General Anesthesia

Pregnancy

Fisher et al. used National Birth Defects Prevention Study data for pregnancies ending in 1997-2011, in a population-based case-control study with 24,337 birth-defect cases and 11,501 controls. For periconceptional general anesthesia, 226 case and 73 control women reported exposure. The study assessed 25 birth defects with at least five exposed cases and found no significant associations; odds ratios were generally close to null and imprecise. Exposure was rare, so the study was underpowered for moderate effects, and self-report, missing procedure and drug details, and confounding by the reason for surgery limit causal interpretation. Yu et al. (2018) analyzed Taiwan insurance claims from 1997-2013. After exclusions, 114,852 delivery records and 3,999 abortion records remained; 462 pregnancies involved nonobstetric surgery under general or regional anesthesia. Compared with no surgery, the study reported spontaneous abortion in 4.23% versus 2.43% (adjusted odds ratio 1.53, 95% CI 1.01-2.31), preterm birth in 9.06% versus 4.90% (adjusted odds ratio 3.31, 95% CI 2.54-4.31), and preeclampsia/eclampsia in 2.60% versus 1.01% (adjusted odds ratio 2.35, 95% CI 1.30-4.23). Those comparisons combine surgery, illness, and general or regional anesthesia. In the surgery group, general anesthesia was not significantly associated with premature labor compared with regional anesthesia (adjusted odds ratio 1.78, 95% CI 0.89-3.57). The claims database lacked fetal details and some key confounders, including BMI, smoking, alcohol, and condition severity. Haataja et al. (2023) reviewed 60 mostly retrospective studies of nonobstetric surgery, including 80,205 pregnancies with surgery and 16,655,486 without surgery. The review reported higher odds of stillbirth and preterm birth after surgery, but no overall increase in miscarriage, low five-minute Apgar scores, or congenital anomalies. These results concern surgery as a whole and cannot be assigned to anesthetic drugs alone. The included studies were heterogeneous, and fewer than half had a no-surgery comparison group. ACOG's 2016 advisory quotes the FDA's warning that repeated or lengthy exposure to general anesthetic or sedation drugs in the third trimester may affect a child's developing brain. The advisory says the warning cited animal and pediatric studies but no data in pregnant women, and its clinical significance is unknown; it describes pregnant-primate studies involving 24-hour ketamine or five-hour isoflurane plus nitrous oxide or propofol exposures. ACOG Committee Opinion 775 states that currently used anesthetics have not been shown teratogenic in humans at standard concentrations, that no human in-utero fetal-brain effect has been shown, and that animal data do not support an effect with exposures under three hours; it says medically necessary surgery should not be denied or delayed and elective surgery should be postponed until after delivery. Human primary neurodevelopment cohorts are mixed. Ing et al. (2021) assessed 2,024 children, of whom 22 were prenatally exposed to general anesthesia; after correction for multiple outcomes, only one of 11 scores, externalizing behavior, remained significantly different, and other scores did not. Bleeser et al. (2023) compared 129 exposed children with 453 unexposed children; 90% of exposures were a single session with a mean duration of 91 minutes, and there were no significant overall differences in primary or secondary outcomes, although exploratory subgroups showed differences. Courteille et al. (2025) compared 40 children exposed to general anesthesia, 13 exposed to regional anesthesia, and 41 controls; the 94-child pilot found no difference in parent-rated executive-function scores. Ing et al. (2024) matched 34,271 children exposed to general anesthesia during maternal appendectomy or cholecystectomy with 171,355 controls and reported higher behavioral-disorder diagnoses (HR 1.31, 95% CI 1.23-1.40). The claims-based observational study cannot separate anesthesia from surgery or underlying illness. These human studies do not directly test the repeated or lengthy third-trimester exposures in the FDA warning and do not establish causation.

Breastfeeding

MotherToBaby says anesthetic medicines are generally processed quickly and that evidence after procedures is limited. ASA's statement, last amended 2024-10-23, says anesthetic and analgesic medicines transfer into milk in low concentrations and recommends resuming breastfeeding as soon as desired after surgery rather than discarding milk. ASA says codeine and tramadol are not recommended during breastfeeding and calls for monitoring for infant sedation with other opioid pain medicines, especially for newborns and infants at risk for apnea. ABM Protocol No. 15 (revised 2017) says healthy term or older infants can generally feed when the parent is awake, stable, and alert; it notes that infants at risk for apnea, hypotension, or hypotonia may benefit from a brief interruption. The Association of Anaesthetists' 2026 multidisciplinary guideline used a targeted literature review and three-round modified Delphi process; it says perioperative medicines pass into milk in very small amounts, that evidence of adverse effects is absent for almost all perioperative medicines, and that routine milk disposal is not needed. Hodge et al. (2026) studied 10 term patients having scheduled caesarean delivery under general anesthesia. Sugammadex was detected in 12.5% of milk samples taken within six hours, but all detected samples were below the limit of quantification; the study did not assess infant clinical outcomes and applies to sugammadex, not every anesthetic drug. ASA, ABM, and the Association guideline support resuming breastfeeding after recovery, but their guidance does not establish one regimen for every anesthetic and postoperative medicine. Kelly et al. (2013) followed 238 breastfeeding mothers taking codeine after caesarean delivery; neonatal sedation was reported for 5 infants (2.1%), and those mothers had used codeine longer on average, beyond the study's four-day guidance. This codeine-specific result does not estimate infant effects after a single anesthesia exposure or other opioids. Zipursky et al. (2023) matched 85,675 mothers who filled an opioid prescription within seven days after delivery with 85,675 controls; infant hospital readmission was similar (3.5% in each group; HR 0.98, 95% CI 0.93-1.03), and no infant deaths occurred. The claims did not confirm breastfeeding status and could not rule out infant sedation that did not lead to medical care.

What was tested

Fisher et al. studied specific birth defects after periconceptional general or local anesthesia for surgery. Yu et al. compared outcomes after nonobstetric surgery under general or regional anesthesia with outcomes after no surgery, so that study cannot isolate the effect of general anesthesia. Haataja et al. reviewed pregnancy outcomes after nonobstetric surgery. For breastfeeding, Hodge et al. measured sugammadex in plasma, urine, and breastmilk after caesarean delivery under general anesthesia; the study did not measure infant outcomes. Kelly et al. prospectively recorded neonatal sedation among breastfeeding mothers taking codeine, and Zipursky et al. studied claims-based infant outcomes after maternal postpartum opioid prescriptions without confirming breastfeeding status.