LAUSR

BACKGROUND Previous studies suggest that Supplemental Nutrition Assistance Program (SNAP) participants purchase less produce than nonparticipants. Whether this is due to buying smaller amounts or to being less likely to buy any produce is unclear. Purchase patterns may also differ over the monthly distribution cycle. OBJECTIVE To examine differences in the likelihood and amounts of fruits and vegetables purchased between SNAP household compared with nonparticipant households and to determine differences in produce purchases among SNAP households at different time points in the monthly distribution cycle. DESIGN Cross-sectional. PARTICIPANTS/SETTING Data from 4708 households in the National Household Food Acquisition and Purchase Survey (April 2012 to January 2013). Participants recorded all foods acquired over 7 days. MAIN OUTCOME MEASURES Fruits and vegetables acquired over a 7-day period. STATISTICAL ANALYSES PERFORMED Weighted logistic and linear regression models adjusting for household and primary respondent characteristics were used to compare odds of purchasing fruits and vegetables and amounts purchased across 3 categories: SNAP participants, SNAP-eligible nonparticipants, and ineligible nonparticipants. SNAP participants were further subdivided according to weeks since last receiving benefits. RESULTS In adjusted analyses, SNAP participants and nonparticipants were similarly likely to purchase fruits and vegetables. However, SNAP households within a week of receiving benefits were more likely than SNAP households later in the benefit cycle to buy fruit overall, especially frozen or canned fruit, and vegetables overall, including fresh, frozen or canned, starchy, and nonstarchy vegetables (fruit odds ratio [OR] 1.68, 95% confidence interval [CI] 1.12, 2.53; vegetable OR 1.63, 95% CI 1.04, 2.55 vs households in middle of cycle). In contrast, those in the last week of the benefit cycle were less likely to purchase fruit, especially fresh fruit, and vegetables, especially fresh and nonstarchy vegetables (fruit OR 0.58, 95% CI 0.35, 0.94; vegetable OR 0.58, 95% CI 0.42, 0.79 vs. households in middle of cycle), and when they bought vegetables, they bought significantly less. CONCLUSION Considering all SNAP households together at different points in their distribution cycle masks substantial declines in purchasing fruits and vegetables over the monthly cycle.