In utero effects of opioids on fetal growth

Abstract

Background

Studies indicate antenatal opioid use is associated with birth size deficits, as evidenced by reductions in birth weight and head circumference. However, there remains a limited understanding of how early this growth restriction occurs, and what specific parameters are affected. This novel study evaluated global and specific growth deficits associated with prenatal opioid exposure between 18–22 weeks’ gestation as assessed during anatomy ultrasounds.

Methods

Pregnant women who completed an anatomy ultrasound were identified via electronic medical records from a large academic obstetric practice. The study group used opioids, with tobacco and/or marijuana use permitted (n = 41). The control group could have used tobacco and/or marijuana, but not opioids (n = 308). Neither group had alcohol or other drug exposure. Records were reviewed for medical history and ultrasound size parameters, coded as percentiles for gestational age.

Results

Demographics and medical histories were compared with several significant differences noted. After controlling for these differences, significant (p < 0.05) growth deficits were identified in opioid-exposed fetuses. Specifically, reductions >10 percentile points were observed in head circumference, biparietal diameter, and humerus length for opioid-exposed fetuses compared to controls. Additionally, intrauterine growth restriction (IUGR) was diagnosed five times more often. Femur length was significantly reduced in opioid-exposed fetuses prior to adjustment for confounding (p = .016), but this reduction was not significant (p = .072) after controlling for background differences. Estimated fetal weight (p = .274) and abdominal circumference (p = .633) were not significantly different between exposure groups.

Conclusion

Fetal opioid exposure predicted various bone growth deficits during routine anatomy ultrasound, indicating the effects of opioid exposure on size deficits may be evident as early as 18–22 weeks’ gestation. These findings may also suggest that in utero opioid exposure negatively impacts bone growth specifically rather than weight or fat/muscle mass. Additional studies with larger sample sizes may also reveal significantly reduced femur length, further supporting a negative impact on bone growth. Future studies evaluating bone health and immune function in children after antenatal opioid exposure may help clarify this specific effect of opioids on bone development.

Keywords:

• Maternal opioid use
• fetal growth
• bone

Introduction

Recent studies suggest that 14–22% of pregnant women nationwide are prescribed at least one opioid during their pregnancy [1,2]. Of women that receive an opioid prescription during pregnancy, an estimated 39% receive one during the first trimester, which is known to be a significant period for teratogenic influence and specifically makes the fetus vulnerable to structural abnormalities [1]. National trends demonstrate that the third wave of the opioid epidemic has also led to a rapid increase in the illicit use of synthetic opioids such as Tramadol and Fentanyl [3,4]. Though there is limited information on illicit opioid use - especially synthetic opioid use - during pregnancy, studies have shown that most women that use opioids are of reproductive age [5]. The alarming rate of synthetic opioid use in the United States by women of childbearing age, along with the reality that 45% of pregnancies are unintended, underscores the need to examine the effects of opioid use on fetal development [6].

Because rates of antenatal maternal opioid use are difficult to ascertain, diagnosis of neonatal opioid withdrawal syndrome (NOWS), which occurs in about half of exposed pregnancies, may inform prevalence estimates of fetal opioid exposure. From 2010 to 2017, the incidence of NOWS increased from 4 to 7.3 per 1000 hospital births [7]. This increase in neonatal opioid exposure is indicative of an increase in prescriptions, the evolving opioid epidemic, and the prevalence of opioid medication-assisted treatment (MAT) with methadone or suboxone.

Of women who were prescribed an opioid during pregnancy, nearly one-third reported that they never received counselling from their provider regarding the risks of opioid use on their fetus [2]. Impacts include an increased rate of preterm delivery, low birth weight (LBW), small for gestational age (SGA), neonatal opioid withdrawal syndrome (NOWS), and intrauterine growth restriction (IUGR) [8–10]. IUGR, also referred to as Fetal Growth Restriction (FGR), describes a fetus that is not growing at expected rates given gestational age. Size at birth is the strongest predictor of infant mortality, highlighting an additional risk of opioid use during pregnancy [11].

Research focused on reductions in head circumference in opioid-exposed fetuses indicated humerus and femur lengths may also be reduced, leading to the suggestion that in utero opioid exposure may specifically impact bone growth [12]. Animal models revealed that fetal morphine exposure may impair intra-cartilaginous osteogenesis in rats, evidenced by reductions in growth plate thickness and cell density, as well as reductions in longitudinal bone growth [13]. Although some in utero opioid exposure-related size deficits have been established, there remains a limited understanding of how early this growth restriction occurs, and what specific parameters are impacted. This study evaluated the association between prenatal opioid exposure and both global and specific growth deficits at 18 to 22 weeks gestation, as assessed during routine second-trimester anatomy ultrasound.

Materials and Methods

This retrospective cohort study was approved by the Institutional Review Board of the hospital system where the project took place. Participants were pregnant women who delivered between January 2012 and May 2021. Women who completed routine anatomy ultrasounds were identified via electronic medical records from a single academic obstetrics practice.

Prenatal records were screened for eligibility based on pre-defined criteria. Excluded were women under the age of 18, those pregnant with multiples, and those who had major medical diagnoses. Exclusionary conditions identified were major heart defects, systemic lupus erythematosus, current cancer, uncontrolled epilepsy, and those whose fetuses were diagnosed with or suspected to have genetic anomalies. Additionally excluded were those with unknown pregnancy drug use status, those lacking a second-trimester ultrasound, and those prescribed medications with known teratogenic effects.

During chart review, substance use was identified based on patient self-report at any prenatal visit or at delivery, and via biologic testing, including urine drug screen (UDS) and cord blood tests. Only those with biochemical verification of substance exposure via UDS or cord blood tests were included in the sample. Patients included in the study group used opioids either illicitly or via prescription (n = 41), with tobacco and/or marijuana use permitted. Control group women may also have used tobacco and/or marijuana, but not opioids (n = 308). Because many patients that used opioids simultaneously used tobacco and/or marijuana, the use of either substance was permitted in both study and control groups. Women with any evidence of alcohol or other illicit drug use were excluded. It was originally intended to further stratify opioid use during pregnancy based on type (illicit vs prescription), duration of exposure, dosing, and timing during pregnancy. However, this information was not readily available in all records and thus could not be included as study variables.

To account for the impact of socioeconomic status on fetal growth, demographic data and social determinants of health were recorded during chart review and included in data analyses. Medical records were further reviewed for medical conditions, prenatal care, and delivery information to describe the sample and to control for potential confounding. The primary outcomes analyzed in this study were fetal size measurements at the second-trimester anatomy ultrasound (18–22 weeks’ gestation), which included head circumference (HC), biparietal diameter (BPD), humerus length (HL), femur length (FL), estimated fetal weight (EFW), and abdominal circumference (AC). All measurements were converted to percentiles for gestational age. Also documented during chart review was a diagnosis of IUGR, defined as a fetus with an estimated fetal weight below the 10^th percentile for a given gestational age [14–16]. Finally, infants diagnosed with and/or pharmacologically treated for NOWS were also recorded.

Statistical analysis was conducted using IBM SPSS 28.0.0 (SPSS Inc., Chicago, Illinois, United States). Chi-square and t-tests were used to compare background characteristics between mothers in the control and study groups. Significant demographic, socioeconomic, medical, and prenatal care differences were then controlled for in the adjusted linear regression analyses predicting each size parameter from opioid exposure status. A value of p < .10 was considered significant for determining potential confounders, while a p < .05 was considered significant for primary analyses.

Results

The study sample contained 349 eligible maternal-fetal dyads. Of these, 41 fetuses were exposed to opioids and included in the study group. Within the 41 dyads included in the study group, 25 patients were receiving MAT for opioid use disorder, while an additional 4 patients were prescribed opioids for pain during pregnancy. Of the 41 opioid-exposed fetuses, 21 were diagnosed with NOWS, and 14 required pharmacologic treatment. The control group consisted of 308 fetuses who were biochemically confirmed to not have been exposed to opioids. Mothers who used opioids were significantly older, more likely to be white, with a lower BMI, greater parity, later entry to prenatal care, and had higher rates of tobacco use compared to controls. No significant differences were found between the study and control groups regarding education level, marital status, medical insurance coverage, diabetes, hypertension, gestational age at ultrasound, and fetal sex (Table 1).

Table 1. Maternal characteristics by opioid exposure status.

	Non-opioid exposed (n = 308)	Opioid exposed (n = 41)	t/χ^2	p
Age (yrs)	24.3 (4.0)	27.8 (4.5)	5.11	<.001
Education (% beyond H.S.)	19.5%	19.5%	.00	.999
Race (% white, non-Hispanic)	48.4%	78.0%	12.76	<.001
Marital status (% married)	12.3%	17.1%	.84	.656
Medical insurance (% Medicaid/none)	90.6%	90.2%	.01	.944
BMI	29.1 (7.5)	27.0 (5.6)	2.21	.031
Parity	1.2 (1.3)	1.9 (1.6)	2.89	.004
Gestational age at initial prenatal visit	14.3 (3.9)	15.9 (4.2)	2.49	.013
Diabetes (% existing or gestational)	6.8%	9.8%	.94	.333
Hypertension (% existing or gestational)	16.6%	14.6%	.10	.754
Gestational age at ultrasound	20.2 (1.3)	20.6 (1.7)	.97	.339
Fetal sex (% male)	46.8%	48.8%	.06	.807

Both before and after controlling for significant maternal background differences noted in Table 1 with a linear regression, analysis revealed significant growth deficits across several measurements in opioid-exposed fetuses (Table 2). After adjustment, significant reductions greater than 10 percentile points were observed in HC, BPD, and HL in opioid-exposed fetuses compared to controls. There was also a more than 5-fold significant increase in the risk of IUGR diagnosis for those exposed to opioids. While FL was significantly reduced in opioid-exposed fetuses before controlling for maternal background differences, this effect was no longer significant after adjustment for confounding. Finally, those with and without opioid exposure did not differ significantly in EFW nor AC (Table 2).

Table 2. Fetal growth percentiles at second-trimester ultrasound by opioid exposure status.

	Non-opioid exposed (n = 308)	Opioid exposed (n = 41)	Unadjusted^a regression coefficient	p	Adjusted^b regression coefficient	p
Estimated fetal weight	55.4	48.8	−6.6 (4.6)	.161	−5.5 (5.0)	.274
Femur length	50.3	38.4	−11.8 (4.9)	.016	−9.3 (5.2)	.072
Head circumference	56.8	46.3	−10.2 (4.7)	.029	−11.7 (5.0)	.020
Abdominal circumference	55.0	53.9	−1.1 (4.4)	.805	−2.3 (4.7)	.633
Biparietal diameter	54.7	40.2	−14.4 (4.6)	.002	−14.7 (4.9)	.003
Humerus length	57.0	33.4	−23.6 (10.8)	.031	−21.0 (10.7)	.042
IUGR diagnosis (%)^c	5.9%	14.6%	2.87 (1.06-7.77)	.038	5.18 (1.51-17.72)	.009

^a Unstandardized regression coefficient (standard error) for opioid exposure status from the first step of linear regression with no other predictors included.

^b Unstandardized regression coefficient (standard error) for opioid exposure status from the final step of linear regression with all significant background factors (see Table 1) included: maternal age, race, pre-pregnancy BMI, parity, and gestational age at a first prenatal visit.

^c Odds ratio (95% confidence interval) for opioid exposure.

Discussion

Fetal opioid exposure remains an area of concern as increases in opioid prescriptions, the evolving opioid epidemic, and the prevalence of MAT present risks to fetal development. Studies indicate antenatal opioid use is associated with birth size deficits, as evidenced by reductions in birth weight and head circumference. However, there remains a limited understanding of how early this growth restriction occurs, and what specific aspects of growth are affected. To our knowledge, this study is the first to examine these global and specific growth parameters associated with prenatal opioid exposure as early as 18–22 weeks’ gestation.

FGR presents a significant challenge in contemporary antenatal care. The aetiology of FGR is complicated, with multiple genetic and epigenetic factors that may interact to contribute to the condition. Chromosomal abnormalities associated with FGR include trisomy 13, 18, 21, triploidy, copy number variants (CNVs), uniparental disomy (UPD), and loss of heterozygosity [17–20]. Confined placental mosaicism (CPM) has also been associated with FGR [21]. Additionally, aberrant epigenetic patterns in DNA methylations and histone modifications have been observed in the placentas and fetal tissues of growth-restricted fetuses [18,19].

FGR is often evaluated using an HC/AC ratio. Symmetric (Type 1) FGR is characterized by proportional global growth deficits across BPD, HC, AC, and FL, suggesting intrinsic factors restricting growth. Symmetric FGR is predominantly associated with genetic or chromosomal abnormalities. Asymmetric (Type 2) FGR is characterized by more prominent reductions in AC and is associated with placental insufficiency [22]. Asymmetric FGR tends to be seen in cases of prenatal cigarette smoking.

In this study, fetal opioid exposure predicted various growth deficits, suggesting that the effects of opioid exposure on size are evident as early as the second trimester. Over half of the opioid-exposed newborns were diagnosed with NOWS, with a significant proportion of those diagnosed requiring pharmacological treatment for NOWS. While there was a trend for those diagnosed with NOWS, and especially those who required pharmacologic treatment for NOWS to have greater growth restriction, our study was not powered to examine this effect. In addition, opioid exposure predicted some but not all growth parameters. Examination of those impacted points more specifically to an effect on bone growth rather than overall size or weight. The findings of this study show significant growth deficits in BPD, HC, and HL, but not AC or estimated fetal weight. Because these findings do not necessarily fit the typical framework of either Type 1 or Type 2 FGR, we believe that in utero opioid exposure negatively impacts bone growth compared to other aspects of physical development.

The current findings of decreased head circumference, biparietal diameter and humerus length in opioid-exposed human fetuses lend support to a recent animal study that found fetal morphine exposure to impair intra-cartilaginous osteogenesis in rats. This impaired osteogenesis was evidenced by reductions in growth plate thickness, cell density, and reduced longitudinal bone growth [13]. Combined, these findings do suggest that in utero opioid exposure negatively impacts bone growth specifically.

Our study is not without limitations. We selected prenatal records for inclusion into the study group based on identification by patient reporting that was confirmed in the clinical setting via biological testing, including urine drug screen (UDS) and cord blood tests. The inclusion of self-reported drug use does have its limitations, as we recognize that this tends to disproportionately target BIPOC patients and can lead to significant under-reporting. However, self-reported drug use was used as an initial screener, and only patients with biochemically verified substance use via UDS or cord blood test were included in the sample. When analyzing the demographic data, the opioid-exposed group did not overrepresent BIPOC or lower SES populations compared to the patient population as a whole. Additionally, while care was taken to select records for inclusion into the control and study groups, participants in this retrospective chart review did not universally submit to serial urine drug screens throughout pregnancy. Therefore, we cannot rule out the possibility that members of either group used substances that were in the exclusion criteria and did not disclose them to their respective providers.

While it was intended to further stratify opioid use during pregnancy based on type (illicit vs prescription), duration of exposure, dosing, and timing during pregnancy, this information was not readily available in all records and thus could not be included as study variables. In regard to opioid type, we found that many of the patients that were prescribed opioids during pregnancy also misused prescription opioids. Conversely, many patients who were not on MAT or did not have an opioid prescription were found to be positive for prescription opioids. This suggested that either the prescription was not documented, or the patients were using prescription opioids illicitly, in addition to being positive for street opioids such as heroin. As such, we were unable to specifically examine the type of opioid in our study. Additionally, due to the nature of the retrospective chart review, duration of exposure, dosing, and timing were not readily available in all medical records, and thus could not be included and analyzed in this study.

In conclusion, in utero exposure to opioids was associated with significant reductions greater than 10 percentile points in biparietal diameter, head circumference, and humerus length, while reductions in abdominal circumference and estimated fetal weight were not observed. Additionally, femur length differences approached significance, further supporting a negative impact on bone growth. While this study reveals a relationship between maternal opioid use and fetal bone growth, the relationship may be further explained in a larger study that specifically examines the dosing, type, and timing of opioids. Future studies evaluating bone health and immune function in children after antenatal opioid exposure may help clarify this specific effect of opioids on bone development.

Acknowledgements

This research was made possible by the Central Michigan University College of Medicine Summer Research Scholars Program.

Disclosure statement

The authors report that there are no competing interests to declare.

Data availability statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Additional information

Funding

SD, MW, PD, HK, and KN received funding from the Central Michigan University College of Medicine Summer Research Scholars Program.