Abstract
OBJECTIVES: This study identifies sociodemographic predictors of prescription opioid use among older adults (age 65+) during the peak decade of U.S. opioid prescription, and tests whether pain level and Medicaid coverage mediate the association between low wealth and opioid use. Predictors of prescription nonsteroidal anti-inflammatory drug (NSAID) use, and of opinions of both drug classes, are also examined. METHOD: Regressions of opioid and NSAID use on sociodemographic characteristics, pain level, and insurance type were conducted using Health and Retirement Study 2004 core and 2005 Prescription Drug Study data (n = 3,721). Mediation analyses were conducted, and user opinions of drug importance, quality, and side effects were assessed. RESULTS: Low wealth was a strong, consistent predictor of opioid use. Both pain level and Medicaid coverage significantly, but only partially, mediated this association. Net of wealth, there were no significant associations between education and use of, or opinions of, either class of drugs. DISCUSSION: Among older American adults, the poorest are disproportionately likely to have been exposed to prescription opioid analgesics. Wealth, rather than education, drove social class differences in mid-2000s opioid use. Opioid-related policies should take into account socioeconomic contributors to opioid use, and the needs and treatment histories of chronic pain patients.