Alcohol use has been hypothesized to be a maladaptive coping response to pain. This study examines the association between pain and heavy alcohol use over time among HIV-infected drinkers using longitudinal data from the HERMITAGE study performed in St Petersburg, Russian Federation. Participants were HIV-infected adults with recent heavy drinking and unprotected sex. The main predictor was pain in the past month that at least moderately interfered with daily living, based on a single question from the SF-12. The primary outcome was past month heavy drinking (i.e., NIAAA risky drinking criteria). General estimating equations (GEE) logistic regression was used to calculate odds ratios and 95% confidence intervals for the association between pain interference and heavy alcohol use over time (i.e., baseline, 6- and 12-months), adjusting for other covariates. Pain, substance use, and depressive symptoms were modeled as time-dependent variables; all other covariates (age, marital status, gender, education, illicit drug use and randomization group) were based on baseline values. Participants (n=699) baseline characteristics were mean age of 30 (SD±5) years, 41% female, and 22% < 9th grade education. Approximately one quarter of the sample had a CD4 cell count <200 cells/μ/l, and only 17% were on antiretroviral therapy. Nearly half (46%) reported at least moderate pain interference in the past month. At baseline, 81% were drinking heavy amounts, which decreased to 41% and 40% at 6 and 12 months, respectively. In adjusted longitudinal GEE models, pain was significantly associated with greater odds of reporting past month heavy drinking (AOR=1.34, 95% CI: 1.05-1.71, p-value=0.02). We conclude that among a cohort of HIV-infected Russian drinkers, pain that at least moderately interfered with daily living was associated with higher odds of reporting heavy drinking over time. Pain may be an important risk factor for unhealthy alcohol use in the setting of HIV.
© 2013 Published by Elsevier Inc.