LAUSR

Taking ecological perspectives to overweight and obesity, the current study applies data mining approach to examine the association between information and social environments and regional prevalence of overweight and obesity. In particular, we focus on online search and social media data since the increasing popularity of location-based geo-targeting could be an influential source of regional differences in health information and social environment. In Study 1, we calculated the correlation between regional overweight and obesity rates with regional Google searches for a time period of 12 years (2004 to 2016). The findings showed that in regions with high overweight and obesity rates, people were looking for and obtaining information on weight-loss and diet;, but in regions with low overweight and obesity rates, people were looking for and obtaining information on fitness services and facilities. In Study 2, we analyzed and compared 4010 tweets from Houston, a city with a high overweight and obesity rate, and 3281 tweets from San Diego, a city with a low overweight and obesity rate. The tweets were collected from August 2015 to August of 2016. We analyzed the textual content of tweets by word frequency analysis and topic modeling. The findings suggest that San Diego has a social environment that focuses on fitness and combining exercising with dieting. In contrast, Houston’s social environment emphasizes dieting. The implication of these findings is that health practitioners should push a paradigm shift to a stronger focus on “healthy life” (combining exercising and dieting) in regions with high overweight and obesity rates.