Cannabis-only use in the USA: prevalence, demographics, use patterns, and health indicators

The prevalence of adults who consume cannabis while abstaining from other substances is not known in the USA. This study used nationally representative data to estimate the prevalence and explore the demographic characteristics, cannabis use behaviors, and self-reported health of US adults with past 30-day cannabis-only use, as compared with adults who used cannabis as well as other substances.

Data came from adults 18 years and older who responded to the 2017 or 2018 National Survey on Drug Use and Health and reported past 30-day cannabis use (n = 12,143). Cannabis-only use was defined as past 30-day cannabis use with no past 30-day use of other substances (alcohol, tobacco, other illicit substances, non-prescribed controlled medications). Weighted frequencies and 95% confidence intervals (CI) were computed for all sociodemographic and cannabis-related variables, overall, and across the two categories of cannabis consumers, stratified by age.

The prevalence of past 30-day cannabis-only use among US adults was 0.9% (95% CI: 0.8, 1.0) and varied by age (2.0% of 18-25 years old; 0.7% of 26-49 year olds, and 0.6% of those ≥ 50 years). Among adults with any past 30-day cannabis use, 8.4% (95% CI: 7.6, 9.2; n = 980) reported cannabis-only use. Age was significantly associated with past 30-day cannabis-only use, with adults 18-25 years and 26-49 years having higher odds of cannabis-only use compared with older adults. Past year cannabis dependence was lowest among cannabis-only consumers aged ≥ 50 years (0.2%; 95% CI: 0.1, 0.5) and highest among young adult cannabis and other substance consumers (16.7%; 95% CI: 15.3, 18.2). Past year prevalence of any mental illness was generally similar across cannabis use groups and by age.

The prevalence of adult cannabis-only use in the US is low - most cannabis consumers report using other substances in the past 30 days as well. While cannabis-only use among older adults is rare, it does not appear to be accompanied by a high prevalence of dependence. These findings should guide future research and policymaking.