Household Food Sourcing Patterns and Their Associations With Food Insecurity in Appalachian Ohio

Abstract

Background: Households experiencing food insecurity may use dynamic strategies to meet food needs. Yet, the relationship between household food sourcing behaviors and food security, particularly in rural settings, is understudied.

Objective: To identify food sourcing patterns and their associations with food insecurity among households in rural Appalachian Ohio during the COVID-19 pandemic.

Design: Survey data were collected from a cohort of households in Athens County, OH, in July 2020, October 2020, January 2021, and April 2021.

Participants/setting: The sample included 663 households with household food sourcing and food security information for ≥1 survey wave.

Main outcome measures: Household food sourcing patterns. Households reported the frequency with which they obtained food from various retailers and charitable sources, classified as supercenters, supermarkets, convenience stores, farmers markets, or charitable sources.

Statistical analyses: Principal component analysis was used to identify food sourcing patterns. Linear mixed models were used to assess changes in food sourcing behaviors over the study period and to determine whether food sourcing behaviors differed according to food security status.

Results: Two patterns were identified: Convenience Stores and Charitable Food and Supermarkets and Farmers Markets, not Supercenters. Relative to July 2020, alignment of households' food sourcing behaviors with the Supermarkets and Farmers Markets, not Supercenters pattern was higher in October 2020 (β .07, 95% CI .02 to .12) and alignment with the Convenience Stores and Charitable Food pattern was lower in April 2021 (β -.06, 95% CI -.11 to -.02). Compared with food-secure households, food sourcing behaviors of food-insecure households were more closely aligned with the Convenience Stores and Charitable Food pattern (β .07, 95% CI .00 to .13); no statistically significant difference in scores was observed for the Supermarkets and Farmers Markets, not Supercenters pattern (β -.07, 95% CI -.15 to .02).

Conclusions: These findings support efforts to increase access to healthy, affordable options at venues where food-insecure households may be likely to obtain food, such as convenience stores and charitable sources.

Keywords: COVID-19; Food access; Principal component analysis; Rural.