Tuesday, December 9, 2014

Preventing obesity in the supermarket aisle

Anne Thorndike, MD, MPH; General Medicine Division, Department of Medicine, Massachusetts General Hospital; Assistant Professor of Medicine, Harvard Medical School.

A year ago while I was rearranging the beverage display at Compare Supermarket in Chelsea, I thought to myself:  “How did I get from medical school to the supermarket aisle?”  The complete story is long- too long for this article- but my path from internal medicine residency and primary care practice to community-based research reflects an evolution in my thinking about prevention and health. In primary care, counseling patients about diet and exercise requires time, persistence, and an understanding of behavior change.  Although I find counseling individual patients to be rewarding, I am increasingly skeptical about the effectiveness of efforts to prevent obesity and obesity-related diseases within the confines of office-based practice. 

The “epidemiological triad” is a model originally designed to describe infectious disease epidemics but can also been applied to the “obesity epidemic.”1  This model describes three factors that contribute to obesity (host, vector, and environment), and it provides a framework for developing strategies to reverse the epidemic.  Host factors are individual characteristics, including non-modifiable factors (age, genes, and gender) and modifiable factors (behaviors).  Vectors are factors that contribute to energy-dense diets and sedentary behavior, such as large portion sizes, fast-food restaurants, computers, and cars.  Environmental factors include the social, cultural, and policy environments that surround us.  In medicine, we treat “host” factors with counseling, medication, and surgery.  However, if we do not address the other factors (vectors and environments), it is unlikely that we will see significant progress in lowering the prevalence of obesity in the near future.

In 2006, I began my first venture working at the “population-level” when I became the principal investigator of the MGH Be Fit employee wellness program.  In the process of evaluating outcomes of Be Fit, it became clear that the hospital environment should support employees’ efforts to maintain healthy lifestyle habits.  I began working with MGH Food and Nutrition Services in 2009 to launch the Choose Well Eat Well program in the main cafeteria.  This program introduced traffic light labels and “choice architecture” (product placement) to promote healthier food choices.2  Our research evaluation demonstrated that the traffic light labels and choice architecture in the cafeteria resulted in healthier food choices over two years, including among a longitudinal cohort of 2,285 employees who visited the cafeteria regularly.3  We also found that employees from lower educated job types and from black and Hispanic racial/ethnic groups had the least healthy cafeteria purchases at baseline, but all groups improved healthy choices with the cafeteria program.4

Results from research in the MGH cafeteria inspired me to start thinking about testing these types of strategies in low-income communities that have the highest rates of obesity and obesity-related diseases.  The city of Chelsea is a mostly Latino community located north of Boston, and approximately half of school-age children are overweight or obese. The MGH Chelsea Healthcare Center provides primary care for approximately 30% of all Chelsea residents, and the MGH Center for Community Health Initiatives has been active in Chelsea since 1995.  In 2012, I began collaborating with Melissa Dimond, ScM, the Manager of Community Initiatives and Healthy Living at MGH Chelsea, to develop new strategies to address childhood obesity by improving the local food environment.  This collaboration has thus far resulted in two research projects (and explains how I ended up rearranging the supermarket soda display).

Anne Thorndike, MD, MPH: Figure 1

Our first project in Chelsea focused on promoting fresh produce in corner stores.  Melissa and her team conducted a comprehensive Chelsea store inventory survey in 2011 and found that 46 of the 49 stores that sold food in Chelsea had limited availability of healthy foods. One-quarter of these stores carried no produce, and half carried fewer than 4 varieties of fruits or vegetables. Based on these findings, we initiated the Chelsea Corner Store Connection research project to test whether altering the “choice architecture” of corner stores by improving the visibility, availability, and quality of fresh produce would increase fruit and vegetable purchases by customers using Women, Infants, and Children (WIC) nutrition assistance to purchase groceries. WIC is a federally-funded nutrition program for low-income pregnant and breastfeeding women, their infants, and children ages 1-5 years. In 2010, approximately 10 million people in the US participated in WIC, over half of participants were children, and 42% were Hispanic.5 In 2009, the WIC program began providing vouchers specifically for the purchase of fruits and vegetables.  We hypothesized that WIC customers would be more likely to use their fruit and vegetable vouchers in corner stores with more accessible and attractive produce.

We enrolled six WIC-approved corner stores and randomized three stores to the produce intervention. The layout of the three intervention stores was rearranged to display fresh produce prominently at the front of the store rather than on a back shelf.  We purchased new shelving, baskets, and signage for the produce displays (see photo at right). In addition, a “produce consultant” met with each of the corner store owners to provide advice about stocking and purchasing new produce and discarding fruits and vegetables that were starting to rot. To evaluate outcomes for this project, we will compare WIC sales of fruit and vegetables in the intervention and control stores using WIC redemption data obtained from the state WIC office. 

The second research project that we recently completed focused on reducing consumption of sugar-sweetened beverages by families in Chelsea.  The setting for this research was Compare Supermarket, a mid-size grocery store in the center of Chelsea that is well-known for its Latino specialty items.  The study was a randomized controlled trial testing the effectiveness of in-store traffic light labels, beverage education, and financial incentives to reduce purchases of sugar-sweetened beverages. We recruited about 200 regular customers who had at least one child at home.  Each study subject received a store loyalty card that provided her/him with a 5% discount on all store purchases and linked the individual to their purchases in our database.  After collecting baseline sales data for two months, we posted in-store traffic light shelf labels (in English and Spanish) on the 750 different beverage items sold in the store to indicate healthy (green), less healthy (yellow), and unhealthy (red) beverage choices.  Subjects randomized to the intervention group received monthly letters over five months that provided education about healthy beverage choices plus a $25 store gift card if they did not purchase red-labeled beverages during the previous month. Subjects randomized to the control group received monthly letters with general dietary guidelines and no financial incentive. 

Anne Thorndike, MD, MPH: Figure 1

The study at Compare would not have been possible without the generosity and enthusiastic support of the store owner, Alberto Calvo.  He donated the 5% discount for the loyalty card and supported our efforts to label all beverages with traffic-light labels (despite the fact that the majority were “red”).  To assess outcomes, he provided us with a daily download of store sales data throughout the entire 7-month study.  Our preliminary results indicate that over the 5 month intervention, subjects in the intervention group reduced their purchases of sugar-sweetened beverages more than subjects in the control group.  A limitation is that subjects could have purchased sugar-sweetened beverages at other grocery stores.  However in a survey at the end of the study, three-quarters of all subjects stated that they had not purchased beverages at stores other than Compare during the study.

There is no question that research in the community can be messy.  In a “real-world” setting, “real-world” problems occur (a lot). For example, two of our six corner stores lost their WIC accounts during the intervention period.  At Compare, almost one-third of subjects we enrolled did not use their loyalty card, and therefore we could not track their beverage purchases.  Nonetheless, I remain optimistic that these small pilot studies will provide important preliminary data to support future larger scale efforts to prevent obesity in the community.  This type of research is critically important to inform government agencies, non-profit organizations, and local communities about where to direct their limited resources.

So, why should physicians be involved in this kind of work?  As physicians, we have the privilege of learning directly from our patients about their personal barriers to making healthy choices, including (but not limited to): lack of knowledge, lack of resources, poor social support, cultural preferences, and neighborhood environment.  We understand that our patientswantto be healthy but often do not live or work in environments that supports this goal.  We need to convey this message to the local business communities and work with them to create healthy environments that are not only good for patients but also good for business. Preventing obesity can, and should, start in the supermarket aisle.


1. Swinburn B, Egger G. Preventive strategies against weight gain and obesity. Obes Rev. 2002;3(4):289-301.

2. Thorndike AN, Sonnenberg L, Riis J, Barraclough S, Levy DE. A 2-phase labeling and choice architecture intervention to improve healthy food and beverage choices. Am J Public Health. 2012;102(3):527-533.

3. Thorndike AN, Riis J, Sonnenberg LM, Levy DE. Traffic-light labels and choice architecture: Promoting healthy food choices. Am J Prev Med. 2014;46:143-149.

4. Levy DE, Riis J, Sonnenberg LM, Barraclough SJ, Thorndike AN. Food choices of minority and low-income employees: A cafeteria intervention. Am J Prev Med. 2012;43(3):240-248.

5. U.S. Department of Agriculture, Food and Nutrition Service, Office of Research and Analysis, WIC Participant and Program Characteristics 2010, WIC-1-PC, by Patty Connor, Susan Bartlett, Michele Mendelson, Kelly Lawrence, Katherine Wen, et al. Project Officer, Fred Lesnett Alexandria, VA: December 2011.              

Browse the Entire Newsroom Archive

Back to Top