Cannabis Use and Breastfeeding Initiation: Associations with Patterns, Perceived Risk, and Clinical Advice in a Multistate Pregnancy Risk Assessment Monitoring System Study (2017-2021).

Skelton K.; Iobst S.; Benjamin-Neelon S. · Breastfeeding medicine : the official journal of the Academy of Breastfeeding Medicine · 2026

**Background & Methods** This repeated cross-sectional study analyzed 2017-2021 Pregnancy Risk Assessment Monitoring System data from 10 states to examine associations between maternal cannabis use patterns, perceived safety of cannabis during breastfeeding, prenatal counseling, and breastfeeding initiation rates. The analytic sample included 52,577 live births among women who reported any perinatal cannabis use, with weighted prevalence estimates calculated using survey-weighted chi-square tests. **Key Findings** • Breastfeeding initiation rates varied significantly by perinatal cannabis use timing: postpartum-only use showed the highest initiation rate (89.52%), while concurrent prenatal and postpartum use showed the lowest (71.88%), representing an 18-percentage-point difference. • Among women using cannabis postpartum only, receipt of prenatal advice against cannabis use while breastfeeding was associated with significantly lower breastfeeding initiation (81.95% vs. 94.07%, p = 0.004), suggesting that clinical counseling may paradoxically discourage breastfeeding initiation in this population. • The findings highlight a critical gap in clinical communication: counseling that emphasizes cannabis risks during breastfeeding may inadvertently reduce breastfeeding uptake rather than promoting evidence-based infant feeding practices. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** This study provides moderate-level evidence with several limitations. As a cross-sectional observational study, it cannot establish causality or determine whether the association between receiving advice and lower breastfeeding initiation reflects reverse causation (women already intending not to breastfeed may have received more counseling). The restriction to women reporting perinatal cannabis use introduces potential selection bias. Self-reported data on cannabis use and counseling receipt may be subject to recall bias and social desirability bias. The study does not examine actual cannabis concentrations in breast milk or infant outcomes. Findings are limited to 10 states participating in the Marijuana Supplement, potentially limiting generalizability. The lack of information on counseling content quality, clinician training, or communication approaches prevents identification of effective intervention strategies. Future research should employ prospective designs with validated measurement tools and examine mechanisms underlying the observed counseling-initiation association.

More recent cannabis research