6 12 0 2 6 12 Child age (month) Child age (month)

Mean BMI for age z-score (BAZ) of infant with (CI) Figure 4.1 Infant growth status showing mean for weight, weight for age z-score (WAZ), length for age z-score (LAZ), weight for length z-score (WLZ) and body mass index for age z-score (BAZ) at 0, 2, 6 and 12 month according to parental a) paternal and b) maternal BMI categories (<25 kg/m2 and 25 kg/m2)

99

Table 4.1 Distribution (mean and standard deviation) of infant WAZ, LAZ, WLZ and BAZ at 12 months of age by selected background factors

N Mean WAZ P Mean LAZ P Mean WLZ P Mean BAZ P (SD) value (SD) value (SD) value (SD) value Maternal age (years) <30 91 –0.79(0.89) 0.99 –1.36(0.90) 0.57 –0.18(0.92) 0.69 0.01(0.94) 0.67 >30 54 –0.79(1.04) –1.45(0.96) –0.11(1.08) 0.08(1.10)

Maternal pre-pregnancy BMI (kg/m2) <25 106 –0.85(0.85) 0.02 –1.39(0.88) 0.28 –0. 24(0.94) 0.40 –0.06(0.96) 0.46 >25 39 –0.62(1.15) –1.40(1.02) 0.09(1.07) 0.28(1.06)

Total GWG Inadequate 42 –1.00(0.87) 0.05 –1.33(1.00) 0.14 –0.00(1.04) 0.15 0.19(1.07) 0.21 Adequate 63 –0.64(0.94) –1.55(0.85) –0.33(1.00) –0.13(0.98) Excess 40 –0.60(1.00) –1.20(0.90) 0.03(0.94) 0.14(0.92)

Paternal BMI (kg/m2) <25 66 –0.79(0.88) 0.26 –1.40(0.95) 0.69 –0.14(0.90) 0.35 0.05(0.91) 0.11 >25 50 –0.62(1.03) –1.34(0.87) 0.04(1.11) 0.22(1.13)

Maternal educational level Primary, secondary 49 –1.05(0.81) 0.03 –1.59(0.88) 0.17 –0.37(0.88) 0.10 –1.16(0.92) 0.14 Diploma 59 –0.74(1.03) –1.33(1.00) –0.12(0.98) 0.05(0.97) Tertiary (undergraduate, 37 –0.52(0.90) –1.24(0.80) 0.09(1.10) 0.27(1.10) postgraduate)

Household monthly income (RM) <2300 61 –0.76(1.00) 0.03 –1.38(0.95) 0.80 –0.13(0.97) 0.01 0.06(0.98) 0.01 2300–5599 70 –0.93(0.88) –1.43(0.90) –0.32(0.96) –0.13(1.00) >5600 14 –0.19(0.80) –1.24(0.88) 0.54(0.82) 0.74(0.83)

Parity 100

Nulliparous 40 –0.56(0.93) 0.08 –1.20(0.87) 0.11 0.00(1.01) 0.24 0.18(1.03) 0.28 >1 105 –0.88(0.94) –1.47(0.92) –0.21(0.97) –0.02(0.98)

Infant sex Female 80 –0.70(0.88) 0.20 –1.20(0.84) 0.01 –0.16(0.94) 0.91 0.00(0.96) 0.64 Male 65 –0.90(1.01) –1.62(.096) –0.14(1.03) 0.08(1.05)

Birth weight categories (kg) <2.5 10 –0.98(0.87) 0.52 –1.61(1.20) 0.35 –0.24(0.74) 0.85 –0.04(0.78) 0.92 2.5–4 132 –0.79(0.95) –1.39(0.88) –0.15(1.01) 0.03(1.02) >4 3 –0.26(0.57) –0.73(1.55) 0.13(0.26) 0.24(0.48)

Maternal walking at 18 weeks Nil days/week 38 –0.90(0.89) 0.20 –1.53(0.96) 0.03 –0.21(0.92) 0.79 –0.01(0.94) 0.88 1–2 days/week 38 –0.89(0.94) –1.46(0.92) –0.24(1.03) –0.05(1.06) 3–4 days/week 19 –0.38(1.22) –0.82(0.94) 0.01(1.34) 0.12(1.35) 5 or more days/week 50 –0.78(0.85) –1.46(0.82) –0.10(0.84) 0.10(0.85)

Maternal moderate activity at 18 weeks Nil days/week 69 –0.81(1.03) 0.60 –1.39(1.02) 0.90 –0.18(1.03) 0.39 –0.00(1.04) 0.38 1–2 days/week 50 –0.81(0.80) –1.50(0.76) –0.12(0.94) 0.09(0.97) 3–4 days/week 13 –0.60(1.02) –1.21(0.89) –0.01(0.89) 0.16(0.86) 5 or more days/week 13 –0.76(0.96) –1.17(0.97) –0.28(1.05) 0.07(0.00)

Maternal physical activity level at 18 weeks Low 91 –0.83(0.96) 0.68 –1.45(0.90) 0.53 –0.16(1.00) 0.89 0.03(1.02) 0.91 Moderate 37 –0.67(0.88) –1.25(0.95) –0.09(0.88) 0.08(0.89) High 17 –0.84(1.01) –1.40(0.93) –0.22(1.10) –0.04(1.13)

101

Maternal calories intake (RNI) at 18 weeks Inadequate 129 –0.78(0.94) 0.63 –1.42(0.94) 0.30 –0.11(0.94) 0.17 0.08(0.95) 0.14 Adequate 16 –0.90(0.97) –1.17(0.68) –0.47(1.22) –0.32(1.26)

Maternal protein intake (RNI) at 18 weeks Inadequate 88 –0.83(0.94) 0.50 –1.54(0.91) 0.01 –0.10(0.95) 0.45 0.10(0.96) 0.30 Adequate 57 –0.72(0.95) –1.16(0.88) –0.22(1.03) –0.07(1.04)

Abbreviation: GWG, gestational weight gain; BMI, body mass index; Mets, metabolic equivalent of task; RNI, recommended nutrient intake

102

Table 4.2 Distribution (mean and standard deviation) of infant skinfold thickness at 12 months of age by parental BMI in two categories.

N Mean P Mean BSF P Mean SSS P Mean SIS P Mean P TSF (SD) value (SD) value (SD) value (SD) value %FM (SD) value Maternal pre-pregnancy BMI (kg/m2) <25 106 7.58(1.53) 0.49 5.50(1.24) 0.87 6.24(1.28) 0.55 5.09(1.25) 0.72 22.55(6.97) 0.15 >25 39 7.42(1.67) 5.34(1.21) 6.51(1.39) 5.18(1.31) 20.70(8.42)

Paternal BMI (kg/m2) <25 66 7.73(1.39) 0.02 5.52(1.26) 0.52 6.63(1.41) 0.99 5.20(1.37) 0.60 22.50(8.22) 0.11 >25 50 7.45(1.89) 5.40(1.37) 6.47(1.42) 5.30(1.27) 22.39(6.62)

BMI, body mass index, Mets, metabolic equivalent of task; RNI, recommended nutrient intake; TSF, triceps skinfold; BSF, biceps skinfold; SSS, subscapular skinfold; SIS, supra-iliac skinfold; FM, fat mass

103

Table 4.3 Association of parental pre-pregnancy BMI and offspring growth factors throughout 12 months

Infant outcomes β (95% CI) WAZ P value LAZ P value WLZ P value BAZ P value Maternal pre-pregnancy BMI (kg/m2) Unadjusted 0.03(–0.00,0.06) 0.050 –0.00(0.03,0.04) 0.953 0.04(0.01,0.07) 0.012 0.04(0.01,0.07) 0.007 Model1 0.03(–0.00,0.06) 0.070 –0.00(0.03,0.04) 0.865 0.04(0.01,0.07) 0.015 0.04(0.01,0.07) 0.009 Model2 0.03(–0.01,0.06) 0.100 0.01(–0.02,0.05) 0.453 0.02(–0.01,0.06) 0.204 0.03(–0.00,0.06) 0.095 Model3 0.03(0.00,0.06) 0.040 0.00(–0.03,0.04) 0.841 0.03(–0.01,0.07) 0.106 0.03(–0.01,0.07) 0.060 Model4 0.03(0.00,0.06) 0.087 0.01(–0.03,0.05) 0.596 0.03(–0.00,0.07) 0.205 0.04(–0.00,0.07) 0.121

Paternal BMI (kg/m2) Unadjusted 0.02(–0.00,0.04) 0.106 0.00(–0.02,0.03) 0.928 0.03(–0.00,0.05) 0.068 0.02(0.00,0.05) 0.047 Model1 0.01(–0.01,0.04) 0.211 –0.00(0.03,0.02) 0.971 0.02(–0.01,0.04) 0.135 0.02(–0.00,0.04) 0.112 Model2 0.01(–0.01,0.04) 0.240 –0.00(0.02,0.02) 0.825 0.01(–0.01,0.04) 0.249 0.02(–0.01,0.04) 0.184 Model3 0.01(–0.01,0.03) 0.400 –0.00(0.03,0.02) 0.798 0.02(–0.01,0.04) 0.219 0.01(–0.01,0.04) 0.216 Model4 0.00(–0.02,0.03) 0.727 –0.01(0.03,0.02) 0.621 0.01(–0.01,0.04) 0.250 0.01(–0.01,0.03) 0.433 BMI, body mass index; WAZ, weight for age z-score; LAZ, length for age z-score, WLZ, weight for length z-score; BAZ, body mass index for age z-score. 1adjusted for maternal education, household monthly income and maternal age at birth; 2adjusted for 1 + maternal physical activity, calories intake at 18 weeks of gestation and prenatal less-healthy dietary pattern 3adjusted for 2 + total gestational weight gain and gestational age 4adjusted for 3 + other parents’ BMI (paternal pre-pregnancy BMI or maternal pre-pregnancy BMI)

104

Discussion We investigated whether parental pre-pregnancy BMI and fat distribution was associated with infant growth status and body composition in early life. We found maternal pre-pregnancy BMI was positively associated with infant growth status independent of parental socio-economic factors. However, this association were not statistically significant after incorporating maternal diet, physical activities during pregnancy, child factors and paternal BMI.

Contribution of parental pre-pregnancy BMI on the child’s increased anthropometric and weight status in early life was reported in previous studies [26-28]. Parental factors may determine offspring growth and body composition at an early age through intrauterine development from mothers [29]. The intrauterine development or over-nutrition hypothesis refers to the risk that develops in the womb throughout pregnancy when mothers are exposed to a high fat diet and consume high-energy dense food which contributes to unhealthy outcomes for the offspring, such as obesity [7, 9]. To support this hypothesis, the effect of maternal pre-pregnancy BMI has been demonstrated and we found a positive association with offspring anthropometrics [15, 30, 31]. Deierlein et al. [31] found that for every unit increase of maternal pre-pregnancy BMI, offspring weight-for-length at 6 months increased 0.02 (95% CI: 0.003–0.05) unit. However, they did not include mother’s lifestyle during pregnancy in their model. Other studies also suggested that maternal pre-pregnancy BMI, weight gained, and body composition have an effect on offspring body composition, such as fat or fat mass [16, 32]. Maternal pre-pregnancy weight played a significant role in infant fat mass deposition [33]. We found a significant association between maternal pre-pregnancy BMI and infant longitudinal growth status including WAZ, WLZ and BAZ. These associations were explained by the dietary intake and physical activities of mothers during pregnancy. We found no significant association between pre-pregnancy BMI and maternal body fat percentage at early pregnancy on infant’s body fat percentage. Similarly, a study by Ay et al. [34] found that parental pre-gravid BMI had no influence on early childhood fat mass. The consideration and contribution of fathers weight status and body composition on offspring growth and body composition have been demonstrated in previous studies [14, 27, 35]. We examined the association of paternal BMI factors on infant anthropometrics outcomes and body composition in difference models. However, the associations we observed were weaker and not significant.

Some studies suggested environmental factors such as parental prenatal lifestyle factors, including food intake, physical activity, and smoking habit may contribute to offspring growth and body fat distribution [36-38]. Our study supported this as we found maternal dietary intake and physical 105

activities during pregnancy influenced the association between maternal BMI and offspring WLZ and BAZ association. This finding is consistent with a previous study about diet and physical activity intervention during pregnancy. The researchers found that maternal lifestyle—consisting of healthy diet, dietary consultation and physical activity at the prenatal stage—may influence offspring anthropometrics during early childhood [39]. Most of these studies only had short-term follow-ups. Therefore, a further longitudinal study is warranted to investigate whether maternal lifestyle during pregnancy, which includes details of dietary consumption, dietary pattern and the intensity of physical activity, may influence offspring anthropometrics in the long term. As for parental smoking habit, this was irrelevant to our findings since, in our study, mothers were non- smokers during pregnancy.

Strengths and limitations Our findings highlight the early-stage environment in which positive association was found between mothers and infants. This association is important as it may predict continuity and become stronger as the children grow older. In accordance with these results, many previous studies have suggested that the direction of association between parental BMI and offspring is present since the early life stage and increases slightly throughout life [40, 41].

The strength of this study was the inclusion of various potential covariates in the models, including parental background and lifestyle factors (e.g. dietary intake, dietary pattern and physical activities of mothers throughout pregnancy). Using these validated data, we were able to estimate the energy in and out during pregnancy and whether that imbalance was associated with offspring growth in early life. Another strength was the availability of data for paternal weight and height that were collected at baseline. Moreover, comprehensive information on body composition such as fat percentage and fat mass were used in this study through BIA. Although obtaining maternal body composition by using singular BIA during pregnancy may be inaccurate due to the influence of the large amount of fluid a woman carries during pregnancy, it is better than self-reporting and was not in a clinical setting [42]. Nonetheless, there were a few issues that needed to be considered. Firstly, the sample size was small, thus the magnitude of the effects size is small and marginally significant, and some women were lost throughout the study. Secondly, pre-pregnancy weight was based on self-reported data which may induce bias. Nevertheless, among 52 women who visited the clinic in the first trimester, high correlation (Pearson’s correlation coefficient = 0.95) was observed between the self-reported and measured weight. This reported data was found to be reliable to replace weight measurement when it was impossible to measure weight itself [43]. Lastly, there was lack of 106

information about parental genetic factors—suggested as contributing to child growth and body composition that leads to OW/OB later in life.

Conclusion We found that maternal pre-pregnancy BMI was prospectively associated with offspring WLZ and BAZ throughout the first 12 months of life. This association was independent of parental socio- economic status, but partly explained by maternal lifestyle factors, including diet and physical activity during pregnancy. Our findings suggest that, compared with mothers with normal pre- pregnancy BMI, mothers with higher pre-pregnancy BMI had offspring at greater risk of higher BAZ and WAZ in early childhood. Future studies with larger sample sizes are needed to replicate this finding in the low- to middle-income countries.

References [1] World Health Organization (WHO). (2015, 30 April). Obesity and overweight, Fact Sheet. Available: http://www.who.int/mediacentre/factsheets/fs311/en/ [2] C. Murrin, G. Kelly, R. Tremblay, and C. Kelleher, "Body mass index and height over three generations: evidence from the Lifeways cross-generational cohort study," BMC Public Health, vol. 12, p. 81, 2012. [3] R. Cooper, E. Hyppönen, D. Berry, and C. Power, "Associations between parental and offspring adiposity up to midlife: the contribution of adult lifestyle factors in the 1958 British Birth Cohort Study," American Journal of Clinical Nutrition, vol. 92, pp. 946-953, 2010. [4] A. A. Mamun, D. A. Lawlor, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Family and early life factors associated with changes in overweight status between ages 5 and 14 years: findings from the Mater University Study of Pregnancy and its outcomes," International Journal of Obesity and Related Metabolic Disorders, vol. 29, pp. 475-482, 2005. [5] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357- 1359, Jun. 2005. [6] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: a 16- year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct. 2011. [7] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater-university

107

study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb 15 2007. [8] G. Davey Smith, C. Steer, S. Leary, and A. Ness, "Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC)," Archives of Disease in Childhood, vol. 92, pp. 876-880, 2007. [9] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child Cohort Study: A Family-based Approach to Studying the Role of the Intrauterine Environment in Childhood Adiposity," American Journal of Epidemiology, vol. 176, pp. 83-92 July 15, 2012 2012. [10] A. M. Siega-Riz, M. Viswanathan, M.-K. Moos, A. Deierlein, S. Mumford, J. Knaack, et al., "A systematic review of outcomes of maternal weight gain according to the Institute of Medicine recommendations: birthweight, fetal growth, and postpartum weight retention," American Journal of Obstetrics and Gynecology, vol. 201, pp. 339.e1-339.e14, 10// 2009. [11] Z. Yu, S. Han, J. Zhu, X. Sun, C. Ji, and X. Guo, "Pre-Pregnancy Body Mass Index in Relation to Infant Birth Weight and Offspring Overweight/Obesity: A Systematic Review and Meta-Analysis," PLoS ONE, vol. 8, p. e61627, 2013. [12] M. Mannan, S. A. R. Doi, and A. A. Mamun, "Association between weight gain during pregnancy and postpartum weight retention and obesity: a bias-adjusted meta-analysis," Nutrition Reviews, vol. 71, pp. 343-352, 2013. [13] J. Baird, D. Fisher, P. Lucas, J. Kleijnen, H. Roberts, and C. Law, "Being big or growing fast: systematic review of size and growth in infancy and later obesity," BMJ, vol. 331, p. 929, 2005-10-20 22:47:46 2005. [14] N. C. Harvey, J. R. Poole, M. K. Javaid, E. M. Dennison, S. Robinson, H. M. Inskip, et al., "Parental Determinants of Neonatal Body Composition," Journal of Clinical Endocrinology & Metabolism, vol. 92, pp. 523-526, February 1, 2007 2007. [15] I. Mesman, T. J. Roseboom, G. J. Bonsel, R. J. Gemke, M. F. van der Wal, and T. G. M. Vrijkotte, "Maternal pre-pregnancy body mass index explains infant's weight and BMI at 14 months: results from a multi-ethnic birth cohort study," Archives of Disease in Childhood, vol. 94, pp. 587-595, Aug. 2009. [16] J. L. Josefson, J. A. Hoffmann, and B. E. Metzger, "Excessive weight gain in women with a normal pre-pregnancy BMI is associated with increased neonatal adiposity," Pediatric Obesity, vol. 8, pp. e33-e36, 2013.

108

[17] S. Loy and J. Hamid Jan, "The Universiti Sains Malaysia Pregnancy Cohort Study: Maternal-infant Adiposity Development until the First Year of Life," Health and the Environment Journal, vol. 5, pp. 50-64, 2014. [18] World Health Organization (WHO), "Obesity: Preventing and Managing a Global Epidemic, Report of a WHO Consultant on Obesity. Geneva: World Health Organization.," 1998. [19] K. Mirnalini, M. S. Zalilah, Y. M. Chan, and A. S. Hazizi, Handbook on Nutritional Assessment Methods. Selangor: August Publishing Sdn. Bhd., 2007. [20] M. de Onis, A. W. Onyango, J. Van den Broeck, W. C. Chumlea, and R. Martorell, "Measurement and standardization protocols for anthropometry used in the construction of a new international growth reference," Food Nutrition Bulletin, vol. 25, pp. S27-36, Mar. 2004. [21] World Health Organization (WHO). (2008, 6 may). World Health Organization Child Growth Standards: Training course on child growth assessment - interpreting growth indicators. Available: http://www.who.int/childgrowth/training/en/ [22] B. Sen, K. Bose, S. Shaikh, and D. Mahalanabis, "Prediction equations for body-fat percentage in Indian infants and young children using skinfold thickness and mid-arm circumference," Journal of Health Population Nutrition, vol. 28, pp. 221-9, Jun. 2010. [23] S. Suzana, M. Y. Noor Aini, S. Nik Shanita, G. Rafidah, and A. Roslina, Atlas of food exchanges and portion sizes, 2nd ed. Kuala Lumpur: MDC Publishers Sdn Bhd (91168-A), 2009. [24] National Coordinating Committee on Food & Nutrition (NCCFN), Recommended nutrient intakes for Malaysia. A report of the Technical Working Group on Nutritional Guidelines. Putrajaya, Malaysia: Ministry of Health Malaysia, 2005. [25] S. L. Loy and J. M. Hamid Jan, "Relative validity of dietary patterns during pregnancy assessed with a food frequency questionnaire," International Journal of Food Sciences and Nutrition, vol. 64, pp. 668-673, 2013/09/01 2013. [26] L. Gomez-Lopez, Van Hulst, A., Barnett, T.A., Roy-Cagnon, M.-H., Tremblay, A., O'Loughlin, J., Lambert, M., "Does parental body mass index status modify the associations among birth weight, early growth and childhood adiposity?," Paediatrics and Child Health (Canada), vol. 18, 2013. [27] B. Durmus, L. R. Arends, L. Ay, A. C. Hokken-Koelega, H. Raat, A. Hofman, et al., "Parental anthropometrics, early growth and the risk of overweight in pre-school children: the Generation R Study," Pediatric Obesity, vol. 8, pp. 339-50, Oct. 2013. [28] U. M. Stamnes Koepp, L. Frost Andersen, K. Dahl-Joergensen, H. Stigum, O. Nass, and W. Nystad, "Maternal pre-pregnant body mass index, maternal weight change and offspring birthweight," Acta Obstetricia et Gynecologica Scandinavica, vol. 91, pp. 243-9, Feb. 2012. 109

[29] E. Oken and M. W. Gillman, "Fetal origins of obesity," Obesity Research, vol. 11, pp. 496- 506, 2003. [30] H. R. Hull, M. K. Dinger, A. W. Knehans, D. M. Thompson, and D. A. Fields, "Impact of maternal body mass index on neonate birthweight and body composition," American Journal of Obstetrics and Gynecology, vol. 198, pp. 416.e1-416.e6, 4// 2008. [31] A. L. Deierlein, A. M. Siega-Riz, L. S. Adair, and A. H. Herring, "Effects of pre-pregnancy body mass index and gestational weight gain on infant anthropometric outcomes," Journal of Pediatrics, vol. 158, pp. 221-226, 2011. [32] M. Thame, C. Osmond, and H. Trotman, "Fetal growth and birth size is associated with maternal anthropometry and body composition," Maternal & Child Nutrition, pp. n/a-n/a, 2013. [33] M. F. Sewell, L. Huston-Presley, D. M. Super, and P. Catalano, "Increased neonatal fat mass, not lean body mass, is associated with maternal obesity," American Journal of Obstetrics and Gynecology, vol. 195, pp. 1100-1103, 10// 2006. [34] L. Ay, C. J. Kruithof, R. Bakker, E. A. P. Steegers, J. C. M. Witteman, H. A. Moll, et al., "Maternal anthropometrics are associated with fetal size in different periods of pregnancy and at birth. The Generation R Study," Bjog-an International Journal of Obstetrics and Gynaecology, vol. 116, pp. 953-963, Jun. 2009. [35] Y. P. Chen, X. M. Xiao, J. Li, C. Reichetzeder, Z. N. Wang, and B. Hocher, "Paternal Body Mass Index (BMI) Is Associated with Offspring Intrauterine Growth in a Gender Dependent Manner," Plos One, vol. 7, May. 2012. [36] V. Burke, Beilin, L.J., Dunbar, D., "Family lifestyle and parental body mass index as predictors of body mass index in Australian children: A longitudinal study," International Journal of Obesity, vol. 25, 2001. [37] B. Durmus, C. J. Kruithof, M. H. Gillman, S. P. Willemsen, A. Hofman, H. Raat, et al., "Parental smoking during pregnancy, early growth, and risk of obesity in preschool children: the Generation R Study," American Journal of Clinical Nutrition, vol. 94, pp. 164-171, Jul 2011. [38] L. Chu, R. Retnakaran, B. Zinman, A. J. G. Hanley, and J. K. Hamilton, "Impact of Maternal Physical Activity and Infant Feeding Practices on Infant Weight Gain and Adiposity," International Journal of Endocrinology, vol. 2012, p. 9, 2012. [39] R. D. Potdar, S. A. Sahariah, M. Gandhi, S. H. Kehoe, N. Brown, H. Sane, et al., "Improving women's diet quality preconceptionally and during gestation: effects on birth weight and prevalence of low birth weight-a randomized controlled efficacy trial in India (Mumbai Maternal Nutrition Project)," American Journal of Clinical Nutrition, vol. 100, pp. 1257-68, Nov. 2014. 110

[40] P. Chivers, H. Parker, M. Bulsara, L. Beilin, and B. Hands, "Parental and early childhood influences on adolescent obesity: a longitudinal study," Early Child Development and Care, vol. 182, pp. 1071-1087, 2013/04/21 2012. [41] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166- 1175, 2011. [42] A. C. Buchholz, C. Bartok, and D. A. Schoeller, "The validity of bioelectrical impedance models in clinical populations," Nutrition and Clinical Practice, vol. 19, pp. 433-46, Oct. 2004. [43] A. A. Mamun, M. Kinarivala, M. J. O'Callaghan, G. M. Williams, J. M. Najman, and L. K. Callaway, "Associations of excess weight gain during pregnancy with long-term maternal overweight and obesity: evidence from 21 y postpartum follow-up," American Journal of Clinical Nutrition, vol. 91, pp. 1336-1341, May. 2010.

111

Chapter 5: Prenatal parental overweight and obesity and the risk of offspring overweight and obesity at 21 years.

5.1: Context The association between parental pre-pregnancy BMI, body composition and offspring growth characteristics in the short term was observed in the previous chapter (Chapter 4.1). Whether parental pre-pregnancy BMI and obesity have influence on the development of adult offspring BMI, WC and the risk of obesity is explored in this chapter using the MUSP birth cohort data. The strength of the effect size of this association was determined after adjustment of covariates that may confound the link. The influence of parental BMI on offspring waist circumference in adulthood was assessed.

In this chapter, a significant association between parental pre-pregnancy BMI and the development of offspring BMI, WC and obesity in adulthood was found. The association was robust after adjusting for other covariates. Therefore, parental BMI and obesity before pregnancy is one of the crucial early life factors that influence the increase risk of adulthood obesity. This chapter forms a manuscript which has been accepted by the Australian and New Zealand Public Health Journal.

112

5.2: Prenatal parental overweight and obesity and the risk of offspring overweight and obesity at 21 years.

N Zalbahar, Najman MJ, McIntyre HD, Mamun AA

Abstract Objectives: To investigate the prospective association between parental pre-pregnancy BMI and adult male and female offspring BMI and waist circumference (WC). Methods: Sub-sample of 2229 parent-offspring pairs with parental pre-pregnancy BMI and offspring BMI and WC at 21 years were used from the MUSP (Mater – University of Queensland Study of Pregnancy cohort). Multivariable results were adjusted for maternal factors around pregnancy (e.g. gestational weight and smoking during pregnancy) and offspring factors in early life (e.g. birth weight) and 14 years (e.g. sports participation and meals with family). Results: After adjustments for confounders, each unit increase in paternal, and maternal BMI, the BMI of young adult offspring increased by 0.33 kg/m2 and 0.35 kg/m2, and the WC increased by 0.76 cm and 0.62 cm, respectively. In the combination of parents’ weight status, offspring at 21 years was six times the risk of being overweight/obese (OW/OB) when both parents were OW/OB, compared to offspring of healthy weight parents. Conclusions: Prenatal parental BMI are independently related to adult offspring BMI and WC. Implications: Prenatal paternal–maternal weight status is an important determinant of offspring weight status in the long term. Further studies are warranted to investigate the underlying mechanisms.

113

Introduction The prevalence of obesity (mostly measured using BMI) has increased from 3.2% to 10.8% for adult men and from 6.4% to 14.9% in adult women in the last four decades [1]. It is a major public health problem in most countries in the world including Australia [2]. There is a high degree of continuity of overweight and obesity across the life course indicating that overweight and obesity in childhood tends to persist into adulthood [3]. Progression in overweight and obesity from childhood to adulthood was associated with multiple factors such as socio-cultural, familial and environmental factors throughout life [4-7]. Parental overweight and obesity status have been consistently found to be one of the strongest predictors of the development of offspring overweight and obesity [8-13]. However, the strength of the association between parental pre-pregnancy BMI categories with adult offspring overweight and obesity, in particular central obesity, is relatively less studied as there are few studies that prospectively collected BMI data from parents before the mother became pregnant and also collected adult offspring’s BMI.

The association of parental overweight and obesity with offspring overweight and obesity may vary by sex [14]. The differences between parental weight status association and male and female offspring may explain whether parental–offspring links are influenced by environmental factors or genetic factors [14]. Results are inconclusive for the association of mother–daughter and father–son overweight or obesity. Some studies reported that the mother–daughter association was stronger than the father–son association [13, 15], others reported the opposite (i.e. father–son association was stronger than mother–daughter [14, 16] or no difference by offspring sex) [17]. However, these findings were limited to childhood and adolescent BMI and had not extended to adulthood [18]. It is well established that overweight and obesity can also be assessed by using WC, which is commonly considered as a better measure of abdominal or central obesity [19]. Sex differences of adult offspring OW/OB status with parental BMI score have not been studied for central obesity.

The aim of this study was to assess the long-term association of parental BMI prior to pregnancy with offspring BMI and WC at 21 years using data from a large community-based

114

birth cohort. We also examined whether parent-offspring associations differed by sex of the offspring.

Methods Study participants We used the MUSP cohort which is a prospective study of 7223 mother–offspring pairs. The sample comprised mothers who received antenatal care at a major public hospital in Brisbane, Australia, between 1981 and 1983 and delivered a live singleton child who was not adopted out before leaving the hospital. Mothers and children have been followed-up prospectively with maternal questionnaires being administered at the first antenatal clinic visit, 3–5 days post-delivery, at six months, and 5, 14 and 21 years after birth [20]. Ethical approval was obtained from the ethics committees at The University of Queensland and the Mater Hospital at baseline data collection and at all subsequent follow-ups. Mothers provided a signed consent form for themselves and for their index offspring. At 21 years, offspring provided their own consent. Our analytical sample comprised 2229 mother–offspring pairs from whom we collected pre-pregnancy parental BMI and measured offspring BMI and WC and WHR at 21 years.

Measurements Parental-pregnancy BMI The main exposure was parental BMI before pregnancy. At the first antenatal clinic visit, women were asked to report their pre-pregnancy weight; women were also weighed at the clinic. There was high correlation between these two measures (Pearson’s correlation coefficient = 0.95). Women’s height was measured at the first clinic visit (on average 18 weeks of gestation). Paternal weight and height were self-reported by the mother. Both parental BMIs were categorised using WHO guidelines [21]: underweight (BMI <18.5 kg/m2), normal weight (BMI 18.5–24.9kg/m2), overweight (BMI 25.0–29.9 kg/m2) and obese (BMI >30.0 kg/m2). These were then collapsed into BMI <25.0 kg/m2 ‘normal and underweight’ (HW) and BMI >25.0 kg/m2 ‘OW/OB’ group for further analysis. To further determine the combination association of parental BMI status on adult offspring OW/OB, we categorised parental BMI status into four combinations: Both parents HW; OW/OB father, HW mother; OW/OB mother, HW father and both parents OW/OB.

115

Offspring anthropometry At the 21-year follow-up, offspring’s height without shoes was measured to the nearest centimetre, using a portable stadiometer. Their weight was measured in light clothing with a scale accurate to 0.2 kg. Two measures of weight were taken with a five-minute interval between assessments, and the mean of these two measures was used for all analyses. Waist and hip circumference was measured at 21 years. At 21 years, offspring’s BMI was categorised into four categories [21] and collapsed into HW and OW/OB. WC was measured horizontally using a measuring tape against the skin, in line with the participant’s umbilicus. As for the hip circumference, the participant was asked to lower the tape down to the hips and record the measurement at the point of greatest width. Both waist and hip measurements were taken twice, and an average of the measures was recorded. Waist circumference was categorised for males as follows: <94 cm normal, 94 to <102 cm high (overweight) and >102 cm very high (obesity), while for females as follows: <80 cm normal, 80 to <88cm high (overweight) and >88 cm very high (obese) [22]. Similarly, for the analysis, we collapsed the three WC categories into two categories: HW and OW/OB.

Covariates The covariates were paternal and maternal educational attainment obtained from a questionnaire administered at study initiation (classified from seven categories and collapsed into three categories: incomplete education (no school, complete primary or incomplete secondary school), completed secondary education (grade 10 and 12) and completed further/higher education (college, university or other higher advanced education institution), family annual income (mother reported family income in response to a question with seven categories: A$0–2599; A$2600–5199; A$5200–10,399; A$10,400–15,599; A$15,600– 20,799; A$20,800–25,999; and more than A$26,000, which later we collapsed it into two categories (A$10,400), maternal age at first clinit visit (in years), maternal smoking habits during pregnancy (current smoker and non-smoker) and gestational age (preterm gestational age <37 weeks and term). Lastly, total maternal weight gained in pregnancy was considered as a continuous variable. Maternal gestational weight gain was derived by subtracting pre-pregnancy maternal weight from the maximum measured weight in pregnancy. Maternal maximum measured weight in pregnancy was obtained from the obstetric data.

116

For offspring, early life factors such as sex (male, female), birth weight (measured in grams), breastfeeding duration and lifestyle (including sport participation, time spent watching TV, eating together with family and fast food consumption) at the 14-year follow-up were used. At the 6-month follow-up, duration of breastfeeding was collected using a question with the following response options: never, 2 weeks or less, 3–6 weeks, 7 weeks–3 months, 4–6 months, and still breastfeeding. Due to small numbers in the categories 2 weeks or less, 3–6 weeks and 7 weeks–3 months, breastfeeding categories were collapsed into three categories (never, less than 4 months and more than 4 months). Offspring lifestyle at 14 years was obtained through a questionnaire. Hours of TV watching per day were based on six response options (never watched TV, less than 1 hour, 1 to less than 3 hours, 3 to less than 5 hours, 5 to less than 7 hours, and 7 hours or more) and categorised into less than 3 hours and more than 3 hours daily. The number of days of participation in sports in every week were obtained with five response options (Not at all, 1 day, 2/3 days, 4/5 days and 6/7 days) and classified into less than 4 days and more than 4 days per week. Mothers responded as to whether family ate together with the following options: at least once a day, a few times a week, about once a week, and less than once a week. These were further collapsed into two categories (not daily and at least once a day). Lastly, how often they ate fast food (daily, 2–3 times per week and once per week or less) [23].

Statistical analysis Statistical analysis was carried out using Stata version 13.0 (Stata Inc, College Station, TX). All continuous data are reported as mean and standard deviation and categorical data are reported as percentages. We used chi square test to compare the percentages or proportions and F test to compare mean values between HW and OW/OB. Multiple linear regression was used to examine the association between parental BMI with offspring BMI and WC at 21 years. A series of multiple linear regression models was used to examine the association between parental BMI and offspring BMI and WC at 21 years. In the first model, we have adjusted for the other parent’s BMI and offspring’s sex. In the second model, we have added maternal factors around pregnancy (family annual income, maternal educational attainment, maternal age at birth, smoking during pregnancy and gestational weight gain) and paternal educational attainment. In the third model, we have additionally adjusted for offspring early life factors (gestational age, offspring’s birth weight and breastfeeding duration). In the final model, the lifestyle factors of offspring when they were 14 years old were added. The main 117

purpose of this modelling strategy was to examine whether maternal factors around pregnancy or offspring factors at 14 years explain the parent–offspring BMI long-term association. Interactions between parental BMI and offspring’s sex were tested. Similarly, a series of logistic regressions were used to evaluate the independent association of paternal, maternal and parental combination of BMI categories with offspring BMI or WC in categories at 21 years adjusting for the maternal and offspring factors. We computed an F statistic to formally compare the adjusted coefficients between maternal and paternal associations. All results were interpreted as regression coefficients or OR with 95% confidence intervals.

Results Of 7223 participants, a sub-sample of 2229 (50.6% female offspring) provided the information on pre-pregnancy parental BMI, offspring’s BMI, and WC at 21 years. The participants who provided information on parental BMI and offspring anthropometry at 21 years had relatively higher socio-economic status, were older mothers, and were non-smokers during pregnancy, compared to participants who were excluded from the analyses. However, our analytical sample was not substantially different from the excluded sample by parental pre-pregnancy anthropometry and offspring anthropometric characteristics (Table 5.1).

Pre-pregnancy mean BMI of mothers and fathers was 23.7 (3.3) kg/m2 and 22.0 (3.9) kg/m2, respectively. Mean offspring BMI and WC at 21 years were 24.2 (4.9) kg/m2 [24.2 (4.4) for males; 24.2 (5.3) for females] and 81.6 (12.3) cm [85.2 (11.2) for males, 77.9 (12.2) for females], respectively. We plotted the mean BMI and WC of offspring at 21 years by maternal and paternal pre-pregnancy BMI. We found mean BMI and WC of offspring at 21 years were linearly associated with maternal and paternal pre-pregnancy BMI (Figure 5.1).

Table 5.2 shows the distribution of offspring BMI and WC at 21 years in categories by parental and offspring factors. A total of 16.3% mothers and 28.5% fathers were OW/OB before pregnancy. One third of offspring at 21 years were OW/OB based on BMI and 27.1% based on WC. As expected both maternal and paternal BMI categories were associated with offspring BMI and WC categories at 21 years (all P-value <0.001). Among all the maternal factors, maternal smoking during pregnancy was consistently associated with offspring BMI as well as WC categories. Among offspring factors, only sex was associated with both BMI 118

and WC of the offspring; sports participation was associated with offspring WC (P-value <0.001) and birth weight with BMI.

In Table 5.3, we have presented unadjusted and adjusted models of associations between parent–offpsring BMI, WC and their categories. In the unadjusted regression models, for every one unit increase in paternal BMI before pregnancy, the BMI of the offspring increased by 0.36 kg/m2 (95% CI: 0.30–0.42) at 21 years as compared to 0.35 kg/m2 (95% CI: 0.31– 0.40) for maternal BMI (Table 3). Similarly, for one unit increase in paternal BMI, offspring WC increased by 0.78 cm (95% CI: 0.62–0.93) and this increment was 0.73 cm (95% CI: 0.61–0.86) for maternal BMI. The magnitude and direction of these associations did not attenuate substantially adjusting for a range of parental and offspring factors. The odds of being OW/OB at 21 years was 2.36 (95% CI:1.69–3.29) times greater for offspring having OW/OB pre-pregnancy mothers compared to those offspring with HW mothers; and it was 2.39 (95% CI:1.82–3.14) times greater for having OW/OB fathers than healthy-weight fathers.A similar direction and strength of association were observed with WC categories (Table 5.3). The odds increased up to six times of becoming OW/OB during young adulthood in the fully adjusted model when both father and mother were OW/OB compared to both HW parents. A similar direction and strength of associations was observed for the WC categories. These associations were consistent for adjustment of parental and offspring factors. Although the association of maternal–offspring was slightly stronger than paternal–offspring, the direction and strength of associations were not substantially different (all P-values >0.001, data not shown).

We have repeated all multivariable regression analyses to examine whether paternal– offspring and maternal–offspring associations differed by offspring sex (Table 5.4). In the fully adjusted models, we found mother–daughter, mother–son, father–son and father– daughter associations were all statistically significant (all P-values<0.05). However, there was no differential link between paternal–offspring sex associations and maternal–offspring sex associations.

119

Table 5.1 Comparison of the characteristics of participants who provided information on pre- pregnancy parental BMI, offspring BMI and WC at 21 years versus the participants who were excluded from the analyses

Covariates Incomplete Complete information *P-value information on pre- on pre-pregnancy pregnancy parental parental BMI, BMI, offspring BMI at offspring BMI at 21 21 years and offspring years and offspring waist circumference (n waist circumference = 4994) (n= 2229)

Fathers BMI (kg/m2) 23.6 (3.7) 23.7 (3.3) 0.456 Mothers pre-pregnancy BMI 21.8 (4.1) 22.0 (3.9) 0.439 (kg/m2) Maternal total gestational 14.7 (5.7) 14.6 (5.4) 0.562 weight gained (kg) Offspring BMI at 21 (kg/m2) 24.7 (5.5) 24.2 (4.9) 0.053 Offspring birth weight (g) 3378.8 (526.5) 3401.0 (500.8) 0.096

Family annual income at baseline Less than A$10399 1698 (36.9) 610 (28.5) More than A$10400 2908 (63.1) 1533 (71.5) <0.001 Maternal Education at birth incomplete higher education 964 (19.5) 341 (15.3) Completed higher education 3196 (64.5) 1414 (63.7) Post high 791 (16.0) 465 (21.0) <0.001 Mother’s age at birth (year) 25.2 (5.1) 26.1 (5.0) <0.001 13-19 238 (4.8) 45 (2.0) 20-34 4447 (89.0) 2009 (90.2) > 34 311 (6.2) 173 (7.8) <0.001 Maternal smoking habit during pregnancy Nil Smoker 2771 (60.3) 1358 (65.5) Smoker 1823 (39.7) 714 (34.5) <0.001

Data are presented as number (%), mean (standard deviation). * We used F test and Chi square for comparing percentages/proportion and mean differences between incomplete and complete data sample.

120

Table 5.2: Parental and offspring factors by offspring BMI and WC at 21 years (n = 222)

Characteristics Body Mass Index (kg/m2) Waist circumference (cm) Parental factors Normal OW/OB *P-value Normal OW/OB *P-value N (%) 1480 (66.5) 747 (33.5) <0.001 1623 (72.9) 604 (27.1) <0.001 Father’s BMI Healthy weight (<25 kg/m2) 1142 (77.2) 461 (61.7) <0.001 1229 (75.7) 374 (61.9) <0.001 Overweight/obese 338 (22.8) 286 (38.4) 394 (24.3) 230 (38.1) Mother’s BMI Healthy weight (<25 kg/m2) 1314 (88.8) 561 (75.1) <0.001 1423 (87.7) 452 (74.8) <0.001 Overweight/obese 166 (11.2) 186 (24.9) 200 (12.3) 152(25.2) Parents’ weight status combination Both parents HW 1026 (69.3) 373 (49.9) <0.001 1094 (67.3) 305 (50.5) <0.001 OW/OB father, HW mother 289 (19.5) 188 (25.2) 330 (20.3) 147 (24.3) HW father, OW/OB mother 117 (7.9) 88 (11.8) 136 (8.4) 69 (11.4) Both parents OW/OB 49 (3.3) 99 (13.1) 64 (3.9) 84 (13.7) Family annual income at FCV Less than A$10400 397 (27.9) 213 (29.5) 0.483 427 (27.4) 183 (31.1) 0.094 More than A$10400 1023 (72.1) 509 (70.5) 1130 (72.6) 403 (68.8) Maternal education at FCV Did not complete secondary school 222 (15.1) 119 (16.0) 0.524 236 (14.6) 105 (17.4) 0.239 Completed secondary school 933 (63.3) 480 (64.4) 1042 (64.5) 371 (61.5) Completed further/higher education 318 (21.6) 147 (19.6) 337 (20.9) 127 (21.1) Paternal education at FCV Did not complete secondary school 265 (18.1) 149 (20.2) 0.437 279 (17.3) 135 (22.7) 0.002 Completed secondary school 865 (58.8) 429 (58.1) 942 (58.6) 349 (58.7) Completed further/higher education 339 (23.1) 160 (21.7) 388 (24.1) 111 (18.7) Maternal smoking habit during pregnancy Nil smoker 922 (67.6) 434 (61.5) 0.005 1016 (67.6) 340 (59.9) <0.001 Smoker 442 (32.4) 272 (38.5) 486 (32.4) 228 (40.1) Maternal age at birth 26.2 (5.0) 25.9 (5.1) 0.148 26.2 (5.0) 25.8 (4.9) 0.167 Total weight gained (kg) 14.6 (5.1) 14.8 (5.9) 0.370 14.6 (5.2) 14.7 (5.8) 0.568

121

Offspring factors Sex Male 728 (49.1) 386 (51.6) <0.001 887 (54.6) 227 (37.5) <0.001 Female 753 (50.9) 361 (48.4) 737 (45.4) 377 (62.5) Breastfeeding duration Less than 4 months 763 (53.1) 410 (57.1) 0.178 841 (53.4) 332 (57.2) 0.140 4–6 months 160 (11.1) 78 (10.9) 171 (10.9) 67 (11.6) 6 months 515 (35.8) 230 (32.0) 564 (35.8) 181 (31.2) Meal time at 14 years Not daily 302 (21.4) 152 (21.5) 0.959 327 (21.1) 127 (22.2) 0.583 At least once a day 1109 (78.6) 555 (78.5) 1220 (78.9) 444 (77.8) Watching TV weekdays at 14 years Less than 3 hours per day 520 (37.0) 247 (35.0) 0.367 572 (37.2) 195 (34.2) 0.197 More than 3 hours per day 884 (63.0) 458 (65.0) 966 (62.8) 376 (65.9) Sports at 14y Less than 3 days per week 696 (49.5) 366 (51.8) 0.303 728 (47.2) 334 (58.5) <0.001 More than 3 days per week 711 (50.5) 340 (48.2) 814 (52.8) 237 (41.5) Fast Food Once a day 32 (3.2) 19 (4.2) 0.493 39 (3.6) 12 (3.3) 0.258 2/3 times a week 249 (25.1) 105 (22.9) 276 (25.4) 78 (21.3) Once a week or less 713 (71.7) 334 (72.9) 771 (71.0) 276 (75.4) Birth weight (g) 3383.9 (487.8) 3435.4 (524.6) 0.022 3396.5 (489.4) 3413.9 (530.8) 0.465 Data are presented as number (%), mean (standard deviation). Abbreviation: OW/OB, overweight and obese; NW/UW, normal and underweight; BMI, body mass index. * We used t-test and chi square for comparing percentages/proportions and mean differences by offspring BMI and WC in two categories (HW and OW/OB)

122

Table 5.3 Unadjusted and adjusted regression coefficient and odds ratios between paternal, maternal BMI and parental weight status with offspring outcomes at 21 years (n = 2229)

BMI (kg/m2) Waist circumference (cm) OW/OB OW/OB β 95% CI OR 95% CI β 95% CI OR 95% CI Paternal BMI Unadjusted Model 0.36 0.30–0.42 2.11 1.74–2.55 0.78 0.62–0.93 1.93 1.58-2.35 Model 1 0.29 0.24–0.35 1.93 1.60–2.36 0.64 0.63–0.93 1.84 1.49-2.56 Model 2 0.36 0.29–0.42 2.05 1.66–2.53 0.77 0.61–0.92 2.04 1.63-2.55 Model 3 0.36 0.29–0.42 2.09 1.69–2.59 0.76 0.61–0.92 2.07 1.65-2.61 Model 4 0.36 0.28–0.43 2.39 1.82–3.14 0.76 0.45–0.79 2.23 1.66-2.98

Maternal BMI Unadjusted Model 0.35 0.31–0.40 2.64 2.09–3.33 0.73 0.61–0.86 2.41 1.90-3.05 Model 1 0.31 0.26–0.36 2.41 1.91–3.05 0.63 0.51–0.76 2.31 1.81-2.95 Model 2 0.38 0.32–0.43 2.57 2.00–3.32 0.74 0.60–0.87 2.29 1.75-2.99 Model 3 0.37 0.32–0.43 2.51 1.93–3.26 0.72 0.58–0.85 2.25 1.71-2.96 Model 4 0.34 0.27–0.41 2.36 1.69–3.29 0.62 0.45–0.79 1.96 1.38-2.80

Parents’ weight status combination Final Model** Both parents HW Reference OW/OB father, HW mother 2.38 1.76–3.22 2.25 1.62–3.12 HW father, OW/OB mother 2.33 1.54–3.51 2.00 1.28–3.12 Both parents OW/OB 5.73 3.34–9.83 4.28 2.48–7.38

BMI, body mass index; WC, waist circumference; NW/UW, normalweight, and underweight; OW/OB, overweight and obese Model 1: Adjusted for other parent’s BMI and offspring’s sex Model 2: Adjusted for maternal factors (gestational weight gain, maternal age at birth, maternal smoking during pregnancy), maternal and paternal education attainment, annual family income Model 3: Model 2 + offspring early life factors (gestational age, birth weight and breastfeeding duration Model 4: Model 3 + offspring lifestyle factors at 14 years (TV watching, sports participation, family eats together and fast food consumption) **Final model (Model 4)

123

Table 5.4: Associations of parental BMI and offspring BMI and WC by sex at 21 years, male (n =1114) and female (n = 1115)

BMI WC Son Daughter Son Daughter β 95% CI β 95% CI *P-value Β 95% CI β 95% CI *P-value Paternal Unadjusted 0.30 0.23–0.38 0.41 0.32–0.50 0.091 0.73 0.53–0.93 0.80 0.59–1.02 0.608 Model 1 0.28 0.20–0.35 0.30 0.21–0.39 0.334 0.67 0.46–0.86 0.59 0.38–0.81 0.781 Model 2 0.31 0.23–0.39 0.39 0.29–0.49 0.261 0.73 0.51–0.94 0.79 0.57–1.01 0.903 Model 3 0.32 0.23–0.40 0.39 0.29–0.48 0.370 0.73 0.51–0.94 0.78 0.56–1.00 0.977 Model 4 0.31 0.22–0.41 0.43 0.32–0.55 0.074 0.67 0.40–0.93 0.86 0.59–1.13 0.647

Maternal Unadjusted 0.25 0.19–0.32 0.47 0.39–0.54 0.000 0.57 0.41–0.74 0.89 0.71–1.07 0.011 Model 1 0.23 0.17–0.29 0.41 0.33–0.48 0.000 0.52 0.36–0.68 0.77 0.58–0.95 0.051 Model 2 0.30 0.22–0.37 0.44 0.35–0.52 0.015 0.62 0.43–0.81 0.83 0.64–1.02 0.180 Model 3 0.29 0.22–0.37 0.43 0.35–0.51 0.017 0.59 0.39–0.78 0.81 0.61–1.00 0.160 Model 4 0.31 0.22–0.41 0.34 0.24–0.44 0.493 0.61 0.37–0.86 0.58 0.34–0.82 0.713

*P-value, test for interaction of parental pre-pregnancy BMI and offspring’s sex Model 1: Adjusted for other parent’s BMI Model 2: Adjusted for maternal factors (gestational weight gain, maternal age at birth, maternal smoking during pregnancy), maternal and paternal education attainment, annual family income Model 3: Model 2 + offspring early life factors (gestational age, birth weight and breastfeeding duration) Model 4: Model 3 + offspring lifestyle factors at 14 years (TV watching, sports participation, family eats together and fast food consumption)

124

40

35

30

25

Offspring21 atBMI years 20 10 20 30 40 Parental pre-pregnancy BMI

paternal-offspring BMI Maternal-offspring BMI Linear Linear

Figure 5.1 Distribution of mean offspring BMI at 21 years by prenatal parental BMI (n = 2229)

125

Discussion Our findings demonstrate that parental BMI before pregnancy is positively associated with adult offspring BMI, WC and its categories. This association was independent of a range of maternal factors around pregnancy, offspring factors in early life and offspring lifestyle at 14 years (sports participation, TV watching, eating together with family and frequency of fast food consumption). The risk was stronger when both parents were OW/OB compared to one parent. Parent–offspring associations did not differ by sex of the offspring.

Overall, our findings are consistent with previous studies where parental weight status was found to be associated with the development of offspring OW/OB [9,13,24,25]. A significant association between parents’ weight status and offspring’s weight status in early life is evident. Parents’ weight status before or during pregnancy has been previously reported as having an impact on their offspring being OW/OB in childhood and adolescence [11,25]. Accordingly, the risk of offspring being OW/OB in childhood was stronger when both parents were OW/OB [9]. Our study further suggests that parental weight status before pregnancy continues to contribute significantly to their offspring’s weight status up to adulthood. Some previous studies suggest that maternal BMI is more strongly associated with male and female adult offspring weight status [26-28]. In contrast, our study provides evidence that both parents’ BMI is associated with adult offspring weight status with no statistical differences between male and female offspring. Our findings add to the literature, showing that both parents BMI before pregnancy may have an impact on their child in the long term, regardless of their child’s age and sex.

Parental OW/OB impact on offspring weight status may be explained by genetic, epigenetic, or environmental factors. Although previous studies suggested that parental–offspring overweight or obesity links may contribute through environmental factors, intrauterine over-nutrition [25,29,30] has been identified as a possible maternal specific mechanism for this association. Some studies reported that one parent, particularly mothers, have a stronger link than fathers [11,12,31]. It is suggested that maternal excess of glucose, amino acids, and free fatty acids transmitted to the baby in utero may impact on the development of the physical and metabolic function in the offspring. In addition, mothers who are OW/OB before pregnancy have high fat stores, high glucose and insulin resistance that could affect the child in the womb and result in increased fat in the infant and

126

increase the risk of a number of other diseases [32-34]. More recent literature has suggested that fathers who are OW/OB before or during pregnancy may also contribute to their child’s risk of being OW/OB, and other health outcomes [35-38]. In contrast to these earlier findings, our study showed no differences in effect size between fathers and mothers on offspring BMI at 21 years after controlling for all potential confounding factors. Further, the odds of the child being OW/OB in young adulthood were the same when either one of their parents (mother or father) were OW/OB. Our findings are consistent with previous reports from longitudinal studies such as the Northern Finland Birth Cohort [13], and Avon Longitudinal Study Parents and Children (ALSPAC) [24]. Both of these reported that parents’ weight status was a stronger predictor of long-term offspring weight status

Parents may influence their child’s weight through shared familial environments [23, 39]. The present study also raises the possibility that parents’ weight status may also affect their offspring weight status beyond familial shared factors. Multiple mechanisms have been suggested to explain this association other than the influence of parental environmental factors. We did manage to control for maternal smoking habit during pregnancy, parental education attainment, and family annual income. Although we do not have parental and offspring dietary patterns and physical activity at early stages of the MUSP study to consider these possible associations with the outcomes, we do have information on the broader perspective of offspring lifestyle which includes eating together with family at the 14-year follow-up, offspring’s fast food consumption, sports participation and time spent watching TV.

Abdominal obesity is a strong predictor of all non-communicable chronic diseases [40]. Previous evidence does suggest that higher maternal BMI is associated with greater offspring waist circumference at age 30 [41]. With limited studies available on parental BMI/ offspring WC associations up to adulthood, our study also highlights the importance of maintaining healthy weight status for both parents around pregnancy to reduce the risk of their adult offspring having abdominal obesity. We did find a significant association between both maternal and paternal weight status and a higher waist circumference of their young adult offspring.

127

This study, however, does have several limitations. There was high attrition from baseline to the later stages of the study due to the long-term follow-up. Therefore, the data may not be representative of the population. The response rate at the six-month follow-up was 93%; at five- year follow-up 73%; at 14-year follow-up 72%, and at 21-year follow-up 57%. In general, participants lost to follow-up were more likely to be less educated, depressed, and smokers [42]. Attrition bias may have affected the sub-groups who had data included in the analysis. The mother reported both her own and the father’s weight and height; this may introduce bias. However, the correlation between maternal reported pre-pregnancy weight and actual weight at the first antenatal visit was very high (0.95) [43]. Multiple imputation and inverse-probability weighting were used to examine potential attrition bias in the MUSP and we found missing values did not make major differences to our conclusions [20, 42].

As for the environmental factors of parents and child, there were suggestions that parental and child dietary factors may confound and contribute to the weight status of the child [23,44,45]. From our dataset, we were able to control for offspring lifestyle factors at 14 years (such as sports, screen time, family mealtime and fast food consumption) which may influence offspring weight at age 21 years. However, dietary factors such as meal patterns, caloric intake and physical activity levels may need to be included in future studies to determine that influence the parent–offspring association has on weight status.

Conclusion Pre-pregnancy parental BMI, both maternal and paternal, is prospectively associated with adult offspring BMI and WC status in the long term. Our findings show that this association is independent of maternal factors around pregnancy (gestational weight gain, maternal age at birth, education attainment, annual income and smoking during pregnancy) and offspring factors in early life and adolescent period (gestational age, breastfeeding duration, birth weight, sex; and at the 14- year follow-up: sports participation, TV watching hours, family eating together and fast food). Maintaining healthy parental weight status before pregnancy appears important. Strategies to improve lifestyle among parents and families with overweight and obesity may provide a means of improving healthy lifestyles and potentially overall health outcomes in the long term.

128

References: [1] N. R. F. C. (NCD-RisC). "Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19.2 million participants," The Lancet, vol. 387, pp. 1377-1396, 2016. [2] World Health Organization (WHO). (2015, 30 April). Obesity and overweight, Fact Sheet. Available: http://www.who.int/mediacentre/factsheets/fs311/en/ [3] A. S. Singh, C. Mulder, J. W. R. Twisk, W. Van Mechelen, and M. J. M. Chinapaw, "Tracking of childhood overweight into adulthood: a systematic review of the literature," Obesity Reviews, vol. 9, pp. 474-488, 2008. [4] S. F. Weng, S. A. Redsell, J. A. Swift, M. Yang, and C. P. Glazebrook, "Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy," Archives of Diseases in Childhood, vol. 97, pp. 1019-1026, Oct 29 2012. [5] P. M. Catalano and H. M. Ehrenberg, "Review article: The short- and long-term implications of maternal obesity on the mother and her offspring," BJOG: An International Journal of Obstetrics & Gynaecology, vol. 113, pp. 1126-1133, 2006. [6] T. D. Brisbois, A. P. Farmer, and L. J. McCargar, "Early markers of adult obesity: a review," Obesity Reviews, vol. 13, pp. 347-367, Apr 2012. [7] K. Silventoinen, B. Rokholm, Kaprio, J., and T. I. A. Sorensen, "The genetic and environmental influences on childhood obesity: a systematic review of twin and adoption studies," International Journal of Obesity, vol. 34, pp. 29-40, 2009. [8] M. Kanciruk, J. J. Andrews, and T. Donnon, "Family history of obesity and risk of childhood overweight and obesity: A meta-analysis," International Journal of Medicine Health Pharmacology and Biomedical Engineering, vol. 8, pp. 244-56, 2014. [9] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357-1359, Jun 2005. [10] T. Parsons, C. Power, S. Logan, and C. Summerbell, "Childhood predictors of adult obesity: a systematic review," International Journal of Obestatric Related Metabolic Disorders, vol. 23, pp. S1 - S107, 1999. [11] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child

129

Cohort Study: A Family-based Approach to Studying the Role of the Intrauterine Environment in Childhood Adiposity," American Journal of Epidemiology, vol. 176, pp. 83- 92 July 15, 2012 2012. [12] C. Murrin, G. Kelly, R. Tremblay, and C. Kelleher, "Body mass index and height over three generations: evidence from the Lifeways cross-generational cohort study," BMC Public Health, vol. 12, p. 81, 2012. [13] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: a 16-year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct 2011. [14] E. M. Perez-Pastor, B. S. Metcalf, J. Hosking, A. N. Jeffery, L. D. Voss, and T. J. Wilkin, "Assortative weight gain in mother-daughter and father-son pairs: an emerging source of childhood obesity. Longitudinal study of trios (EarlyBird 43)," International Journal of Obesity, vol. 33, pp. 727-735, Jul 2009. [15] M. Mostazir, Jeffery, A., Voss, L., Wilkin, T.J., "Gender-assortative waist circumference in mother-daughter and father-son pairs, and its implications. An 11-year longitudinal study in children (EarlyBird 59)," Pediatric Obesity, vol. 9, 2014. [16] V. Burke, Beilin, L.J., Dunbar, D., "Family lifestyle and parental body mass index as predictors of body mass index in Australian children: A longitudinal study," International Journal of Obesity, vol. 25, 2001. [17] V. Svensson, J. A. Jacobsson, R. Fredriksson, P. Danielsson, T. Sobko, H. B. Schiöth, et al., "Associations between severity of obesity in childhood and adolescence, obesity onset and parental BMI: A longitudinal cohort study," International Journal of Obesity, vol. 35, pp. 46-52, 2011. [18] A. Magarey, L. Daniels, T. Boulton, and R. Cockington, "Predicting obesity in early adulthood from childhood and parental obesity," International Journal of Obestatric Related Metabolic Disorders, vol. 27, pp. 505 - 513, 2003. [19] S. Zhu, Z. Wang, S. Heshka, M. Heo, M. Faith, and S. Heymsfield, "Waist circumference and obesity-associated risk factors among whites in the third National Health and Nutrition Examination Survey: clinical action thresholds," American Journal of Clinical Nutrition, vol. 76, pp. 743 - 749, 2002.

130

[20] J. M. Najman, R. Alati, W. Bor, A. Clavarino, A. Mamun, J. J. McGrath, et al., "Cohort profile update: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology vol. 44, pp. 78-78f, December 16, 2014 2014. [21] World Health Organization (WHO), "Obesity: Preventing and Managing a Global Epidemic, Report of a WHO Consultant on Obesity. Geneva: World Health Organization.," 1998. [22] WHO, "Obesity: Preventing and Managing The Global Epidemic. Report of a WHO Consultation," 2000. [23] S. Babajafari, G. C. Marks, A. A. Mamun, M. J. O'Callaghan, and J. M. Najman, "Family food behaviours and adolescents' overweight status: a mother-offspring link study," Iran Red Crescent Medical Journal, vol. 13, pp. 783-94, Nov 2011. [24] G. Davey Smith, C. Steer, S. Leary, and A. Ness, "Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC)," Archives of Disease in Childhood, vol. 92, pp. 876-880, 2007. [25] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater- university study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb 15 2007. [26] J. Laitinen, C. Power, and M. Järvelin, "Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity," American Journal of Clinical Nutrition, vol. 74, pp. 287-294, 2001. [27] A. L. L. Tequeanes, D. P. Gigante, M. C. F. Assuncao, D. A. G. Chica, and B. L. Horta, "Maternal Anthropometry Is Associated with the Body Mass Index and Waist:Height Ratio of Offspring at 23 Years of Age," Journal of Nutrition, vol. 139, pp. 750-754, Apr 2009. [28] R. Cooper, E. Hyppönen, D. Berry, and C. Power, "Associations between parental and offspring adiposity up to midlife: the contribution of adult lifestyle factors in the 1958 British Birth Cohort Study," American Journal of Clinical Nutrition, vol. 92, pp. 946-953, 2010. [29] D. A. Lawlor, A. Hübinette, P. Tynelius, D. A. Leon, G. D. Smith, and F. Rasmussen, "Associations of gestational age and intrauterine growth with systolic blood pressure in a family-based study of 386,485 men in 331,089 families," Circulation, vol. 115, pp. 562-568, 2007.

131

[30] P. D. Taylor and L. Poston, "Developmental programming of obesity in mammals," Experimental Physiology, vol. 92, pp. 287-298, March 1, 2007 2007. [31] D. A. Lawlor and N. Chaturvedi, "Treatment and prevention of obesity—are there critical periods for intervention?," International Journal of Epidemiology, vol. 35, pp. 3-9, February 1, 2006. [32] R. C. Whitaker and W. H. Dietz, "Role of the prenatal environment in the development of obesity," Journal of Pediatrics, vol. 132, pp. 768-776, 5// 1998. [33] M. Alfaradhi and S. Ozanne, "Developmental programming in response to maternal over- nutrition," Frontiers in Genetics, vol. 2, 2011-June-3 2011. [34] E. Oken and M. W. Gillman, "Fetal origins of obesity," Obesity Research, vol. 11, pp. 496- 506, 2003. [35] A. Soubry, S. K. Murphy, F. Wang, Z. Huang, A. C. Vidal, B. F. Fuemmeler, et al., "Newborns of obese parents have altered DNA methylation patterns at imprinted genes," International Journal of Obesity (London), Oct 25 2013. [36] N. O. McPherson, T. Fullston, R. J. Aitken, and M. Lane, "Paternal obesity, interventions, and mechanistic pathways to impaired health in offspring," Annals of Nutrition and Metabolism, vol. 64, pp. 231-238, 2014. [37] Y. Slyvka, Y. Zhang, and F. Nowak, "Epigenetic effects of paternal diet on offspring: emphasis on obesity," Endocrine, pp. 1-11, 2014/07/06 2014. [38] Y. P. Chen, X. M. Xiao, J. Li, C. Reichetzeder, Z. N. Wang, and B. Hocher, "Paternal Body Mass Index (BMI) Is Associated with Offspring Intrauterine Growth in a Gender Dependent Manner," Plos One, vol. 7, May 2012. [39] P. Monteiro, C. Victora, and F. Barros, "[Social, familial, and behavioral risk factors for obesity in adolescents]," Rev Panam Salud Publica, vol. 16, pp. 250-8, Oct 2004. [40] C. Zhang, K. M. Rexrode, R. M. van Dam, T. Y. Li, and F. B. Hu, "Abdominal obesity and the risk of all-cause, cardiovascular, and cancer mortality sixteen years of follow-up in US women," Circulation, vol. 117, pp. 1658-1667, 2008. [41] R. M. Reynolds, C. Osmond, D. I. W. Phillips, and K. M. Godfrey, "Maternal BMI, Parity, and Pregnancy Weight Gain: Influences on Offspring Adiposity in Young Adulthood," Journal of Clinical Endocrinology & Metabolism, vol. 95, pp. 5365-5369, Dec 2010.

132

[42] R. S. Ware, G. M. Williams, and R. L. Aird, "Participants Who Left a Multiple-Wave Cohort Study Had Similar Baseline Characteristics to Participants Who Returned," Annals of Epidemiology, vol. 16, pp. 820-823, 11// 2006. [43] A. A. Mamun, D. A. Lawlor, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Family and early life factors associated with changes in overweight status between ages 5 and 14 years: findings from the Mater University Study of Pregnancy and its outcomes," International Journal of Obesity and Related Metabolic Disorders, vol. 29, pp. 475-482, 2005. [44] A. J. Hammons and B. H. Fiese, "Is Frequency of Shared Family Meals Related to the Nutritional Health of Children and Adolescents?," Pediatrics, vol. 127, pp. e1565-e1574, 02/04/accepted 2011. [45] N. Shloim, L. R. Edelson, N. Martin, and M. M. Hetherington, "Parenting Styles, Feeding Styles, Feeding Practices, and Weight Status in 4-12 Year-Old Children: A Systematic Review of the Literature," Front Psychology, vol. 6, p. 1849, 2015.

133

Chapter 6: Parental pre-pregnancy obesity and the risk of offspring weight and body mass index change from childhood to adulthood.

6.1 Context

In the previous two chapters (chapters 4 and 5), we have explored the association between parental pre-pregnancy BMI, and obesity with the development of growth characteristics in early childhood (Chapter 4) and BMI, WC and obesity in young adulthood (Chapter 5). An independent association was observed between maternal and paternal pre-pregnancy BMI and obesity with the development of offspring obesity. In this chapter, we have further explored whether parental pre-pregnancy BMI and obesity prospectively associated with the changing patterns of offspring weight, BMI and obesity from childhood to adolescence and then to adulthood. We used the MUSP 21 year follow-up study.

Findings from this chapter showed that parental pre-pregnancy BMI associated with offspring weight and BMI change from childhood to adolescence and adolescence to adulthood. The strength of the association increased with the increase of offspring age. This chapter also present that parents who were OW/OB prior to pregnancy were associated with their offspring at higher risk of being OW/OB in childhood, adolescence and adulthood compared to offspring of normal weight parents. Thus, parental pre-pregnancy BMI and OW/OB is a strong predictor of offspring weight and BMI change from childhood to adulthood. This chapter forms a manuscript which has been under review for publication in the Clinical Obesity Journal.

134

6.2: Parental pre-pregnancy obesity and the risk of offspring weight and body mass index change from childhood to adulthood

Zalbahar N, Najman MJ, McIntyre HD, Mamun AA

Abstract We examined the association of parental pre-pregnancy weight and body mass index (BMI) on offspring weight and BMI change from childhood to adulthood. We analysed BMI data from a subsample of parents (n = 1494) from the Mater–University of Queensland Study of Pregnancy cohort that started in the early 1980s in Brisbane, Australia. Data was collected at pre-pregnancy and then also for offspring at 5, 14 and 21-year follow-ups. Multiple regression for continuous outcomes and multinomial regression for categorical outcomes was performed. A total of 14.7% of offspring experienced BMI change from normal at 5 years to overweight or obese (OW/OB) at 14 years, 15.3% from normal at 14 years to OW/OB at 21 years, and 22.8% from normal at 5 years to OW/OB at 21 years. We observed a significant and strong association between parents who were OW/OB prior to pregnancy and offspring weight and BMI change from childhood to adolescence or adulthood, but not adolescence to adulthood. Overall, the strength of the association of parental BMI with offspring BMI was stronger as offspring become older. When both parents were OW/OB, these associations were stronger than with one parent. Parental pre-pregnancy BMI and OW/OB is a strong predictor of offspring weight and BMI change from early life to adulthood.

Keywords: maternal obesity, obesity, pregnancy, parent, offspring, birth cohort

Introduction Obesity is a global public health problem and some countries are predicted to face increased rates of severe obesity [1, 2]. Irrespective of age and gender, once an individual is OW/OB, it is unusual for that person to return to a weight in the normal range. There is a high degree of continuity of OW/OB across the life course [3]. What is known about the natural history of OW/OB suggests that it rises with age [4, 5]. Many factors including genetic susceptibility of offspring; parental weight; and familial, social and environmental factors might play an important role in the development of an individual’s weight trajectory [6-8].

135

Parental excess weight is associated with offspring excess weight across the life course, although, some have suggested that the maternal–offspring link is stronger than the paternal– offspring link [9,10]. Studies have consistently found that when both parents are OW/OB their offspring are at greater risk of becoming OW/OB across the life course compared with one parent being OW/OB [3,6,11,12]. Most studies have examined these associations at only one time point in the offspring’s life; a few studies have examined multiple time points [11,13,14]. Only a handful of studies have examined these associations in relation to the changing pattern of offspring OW/OB [15-17].

In life-course epidemiology, three major stages have been identified as having important influences on subsequent body weight: pregnancy, childhood, and adolescence [6,18,19]. Variation in childhood development may relate to socio-economic background, environmental factors and puberty [20, 21]. These factors may influence parental-offspring BMI associations, which might be stronger or weaker at different development stages of the offspring life. A recent systematic review comparing genetic and environmental factors linking parental and offspring obesity demonstrated that childhood obesity was genetically and environmentally influenced by obese parents. However, the common environment effect of parents on children was attenuated by the time children had reached adolescence [22]. This suggested that parental–offspring obesity associations may differ with the developmental stages of children. We found only one study that reported parental-offspring BMI associations which increased from childhood to adulthood, but this report did not adjust for the correlated nature of the outcome and potential parental and offspring lifestyle factors [23]. To our knowledge, there is no prospective cohort study which describes the differential effect of pre-pregnancy parental BMI on offspring BMI at multiple developmental stages.

We hypothesised that parental pre-pregnancy BMI may prospectively predict offspring weight and BMI change across their life course from childhood to adolescence, and then to adulthood. Thus, we aimed to examine whether parental pre-pregnancy weight and BMI predict the weight change and BMI change of offspring from childhood to adolescence and then to young adulthood using a large community-based MUSP cohort that captures all three important offspring developmental stages and their subsequent impact on young adult BMI.

136

Methodology Study population The MUSP is a prospective study of 7223 mother–offspring pairs started in 1981–83 at a major public hospital in Brisbane, Australia. MUSP prospectively collected a range of socio- economic, mental and physical health data from mothers and offspring. Both mothers and offspring were followed-up prospectively at delivery, 6-months, and 5, 14 and 21 years post- delivery. Recently, mothers were followed-up at 27 years postpartum and offspring were followed up at 30 years. Information about the mothers and offspring was obtained by a questionnaire administered to the mothers at 3–5 days post-delivery and then at the 6-month follow-up (93%). At the 5-year follow-up 73% of mothers provided data for their offspring and 56% of the offspring attended physical assessment. At the 14-year follow-up, 72% of mothers and their offspring responded, and 53% of offspring attended physical assessment. At the 21-year follow-up, only 37% (2664) of offspring attended physical assessment. The reduction in participants undertaking physical assessment was due to limited funding [24-26].

Ethics approvals were obtained from the Mater Hospital and The University of Queensland ethics committees. Participants gave their signed informed consent for themselves and their offspring. At 21 years offspring provided their own consent. Details of the cohort recruitment and procedures are available elsewhere [24, 25]. A total sample of 1494 mothers, fathers, and offspring provided information for this study.

Measurements Paternal weight and height and the pre-pregnancy weight of mothers were self-reported by the mother. Maternal height was measured by an obstetrician using a portable stadiometer to the nearest 0.1 cm during the first clinic visit. Women were also weighed at this clinic (average gestational age was 18 weeks). There was high correlation between the two measures of weight (Pearson’s correlation coefficient = 0.95). Weight and height of the parents were used to derive BMI (weight (kg) / height (m2). For the purpose of analysis, parental BMI was categorised as: normal weight (less than 25 kg/m2) and OW/OB (more than 25 kg/m2).

Offspring’s’ BMI was measured at 5, 14 and 21-year follow-ups during physical assessment. At 5 and 14 years, weight was measured using scales and rounded to the nearest 0.2 kg. 137

Height was measured using a measuring tape and rounded to the nearest 0.2 cm. BMI categorisation at 5 and 14 years was based on a standard definition [27]. This reference was used which corresponded to the adult cut-off points for overweight and obesity. Centile curves were derived using data from six different reference populations (Great Britain, Brazil, the Netherlands, Hong Kong, Singapore and the USA). The curves passed through the points of 25 kg m2 and 30 kg m2 at age 18 years, specific to age and gender. At age 21 years, BMI categories were derived using the WHO recommendation [28]. Subjects were categorised as: normal weight (less than 25 kg/m2) or OW/OB (more than 25k g/m2).

Potential covariates were included in the multivariable analyses based on the published literature [29-31]. Available covariates in the MUSP data are maternal and paternal educational attainment, family annual income, maternal gestational weight gain, maternal smoking habit, offspring’s birth weight, offspring’s sex, gestational age, breastfeeding duration, offspring’s lifestyle at 14 years (TV watching, sports participation, and eats together with family). Maternal gestational weight gain was derived by subtracting pre-pregnancy weight from the maximum weight measured during pregnancy, which was obtained from obstetric data [32]. Data about smoking during pregnancy was obtained by self-reported average smoking at first clinical visit and late pregnancy, and divided into smoking and non- smoking groups. Family annual income and parental education were reported by mothers in a questionnaire administered at the initial data collection. At the first clinic visit, mothers reported family income on a questionnaire with seven categories: A$0–2599, A$2600–5199, A$5200–10399, A$10400–15599, A$15600–20799, A$20800–25999 and more than A$26000; later we collapsed tthese into two arbitrary categories (A$10400). Maternal and paternal education were categorised into three categories: incomplete education (no school, complete primary or incomplete secondary school), completed secondary education (grade 10 and 12) and completed further/higher education (college, university or other institution of advanced education). Offspring’s birth weight was measured immediately after birth and recorded in grams. At the 6-month follow-up, duration of breastfeeding was collected using a questionnaire with the following response options: never, 2 weeks or less, 3–6 weeks, 7 weeks–3 months, 4–6 months and still breastfeeding. Due to small numbers in the categories 2 weeks or less, 3–6 weeks, and 7 weeks–3 months, they were collapsed into three categories (never, less than 4 months, and more than 4 months).

138

Statistical analysis All statistical analyses were conducted using STATA version 13.0 (Stata Inc, College Station, TX). Chi square tests were used for categorical variables and t-tests for comparing means between groups. Means and proportions between complete and incomplete datasets were tabulated. Results are presented as mean and standard deviation (SD), and percentages for descriptive analyses where appropriate. The regression coefficients and odds ratio (OR) with 95% confidence interval (CI) were reported for multivariable analyses.

Parental BMI status was classified into four categories: (1) normal-weight mother and father; (2) normal-weight father, OW/OB mother; (3) normal-weight mother, OW/OB father; and (4) OW/OB mother and father. Offspring weight changes over the stages of life were categorised into four groups: 5 and 21 years: (1) normal weight at 5 and 21 years; (2) normal weight at 5 years, OW/OB at 21 years; (2) OW/OB at 5 years, normal weight at 21 years; and (4) OW/OB at 5 and 21 years. 14 and 21 years: (1) normal weight at 14 and 21 years; (2) normal weight at 14 years, OW/OB at 21 years; (3) OW/OB at 14 years, normal weight at 21 years; and (4) OW/OB at 14 and 21 years. 5 and 14 years: (1) normal weight at 5 and 14 years; (2) normal weight at 5 years, OW/OB at 14 years; (2) OW/OB at 5 years, normal weight at 14 years; and (4) OW/OB at 5 years and 14 years. For the purpose of testing the models, combination (1) for exposure and outcomes act as the reference group.

Multiple imputation was used to impute missing covariates in all models. Multinomial logistic regression was used to determine the association of parental BMI categories with offspring BMI change in categories from one stage to another. All potential covariates were included in the adjusted model. To determine whether parental pre-pregnancy weight and BMI were prospectively associated with offspring’s relative weight and BMI change from one stage to another, linear regression was used. Relative change from one stage to another was computed by dividing the absolute change between two stages with the baseline-stage value. Then it was multiplied by 100 to express the percentage change. Using GEE, we examined the differential associations of parental pre-pregnancy BMI with their offspring, 139

developmental stages of weight, and BMI at 5, 14 and 21 years. For every unit BMI change in mother and father, the average rate of offspring BMI change across three age groups was estimated using GEE for correlated BMI outcomes for offspring. To estimate the differential effects, we used contrast coding for the interaction of parental BMI and offspring BMI at three developmental stages.

Results A total of 1494 parent–offspring pairs provided parental pre-pregnancy BMI and offspring BMI at 5 years (mean age 4.7, SD 0.5), 14 years (mean age 13.3, SD 0.4), and 21 years (mean age 20.0, SD 0.8). Offspring with incomplete information disproportionately had a lower income, less education and a younger mother. However, offspring with complete information for weight and height were no different to those who were excluded due to missing values (Table 6.1). Table 6.2 presents the prevalence of OW/OB and its patterns from one stage to the next among male and female offspring. Of the total offspring at 5, 14 and 21 years, 15.7%, 24.5% and 33.5% were OW/OB, respectively. Male–female difference across different stages was not substantial. As for BMI-categories change among offspring, from childhood to adulthood, a quarter of the offspring were normal weight at 5 years and became overweight at 21 years. Approximately 15% of offspring who had normal BMI at 5 years became overweight at 14 years, and it was the same from age 14 to 21 years. Only 3% to 7% of the children who were OW/OB in the previous stage were in the normal weight range in the next stage. Approximately 10% of the offspring persistently remained OW/OB from 5 to 14 years or 5 to 21 years. It was double from 14 to 21 years. Figure 6.1 shows the patterns of offspring BMI change—(A) from 5 to 14 years and then (B) from 14 to 21 years—by the combinations of parental BMI categories. If both parents were OW/OB, 27.3% of offspring were OW/OB at 5 and 14 years and 30.9% become OW/OB during this period. Similarly, when both parents were OW/OB, 49.1% were OW/OB at age 14 and 21 years and only 16.4% become OW/OB from adolescence to young adulthood. When both parents had normal BMI, only 6.4% at 5 and 14 years, and 12.4% at 14 and 21 years were OW/OB. The odds of offspring being a normal weight at 5 years and OW/OB at 21 years were three times higher for those having a normal BMI father but an OW/OB mother compared with both parents having a normal BMI (Table 6.3). The odds were increased to 5.14 (95% CI: 140

3.11–8.50) when both parents were OW/OB. The odds of a child being a normal weight at 14 years and becoming OW/OB at 21 years were 3.22 (95% CI: 1.70–6.11) when both of their parents were OW/OB compared with those with normal-weight parents. As for the child being OW/OB at 14 years from normal weight at 5 years, the odds after adjustment were 6.59 (95% CI: 3.90-11.13) when both of their parents were OW/OB. The risk of the offspring continuing to be OW/OB at 5 to 21 years were also higher for those having OW/OB parents compared with those having normal-weight parents. The chances that offspring at 5 and 21 years maintained being OW/OB were 9.03 (95% CI: 5.05-16.16) compared with offspring with normal-weight parents, after adjusting for other factors. Considering the association between offspring weight change between 5 and 14 years with pre-pregnancy parental BMI, the odds of having offspring who remained OW/OB at both 5 and 14 years were 9.95 (95% CI: 5.60–17.69) when both parents were OW/OB. Similarly, when we examined the association of parental BMI (continuous and categorical scale) with offspring relative BMI change (continuous and tertile) from one stage to the next stage, direction and strength of the associations are similar (Table 6.4). Adjustment for potential confounders did not substantially change any of the associations between parental BMI and changing patterns of offspring BMI from one stage to the next.

The differential associations of parental pre-pregnancy BMI with their offspring’s BMI at the developmental stages of 5, 14 and 21 years is presented in Figure 6.2. Results are presented for both BMI and the BMI z-scores. For every unit increase in paternal and maternal BMI z- score, offspring BMI z-score increased on average by between 0.15% (kg/m2) and 0.24% (kg/m2), respectively, throughout all three stages of life (Figure 6.2). When we separated this average effect by the three developmental stages, the strength of parental–offspring BMI score associations increased with offspring age. We found no differential associations of parental BMI score with offspring BMI score across three different stages by gender (results are not shown).

141

OW/OB at 5yrs & 14yrs (A) At 5 and 14 years OW/OB at 5yrs & NW at 14yrs NW at 5yrs & OW/OB at 14yrs NW at 5yrs & 14yrs

6.4 12.3 5.9 14 27.3 10.9 6.8 4.4

14.5 4.6 27.9

30.9

Percentage (%) Percentage 76.9 66.5 53.7 37.3

Normal weight parents Normal weight mother, Overweight or obese Both parents, overweight or obese mother, normal father overweight and obese father

OW/OB at 14yrs & 21yrs (B) At 14 and 21 years OW/OB at 14yrs & NW at 21yrs NW at 14yrs & OW/OB at 21yrs NW at 14yrs & 21yrs

12.4 21 4.9 29.4

14.1 5.8 49.1

19 12.5

14.7

9.1 Percentage (%) Percentage 68.7 16.4 54.2 43.3 25.5

Normal weight parents Normal weight mother, Overweight or obese Both parents, overweight or obese mother, normal father overweight and obese father

Figure 6.1Distribution of offspring BMI change pattern in categories: (A) 5 and 14 years and (B) 14 and 21 years by prenatal parental BMI category in combination: Both NW parents; NW mother and OW/OB father; OW/OB mother and NW father; both OW/OB parents.

Legend: (A) Lightest grey, NW at 5 and 14; grey, NW at 5 years and OW/OB at 14 years; dark grey, OW/OB at 5 years and NW at 14 years; black, OW/OB at 5 and 14 years. (B) Lightest grey, NW at 14 and 21 years; grey, NW at 14 years and OW/OB at 21 years; dark grey, OW/OB at 14 years and NW at 21 years; black, OW/OB at 14 and 21 years.

142

Table 6.1 Comparison of the characteristics of participants who provided complete information and those who did not provide complete information.

Covariates Complete information of Incomplete information of P-value parental pre-pregnancy parental pre-pregnancy BMI and offspring’s BMI and offspring’s BMI BMI at 5, 14 and 21 at 5, 14 and 21 years years (n = 5729) (n = 1494) Mean (SD) or n (%) Mean (SD) or n (%) P value Fathers BMI (kg/m2) 23.7 (3.3) 23.6 (3.6) 0.387 Mothers pre-pregnancy BMI 22.1 (4.0) 21.9 (4.0) 0.101 (kg/m2) Maternal total gestational 14.5 (5.2) 14.7 (5.7) 0.174 weight gained (kg) Offspring BMI at 5 yrs (kg/m2) 16.0 (1.5) 16.0 (1.7) 0.180 Offspring BMI at 14 yrs 20.5 (3.6) 20.7 (4.0) 0.081 (kg/m2) Offspring BMI at 21 yrs 24.2 (4.8) 24.4(5.2) 0.280 (kg/m2) Offspring birth weight (g) 3405.6 (507.4) 3380.6 (521.6) 0.097

Offspring sex Male 745 (49.9) 3013 (52.6) 0.060 Female 749 (50.1) 2716 (47.4) Family income at baseline Less than A$10400 395 (27.3) 1913 (36.1) <0.001 More than A$10400 1051 (72.7) 3390 (63.9) Maternal education at birth Incomplete higher education 214 (14.4) 1091 (19.2) <0.001 Completed higher education 942 (63.3) 3667 (64.5) Post higher education 332 (22.3) 924 (16.3) Mother’s age at birth 13–19 120(8.0) 876 (15.3) <0.001 20–34 1287 (86.1) 4583 (80.0) >34 87 (5.8) 270 (4.7) Maternal smoking habit during pregnancy Nil smoker 962 (68.9) 3167 (60.1) <0.001 Smoker 434 (31.1) 2103 (39.9)

*P-value: describes the difference in means (and proportions) between incomplete and complete information

143

Table 6.2 Prevalence and change in overweight or obesity status between ages 5, 14 and 21 years.

Males (%) Females (%) Total (%) Mean (SD) Mean (SD) Mean (SD) N N = 745 N = 749 N = 1494 BMI at 5 years Normal 1260 85.9(15.7;1.0) 82.8(15.3;1.1) 84.3(15.5;1.0) OW/OB 234 14.1(18.6;1.1) 17.2(18.4;1.3) 15.7(18.5;1.2)

BMI at 14 years Normal 1128 75.7(18.7;1.9) 75.3(19.0;2.1) 75.5(18.9;2.0) OW/OB 366 24.3(25.1;2.6) 24.7(25.8;3.0) 24.5(25.5;2.8)

BMI at 21 years Normal 990 63.8(21.7;2.0) 68.8(21.2;2.2) 66.3(21.5;2.1) OW/OB 504 36.2(29.0;3.5) 31.2(30.2;4.7) 33.7(29.5;4.1)

BMI at 5 and 14 years Normal both ages 1041 70.7(17.1;1.2) 68.6(17.0;1.3) 69.7(17.1;1.3) Normal at 5, Ow/OB at 14 219 15.7(20.4;1.1) 14.2(20.3;1.2) 14.7(20.4;1.1) OW/OB at 5, normal at 14 87 5.0(19.3;0.6) 6.7(19.3;0.9) 5.8(19.3;0.8) OW/OB at 5 and 14 147 9.1(22.5;2.0) 10.6(23.0;2.2) 9.8(22.8;2.1)

BMI at 14 and 21 years Normal both ages 899 58.0(19.9;1.6) 62.4(19.8;1.8) 60.2(19.9;1.7) Normal at 14, OW/OB at 21 229 17.7(24.0;1.6) 13.0(24.4;1.6) 15.3(24.2;1.6) OW/OB at 14, normal at 21 91 5.8(23.4;1.0) 6.4(23.9;1.0) 6.1(23.7;1.0) OW/OB at 14 and 21 275 18.5(27.9;2.7) 18.3(29.0;3.7) 18.4(28.4;3.3)

BMI at 5 and 21 years Normal both ages 919 60.4(18.6:1.2) 62.6(18.1;1.4) 61.5(18.3;1.3) Normal at 5, OW/OB at 21 341 25.5(22.4;1.7) 20.2(22.5;2.1) 22.8(22.4;1.8) OW/OB at 5, normal at 21 71 3.4(20.6;0.7) 6.1(20.5;1.1) 4.8(20.5;1.0) OW/OB at 5 and 21 163 10.7(24.3;2.3) 11.1(25.1;3.1) 10.9(24.7;2.7)

144

Table 6.3 Unadjusted and adjusted ORs with 95% CI of offspring’s changing patterns of BMI categories from stage to stage by parental BMI categories

Offspring BMI categories change (OR and 95% CI), N = 1494 Parental BMI combination Normal at 5 yrs, OW/OB at 14 OW/OB at 5 yrs, normal weight OW/OB at 5 and 14 yrs yrs at 14 yrs Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted Normal-weight mother and father Reference Normal-weight mother, OW/OB father 1.54 1.62 1.34 1.42 2.22 2.34 (1.05–2.27) (1.10–2.39) (0.79–2.26) (0.83–2.44) (1.43-3.42) (1.50-3.65) Normal-weight father, OW/OB mother 3.68 3.91 1.08 1.06 3.13 3.06 (2.36–5.73) (2.46–6.21) (0.45–2.59) (0.43–2.62) (1.78-5.53) (1.69-5.54) Both parents OW/OB 5.86 6.59 1.60 1.78 8.79 9.95 (3.56–9.66) (3.90–11.13) (0.61–4.21) (0.65–4.84) (5.13-15.08) (5.60-17.69) Normal at 14 yrs, OW/OB at 21 OW/OB at 14 yrs, normal OW/OB at 14 and 21 yrs yrs weight at 21 yrs Normal-weight mother and father Reference Normal-weight mother, OW/OB father 1.71 1.71 1.50 1.53 2.15 2.27 (1.21-2.43) (1.20-2.45) (0.85-2.65) (0.86-2.73) (1.52-3.04) (1.60-3.24) Normal-weight father, OW/OB mother 1.65 1.66 4.03 3.88 3.76 4.00 (0.96-2.84) (0.65-2.90) (2.18-7.48) (2.03-7.41) (2.41-5.89) (2.50-6.36) Both parents overweight or obese 3.14 3.22 5.00 5.16 10.71 12.47 (1.69-5.84) (1.70-6.11) (2.29-10.92) (2.28-11.68) (6.51-17.61) (7.40-21.03)

Adjusted for maternal age at birth, maternal education attainment, baseline family income, maternal smoking habit, maternal gestational weight gained, offspring’s birth weight, sex and lifestyle at 14 years.

145

Table 6.4 Adjusted associations of parental BMI and offspring’s relative BMI change (continuous scale and tertiles) from 5 to 14 years, 14 to 21 years, and 5 to 21 years

Offspring’s relative BMI (kg/m2) change (%)

5 to 14 years 14 to 21 years 5 to 21 years Paternal BMI& Regression coefficient (95% CI)* Continuous 0.74 0.28 1.23 (0.47–1.01) (0.03–0.52) (0.84–1.64)

Odd ratios (95% CI)* Categories&& Q1 Q2 Q3 Q1 Q2 Q3 Q1 Q2 Q3 <25 kg/m2 Reference ≥25 kg/m2 1.00 1.39 1.74 1.00 1.13 1.20 1.00 1.22 2.03 (1.03–1.87) (1.30–2.35) (0.84–1.50) (0.90–1.60) (0.90–1.64) (1.52–2.71)

Maternal BMI& Regression coefficient (95% CI)* Continuous 1.19 0.05 1.47 (0.96–1.42) (–0.16–0.26) (1.12–1.81)

Odd ratios (95% CI)* Categories && Q1 Q2 Q3 Q1 Q2 Q3 Q1 Q1 Q3 <25 kg/m2 Reference ≥25 kg/m2 1.00 1.75 3.52 1.00 0.77 1.03 1.00 0.97 2.28 (1.16–2.63) (2.38–5.20) (0.54–1.12) (0.73–1.46) (0.66–1.45) (1.60–3.25)

*Adjusted for other parent’s BMI, maternal age at birth, maternal educational attainment, family income, maternal smoking habit, maternal gestational weight gain, offspring birth weight, offspring sex and offspring lifestyle at 14 (family meal time, sports and TV watching) & Used multiple linear regression adjusting for the potential covariates as above && Used multinomial logistic regression adjusting for the potential covariates as above

146

A) Paternal-offspring B) Maternal-offspring  Maternal–offspring BMI and BMI z-scores . Paternal–offspring BMI and BMI z-scores 0.5 0.5

Adjusted BMI z-scores

CI 0.4 0.4 Adjusted BMIz-scores 0.3 0.3

0.2 0.2

0.1 0.1 Adjusted BMI

Adjusted BMI Regression coefficient, 95% Regression 95% coefficient,

0 0

-0.1 -0.1 BMI BMI-z 5 yrs 14 yrs 21 yrs 5 yrs 14 yrs 21 yrs BMI BMI-z 5 yrs 14 yrs 21 yrs 5 yrs 14 yrs 21 yrs

Figure 6.2 Adjusted regression coefficient of overall parental–offspring BMI and BMI-z score association and the separated effect at three developmental stages (5, 14 and 21 years). (A) Paternal–offspring association; (B) Maternal–offspring association. Adjusted for maternal age at birth, baseline family income, maternal smoking habit, maternal gestational weight gained, offspring’s birth weight, offspring’s sex, and offspring’s lifestyle at 14 years (family meals, TV watching and sports participation).

147

Discussion In this study we found that parental pre-pregnancy BMI was independently associated with offspring BMI, and change in BMI from childhood to adolescence and adolescence to adulthood. We also found that parental pre-pregnancy BMI has a differential impact on offspring BMI at childhood, adolescence, and early young adulthood. These associations became stronger as the offspring became older. Importantly, the differential associations were not confounded or mediated by a range of parental, family and offspring factors such as family annual income, parental education, maternal factors (gestational weight gain, maternal age at birth) and offspring factors (birth weight, sex, and lifestyle at 14 years).

We are unaware of any study that has examined these associations with offspring BMI change at three significant time points. Therefore, it is difficult to make direct comparisons. However, like ours, previous studies found that parental pre-pregnancy BMI has a strong influence on offspring BMI later in life [11,12,33,34] and this risk was substantially greater when both parents were obese [3,6]. We also found that when both parents are OW/OB before pregnancy, there is a greater chance that their offspring will continue being OW/OB during their adolescence and in young adulthood compared with offspring of parents with normal pre-pregnancy weight. This association does not change after adjustment for other parental and offspring factors. Previous studies had not considered the association of the BMI of both parents in early life with offspring BMI change across the life course [11,35,36]. We found a study by Lawrence [16] that showed that for each unit of maternal pre-pregnancy BMI, offspring BMI changed 0.83 kg/m2 from age 17 to 32 years. It was not explained by the potential confounders they included in the adjusted model.

Our results demonstrate that parental pre-pregnancy BMI are strong predictors of the development of offspring BMI across different stages of life. If high parental BMI at time of birth predicts a lifestyle over the next 21 years that leads to OW/OB across offspring life course, focusing on parental health and weight before pregnancy might be an option for interventions which aim to prevent the continuation of the obesogenic inter-generational cycle [37,38]. Recently, the links between maternal pre-pregnancy BMI and the development of obesity and non-communicable disease have been widely elaborated [39-41]. Several pathways have been suggested and examined which may influence significant future health outcomes in offspring [37]. For example, the over- nutrition hypothesis supposedly explains how the maternal intra-uterine environment modifies offspring nutrition relating to the supply of glucose, fat and protein in the womb [42,43]. Any changes in nutrition and biochemical status of the mother affect her offspring’s risk of an unhealthy outcome in the short and long term. Paternal factors are now gaining attention [37, 44]. With 148

combined interventions involving both parents before pregnancy, in-utero environmental programming may be controlled. The combination of parent-weight status before pregnancy has a significant impact not only on birthweight but on offspring-weight change and BMI change throughout life.

In our study, we found a significant association of parental BMI and relative offspring-BMI change from childhood to adulthood, childhood to adolescence, and adolescence to adulthood. The associations are robust after adjustment for other parental and offspring lifestyle factors. Our findings raise the possibility that early childhood may act as the most vulnerable stage of life, which may be easily influenced by environmental and biological factors. For example, a child’s eating traits and behaviours are likely to be moulded and influenced mainly by their parents [45,46]. Kral and Faith [47] reported that parental feeding practices do influence child eating habits which, when compounded with existing genetic factors, increase the chances of offspring being OW/OB later in life. Nonetheless, at this young age, the offspring may also be vulnerable to modification of phenotypes related to obesogenic genes [33]. Further, epigenetic mechanisms may also be involved in these changes and may be largely fixed at the beginning of life [48,49]. Targeting the early stages of life may slow down or reduce the chances of being OW/OB later in life. Supporting this idea, previous studies have shown a high risk of being OW/OB among individuals later in life when they were OW/OB at a younger age [3,9,12,50]. The mechanisms and pathways leading to OW/OB are still the subject of ongoing research and remain poorly understood. The degree of parental– offspring BMI association we observed became stronger throughout the offspring’s life. This is consistent with the Finnish prospective study of young people [23]. Various factors surrounding offspring life such as peer pressure during adolescent years, different working environment during adulthood, and other behavioural changes may influence and modify parental–offspring BMI association. When the child grows older, the influence of parents on their offspring is prominent. Our findings confirm the parallel evolution of parent–child weight status that may suggest underlying biological, physical and psychological influences. Consistent with intergenerational studies, stable transmission of OW/OB throughout families in between generations was found [51,52]. Even within generations, the continuity of increasing BMI in families persists and is apparent in the long term. These predisposed children—with obese genes, environmental factors and over-nutrition programming—subsequently develop increased BMI later in life. Therefore, we suggest that potential interventions among parents should commence before conception and also involve offspring very early in their lives.

149

The strength of this study is the prospective longitudinal study design with a combination of complete measures of weight and height of parental BMI before pregnancy and offspring at 5, 14 and 21 years. Other advantages include a large sample size and the availability of information collected in early life, including parental before and during pregnancy lifestyle and socio- demographic characteristics, offspring lifestyle during adolescence allowing adjustment for these factors.

Study limitations In this study, parental weight and height before pregnancy were self–reported; this may contribute a bias although we found a strong correlation (r = 0.95) between self-reported weight and actual weight. BMI change may also be influenced by physical inactivity and unhealthy dietary patterns of the offspring later in life. We could not adjust for the validated measures of offspring dietary patterns and physical inactivity across all age groups. However, we could not find any longitudinal studies similar to ours that have examined the impact of offspring dietary pattern on parental– offspring BMI links. Therefore, further studies are warranted for cohort follow-up, with concomitant measurement of diet and lifestyle across different age groups. Such detailed information on physical activity and dietary patterns may shed further light on whether these factors influence BMI and weight change throughout the development of the child. Important offspring factors such as family meals, TV watching, and sports participations at 14 years were available and were used in our analyses. We adjusted for offspring’s breastfeeding duration at the 6-month follow-up. Loss to follow-up in a cohort study such as MUSP may introduce bias in the findings; however, strategies used to manage attrition bias in this cohort have been reported elsewhere [53] and found attrition did not bias our findings substantially. In this study we have also used multiple imputation to impute the missing covariates and found results are not substantially different from the complete case analysis.

Conclusion We conclude that offspring BMI change from childhood to any stage of life is influenced by parental pre-pregnancy BMI. The stronger relationship seen between parental BMI and offspring BMI in adulthood suggests the relevance of these associations in the longer term. Interventions in early life and among parents before pregnancy may have the potential to reduce the risk of familial obesity aggregation in future. Interventions focusing on early life factors involving maintaining healthy weight among parents prior to pregnancy and in their offspring’s early childhood may be crucial to help curb the continuation of the obesity cycle in future life.

150

References [1] M. M. Finucane, G. A. Stevens, M. J. Cowan, G. Danaei, J. K. Lin, C. J. Paciorek, et al., "National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9·1 million participants," The Lancet, vol. 377, pp. 557-567, // 2011. [2] N. R. F. C. (NCD-RisC). "Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19.2 million participants," The Lancet, vol. 387, pp. 1377-1396, 2016. [3] R. C. Whitaker, Wright, J.A., Pepe, M.S., Seidel, K.D., Dietz, W.H., "Predicting obesity in young adulthood from childhood and parental obesity," New England Journal of Medicine, vol. 337, pp. 869 - 873, 1997. [4] C. L. Ogden, M. D. Carroll, L. R. Curtin, M. A. McDowell, C. J. Tabak, and K. M. Flegal, "PRevalence of overweight and obesity in the united states, 1999-2004," JAMA, vol. 295, pp. 1549-1555, 2006. [5] M. Ng, T. Fleming, M. Robinson, B. Thomson, N. Graetz, C. Margono, 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," The Lancet, vol. 384, pp. 766-781, 2014. [6] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357-1359, Jun 2005. [7] C. Semmler, J. Ashcroft, C. H. M. van Jaarsveld, S. Carnell, and J. Wardle, "Development of overweight in children in relation to parental weight and socioeconomic status," Obesity, vol. 17, pp. 814-820, Apr 2009. [8] J.-P. Chaput, L. Pérusse, J.-P. Després, A. Tremblay, and C. Bouchard, "Findings from the Quebec Family Study on the Etiology of Obesity: Genetics and Environmental Highlights," Current Obesity Reports, vol. 3, pp. 54-66, 2014/03/01 2014. [9] R. M. Reynolds, C. Osmond, D. I. W. Phillips, and K. M. Godfrey, "Maternal BMI, Parity, and Pregnancy Weight Gain: Influences on Offspring Adiposity in Young Adulthood," Journal of Clinical Endocrinology & Metabolism, vol. 95, pp. 5365-5369, Dec 2010. [10] K. L. Whitaker, M. J. Jarvis, R. J. Beeken, D. Boniface, and J. Wardle, "Comparing maternal and paternal intergenerational transmission of obesity risk in a large population- based sample," American Journal of Clinical Nutrition, vol. 91, pp. 1560 - 1567, 2010. [11] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: 151

a 16-year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct 2011. [12] P. Chivers, H. Parker, M. Bulsara, L. Beilin, and B. Hands, "Parental and early childhood influences on adolescent obesity: a longitudinal study," Early Child Development and Care, vol. 182, pp. 1071-1087, 2013/04/21 2012. [13] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166-1175, 2011. [14] C. M. Murrin, G. E. Kelly, R. E. Tremblay, and C. C. Kelleher, "Body mass index and height over three generations: evidence from the Lifeways cross-generational cohort study," BMC Public Health, vol. 12, p. 81, 2012. [15] A. A. Mamun, D. A. Lawlor, M. J. O'Callaghan, G. M. Williams, and J. M. Najman, "Family and early life factors associated with changes in overweight status between ages 5 and 14 years: findings from the Mater University Study of Pregnancy and its outcomes," International Journal of Obesity and Related Metabolic Disorders, vol. 29, pp. 475-482, 2005. [16] G. M. Lawrence, Shulman, S., Friedlander, Y., Sitlani, C.M., Burger, A., Savitsky, B., Granot-Hershkovitz, E., Lumley, T., Kwok, P.-Y., Hesselson, S., Enquobahrie, D., Wander, P.L., Manor, O., Siscovick, D.S., Hochner, H., "Associations of maternal pre-pregnancy and gestational body size with offspring longitudinal change in BMI," Obesity, vol. 22, 2014. [17] K. Hesketh, J. Carlin, M. Wake, and D. Crawford, "Predictors of body mass index change in Australian primary school children," International Journal of Pediatric Obesity, vol. 4, pp. 45-53, 2009. [18] A. J. Venn, R. J. Thomson, M. D. Schmidt, V. J. Cleland, B. A. Curry, H. C. Gennat, et al., "Overweight and obesity from childhood to adulthood: a follow-up of participants in the 1985 Australian Schools Health and Fitness Survey," Medical Journal of Australia, vol. 186, pp. 458-60, May 7 2007. [19] S. F. Weng, S. A. Redsell, J. A. Swift, M. Yang, and C. P. Glazebrook, "Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy," Archives of Diseases in Childhood, vol. 97, pp. 1019-1026, Oct 29 2012. [20] J. Araújo, H. Barros, M. Severo, C. Lopes, and E. Ramos, "Longitudinal changes in adiposity during adolescence: A population-based cohort," BMJ Open, vol. 4, 2014. [21] K. K. L. Ong, M. L. Ahmed, P. M. Emmett, M. A. Preece, D. B. Dunger, T. A. L. S. o. P. Team, et al., "Association between postnatal catch-up growth and obesity in childhood: prospective cohort study," BMJ: British Medical Journal, vol. 320, pp. 967-971, 2000. 152

[22] K. Silventoinen, B. Rokholm, Kaprio, J., and T. I. A. Sorensen, "The genetic and environmental influences on childhood obesity: a systematic review of twin and adoption studies," International Journal of Obesity, vol. 34, pp. 29-40, 2009. [23] M. Kivimäki, D. A. Lawlor, G. D. Smith, M. Elovainio, M. Jokela, L. Keltikangas-Järvinen, et al., "Substantial intergenerational increases in body mass index are not explained by the fetal overnutrition hypothesis: The Cardiovascular Risk in Young Finns Study," American Journal of Clinical Nutrition, vol. 86, pp. 1509-1514, 2007. [24] J. M. Najman, R. Alati, W. Bor, A. Clavarino, A. Mamun, J. J. McGrath, et al., "Cohort profile update: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology vol. 44, pp. 78-78f, December 16, 2014 2014. [25] J. Keeping, J. Najman, J. Morrison, J. Western, M. Andersen, and G. Williams, "A prospective longitudinal study of social, psychological and obstetric factors in pregnancy: response rates and demographic characteristics of the 8556 respondents," British Journal of Obstetrics and Gynaecology, vol. 96, pp. 289 - 97, 1989. [26] A. A. Mamun, M. J. O'Callaghan, S. M. Cramb, J. M. Najman, G. M. Williams, and W. Bor, "Childhood behavioral problems predict young adults' BMI and obesity: evidence from a birth cohort study," Obesity (Silver Spring), vol. 17, pp. 761-6, Apr 2009. [27] T. Cole, M. Bellizzi, K. Flegal, and W. Dietz, "Establishing a standard definition for child overweight and obesity worldwide: international survey," British Medical Journal, vol. 320, pp. 1240 - 1243, 2000. [28] W. H. O. (WHO). "Obesity: Preventing and Managing a Global Epidemic, Report of a WHO Consultant on Obesity. Geneva: World Health Organization.," 1998. [29] A. M. F. M. R. M. K. B. Stuebe, "Maternal-recalled gestational weight gain, pre-pregnancy body mass index, and obesity in the daughter," International Journal of Obesity, 2009. [30] M. B. Terry, Wei, Y., Esserman, D., McKeague, I.W., Susser, E., "Pre-and postnatal determinants of childhood body size: Cohort and sibling analyses," Journal of Developmental Origins of Health and Disease, vol. 2, pp. 99-111, 2011. [31] L. Fairley, G. Santorelli, D. A. Lawlor, M. Bryant, R. Bhopal, E. S. Petherick, et al., "The relationship between early life modifiable risk factors for childhood obesity, ethnicity and body mass index at age 3 years: findings from the Born in Bradford birth cohort study," BMC Obesity, vol. 2, p. 9, 2015. [32] A. A. Mamun, M. O'Callaghan, L. Callaway, G. Williams, J. Najman, and D. A. Lawlor, "Associations of gestational weight gain with offspring body mass index and blood pressure at 21 years of age evidence from a birth cohort study," Circulation, vol. 119, pp. 1720-U52, Apr 2009. 153

[33] S. R. Veena, G. V. Krishnaveni, S. C. Karat, C. Osmond, and C. H. Fall, "Testing the fetal overnutrition hypothesis; the relationship of maternal and paternal adiposity to adiposity, insulin resistance and cardiovascular risk factors in Indian children," Public Health Nutrition, vol. 16, pp. 1656-66, Sep 2013. [34] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-Offspring Body Mass Index Associations in the Norwegian Mother and Child Cohort Study: A Family-based Approach to Studying the Role of the Intrauterine Environment in Childhood Adiposity," American Journal of Epidemiology, vol. 176, pp. 83- 92 July 15, 2012 2012. [35] B. Patro, A. Liber, B. Zalewski, L. Poston, H. Szajewska, and B. Koletzko, "Maternal and paternal body mass index and offspring obesity: a systematic review," Annals of Nutrition Metabolism, vol. 63, pp. 32-41, 2013. [36] E. Pomeroy, J. C. Wells, T. J. Cole, M. O'Callaghan, and J. T. Stock, "Relationships of maternal and paternal anthropometry with neonatal body size, proportions and adiposity in an Australian cohort," American Journal of Physical Anthropology, vol. 156, pp. 625-636, Dec 13 2014. [37] S. C. Langley-Evans, "Nutrition in early life and the programming of adult disease: a review," Journal of Human Nutrition and Dietetics, vol. 28, pp. 1-14, 2015. [38] J. Moss and K. Harris, "Impact of maternal and paternal preconception health on birth outcomes using prospective couples’ data in Add Health," Archives of Gynecology and Obstetrics, vol. 291, pp. 287-298, 2015/02/01 2015. [39] H. Hochner, Friedlander, Y., Calderon-Margalit, R., Meiner, V., Sagy, Y., Avgil-Tsadok, M., Burger, A., Savitsky, B., Siscovick, D.S., Manor, O., "Associations of maternal prepregnancy body mass index and gestational weight gain with adult offspring cardiometabolic risk factors: The Jerusalem perinatal family follow-up study," Circulation, vol. 125, pp. 1381-9, 2012. [40] M. A. Tsadok, Y. Friedlander, O. Paltiel, O. Manor, V. Meiner, H. Hochner, et al., "Obesity and blood pressure in 17-year-old offspring of mothers with gestational diabetes: insights from the Jerusalem perinatal study," Experimental Diabetes Research, 2011. [41] X. Liu, Q. Chen, H. J. Tsai, G. Wang, X. Hong, Y. Zhou, et al., "Maternal preconception body mass index and offspring cord blood DNA methylation: exploration of early life origins of disease," Environmental and Molecular Mutagenesis, vol. 55, pp. 223-30, Apr 2014. [42] E. Oken and M. W. Gillman, "Fetal origins of obesity," Obesity Research, vol. 11, pp. 496- 506, 2003. 154

[43] R. C. Whitaker and W. H. Dietz, "Role of the prenatal environment in the development of obesity," Journal of Pediatrics, vol. 132, pp. 768-776, 5// 1998. [44] N. O. McPherson, T. Fullston, R. J. Aitken, and M. Lane, "Paternal obesity, interventions, and mechanistic pathways to impaired health in offspring," Annals of Nutrition and Metabolism, vol. 64, pp. 231-238, 2014. [45] M. S. Faith, R. I. Berkowitz, V. A. Stallings, J. Kerns, M. Storey, and A. J. Stunkard, "Parental feeding attitudes and styles and child body mass index: prospective analysis of a gene-environment interaction," Pediatrics, vol. 114, pp. e429-e436, 2004. [46] M. Golan and S. Crow, "Parents are key players in the prevention and treatment of weight related problems.," Nutrition Reviews, vol. 62, pp. 39-50, 2004. [47] T. V. E. Kral and M. S. Faith, "Influences on child eating and weight development from a behavioral genetics perspective," Journal of Pediatric Psychology, vol. 34, pp. 596-605, July 1, 2009 2009. [48] D. A. Lawlor, N. J. Timpson, R. M. Harbord, S. Leary, A. Ness, M. I. McCarthy, et al., "Exploring the developmental overnutrition hypothesis using parental-offspring associations and FTO as an instrumental variable," PLoS Medicine, vol. 5, p. e33, Mar 11 2008. [49] P. D. Gluckman and M. A. Hanson, "Developmental and epigenetic pathways to obesity: an evolutionary-developmental perspective," International Journal of Obesity, vol. 32, pp. S62-S71, //print 2008. [50] S. Williams, "Overweight at age 21: the association with body mass index in childhood and adolescence and parents' body mass index. A cohort study of New Zealanders born in 1972- 1973," International Journal of Obesity and Related Metabolic Disorders, vol. 25, pp. 158- 63, Feb 2001. [51] T. A. Ajslev, L. Angquist, K. Silventoinen, J. L. Baker, and T. I. A. Sorensen, "Stable intergenerational associations of childhood overweight during the development of the obesity epidemic," Obesity, vol. 23, pp. 1279-1287, Jun 2015. [52] P. C. D. Johnson, Logue, J., McConnachie, A., Abu-Rmeileh, N.M.E., Hart, C., Upton, M.N., Lean, M., Sattar, N., Watt, G., "Intergenerational change and familial aggregation of body mass index," European Journal of Epidemiology, vol. 27, pp. 53-61, 2012. [53] J. Najman, W. Bor, M. O'Callaghan, G. Williams, R. Aird, and G. Shuttlewood, "Cohort Profile: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology, vol. 34, pp. 992 - 997, 2005.

155

Chapter 7: Impact of maternal anthropometric change from pre-pregnancy to 21 years postpartum on adult offspring anthropometry and central obesity.

7.1 Context

In previous three chapters (Chapters 4, 5 and 6), the association between parental BMI and adiposity and offspring BMI and obesity over the life course has been explored. The association between parental pre-pregnancy BMI and offspring BMI increased with the increase of the offspring aged (Chapter 6). Also, when both parents were OW/OB prior to pregnancy increased the risk of offspring BMI change in childhood to adolescence and into adulthood was observed. In the natural history offspring follow mother’s weight change trajectory. In this chapter, changes in maternal weight and BMI from pre-pregnancy to 21 years after pregnancy and its impact on offspring weight, BMI, WC and obesity have been explored using the MUSP 21 year study.

Overall, maternal 21-year postpartum weight and BMI change was identified to associate with the risk of offspring being OW/OB in adulthood. Association between maternal 21-year postpartum BMI change and offspring BMI and WC in adulthood was significant after adjustment for potential covariates (i.e. maternal gestational age, total gestational weight gain, smoking during pregnancy, and offspring’s lifestyle at 14 years). Thus, this will provide evidence that if parents control their weight may influence their offspring weight change in the life course. This chapter forms a manuscript which has been resubmitted to the European Journal of Epidemiology.

156

7.2 Impact of maternal anthropometric change from pre-pregnancy to 21 years postpartum on adult offspring anthropometry and central obesity

Zalbahar N, Najman MJ, McIntyre HD, Mamun AA

Abstract There is a need to know more about whether parental weight and BMI change have an impact on offspring weight status over the child’s early life course. A total of 1981 mother–offspring dyads from the Mater–University of Queensland Study of Pregnancy (MUSP) cohort were studied with complete information on weight, height and waist circumference. Linear regression and multivariate logistic regression were used to examine the independent association between maternal absolute and relative postpartum weight change (PPWC) and BMI change (PPBMIC) over a 20-year period with changes in their offspring BMI and waist circumference (WC) over that same period of time. A linear relationship between maternal post-partum weight change and offspring BMI and WC were observed. Overall, relative PPWC in the highest quintiles is associated with adult offspring’s risk of overweight/obesity with the odds 1.47 (95%CI: 1.08, 2.00) and 1.50 (95%CI: 1.10, 2.02), respectively. Our findings suggest that increased maternal weight and BMI for 21 years following a pregnancy predict offspring overweight or obesity in young adulthood.

Introduction A review found that it might prove effective to involve parents in encouraging behavioural and weight change among their offspring [1,2]. The review includes evidence of a link between changes in parental BMI and changes in their offspring BMI. In one study, a decrease in each unit of parental BMI is associated with a reduction of 0.255 kg/m2 child BMI after 11 months follow-up in family-based behavioural obesity treatment [3]. Another study found that children of parents who had a greater reduction of BMI after two years, had significantly greater reduction in BMI than children of parents who had lowest BMI change, and the proportion of overweight children was 16.8% lower than those parents who had less BMI reduction (10.5% less) [4]. However, these studies demonstrated only a short-term impact (up to 24 months) of parental weight change on childhood weight status.

There is no available data regarding the potential long-term impact of such interventions. To supplement the limited available body of data, it is possible to use longitudinal anthropometric data from follow-up birth-cohort studies to assess the long-term impact of parental weight and BMI

157

changes as these are associated with offspring weight and BMI. Data from population-based studies suggests that when parents increase in weight, many children eventually gain weight [5,6]. In prospective studies, increases in parental or maternal BMI have been found to be strongly associated with offspring BMI and obesity across the life course [7-10]. However, there are relatively few studies which have examined long-term changes in maternal BMI as these are associated with offspring changes in BMI over the same period of time. Recent findings suggest that mothers gaining weight of more than 5 kg after 3 years post-delivery has been linked with the risk (OR: 1.12; 95% CI: 1.04-1.21) of their offspring being overweight in childhood [11]. Two other prospective studies have reported that maternal postpartum weight gained in the long term appears to be associated with offspring risk of OW/OB in adolescence and adulthood [12,13]. Mamun et al. reported that, over two decades of postpartum follow-up, change in maternal weight category from normal to OW/OB was associated with 1.72 ( 95%; CI: 1.20-2.47) odds of their adult offspring being obese [13]. Another study, with 16 years of follow-up of adolescents and their parents after birth, found that mothers who become overweight had offspring with a higher risk of overweight in adolescence compared to mothers who maintained a normal BMI, even after adjustment for parental age and education level [12]. These studies have examined the long-term impact of maternal BMI change in the OW/OB categories on the risk of offspring OW/OB. Whether, and to what extent, the change in weight and BMI from pre-pregnancy to long-term postpartum are independently associated with offspring weight, BMI and WC are unknown.

Using the MUSP cohort data, we add to the above literature by examining the associations of maternal relative weight and BMI change from pre-pregnancy to 21 years post-partum with their offspring weight, BMI and WC at 21 years. We further examine maternal weight and BMI change on the risk of adult offspring OW/OB and abdominal obesity, adjusting for maternal gestational weight and other maternal lifestyle factors during pregnancy and offspring lifestyle factors in adolescence.

Study population Data are derived from 7223 mothers and children who are participants in MUSP. This study began in 1981–1984 and includes information about mothers and their offspring. The sample was followed up at 6 months, 5 years, 14 years, and 21 years. The study is ongoing. The birth cohort study overall methodology has previously been described elsewhere [14,15]. Informed consent was gained from the participants and ethical approval was granted by ethics committees at the Mater Hospital and The University of Queensland, Australia. For the purpose of this study, complete data for 1981 maternal–offspring pairs were used and analysed. 158

Exposure Maternal postpartum weight change (PPWC) and BMI change (PPBMIC) at 21 years were used as the exposure for this study. Maternal pre-pregnancy self-reported weight during the first clinical visit (FCV) was used as the baseline measure. Also, at the FCV mother’s weight was measured in light clothing with a scale to the nearest 0.2 kg while height was measured without shoes to the nearest 0.1 cm using a portable stadiometer. The average of the two weight measurements was recorded. There was a high correlation between maternal self-reported weight and FCV weight (Pearson’s correlation coefficient = 0.95). Maternal postnatal weight and height at 21 years were also obtained. The mother’s pre-pregnancy BMI and maternal postpartum BMI at 21 years were calculated based on the BMI formula: weight (kilograms) / height (metres)2. Mother’s PPWC at 21 years was derived from subtracting pre-pregnancy weight from postnatal weight at 21 years. Change in maternal PPBMIC was calculated by subtracting pre-pregnancy BMI from maternal BMI at 21 years. To determine the relative weight and BMI change, total maternal weight or BMI change in postpartum at 21 years was divided by maternal weight or BMI at baseline (x100).

Outcomes The offspring’s weight, height and WC at 21 years were the outcomes for this study. Offspring’s weight in light clothing at 21 years was measured using a scale (Wedderburn Scales, Japan) to the nearest 0.2 kg. Similar protocols to those above were used to measure participant’s height. Offspring BMI was calculated using weight and height measurements as noted for the mothers, and categorised into a normal (<25 kg/m2) and OW/OB (>25 kg/m2) group. We combined overweight and obese groups due to small numbers in both groups. WC was measured horizontally at the umbilicus using a disposable tape with two or more measurements taken. Offspring WC was divided into two categories based on gender: without abdominal obesity (males: <102 cm; females <88 cm); with abdominal obesity (males >102 cm; females >88cm). [16]

Covariates Potential covariates that might confound associations with PPWC and PPBMIC with offspring BMI and WC were included in our regression models. These include prenatal factors such as maternal age at FCV, maternal educational attainment (incomplete secondary education, completed secondary education, and completed further or higher education), family annual income at baseline (less than A$10400 and more than A$10400), maternal smoking during pregnancy (smoker and non-smoker) and maternal gestational weight gain. Offspring early-life factors and other lifestyle factors were included: offspring birth weight (grams), duration of breastfeeding obtained at 6-month 159

follow-up, TV watching at 14 years (3 hours and more or less than 3 hours of watching TV during weekdays and weekends), sports participation at 14 years (3 hours and more spent on sports/exercise per week or less than 3 hours), and frequency of the family eating meals together (at least once a day, a few times a week, and less than that).

Statistical analysis Chi square tests were used to determine the differences between mothers and offspring who had complete and incomplete data. Mean and standard deviation were derived to describe the main measurements of mothers and offspring. Comparison of means between groups was tested by using t-test with significant P-value <0.05. Linear regression analysis was used to determine the association of PPWC and PPBMIC with offspring weight, BMI and WC in adulthood. To further illustrate the effect size of maternal postpartum absolute and relative weight and BMI change, maternal PPWC and PPBMIC were examined as categorical variables, grouped by quintiles of distributions. We used multivariate logistic regression to examine the independent association between maternal absolute and relative PPWC and PPBMIC with offspring BMI and WC at 21 years of age. Final adjusted models include child characteristics (sex, birth weight, breastfeeding, sports/exercise participation, TV watching and family meals together at 14 years), maternal and household characteristics (education attainment, household income, smoking during pregnancy). Separate analyses were done for male and female offspring. We report beta co-efficient or odd ratios and 95% confidence interval (CIs) for the outcomes. All the analysis was conducted using STATA version 12.0 (Stata Inc. Texas).

Results Incomplete information for both mothers and offspring was associated with being younger, lower maternal education and household income, single or separated status, and Asian and Aboriginal and Torres Strait Islander peoples ethnicity (Table 7.1). Data from a total of 1981 mothers and offspring with complete data were analysed. Measurements in mean (SD) at baseline and the 21- year follow-up, and maternal weight and BMI change are shown in Table 7.2. Mothers were on average 16.02 (11.75) kg and 11.27 (11.53) kg heavier at 20 years postpartum than their pre- pregnancy weight and FCV weight, respectively. An increase of 30% in mean weight and BMI from baseline was noted (pre-pregnancy and FCV). As for maternal postpartum BMI, an average increase of 4.48 (4.41) kg/m2 and 6.27 (4.49) kg/m2 from pre-pregnancy and FCV BMI was noted. There was a significant relationship between 21-year maternal weight and BMI change with offspring weight, BMI and WC in young adulthood; however, these association were weak (not shown). When we divided maternal weight and BMI relative change in quintiles, there was an 160

increasing pattern for male and female offspring to have high BMI outcomes across the quintiles (Table 7.3).

There was a significant difference in the proportion of offspring-BMI categories between quintiles of maternal relative weight and BMI change as shown in Table 7.4. Long term maternal relative PPWC and PPBMIC in the highest quintile was a significant predictor for OW/OB and abdominal obesity (BMI and WC) in adult offspring (OR: 1.47, 95% CI: 1.08–2.00 and OR: 1.50, 95% CI: 1.10–2.02, respectively). These risks did not change after adjustment for other covariates. However, this only remained significant for female offspring when we separately measured the predictors between offspring sex. The odds of female adults being OW/OB were 1.68 (95% CI: 1.08–2.60) higher when their mothers were in the highest quintile of the 21-year PPWC. Similarly, female adult offspring were 1.63 times (95% CI: 1.05–2.52) at risk of being overweight or obese when their mothers were in the highest quintile of 21-year PPBMIC, compared to offspring with mothers in the lowest quintile. Similar trends were found between maternal 21-year relative PPWC and PPBMIC and offspring abdominal obesity.

161

Table 7.1 Complete and incomplete information on maternal weight pre-pregnancy and at 21 years postpartum, and on offspring weight and WC at 21 years

Information on maternal No information on pre-pregnancy weight, maternal pre-pregnancy 21-year post-partum weight, 21-year post- Background factors P-value weight, offspring weight partum weight, offspring and WC at 21 years weight and WC at 21 years (n = 1981) (n = 5242) Maternal age at FCV 13–19 240 (12.12) 944 (18.01) <0.001 20–35 1648 (83.19) 4078 (77.79) 35+ 93 (4.69) 220 (4.20) Maternal education at FCV Did not complete secondary 312 (15.84) 993 (19.10) <0.001 education Complete secondary 1266 (64.26) 3343 (64.29) education Complete further or higher 392 (19.90) 864 (16.62) education Income at FCV A$10400 or more 1352 (71.53) 3089 (63.57) <0.001 Less than A$10400 538 (28.47) 1770 (36.43) Marital status Married 1643 (83.44) 3744 (72.11) <0.001 Living together 156 (7.92) 688 (13.25) Single 138 (7.01) 598 (11.52) Separated/divorced/widowed 32 (1.63) 162 (3.12) Origin White 1827 (95.11) 4616 (91.23) <0.001 Asian 53 (2.76) 196 (3.87) Indigenous 41 (2.13) 248 (4.90) Maternal smoking at FCV Non smoker 1291 (65.67) 3131 (60.33) <0.001 Smoker 675 (34.33) 2059 (39.67) Pre-pregnancy BMI Normal + underweight 1688 (85.21) 3911 (83.02) 0.027 Overweight 215 (10.85) 571 (12.12) Obese 78 (3.94) 229 (4.86)

P-value: significant level of difference by characteristics of participants who did and did not have information on FCV weight, 21 years post-partum weight and offspring weight and waist circumference at 21 years.

162

Table 7.2 Maternal (pre-pregnancy, first clinical visit, 21 years postpartum and changes at 21 years postpartum) and offspring (weight, BMI, WC and its categories) measurements

21 years postpartum change Maternal Measurement Pre-pregnancy FCV 21 years postpartum from pre-pregnancy

Mean (SD), N = 1981 Weight (kg) 57.90 (10.68) 73.97 (16.57) Height (cm) 162.84 (6.15) 162.65 (10.82) 162.21 (6.00) BMI (kg/m2) 21.81 (3.81) 23.60 (3.85) 28.12 (6.18) WC (cm) 87.63 (13.70)

Absolute Weight (kg) 16.02 (11.75) BMI (kg/m2) 4.48 (4.41) *Relative

Weight (%) 28.10 (19.72) BMI (%) 29.08 (19.90)

Offspring Measurement Overall Male Female

(n = 1013) (n = 968) Mean (SD)/n (%) Weight (kg) 71.27 (15.77) 76.60 (15.01) 65.70 (14.59) Height (cm) 171.95 (9.27) 178.36 (6.89) 165.25 (6.19) BMI (kg/m2) 24.05 (4.76) 24.05 (4.34) 24.06 (5.17)

Underweight and normal 1329 (67.09) 679 (67.03) 650 (67.15) Overweight 424 (21.40) 223 (22.01) 201 (20.76) Obese 228 (11.51) 111 (10.96) 117 (12.09) WC (cm) 81.48 (12.00) 84.97 (10.92) 77.83 (11.99)

Normal 1453 (73.35) 819 (80.85) 634 (65.50) Abdominal Overweight 278 (14.03) 104 (10.27) 174 (17.98) Abdominal Obesity 250 (12.62) 90 (8.88) 160 (16.53)

*Relative weight/BMI formula: (21-year weight–baseline weight)/baseline weight x100

163

Table 7.3 Distribution of the mean (SD) of offspring weight, WC and BMI at 21years by maternal relative weight and BMI change at 21 years postpartum

Offspring outcomes at 21 years, mean (SD) Relative change Weight (kg) BMI (kg/m2) WC (cm) Maternal weight from pre Overall Male Female Overall Male Female Overall Male Female – 21 years postpartum Q1 70.5 (15.0) 76.0 (13.6) 65.5 (14.4) 23.9 (4.6) 23.7 (3.8) 24.1 (2.3) 81.1 (11.4) 84.2 (9.6) 78.3 (12.1) Q2 70.1 (14.3) 76.3 (13.1) 64.5 (12.9) 23.6 (4.2) 23.7 (3.9) 23.4 (4.5) 80.3 (11.2) 84.5 (9.8) 76.5 (11.1) Q3 70.3 (16.5) 75.5 (16.0) 64.9 (15.3) 23.8 (5.0) 23.9 (4.6) 23.6 (5.5) 80.7 (12.2) 84.3 (11.1) 77.1 (12.3) Q4 71.6 (16.2) 77.4 (16.0) 64.7 (13.5) 24.0 (4.7) 24.2 (4.6) 23.9 (4.7) 81.8 (12.4) 85.7 (11.9) 77.2 (11.5) Q5 73.8 (16.5) 77.6 (15.7) 69.2 (16.4) 25.0 (5.1) 24.6 (4.6) 25.5 (5.6) 83.5 (12.5) 86.0 (11.7) 80.3 (12.7)

Maternal BMI from pre – 21 years postpartum Q1 70.5 (15.0) 75.7 (13.3) 65.7 (15.0) 23.9 (4.8) 23.7 (3.9) 24.1 (5.4) 81.2 (11.6) 84.2 (9.8) 78.3 (12.5) Q2 70.0 (14.2) 76.3 (13.6) 64.5 (12.3) 23.5 (4.1) 23.7 (3.8) 23.4 (4.3) 80.2 (10.8) 84.2 (9.5) 76.8 (10.6) Q3 70.0 (15.9) 75.8 (15.8) 63.7 (13.4) 23.7 (4.8) 24.0 (4.6) 23.3 (4.9) 80.5 (11.8) 84.6 (11.2) 76.2 (11.0) Q4 72.5 (16.6) 77.0 (15.3) 66.7 (16.5) 24.2 (5.0) 24.0 (4.5) 24.4 (5.7) 82.3 (12.9) 85.3 (11.5) 78.3 (13.5) Q5 73.6 (16.7) 78.0 (16.4) 68.3 (15.6) 24.9 (5.8) 24.7 (4.6) 25.2 (5.4) 83.3 (12.5) 86.3 (12.0) 79.8 (12.2)

164

Table 7.4 Distribution n (%) of offspring adulthood overweight/obesity and abdominal obesity by maternal weight and BMI change in quantiles

Offspring outcomes at 21 years, n (%)

BMI (kg/m2) Waist circumference (cm) Relative change Normal and Overweight Chi Male <102cm; Abdominal Chi underweight and obese square female <88cm obesity: square Male > 102cm; female > 88cm Maternal weight from Pre – 21 years post-partum Q1 267 (20.1) 129 (19.8) 346 (20.0) 50 (20.0) Q2 279 (21.0) 117 (17.9) 355 (20.5) 41 (16.4) Q3 284 (21.4) 112 (17.2) 0.001 349 (20.2) 47 (18.8) 0.209 Q4 265 (19.9) 131 (20.1) 347 (20.1) 49 (19.6) Q5 234 (17.6) 163 (25.0) 334 (19.3) 63 (25.2)

Maternal BMI from Pre – 21 years post-partum Q1 267 (20.1) 129 (19.8) 346 (20.0) 50 (20.0) Q2 279 (21.0) 117 (17.9) 356 (20.6) 40 (16.0) Q3 282 (21.2) 114 (17.5) 0.001 351 (20.3) 45 (18.0) 0.179 Q4 268 (20.2) 128 (19.6) 343 (19.8) 53 (21.2) Q5 233 (17.5) 164 (25.2) 335 (19.4) 62 (24.8)

165

Table 7.5 The OR of OW/OB (BMI) and abdominal obesity of offspring at 21 years by maternal postpartum relative weight and BMI change in quantiles at 21 years OW/OB; OR (95% CI) Relative Maternal Pre – Overall Male Female Postpartum Change Unadjusted Adjusted* Unadjusted Adjusted* Unadjusted Adjusted* Weight Q1 Reference Q2 0.87 (0.64,1.17) 0.87 (0.65,1.19) 1.09 (0.71,1.69) 1.10 (0.71,1.72) 0.70 (0.46,1.06) 0.70 (0.46,1.08) Q3 0.82 (0.60,1.11) 0.83 (0.61,1.13) 0.98 (0.63,1.50) 1.02 (0.66,1.58) 0.68 (0.44,1.05) 0.68 (0.44,1.05) Q4 1.02 (0.76,1.38) 1.03 (0.76,1.40) 1.15 (0.76,1.75) 1.22 (0.79,1.89) 0.91 (0.60,1.39) 0.89 (0.57,1.37) Q5 1.44 (1.08,1.93) 1.47 (1.08,2.00) 1.28 (0.85,1.94) 1.32 (0.85,2.05) 1.70 (1.13,2.57) 1.68 (1.08,2.60) BMI Q1 Reference Q2 0.87 (0.64,1.17) 0.87 (0.64,1.18) 0.91 (0.59,1.42) 0.91 (0.58,1.43) 0.83 (0.55,1.25) 0.82 (0.54,1.25) Q3 0.84 (0.62,1.13) 0.86 (0.63,1.17) 1.02 (0.67,1.56) 1.07 (0.70,1.66) 0.67 (0.43,1.04) 0.66 (0.42,1.03) Q4 0.99 (0.73,1.33) 1.00 (0.73,1.35) 1.00 (0.66,1.52) 1.06 (0.69,1.63) 0.98 (0.64,1.51) 0.93 (0.60,1.44) Q5 1.46 (1.09,1.95) 1.50 (1.10,2.02) 1.30 (0.87,1.96) 1.37 (0.88,2.12) 1.67 (1.10,2.52) 1.63 (1.05,2.52) Abdominal obesity Weight Q1 Reference Q2 0.80 (0.52,1.24) 0.80 (0.51,1.25) 0.91 (0.41,2.06) 0.89 (0.39,2.02) 0.75 (0.44,1.28) 0.76 (0.45,1.31) Q3 0.93 (0.61,1.43) 0.99 (0.64,1.53) 1.49 (0.72,3.08) 1.60 (0.76,3.36) 0.74 (0.43,1.27) 0.75 (0.43,1.30) Q4 0.98 (0.64,1.49) 1.00 (0.65,1.55) 1.46 (0.71,3.00) 1.32 (0.63,2.80) 0.85 (0.49,1.45) 0.83 (0.48,1.44) Q5 1.31 (0.87,1.95) 1.39 (0.91,2.14) 1.65 (0.81,3.34) 1.61 (0.76,3.41) 1.31 (0.79,2.16) 1.28 (0.75,2.19) BMI Q1 Reference Q2 0.78 (0.50,1.21) 0.77 (0.49,1.20) 1.07 (0.46,2.44) 1.05 (0.45,2.44) 0.65 (0.38,1.11) 0.66 (0.38,1.13) Q3 0.89 (0.58,1.36) 0.95 (0.61,1.48) 1.71 (0.82,3.58) 1.88 (0.89,4.00) 0.63 (0.36,1.09) 0.64 (0.36,1.12) Q4 1.07 (0.71,1.62) 1.11 (0.72,1.70) 1.56 (0.75,3.26) 1.46 (0.68,3.13) 0.99 (0.59,1.67) 0.95 (0.55,1.62) Q5 1.28 (0.86,1.91) 1.37 (0.89,2.11) 1.86 (0.91,3.84) 1.84 (0.85,3.97) 1.17 (0.71,1.93) 1.16 (0.68,1.97)

*adjusted for maternal factors (gestational weight gained, age at birth, smoking during pregnancy, education, income) and offspring birth weight and lifestyle factors at 14 years (TV watching, meal with family, sports participation)

166

Table 7.6 (Supplement): Association between maternal 21-year postpartum weight and BMI change with offspring weight, BMI and waist circumference at 21 years Weight (kg); β (95% CI) Absolute maternal pre – Overall Male Female postpartum change Unadjusted Adjusted* Unadjusted Adjusted* Unadjusted Adjusted* Weight (kg) 0.17 (0.11,0.23) 0.15 (0.09,0.20)a 0.13 (0.05,0.21) 0.13 (0.04,0.21)b 0.17 (0.09,0.25) 0.16 (0.08,0.24)a BMI (kg/m2) 0.43 (0.28,0.58) 0.36 (0.21,0.51)a 0.32 (0.11,0.52) 0.30 (0.08,0.51)b 0.43 (0.22,0.63) 0.42 (0.20,0.63)a

Body Mass Index (kg/m2) Weight (kg) 0.05 (0.03,0.07) 0.05 (0.03,0.07)a 0.04 (0.02,0.06) 0.04 (0.02,0.06)b 0.06 (0.03,0.09) 0.06 (0.03,0.09)a BMI (kg/m2) 0.14 (0.09,0.18) 0.15 (0.09,0.19)a 0.11 (0.05,0.17) 0.11 (0.05,0.17)b 0.17 (0.10,0.24 0.17 (0.09,0.24)a

Waist Circumference (cm) Weight (kg) 0.12 (0.08,0.16) 0.10 (0.06,0.15)a 0.09 (0.04,0.15) 0.09 (0.03,0.15)b 0.12 (0.05,0.18) 0.11 (0.04,0.18)b BMI (kg/m2) 0.31 (0.20,0.43) 0.27 (0.16,0.39)a 0.25 (0.10,0.40) 0.24 (0.08,0.40)b 0.30 (0.13,0.47) 0.30 (0.12,0.48)b

Weight (kg); β (95% CI) Relative maternal pre – Overall Male Female costpartum change Unadjusted Adjusted* Unadjusted Adjusted* Unadjusted Adjusted* Weight (kg) 0.06 (0.02,0.09)b 0.04 (0.01,0.07)b 0.03 (-0.01,0.08) 0.03 (-0.02,0.08) 0.05 (0.00,0.10)b 0.05 (-0.00,0.10) BMI (kg/m2) 0.05 (0.02,0.09)b 0.04 (0.01,0.08)b 0.04 (-0.01,0.08) 0.03 (-0.02,0.08) 0.05 (0.01,0.10)b 0.05 (0.00,0.10)b

Body Mass Index (kg/m2) Weight (kg) 0.02 (0.01,0.03)b 0.02 (0.01,0.03)b 0.02 (0.00,0.03)b 0.01 (-0.00,0.03) 0.02 (0.00,0.04)b 0.02 (0.00,0.04)b BMI (kg/m2) 0.02 (0.01,0.03)b 0.02 (0.01,0.03)b 0.01 (0.00,0.03)b 0.01 (-0.00, 0.03) 0.02 (0.01,0.04)b 0.02 (0.00,0.04)b

Waist Circumference (cm) Weight (kg) 0.05 (0.02,0.07)b 0.03 (0.01,0.06) 0.03 (-0.00,0.07) 0.03 (-0.00,0.07) 0.04 (0.00,0.08)b 0.04 (-0.00,0.08) BMI (kg/m2) 0.05 (0.02,0.07)b 0.03 (0.01,0.06)b 0.03 (-0.00,0.07) 0.03 (-0.00,0.07) 0.04 (0.00,0.08)b 0.04 (-0.00,0.08) a: p<0.001 b: p <0.05 *Adjusted for maternal gestational age, gestational weight gained, smoking habit during pregnancy, education, income, offspring’s sex (for overall), birth weight, offspring lifestyle at 14 years.

167

Table 7.7(Supplement) Distribution of the mean (SD) of offspring weight, waist circumference and BMI at 21years by maternal absolute weight and BMI change

Offspring outcomes at 21 years, mean (SD) Absolute Change Weight (kg) BMI (kg/m2) Waist circumference (cm) Maternal weight from pre – Overall Male Female Overall Male Female Overall Male Female 21 years post-partum Q1 69.7(14.7) 75.0(13.3) 64.7(14.2) 23.7(4.6) 23.5(3.8) 23.9(5.3) 80.5(11.3) 83.6(9.7) 77.6(11.9) Q2 69.7(14.5) 75.7(13.3) 63.8(13.1) 23.4(4.2) 23.6(3.8) 23.2(4.5) 80.0(10.9) 84.1(9.3) 76.2(10.9) Q3 68.4(14.5) 74.2(14.6) 63.5(12.4) 23.3(4.3) 23.4(4.2) 23.2(4.4) 79.5(10.8) 83.3(10.3) 76.3(10.3) Q4 72.9(16.6) 77.7(15.7) 66.7(15.7) 24.5(5.0) 24.5(4.6) 24.5(5.5) 82.8(13.0) 86.2(11.9) 78.5(13.0) Q5 75.7(17.3) 79.6(16.8) 70.8(16.7) 25.3(5.3) 24.9(4.9) 25.9(5.8) 84.5(13.1) 87.0(12.3) 81.3(13.4)

Maternal BMI from Pre – 21 y post-partum Q1 69.3(14.3) 74.7(12.9) 64.4(13.6) 23.6(4.5) 23.4(3.8) 23.7(8.1) 80.4(11.2) 83.6(9.7) 77.5(11.6) Q2 70.0(14.8) 75.9(13.4) 64.7(14.0) 23.5(4.3) 23.6(3.7) 23.4(4.8) 80.1(11.1) 84.0(9.1) 76.6(11.6) Q3 69.0(15.0) 75.0(15.2) 63.0(12.1) 23.4(4.5) 23.7(4.6) 23.1(4.3) 79.8(11.0) 83.9(10.7) 75.8(9.6) Q4 72.6(16.3) 77.2(15.6) 66.8(15.4) 24.3(4.8) 24.2(4.3) 24.5(5.3) 82.5(12.6) 85.6(11.4) 78.4(12.8) Q5 75.4(17.4) 79.7(16.7) 70.3(16.9) 25.5(5.3) 25.1(4.9) 26.0(5.8) 84.6(13.3) 87.4(12.5) 81.3(13.6)

168

25.5 26 26

25 25.5 25.5 25 25 24.5 24.5 24.5 24 24 24 23.5

23.5 23.5

21y, mean, 95% CI 95% mean, 21y, 21y,mean, 95%CI

23 23 CI 95% mean, 21y, 23

Male offspring's BMI at offspring's Male Overall offspring'sOverall BMi at 22.5 22.5 BMI at offspring's Female 22.5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

81 81 81 80 80 80 79 79 79

78 78 78

77 77 77 76 76 76 75 75 75 74 74 74 73 73 73 72 72 72 71 71 71 70 70 70 69 69 69 68 68 68 67 67 67

66 66 66

21y, mean, 95% CI 95% mean, 21y, 21y, mean, 95% CI 95% mean, 21y, 65 65 CI 95% mean, 21y, 65 64 64 64 63 63

63 at weight offspring's Male 62 62 Overall offspring's weight at weight offspring's Overall 62 61 61 at weight offspring's Female 61 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

89 88 89 88 87 88 87 86 87

86 86

85 85 84 85 84 83 84 83 83 82 82 82 81 81 81 80 80 80

79 79 79

mean, 95%CI mean, 95%CI mean, mean, 95%CI mean, 78 78 78

77 77 77

circumference at 21y, at circumference 21y, at circumference Male offspring's waist waist offspring's Male

circumference at 21y, at circumference 76 76 76 Female offspring's waist offspring's Female Overall offspring's waist offspring's Overall 75 75 75 74 74 74 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Figure 7.1(Supplement) Overall, male and female mean, 95% CI of offspring BMI, weight and WC at 21years by maternal absolute BMI change from pre-pregnancy to 21 years post partum in quantiles

169

26 26 26

25.5 25.5 25.5 25 25 25 24.5 24.5 24.5 24 24 24

23.5 23.5 23.5

21y, mean, 95% CI 95% mean, 21y, CI 95% mean, 21y,

21y, mean, 95% CI 95% mean, 21y, 23 23 23

Male offspring's BMI at offspring's Male Overall offspring's BMI at offspring's Overall 22.5 22.5 BMI at offspring's Female 22.5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

81 81 81 80 80 80 79 79 79

78

78 78 77 77 77 76 76 76 75 75 75 74 74 74 73 73 73 72 72 72 71 71 71 70 70 70 69 69 69 68 68 68 67 67 67 66 66 66

65 CI 95% mean, 21y, 65 65 21y, mean, 95% CI 95% mean, 21y,

64 CI 95% 21y, mean, at 63 64 64

Male offspring weight at weight offspring Male 63 63 62 62 weight offspring's Female 62

Overall offspring's weight at weight offspring's Overall 61 61 61 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

88 88 88 87 87 87

86 86 86

85 85 85 84 84 84 83 83 83 82 82 82 81 81 81 80 80 80

79 79 79

mean, 95%CI mean, mean, 95%CI mean, mean, 95%CI mean, 78 78 78

77 77 77

Male offspring waist offspring Male

circumference at 21y, at circumference 21y, at circumference

cirdumference 21y, at cirdumference 76 76 76 Overall offspring's waist offspring's Overall 75 75 waist offspring's Female 75 74 74 74 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Figure 7.2 (supplement) Overall, male and female mean, 95% CI of offspring BMI, weight and WC at 21years by maternal relative BMI change from pre-pregnancy to 21 years post partum in quantiles

170

Discussion In this study we have found that changes in maternal post-partum weight and BMI at the 21 year offspring follow-up are associated with their adult offspring’s BMI and WC. A linear relationship between maternal post-partum weight change in two decades with offspring BMI and WC were observed. Significant associations were obtained after adjustment for maternal gestational age, total gestational weight gain, smoking during pregnancy, education level, annual income, offspring’s sex and birth weight and offspring lifestyle at 14 years. Adjusted changes in weight and BMI at 21 years postpartum were associated with the odds of adult offspring OW/OB.

Consistent with previous studies, the odds of having overweight offspring were higher among mothers who changed weight after pregnancy [11,17,18]. However, the association was observed between maternal postpartum-weight change and offspring in childhood and adolescence [11,17]. Van Rossem et al. showed no significant risk of overweight in adolescent offspring with postpartum weight gained (high: >1.0 kg/ per year) within 14 years. They classified weight gained after pregnancy into three categories: low (<0.5 kg/per year), moderate (0.5–1.0 kg/per year) and high (> 1.0 kg/per year) [17]. Another study reported that maternal weight gain of more than 8 kg after 16 years postpartum was associated with a higher risk of overweight in adolescent offspring compared to those gaining less than 3 kg [18]. Since the cut-off points for postpartum weight gain were not available in previous studies, it is difficult to make a comparison. However, in our study relative weight and BMI change were used and categorised into quintiles. We observed that offspring of mothers in the highest quintiles of relative weight and BMI change carried the highest risks of adult OW/OB. The associations were robust after adjustment for other potential covariates. We also found that changes in maternal weight and BMI over 20 years postpartum were associated with offspring WC. No previous study has reported this finding, but a few studies have previously examined the relationship between maternal pre-pregnancy BMI and gestational weight gain with adolescent and adult offspring WC [8,19,20].These studies found a positive association between maternal adiposity with offspring adiposity later in life. Our findings suggest that prolonged maternal post-partum weight and BMI change in 21 years may be associated with adiposity in early adulthood. Abdominal adiposity in adulthood is crucial as one of the cardio-metabolic factors that may increase the risk of cardiovascular diseases and type 2 Diabetes [21,22]. Since we do not have maternal pre-pregnancy WC, we are unable to explore the maternal postpartum adiposity change and its influence on offspring adiposity.

171

Consistent with previous studies, parental weight change was independently associated with offspring weight change [3,4,23]. A review by Yavuz and his colleague, involving parents aiming to reduce obesity among their children, has shown a positive impact with reduction of weight in children in the short term [1]. Added to that, family-based obesity interventions may show promising positive outcomes over the long term, although there were no studies examine the long- term effect and parents were the primary mediator of change [24]. Parents are believed to influence children’s weight status through shared familial environment, genetics and psychological factors [25]. Consistent positive changes in parental weight may influence children to change their weight. Mothers are generally considered to have more influence than fathers on their offspring’s behaviour and lifestyle. They are more frequently in contact with their offspring, often more involved in food preparation and purchasing, and may also shape perceptions of body weight [26-28]. These are possible modifiable mechanisms that may underlie a potential causal association between maternal postpartum weight and BMI change on changes in their adult offspring’s weight, BMI and WC. We lack information on maternal physical activity and dietary patterns and were unable to adjust for these. However, the role of father should not be neglected, as fathers also contribute through genetics and shared familial environment. [28,29].

Our study has several strengths, including a relatively large sample size of longitudinal data. We managed to adjust for a range of maternal pregnancy factors (gestational age, socio- demographic factors, smoking during pregnancy, and gestational weight gained) and offspring lifestyle factors at 14 years of age. Our study also has several limitations. Firstly, we do not have maternal and offspring measures of dietary habits and physical activity patterns throughout 21 years post-partum. Secondly, we have no measures of maternal pre-pregnancy WC measurements to examine the association between maternal WC weight change and offspring WC in adulthood. Abdominal obesity measurement by WC is generally used to examine the cardiovascular risk factors. With this measurement one could demonstrate the association between early-life maternal factors and long- term risk of offspring cardiovascular disease. Finally, as noted previously, other studies have lacked serial measurements of maternal BMI over a 20-year postpartum period, and have been unable to define associations between maternal postpartum weight change and offspring BMI [13].

Conclusions The mother’s postpartum relative weight and BMI change in 21 years is independently associated with adult offspring weight, BMI and WC. In addition, this study suggests that the risk of OW/OB

172

is particularly high for adult offspring of mothers with highest BMI change in 21 years post-partum. Future family obesity interventions should consider encouraging parents, and perhaps especially mothers, to maintain their weight and BMI in the healthy range over the long term. Interventions with childbearing-age women and/or those with children rather than pregnant mothers may be feasible and reduce the risk of subsequent offspring becoming overweight or obese in adulthood.

References [1] H. M. Yavuz, M. H. van Ijzendoorn, J. Mesman, and S. van der Veek, "Interventions aimed at reducing obesity in early childhood: a meta-analysis of programs that involve parents," Journal of Child Psychology and Psychiatry, pp. n/a-n/a, 2014. [2] D. Berry, R. Sheehan, R. Heschel, K. Knafl, G. Melkus, and M. Grey, "Family-Based Interventions for Childhood Obesity: A Review," Journal of Family Nursing, vol. 10, pp. 429-449, November 1, 2004 2004. [3] K. N. Boutelle, G. Cafri, and S. J. Crow, "Parent Predictors of Child Weight Change in Family Based Behavioral Obesity Treatment," Obesity, vol. 20, pp. 1539-1543, 2012. [4] B. H. Wrotniak, L. H. Epstein, R. A. Paluch, and J. N. Roemmich, "Parent weight change as a predictor of child weight change in family-based behavioral obesity treatment," Archives of Pediatrics & Adolescent Medicine, vol. 158, pp. 342-347, 2004. [5] M. H. Fasting, T. I. Nilsen, T. L. Holmen, and T. Vik, "Changes in parental weight and smoking habits and offspring adiposity: data from the HUNT-study," International Journal of Pediatric Obesity, vol. 6, pp. e399-407, Jun 2011. [6] R. C. Whitaker, Wright, J.A., Pepe, M.S., Seidel, K.D., Dietz, W.H., "Predicting obesity in young adulthood from childhood and parental obesity," New England Journal of Medicine, vol. 337, pp. 869 - 873, 1997. [7] D. H. Heppe, J. C. Kiefte-de Jong, B. Durmus, H. A. Moll, H. Raat, A. Hofman, et al., "Parental, fetal, and infant risk factors for preschool overweight: the Generation R Study," Pediatric Research, vol. 73, pp. 120-7, Jan 2013. [8] H. Hochner, Friedlander, Y., Calderon-Margalit, R., Meiner, V., Sagy, Y., Avgil-Tsadok, M., Burger, A., Savitsky, B., Siscovick, D.S., Manor, O., "Associations of maternal prepregnancy body mass index and gestational weight gain with adult offspring cardiometabolic risk factors: The Jerusalem perinatal family follow-up study," Circulation, vol. 125, pp. 1381-9, 2012.

173

[9] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166-1175, 2011. [10] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater- university study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb 15 2007. [11] C. A. Robinson, A. K. Cohen, D. H. Rehkopf, J. Deardorff, L. Ritchie, R. T. Jayaweera, et al., "Pregnancy and post-delivery maternal weight changes and overweight in preschool children," Preventive Medicine, vol. 60, pp. 77-82, Mar 2014. [12] A. Jaaskelainen, J. Pussinen, O. Nuutinen, U. Schwab, J. Pirkola, M. Kolehmainen, et al., "Intergenerational transmission of overweight among Finnish adolescents and their parents: a 16-year follow-up study," International Journal of Obesity, vol. 35, pp. 1289-1294, Oct 2011. [13] A. Mamun, M. O’Callaghan, G. Williams, and J. Najman, "Change in maternal body mass index is associated with offspring body mass index: a 21-year prospective study," European Journal of Nutrition, vol. 52, pp. 1597-1606, 2013/09/01 2013. [14] J. Najman, W. Bor, M. O'Callaghan, G. Williams, R. Aird, and G. Shuttlewood, "Cohort Profile: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology, vol. 34, pp. 992 - 997, 2005. [15] J. M. Najman, R. Alati, W. Bor, A. Clavarino, A. Mamun, J. J. McGrath, et al., "Cohort profile update: The Mater-University of Queensland Study of Pregnancy (MUSP)," International Journal of Epidemiology vol. 44, pp. 78-78f, December 16, 2014 2014. [16] WHO, "Obesity: Preventing and Managing The Global Epidemic. Report of a WHO Consultation," 2000. [17] L. van Rossem, A. H. Wijga, U. Gehring, G. H. Koppelman, and H. A. Smit, "Maternal Gestational and Postdelivery Weight Gain and Child Weight," Pediatrics, vol. 136, pp. e1294-e1301, 2015. [18] J. Laitinen, Jääskeläinen, A., Hartikainen, A.-L., Sovio, U., Vääräsmäki, M., Pouta, A., Kaakinen, M., Järvelin, M.-R., "Maternal weight gain during the first half of pregnancy and offspring obesity at 16 years: A prospective cohort study," BJOG: An International Journal of Obstetrics and Gynaecology, vol. 119, 2012.

174

[19] R. Gaillard, E. A. Steegers, O. H. Franco, A. Hofman, and V. W. Jaddoe, "Maternal weight gain in different periods of pregnancy and childhood cardio-metabolic outcomes. The Generation R Study," International Journal of Obesity (London), vol. 39, pp. 677-85, Apr 2015. [20] R. M. Reynolds, C. Osmond, D. I. W. Phillips, and K. M. Godfrey, "Maternal BMI, Parity, and Pregnancy Weight Gain: Influences on Offspring Adiposity in Young Adulthood," Journal of Clinical Endocrinology & Metabolism, vol. 95, pp. 5365-5369, Dec 2010. [21] H. Ghandehari, V. Le, S. Kamal-Bahl, S. L. Bassin, and N. D. Wong, "Abdominal obesity and the spectrum of global cardiometabolic risks in US adults," International Journal of Obesity, vol. 33, pp. 239-248, 12/09/online 2008. [22] S. Klein, D. B. Allison, S. B. Heymsfield, D. E. Kelley, R. L. Leibel, C. Nonas, et al., "Waist Circumference and Cardiometabolic Risk: A Consensus Statement from Shaping America's Health: Association for Weight Management and Obesity Prevention; NAASO, The Obesity Society; the American Society for Nutrition; and the American Diabetes Association," Obesity, vol. 15, pp. 1061-1067, 2007. [23] H. Andriani, C. Y. Liao, and H. W. Kuo, "Parental weight changes as key predictors of child weight changes," BMC Public Health, vol. 15, p. 645, 2015. [24] M. Golan, V. Kaufman, and D. R. Shahar, "Childhood obesity treatment: targeting parents exclusively v. parents and children," British Journal of Nutrition, vol. 95, pp. 1008-1015, 2006. [25] T. V. Kral and E. M. Rauh, "Eating behaviors of children in the context of their family environment," Physiological Behaviour, vol. 100, pp. 567-73, Jul 14 2010. [26] L. L. Birch and K. K. Davison, "Family Environmental Factors Influencing The Developing Behavioral Controls Of Food Intake And Childhood Overweight," Pediatric Clinics, vol. 48, pp. 893-907, 2001. [27] A. L. Thompson, "Intergenerational impact of maternal obesity and postnatal feeding practices on pediatric obesity," Nutrition Reviews, vol. 71, pp. S55-S61, 2013. [28] J. Blissett, C. Meyer, and E. Haycraft, "Maternal and paternal controlling feeding practices with male and female children," Appetite, vol. 47, pp. 212-219, 9// 2006. [29] P. Jacobson, J. S. Torgerson, L. Sjostrom, and C. Bouchard, "Spouse resemblance in body mass index: effects on adult obesity prevalence in the offspring generation," American Journal of Epidemiology, vol. 165, pp. 101-8, Jan 1 2007.

175

Chapter 8: Discussion and conclusions

8.1 Introduction

Parental obesity is a strong early life determinant of childhood obesity, and recently it has been suggested that it influences offspring obesity in later life [1-3]. In the literature review (Chapter 2), I found relatively few cohort studies that examined the longitudinal association between parental BMI before pregnancy and offspring BMI. Thus, this thesis further investigated the contribution of parental pre-pregnancy obesity to the development of offspring obesity throughout life, particularly in childhood, adolescence and adulthood. Parental BMI scores and their categories were used to investigate these associations with offspring BMI and WC scores and categories. Two cohort studies were used to investigate the overall association between pre-pregnancy parental BMI and offspring BMI in childhood and adulthood.

Although I and my co-authors discussed specific findings in each of the result chapters (Chapter 4– 7), in this chapter I have summarised the major findings of this thesis (see Section 8.2). Likely explanations for the results are summarised in Section 8.3. Then, I briefly discuss the overall implications of the findings in terms of public health in Section 8.4 and the strengths and limitations of the thesis in Section 8.5. Finally, conclusions and recommendations for the direction of future research are in Section 8.6 and 8.7.

8.2 Main findings

1. Systematic review (Chapter 2): A cohort study design is optimal to study the relationship between parental BMI and adiposity and the development of offspring BMI and obesity.  A comprehensive search found that there is extensive publications which provide evidence about the relationship between parental BMI and offspring BMI (Chapter 2, Section 2.2). However, there were a limited number of studies that examined this relationship with pre- pregnancy parental BMI [4-9].  Together, the previous studies in this review showed that the effect size between parental and offspring BMI increases with increasing offspring age. However, there were few studies examining the longitudinal association [10-12] and there were no studies including both parents before pregnancy and offspring BMI and obesity across the life course.

176

 There were limited studies measuring the associations of adiposity for parental–offspring obesity with measures other than BMI (i.e. fat mass index, fat percentage, skinfold thickness and WC).  Although associations between parental BMI after birth and offspring BMI and obesity were found, these could not be meta-analysed due to the heterogeneity of the exposures and the differing outcome definitions used.  Through the process of systematic review and meta-analysis, the focus of this thesis was narrowed to identify associations between parental pre-pregnancy BMI and adiposity and the development of BMI and adiposity by offspring throughout the life course. 2. The aim of this thesis was to identify the association between parental BMI and obesity and offspring BMI and obesity over the life course (Chapter 4–7). These associations have been described in terms of linear regression or effect size between continuous measures (BMI and BMI z-score) and through the impact (odds ratios) of overweight and obesity divided according to paternal and maternal BMI categories and their combinations. The findings from the first study (Chapter 4) cannot be compared directly with other findings as presented in subsequent chapters (Chapter 5–7), as they were carried out in different populations and results were adjusted for different factors. Nonetheless, they will be discussed together to sum up the overall effect of parental BMI and obesity before pregnancy on offspring weight, BMI and changes from childhood to adulthood, and whether the BMI of both parents is of equal importance. Parental and offspring weight and BMI become more closely related as the child grow older, and this seems especially evident when both parents are considered. Parental BMI before pregnancy is likely to operate through a combination of biological and environmental mechanisms as further discussed in the next section. Chapter 4: Association of parental body mass index before pregnancy on infant growth and body composition: evidence from a pregnancy cohort study in Malaysia  Studies have shown that weight and parental BMI has a distinct impact on children’s growth during the first year of life; however, this association was not significant after adjustment for maternal dietary patterns and physical activity of mothers during pregnancy.  Parental BMI and obesity affects offspring BMI as early as 1 year of age. Chapter 5: Prenatal parental overweight and obesity and the risk of offspring overweight and obesity at 21 years  Our findings demonstrate that parental BMI before pregnancy is positively associated with offspring OW/OB during young adulthood and that this association is independent of a

177

range of parental, family and offspring factors, including family income, parental education, maternal gestational weight gain, maternal age at birth, offspring’s birth weight, and offspring lifestyle at 14 years (TV watching, sports participation, eating with family, and fast-food consumption).  There were no differences in parental–offspring overweight and obese associations between genders. Chapter 6: Prenatal pre-pregnancy obesity and the risk of offspring weight and body mass index change from childhood to adulthood  Parental BMI is associated with offspring BMI and weight change from childhood to adulthood (Chapter 6).  Having two parents who were OW/OB was associated with higher risks of the offspring being/maintaining OW/OB status across childhood, through adolescence and into adulthood compared to having one parent or both parents of normal weight.  I am unaware of any previous study examining these associations with offspring weight change at these three significant time points.  Parental BMI has a distinct effect on offspring BMI throughout the life course and the associations were slightly higher between maternal rather than paternal BMI; although these minor differences were not statistically significant.  Both paternal and maternal BMI before pregnancy had a higher impact on the child as it grew older than at a younger age (Chapter 6). Chapter 7: Impact of maternal anthropometric change from pre-pregnancy to 21 years postpartum on adult offspring anthropometry and central obesity  In Chapter 7, adult offspring of mothers who had the greatest change in weight and BMI showed a higher risk of OW/OB compared to offspring of mothers in other lower quintiles.  A linear relationship between maternal post-partum weight change over two decades with offspring BMI and WC were observed. Significant associations were obtained after adjustment for maternal gestational age, total gestational weight gain, smoking habit during pregnancy, education level, annual income, offspring’s sex and birth weight, and offspring lifestyle at 14 years.  Adjusted changes in weight and BMI at 20 years postpartum were associated with the odds of adult offspring being OW/OB.

178

 Although the findings were not significant for adult abdominal obesity risk, it appears that the odds of this outcome may be increased in the highest quintile of maternal postpartum relative weight and BMI change.

Overall, our findings were in accordance with previous studies that parental BMI and obesity are important determinants of offspring obesity in childhood, adolescence and adulthood [3,12,13]. Findings from this study further suggest that parental BMI contributes to offspring development of BMI when measured prior to pregnancy, and that the effect size increases with stages of life. Previous studies have documented that prenatal parental BMI and adiposity were associated with offspring BMI and adiposity at one stage of life [5,8,14-18]. Some studies have reported that there is significant increase in the effect size at different developmental stages [1,10,17]. However, in these reports offspring associations with maternal obesity were highlighted [17] and the effect size or logistic regression was conducted to predict obesity at each developmental time frame separately. Due to the long-term follow-up of the cohort used in this thesis, we were able to capture the longitudinal effect of prenatal parental BMI on overall offspring BMI development at 5, 14 and 21 years.

Few studies address associations between pre-pregnancy parental BMI and offspring BMI and WC [19,20]. Recently, a large cohort study from Western Australia found that parental pre-pregnancy BMI was the strongest determinant of offspring BMI in adulthood [1]. Our findings are in agreement with this study, confirming that parental BMI before pregnancy is independently associated with offspring BMI. We further suggest that parental BMI has a significant influence on offspring WC in adulthood after adjustment for maternal factors and other offspring early life factors (Chapter 5). Parental lifestyle choices pre-pregnancy (including smoking, diet and physical activity) may, in part, explain the parental–offspring obesity association [21-26]. To our knowledge, however, no previous study has considered whether other markers of parental nutritional status (for example dietary intake and physical activity prior to pregnancy) influence offspring nutritional status throughout life. The effect of maternal pre-pregnancy BMI has been demonstrated previously in large cohort studies such as the Amsterdam Born Children and Development (ABCD) study and the Pregnancy, Infection and Nutrition study [27,28]. Pre-pregnancy parental BMI, particularly maternal BMI, is linearly associated with infant BMI. Furthermore, our study suggests that this association is explained in part by maternal lifestyle factors during pregnancy, including dietary pattern and physical activity.

179

8.3 Biological and environmental mechanisms which may explain the associations

Multiple mechanisms have been reported to potentially explain the associations observed. These principally concern the biological and environmental factors which contribute to the BMI development in children. Although it is beyond the scope of this thesis to specifically identify which mechanisms contribute most to the observed associations, our findings provide insights into their effects through parental pre-pregnancy BMI and offspring BMI relationships.

Biological factors contributing to offspring BMI may potentially be explained through the influences of genetic, epigenetic and in utero programming. Genetically, adiposity from parents may be transferred by ‘fat genes’ to the child and associated with an increase in the child’s BMI later in life [29]. Other studies have reported that an insulin-like growth factor is imprinted and expressed from the paternal copy of the gene, and regulates foetal growth and weight gain during conception or later after birth [30]. Although genetic factors may play a significant role contributing to the increase in BMI [31-33], they cannot fully explain parental–offspring obesity associations. Recently, altered genetic expression (without change to underlying DNA sequences) due to interactions with the environment (referred to as ‘epigenetic’ effects) have also been identified as potential contributors to offspring overweight and obesity [34]. When considering the early life course determinants of the development of offspring obesity, there is a need to also consider parents and the possibility that in utero over-nutrition or ‘programming’ partly contributes to the maternal– offspring obesity link. Over-nutrition programming has been hypothesised to relate to maternal obesity, which contributes to an excessive supply of nutrients such as glucose, fat and protein, and also changes the hormonal system. Through the manipulation of in utero metabolic functions, obese mothers with excess fat, glucose and protein may expose their fetus to an ‘obesogenic environment’ in the womb. These changes and modifications are associated with fetal adiposity and subsequently increase birth weight and susceptibility to development of excess adiposity later in life. In this thesis, maternal factors were shown to be strongly associated with offspring BMI for age during the first year of life (Chapter 4). Our findings also showed that maternal nutritional status (including maternal dietary pattern and physical activity) during pregnancy and/or before pregnancy may explain this association. While adjustment of these factors showed no significant change in the observed associations, this may be due to the relatively small sample size of the cohort used. Thus, physiological influences and maternal nutritional status in early life directly affects the crucial developmental environment experienced by the fetus and this extends to the infant in early postnatal life [35-37]. Under- or over-nutrition during this early stage potentially predisposes offspring to

180

develop overweight and obesity later in life. However, using a larger cohort, which prospectively examined the link between parental BMI and offspring BMI in adulthood (Chapter 5), we found no statistically different effect size between paternal and maternal BMI on offspring outcomes. Instead, the combination of parental higher BMI status, such as both parents being OW/OB, showed higher risks compared to when both parents were normal-weight status or when only one parent was OW/OB. This is at variance with the findings from some previous studies; it was, however, consistent with other previous large cohort studies, in which parental pre-pregnancy BMI was associated with offspring BMI in childhood and adolescence [5-9, 19]. Further, we added to the current literature by demonstrating that parental prenatal BMI influences their offspring BMI in adulthood [1]. Also, we noted that the effect size increased with age (Chapter 6).

It is clear that biological and environmental interactions are not mutually exclusive, throughout the normal life cycle. The dynamics of biological factors are greatly influenced by environmental factors, depending on the specific stage of life [38]. As the child grows older, the interaction between biological and environmental factors and the relative strength of each factor may change. Parental lifestyle during early life, either before pregnancy or after birth, may shape their offspring’s lifestyle. There is increasing evidence that paternal health and behaviour before pregnancy may influence offspring outcomes [21,29,30,39,40]. In childhood, influences from both parents are highly significant. These include parental feeding practices, such as breastfeeding, introduction to food, food choices and eating habits. All of these may confound or modify the parental–offspring associations [41]. Previous studies found that breastfeeding has potential benefits in reducing the risk of obesity in later life [42-45]. However, we found that incorporating breastfeeding duration into our model did not attenuate the associations. As the child develops, dramatic changes in food preferences may be transmitted from parents to their children. Shared familial environment and an ‘obesogenic lifestyle’ may influence offspring from early childhood and persist into adulthood [46,47]. Such a pattern of excess consumption of energy-dense foods may influence childhood and later obesity independently of other genetic and intrauterine factors.

Similarly, poor health behaviours among children are likely to resemble those of their parents, such as poor diet and eating patterns [26,48,49] and sedentary life [13,50-52]. With regards to inactivity, lifestyle may be worsened due to advancements in technology available to families: electronic devices and video/computer games rather than outdoor activities, ease of commuting and access to transport which reduces energy expenditure. These behaviours may be indirectly transmitted from

181

parents to their offspring through familial shared environments. Previous studies have showed that changes in parental nutritional status have significant influences on the changes in offspring nutritional status [53-55]. However, these studies used BMI alone rather than more detailed measures for estimation of adiposity and follow-up in the long-term overall lifestyle changes of parents and offspring. There is a need for further investigation to explore the influence of parental behaviour and lifestyle on offspring behaviour and lifestyle, as potential environmental mechanisms may explain parental–offspring BMI associations. From our dataset we were able to adjust for some potentially relevant offspring lifestyle factors, such as TV watching, sports participation, eating with family, and fast food consumption at the 14-year follow-up. These factors provide a general perspective/overview of offspring lifestyle factors that may influence the link between parental– offspring BMI associations. However, our findings showed that these factors did not explain the prenatal parental and offspring BMI associations, which remained significant after adjustment for such factors (Chapters 5 and 6).

8.4 Implications

Our research has important practical implications for identifying early-life factors associated with offspring obesity throughout life: firstly, it appears that parental overweight and obesity are significant factors contributing to offspring overweight and obesity, irrespective of potential mechanisms such as in utero programming, epigenetic or shared environment. Secondly, pre- pregnancy parental BMI contributes significantly to the development of offspring BMI from childhood to adulthood, and its impact increases with age. It provides an insight into the role of parental nutritional status in early life, particularly prior to pregnancy that may predict offspring BMI and risk of overweight and obesity later in life. There is a need of a huge shift in prevention strategies [56]. When considering the optimal time for preventing obesity, our data suggests that interventions should start before pregnancy. If an obese mother loses weight before pregnancy then her baby is likely to be at less risk of obesity; although this suggestion is as yet untested in clinical or population studies. Currently, intervention studies involving lifestyle changes in pregnant women show some results in the short term [57,58] (e.g. reduced weight or slow weight gain in pregnancy), and these may potentially benefit pregnancy outcomes. Our study suggests that earlier interventions involving the parents may have more potential to improve long-term outcomes. Small changes in parental BMI in the long-term have impact on adult offspring weight status. Our findings also suggest that maintaining healthy weight gain among parents over 20 years postpartum might help to reduce the increasing impact of adult–offspring obesity, particularly among mothers. At risk 182

mothers should be educated regarding eating a healthy and balanced diet. This advice should also extend to partners (fathers), to minimise the potential risk associated with increased BMI and pregnancy outcomes. Helping parents, particularly in their child’s early life, to improve their diet and physical activity may promote improvements in their offspring’s health in the long term.

Finally, our findings add to the existing literature that in an offspring’s early life, both parents significantly influence the risk of OB/OW in later life. Future obesity interventions programs to reduce the continuing increase in obesity prevalence should focus on families with offspring who are at risk in early life [56]. Tackling adult obesity, in general, is likely to be important both for the health benefit of adult and for that of their offspring.

8.5 Strengths and limitations

Strengths

A systematic review and meta-analysis was undertaken to explore the contribution of parental BMI and adiposity on offspring BMI and adiposity. The present review was the first attempt to systematically search the substantial research on parental and offspring obesity associations, covering the period from infancy to adulthood. Addressing parents as one of the early life factors contributing to offspring obesity is important. Current obesity research has not yet defined how parents in the early life of their offspring contribute to offspring obesity through familial components as well as genetic and/or environmental determinants [59].

Data from a prospective cohort study were analysed for this thesis and were optimal for evaluating the parental–offspring association over time, although this design could not confirm causality. Two cohort study were used and both are unique in their own way. The first cohort used was the Universiti Sains Malaysia Pregnancy Cohort Study, which documents comprehensive information on parental factors and lifestyle before pregnancy and includes objectively measured anthropometric data for both parents and offspring. The second cohort used was the Mater- University Study of Pregnancy, which has longitudinal follow-up of information from childhood to adulthood. This is the first study conducted to evaluate the longitudinal association between pre- pregnancy parental BMI and offspring BMI in childhood, adolescence and adulthood.

183

To our knowledge, this was the first prospective cohort study reporting the associations of pre-birth parental BMI and offspring BMI across three important time periods of follow-up— childhood, adolescence, and adulthood. Adjustment was possible for many potential factors that may confound and mediate the associations, such as smoking during pregnancy, gestational weight gain, birth weight, and offspring lifestyles. The results of the various analyses (univariate analysis, multivariate analysis, and use imputed values for missing data) led to similar conclusions. Additionally, GEE models were used to estimate the effect size of parental factors before pregnancy on offspring outcomes.

Limitations

Longitudinal follow-up data tend to be exposed to loss to follow-up of study participants and missing values throughout the collection processes. Comparisons between complete information and incomplete information were used to report factors associated with lost to follow-up. Multiple imputation was used to estimate the relationships among variables having missing values by using known data.

BMI was used as an indirect definition of overweight and obesity. This is acceptable and widely used in large population studies and global surveillance, but it reflects both lean and fat mass, and this may contribute to bias. BMI generally provides a reasonable pragmatic indicator of OW/OB status among adults. However, at present no definition of obesity in childhood and adolescence is ideal. We did consider parental body fat percentage and skinfold thickness measured in the first cohort, due to small sample size, but no significant associations were found. For the second cohort, we incorporated offspring WC at 21 years as these data were available.

In this thesis, some details of offspring lifestyle such as dietary intake and physical activity were not available. However, we were able to incorporate offspring lifestyle factors, such as breastfeeding duration at 6-month follow-up, at 14-year follow-up, TV watching hours spent on weekends and weekdays, sports participation, eating with family, and fast-food consumption frequency. These items give a general overview of the offspring lifestyle. We did not have frequent follow-up measures of parental weight and height after birth to explore the changes of parental BMI throughout the changes of offspring BMI from childhood to adulthood. No data was available for fathers and only BMI at the 21-year follow-up was available for the mothers.

184

8.6 Conclusion

Overall, parental OW/OB have significant associations with the development of OW/OB in their offspring from childhood to adulthood. We found mother–offspring and father–offspring associations were of similar strength and direction. There was no offspring male–female differential association. Our study further demonstrated that parental pre-pregnancy nutritional status is an important predictor in early life that influences offspring overweight and obesity throughout life. Intervening to reduce the prevalence of overweight or obesity in early life by targeting parents may have potential benefits. Encouraging adults to maintain a healthy lifestyle in this challenging obesogenic environment may benefit both themselves and subsequent generations.

8.7 Future research and recommendations

 Although the USM Pregnancy Cohort Study was small in sample size, it collected validated information on parental dietary intakes and physical activity during and after pregnancy. We found that after adjusting for this information, the association between parental BMI score and offspring early growth characteristics attenuated. Replication of this analysis in large birth- cohort studies that collect this type of data would be important.  Future study designs should incorporate a longitudinal study, complete with parental lifestyle factors (dietary consumption and physical activities). Objective measures of dietary intake (micro and macronutrient) and energy expenditure (IPAQ or other tools) would be valuable to identify whether these factors influence the mechanisms that explain the parental–offspring relationship association. Perhaps we can improve our understanding of how biological and environmental factors interact with each other to achieve energy balance. This may help to identify the key mechanisms underlying the parental influence on offspring weight development.  Using other direct measures of obesity would improve the reliability of findings. We used BMI, which is an indirect measure of obesity that has been used widely in large observational studies which are population based. Using more direct measures such as bio impedance or dual energy X ray absorptiometry (DEXA) would be useful in future studies.  There is a strong argument that the prevention of obesity should start as early as possible, even before pregnancy. Although we have shown that parental per-pregnancy self-reported BMI strongly predicts offspring BMI and obesity across early childhood, adolescent and

185

adulthood, future studies with measured parental pre-pregnancy anthropometry data and its prospective relationship with offspring obesity could validate our findings.  Given the importance of paternal nutritional status, considering paternal weight status, paternal BMI and adiposity measures prior to pregnancy could be undertaken and examined.  In both the MUSP and USM studies, we do not have biological data (e.g. genetic) in pregnancy or early life; future studies that have biological data might examine different mechanisms of the association between parental and offspring obesity/adiposity.  Intervention studies focusing on parents at high risk before pregnancy and offspring long- term outcomes would be beneficial to further explore the mechanisms of parental–offspring obesity associations.

References: [1] S. R. Rath, J. A. Marsh, J. P. Newnham, K. Zhu, H. C. Atkinson, J. Mountain, et al., "Parental pre-pregnancy BMI is a dominant early-life risk factor influencing BMI of offspring in adulthood," Obesity Science & Practice, pp. 2016. [2] J. J. Reilly, J. Armstrong, A. R. Dorosty, P. M. Emmett, A. Ness, I. Rogers, et al., "Early life risk factors for obesity in childhood: cohort study," British Medical Journal, vol. 330, pp. 1357- 1359, Jun. 2005. [3] R. Cooper, E. Hyppönen, D. Berry, and C. Power, "Associations between parental and offspring adiposity up to midlife: the contribution of adult lifestyle factors in the 1958 British Birth Cohort Study," American Journal of Clinical Nutrition, vol. 92, pp. 946-953, 2010. [4] G. Davey Smith, C. Steer, S. Leary, and A. Ness, "Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC)," Archives of Disease in Childhood, vol. 92, pp. 876-880, 2007. [5] D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, et al., "Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater-university study of pregnancy and its outcomes," American Journal of Epidemiology, vol. 165, pp. 418-24, Feb 2007. [6] S. Leary, G. Davey Smith, and A. Ness, "No evidence of large differences in mother- daughter and father-son body mass index concordance in a large UK birth cohort," International Journal of Obesity, vol. 34, pp. 1191-1192, 2010.

186

[7] P. Chivers, H. Parker, M. Bulsara, L. Beilin, and B. Hands, "Parental and early childhood influences on adolescent obesity: a longitudinal study," Early Child Development and Care, vol. 182, pp. 1071-1087, 2013/04/21 2012. [8] C. Fleten, W. Nystad, H. Stigum, R. Skjærven, D. A. Lawlor, G. Davey Smith, et al., "Parent-offspring body mass index associations in the Norwegian Mother and Child Cohort Study: a family-based approach to studying the role of the intrauterine environment in childhood adiposity," American Journal of Epidemiology, vol. 176, pp. 83-92 Jul. 2012. [9] S. R. Veena, G. V. Krishnaveni, S. C. Karat, C. Osmond, and C. H. Fall, "Testing the fetal overnutrition hypothesis; the relationship of maternal and paternal adiposity to adiposity, insulin resistance and cardiovascular risk factors in Indian children," Public Health Nutrition, vol. 16, pp. 1656-66, Sep. 2013. [10] M. Kivimäki, D. A. Lawlor, G. D. Smith, M. Elovainio, M. Jokela, L. Keltikangas-Järvinen, et al., "Substantial intergenerational increases in body mass index are not explained by the fetal overnutrition hypothesis: The Cardiovascular Risk in Young Finns Study," American Journal of Clinical Nutrition, vol. 86, pp. 1509-1514, 2007. [11] J. K. Lake, C. Power, and T. J. Cole, "Child to adult body mass index in the 1958 British birth cohort: associations with parental obesity," Archives of Disease in Childhood, vol. 77, pp. 376- 381, Nov. 1997. [12] R. C. Whitaker, Wright, J.A., Pepe, M.S., Seidel, K.D., Dietz, W.H., "Predicting obesity in young adulthood from childhood and parental obesity," New England Journal of Medicine, vol. 337, pp. 869 - 873, 1997. [13] E. Kvaavik, G. S. Tell, and K. I. Klepp, "Predictors and tracking of body mass index from adolescence into adulthood - Follow-up of 18 to 20 years in the Oslo youth study," Archives of Pediatrics & Adolescent Medicine, vol. 157, pp. 1212-1218, Dec. 2003. [14] M. J. O'Callaghan, G. M. Williams, M. J. Andersen, W. Bor, and J. M. Najman, "Prediction of obesity in children at 5 years: a cohort study," Journal of Paediatric Child Health, vol. 33, pp. 311-6, Aug. 1997. [15] J. J. Reilly, E. Methven, Z. C. McDowell, B. Hacking, D. Alexander, L. Stewart, et al., "Health consequences of obesity," Archives of Disease in Childhood, vol. 88, pp. 748-752, Sep. 2003. [16] P. J. Salsberry, Reagan, P.B., "Taking the long view: the prenatal environment and early adolescent overweight," Research in Nursing and Health, vol. 30, 2007.

187

[17] B. Rooney, M. Mathiason, and C. Schauberger, "Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort," Maternal & Child Health Journal, vol. 15, pp. 1166- 1175, 2011. [18] B. Durmus, L. R. Arends, L. Ay, A. C. Hokken-Koelega, H. Raat, A. Hofman, et al., "Parental anthropometrics, early growth and the risk of overweight in pre-school children: the Generation R Study," Pediatric Obesity, vol. 8, pp. 339-50, Oct. 2013. [19] B. Patro, A. Liber, B. Zalewski, L. Poston, H. Szajewska, and B. Koletzko, "Maternal and paternal body mass index and offspring obesity: a systematic review," Annals of Nutrition Metabolism, vol. 63, pp. 32-41, 2013. [20] M. Kanciruk, J. J. Andrews, and T. Donnon, "Family history of obesity and risk of childhood overweight and obesity: A meta-analysis," International Journal of Medical Health, Pharmaceutical and Biomedicla Engineering, vol. 8, pp. 244-56, 2014. [21] J. Moss and K. Harris, "Impact of maternal and paternal preconception health on birth outcomes using prospective couples’ data in Add Health," Archives of Gynecology and Obstetrics, vol. 291, pp. 287-298, 2015/02/01 2015. [22] V. Burke, Beilin, L.J., Dunbar, D., "Family lifestyle and parental body mass index as predictors of body mass index in Australian children: A longitudinal study," International Journal of Obesity, vol. 25, 2001. [23] S. Scaglioni, M. Salvioni, and C. Galimberti, "Influence of parental attitudes in the development of children eating behaviour," British Journal of Nutrition, vol. 99 Suppl 1, pp. S22-5, Feb 2008. [24] I. K. Kim, H. J. Lee, J. H. Kang, and J. Song, "Effect of parental overweight and serum leptin levels on the manifestation of overweight in 7-year-old Korean children," Public Health Nutrition, vol. 13, pp. 384-392, Mar. 2010. [25] M. E. Cerf, "Parental high-fat programming of offspring development, health and beta- cells," Islets, vol. 3, pp. 118-20, May-Jun. 2011. [26] G. Hendrie, G. Sohonpal, K. Lange, and R. Golley, "Change in the family food environment is associated with positive dietary change in children," International Journal of Behavioural Nutrition and Physical Activity, vol. 10, p. 4, 2013. [27] I. Mesman, T. J. Roseboom, G. J. Bonsel, R. J. Gemke, M. F. van der Wal, and T. G. M. Vrijkotte, "Maternal pre-pregnancy body mass index explains infant's weight and BMI at 14 months: results from a multi-ethnic birth cohort study," Archives of Disease in Childhood, vol. 94, pp. 587-595, Aug. 2009.

188

[28] A. L. Deierlein, A. M. Siega-Riz, L. S. Adair, and A. H. Herring, "Effects of pre-pregnancy body mass index and gestational weight gain on infant anthropometric outcomes," Journal of Pediatrics, vol. 158, pp. 221-226, 2011. [29] S. N. Yazbek, S. H. Spiezio, J. H. Nadeau, and D. A. Buchner, "Ancestral paternal genotype controls body weight and food intake for multiple generations," Human Molecular Genetics, vol. 19, pp. 4134-4144, Nov. 2010. [30] L. M. E. McCowan, R. A. North, E. M. Kho, M. A. Black, E. H. Y. Chan, G. A. Dekker, et al., "Paternal contribution to small for gestational age babies: a multicenter prospective study," Obesity, vol. 19, pp. 1035-1039, May. 2011. [31] D. Albuquerque, E. Stice, R. Rodriguez-Lopez, L. Manco, and C. Nobrega, "Current review of genetics of human obesity: from molecular mechanisms to an evolutionary perspective," Molecular Genetic and Genomics, vol. 290, pp. 1191-221, Aug. 2015. [32] E. K. Speliotes, C. J. Willer, S. I. Berndt, K. L. Monda, G. Thorleifsson, A. U. Jackson, et al., "Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index," Nature Genetics, vol. 42, pp. 937-48, Nov 2010. [33] S. I. Berndt, S. Gustafsson, R. Magi, A. Ganna, E. Wheeler, M. F. Feitosa, et al., "Genome- wide meta-analysis identifies 11 new loci for anthropometric traits and provides insights into genetic architecture," Nature Genetics, vol. 45, pp. 501-12, May. 2013. [34] E. R. Kyung, P. Suzanne, and M. Jeanne, "Early Determinants of Obesity: Genetic, Epigenetic, and In Utero Influences," International Journal of Pediatrics, vol. 2012, p. 9, 2012. [35] J. M. D. Thompson, C. Wall, D. M. O. Becroft, E. Robinson, C. J. Wild, and E. A. Mitchell, "Maternal dietary patterns in pregnancy and the association with small-for-gestational-age infants," British Journal of Nutrition, vol. 103, pp. 1665-1673, Jun. 2010. [36] L. Higgins, S. L. Greenwood, M. Wareing, C. P. Sibley, and T. A. Mills, "Obesity and the placenta: A consideration of nutrient exchange mechanisms in relation to aberrant fetal growth," Placenta, vol. 32, pp. 1-7, Jan. 2011. [37] C. Murrin, A. Shrivastava, and C. C. Kelleher, "Maternal macronutrient intake during pregnancy and 5 years postpartum and associations with child weight status aged five," European Journal of Clinical Nutrition, vol. 67, pp. 670-9, Jun. 2013. [38] K. Harrison, K. K. Bost, B. A. McBride, S. M. Donovan, D. S. Grigsby-Toussaint, J. Kim, et al., "Toward a Developmental Conceptualization of Contributors to Overweight and Obesity in Childhood: The Six-Cs Model," Child Development Perspectives, vol. 5, pp. 50-58, 2011.

189

[39] Y. P. Chen, X. M. Xiao, J. Li, C. Reichetzeder, Z. N. Wang, and B. Hocher, "Paternal body mass index (BMI) is associated with offspring intrauterine growth in a gender dependent manner," Plos One, vol. 7, May. 2012. [40] N. O. McPherson, T. Fullston, R. J. Aitken, and M. Lane, "Paternal obesity, interventions, and mechanistic pathways to impaired health in offspring," Annals of Nutrition and Metabolism, vol. 64, pp. 231-238, 2014. [41] S. L. Anzman, B. Y. Rollins, and L. L. Birch, "Parental influence on children's early eating environments and obesity risk: implications for prevention," International Journal of Obesity, vol. 34, pp. 1116-1124, 07//print 2010. [42] T. Harder, R. Bergmann, G. Kallischnigg, and A. Plagemann, "Duration of breastfeeding and risk of overweight: a meta-analysis," American Journal of Epidemiology, vol. 162, pp. 397-403, Sep. 2005. [43] A. A. Mamun, M. J. O’Callaghan, G. M. Williams, J. M. Najman, L. Callaway, and H. D. McIntyre, "Breastfeeding is protective to diabetes risk in young adults: a longitudinal study," Acta Diabetologica, vol. 52, pp. 837-844, 2014. [44] W. H. Oddy, T. A. Mori, R.-C. Huang, J. A. Marsh, C. E. Pennell, P. T. Chivers, et al., "Early infant feeding and adiposity risk: from infancy to adulthood," Annals of Nutrition and Metabolism, vol. 64, pp. 262-270, 2014. [45] D. S. Portela, T. O. Vieira, S. M. A. Matos, N. F. de Oliveira, and G. O. Vieira, "Maternal obesity, environmental factors, cesarean delivery and breastfeeding as determinants of overweight and obesity in children: results from a cohort," BMC Pregnancy and , vol. 15, Apr. 2015. [46] T. V. Kral and E. M. Rauh, "Eating behaviors of children in the context of their family environment," Physiological Behavior, vol. 100, pp. 567-73, Jul. 2010. [47] R. Patel, R. M. Martin, M. S. Kramer, E. Oken, N. Bogdanovich, L. Matush, et al., "Familial associations of adiposity: findings from a cross-sectional study of 12,181 parental-offspring trios from Belarus," Plos One, vol. 6, Jan. 2011. [48] H. R. Clark, E. Goyder, P. Bissell, L. Blank, and J. Peters, "How do parents' child-feeding behaviours influence child weight? Implications for childhood obesity policy," Journal of Public Health, vol. 29, pp. 132-141, Jun. 2007. [49] L. L. Birch and K. K. Davison, "Family environmental factors influencing the developing behavioral controls of food intake and childhood overweight," Pediatric Clinics, vol. 48, pp. 893- 907, 2001.

190

[50] C. Maffeis, G. Talamini, and L. Tato, "Influence of diet, physical activity and parents' obesity on children's adiposity: a four-year longitudinal study," International Journal Obesity Related Metabolic Disorders, vol. 22, pp. 758-64, Aug. 1998. [51] I. Ornelas, K. Perreira, and G. Ayala, "Parental influences on adolescent physical activity: a longitudinal study," International Journal of Behavioral Nutrition and Physical Activity, vol. 4, p. 3, 2007. [52] M. Sotos-Prieto, "Parental and self-reported dietary and physical activity habits in pre- school children and their socio-economic determinants," Public Health Nutrition, vol. 18, pp. 275- 285, 2015. [53] B. H. Wrotniak, L. H. Epstein, R. A. Paluch, and J. N. Roemmich, "Parent weight change as a predictor of child weight change in family-based behavioral obesity treatment," Archives of Pediatrics & Adolescent Medicine, vol. 158, pp. 342-347, 2004. [54] M. H. Fasting, T. I. Nilsen, T. L. Holmen, and T. Vik, "Changes in parental weight and smoking habits and offspring adiposity: data from the HUNT-study," Internation Journal of Pediatric Obesity, vol. 6, pp. e399-407, Jun. 2011. [55] H. Andriani, C. Y. Liao, and H. W. Kuo, "Parental weight changes as key predictors of child weight changes," BMC Public Health, vol. 15, p. 645, 2015. [56] D. Haire-Joshu and R. Tabak, "Preventing obesity across generations: evidence for early life intervention," Annual Review of Public Health, vol. 37, pp. 253-71, Mar. 2016. [57] C. A. Vinter, D. M. Jensen, P. Ovesen, H. Beck-Nielsen, and J. S. Jorgensen, "The LiP (Lifestyle in Pregnancy) study: a randomized controlled trial of lifestyle intervention in 360 obese pregnant women," Diabetes Care, vol. 34, pp. 2502-7, Dec. 2011. [58] M. Tanvig, C. A. Vinter, J. S. Jørgensen, S. Wehberg, P. G. Ovesen, R. F. Lamont, et al., "Anthropometrics and body composition by dual energy X-ray in children of obese women: a follow-up of a randomized controlled trial (the Lifestyle in Pregnancy and Offspring [LiPO] study)," PloS one, vol. 9, p. e89590, 2014. [59] C. S. Fox, M. J. Pencina, N. L. Heard-Costa, P. Shrader, C. Jaquish, C. J. O’Donnell, et al., "Trends in the association of parental history of obesity over 60 Years," Obesity (Silver Spring, Md.), vol. 22, pp. 919-924, 02/06 2014.

191

Appendix A

Figure 9.1: Histogram with normal curves for continuous variables used from USM

Pregnancy Cohort in this study.

40

40

30

30

20

20

Frequency

Frequency

10

10

0 0 2 2.5 3 3.5 4 4.5 -2 0 2 4

Birth weight (kg) weight for height z-score (WHZ) at 2 months

30

40

30

20

20

Frequency

Frequency

10

10

0 0 -3 -2 -1 0 1 2 -4 -2 0 2 4 Height for age z-score (HAZ) at 2 months

weight for age z-score (WAZ) at 2 months

25

40

20

30

15

20

Frequency

Frequency

10

10

5

0 0 -2 -1 0 1 2 3 -2 -1 0 1 2 BMI for age z-score (BAZ) at 2 months Weight for height z-score at 6 months

192

40

30

30

20

20

Frequency

Frequency

10

10

0 0 -4 -2 0 2 -4 -2 0 2

Weight for age z-score (WAZ) at 6 months Height for age z-score (HAZ) at 6 months

30

20

15

20

10

Frequency

Frequency

10

5

0 0 -2 -1 0 1 2 -4 -2 0 2 4

BMI for age z-score (BAZ) at 6 months Weight for height z-score (WHZ) at 12 months

25

30

20

20

15

Frequency

Frequency

10

10

5

0 0 -3 -2 -1 0 1 2 -4 -3 -2 -1 0 1

Weight for age z-score (WAZ) at 12 months Height for age z-score (HAZ) at 12 months

30

20

Frequency

10 0 -4 -2 0 2 4 BMI for age z-score (BAZ) at 12 months

193

Appendix B List of USM Pregnancy cohort study Questionnaires item used for the study variables. Questionnaire Items variables Baseline data collection 1. Umur (age) : ...... tahun (year) Maternal age 2. Jumlah pendapatan isi rumah sebulan (monthly Household income at baseline household income): RM...... 3. Pendidikan tertinggi (Highest education level): Maternal education Tidak bersekolah (No school)

Sekolah rendah (Primary school)

Sekolah menengah (Secondary school) Sijil/Diploma (Certificate/Diploma)

Ijazah (Degree)

Ijazah sarjana/Doktor falsafah (Master degree/PhD)

4. Pendidikan tertinggi suami (husband’s highest Paternal education education level): Tidak bersekolah (No school)

Sekolah rendah (Primary school)

Sekolah menengah (Secondary school) Sijil/Diploma (Certificate/Diploma)

Ijazah (Degree)

Ijazah sarjana/Doktor falsafah (Master degree/PhD)

5. BIlangan kehamilan (para)(parity):...... Parity 6. Usia kehamilan sekarang (pregnancy’s week): Gestational week 7. Tarikh jangkaan kelahiran (Expected due Due date for birth date):...... 8. Tinggi (Height)(m): ...... Maternal height (m) 9. Berat badan sebelum mengandung (Pre- Maternal pre-pregnancy weight (kg) pregnancy weight) (kg):...... 10. Berat badan sekarang (Current weight) Maternal pregnancy weight (kg) (kg):......

194

11. Ketebalan lipatan kulit (Skinfold thickness): Maternal skinfold thickness (mm) Bahagian Bacaan Bacaan Min badan (cm) 1 2 Bacaan

Triceps

Biceps

12. Komposisi badan (Body composition)(Tanita): Maternal body composition Peratus lemak (Fat percentage):...... % Berat lemak (Fat free mass):...... kg Berat bukan lemak (Fat mass):...... kg Berat otot (Muscle mass):...... kg Peratus air dalam badan (Body water percentage):...... % Berat air dalam badan (Body water mass):...... kg BMR:...... Kcal Kadar lemak visceral (Visceral fat rate):......

13. Tinggi suami (Husband’s height) (m):...... Paternal height 14. Berat badan suami (Husband’s weight) Paternal weight (kg):...... 15. Borang Ingatan Diet 24 jam (24-hours dietary Maternal dietary intake recall form) 16. Borang IPAQ (7 days) (IPAQ form) Maternal physical activity 17. Tarikh lahir (date of birth) Offspring’s date of birth 18. Berat lahir (kg) Offspring’s birth weight 19. Panjang (cm) Offspring’s birth length 20. Ukur lilit (Circumference)cm Offspring’s circumference 2-month follow-up 21. Berat badan (birth weight)(kg) Offspring’s birth weight at 2 month 22. Panjang (length)(cm): Panjang Bacaan Bacaan Min (cm) 1 2 bacaan

Lengan atas

Lengan bawah

195

23. Ketebalan lipatan kulit (skinfold thickness): Offspring’s skinfold thickness Bahagian Bacaan Bacaan Min Badan(mm) 1 2 bacaan

Triceps

Biceps

Subscapular

Suprailiac

24. Maklumat pemakanan bayi (Infant nutrition Infant feeding practice at 2 months information): Susu ibu (breastfeeding)

Susu formula (formula milk) Brand name:......

Minuman (tanpa susu formula) (Non formula milk drinks; juice, goats milk) Brand name:......

Air kosong (Plain water)

Makanan (Food)

6 month follow-up 25. Berat badan (birth weight)(kg) Offspring’s birth weight at 6 month 26. Panjang (length)(cm): Panjang Bacaan Bacaan Min (cm) 1 2 bacaan

Lengan atas

Lengan bawah

27. Ketebalan lipatan kulit (skinfold thickness): Offspring’s skinfold thickness at 6 Bahagian Bacaan Bacaan Min months Badan(mm) 1 2 bacaan

Triceps

Biceps

Subscapular

Suprailiac

196

28. Tempoh penyusuan susu ibu, sejak lahir hingga Breastfeeding information sekarang (Breastfeeding period, since birth until now): Bulan/hari......

12-month follow-up 29. Berat badan (birth weight)(kg) Offspring’s birth weight at 12 month 30. Panjang (length)(cm): Panjang Bacaan Bacaan Min (cm) 1 2 bacaan

Lengan atas

Lengan bawah

31. Ketebalan lipatan kulit (skinfold thickness): Offspring’s skinfold thickness at 12 Bahagian Bacaan Bacaan Min Badan(mm) 1 2 bacaan

Triceps

Biceps

Subscapular

Suprailiac

32. Tempoh penyusuan susu ibu, sejak lahir hingga Breastfeeding information sekarang (Breastfeeding period, since birth until now): Bulan/hari......

197