Housing costs and food purchasing characteristics: The role of SNAP participation and SNAP purchasing power.

Housing costs account for a significant proportion of low-income households' budgets, limiting the amount of money left for food purchases. We aimed to assess 1) the association of area- and household-level housing costs with household-level food purchasing characteristics and 2) explore the moderating effect of Supplemental Nutrition Assistance Program (SNAP) participation and SNAP purchasing power in these associations. Household-level data come from the National Household Food Acquisition and Purchase Survey (FoodAPS; n = 3707 households), including the outcomes analyzed: spending per 500 kcal, total kcal purchased, and quality of food purchased. Exposures include being housing-cost burdened (i.e., spending >30 % of monthly income on housing) and the U.S. Department of Housing and Urban Development's Small Area Fair Market Rent (FMR) as a measure of area-level (i.e., zip code in metro areas, county in non-metro areas) housing costs. Moderators were SNAP participation and SNAP purchasing power (i.e., the ratio of the max SNAP benefit amount to the local cost of the Thrifty Food Plan). We ran linear regression models adjusted for household and area-level covariates. In areas with low SNAP purchasing power, a higher FMR was associated with higher spending per 500 kcal for SNAP participants (B = 0.039, 95 %CI = 0.009,0.069), eligible non-SNAP participants (B = 0.030, 95 %CI = 0.012,0.048), and ineligible non-SNAP participants (B = 0.020, 95 %CI = 0.004,0.036). In areas with high SNAP purchasing power, a higher FMR was associated with lower total kcal purchased for SNAP participants (B = -0.084, 95 %CI = -0.141,-0.028). No significant associations were identified between either area- nor household-level housing costs and quality of food purchased. SNAP participation may not fully buffer low-income households from the nutritional consequences of living in areas with high housing costs.