Article
For poor communities of color, weight goes up when restaurants add calories
By Lia Novotny | April 9, 2020
Research spanned diverse restaurant types and locations
Creation of the dataset used in this study was made possible through a partnership between Harvard, athenahealth, and the Robert Wood Johnson Foundation. The study looked at the time period from 2012 to 2015 and analyzed the connection between per capita calorie changes by county and changes in the body mass index (BMI) of individuals who received care in those counties. In other words, as calories per capita went up in restaurants within the county, did residents experience an increase in BMI?
The restaurants included in the study were large chain restaurants in three categories: fast food (e.g., McDonald’s), fast casual (e.g., Panera), and full service (e.g., Cheesecake Factory). According to Sara Bleich, PhD, professor of public policy at Harvard’s T.H. Chan School of Public Health and lead investigator on the study, “pretty much any big chain you can think of was included — 66 of the 100 largest revenue-generating chains in the country.”
The patient sample was 447,873 adult patients who were seen by providers using athenahealth’s EHR and whose records included BMI measurements in both 2012 and 2015. Patients were also grouped by weight category (i.e., healthy weight or overweight/obese) and by race/ethnicity. The data included patients in 207 counties, each identified as lower or higher income.
More precise and comprehensive data paints a clearer picture
Many previous studies have sought to understand the connection between patients’ food environment and obesity risk. What distinguishes this study is the quality and quantity of the data. Most earlier studies were cross-sectional, looking at population slices over time, but not following the same individuals. But the athenahealth sample is longitudinal, following the same individuals. If you follow the same person over time, you can be much more certain that the changes you see are real and not just due to differences in each study population.
In addition, similar studies often rely on self-reported height and weight measurements, whereas the athenahealth sample provided height and weight measured by a healthcare professional. According to Bleich, “having measured information allows for more precise data and more precise estimates.”
Finally, the athenahealth sample covers a huge swath of the country, encompassing 7 of the 10 most populous cities, 39 out of the 50 states, and 207 out of the 3,007 counties in the U.S. The total population in these 39 states represents 89% of the total population in the U.S. in 2012. “It really gives us a sense at the national level of the association between restaurant exposure and obesity risk,” says Bleich.
Study findings revealed differences by race/ethnicity and income
The study found that a change in restaurant calories per capita was not associated with a change in overall BMI. However, when the data were analyzed separately for each race/ethnicity and income group, it became clear that increased restaurant calories were associated with an increase in BMI among Black and Hispanic patients. For this group, a 10-percent increase in exposure to chain restaurant calories per capita was associated with a .16-percentage point increase in BMI, or .89 pounds for an average-weight woman living a in county in the 90th percentile of calorie change.
And living in counties with greater increases in per capita chain restaurant calories significantly increased BMI for Black or Hispanic adults receiving healthcare services in lower-income counties (.26 percent) and those already overweight or obese (.16 percent). This last finding doesn’t surprise Bleich, who notes, “as people move up the weight trajectory, they’re more likely to get heavier,” so it is important to track the impact on this group in particular.
Subtle changes could impact these trends for the better
“The take-home point,” according to Bleich, “is that if you’re poor or if you’re heavier, your weight is more likely to be affected by the restaurant environment around you.” So, what are the implications for public health initiatives, public policy, patient education, and other population health initiatives?
One basic thing to note is that exposure matters. Although educating people about healthy eating is important, history has shown that it is difficult for individuals to make sustainable change to their diets on their own. So, it is increasingly important for restaurants to consider the options that they are offering, to find opportunities to reduce calories in ways that are largely invisible to customers. For example, Bleich references blended burgers that are part beef and part mushroom, but have a similar texture to all-beef burgers as one small way to reduce some excess calories but not diminish the customer experience.
Another strategy Bleich suggests is to think about the default option on menus, the sides and beverages that patrons don’t even think about. She cites the example of kids’ meals that swap out juice or water for soda as the default option, “because defaults are a powerful behavioral motivator, changing the default items in restaurant meals could have an impact on the calories that people are taking in.”
Finally, given that the study findings show a disproportionate impact on lower-income and minority populations, effective public policy may have implications for health disparities as well. Bleich mentions beverage taxes as a way to effect change in lower-income populations, which tend to include heavier soda drinkers. Cities and states could also use this understanding to drive zoning regulations that govern the number and kind of restaurants in particular neighborhoods.
“Given that on a typical day about a third of Americans are eating in restaurants,” says Bleich, “I think we need to take a careful look at the restaurant environment and identify effective ways to reduce over consumption of calories.”
