The results of this study indicate a very high proportion (> 40%) of pregnant women, who are in treatment for OUD with or without concurrent alcohol consumption, used marijuana early in pregnancy. These estimates are substantially higher than recent (2017) estimates of past month marijuana use in pregnant and non-pregnant women in the general population surveyed in the NSDUH (7.0% and 11.2%, respectively [18], and the 3.4% rate of any use during pregnancy in Northern California [19]). A study in Washington state examined the effect of marijuana legalization on alcohol and drug use in pregnant/parenting women. Subjects were women attending a three-year program serving those who use alcohol and/or drugs during pregnancy, and the study found women exiting the program in the post-legalization period were twice as likely to report using marijuana in the 30 days before exiting than women in the pre-legalization cohort [24]. Moreover, post-legalization, women were twice as likely to continue use from program enrollment to exit.

Increased marijuana use among pregnant women may be associated with low perceived risk [25, 26]. Women often have contradictory views: on the one hand, trying to decrease use when they become pregnant, but at the same time believing that marijuana is safe, ‘natural,’ and a way to mitigate nausea and vomiting during pregnancy [26–28]. Women report their information sources about marijuana use include non-medical and medical sources [29, 30]. Some women with OUD also believe marijuana may help manage opioid dependence and withdrawal [31]. The prevalence of marijuana use among pregnant women with OUD and some alcohol use in this study was twice that reported by people with a chronic medical condition surveyed in the 2016 and 2017 Behavioral Risk Factor Surveillance System (BRFSS): 21.9% of those aged 18–34 with any chronic condition reported marijuana use versus 13.1% of people surveyed with no medical condition [32]. In general, among people with and without medical conditions, marijuana use has become more acceptable for both recreational and medical reasons; 7.3% of adult respondents in a national online survey agreed that marijuana use is somewhat or completely safe during pregnancy [33].

In light of the paucity of studies demonstrating safety in pregnancy, particularly concerning is the substantial proportion of women (~ 14% in both groups) who used marijuana almost daily, as contrasted to ~ 31% of patients using once a week or less during pregnancy. It should be noted that women in the Alcohol group reported light-to-moderate/intermittent alcohol use during pregnancy (1.5 drinks/week, on average, 30 days before enrollment).

Interestingly, prevalence estimates were higher per maternal self-report than urinalysis. We attribute this difference to a shorter detection window on urinalysis vs. the reporting timeframe (use between the LMP and the baseline visit – on average 22.98 (SD = 7.09) weeks). Urinary elimination of THC might vary depending on the quantity and frequency of marijuana use, creatinine levels, body mass index, and individual differences in metabolism. Typically, a rapid decrease is seen during the first three days of abstinence; however, subjects with high exposure levels might have detectable THC concentrations for several weeks [34]. Self-reported measures have the advantage of being noninvasive and enabling evaluation over longer time periods in contrast to biomarker assessments. The validity of self-reported drug use can vary by setting, population, and interviewer traits [35]. Validity of self-reported marijuana use has been shown to be high (agreement 81.1% and sensitivity 91.2% [36]. And, in drug treatment settings, self-reported opioid use has been shown to be high (sensitivity 99%) [37]. We have previously shown that pregnant women in a substance use treatment program generally under-report substance use, but opioids and marijuana are more accurately reported [38]. Specifically, sensitivity of self-report was 100% for methadone, 83.3% for buprenorphine, 58.3% for other opioids, and 57.9% for marijuana as opposed to amphetamines (37.5%), benzodiazepines (40%), and cocaine (47.4%) [38].

Younger maternal age, using other substances, and partnership status were independently associated with marijuana use in our study, consistent with other studies [18, 19]. Un-partnered women with OUD, with and without alcohol use during pregnancy, demonstrated a lower prevalence of marijuana use compared to women who just used alcohol alone. We observed, among women who were partnered (married or cohabitating), those with a history of both OUD and alcohol use were more likely to use marijuana than women who used alcohol alone. Few studies have examined marijuana use among pregnant women who have used other substances; however, our findings differ from other studies of pregnant women, wherein married women were less likely to use marijuana [25, 39] or opioids [40] compared to unpartnered (unmarried, divorced, separated, or widowed) women. Differences in marijuana use by marital status may be related to reasons for use, such as recreational, medical, or ease of access. In a study examining alcohol use in cannabis users (with and without a medical recommendation) Subbaraman et al. [41] found no differences in cannabis use by marital status. Partnership factors are known to influence substance use patterns. Some studies have found newly married men and women tend to reduce substance use over time [42, 43]. However, women in relationships with drug-using men may be more likely to use and misuse substances [44, 45]. Our findings point to the need to examine social factors specific to marijuana use in pregnant women with a history of substance and/or alcohol use.

Women in our study demonstrated a high prevalence of polysubstance use (36.3%), which is consistent with other reports [25, 40], and marijuana use was over twice as likely for women who reported polysubstance use in comparison to women who did not report polysubstance use. Concurrent tobacco use was particularly high (70.5%) in all women, and both tobacco and estimated alcohol consumption were higher among those who concurrently used marijuana. A study of over 160,000 pregnant women admitted for substance use treatment demonstrated that alcohol and marijuana were the most common co-exposures [46]. Polysubstance use is of significant concern among pregnant women. We have previously shown that the risk of growth deficiencies (i.e., head circumference < 10^th percentile) was substantially increased in polydrug users compared to abstinent controls [47]. Additionally, polysubstance exposure was found to be the most significant predictor of NOWS severity in buprenorphine-exposed neonates [48].

The findings of this analysis should be interpreted in light of several limitations. First, while the sample size of this study was large (251 participants), results cannot be generalized to all pregnant women who use substances in pregnancy since the patients were recruited from a comprehensive multi-disciplinary prenatal clinic at UNM focused on SUD. Further, results cannot be generalized to patients with OUD who do not receive MOUD who may have even higher rates of marijuana use. Second, we could not examine secular trends in the prevalence of marijuana use over time. In New Mexico, medical marijuana use was legalized in 2007, while recreational use remained illegal but was decriminalized in 2019. (In early 2021, recreational use was legalized.) Thus, since most data were obtained before 2019, estimates are likely to be conservative. Third, we acknowledge a relatively small sample size in the OUD + Alcohol and Alcohol groups and limited generalizability to patients with different alcohol consumption patterns, including women with alcohol use disorder. Strengths of the study include a prospective cohort design and state-of-the-art characterization of prenatal substance use with repeated prospective interviews and comprehensive batteries of biomarkers for alcohol, tobacco, and illicit drug use.

Future studies should specifically examine the prevalence of marijuana use in pregnant women with chronic use of opioid analgesics and those with OUD who are not receiving methadone or buprenorphine. Opioid use disorder has been listed (since 2018 and not overlapping with enrollment in this study) as one of the 28 indications for medical marijuana in New Mexico, and pregnancy is not an exclusion criteria, which may result in pregnant woman with OUD continuing or initiating medical marijuana [49]. Given the unique impact of the opioid epidemic on rural/under-served areas, it would be important to characterize the pattern of substance use in rural vs. urban areas and examine trends and motivations over time and as a function of marijuana legalization laws. A new preclinical research finding indicates that simultaneous exposure to alcohol and cannabinoid compounds increases the risk of birth defects in a murine model, indicative of a compounding interaction [50]. This research also demonstrated common signaling pathways for such interactive effects on fetal mice development. Future studies of adverse perinatal and pediatric neurodevelopmental outcomes associated with intrauterine exposure to opioids and alcohol need to characterize marijuana co-exposure carefully, given the potential additive effects.

The American College of Obstetricians and Gynecologists (ACOG) recommends that pregnant women “reporting marijuana use should be counseled about concerns regarding potential adverse health consequences of continued use during pregnancy… [and] should be encouraged to discontinue marijuana use” [51]. While little is known about the direct effects of marijuana use on the developing brain (based on restricted access to marijuana for research studies), cannabinoid receptors are found throughout the brain, and the scant work that has been performed demonstrates broad alterations to brain function with cannabinoid exposure [52–55]. As legalization of marijuana becomes more widespread and the percentage of THC in marijuana products continues to increase, additional information about the effects of prenatal marijuana exposure will need to be obtained. Moreover, ACOG also recommends that alternative therapeutic approaches with better pregnancy-specific safety data be considered instead of medical marijuana use. The American Academy of Pediatrics (AAP) also advises against the use of marijuana or other cannabinoid-containing products either for recreational or medical reasons, including nausea and vomiting during pregnancy. AAP recommends universal screening, brief intervention, and referral for treatment [56]. Currently, the ACOG, AAP and the Surgeon General recommend women to abstain from marijuana use completely if they are trying to conceive, pregnant or lactating. There remain significant gaps in knowledge and research however. For instance, there is not substantial data to conclude that abstinence from marijuana use will decrease negative perinatal/pediatric outcomes as it does for cigarette [57] and alcohol exposure [58]. Nevertheless, a harm reduction approach would include consideration of decreasing use in the high proportion of women with near-daily use.