Food insecurity inequities among US adults: An application of intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy (I-MAIHDA) with random effects.

Food insecurity has increased with inequities likely patterned by intersecting social identities/positions such as education and race/ethnicity. Anxiety or depression symptoms may restrict material resources and increase food insecurity. Using intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy (I-MAIHDA) via logistic regression with random effects, we examined whether inequities in persistent food insecurity based on age, gender, race/ethnicity, and education were modified by anxiety or depression symptoms. In a national cohort of adults (n = 4933) followed from 2021 to 2023, 23% experienced food insecurity, with a higher prevalence among those with symptoms of anxiety or depression (49%). In the null model, 20% of the variance in food insecurity lies in the intersectional social strata. Age, gender, race/ethnicity, and education explained 99% of the between-stratum variance among adults with symptoms of anxiety or depression and 94% among adults without symptoms. Predicted probabilities of food insecurity were highest among adults younger than 55 years, non-college-educated, identified as Black and Hispanic. Symptoms of anxiety or depression were associated with 14-38 percentage point increase in predicted food insecurity. One stratum comprising White women aged 40-54 years without a college education or symptoms of anxiety or depression experienced food insecurity that deviated from the expected average effects. Findings suggest addressing food insecurity inequities tied to social positioning using universal proportionate interventions that integrate mental health support.