Ambient temperature and prevalence of obesity in the Spanish population: The Di@bet.es study

Objective The aim of this study was to examine possible associations between ambient temperature and obesity in the Spanish population using an ecological focus. Methods The Di@bet.es study is a national, cross-sectional, population-based survey of cardiometabolic risk factors and their association with lifestyle. Sample: 5,061 subjects in 100 clusters. Variables: Clinical, demographic and lifestyle survey, physical examination, and blood sampling. The mean annual temperature (°C) for each study site was collected from the Spanish National Meteorology Agency (1971-2000). Results The prevalence rates of obesity in the different geographical areas divided according to mean annual temperature quartiles were 26.9% in quartile 1 (10.4-14.5°C), 30.5% in quartile 2 (14.5-15.5°C), 32% in quartile 3 (15.5-17.8°C), and 33.6% in quartile 4 (17.8-21.3°C) (P = 0.003). Logistic regression analyses including multiple socio-demographic (age, gender, educational level, marital status) and lifestyle (physical activity, Mediterranean diet score, smoking) variables showed that, as compared with quartile 1, the odd ratios for obesity were 1.20 (1.01-1.42), 1.35 (1.12-1.61), and 1.38 (1.14-1.67) in quartiles 2, 3, and 4, respectively (P = 0.001 for difference, P