Socioecological Path Analytic Model of Diet Quality among Residents in Two Urban Food Deserts.

Abstract

BACKGROUND\nDiet is critical to chronic disease prevention, yet there are persistent disparities in diet quality among Americans. The socioecological model suggests multiple factors, operating at multiple levels, influence diet quality.\n\n\nOBJECTIVE\nThe goal was to model direct and indirect relationships among healthy eating identity, perceived control of healthy eating, social support for healthy eating, food retail choice block scores, perceptions of healthy food availability, and food shopping behaviors and diet quality measured using Healthy Eating Index-2010 scores (HEI-2010) for residents living in two urban communities defined as food deserts.\n\n\nDESIGN\nA cross-sectional design was used including data collected via self-reported surveys, 24-dietary recalls, and through objective observations of food retail environments.\n\n\nPARTICIPANTS/SETTING\nData collection occurred in 2015-2016 in two low-income communities in Cleveland (n=243) and Columbus (n=244), OH.\n\n\nMAIN OUTCOME MEASURE\nHEI-2010 scores were calculated based on the average of three 24-hour dietary recalls using the Nutrition Data System for Research.\n\n\nANALYSIS\nSeparate path models, controlled for income, were run for each community. Analysis was guided by a conceptual model with 15 hypothesized direct and indirect effects on HEI-2010 scores. Associations were considered statistically significant at P<0.05 and P<0.10 because of modest sample sizes in each community.\n\n\nRESULTS\nAcross both models, significant direct effects on HEI-2010 scores included healthy eating identity (β=.295, Cleveland; β=.297, Columbus, P<0.05) and distance traveled to primary food store (β=.111, Cleveland, P<0.10; β=.175, Columbus, P<0.05). Perceptions of healthy food availability had a significant, inverse effect in the Columbus model (β=-.125, P<0.05). The models explained greater variance in HEI-2010 scores for the Columbus community compared with Cleveland (R2=.282 and R2=.152, respectively).\n\n\nCONCLUSIONS\nFindings highlight the need for tailored dietary intervention approaches even within demographically comparable communities. Interventions aimed at improving diet quality among residents living in food deserts may need to focus on enhancing healthy eating identity using culturally relevant approaches while at the same time addressing the need for transportation supports to access healthy food retailers located farther away.