Effects of Inclusion of Food Purchase Restrictions and Incentives in a Food Benefit Program on Diet Quality and Food Purchasing: Results From a Randomized Trial

Abstract

BackgroundThere is interest in reshaping the Supplemental Nutrition Assistance Program (SNAP) to better support family nutrition. ObjectiveThe Grocery Assistance Program Study (GAPS) for Families evaluated the effects of prohibiting using program funds for the purchase of certain sugary foods on the nutritional quality of foods purchased and consumed by program participants. DesignA randomized experimental trial was carried out with participants randomized to one of three food benefit conditions. Baseline and follow-up measures collected included interviewer-administered 24-hour dietary recalls, food purchase receipts, food security, height, and weight. Participant/settingAdult–child dyads in households eligible for SNAP but currently not enrolled were recruited from the Minneapolis/St Paul MN metropolitan area from May of 2018 through May of 2019. A total of 293 adult–child dyads received the intervention as allocated. Of these dyads, 233 adults completed follow-up measures and met criteria for inclusion in the analytic sample, resulting in an attrition rate of 20.5%. A total of 224 children completed follow-up measures and met criteria for inclusion in the analytic sample, resulting in an attrition rate of 23.5%. InterventionParticipants were randomized to 1 of 3 conditions: restriction (not allowed to buy sugar-sweetened beverages [SSB], sweet baked goods, or candy with program funds); restriction paired with incentive (30% incentive for fruits and vegetables [FV] purchased with funds); and control (funds provided with no restrictions or incentives). Funds were provided on a 4-week cycle for 20 weeks via a study-provided debit card. Main outcome measuresThe primary outcome was the Healthy Eating Index (HEI)-2015 total score. Additional outcomes included selected HEI-2015 component scores; energy intake; food security; body weight; and purchasing of SSB, sweet baked goods, candies, fruits, and vegetables. Statistical analysisLinear regression analyses were conducted with change in the outcome regressed on treatment condition for the primary outcome analyses. ResultsNo differences were observed between conditions in change in the nutrition and food security measures examined. Purchases of SSB and sweet baked goods and candies significantly differed by experimental condition. Purchase of restricted foods was lower at follow-up in the restriction and restriction paired with incentive conditions compared with the control condition. For example, spending on SSB at follow-up was significantly lower in the restriction ($2.66/week) and restriction paired with incentive ($2.06/week) conditions in comparison with control condition ($4.44/week) (P < 0.0003 and P < 0.0001, respectively). ConclusionsThis study failed to find evidence in support of prohibiting the purchase of sugary foods with food program funds as a strategy to improve program participant nutrition, even when paired with an FV incentive. Research carried out in the context of the SNAP program is needed for a more robust evidence base.