**Background:** Indonesia has experienced rapid urbanization over the past three decades, with the urban population share growing from 30.5% to 56.0% between 1990 and 2019. During the same period, the prevalence of overweight and obesity has nearly doubled. While cross-sectional studies in low- and middle-income countries have shown positive associations between urbanization and obesity, longitudinal evidence examining how changes in the built environment affect individual weight trajectories over time is scarce. This study aimed to determine whether increases in the built environment contributed to rising BMI and overweight/obesity in Indonesia using 21 years of panel data.
**Methods:** The authors analyzed four waves (1993, 2000, 2007, 2014) of the Indonesian Family Life Survey (IFLS), a national longitudinal survey covering approximately 83% of the Indonesian population. The analytic sample included 3,770 adults (2,306 women, 1,464 men) who were age 18 or older at baseline, had complete anthropometric data across all four waves, and were never underweight. Height and weight were measured by trained nurses. BMI was calculated as kg/m², and overweight/obesity was defined using the Asian-specific cutoff of BMI ≥ 23 kg/m². The built environment was measured using the Global Human Settlement Layer (GHSL) dataset, which provides the percentage of land area that is built up at approximately 300-meter resolution. Built-up values were calculated within 2km buffers around each IFLS survey cluster for 1990, 2000, and 2014, with annual growth rates used to impute values for 1993 and 2007. The authors estimated two types of longitudinal regression models: "total input" models predicting BMI and overweight/obesity from contemporaneous individual, household, and geographic factors, and "value-added" models that additionally controlled for individuals' lagged BMI from the previous wave (7 years earlier). Covariates included age, age-squared, sex, education, marital status, religion, smoking status, island of residence, and period dummies.
**Key Results:** Mean BMI rose steadily from 22.4 in 1993 to 24.8 in 2014, and the proportion classified as overweight or obese (Asian cutoff) increased from 33.5% to 61.8% over the same period. Women had higher mean BMIs than men in every panel. Average built-up percentage increased from 19.8% in 1993 to 25.8% in 2014. In total input models, a 10-percentage point increase in built-up was associated with 0.22 higher BMI (p < 0.001). The urban-rural gap in mean BMI was 1.3 points. In value-added models that controlled for lagged BMI, the association was substantially smaller but remained statistically significant: a 10-percentage point increase in built-up was associated with a 0.042 BMI point increase (Model 5, p < 0.001). Based on these estimates, the 6-percentage point average increase in built-up over 21 years contributed approximately 1.1% of the observed rise in mean BMI (0.8% in the fully adjusted model). Results for overweight/obesity showed similar patterns. The built-up measure and the binary urban-rural classification performed similarly in terms of model fit (R² ≈ 0.15 for total input models). Sex-stratified analyses showed broadly similar patterns for men and women.
**Clinical Implications:** While the contribution of the built environment to rising obesity in Indonesia was statistically significant, it was small in magnitude, accounting for roughly 1% of the observed increase. This suggests that other factors—such as nutritional transitions, rising incomes, declining physical activity, and technological changes affecting food consumption—likely play larger roles. However, as urbanization continues and built-up percentages increase beyond the modest 6-percentage point change observed in this study period, the contribution of the built environment may grow. The findings highlight the need for healthy urban planning in rapidly developing countries, including attention to walkability, green spaces, and infrastructure. The study also demonstrates the value of satellite-derived measures of built environment over simple urban/rural dichotomies, though the authors note that such measures do not fully capture how people interact with their environment. The rising obesity prevalence in Indonesia (from 33.5% to 61.8% over 21 years) underscores that overweight and obesity are becoming major public health concerns in poorer countries, with implications for future chronic disease burden.