Food Security and Diet Quality Improvements among At-Risk, Low-Income, Older Adults following a Free Meal Pilot Program

Abstract

Older adults are at a greater risk for food insecurity compared to the general population. This study aimed to describe changes in diet quality and food insecurity following a free meal program at low-income, older adult housing complexes. Study participants were recruited from two low-income older adult public housing complexes in Kentucky. Fifty participants were enrolled and received 1–4 free weekly meals across 3 months as part of the Meals on Wings pilot program. Surveys and a 24 h dietary recall were completed at baseline and 3 months. Participants were predominately female, 69 years of age, and Black or African American race. Food security status (FSS) and dietary quality through the Healthy Eating Index (HEI) were assessed as primary measures. Participants were placed in “at-risk” categories of the (1a) lower quartile of the HEI and (1b) lower than 51% HEI, or (2) “low” or “very low” food security status based on the USDA Household Food Security Screener (FSS). Multiple linear regression (MLR) models were conducted for HEI and FSS scores to include time, group, time*group interaction, and control of meal utilization percentage (i.e., dose). The MLR for the HEI quartile had a significant time by quartile interaction that was present for an increase in the HEI in Q1 (n = 12; 32.42 ± 5.95% to n = 6; 46.10 ± 10.62%; p < 0.0001) and a decrease in the HEI for Q4 (n = 12; 70.68 ± 7.13% to n = 9; 52.36 ± 11.57%; p < 0.0001). For those low food security participants (n = 24; 48.0%), the average food insecurity score from the 6-item USDA screener improved from 4.09 ± 1.62 at baseline to 2.63 ± 2.41 at 3 months (p = 0.0064). The MLR for the FSS had a significant group*time interaction (p = 0.0071). In our population, particularly those vulnerable with lowest dietary quality and food insecurity status, we did see improvements across the free meal pilot program. However, a small sample, limited generalizability, and limited data collection measures urge caution when using these results to extrapolate for the general population. However, the current results are promising and should encourage further investigation of the effects of meal assistance programs on the health and well-being of older adults.