With the legalization of recreational marijuana in many states, we anticipate more women will be using and self-reporting marijuana use in pregnancy. Marijuana is the most common illicit drug used in pregnancy, with a prevalence of use ranging from 3% to 30% in various populations. Marijuana freely crosses the placenta and is found in breast milk. It may have adverse effects on both perinatal outcomes and fetal neurodevelopment. Specifically, marijuana may be associated with fetal growth restriction, stillbirth, and preterm birth. However, data are far from uniform regarding adverse perinatal outcomes. Existing studies are plagued by confounding by tobacco and other drug exposures as well as sociodemographic factors. In addition, there is a lack of quantification of marijuana exposure by the trimester of use and a lack of corroboration of maternal self-report with biological sampling, which contributes to the heterogeneity of study results. There is an emerging body of evidence indicating that marijuana may cause problems with neurological development, resulting in hyperactivity, poor cognitive function, and changes in dopaminergic receptors. In addition, contemporary marijuana products have higher quantities of delta-9-tetrahydrocannabinol than in the 1980s when much of the marijuana research was completed. The effects on the pregnancy and fetus may therefore be different than those previously seen. Further research is needed to provide evidence-based counseling of women regarding the anticipated outcomes of marijuana use in pregnancy. In the meantime, women should be advised not to use marijuana in pregnancy or while lactating.
Marijuana is the most frequently used illicit drug in Western countries. In 2013, 19.8 million, or 7.5% of the US population, reported its use within the last month, an increase from 2007 when only 5.8% of the population had used marijuana within the past month.
Reported prevalence rates of marijuana use in pregnancy vary from as low as 3% to as high as 34%. We anticipate an increase in marijuana use in pregnancy as legalization of marijuana increases throughout the United States. This review is intended to provide practicing clinicians with an understanding of existing literature and recommendations for managing women who use marijuana during pregnancy because this will be an increasingly encountered clinical scenario.
The term marijuana is used throughout this article to represent cannabis use globally. Technically the active psychogenic component of marijuana is a cannabinoid called delta-9-tetrahydrocannibinol (THC).
Search methodology
Ovid Medline (PubMed) and Embase were searched on Dec. 11, 2014, for relevant articles. A focused search was conducted with the search terms marijuana and marihuana or cannabinoids and pregnancy, lactation, and outcomes including adverse perinatal outcomes and neurodevelopment. A search without any language or year limits yielded 615 unique citations. Abstracts were reviewed by the authors, and all pertinent articles were obtained and reviewed individually. In addition, references of pertinent articles were reviewed to find any additional articles that were not identified with the initial search (n = 43). All pertinent articles are summarized here. Our goal was not to provide a systematic review of a specific research question but rather to provide practicing clinicians with a comprehensive overview of the existing marijuana in pregnancy and lactation literature.
Legalization of marijuana
Currently both recreational and medical marijuana remain illegal by federal law in the United States. However, the legalization of medical and recreational marijuana at the state level is increasing throughout the United States. At this point, 23 states have legalized the use of medical marijuana, and 4 states have legalized both medical and recreational marijuana ( Figure ).
The Colorado experience
Medical marijuana was legalized in Colorado in the year 2000. However, it was not until 2009 when the US Attorney General issued a statement passing the jurisdiction of marijuana law enforcement to state governments that we saw a sharp increase in the number of medical marijuana users in the state. In 2012, recreational marijuana was legalized in the state of Colorado with the passing of Amendment 64. There is no stipulation in the law stating that pregnant women cannot purchase or possess marijuana.
Sales of recreational marijuana have been steadily increasing since the opening of the first recreational dispensaries on Jan. 1, of 2014. The state of Colorado does not publish overall sales amounts but does publish tax revenue on a monthly basis. In January 2014, the revenue was 3.5 million dollars. The monthly tax revenue is now up to 7.6 million dollars for the month of October 2014, showing a steady increase in sales and consumption. In addition, there has been an increase in the use of alternative forms of consumption such as vaping (heating the cannabis to release THC and cannabinoids without making it smoke), lotions, and edibles.
Following the legalization of marijuana, we have noted several unanticipated adverse consequences of the increase in marijuana availability including an increase in pediatric overdoses and emergency visits for marijuana toxicity.
Attitudes and beliefs
When women have been followed up longitudinally during pregnancy, a decrease in marijuana use has been noted across trimesters of pregnancy. In a 1 year prospective cohort study, marijuana use in pregnancy declined from 32% in the first trimester to 16% in the third trimester.
Similarly, a longitudinal prospective study on drug use in pregnancy (n = 86), the Development and Infancy Study, found that the percentage of women who used marijuana throughout the pregnancy declined. However, approximately 60% of women who used marijuana in the year prior to pregnancy continued to use more than 10 joints per week, indicating that many women continue use throughout pregnancy. It should be noted that the women in the Development and Infancy Study smoked an average of 21 joints per week in the month prior to pregnancy and may not be representative of less frequent users of marijuana.
Two thirds of adults surveyed in a UK study noted that cannabis was either “not very harmful” or “not at all harmful.” This is in contrast to other recreational drugs such as heroin or cocaine in which less than 5% of adults surveyed perceived them to be either “not very harmful” or “not at all harmful.” The perceived safety likely contributes to the high prevalence of its use in pregnancy.
Legalization of marijuana
Currently both recreational and medical marijuana remain illegal by federal law in the United States. However, the legalization of medical and recreational marijuana at the state level is increasing throughout the United States. At this point, 23 states have legalized the use of medical marijuana, and 4 states have legalized both medical and recreational marijuana ( Figure ).
The Colorado experience
Medical marijuana was legalized in Colorado in the year 2000. However, it was not until 2009 when the US Attorney General issued a statement passing the jurisdiction of marijuana law enforcement to state governments that we saw a sharp increase in the number of medical marijuana users in the state. In 2012, recreational marijuana was legalized in the state of Colorado with the passing of Amendment 64. There is no stipulation in the law stating that pregnant women cannot purchase or possess marijuana.
Sales of recreational marijuana have been steadily increasing since the opening of the first recreational dispensaries on Jan. 1, of 2014. The state of Colorado does not publish overall sales amounts but does publish tax revenue on a monthly basis. In January 2014, the revenue was 3.5 million dollars. The monthly tax revenue is now up to 7.6 million dollars for the month of October 2014, showing a steady increase in sales and consumption. In addition, there has been an increase in the use of alternative forms of consumption such as vaping (heating the cannabis to release THC and cannabinoids without making it smoke), lotions, and edibles.
Following the legalization of marijuana, we have noted several unanticipated adverse consequences of the increase in marijuana availability including an increase in pediatric overdoses and emergency visits for marijuana toxicity.
Attitudes and beliefs
When women have been followed up longitudinally during pregnancy, a decrease in marijuana use has been noted across trimesters of pregnancy. In a 1 year prospective cohort study, marijuana use in pregnancy declined from 32% in the first trimester to 16% in the third trimester.
Similarly, a longitudinal prospective study on drug use in pregnancy (n = 86), the Development and Infancy Study, found that the percentage of women who used marijuana throughout the pregnancy declined. However, approximately 60% of women who used marijuana in the year prior to pregnancy continued to use more than 10 joints per week, indicating that many women continue use throughout pregnancy. It should be noted that the women in the Development and Infancy Study smoked an average of 21 joints per week in the month prior to pregnancy and may not be representative of less frequent users of marijuana.
Two thirds of adults surveyed in a UK study noted that cannabis was either “not very harmful” or “not at all harmful.” This is in contrast to other recreational drugs such as heroin or cocaine in which less than 5% of adults surveyed perceived them to be either “not very harmful” or “not at all harmful.” The perceived safety likely contributes to the high prevalence of its use in pregnancy.
Screening and testing for marijuana use
The American College of Obstetricians and Gynecologists and the American Academy of Pediatrics support screening all women for drug use at the time of entry to prenatal care. Verbal screening for self-reported use was noted to be acceptable to patients in one study. Women who report use should then be encouraged to stop and referred to local substance use disorder programs if needed.
Unfortunately, maternal and fetal testing for marijuana exposure is fraught with error. Maternal urine testing is the most accurate method of testing. However, the duration of a positive urine toxicology result from the last use depends on many factors including chronicity of use ( Table 1 ). Despite its limitations, urine is easy to obtain, has a high concentration of metabolites, and is therefore the preferred method of screening.
| Biological sample | Duration of positive result | Test limitations |
|---|---|---|
| Maternal urine | 2–3 days in occasional users Several weeks in chronic users | Chronicity of use determines duration of positive result |
| Maternal serum | 2–3 days in occasional users Several weeks in chronic users | Chronicity of use determines duration of positive result Invasive sample Shorter half-life than urine |
| Maternal hair | Several weeks | Less accurate for marijuana than other drugs False positives from passive exposure Not clinically used due to cost and inaccuracy |
| Meconium | Positive result indicates second- and third-trimester exposure | Small amount of detectable THC in the samples High false-positive rate (up to 43%) Send out to reference laboratory Costly and impractical at many sites |
| Neonatal hair | Positive result indicates third-trimester exposure | Costly and impractical at many sites Less sensitive than meconium |
Testing of maternal hair samples is inaccurate and may remain positive despite no recent use. Neonatal hair and meconium can also be tested ( Table 1 ); however, because of the cost of testing, delay in results, and a high false-positive rate in laboratory testing of meconium by different techniques, neonatal testing is rarely used in clinical practice.
There are no good methods to quantify the amount of marijuana ingested using biological sampling in a clinical setting. The amount of THC in various forms of marijuana varies by the extraction process from the plant, Cannabis sativa, which also results in challenges in quantifying self-reported use. In addition, the various forms of consumption result in different rates of absorption and peak blood concentrations. In a report from the University of Mississippi’s Potency Monitoring Project, the average concentration of THC in seized samples in the United States in 2008 was 13.0%, which was an increase from 3.2% in 1983.
Nausea and vomiting in pregnancy
As with any drug or medication in pregnancy, possible benefits must be weighed against possible adverse effects. There are few data on the possible benefits of marijuana use in pregnancy. Interest in the use of marijuana as an antiemetic has been propagated by its efficacy in oncology patients.
There are 2 studies investigating the relationship between marijuana use and nausea and vomiting of pregnancy. Roberson et al used the pregnancy risk assessment monitoring system data (n = 4375) and found that women who used marijuana in pregnancy were more likely to report severe nausea (3.7% vs 2.3%; prevalence ratio, 1.63; 95% confidence interval [CI], 1.08–2.44). The treatment of nausea with marijuana was not specifically addressed in the study.
Westfall et al reported on the prevalence of nausea among 79 women who used medicinal marijuana in pregnancy. Forty of these women (51%) used marijuana to treat nausea and vomiting of pregnancy, and 92% of them believed it was effective. There was no control group, no documentation of quantity used, or a demonstration of effect on symptoms of nausea other than subjective report by survey after the pregnancy.
In summary, the effect of marijuana use on nausea and vomiting of pregnancy is unknown.
Anesthetic considerations
Marijuana use can affect the safety and administration of anesthesia surrounding delivery. In high doses, marijuana can cause bradycardia and hypotension, but more commonly, low or moderate doses can cause tachycardia. If tachycardia is present or marijuana use is suspected, drugs that increase heart rate such as ketamine, pancuronium, and epinephrine should be avoided. Because marijuana is often inhaled, it can also cause upper-airway irritation and edema, making anesthetic administration more complicated.
Adverse perinatal outcomes
Fetal growth
Many of the human studies of marijuana in pregnancy focus on fetal growth ( Table 2 ). Abnormalities in growth are biologically plausible, given the passage of cannabinoids across the placenta. There are some data suggesting that cannabis affects glucose and insulin regulation and therefore may affect the fetal growth trajectory. However, data regarding fetal growth with marijuana exposure are mixed, with some studies demonstrating a decrease in birthweight and/or growth and others demonstrating no association ( Table 2 ). In part, the controversy may be a result of differing methodology for the ascertainment of marijuana exposure, varying from a single question about self-reported use at study entry to detailed longitudinal frequency of use data and biological sampling ( Table 2 ). In addition, many early studies did not account for concurrent exposure to tobacco.
| Study and number in cohort | Marijuana-exposed women, n (%) | Setting | Data source | Marijuana measure | Other variables considered in analysis | Findings (adjusted ORs or regression coefficients with 95% CIs reported when available) | Limitations and comments |
|---|---|---|---|---|---|---|---|
| Prospective cohort studies a | |||||||
| Day et al, 1991 (n = 519) | 324 (62) | Single institution | Self-report by prenatal interview in each trimester of pregnancy | Frequency: light (0–2.9 joints/wk), moderate (3–6.9/wk), and heavy (≥1 joints/d) | SES, obstetric hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with SGA Isolated higher birthweight in heavy third-trimester users compared with nonusers (3357 g vs 3215 g; P = .04) | Increase in birthweight in marijuana users compared with nonusers Women who use marijuana were intentionally oversampled |
| El Marroun et al, 2009 (n = 7452) | 459 (6) | Population-based study in The Netherlands | Self-report at study enrollment | Frequency: daily, weekly, monthly Reported use: only before pregnancy, use in early pregnancy, or ongoing use | Standard demo, psych hx, EtOH, fetal sex, tobacco Excluded women with other drugs | Use before pregnancy did not affect growth Early pregnancy use decreased growth 11.18 g (–15.26 to –7.10)/wk Ongoing marijuana use decreased growth 14.44 g (–22.94 to –5.94)/wk | Only study with serial ultrasounds to assess fetal growth (detailed in fetal growth section of text) Marijuana use not well quantified |
| Fergusson et al, 2002 (n = 12,129) | 606 (5) | Population-based study in United Kingdom | Self-completed questionnaire at 18–20 wks gestation | Frequency: 1 time/day, 2–4 time/wk, 1 time/wk, <1 time/wk before pregnancy, first trimester and ongoing | Standard demo, other drugs, EtOH, tobacco | Ongoing use 1 or more/wk throughout pregnancy was not associated with lower birthweight –84.20 g (–174.70 to 6.40) | Self-report data collected at 18–20 wks’ gestation, no later pregnancy data |
| Fried et al, 1984 (n = 583) | 84 (14) | Referred to study by primary obstetrician/study ads | Self-report by prenatal interview in each trimester of pregnancy | Frequency: irregular users (≤1 joint/wk), moderate (2-5/wk), heavy (>5/wk) | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with LBW | Marijuana use not well quantified, averaged over the course of pregnancy |
| Gray et al, 2010 (n = 86) | 38 (44) | Single institution | Self-report by prenatal interview in each trimester of pregnancy Biological samples | Frequency: number of joints/day by trimester Presence of THC in maternal saliva and meconium | Standard demo, OB hx (parity only), tobacco Excluded women with other drugs, or heavy EtOH | THC in meconium associated with lower birthweight (3429 g vs 2853 g; P < .001), persistent effect in multivariable logistic regression Self-report alone was not associated with lower birthweight | Study designed to assess tobacco exposure primarily Sampling strategy for high prevalence of use not reported |
| Hatch et al, 1986 (n = 3857) | 366 (10) | Planned delivery at single institution | Self-report by structured interview early in pregnancy | Frequency: none, occasional (≤1 times/mo), regular (≥2 times/mo) | OB hx, standard demo, other drugs EtOH, tobacco | Regular use in white women associated with LBW (OR, 2.6; 95% CI, 1.1–6.2) Regular use in white women associated with SGA (OR, 2.3; 95% CI, 1.3–4.1) | Self-report data collected early in pregnancy, no later pregnancy data Differing results by racial group Marijuana use not well quantified |
| Hingson et al, 1982 (n = 1690) | 237 (14) | Single institution | Self-report by structured interview postpartum | Frequency: <1 time/mo, <1/wk, 1–2 times/wk, ≥3 per week | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | Neonates 95 g smaller than nonusers with use <3 times/wk ( P < .01) Neonates 139 g smaller than nonusers with use ≥3 times/wk ( P < .01) | Possible recall bias, most exposure data collected postpartum, small subset with prenatal interview (n = 328) Marijuana use not well quantified |
| Hurd et al, 2005 (n = 139) | 44 (32) | Women undergoing elective termination at a single center at 17–22 wks | Self-report by structured interview at time of termination Biological samples | Frequency: light (0–0.4 joints/d), moderate (0.41–0.88/d), and heavy (≥0.89 joints/d), THC in maternal urine or meconium | Standard demo, gestational age at termination, EtOH, tobacco Excluded women with cocaine/opiates | Increasing self-reported use not associated with decreasing weight Decreased birthweight in marijuana-exposed (either urine/meconium positive toxicology or self-report) fetuses by 14.53 g (–28.21 to –0.86) | Growth assessed in midgestation prior to presentation of most growth abnormalities Women in study were undergoing elective termination of pregnancy |
| Kliegman et al, 1994 (n = 425) | 34 (8) | Single institution | Self-report by structured interview at time of delivery Biological samples | THC in maternal urine at time of delivery | SES, OB hx, standard demo, other drugs, EtOH, tobacco | No association with LBW (OR, 2.28; 95% CI, 0.27–19.5) | Study designed to assess cocaine exposure primarily Marijuana use not quantified |
| Linn et al, 1983 (n = 12,424) | 1246 (10) | Single institution | Self-report by structured interview postpartum | Frequency: occasional, weekly or daily use | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with LBW for any use of marijuana (OR, 1.07; 95% CI, 0.87–1.31) | Possible recall bias, exposure data collected postpartum Marijuana use not well quantified |
| Tennes et al, 1985 (n = 756) | 257 (34) | Two affiliated institutions | Self-report by structured interview at 1 prenatal visit and postpartum | Frequency quantified by trimester: light (≤1 times/wk), moderate (>1 time/wk but <1 time/d), heavy (≥1 times/d) | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No effect on birthweight when considered by trimester or as a total amount consumed during pregnancy | Possible recall bias, only 2 sessions of self-report, which was then reported by trimester of use Marijuana use not well quantified |
| Zuckerman et al, 1989 (n = 1226) | 331 (27) | Single institution | Self-report by structured interview at 1 prenatal visit and postpartum Biological sampling | Reported use: yes/no THC in maternal urine at time of prenatal or postpartum interview | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | Positive urine toxicology screen for THC associated with 79 g decrease in birthweight ( P = .04) No association when only self-report considered | Possible recall bias, only 2 sessions of self-report Marijuana use not well quantified |
| Secondary analysis of prospective cohort a | |||||||
| Bada et al, 2005 (n = 8637) | 812 (9) | Multicenter, 4 university-based centers | Self-report by structured interview prior to delivery | Reported use: yes/no | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with LBW (OR, 1.08; 95% CI, 0.85–1.36) or SGA (OR, 0.9; 95% CI, 0.73–1.11) | Not designed to assess marijuana specifically (Maternal Lifestyle Study ) Marijuana use not quantified |
| Gibson et al, 1983 (n = 7301) | 392 (5) | Two affiliated institutions | Self-report by structured interview at 1 prenatal visit and postpartum | Frequency: ≤1 times/wk, >1 time/wk | Standard demo, OB hx (parity only), EtOH, tobacco | No association with LBW after excluding premature neonates | Marijuana use not well quantified |
| Janisse et al, 2014 (n = 3090) | 748 (24) | Single institution | Self-report by structured interview at each prenatal visit | Proportion of prenatal visits with reported use: 1–33%, 34-66%, or 67-100% | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | 55 g decrease in fetal growth with ongoing marijuana use (reported at 67–100% of visits) ( P < .004) | Study designed to assess EtOH exposure Population limited to African Americans Marijuana use not well quantified |
| Kline et al, 1987 (n = 2815) | 275 (10) | Two overlapping prospective cohorts at 3 urban hospitals | Self-report by structured interview at 1 prenatal visit | Frequency: <1 time/mo, 2–3 times/mo, 2–3 times/wk, 4–6 times/wk, daily | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with FGR in early cohort Decreased growth with increased use (127 g less with 2–3 times/wk, 143 g less with 4–6 times/wk and 230 g less with daily) in late cohort | Study designed as a case-control study with SAB as primary outcome Differing results for 2 overlapping prospective cohorts Marijuana use not well quantified |
| Saurel-Cubizolles et al, 2014 (n = 13,545) | 156 (1) | Population-based study, all births in France during a single week | Self-report by structured interview 2–3 d postpartum | Frequency: <1 time/mo, 1–9 times/mo, ≥10 times/mo | SES, standard demo, EtOH, tobacco | No association with SGA for <1 time/mo use (OR, 1.29; 95% CI, 0.61–2.72) or for use ≥1 times/mo use compared with nonusers (OR, 1.30; 95% CI, 0.66–2.56) Also no association with SGA for non-tobacco users, marijuana only | Recall bias Low prevalence of use concerning for ascertainment bias for marijuana exposure Marijuana use not well quantified |
| Shiono et al, 1995 (n = 7470) | 822 (11) | Multicenter, 7 university-based clinics | Self-report by structured interview at 1 prenatal visit Biological samples | Frequency: number of times/wk THC in maternal serum | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with LBW when marijuana use assessed by self-report or positive serum assay for THC (OR, 1.1; 95% CI, 0.9–1.5) Increased odds of LBW with positive serum assay in isolation but not with self-report | Study designed to assess association between vaginal infections and PTB Marijuana use not well quantified |
| Teitelman et al, 1990 (n = 1206) | 95 (8) | Planned delivery at single institution | Self-report by structured interview early in pregnancy | Reported use: yes/no | OB hx, standard demo, other drugs, EtOH, tobacco | No association with LBW (OR, 1.57; 95% CI, 0.54–4.52) | Study designed to assess associations between maternal work activity and LBW Same cohort as Hatch et al study with different inclusion criteria (employed women) No quantification of marijuana use |
| van Gelder et al, 2010 (n = 5871) | 189 (3) | Population-based, US National Birth Defects Prevention Study | Self-report by structured interview 6 wks to 24 mo after delivery | Reported use: yes/no by trimester | SES, OB hx, medical hx, standard demo, other drugs, EtOH, tobacco | No association with LBW (OR, 0.7; 95% CI, 0.3–1.6) No difference in mean birthweight (–17 g; P = .65) No difference by trimester of use | Recall bias, interviews up to 2 years postpartum Not designed for marijuana exposure specifically (birth defects registry) Marijuana use not well quantified |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
