Marijuana smoking and outcomes of infertility treatment with assisted reproductive technologies

Abstract

Study question: What is the association of female and male partner marijuana smoking with infertility treatment outcomes with ART?

Summary answer: Women who were marijuana smokers at enrollment had a significantly higher adjusted probability of pregnancy loss during infertility treatment with ART whereas, unexpectedly, there was a suggestion of more favorable treatment outcomes in couples where the man was a marijuana smoker at enrollment.

What is known already: Data on the relation of female and male partner marijuana use with outcomes of infertility treatment is scarce despite increased use and legalization worldwide.

Study design, size, duration: We followed 421 women who underwent 730 ART cycles while participating in a prospective cohort (the Environment and Reproductive Health Study) at a fertility center between 2004 and 2017. Among them, 200 women (368 cycles) were part of a couple in which their male partner also enrolled in the study.

Participants/materials, setting, methods: Participants self-reported marijuana smoking at baseline. Clinical endpoints were abstracted from electronic medical records. We used generalized linear mixed models with empirical standard errors to evaluate the association of baseline marijuana smoking with ART outcomes adjusting for participants' age, race, BMI, tobacco smoking, coffee and alcohol consumption, and cocaine use. We estimated the adjusted probability of implantation, clinical pregnancy, and live birth per ART cycle, as well as the probability of pregnancy loss among those with a positive B-hCG.

Main results and the role of chance: The 44% of the women and 61% of the men had ever smoked marijuana; 3% and 12% were marijuana smokers at enrollment, respectively. Among 317 women (395 cycles) with a positive B-hCG, those who were marijuana smokers at enrollment (N = 9, cycles = 16) had more than double the adjusted probability of pregnancy loss than those who were past marijuana smokers or had never smoked marijuana (N = 308, 379 cycles) (54% vs 26%; P = 0.0003). This estimate was based on sparse data. However, couples in which the male partner was a marijuana smoker at enrollment (N = 23, 41 cycles) had a significantly higher adjusted probability of live birth than couples in which the male partner was a past marijuana smoker or had never smoked marijuana (N= 177, 327 cycles) (48% vs 29%; P = 0.04), independently of the women's marijuana smoking status. Treatment outcomes of past marijuana smokers, male and female, did not differ significantly from those who had never smoked marijuana.

Limitations, reasons for caution: Marijuana smoking was self-reported with possible exposure misclassification. Chance findings cannot be excluded due to the small number of exposed cases. The results may not be generalizable to couples from the general population.

Wider implications of the findings: Even though marijuana smoking has not been found in past studies to impact the ability to become pregnant among pregnancy planners in the general population, it may increase the risk of pregnancy loss among couples undergoing infertility treatment. Marijuana smoking by females and males may have opposing effects on outcomes of infertility treatment with ART.

Study funding/competing interest(s): The project was financed by grants R01ES009718, P30ES000002, and K99ES026648 from the National Institute of Environmental Health Sciences (NIEHS). None of the authors has any conflicts of interest to declare.

Keywords: assisted reproductive technologies; couple; infertility; live birth; marijuana; pregnancy loss.

Figure 1

Adjusted probability (95% CI) of pregnancy loss associated with women’s marijuana smoking among 317 women who had 395 pregnancies in the EARTH study. Abbreviations: EARTH, the Environment and Reproductive Health Study; N, number of women; n, number of ART cycles. ¹Defined as a positive B-hCG that did not result in live birth. ²Data is presented as predicted adjusted probabilities with 95% confidence intervals adjusted for women’s age, BMI, race, tobacco smoking status, coffee intake, alcohol intake, and cocaine use. Numbers shown below columns represent numbers of pregnancy losses/total number of pregnancies and total number of women (N) across marijuana smoking categories. Analysis was done using generalized linear mixed models with random intercepts, binary distribution and logit link function, and empirical standard error. The marginal covariate-adjusted probabilities were used to present the results adjusted for the covariates at their average levels for continuous variables and weighted average level of categorical variable in the model. ^**P-value < 0.005 compared to never marijuana smokers.

Figure 2

Adjusted probability of clinical ART outcomes associated with joint male and female partners’ marijuana smoking at enrollment among 200 couples (368 cycles) in the EARTH study. Abbreviations: EARTH, the Environment and Reproductive Health Study. Couples with not at enrollment marijuana smoker woman and not at enrollment marijuana smoker man were 175 couples and 324 cycles. Couples with not at enrollment marijuana smoker woman and marijuana smoker at enrollment man were 21 couples and 35 cycles. Couples with marijuana smoker woman at enrollment and not at enrollment marijuana smoker man were 2 couples and 3 cycles. Couples with marijuana smoker woman at enrollment and marijuana smoker man at enrollment were 2 couples and 6 cycles. Data is presented as covariate-adjusted marginal means adjusted for both men’s and women’s age, BMI, race, tobacco smoking status, coffee intake, alcohol intake, and cocaine use. Analysis was done using generalized linear mixed models with random intercepts, binary distribution and logit link function, and empirical standard error. The marginal covariate-adjusted probabilities were used to present the results adjusted for the covariates at their average levels for continuous variables and weighted average level of categorical variable in the model. *Indicates <0.05 compared to the couples with not at enrollment marijuana smoker woman and marijuana smoker man at enrollment.