Abstract
Background
Cannabis use during pregnancy is increasingly common and is associated with adverse health outcomes for pregnant individuals and their offspring. Identifying preconception factors that are associated with prenatal cannabis use is critical to inform early prevention and intervention. This study tested whether frequency of preconception cannabis use was associated with cannabis use during early pregnancy using data from a large healthcare system with universal screening for cannabis use.
Methods
This observational study included pregnant individuals in Kaiser Permanente Northern California (KPNC) who self-reported any cannabis use (daily, weekly, or monthly or less) during the year before pregnancy from 2011 to 2022 (excluding 2020). Prenatal cannabis use was based on self-report and/or a positive toxicology test during the first trimester of pregnancy at entrance to prenatal care (at ~ 8 weeks gestation). Modified Poisson models were fit to assess associations between frequency of preconception cannabis use and prenatal cannabis use, adjusting for covariates.
Results
The sample of 40,806 pregnancies from 36,622 unique individuals who self-reported any preconception cannabis use was 65.7% non-White; 27.6% were aged < 25 years. Nearly half (45.1%) screened positive for prenatal use, including 23.7% by self-report and 36.6% by positive toxicology test. Compared to monthly or less preconception cannabis use, daily use (adjusted prevalence ratio [aPR] = 2.66, 95% CI 2.59–2.73) or weekly use (aPR = 1.99, 95% CI 1.93–2.05) was associated with greater risk of prenatal cannabis use. Results were similar when prenatal cannabis use was based on self-report only or on toxicology testing only.
Conclusions
Greater frequency of preconception cannabis use was associated with substantially increased risk of prenatal cannabis use. Findings reinforce the need for early harm prevention efforts focused on reducing the frequency of cannabis use among women of reproductive age, including screening, education, and early linkage to intervention.