Since the introduction of the term ‘food desert’ in the 1990s, public health efforts have increasingly focused on providing mostly low-income neighborhoods with access to fresh, healthy foods. The logic is that if people live or work in an area that lacks food outlets with healthy options, they are more likely to eat foods high in fat, sugar and sodium. Ultimately this leads to higher levels of overweight and obesity for those residents and widens the health disparities gap between neighborhoods. First lady Michelle Obama has tackled food access as part of her Let’s Move campaign and 22 states now offer financing to supermarkets and other outlets willing to open stores in low-income neighborhoods. However, is there any evidence these initiatives actually make residents healthier?
The causal link between obesity and the food environment is receiving some attention in Los Angeles County after a study found virtually no link between the types of food and drinks consumed by County residents and the proximity of their home to fast-food outlets, grocery stores and convenience stores. They argue that other variables, such as cultural preferences, cost and food marketing play a larger role in obesity. Opponents of the study’s findings claim that the methodology was flawed because they examined too few variables and only focused on food access around the home and not work. They worry that the findings downplay the importance and impact of healthy food availability in L.A. neighborhoods. But, in a large city like L.A. with highly mobile residents from diverse cultures, how much does the food environment vary across neighborhoods and does it even matter? Using food establishment data from the L.A. County Public Health Department, we took a closer look at the retail food environment in Los Angeles County to examine access to healthy foods across neighborhoods and income levels.
Take a look below at the point density and choropleth maps of L.A. County. Both indicate some variability in the retail food environment between the poorest areas of L.A. in comparison to the higher-income areas of West L.A.
Now take a closer look at the scatterplot below, showing the relationship between the retail food environment index (RFEI) and median income. Although there is a slight trend upwards, there is clearly a great amount of variation. Basically, the food environment in L.A. does not vary that much across L.A. neighborhoods in ways that are systematically associated with income.
So what does this say about the debate between the study authors and other public health experts? These results are a good example of an issue we were warned about 30 years ago. Epidemiologist Geoffrey Rose pointed out that the determinants of individual cases are not always the same determinants of incidence rate for a population. He explains that if you study a risk factor that is homogeneous within a population – in L.A.’s case, the food environment – then an individual-centered study approach will fail to detect a causal link. Individual-level analysis is likely just picking up individual susceptibility. So basically, determining the causal impact of the L.A. food environment on obesity is meaningless as long as the food environment remains mostly equal across neighborhoods.
This Thanksgiving, let’s be thankful for Rose’s insights three decades earlier. With it, we can do better at public health.