Toxicology Screening for Marijuana and Impact on Breast Milk Feeding Policies in Neonatal Intensive Care Units

Introduction

Marijuana use in pregnant and lactating women is becoming more prevalent in the United States. ¹ Although many women abstain or reduce marijuana intake during pregnancy, some maintain or even increase frequency of use.^2,3 Many women believe that cannabis is safer, more effective, and more affordable than prescribed medications for common pregnancy-related conditions including nausea, insomnia, and anxiety.^4,5

Tetrahydrocannabinol (THC), the major psychoactive cannabinoid of marijuana, transfers readily from mother to baby, crossing the placenta prenatally and via maternal breast milk (MBM) postnatally.^6–8 THC is a lipophilic molecule that rapidly distributes to the fetal brain and fat after ingestion by a pregnant woman, in fact THC concentrations in fetal blood can reach 10–33% of maternal concentrations.^7,9 Among women with prenatal use, prolonged excretion of THC in MBM has been demonstrated for 6 weeks or greater, even in women who abstain from use after delivery. ¹⁰ The impact of prenatal and postnatal exposure to THC on neonatal development is not fully understood. Prenatal exposure is thought to affect the neurotransmitter systems and functions in the fetus, which may ultimately affect their neuropsychiatric development. ⁹ Infants prenatally exposed to marijuana have lower birth weights and are more likely small for their gestational age.^11,12 Current available literature, however, is inconsistent regarding impacts of THC postnatally on the neonate, especially when it involves safety of concomitant breastfeeding and THC use.

The question of whether the benefits of MBM outweigh the risks of THC exposure remains unanswered, especially in premature neonates. Given the current lack of data, providing evidence-based guidance has been challenging for national medical organizations. The American Academy of Pediatrics (AAP) states that THC use is not a categorical contraindication to breastfeeding and that women should be informed of potential risks of exposure and encouraged to abstain from THC use. ¹³ But it remains unclear how these recommendations are being implemented in hospitals that care for mothers and newborns. A survey of U.S. newborn nursery (NBN) directors found that 29% discouraged but did not restrict breastfeeding, whereas another 16% either selectively or universally restrict MBM from women who test positive for THC. ¹⁴ However, little is known about policies in neonatal intensive care units (NICUs). Breast milk feeding policies have been shown to vary between NBNs and NICUs, ¹⁵ perhaps due to concerns over increased medical complexity of NICU patients. They are at higher risk of medical comorbidities if they do not receive MBM including necrotizing enterocolitis and sepsis, but they may also be at increased risk of complications related to THC exposure if they receive MBM from a THC-positive mother.

Given the inconsistent results of neonatal outcome studies and limited guidance from national organizations, this study aimed to assess variation in NICU policies related to toxicology screening and provision of MBM from mothers who test positive for THC. We hypothesized that there would be variation in policies across the nation, with differences associated with state legalization status.

Methods

We performed a cross-sectional survey of Level II, III, and IV NICUs across the United States on their policies for toxicology screening and MBM provision based on THC screening status. The study was determined to be exempt by the University of Maryland Baltimore (UMB) Institutional Review Board. The survey was created, distributed, and managed through the REDCap system, a secure platform for survey design and database management, maintained at UMB. ¹⁶ We identified participants utilizing the AAP Society of Neonatal Perinatal Medicine (SONPM) member listserv via which an email link to the survey was distributed in June 2023. We identified additional participants by providing a survey link to Neonatology fellowship program leadership or designated representative as available between February 2022 and November 2023. For states without survey responses via these methods, we identified Level III and/or IV NICUs in the state and either called the NICU directly or emailed NICU physicians using contact information via SONPM district websites. If we received multiple responses from one institution, we only included the first complete survey response in our analysis.

The survey was developed by the investigators and requested demographic information including name of institution (to ensure only one response), state of institution, level of NICU, respondent’s job title, academic affiliation, whether a formal MBM feeding policy exists and if so, if it differs between the NICU and NBN, available lactation support, who underwent toxicology screening (mother and/or baby), and indications for screening, whether THC was decriminalized or legal in their state. The remaining survey questions asked about each NICU’s specific policies/guidelines related to THC-positive toxicology and MBM feedings. The survey requested data on specific limitations in place including: restricting or prohibiting MBM feeds, limiting lactation consultations, and/or limiting breast pump prescriptions. The survey consisted of selectable choices or free text options when applicable.

Descriptive statistics were used to describe characteristics of the cohort. Comparison of limitations based on THC state legalization status was performed using Chi-square testing. Analyses were performed using SAS 9.4 (Carey, NC).

Results

Discussion

In this large study of U.S. NICUs, we found wide variation in toxicology screening and MBM feeding policies in the setting of maternal THC usage. We found that many NICUs do not follow current national medical organization guidelines related to MBM feeds from THC-positive mothers (to counsel and discourage THC use but not prohibit MBM use). Although THC legality is not universal and varies by state, we found no significant association between MBM limitations and THC legalization status.

MBM feeding provides short- and long-term medical benefits for both mother and baby, as well as neurodevelopmental benefits for the infant. The AAP supports MBM feeds for 2 years and beyond (as appropriate for each family),^13,18 especially for those born very low birth weight since MBM is associated with reduced risk of common comorbidities of prematurity including necrotizing enterocolitis and chronic lung disease. ¹⁹ Rare contraindications to MBM feeds do exist, including maternal consumption of illicit opioids and cocaine, given concern for neurodevelopmental impacts. Marijuana use during breastfeeding, however, is not an absolute contraindication as the impact of THC exposure is not fully understood. Since it remains unclear whether the risks of THC-exposed MBM outweigh the benefits, the AAP, American College of Obstetricians and Gynecologists, Academy of Breastfeeding Medicine, and Centers for Disease Control and Prevention all currently recommend informing mothers of potential risk and discouraging THC use but not limiting or prohibiting MBM use.^13,20–22

The true prevalence of THC use by pregnant and lactating women is unclear. Mark et al studied a cohort of women who had universal toxicology screening at the time of admission to labor and delivery and found that a range of 3–10% were positive for THC between three different institutions. ²³ In another study of universal toxicology screening of mothers who delivered neonates at <34 weeks gestation admitted to the NICU, 17% were found to be positive for THC. ²⁴

Despite increasing epidemiological data on maternal THC use during pregnancy and breast feeding, the impact of THC exposure on clinical and neurodevelopmental outcomes of neonates remains unknown. We know that cannabinoids are measurable in human milk and that levels can remain elevated throughout the day with repeated use. ⁶ Even after cessation of use, THC can still be detected in MBM for up to 6 weeks. ¹⁰ Cannabis use may alter the composition of maternal milk, decreasing both fat composition and the main immunoglobulin (secretory Immunoglobulin A [IgA]), which could impact the nutritional and immune benefits of MBM.^25–27 Developmental outcome data from human studies are limited, plagued by small sample size, clinical confounders, and conflicting results. ²⁸ However, animal models of THC exposure during early postnatal life, meant to mimic lactation exposure, have demonstrated disruption of prefrontal cortex maturation and social and behavioral abnormalities that may persist into adulthood.^29,30

Although THC exposure is a concern, only 57% of NICUs in our cohort specifically addressed THC in their feeding policy and of those, only 40% followed the previously described guidance by national organizations to recommend abstention from THC but not limit MBM feeds. Of the remaining NICUs, 20% prohibited any MBM from THC-positive mothers, approximately one-third each either prohibited MBM until repeat toxicology was negative or until mother signed a waiver or verbally agreed to abstain from use. Although the goal is to improve health outcomes by minimizing THC exposure to vulnerable neonates, selective or universal prohibition risks alienating mothers and may worsen health care disparities. ³¹ A recent study of NBN policies for MBM feeds from THC-positive women found fewer restrictions, with 16% either universally or selectively prohibiting use, and 23% discouraging but not limiting MBM. ¹⁴ Perhaps restrictions were less stringent in NBNs since infants are generally healthier and the risk of THC exposure to premature infants in the NICU is perceived as more serious. In our study, almost 33% of respondents noted that their NICU and NBN feeding policies differed. Similar differences between NBN and NICUs policies were also noted in a recent survey of perinatal health workers. ¹⁵ But regardless of whether a neonate is admitted to a NBN or NICU, it seems that national guidelines are not being followed. This practice variation means that care is different not based on patient-level clinical differences but based on the hospital where each mother–baby dyad receives care. This risks significant disparities in care across the nation.

We also found wide variation in toxicology screening policies ranging from universal screening of mother and/or baby dyads to formal risk-factor-based policies to informal clinically or provider-based approaches. Of those without universal screening who provided data, fewer than 30% noted they have a formal “risk-factor-based” policy to determine which mothers and/or infants are screened to limit bias. Some noted screening women who receive limited prenatal care or with a “known history of substance use.” Others screened based on responses to a formal questionnaire given to women during their prenatal care. However, use of maternal report may underestimate fetal exposure. In 2019, Metz et al. found that while only 6% of women reported THC use in the prior month, in fact over 22% had demonstrable marijuana metabolites when testing cord blood. ³² Data have shown that selective testing for THC risks bias in care. In a study of mother–baby dyads tested specifically for cannabis use, Sarathy et al. found that women screened were more often young (<25 years), non-Hispanic Black, or Hispanic. ³³ Similarly, Perlman et al. found that Black and Hispanic women were over four times more likely to have toxicology testing for an indication other than reported substance use compared with their White counterparts. ³⁴

Clearly more data are needed on the impact of THC exposure via MBM, as well as prenatally, on infant health and development. Until then, health care providers should educate women on the potential risks not only after delivery but ideally prior to pregnancy and during prenatal care. However, Skelton et al. found that 37% of women reported not being asked during prenatal care about marijuana use, and almost 63% reported that they were not advised against THC use during pregnancy. ³⁵ The lack of perinatal discussion of marijuana use by providers has caused some women to believe that continued use is, in fact, safe. ³ Interestingly, women in THC-legal states were more likely to be asked about and advised against THC use during pregnancy and lactation. ³⁵ Women also note that the fear of child protective services (CPS) involvement resulting in parent–child separation prevented them from discussing their THC use with providers. ³

Education must be informative and nonjudgmental in order to partner with NICU mothers and accomplish shared decision-making. To minimize risk of unconscious and conscious bias, all mothers should be screened for toxicology exposures during their prenatal care, whether this is by urinary testing, screening questionnaire, or a combination. Ideally, this would be based on a conversation between the patient and her health care provider, but for this to be successful, punitive measures such as CPS referrals for isolated THC exposure without other social concerns should be stopped, especially in locations where THC is decriminalized/legal. These referrals make mothers less likely to report use for fear of punishment ³ and therefore less likely to get appropriate medical counseling throughout pregnancy and during lactation. Future studies could evaluate family outcomes after CPS referral for isolated THC-positive toxicology screen to provide more objective data to determine utility of referral for isolated THC status.

Although this is the largest study to date evaluating toxicology screening policies and impact of THC exposure on provision of MBM feedings specifically in U.S. NICUs, there are some limitations that must be noted. We were not able to obtain data from all NICUs, so there is risk to generalizability of the results, although we were able to obtain data from all AAP NICU districts and the vast majority of states as well as DC. Although our cohort is not as well representative of Level II and III NICUs, we did obtain results from almost half the Level IV NICUs across the nation. Risk of recall bias exists since we relied upon survey response. Respondents had the opportunity to submit a copy of their policy, but none provided this. However, we requested that the provider with the best knowledge of their feeding policy fill out the survey in an attempt to obtain the most accurate data and only allowed one response per NICU to avoid duplication.

Conclusion

Substantial variation exists in NICU policies regarding both toxicology screening for mother–baby dyads and the provision of MBM feeds from THC-positive mothers. This variation was not found to be associated with THC state legalization status. In addition, NBN and NICU policies from the same hospital are often different, which could lead to confusion for newborn care providers and families. More research is needed on the effect of THC exposure on neonates, but, in the meantime, we could limit inconsistent care by following current national medical organization guidelines of education and shared decision-making with mothers. Policies should focus on educating women before pregnancy, during prenatal care, and after delivery on the potential risks of fetal and neonatal THC exposure.