Publication
Background/objectives: This study analyzed the temporal evolution of geographic clusters of obesity among employees of the French national gas and electricity company in the GAZEL cohort from 1990 to 2015. Subjects/methods: The study included 12,798 men and 4590 women (aged 40 to 72 years) followed since 1989. Spatiotemporal clusters of obesity were identified using SaTScan software, and spatial dependence was assessed with the spatial relative risk function in the SPARR package. General estimating equations accounted for correlated observations in clustered or repeated measures data. Results: SaTScan identified significant obesity clusters primarily in northern France, persisting throughout the study period. In 1990, 35.8 % of participants were in obesity hotspots, peaking at 36.8 % in 1995, then declining to 21.4 % in 2015, as confirmed by spatial relative risk analysis. Despite a stable adjusted mean body mass index, the prevalence of obesity increased, indicating more individuals crossing into obesity. Models showed alcohol and tobacco use were significantly associated with body mass index and spatial clustering, with longitudinal interactions influenced by sex and age. Conclusions: Although adjusted mean body mass index remained stable from 1990 to 2015, obesity prevalence increased, with persistent clusters in northern France and Paris. The dilution of clusters over time may be influenced by aging and retirement-induced relocations. These findings underscore the importance of monitoring spatial and temporal obesity trends within populations to address the growing obesity problem effectively.