Pregnancy and postpartum cannabis and alcohol use among people with HIV in the United States.

Steinberg J.; Kacanek D.; Broadwell C.; Jao J.; Powis K.; Yao T.; Barr E.; Haddad L.; Siminski S.; Chadwick E. et al. · Pregnancy (Hoboken, N.J.) · 2026

**Background & Methods** This observational study analyzed substance use data from the Surveillance Monitoring for ART Toxicities study, which enrolled pregnant people with HIV across 22 U.S. sites between 2007-2019. The analysis included 1,765 pregnancies with alcohol data and 1,772 with cannabis data, examining patterns of use during pregnancy and postpartum periods, with multivariable regression models identifying associated risk factors. **Key Findings** - Substance use increased substantially postpartum: alcohol use rose from 8% during pregnancy to 44% postpartum, while cannabis use increased from 8% to 14%. Discontinuation rates were low, with most pregnancies showing either never-use or initiation/resumption patterns postpartum. - Alcohol initiation/resumption was associated with younger age (aRR 0.98 per year), Midwest residence (aRR 1.44 vs. Northeast), cesarean delivery (aRR 1.14), and tobacco use during pregnancy (aRR 1.19). Lower income and less education were protective factors. - Cannabis initiation/resumption was associated with younger age (aRR 0.94 per year), lower household income (aRR 1.90), more recent delivery year (aRR 1.10 per year), psychiatric diagnosis during pregnancy (aRR 1.71), and tobacco use during pregnancy (aRR 1.94). **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** This study provides important epidemiological data but has notable limitations. The observational design prevents causal inference. Self-reported substance use is subject to recall bias and potential underreporting due to social desirability bias or concerns about legal/custody implications. The cohort may not be representative of all pregnant people with HIV nationally. The study does not assess clinical outcomes related to substance use (maternal health, infant outcomes, medication adherence). Confounding by unmeasured variables is possible. However, the large, diverse cohort from multiple sites and long enrollment period (2007-2019) strengthen generalizability. The differentiation between alcohol and cannabis risk factors provides clinically relevant insights for targeted interventions in this vulnerable population.

More recent cannabis research