Constantin · Addictive behaviors 2026 · Repeated cross-sectional survey study · n=6,099

Cannabis surpasses tobacco among pregnant women: national trends and predictors of use in the U.S., 2015-2023.

Cited 0 times in the scientific literature.

Level 4 - case-series / case-control

Repeated cross-sectional survey analysis of national survey data

PubMed 42143410 · doi:10.1016/j.addbeh.2026.108740 · record verified 2026-08-26

What was done

Analyzed repeated cross-sectional data from the 2015–2023 National Survey on Drug Use and Health (NSDUH) among pregnant women aged 15–44 (n = 6,099). The authors evaluated weighted prevalence trends for self-reported past-30-day tobacco-only use, cannabis-only use, and co-use across 2015–2019, 2020, and 2021–2023. Multivariable logistic regressions identified sociodemographic predictors of use patterns relative to non-use for the pre-pandemic (2015–2019) and post-pandemic (2021–2023) periods.

What was found

Tobacco-only use among pregnant individuals declined from 12.0% in 2015 to 3.9% in 2023. Cannabis-only use more than doubled pre-pandemic and peaked at 6.3% in 2022, surpassing tobacco-only use before converging in 2023, while co-use fell to 0.8% by 2023. Pre-pandemic, lower education, unmarried status, lower income, and public insurance were linked to higher tobacco-only use and co-use, and Hispanic individuals had lower likelihoods of all use patterns versus non-Hispanic White individuals. Post-pandemic, educational gradients emerged for cannabis-only use, Medicaid and Medicare coverage remained associated with higher tobacco-only use, and uninsured status became associated with higher cannabis-only use.

Why it matters

Prenatal substance use profiles in the U.S. have shifted, with cannabis overtaking tobacco as the more prevalent standalone substance. Clinical screening and public health counseling during pregnancy need to adapt from a primary focus on tobacco to addressing rising cannabis exposure.

Limits

Substance use was measured exclusively via self-report without biochemical validation, which may lead to underreporting. The repeated cross-sectional design cannot establish causality or track individual trajectories across pregnancy. Sample size constraints precluded regression analyses for the year 2020, and specific effect sizes (adjusted odds ratios and marginal effects) were not reported in the abstract.

Cited by