Marijuana and Breastfeeding: A Pilot Survey of Mothers

The study design for this pilot study was a cross-sectional survey of postpartum mothers at a single urban, academic hospital from 2018 to 2019. Participants were recruited in-person during their birth hospitalization by a member of the research team. Mothers were recruited consecutively on days when the research team member was available. Participants consented to a self-administered, paper-based survey, and data abstraction was based on a maternal and infant chart review. Mothers were eligible for inclusion if they were ≥ 18 years with a newborn ≥ 35 weeks’ gestation who was sufficiently healthy to room-in on the routine postpartum unit. Data were not collected on those who were eligible but declined to participate in the study.

The survey (available upon request) contained 48 items, including previously validated questions adapted from published drug use and KAP surveys, assessing KAPs related to marijuana use and breastfeeding. Survey questions on practices asked about past and current use of marijuana, whereas attitude-related questions asked about general feelings toward people who use marijuana. Several items asked about maternal knowledge on specific topics, for example the presence of marijuana in breast milk, its storage in body fat tissue, general harmful effects, and potential for harm to the fetus. A research team member stayed in or near the mother’s room to answer any clarifying questions or address concerns while participants completed the survey. Data abstracted from the linked maternal and infant chart included sociodemographic information such as maternal health insurance and postpartum information, including maternal parity and counseling related to marijuana use and breastfeeding. The study was reviewed and approved by the hospital’s institutional review board.

Responses were aggregated and analyzed using IBM SPSS version 25 (IBM Corp, 2017). Descriptive statistics were tabulated including frequencies and means. This report focuses on the following 3 specific maternal knowledge items: (1) that marijuana use during pregnancy may cause childhood learning or behavior problems, (2) that THC or marijuana is found in breast milk, and (3) that marijuana use may be associated with harmful effects to the breastfed baby. We tested associations between these outcome variables and predictors, consisting of sociodemographic characteristics and other KAP responses. We also tested associations among the outcome variables. χ² and Fischer’s exact tests were used for comparisons.

The 46 participants were largely multiparous and had some college education, and just over 50% were between the ages of 18 and 30 years (Table 1). The second column of Table 1 displays characteristics and KAPs for the total study population. Almost one-half of participants (46%) knew that marijuana or THC is found in breast milk, and most (87%) were aware that use while breastfeeding can be harmful to the infant. The majority (76%) felt “neutral” or approved of people who use marijuana. Nearly one-half (46%) were not confident in their marijuana knowledge (not shown in Table 1), and only 37% knew that marijuana was legal only for medical use in the state where the study took place. Over one-half of participants (57%) reported ever using marijuana, 13% within the past 12 months, and none in the past 30 days. Seven participants had urine drug screening performed during the birth hospitalization, and of these only 1 was positive for cannabinoids. Approximately one-third of participants reported prenatal (35%) or postpartum (30%) health care provider (HCP) counseling regarding the risks of marijuana use while breastfeeding.

Most (n = 31, 67%) knew that marijuana use during pregnancy may cause child learning and/or behavior problems. Table 1 columns 3 and 4 show differences in characteristics and responses among women who reported such knowledge (“aware”) as compared to those who did not (“unaware”). A higher percentage of those who knew that marijuana use during pregnancy may cause learning and/or behavior problems knew that use during pregnancy has somewhat to very harmful potential effects to a fetus (94% vs. 60%, P = .01). Similarly, a higher percentage of those who knew about child learning and/or behavior problems also felt that marijuana use may have somewhat to very harmful effects on a breastfeeding baby (97% vs. 67%, P = .03).

Table 2 provides an overview of participants who knew that marijuana or THC is found in breast milk (left-hand side) and those who knew that use has potential harmful effects to a breastfeeding baby (right-hand side). Those who knew that marijuana or THC was stored in fat when compared to those who did not were also more likely to know that marijuana or THC is found in breast milk (63% vs. 33%, P = .05). Those who knew that marijuana use may be harmful to a fetus when compared to those who did not were also more likely to know that use has potential harmful effects to a breastfeeding baby (100% vs. 29%, P < .001). Marijuana use in the past 12 months was associated with decreased knowledge that marijuana use has potential harmful effects for a breastfeeding baby (50% vs. 95%, P = .03). Similarly, those with tobacco use in the past 12 months when compared to those without use were less likely to know that marijuana use has potential harmful effects to a breastfeeding baby (50% vs. 95%, P = .01). Those reporting prenatal HCP discussion about the risk of marijuana use while breastfeeding were more likely to know that marijuana or THC is found in breast milk (69% vs. 33%, P = .02).

In this pilot survey of mothers in a single birthing hospital, past and recent marijuana use were relatively common. The majority of mothers were aware that marijuana use during pregnancy is associated with potential child learning and behavior problems and knew that use while breastfeeding may be harmful to the breastfeeding baby. Despite knowing that there may be a risk to breastfeeding while using marijuana, over one-half of participants did not know that marijuana or THC can be found in breast milk. Those who received prenatal counseling about the risk of marijuana use while breastfeeding were more likely to know that marijuana or THC is found in breast milk, suggesting that counseling had a positive impact on mothers’ knowledge. However, a minority of participants (35%) reported receiving such counseling.

This study is limited by the small, single-center sample, which limits generalizability and statistical power. Because this pilot study was conducted in a state where recreational marijuana use was not legal at the time of the study, it is possible that there was unmeasured bias in those who chose not to participate and/or in the responses among those who agreed to participate. Furthermore, the number of participants with marijuana use during pregnancy was too small for subgroup analyses. However, these findings provide preliminary evidence for variation in maternal knowledge and attitudes regarding marijuana use and breastfeeding even within 1 institution. Our results suggest that there are opportunities to standardize patient education in the prenatal and postpartum periods. In particular, provider counseling may positively impact knowledge about the potential harmful effects of marijuana use during pregnancy and breastfeeding, and as a result reduce the potential for adverse child neurodevelopmental outcomes. To obtain a larger sample size and allow for regional differences in marijuana legalization and in KAPs among postpartum mothers, a national, multisite study in collaboration with the Academic Pediatric Association Better Outcomes through Research for Newborns (BORN) network is underway as of 2021. We anticipate that this research will improve our understanding of maternal KAPs related to marijuana use and help to inform patient educational materials and quality improvement initiatives related to this timely and important topic.