Abstract
OBJECTIVE: To determine predictors of the association between being a Veteran and adult food security, as well as to examine the relation of potential covariates to this relationship. DESIGN: Data collected during 2011-2012, 2013-2014 and 2015-2016 National Health and Nutrition Examination Survey (NHANES) were pooled for analyses. Veterans (self-reported) were matched to non-Veterans on age, race/ethnicity, sex and education. Adjusted logistic regression was used to determine the odds of Veterans having high food security v. the combination of marginal, low and very low food security compared with non-Veterans. SETTING: 2011-2012, 2013-2014 and 2015-2016 NHANES. PARTICIPANTS: 1227 Veterans; 2432 non-Veterans. RESULTS: Veteran status had no effect on the proportion of food insecurities between Veterans and non-Veterans reporting high (Veterans v. non-Veteran: 79 % v. 80 %), marginal (9 % v. 8 %), low (5 % v. 6 %) and very low (8 % v. 6 %) food security (P = 0·11). However, after controlling for covariates, Veterans tended to be less likely to have high food security (OR: 0·82 (95 % CI 0·66, 1·02), P = 0·07). Further, non-Hispanic White Veterans (OR: 0·72 (95 % CI 0·55, 0·95), P = 0·02) and Veterans completing some college (OR: 0·71 (95 % CI 0·50, 0·99), P < 0·05) were significantly less likely to experience high food security compared with non-Veterans. CONCLUSION: This study supports previous research findings that after controlling for covariates, Veterans tend to be less likely to have high food security. It also highlights ethnicity and level of education as important socio-economic determinates of food security status in Veterans.