Introduction
Marijuana is believed to be the most commonly used dependent substance during pregnancy, with reported rates estimated between 2-11% of gravidae. However, there is wide variation in estimated usage rates, largely due to varying practices of screening. Recently, there have been both societal and legislative shifts in legal use of marijuana, leading to both increased use and reporting as a result of a decline in perceived risk and stigma.
While there are multiple published studies on marijuana use during pregnancy, findings are variable and not necessarily contemporary. Moreover, while concomitant dependent substance use and cigarette smoking is perceived to be common, there are limited studies reporting on concomitant use. The relationship between cigarette use and adverse fetal outcomes is an important component, and potentially a confounding factor, in the clinical picture as the adverse fetal effects of tobacco exposure are well documented.
The potential for adverse fetal outcomes are evident. Animal models demonstrate the ability of the active ingredient of marijuana (delta-9-tetrahydrocannabinol) to cross the placenta, however at much lower levels relative to the serum levels. Additionally, cannabinoid receptors are expressed in the fetal brain, providing a biologic rationale for potential fetal effects of maternal use. There have been several studies reporting adverse fetal outcomes with marijuana use, including lower birthweight babies. One large prospective study (Generation R) demonstrated decreased growth trajectory related to marijuana use during pregnancy, suggesting a dose-dependent response. However, while this study did compare marijuana use to cigarette use and demonstrated that marijuana-related adverse outcomes are more pronounced, it did not account for concurrent cigarette use, thereby limiting its interpretation. In contrast with Generation R, other studies failed to demonstrate a difference among marijuana users alone, including 1 meta-analysis. Similarly, there is conflicting evidence regarding marijuana use and preterm birth, with some older studies suggesting a correlation, including suggestion of dose-related response, with others failing to observe evidence of increased risk. Some studies have noted an increased rate of neonatal adverse events, such as rates of neonatal intensive care unit admissions. The majority of data indicates no association with congenital malformations, with the exception of 1 study that demonstrated findings consistent with fetal alcohol syndrome. While long-term data are limited, there is potential emerging evidence suggesting marijuana exposure in pregnancy may impact propensity for future use among offspring and diminished academic performance.
Notably, few studies have examined the maternal morbidity and mortality associated with marijuana smoking. There is a well-established inverse relationship between cigarette smoking and rates of preeclampsia but there are limited existing data examining whether marijuana would have similar or differing effects. In light of these knowledge gaps, and given the current social and legislative climate, we sought to examine perinatal outcomes in a large, population-based, contemporary cohort with direct query for singular and concurrent use of nicotine, tobacco, and marijuana.
Materials and Methods
Subjects
In the current study we utilized PeriBank, a perinatal database curated by Baylor College of Medicine (institutional review board no. H-26364; H-33382). At the time of admission, skilled PeriBank personnel enrolled gravidae after obtaining written informed consent. Up to 4900 variables of clinical data were sought from the electronic medical record, prenatal records, and in-person interview. The quality of the data was ascertained by regular verification of a subset of the inserted clinical data and by a board-certified maternal-fetal medicine physician scientist (Dr Aagaard) as previously published. Clinical data extracted for this study were patient history, socioeconomic status, and maternal and fetal outcomes. Not all 4900 potential PeriBank variables were employed in this analysis.
After obtaining full and informed consent, all gravidae who chose to participate were surveyed on presentation to labor and delivery regarding use of marijuana, tobacco, and nicotine-containing products. All information was prospectively entered into PeriBank. Following delivery, our trained team of obstetric researcher coordinators abstracted further data surrounding birth information and neonatal outcomes from the electronic medical record to complete the information for each patient. Thus, this database consisted of both patient-reported, interview-based data and data obtained from the electronic medical record.
Prior to asking detailed questions by our team of well-trained personnel, all participants were first consented for inclusion into PeriBank. Subsequently, a team of rigorously trained research coordinators conducted interviews by using both open-ended and directed questions, and utilizing a standardized multipage questionnaire. Care was taken to elucidate both “ever” and current exposure, as well as to quantify the amount used. Importantly, as we practice in a highly Spanish-speaking area, interpreters were consistently employed whenever English was not the preferred language of the participant. Interviews were conducted after the initial assessment of the patient by the physician, to allow for minimal anxiety and optimal comfort of the patient prior to direct questioning.
Inclusion criteria
Inclusion criteria was all women from Jan. 1, 2011, through June 1, 2015, with singleton pregnancies who delivered at one of the tertiary referral hospitals associated with Baylor College of Medicine. Exclusion criteria included any women without information regarding smoking during pregnancy, women who delivered at nonaffiliated hospitals, and those without appropriate information regarding maternal or neonatal data points of interest for the study.
Marijuana use was defined by self-disclosure of reported use during pregnancy. Remote history of marijuana use prior to pregnancy was not included. Smoking status was stratified into 4 distinct categories: nonsmokers, tobacco smokers, marijuana smokers, and tobacco and marijuana users, and perinatal outcomes examined among all groups. Maternal outcomes examined included adequacy of prenatal care, oligohydramnios, preterm birth, preterm premature rupture of membranes, placental abruption, hypertensive disorders of pregnancy, and mental health comorbidities. Preterm birth was evaluated both in totality and substratified to examine only births through 34 weeks’ gestation. Insufficient prenatal care was defined as <11 prenatal visits for full-term pregnancies as previously defined by the American Congress of Obstetricians and Gynecologists ; pregnancies with preterm deliveries were excluded from analysis for this outcome. Neonatal outcomes examined included 5-minute Apgar scores <7, decreased birthweight (defined as <25th percentile), decreased head circumference (defined as <25th percentile), fetal growth restriction (defined as <10th percentile), as well as head circumference <2 SD below the mean.
During the study period, 13,919 gravidae were directly interviewed as part of their inclusion in the database. Of these subjects, 450 were excluded for inadequate information on smoking tendencies during pregnancy, and an additional 1400 patients were removed for insufficient delivery and postnatal information. Thus, in total 12,069 women met inclusion criteria.
Statistical analysis
The frequency of reported marijuana and cigarette use, both individually and in combination, were determined. Maternal characteristics and demographics of the groups were compared using Mann-Whitney U test. We used χ 2 analysis and Fisher exact test to determine significance of determined outcomes between groups. Logistic regression was used to calculate adjusted odds ratios (aOR) after controlling for age, parity, race, marital status, and mode of delivery (only when investigating the use of regional anesthesia). Additionally, we adjusted for chronic hypertension and diabetes (both gestational and pregestational), as these are strongly correlated with development of preeclampsia. All statistical analysis was performed in statistical software (SPSS, Version 22.0; IBM Corp, Armonk, NY). A P value <.05 was considered statistically significant.
Materials and Methods
Subjects
In the current study we utilized PeriBank, a perinatal database curated by Baylor College of Medicine (institutional review board no. H-26364; H-33382). At the time of admission, skilled PeriBank personnel enrolled gravidae after obtaining written informed consent. Up to 4900 variables of clinical data were sought from the electronic medical record, prenatal records, and in-person interview. The quality of the data was ascertained by regular verification of a subset of the inserted clinical data and by a board-certified maternal-fetal medicine physician scientist (Dr Aagaard) as previously published. Clinical data extracted for this study were patient history, socioeconomic status, and maternal and fetal outcomes. Not all 4900 potential PeriBank variables were employed in this analysis.
After obtaining full and informed consent, all gravidae who chose to participate were surveyed on presentation to labor and delivery regarding use of marijuana, tobacco, and nicotine-containing products. All information was prospectively entered into PeriBank. Following delivery, our trained team of obstetric researcher coordinators abstracted further data surrounding birth information and neonatal outcomes from the electronic medical record to complete the information for each patient. Thus, this database consisted of both patient-reported, interview-based data and data obtained from the electronic medical record.
Prior to asking detailed questions by our team of well-trained personnel, all participants were first consented for inclusion into PeriBank. Subsequently, a team of rigorously trained research coordinators conducted interviews by using both open-ended and directed questions, and utilizing a standardized multipage questionnaire. Care was taken to elucidate both “ever” and current exposure, as well as to quantify the amount used. Importantly, as we practice in a highly Spanish-speaking area, interpreters were consistently employed whenever English was not the preferred language of the participant. Interviews were conducted after the initial assessment of the patient by the physician, to allow for minimal anxiety and optimal comfort of the patient prior to direct questioning.
Inclusion criteria
Inclusion criteria was all women from Jan. 1, 2011, through June 1, 2015, with singleton pregnancies who delivered at one of the tertiary referral hospitals associated with Baylor College of Medicine. Exclusion criteria included any women without information regarding smoking during pregnancy, women who delivered at nonaffiliated hospitals, and those without appropriate information regarding maternal or neonatal data points of interest for the study.
Marijuana use was defined by self-disclosure of reported use during pregnancy. Remote history of marijuana use prior to pregnancy was not included. Smoking status was stratified into 4 distinct categories: nonsmokers, tobacco smokers, marijuana smokers, and tobacco and marijuana users, and perinatal outcomes examined among all groups. Maternal outcomes examined included adequacy of prenatal care, oligohydramnios, preterm birth, preterm premature rupture of membranes, placental abruption, hypertensive disorders of pregnancy, and mental health comorbidities. Preterm birth was evaluated both in totality and substratified to examine only births through 34 weeks’ gestation. Insufficient prenatal care was defined as <11 prenatal visits for full-term pregnancies as previously defined by the American Congress of Obstetricians and Gynecologists ; pregnancies with preterm deliveries were excluded from analysis for this outcome. Neonatal outcomes examined included 5-minute Apgar scores <7, decreased birthweight (defined as <25th percentile), decreased head circumference (defined as <25th percentile), fetal growth restriction (defined as <10th percentile), as well as head circumference <2 SD below the mean.
During the study period, 13,919 gravidae were directly interviewed as part of their inclusion in the database. Of these subjects, 450 were excluded for inadequate information on smoking tendencies during pregnancy, and an additional 1400 patients were removed for insufficient delivery and postnatal information. Thus, in total 12,069 women met inclusion criteria.
Statistical analysis
The frequency of reported marijuana and cigarette use, both individually and in combination, were determined. Maternal characteristics and demographics of the groups were compared using Mann-Whitney U test. We used χ 2 analysis and Fisher exact test to determine significance of determined outcomes between groups. Logistic regression was used to calculate adjusted odds ratios (aOR) after controlling for age, parity, race, marital status, and mode of delivery (only when investigating the use of regional anesthesia). Additionally, we adjusted for chronic hypertension and diabetes (both gestational and pregestational), as these are strongly correlated with development of preeclampsia. All statistical analysis was performed in statistical software (SPSS, Version 22.0; IBM Corp, Armonk, NY). A P value <.05 was considered statistically significant.
Results
Of the cohort of 12,069 subjects meeting inclusion criteria, 106 (0.88%) reported marijuana use, with almost half (45%) reporting concurrent tobacco smoking ( n = 48). Tobacco smokers comprised 2% of the population ( n = 242). Demographic baseline data were notably significantly different between groups ( Table 1 ). Marijuana smokers were more likely to be younger ( P < .001), single ( P < .001), African American ( P < .001), and primigravida (P = .004) as compared to nonsmokers. In our population, co-users of marijuana and cigarettes had significantly higher rates of diabetes and chronic hypertension ( P = .005 and P = .002, respectively). There was no difference in body mass index between groups, and subjects using marijuana were no more likely to fall below the poverty line ( P = .6). Based on these univariate differences, all outcomes were adjusted for age, race, marital status, parity, chronic hypertension, and diabetes (both pregestational and gestational).
| Marijuana only | Cigarette only | Co-use | No smoking | Marijuana vs none | Cigarette vs none | Co-use vs none | |
|---|---|---|---|---|---|---|---|
| No. | 58 | 194 | 48 | 11,769 | |||
| Median ± SD or percent | Significance of difference, P value | ||||||
| Maternal age, y | |||||||
| Median age | 24 ± 5.1 | 28.2 ± 9.5 | 27 ± 5.5 | 29.1 ± 9.1 | <.001 a | .131 | .002 a |
| Parity | |||||||
| Nulliparous | 26 (45%) | 53 (26%) | 17 (35%) | 3296 (28%) | .004 a | .83 | .25 |
| Multiparous | 32 (55%) | 141 (73%) | 31 (65%) | 8473 (72%) | |||
| Race/ethnicity | |||||||
| Hispanic | 29 (50%) | 73 (39%) | 14 (29%) | 9013 (77%) | < .001 a | < .001 a | < .001 a |
| African-American | 24 (41%) | 68 (35%) | 24 (50%) | 1140 (10%) | |||
| Caucasian | 5 (9%) | 48 (25%) | 10 (21%) | 1006 (9%) | |||
| Asian | 0 (0%) | 3 (2%) | 0 (0%) | 373 (3%) | |||
| Other | 0 (0%) | 0 (0%) | 0 (0%) | 237 (2%) | |||
| Marital status | |||||||
| Single | 46 (81%) | 118 (63%) | 37 (79%) | 4245 (38%) | < .001 a | < .001 a | < .001 a |
| Separated | 2 (4%) | 9 (5%) | 0 (0%) | 249 (2%) | |||
| Divorced | 0 (0%) | 6 (3%) | 0 (0%) | 67 (1%) | |||
| Married | 9 (16%) | 55 (29%) | 10 (21%) | 6736 (60%) | |||
| Maternal comorbidities | |||||||
| Chronic hypertension | 3 (5%) | 17 (9%) | 7 (15%) | 572 (5%) | .760 | .013 a | .002 a |
| Pregestational diabetes | 1 (1.7%) | 4 (2.1%) | 3 (4.2%) | 449 (4%) | .726 | .332 | .689 |
| Body mass index | 30 ± 6.9 | 32.3 ± 7.3 | 31.9 ± 9.5 | 31.3 ± 6.2 | .500 | .261 | .483 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
