The geography of obesity

Aug 30 2007

Neighborhood property values predict local obesity rates better than education or incomes, according to a study from the University of Washington being published online this week by the journal Social Science and Medicine.

The study, based on analyses of responses to a telephone survey conducted in King County by the local health department and the federal Centers for Disease Control, found six-fold disparities in obesity rates across the Seattle metropolitan area. Obesity rates reached 30 percent in the most deprived areas but were only around 5 percent in the most affluent ZIP codes.

"Obesity is an economic issue,” said Dr. Adam Drewnowski, director of the UW Center for Obesity Research and leader of the study. “Knowing more about the geography of obesity will allow us to identify the most vulnerable neighborhoods.”

Working with the local health agency, Public Health-Seattle & King County, the researchers aggregated multiple-year data from Washington state's Behavioral Risk Factor Surveillance System (BRFSS) to analyze data for more than 8,000 respondents. The Centers for Disease Control and Prevention use the same data to map rising obesity rates in the United States at the state level. However, unlike most states, Washington codes the BRFSS data by the respondents' ZIP code, which permits more detailed analyses of local obesity rates at a finer geographic scale. Other information about the ZIP code areas was provided by data from the U.S. Census.

Residential property values were used as a proxy measure of ZIP code socioeconomic status. “Incomes are not the same as assets and wealth,” said Drewnowski. “The chief financial asset for most Americans is their home.”

Area prosperity can also be a good predictor of access to healthy foods, or opportunities for exercise.

The UW study was the first to examine obesity rates by area-based indexes of poverty and wealth across a metropolitan area. Previous studies have found higher obesity rates among racial and ethnic minorities and groups of lower education and incomes. Analyses of the same BRFSS data for King County showed that obesity rates were higher for African-Americans (26 percent) than for whites (16 percent), and were higher for people with annual incomes below $15,000 (20 percent) than for those with incomes above $50,000 (15 percent), all consistent with national trends. These disparities were much lower than those dependent on ZIP codes and geographic location. The study concluded that social and economic disparities were more important in predicting obesity than previously thought.

Well-known maps of rising obesity rates in the United States, also based on BRFSS data, showed only small differences among the poorest and the richest states.

“Those maps were used to support that argument that the obesity epidemic did not discriminate," said Drewnowski. "Our research shows that geography, social class, and economic standing all play huge roles in the obesity problem. Some of the most disadvantaged areas -- those hardest hit by low income, low education, and low property values -- are also the ones most affected by the obesity epidemic."

http://www.washington.edu/