|| Checking for direct PDF access through Ovid
Despite emerging evidence of a significant adverse relationship between food insecurity and sexual risk-taking, data have been primarily derived from resource-constrained settings and HIV-negative populations. To our knowledge, this study is the first to longitudinally evaluate the relationship between food insecurity and unprotected sex among HIV-seropositive people who inject drugs [injection drug users (IDUs)] both on and not on HAART.Longitudinal analyses were restricted to HIV-positive IDUs who completed baseline and at least one follow-up visit in a prospective cohort (AIDS Care Cohort to evaluate Exposure to Survival Services, 2005–2009).We constructed a multivariate logistic model using generalized estimating equations (GEEs) to assess an independent relationship between severe food insecurity (e.g., hunger due to lack of access or means to acquire food) and unprotected vaginal/anal sex.Among 470 HIV-positive IDUs, the median age was 42 years (interquartile range 36–47) with 61% men and 39% women. The prevalence of severe food insecurity was 71%, with no differences by HAART use. Severe food insecure IDUs were marginally less likely to have a suppressed HIV-1 RNA viral load (31 vs. 39%, P = 0.099). In multivariate GEE analyses, severe food insecurity [adjusted odds ratio = 2.68, 95% confidence interval 1.49–4.82] remained independently correlated with unprotected sex among HIV-positive IDUs, controlling for age, sex/gender, married/cohabitating partner, binge drug use, homelessness, and HAART use.These findings highlight a crucial need for structural HIV interventions that incorporate targeted food assistance strategies for IDUs. Given recent evidence of poor virological response among food insecure individuals on HAART, innovative HIV care models should integrate targeted food security programs and early access to HAART.