METHODS: Using data from 18,101 individuals (aged 18-97 years) from the baseline Census 2012, Health Round 2013, and Verbal Autopsies 2012-2023 of the South East Asia Community Observatory (SEACO) health and demographic surveillance system, latent class analysis was performed on 13 chronic health conditions to identify statistically and clinically meaningful groups. Multinomial logistic regression and Cox proportional hazards regression models were conducted to investigate the adjusted association of multimorbidity patterns with the risk factors and mortality, respectively. HRQoL was analyzed by linear contrasts in conjunction with ANCOVA adjusted for baseline confounders.
RESULTS: Four distinct multimorbidity latent classes were identified: (1) relatively healthy (n = 10,640); (2) cardiometabolic diseases (n = 2428); (3) musculoskeletal, mobility and sensory disorders (n = 2391); and (4) complex multimorbidity (a group with more severe multimorbidity with combined profiles of classes 2 and 3) (n = 699). Significant variations in associations between socio-demographic characteristics and multimorbidity patterns were discovered, including age, sex, ethnicity, education level, marital status, household monthly income and employment status. The complex multimorbidity group had the lowest HRQoL across all domains compared to other groups (p