Identifying determinants of obesity in Athens, Greece through global and local statistical models.

Individual socioeconomic status is linked to obesity risk, though, less is known about the influence of an area's socio-environmental conditions on obesity/overweight prevalence. This association was investigated using obesity/overweight data collected from 2445 individuals in Athens greater area, who were randomly enrolled in ATTICA study, during 2001 to 2002, and factors related to educational and economic level, population density, immigrants and green urban areas. Thematic mapping illustrated the socio-environmental status and highlighted the obesity/overweight prevalence across the municipalities of Athens. Global statistical models revealed that high socio-environmental status, educational and economic level were inversely associated with obesity/overweight (p < 0.001). Furthermore, local Geographically Weighted Regression revealed spatial non-stationarity in the relationship between green urban areas and obesity/overweight. The findings of the study lead to the better understanding of the factors affecting obesity/overweight and may support the development of policies that target to well-being, decrease of obesity/overweight prevalence and improvement of public health.

Association of Socio-Environmental Determinants with Diabetes Prevalence in the Athens Metropolitan Area, Greece: A Spatial Analysis.

OBJECTIVES: The aim of this study was to investigate the spatial variation of diabetes in relation to the geographical variability of socio-environmental characteristics in the urban districts of Athens. METHODS: A sample of 2,445 individuals from the greater area of Athens was randomly enrolled in the ATTICA study between 2001 and 2002. Diabetes was defined according to American Diabetes Association criteria. Geographical and statistical analyses were applied to examine the relationship between diabetes prevalence and factors related to education, economic status, population density, immigrant status, and availability of urban green areas. Diabetes prevalence and socio-environmental factor mapping was based on the Geographic Information Systems (GIS) technology. Variograms and spatial quasi-Poisson regression analysis evaluated the associations of diabetes with the socio-environmental variables at the municipal level. RESULTS: According to the geographical analysis and mapping, the highest proportions of people with diabetes were found in the West sector and in one district of the East and South sector each. Regression analysis revealed that the proportion of inhabitants with higher education is negatively correlated with diabetes prevalence in the regional areas of Athens. CONCLUSIONS: The study revealed that socio-environmental status in residential areas, especially educational and economic levels, is correlated with diabetes prevalence at the aggregate level. These correlations may reflect socio-economic segregation patterns at the district level, and different prevalence rates of diabetes among individuals with higher income and educational levels.