Cannabis Exposure and Adverse Pregnancy Outcomes Related to Placental Function

Abstract

Importance: Cannabis use is increasing among reproductive-age individuals and the risks associated with cannabis exposure during pregnancy remain uncertain.

Objective: To evaluate the association between maternal cannabis use and adverse pregnancy outcomes known to be related to placental function.

Design, setting, and participants: Ancillary analysis of nulliparous individuals treated at 8 US medical centers with stored urine samples and abstracted pregnancy outcome data available. Participants in the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be cohort were recruited from 2010 through 2013; the drug assays and analyses for this ancillary project were completed from June 2020 through April 2023.

Exposure: Cannabis exposure was ascertained by urine immunoassay for 11-nor-9-carboxy-Δ9-tetrahydrocannabinol using frozen stored urine samples from study visits during the pregnancy gestational age windows of 6 weeks and 0 days to 13 weeks and 6 days (visit 1); 16 weeks and 0 days to 21 weeks and 6 days (visit 2); and 22 weeks and 0 days to 29 weeks and 6 days (visit 3). Positive results were confirmed with liquid chromatography tandem mass spectrometry. The timing of cannabis exposure was defined as only during the first trimester or ongoing exposure beyond the first trimester.

Main outcome and measure: The dichotomous primary composite outcome included small-for-gestational-age birth, medically indicated preterm birth, stillbirth, or hypertensive disorders of pregnancy ascertained by medical record abstraction by trained perinatal research staff with adjudication of outcomes by site investigators.

Results: Of 10 038 participants, 9257 were eligible for this analysis. Of the 610 participants (6.6%) with cannabis use, 32.4% (n = 197) had cannabis exposure only during the first trimester and 67.6% (n = 413) had ongoing exposure beyond the first trimester. Cannabis exposure was associated with the primary composite outcome (25.9% in the cannabis exposure group vs 17.4% in the no exposure group; adjusted relative risk, 1.27 [95% CI, 1.07-1.49]) in the propensity score-weighted analyses after adjustment for sociodemographic characteristics, body mass index, medical comorbidities, and active nicotine use ascertained via urine cotinine assays. In a 3-category cannabis exposure model (no exposure, exposure only during the first trimester, or ongoing exposure), cannabis use during the first trimester only was not associated with the primary composite outcome; however, ongoing cannabis use was associated with the primary composite outcome (adjusted relative risk, 1.32 [95% CI, 1.09-1.60]).

Conclusions and relevance: In this multicenter cohort, maternal cannabis use ascertained by biological sampling was associated with adverse pregnancy outcomes related to placental dysfunction.

Figure 1.. Determination of the Study Population

Figure 2.. Quantified Cannabis Exposure and the Primary Composite Adverse Pregnancy Outcome

Each panel simultaneously displays the observed distribution of the 11-nor-9-carboxy-Δ⁹-tetrahydrocannabinol (THC-COOH) exposure levels in urine measured by routine immunoassay and confirmed with quantitative liquid chromatography tandem mass spectrometry and the modeled relationship between the primary composite outcome and the number of pregnancies at each cannabis exposure level for the participants included in this ancillary analysis (n = 9257). The light blue vertical bars indicate the number of pregnancies in each exposure-level group. Each bar represents a doubling of exposure when moving left to right across the x-axis. All pregnancies within each range of exposure are depicted in a single bar (eg, the bar for an exposure level of 16 ng/mL includes all pregnancies with an exposure level from 16-<32 ng/mL). The black line is the estimated proportion of participants with the primary composite adverse pregnancy outcome along with the 95% CIs depicted by the gray shading. The relationship between the prevalence of the primary composite outcome and cannabis exposure was estimated from adjusted logistic regression models with natural cubic splines log transformed with 4 knots. The participants without detectable THC-COOH exposure levels were excluded from the Figure. ^aEstimates are from Poisson regression with multiple imputation. Includes any nicotine use (determined by a urine cotinine level ≥300 ng/mL) at any study visit, age (≥30 years), body mass index (calculated as weight in kilograms divided by height in meters squared; <20, 20-29.9, or ≥30), marital status (married or not married), public insurance, maternal comorbidities (had preexisting diabetes, chronic hypertension, or both). ^bIncludes all covariates in footnote “a” plus level of depression measured by the Edinburgh Postnatal Depression Scale (score ≥11 or <11), level of stress measured by the Perceived Stress Scale (low, moderate, or high), level of anxiety measured by the State-Trait Anxiety Inventory (none or low, moderate, or high), and other illicit drug use (assessed by urine assay at study visit 1).