The Unexpected Way Fixing Food Deserts Changes Diets

Food deserts—areas with a lack of access to healthy and affordable food—are thought by some to be risk factors for obesity. When people don’t have easy access to healthy food, the thinking goes, they can’t eat it and their diet remains poor. However, other researchers question whether a market is really what changes the diets of a neighborhood’s residents.

The U.S. has spent millions of dollars building supermarkets in food deserts since 2011, according to a new study published Monday in the journal Health Affairs. The researchers noted that despite the large investment, there’s a lack of understanding around whether supermarkets directly impact neighborhood diet and health. To look at this closer, RAND researchers turned to Pittsburgh.

The Hill District neighborhood, a predominately low-income African American community, received a new supermarket in 2013. (It hadn’t had a good one in 30 years, the researchers say). The researchers then compared the Hill District to another similar neighborhood called Homewood, which did not get a new supermarket.

In the study, the researchers looked at the effect the market had on the residents’ diets, shopping behaviors, perceived access to healthy food, neighborhood satisfaction and body mass index (BMI). These factors were measured before the supermarket arrived in 2011 and again in 2014.

In many ways, there were improvements among the Hill District residents. They consumed fewer calories from sugar, solid fat and alcohol and ate less added sugar overall. Perceptions of healthy food access also greatly improved for the people living in the Hill District neighborhood. So did their overall neighborhood satisfaction, with 66% of residents saying they were satisfied or very satisfied with their neighborhood as a place to live before the supermarket, and 80% saying so afterward.

Yet the researchers saw no notable changes in fruit, vegetable or whole grain food consumption. Nor did they see improvements related to obesity. BMI among Hill District residents had not changed significantly and there were no significant changes in obesity or overweight rates. In Homewoood, BMI, obesity and rates of overweight residents got a bit worse over the study period.

Interestingly, the researchers concluded that the changes observed in diet were not related to whether the residents in the community used the store or not. The improvements among Hill District residents were similar among people who frequented the store and people who didn’t. When the researchers looked at other factors like weekly food spending, they still couldn’t explain the improvements. The analysis doesn’t look at precisely why that is, but the study authors have some guesses.

“The most significant change in the intervention neighborhood was the opening of the new full service supermarket. There was a tremendous amount of community investment that went into seeing the store open. There was also a lot of dialogue around issues of food access and availability surrounding the store,” said study author Tamara Dubowitz, a senior policy researcher at RAND Corporation in an email to TIME. “It is possible that this, in and of itself, increased awareness among residents in the neighborhood where the store was built.”

Overall, Dubowitz says the researchers saw a great improvement in neighborhood satisfaction. “These elements of hope, neighborhood satisfaction, and investment in community could ultimately have translated into increased awareness around everyday diet,” she says. “But these are also the types of elements that are extremely difficult to capture.”

The researchers conclude that based on their findings, building new supermarkets should continue, but planners should “proceed with caution” and make sure projects are coordinated with continued research on how to impact both BMI and community behaviors and health.