Objective
We examined whether maternal opioid treatment between 1 month before pregnancy and the first trimester was associated with birth defects.
Study Design
The National Birth Defects Prevention Study (1997 through 2005) is an ongoing population-based case-control study. We estimated adjusted odds ratios (ORs) and 95% confidence intervals (CIS) for birth defects categories with at least 200 case infants or at least 4 exposed case infants.
Results
Therapeutic opioid use was reported by 2.6% of 17,449 case mothers and 2.0% of 6701 control mothers. Treatment was statistically significantly associated with conoventricular septal defects (OR, 2.7; 95% CI, 1.1–6.3), atrioventricular septal defects (OR, 2.0; 95% CI, 1.2–3.6), hypoplastic left heart syndrome (OR, 2.4; 95% CI, 1.4–4.1), spina bifida (OR, 2.0; 95% CI, 1.3–3.2), or gastroschisis (OR, 1.8; 95% CI, 1.1–2.9) in infants.
Conclusion
Consistent with some previous investigations, our study shows an association between early pregnancy maternal opioid analgesic treatment and certain birth defects. This information should be considered by women and their physicians who are making treatment decisions during pregnancy.
Major birth defects affect about 3% of the 4 million US live births each year and are a leading cause of infant mortality. Congenital heart defects (CHD) are among the most common birth defects, affecting nearly 1% of US births, and are the main contributor to infant mortality attributable to birth defects.
See related editorial, page 281
For Editors’ Commentary, see Table of Contents
Opioid medications are potent prescription analgesics that are the mainstay for treatment of severe pain. Opioids are often used in combination with nonopioid analgesics, such as acetaminophen, and lower doses are also a component of some cough suppressants. Previous studies have shown that opioid analgesic use and abuse have been increasing in recent years, but the effects of opioid use on the developing fetus during pregnancy are poorly understood.
Associations between maternal first-trimester use of the opioid analgesic codeine and CHD in infants were found in 3 of 4 previous case-control studies, with case counts ranging from 141–390 (and control counts of 176–3002). Most previous studies considered CHD of any type as a single group and had insufficient sample sizes to examine individual heart defects.
Other studies have shown associations between first-trimester codeine use and other birth defects such as orofacial clefts, but these findings have been inconsistent across studies. Effects of maternal use of opioids other than codeine have not been thoroughly studied, but previous reports have shown no increase in risks for birth defects following prenatal exposures to oxycodone, propoxyphene, or meperidine. Neural tube defects have not been associated with maternal opioid treatment in human pregnancy, but experimental studies showed increased frequencies of neural tube defects in the offspring after treatment of pregnant hamsters with high doses of morphine, meperidine, pentazocine, hydromorphone, or propoxyphene.
Despite evidence of adverse fetal effects with maternal codeine use and the paucity of data on the effects of maternal use of other opioids, such treatment is often assumed to be safe during pregnancy. Our study objective was to examine whether maternal therapeutic use of opioid analgesics in early pregnancy is associated with CHD or other birth defects.
Materials and Methods
We analyzed data from the National Birth Defects Prevention Study (NBDPS) for infants born Oct. 1, 1997, through Dec. 31, 2005. NBDPS is an ongoing multisite population-based case-control study of >30 types of major structural birth defects that focuses on exposures immediately before and during pregnancy. Each of the study sites (Arkansas, California, Georgia, Iowa, Massachusetts, New Jersey, New York, North Carolina, Texas, and Utah) ascertains deliveries with birth defects through birth defects surveillance systems using standard, detailed case definitions. The study was approved by institutional review boards of the Centers for Disease Control and Prevention and all other participating centers.
NBDPS focuses on birth defects with unknown etiologies; therefore, infants with recognized chromosomal abnormalities or single-gene disorders are excluded. Some cardiac anomalies that are usually physiological rather than pathological were also excluded, eg, patent ductus arteriosus or patent foramen ovale in premature infants. In addition, some cardiovascular anomalies were excluded from NBDPS either because of their rarity, poor ascertainment in infancy, unclear significance (eg, insufficiency of the tricuspid, mitral, or pulmonary valves), or because they were vascular (noncardiac) defects or arrhythmias (nonstructural defects). All CHD cases were confirmed by echocardiography, cardiac catheterization, surgery, or autopsy.
Classification of NBDPS cases has previously been described. Briefly, the process for NBDPS-eligible defects involves confirmation of the diagnosis from medical record abstracts and determining whether each case is an “isolated” defect, 1 of multiple unrelated major defects, or a component of a syndrome or other complex pattern. For CHD, classification is also performed on a second axis to describe the complexity of cardiac involvement, by clinicians with expertise in pediatric cardiology. “Simple” heart defects are anatomically discrete or well-recognized single entities; “associations” are common, uncomplicated combinations of heart defects; and heart defects that do not fall into either category are considered “complex.” Cardiac classification and subsequent grouping into larger categories are based on clinical and presumed developmental mechanisms that may have relevance in considering the teratogenic effect of exposures.
An annual random sample of approximately 1200 liveborn infants without birth defects (control infants) is selected from the same geographic regions and time period as the cases, either from birth certificates or birth hospitals.
Mothers are invited to participate in an hour-long computer-assisted telephone interview, conducted by interviewers in English or Spanish, between 6 weeks and 2 years after the mother’s estimated date of delivery (EDD) (average is 11 months post-EDD for cases and 9 months for controls). The interview assesses various maternal health factors, pregnancy history information, dietary and other drug exposures, and sociodemographic characteristics. Exposures are assessed for the period from 3 months before conception through the end of the pregnancy. Pregnancy was defined as the time period from conception (ie, 2 weeks after the last menstrual period) to delivery, and pregnancy “months” for this analysis were consecutive 30-day periods.
Mothers are asked about medications used for each specific illness (eg, influenza) or indication (eg, surgery) mentioned within the maternal health section of the questionnaire, and are also encouraged to report any other medications not already mentioned. Respondents are asked to report the start and stop dates, duration, and frequency of medication use using calendar dates or pregnancy months.
All medications reported in each section of the interview are compiled and coded using the Slone Drug Dictionary, which NBDPS licenses from Boston University’s Slone Epidemiology Center. This dictionary links products to their active ingredients. We defined opioid exposure as maternal report of use of ≥1 products with any of the following components taken for therapeutic reasons in any dose, duration, or frequency: codeine, hydrocodone, meperidine, oxycodone, propoxyphene, morphine, tramadol, methadone, hydromorphone, fentanyl, or pentazocine. We included individual and combination products. The exposure window of interest was the period from 1 month before to 3 months after conception.
We conducted multivariable logistic regression analyses to calculate adjusted odds ratios (ORs). We examined birth defect categories that had ≥200 cases or ≥4 exposed cases to limit analyses to defect categories that would likely have adequate statistical power and to enable us to identify potentially elevated risks among rarer defect groups. All models were adjusted for maternal age (continuous), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other), education (<12 years, ≥12 years), prepregnancy obesity (body mass index <30, body mass index ≥30, using prepregnancy weight and height self-reported by the mother during the telephone interview), periconceptional smoking status (no smoking from 1 month before to 1 month after conception, smoking at least once in the same period), and study center (10 sites previously listed). These variables were chosen based on the strength of their relation in univariate analyses with “all NBDPS birth defects.”
There were maternal interviews from 19,059 cases and 6807 controls who met the inclusion dates. The participation rate was 70% for cases and 67% for controls. We excluded participants with missing data on whether they were treated with an opioid analgesic in early pregnancy, due either to an incomplete medication history from the interview or an unknown medication start or stop date. We excluded mothers with preexisting diabetes, which has been found to be a strong independent risk factor for birth defects. To focus our investigation on therapeutic opioid analgesic use, we also excluded mothers who reported opioid exposure in the form of illicit drugs such as heroin any time during pregnancy. Our final sample for analysis included 17,449 cases with the included defects and 6701 controls ( Figure ). Although the total number of cases included was greater than the total number of controls, controls substantially outnumbered cases in most analyses, each of which involved comparisons of infants with 1 particular category or subcategory of birth defects to all controls. Of note, infants with >1 defect were included in multiple birth defect categories.
Our primary analysis tested the hypothesis that maternal therapeutic use of opioid analgesics in early pregnancy was associated with the occurrence of CHD, cleft lip and/or palate, or neural tube defects in infants. Secondary, exploratory analyses were also performed in other birth defect groups.
We conducted subanalyses limiting either the exposure or birth defect groups. First, using the same case groups and exposure definition as the main analysis, we restricted the exposure time period to only the first 2 months after conception. Next, using the original case groups and the original exposure time period (1 month before to 3 months after conception), we estimated the effects of exposure to particular opioid analgesics. Because of the decreased sample size, we estimated crude ORs only for the most commonly reported components. Finally, using the same exposure definition as the main analysis, we examined cases with “isolated” birth defects or “simple, isolated” heart defects separately.
Results
Of the 17,449 mothers of case infants with various included birth defects, 454 (2.6%) reported opioid analgesic treatment between 1 month before and 3 months after conception. Among 6701 control mothers, 134 (2.0%) reported treatment in this time period. The most commonly reported opioids were codeine (34.5%), hydrocodone (34.5%), oxycodone (14.4%), and meperidine (12.9%), with codeine and hydrocodone exposure being slightly more common among cases, and oxycodone and meperidine exposure slightly more common among controls. Lower maternal education level, prepregnancy obesity, and periconceptional smoking were all slightly more frequent among cases than controls ( Table 1 ).
| Variable | Cases n (%) | Controls n (%) | OR (95% CI) |
|---|---|---|---|
| Study site | 17,449 | 6701 | |
| Arkansas | 2328 (13.3) | 832 (12.4) | (referent) |
| California | 2262 (13.0) | 845 (12.6) | 1.0 (0.9–1.1) |
| Georgia | 2082 (11.9) | 728 (10.9) | 1.0 (0.9–1.1) |
| Iowa | 1731 (9.9) | 756 (11.3) | 0.8 (0.7–0.9) |
| Massachusetts | 2386 (13.7) | 851 (12.7) | 1.0 (0.9–1.1) |
| New Jersey | 1513 (8.7) | 573 (8.6) | 0.9 (0.8–1.1) |
| New York | 1258 (7.2) | 592 (8.8) | 0.8 (0.7–0.9) |
| North Carolina | 667 (3.8) | 395 (5.9) | 0.6 (0.5–0.7) |
| Texas | 2176 (12.5) | 762 (11.4) | 1.0 (0.9–1.1) |
| Utah | 1046 (6.0) | 367 (5.5) | 1.0 (0.9–1.2) |
| Maternal age at delivery, y | 17,449 | 6701 | |
| <20 | 1898 (10.9) | 711 (10.6) | 1.1 (1.0–1.2) |
| 20-24 | 4111 (23.9) | 1531 (22.9) | 1.1 (1.0–1.2) |
| 25-29 | 4478 (25.7) | 1777 (26.5) | (referent) |
| 30-34 | 4313 (24.7) | 1743 (26.0) | 1.0 (0.9–1.1) |
| 35-39 | 2139 (12.3) | 803 (12.0) | 1.1 (1.0–1.2) |
| ≥40 | 509 (2.9) | 136 (2.0) | 1.5 (1.2–1.8) |
| Maternal race/ethnicity | 17,397 | 6672 | |
| Non-Hispanic white | 10,472 (60.2) | 3993 (59.9) | (referent) |
| Non-Hispanic black | 1706 (9.8) | 761 (11.4) | 0.9 (0.8–0.9) |
| Hispanic | 4033 (23.2) | 1486 (22.3) | 1.0 (1.0–1.1) |
| Other | 1186 (6.8) | 432 (6.5) | 1.0 (0.9–1.2) |
| Maternal education, y | 17,326 | 6649 | |
| <12 | 3092 (17.9) | 1121 (16.9) | 1.1 (1.0–1.2) |
| ≥12 | 14,234 (82.2) | 5528 (83.1) | (referent) |
| Maternal prepregnancy obesity | 16,722 | 6432 | |
| Body mass index <30 | 13,658 (81.7) | 5399 (83.9) | (referent) |
| Body mass index ≥30 | 3064 (18.3) | 1033 (16.1) | 1.2 (1.1–1.3) |
| Folic acid supplement use a | 17,446 | 6701 | |
| No use/use at another time | 8608 (49.3) | 3293 (49.1) | (referent) |
| Any use | 8838 (50.7) | 3408 (50.9) | 1.0 (0.9–1.0) |
| Smoking a | 17,358 | 6666 | |
| None | 13,728 (79.1) | 5407 (81.1) | (referent) |
| Any | 3630 (20.9) | 1259 (18.9) | 1.1 (1.1–1.2) |
a Reported use from 1 month before to 1 month after pregnancy conception.
The reasons for opioid use were based on the questionnaire topic under which the medication was reported. For the 66% of exposed women whose treatment could be linked to a specific reason, opioids were most commonly reported within the surgical procedures (41%), infections (34%), chronic diseases (20%), and injuries (18%) sections of the questionnaire.
The primary analysis included a total of 7724 infants with ≥1 of 15 different kinds of CHD. Some of these defects were grouped into 1 of 4 larger phenotypic categories (conotruncal defects, left ventricular outflow tract obstruction defects, right ventricular outflow tract obstruction defects, or septal defects) or into 1 of 2 CHD associations (ventricular septal defect + atrial septal defect, ventricular septal defect + pulmonary valve stenosis), and ORs were estimated for each case group and higher level classification in comparison to the same set of 6701 control infants.
Effect estimates were statistically significantly elevated for all eligible CHD combined (OR, 1.4; 95% confidence interval [CI], 1.1–1.7). Statistically significant associations with maternal opioid use were found among infants with conoventricular septal defect, atrioventricular septal defect, atrial septal defect (not otherwise specified), hypoplastic left heart syndrome, tetralogy of Fallot, or pulmonary valve stenosis ( Table 2 ).
| Birth defect | Total no. a | No. exposed | aOR (95% CI) |
|---|---|---|---|
| Hypothesis-testing analysis | |||
| Controls | 6701 | 134 | Referent |
| Anencephaly/craniorachischisis | 340 | 9 | 1.7 (0.84–3.4) |
| Spina bifida | 718 | 26 | 2.0 (1.3–3.2) |
| Any of included heart defects | 7724 | 211 | 1.4 (1.1–1.7) |
| Laterality defects with CHD | 198 | 4 | 1.2 (0.42–3.2) |
| Atrioventricular septal defect | 175 | 9 | 2.4 (1.2–4.8) |
| Anomalous pulmonary venous return | 206 | 4 | 0.71 (0.22–2.3) |
| Single ventricle/complex | 201 | 4 | 1.1 (0.42–3.2) |
| Conotruncal defects | 1481 | 41 | 1.5 (1.0–2.1) |
| Tetralogy of Fallot | 672 | 21 | 1.7 (1.1–2.8) |
| D-transposition of great arteries | 461 | 10 | 1.1 (0.56–2.1) |
| Ventricular septal defect conoventricular | 110 | 6 | 2.7 (1.1–6.3) |
| Left ventricular outflow tract obstruction defects | 1195 | 36 | 1.5 (1.0–2.2) |
| Hypoplastic left heart syndrome | 357 | 17 | 2.4 (1.4–4.1) |
| Coarctation of aorta | 630 | 11 | 0.88 (0.47–1.6) |
| Aortic stenosis | 253 | 7 | 1.3 (0.61–2.9) |
| Right ventricular outflow tract obstruction defects | 1175 | 40 | 1.6 (1.1–2.3) |
| Pulmonary valve stenosis | 867 | 34 | 1.7 (1.2–2.6) |
| Septal defects | 3482 | 87 | 1.2 (0.93–1.6) |
| Ventricular septal defect perimembranous | 1402 | 29 | 0.99 (0.65–1.5) |
| Atrial septal defect secundum | 1507 | 43 | 1.3 (0.94–1.9) |
| Atrial septal defect not otherwise specified | 511 | 17 | 2.0 (1.2–3.6) |
| CHD association: ventricular septal defect + atrial septal defect | 528 | 17 | 1.7 (1.0–2.9) |
| CHD association: pulmonary valve stenosis + ventricular septal defect | 131 | 4 | 1.3 (0.46–3.7) |
| Cleft palate | 936 | 25 | 1.3 (0.84–2.0) |
| Cleft lip with cleft palate | 1162 | 33 | 1.4 (0.96–2.1) |
| Cleft lip without cleft palate | 614 | 9 | 0.68 (0.34–1.3) |
| Exploratory analysis | |||
| Controls | 6701 | 134 | Referent |
| Amniotic band syndrome/limb body wall complex | 203 | 5 | 1.0 (0.37–2.9) |
| Hydrocephaly | 301 | 11 | 2.0 (1.0–3.7) |
| Cataracts | 217 | 7 | 1.6 (0.72–3.5) |
| Glaucoma/anterior chamber defects | 103 | 5 | 2.6 (1.0–6.6) |
| Anotia/microtia | 403 | 4 | 0.77 (0.28–2.1) |
| Esophageal atresia | 434 | 12 | 1.4 (0.76–2.5) |
| Intestinal atresia/stenosis | 266 | 4 | 0.88 (0.32–2.4) |
| Anorectal atresia/stenosis | 623 | 18 | 1.5 (0.87–2.4) |
| Hypospadias second/third degree | 1313 | 29 | 0.92 (0.59–1.4) |
| Bilateral renal agenesis or hypoplasia | 112 | 4 | 1.3 (0.40–4.2) |
| Longitudinal limb deficiency | 269 | 6 | 1.1 (0.49–2.6) |
| Longitudinal preaxial limb deficiency | 157 | 4 | 1.3 (0.48–3.6) |
| Transverse limb deficiency | 415 | 7 | 1.0 (0.46–2.2) |
| Craniosynostosis | 806 | 16 | 0.82 (0.48–1.4) |
| Diaphragmatic hernia | 507 | 12 | 1.2 (0.66–2.2) |
| Omphalocele | 267 | 7 | 1.3 (0.60–2.8) |
| Gastroschisis | 726 | 26 | 1.8 (1.1–2.9) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
