Matched controls: exact matching by maternal presurgery body mass index and diabetes (early-pregnancy body mass index and preconception diabetes used for controls), history of major birth defects in previous pregnancies, delivery year, and a propensity score based on maternal age, alcohol and substance use, smoking, parity, psychiatric drug use, and number of prescription drugs. For first birth after gastric bypass surgery, restriction was performed to avoid clustering effects (13% of mothers had >1 birth after surgery). Weights from coarsened exact matching were used to account for different sizes of matching strata.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Neovius M, Pasternak B, Näslund I, Söderling J, Johansson K, Stephansson O. Association of Maternal Gastric Bypass Surgery With Offspring Birth Defects. JAMA. 2019;322(15):1515–1517. doi:10.1001/jama.2019.12925
Maternal body mass index (BMI) and glucose control are associated with offspring birth defects.1,2 Bariatric surgery results in weight loss and glucose normalization but is also associated with nutritional deficiencies and substance abuse,3 which could cause birth defects as hypothesized based on case series.4
It could be unethical to randomize pregnant women to bariatric surgery, and it is impossible to perform a randomized trial requiring pregnancy after alternative interventions. We conducted a nationwide matched cohort study to investigate major birth defect risk in infants born to women after gastric bypass surgery vs infants born to comparable women without bariatric surgery.
We identified live-born singleton infants in the Swedish Medical Birth Register born in 2007 to 2014 to women receiving Roux-en-Y gastric bypass surgery during the same period (ascertained from the Scandinavian Obesity Surgery Register) and to women without bariatric surgery.
Using coarsened exact matching, controls were matched by major birth defects in previous pregnancies, presurgery BMI and diabetes (early-pregnancy BMI and preconception diabetes used for controls), delivery year, and a propensity score (estimated using logistic regression) including maternal age, smoking, alcohol/substance use, parity, psychiatric drugs, and number of prescription drugs.
Major birth defects, excluding genetic syndromes, defined according to the EUROCAT classification (which does not include minor defects), were identified via the National Patient Register (including inpatient and hospital-based outpatient care) and Causes of Death Register through 2015, permitting 1 year of follow-up.
Using generalized linear models with robust sandwich estimators in SAS version 9.4 (SAS Institute Inc), we estimated risk ratios in infants born after gastric bypass surgery vs control infants assuming a binomial distribution. A 2-sided P < .05 indicates statistical significance. Sensitivity analyses were conducted excluding infants with chromosomal abnormalities, restricting analysis to the first postsurgery birth, and excluding strata with high coarsened exact matching weights (>20).
The study was approved by the regional ethics committee in Stockholm, Sweden. Informed consent was not required.
Matched controls were found for 97.4% (2921/2998) of postsurgery-born infants. The groups were well balanced on maternal characteristics (Table). In the surgery group, mean presurgery BMI (calculated as weight in kilograms divided by height in meters squared) was 43.5 and mean body weight was 122 kg; median surgery-to-conception interval was 1.6 years; mean weight loss was 40 kg, resulting in a body weight of 82 kg; and diabetes drug use decreased from 9.7% before surgery to 1.5% during the 6 months before conception.
Major birth defects were recorded in 3.4% (98/2921) of infants born to mothers with gastric bypass surgery vs 4.9% (1510/30 573) of controls (risk ratio, 0.67 [95% CI, 0.52-0.87]; risk difference, −1.6% [95% CI, −2.7% to −0.6%]) (Figure). Major heart defects accounted for 60% (n = 58) of birth defects among postsurgery-born infants. There were no cases of neural tube defects in the surgery group and 20 cases (0.07%) among controls. The lower risk remained in sensitivity analyses (Figure).
In this nationwide matched cohort study, infants born to women with Roux-en-Y gastric bypass surgery had lower risk of major birth defects than infants born to matched control women.
Obesity is associated with poor glucose control, which is teratogenic.1,2 In this study, after bariatric surgery, women lost weight and diabetes drug use decreased. If the observed association is true, a mechanism could be that surgery-induced improvements in glucose metabolism, and potentially other beneficial physiologic changes, led to a reduction of major birth defect risk to a level similar to that of the general population (3.5%).1
Concerns were raised by case series that gastric bypass–induced folate deficiency could increase risk of neural tube defects.4 In this study, no neural tube defects were found in postsurgery-born infants.
A systematic review of bariatric surgery and birth defects5 could not draw conclusions from available studies because of inadequate statistical power,6 a mix of procedures with different physiologic effects, unclear outcome ascertainment, and limited confounder control. The current study had sufficient statistical power, analyzed gastric bypass surgery only, used a well-established birth defect classification applied to prospectively recorded outcome data including defects detected after delivery hospital discharge, and matched for a range of potential confounders.
Limitations include that pregnancy termination data were not available, stillbirths were not included, and individual birth defects could not be analyzed because of small numbers.
Accepted for Publication: August 5, 2019.
Corresponding Author: Martin Neovius, PhD, Clinical Epidemiology Division, Department of Medicine (Solna), Karolinska Institutet, SE-171 76 Stockholm, Sweden (email@example.com).
Author Contributions: Drs Neovius and Stephansson had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Neovius, Pasternak, Johansson, Stephansson.
Acquisition, analysis, or interpretation of data: Neovius, Pasternak, Näslund, Söderling, Stephansson.
Drafting of the manuscript: Neovius.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Neovius, Söderling.
Obtained funding: Neovius, Johansson.
Administrative, technical, or material support: Neovius, Stephansson.
Conflict of Interest Disclosures: Dr Neovius reported receipt of advisory board fees from Itrim and Ethicon Johnson & Johnson. Dr Pasternak reported receipt of an unrestricted research grant from the Novo Nordisk Foundation for a diabetes project not related to the present work. Dr Näslund reported receipt of consultancy and lecture fees from Baricol Bariatrics AB Sweden and Ethicon Johnson & Johnson and lecture fees from AstraZeneca A/S Denmark. Dr Söderling reported receipt of consultancy fees from Itrim. No other disclosures were reported.
Funding/Support: This study was funded by the National Institute of Diabetes and Digestive and Kidney Diseases, US National Institutes of Health (Dr Neovius [grant R01DK105948]), the Swedish Research Council (Dr Neovius [grant 2013-3770] and Dr Stephansson [grant 2013-2429]), and the Swedish Research Council for Health, Working Life, and Welfare (Dr Johansson [grant 2017-00321]). Drs Pasternak, Johansson, and Stephansson are supported by the Strategic Research Area Epidemiology program at Karolinska Institute. Dr Pasternak is also supported by the Swedish Research Council.
Role of the Funder/Sponsor: The study funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Additional Contributions: We thank health care personnel for prospective data entry into the Medical Birth Register and the Scandinavian Obesity Surgery Registry.
Create a personal account or sign in to: