Essays on Social Disparities in Health Among Older People in Malaysia
Total Page:16
File Type:pdf, Size:1020Kb
Essays on Social Disparities in Health among Older People in Malaysia A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy in the Faculty of Humanities 2019 Syed Zahiruddin bin Syed Zainulabidin School of Social Science Contents Abstract viii Declaration ix Copyright x Acknowledgement xi Abbreviations xii 1 Introduction 1 1.1 Why health? 4 1.2 The concept of disparity 6 1.2.1 Health disparities 7 1.2.2 Social disparities in health 8 1.2.3 Older people and disparities 10 1.3 Malaysia 12 1.3.1 Districts in Malaysia 14 1.3.2 The healthcare system 15 1.3.3 The older people in Malaysia 17 1.4 Description of chapters 19 2 Methodology 23 2.1 Theoretical framework 23 2.1.1 The spatial arrangement 24 2.2 The determinants – context and composition 25 2.3 Conceptual outline of the thesis 27 2.4 Data - The National Health and Morbidity Survey 29 2.4.1 Sampling framework 29 2.4.2 Sample size determination 30 2.4.3 Missing data 30 2.5 Analysis technique: Multilevel and spatial modelling 31 3 Social Determinants and Mental Disorders among Older People in Malaysia 34 3.1 Introduction 35 3.2 Data and methods 38 3.2.1 Data 38 3.2.2 Outcome variable: Mental disorders 38 3.2.3 Covariates 39 3.2.4 Method 40 ii 3.3 Results 41 3.4 Discussion 50 3.5 Conclusion 57 4 Unmet Cardiovascular Care Needs among Older People in Malaysia 58 4.1 Introduction 59 4.2 Methods 62 4.2.1 Data 62 4.2.2 The dependent variable: Unmet cardiovascular care needs 62 4.2.3 FRS and SCORE calculations 63 4.2.4 Covariates 63 4.2.5 Method 65 4.3 Results 66 4.4 Discussion 79 4.5 Conclusion 85 5 Does Socioeconomic Status of Older People Define Geographic Variation of Diabetes in Malaysia? 87 5.1 Introduction 89 5.2 Data and method 93 5.2.1 Data 93 5.2.2 The dependent variables/ outcomes 94 5.2.3 Covariates 94 5.2.4 Method 95 5.3 Results 97 5.4 Discussion 107 5.5 Conclusion 112 6 How Spatial Distribution Informed Us about Undiagnosed Non- Communicable Diseases Risks in a Developing Country 113 6.1 Introduction 114 6.2 Methods 118 6.2.1 Study area 118 6.2.2 Data 119 6.2.3 Outcome variable 119 6.2.4 Independent Variables 120 6.2.5 Modelling 120 6.3 Results 124 6.3.1 Descriptive and bivariate analysis 124 6.3.2 Multivariate analysis 128 6.4 Discussion 132 6.5 Conclusion 137 iii 7 Discussion 139 7.1 Key findings 140 7.2 Surprising non-significant findings 143 7.3 Implications for theory 145 7.4 Policy implications and implementations 147 7.5 Strength of the thesis 148 7.6 Future research 149 7.6.1 The importance of a multilevel perspective 150 7.6.2 A need for longitudinal data and research 151 7.6.3 Considering the spatial context 151 7.7 Concluding remarks 152 Bibliography 154 A. Appendix for Chapter 1 193 B. Appendix for Chapter 3 197 C. Appendix for Chapter 4 200 D. Appendix for Chapter 6 204 E. Syntax E1. Stata code for Chapter 3 207 E2. Stata code for Chapter 4 213 E3. Stata code for Chapter 5 220 E4. R code for Chapter 6 225 iv List of Tables 1.1 Older People (60+) in Malaysia by Ethnicity, 2000 to 2030 19 3.1 Sample characteristics of 5908 respondents in 104 districts 43 3.2 Bivariate analysis of predictor variables 45 3.3 Determinants of the mental disorders, coefficient 48 4.1 Summary statistics of the sample 68 4.2 Bivariate analysis of unadjusted coefficient, odds ratio and marginal effects 73 4.3 Determinants of unmet care, the coefficient and odds ratio 75 5.1 Descriptive analysis of the NHMS dataset 98 5.2 Bivariate logistic regression of known diabetes and undiagnosed diabetes 101 5.3 Multilevel logistic regression of known diabetes 102 5.4 Multilevel logistic regression of undiagnosed diabetes 103 6.1 Descriptive and unadjusted bivariate analysis 127 6.2 Adjusted multivariate analysis using logistic regression and INLA 129 6.3 Summary statistics of posterior fixed effects using INLA: mean, standard deviation, 95% credible interval including the median, mode, and OR based on the posterior mean 130 6.4 DIC value and posterior marginal variance for the fitted models 131 v List of Figures 1.1 Increase in the percentage of the older population aged 65 and above, from 2008 to 2050 in developed and developing countries 3 1.2 Map of Malaysia 12 1.3 The population age pyramids, 2017 and 2050 18 1.4 Old-age dependency ratio in Malaysia 19 2.1 A conceptual framework for understanding how social determinants influence health 28 3.1 The number of older people (%) by districts in 2010 42 3.2 Two-way graphs with standard errors on selected predictor Variables 44 3.3 Estimated probabilities for predictor variables 50 3.4 OADR in quartiles by districts in 2010 55 4.1 Two-way linear prediction plots of 10-year CVD risk score against respondent’s age with 95% CI 69 4.2 Box plots of FRS and SCORE by gender for selected characteristics of ethnicity, education, location and marital status 70 4.3 Median estimates of FRS and SCORE risk scores vs age groups by met and unmet care. Note: the red line indicates the threshold of the high-risk category 71 4.4 Point estimates of logit coefficients with 95% confidence intervals comparing Model 1 (baseline), Model 2 (social determinants) and Model 3 (contextual) 76 4.5 District effects ranking of unmet care needs based on random intercepts in Model 3 78 5.1 Maps of the prevalence of the known diabetes and undiagnosed diabetes among older people in 2015 based on NHMS 99 5.2 The geographic distribution of the adjusted odds ratio of known diabetes between districts at the individual and contextual level 105 5.3 The geographic distribution of the adjusted odds ratio of unknown diabetes between districts at the individual and contextual level 106 6.1 Map of Malaysia 118 6.2 Example of a neighbourhood network in a graph map in the state of Johor, Malaysia 122 6.3 Example of a binary graph for the first three nodes/ districts vi in Johor 122 6.4 The mean prevalence of undiagnosed NCDs risks among older people in the Peninsular Malaysia based on the NHMS 125 6.5 Bar plot of diagnosed and undiagnosed NCDs risks among older people by age group and gender 128 6.6 Posterior mean for the district-specific undiagnosed NCDs risks compared with the whole of the study area and posterior probability 132 vii Abstract In the context of developing countries, the health effects due to increases in the number of older people are still not fully understood. Based on the evidence in developed countries, social determinants that influence health could be used to explain the effects as individuals’ health is both a determinant and an outcome of their socioeconomic circumstances. This study contributes to existing research by investigating associations between health outcomes and social disparities that relate to non-communicable diseases (NCDs) occurrences among older people in Malaysia. In addition, the issue of unmet healthcare needs is also investigated, in which older people with NCDs were not receiving appropriate healthcare interventions that they needed. At this point, policymakers in the country should be aware that older people with greater health needs may also be those with fewer means to access healthcare. In this study, social disparities are found to be an important discourse in understanding health differences at multiple levels - individuals and also districts where they resided. These levels are key areas in understanding how individuals and contextual processes operate as well as how their effects distributed along the spatial scale. Here, the estimation models and geographical maps developed by means of multilevel and spatial regression techniques could offer policymakers with systematic approaches to track health differences due to social disparities across districts in a comparable and interpretable manner. The evidence could also help policymakers to formulate effective health policies that are recent, precise and targeted. Furthermore, the findings may be instrumental in realising multi-sectoral efforts to overcome variations and complexities of health issues among older people that are beyond the control and authority of a single health ministry. viii Declaration No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institutes of learning. ix Copyright Statement 1. The author of this thesis (including any appendices and/ or schedules to this thesis) owns certain copyright or related rights in it (the ‘Copyright’) and he has given The University of Manchester, UK certain rights to use such Copyright, including for administrative purposes. 2. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. 3. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the ‘Intellectual Property’) and any reproductions of copyright works in the thesis, for example graphs and tables (‘Reproductions’), which may be described in this thesis, may not be owned by the author and may be owned by third parties.