Abstract ID: 233
Authors:
Aaron Koay
Tin Tin Su
Edward Fottrell
Daniel Holman
Roshidi Ismail
Sarah Dalglish
Stephen Roberts
Audrey Prost
Affiliations:
Institute for Global Health, University College London, London, United Kingdom; South East Asia Community Observatory, Jeffrey Cheah School of Medicine and Health Sciences, Monash University Malaysia, Selangor, Malaysia; School of Sociological Studies, Politics and International Relations, University of Sheffield, Sheffield, United Kingdom.
Abstract:Background: In Malaysia, the risk of obesity varies across social strata and over the life course. Combining intersectionality examining how social structures interlock to produce compounding (dis)advantages with longitudinal modelling of body-mass index (BMI) trajectories may reveal nuanced inequalities. Objectives: To explore intersectional inequalities in BMI trajectories among adults of the South East Asia Community Observatory (SEACO) in Segamat, Malaysia. Methods: Pooling the SEACO Health Round Surveys 2013, 2018 and 2023, individuals who participated in at least two Surveys were included. Intersectional strata were constructed by cross-stratifying sex (Male, Female), ethnicity (Malay, Chinese, Indian, Orang Asli, Non-citizen), household income level (B40, M40, T20) and cohort (Silent, Baby boomer, X, Y, Z). BMI (kg/m2) trajectories were modelled using two longitudinal I-MAIHDA models: A) Null model; and B) Main effects model adjusting for the stratum variables and their interactions with age. Slope and intercept variance partition coefficients (VPCs) and proportional change in variance (PCVs) between models were calculated. Results: Among 12,941 individuals, 129 strata were constructed. Mean age was 53.3 (SD: 15.4) and mean BMI was 27.2 (SD: 5.6). BMI increased with age, peaking in middle age and declining thereafter, with younger cohorts peaking earlier in life. Intersectional inequalities are evident, e.g., Female Orang Asli consistently had higher BMI across trajectories, whereas Male Non-citizens had lower BMI. In model A, the slope and intercept VPCs were 21.09% and 9.60%, which fell to 11.19% and 2.11% in model B, corresponding to slope and intercept PCVs of 49.81% and 79.73%. This indicates higher within- than between-stratum variation, with most between-stratum variation explained by additive effects. Conclusion: While intersectional inequalities in BMI trajectories were uncovered, substantial within-stratum variation suggests the need for combining universal and targeted interventions to reduce obesity inequalities in Malaysia.
Keywords: Non-Communicable Diseases (NCDs), Obesity, Malaysia, intersectionality, health equity, multi-level modelling
