Researchers reveal the findings of a nationwide study into the relationship between food environment, healthy eating, and nutrition

An Estimated 19 million peopleAmericans live in so-called food stricken areas, which are less likely to have access to healthy and nutritious foods. More than 32 Million PeopleLiving below the poverty line means they are restricted to eating the cheapest food, even if it is closer to healthier options. Numerous studies have shown that diet plays a significant role in early mortality and the development chronic diseases like heart disease, type 2 diabetes, and cancer.

Researchers are only beginning to understand the complex interplay between individual and community characteristics and diet and health. A multidisciplinary team of researchers from Stanford University, University of Washington, and Stanford University conducted the largest ever nationwide study on the relationship between food environments, demographics, and dietary quality in the United States. This was done with the help of a popular app for food journaling that uses smartphones. The five-year-old effort’s results were published Jan. 18, 2018. Nature CommunicationsThis should provide scientists, policymakers, and health care professionals with plenty to think about.

“Our findings show that higher access to grocery shops, lower access to fast foods, higher income, and college education are independently associated to higher consumption of fresh fruits, vegetables, lower intake of fast food and soda, as well as a lower likelihood of being overweight or obese,” said lead author Tim Althoff, UW assistant professor at the Paul G. Allen school of Computer Science & Engineering.

Althoff said that while these results are not surprising, they have been hampered by the small sample sizes, single locations, and non-uniform design of studies. Our quasi-experimental methodology, which is different from traditional epidemiological studies allowed us to examine the impact on a national scale and identify the factors that matter most.

Althoff was a doctoral candidate at Stanford. The study examined data from more then 1.1 million MyFitnessPal app users. It covered approximately 2.3 Billion food entries and more than 9,800 U.S. postal codes. This allowed for insights into how factors like access to grocery and fast food, family income level and education contribute to people’s food consumption and overall health.

The team compared the associations of these variables using data by zip code. They used data from each of four self-reported dietary outcomes that were logged between 2010 to 2016: fresh fruit, vegetable consumption, fast food, soda consumption, incidence of obesity, and body mass index.

The researchers used a matching-based approach to determine how each variable correlated with these outcomes. They divided the zip codes into treatment and controls groups, and then split each input along the median. This allowed them to compare app logs from zip codes statistically above the median (e.g., those with more then 20.3% of the population living within a half-mile of the nearest grocery shop) with those below the median.

Higher education than the median, which is 29.8% of the population who have a college degree or more, was the most positive predictor for a healthier diet. All four inputs had positive effects on dietary outcomes. However, a slightly higher percentage of overweight or obese people were associated with high family income. This is defined as income at or above $70,000.241. These results are only a small part of a complex issue that can vary from one community to the next.

“When we looked into the data further, it became apparent that the population-level results obscure significant differences in the way the food environment and socioeconomic variables correspond with dietary well-being across subpopulations,” stated co-author. Hamed NilforoshanStanford doctoral student,

Nilforoshan gave an example. He pointed out the significantly higher association between above median grocery store access and increased fruits and vegetables in zip codes that have a majority Black population. This was a 10.2% difference. The 7.4% difference was compared to zip code with a majority white non-Hispanic residents. There was only a 1.7% difference between the association between increased fruit & vegetable consumption and greater access to grocery shops.

People assume that eliminating food deserts will lead to healthier eating and that higher income and higher degrees will lead to better nutrition. These assumptions are supported by the data at the entire population level,” Jenna Hua, coauthor and founder of Million Marker Wellness, Inc., said. The complex issue of diet is complicated!”

Hua said, “While policies that improve food access, economic opportunity, and education can and do support healthy consuming, our findings strongly suggest we need to tailor interventions for communities rather than following a one-size fits all approach.”

Althoff stated that both the team’s approach as well as its findings can be used to guide future research on complex topics that have implications for individuals and entire communities.

Althoff, who is also the director for the Behavioral Data Science group, said that the study would have an impact on public health and epidemiological research methods. “Regarding the first, we showed that the growing volume and variety consumer-reported data on health can be used for public health research at unprecedented scales and granularity. We see many opportunities to conduct future research on the mechanisms that drive the disparate diet relationships among subpopulations in the U.S.

Jure Leskovec from Stanford is the senior author.