Diagnosis is not enough: Selective behavior change after chronic disease onset in Thailand

Health shocks-such as new diagnoses of chronic disease-can disrupt daily life and create potential "teachable moments" for behavioral improvement. Yet, existing evidence on post-diagnosis behavior change remains mixed and largely limited to high-income countries. This study addresses a major empirical gap by examining whether chronic disease diagnoses are associated with measurable within-person changes in health behaviors among middle-aged and older adults in Thailand, a rapidly aging middle-income society. We analyzed four waves (2015-2022) of the Health, Aging, and Retirement in Thailand (HART) survey. The analytic sample comprised adults aged ≥45 years with no baseline diagnosis of cancer, diabetes, or cardiovascular disease (N = 2,891; 9,577 person-wave observations). Self-reported physician-diagnosed conditions were used to construct disease-status indicators. Behavioral outcomes included smoking, current alcohol use, sufficient physical activity (≥150 minutes/week), normal body mass index (BMI) (18.5-23), and waist obesity (sex-specific cutoffs). Individual fixed-effects linear probability models with survey-wave fixed effects and adjustment for time-varying covariates (log income, activities of daily living [ADL] score, and marital status) estimated within-person associations, with standard errors clustered at the individual level. Disease-specific models and models including interactions with sex assessed heterogeneity. In pooled analyses, the disease-status indicator was not significantly associated with smoking, alcohol use, physical activity, BMI, or waist obesity. In contrast, a diabetes diagnosis was associated with a reduction of 3.6 percentage-point in current alcohol use, and both diabetes and cancer diagnoses were associated with reductions in frequent drinking. Models including interactions with sex suggested larger diagnosis-associated reductions in selected alcohol-use outcomes among men, although statistical evidence varied by outcome and disease type. Among middle-aged and older Thai adults, chronic disease diagnoses were associated with selective, rather than comprehensive, changes in health behaviors. Diagnosis alone is insufficient to support changes across multiple health behaviors, underscoring the need for structured post-diagnosis counselling, follow-up, and gender-sensitive secondary prevention in aging middle-income populations.