Previous research suggested people who engage in non-medical prescription opioid (NMPO) use alone tend to transition to NMPO and heroin co-use. A new study attempted to shed more light on this trend by examining nationally representative data on NMPO, heroin use, and NMPO and heroin co-use among U.S. adults. Data from National Survey on Drug Use and Health (NSDUH) from 2003 – 2014 were obtained; only responses from individuals aged 18 years and older were examined for this study. Response rates for this period ranged from 82% to 91%. Based on their self-reports of past-year substance use from the 2013-2014 surveys, respondents were categorized into the following mutually exclusive categories: No opioid use, NMPO-only use, heroin-only use, or heroin co-use. Other variables on interest included sociodemographic characteristics, past-year other substance use, psychological distress, and major depressive episodes. The authors used Chi-square tests to determine differences in characteristics among co-users, compared to NMPO-only and heroin-only users. Estimated prevalence rates of each type of opioid use were calculated for each year during the study period, adjusting for potential confounders, and logistic regression was used to examine linear time trends among overall opioid use and NMPO and heroin co-use. Results indicated NMPO-only was the most common type of opioid use among the sample, with a prevalence rate of 3.76% (SE 0.09), compared to only 0.10% (SE 0.01) for heroin-only and 0.23% (SE 0.02) for co-use. Characteristics associated with higher likelihood of co-use included male sex, being younger under 34 years of age (compared to being over 50 years of age), identifying as non-Hispanic White, residing in an urban area, and not having a college degree, compared to NMPO-only (all ps < 0.01). Being unemployed, not having health insurance, and reporting other illicit drug use or psychological distress in the past year were correlated with co-use (all ps < 0.01). Compared to heroin-only users, co-users were more likely to be under 34 years of age (compared to being over 50 years of age), identify as non-Hispanic White, and report past-year psychological distress, major depressive episodes, or other illicit drug use (all ps < 0.01). No significant differences in alcohol use were observed across groups. From 2003 to 2014, the prevalence of overall opioid use among all adults decreased by 6.08% (p < 0.01) from 2003-2004 to 2013-2014, while the prevalence of NMPO-only use decreased by 4.65% (p < 0.001) among all opioid users. The prevalence of heroin-only use increased during the same period; however, this change was not significant. The relative increase in co-use from 2003-2004 to 2013-2014 was 248.17% (p < 0.0001). The highest prevalence of co-use during the study period was observed among opioid users aged 26-34 years who were unemployed and reported illicit drug use, psychological distress, or major depressive episodes. Co-use also increased among women at an above-average rate during the study period.

Take away: Among U.S. adults, prevalence rates of non-medical prescription opioid (NMPO) – only use, heroin-only use, and NMPO and heroin co-use are relatively low; however, the prevalence of co-use increased by nearly 250% among all opioid users from 2003-2004 to 2013-2014.

Mital, S., Windle, M., Cooper, H.L.F. & Crawford, N.D. (2018). Trends in non-medical prescription opioids and heroin co-use among adults, 2003–2014. Addictive Behaviors [published online ahead of print May 16, 2018] doi: 10.1016/j.addbeh.2018.05.005