Epidurals Don't Raise Risk of Brain Injury or Cerebral Palsy: 500,000-Birth BMJ Study
If you have worried that pain relief in labor could harm your baby's brain, a large new study offers a clearer answer. Researchers followed hundreds of thousands of hospital births and found that labor epidurals did not raise the chance of serious newborn brain injury, very low scores on the first newborn exam, or childhood cerebral palsy. That finding matters for anyone making a pain-relief choice under pressure from alarming videos and posts. The BMJ analysis, led by Dr. Rachel Kearns at the University of Glasgow, reviewed 495,695 singleton hospital births in Scotland's National Health Service. After researchers accounted for differences between groups, rates of newborn infection and death in the first 28 days were comparable for babies whose mothers received an epidural and those whose mothers did not.
Based on reporting from MedPage Today / The BMJ: Study Offers Reassurance to Women Considering Epidural During Labor.
If frightening online claims have you second-guessing pain relief, start here
More than 70% of women in United States hospital labor wards choose an epidural for pain relief during childbirth. Yet many expectant parents leave a late-night scroll convinced that the same medicine will starve a baby of oxygen, damage developing brain cells, or cause autism or cerebral palsy.
That fear is common, and it is understandable. Worrying about your child's brain is a protective instinct, not a character flaw. Wanting effective pain relief is not a failure of birth planning.
A nationwide Scottish hospital study published in The BMJ tested those claims against nearly half a million real births. The short version for families: labor epidurals were not linked with more serious newborn brain injury, very low first-exam scores, or childhood cerebral palsy. The longer version, with study design, numbers, and counseling detail, follows below.
What the researchers asked, in plain language
Dr. Rachel Kearns, a consultant in obstetric anaesthesia at Glasgow Royal Infirmary and a clinical researcher at the University of Glasgow, led a population-based cohort study (a review of health records from an entire defined group, not a small volunteer sample). The paper appears as BMJ 2026;394:e343320. MedPage Today summarized the same findings for clinicians counseling people in labor.
Investigators analyzed 495,695 singleton hospital births (one baby, not twins or more) across the Scottish National Health Service from 2007 to 2019. Gestation ranged from 24 weeks plus 0 days to 42 weeks plus 6 days. They included vaginal births and unplanned cesarean births.
Of that group, 114,897 women (23.2%) received labor epidural analgesia (pain medicine placed through a thin catheter near the spinal nerves). Teams then followed infants from the delivery room into childhood, looking for serious neurological harm, low vitality scores, infection, death in the first 28 days, and later cerebral palsy.
How they compared groups that were not alike at the start
People who get epidurals are not a random sample of everyone in labor. In the raw records, women who received epidurals were more often first-time mothers (nulliparous), had a higher body mass index, had longer or induced labors, and more often received oxytocin to strengthen contractions.
Those same factors can also change newborn outcomes. So a simple head-to-head count would mix the effect of the epidural with the effect of a harder labor. Researchers used multivariable regression (a statistical method that holds those other differences as steady as the data allow) before they compared groups.
After that adjustment, the primary analysis produced an adjusted relative risk (aRR, a comparison of how often the outcome occurred in the epidural group versus the no-epidural group, after accounting for measured differences) of 0.87 for serious adverse neonatal neurological outcomes. The 95% confidence interval (the range most compatible with the data) ran from 0.68 to 1.12. Because that range includes 1.0, the result is compatible with no statistical difference. The point estimate sits slightly below 1.0, so the data do not show added harm.
What the numbers show for babies
Serious newborn neurological problems were uncommon in the whole cohort: 434 babies, or about 0.9 per 1,000 births. The primary outcome was a composite that included hypoxic-ischemic encephalopathy (brain injury related to low oxygen around birth) and other early brain-related diagnoses coded within 28 days.
Secondary outcomes stayed comparable after adjustment:
- Childhood cerebral palsy: multi-year tracking found no increase among children whose mothers had an epidural (aRR 0.80, 95% CI 0.60 to 1.06).
- 5-minute Apgar score under 4: this quick test of breathing, heart rate, muscle tone, reflexes, and color showed no added risk (aRR 0.97, 95% CI 0.87 to 1.09).
- Neonatal sepsis: bloodstream infection in the newborn was comparable (aRR 1.11, 95% CI 0.90 to 1.37).
- Death in the first 28 days: rates were not higher after an epidural (aRR 0.81, 95% CI 0.62 to 1.06).
- Other severe newborn illness: a broader bundle of complications also showed no statistically significant increase (aRR 1.17, 95% CI 0.90 to 1.51).
These patterns held in subgroup analyses of vaginal and unplanned cesarean birth, and across higher-risk and lower-risk pregnancies. That is not a promise that every labor is identical. It is strong population-level evidence that the epidural itself was not the driver of these rare harms.
The Science Behind ItWhy epidural medicine stays local
Many people picture an epidural as a large dose of medicine poured into the mother's bloodstream, then shared with the baby through the placenta (the organ that exchanges oxygen and nutrients). That picture does not match how a modern labor epidural works.
An anesthesiologist or nurse anesthetist places dilute local anesthetics (numbing medicines such as bupivacaine 0.0625% to 0.125%, or ropivacaine 0.1% to 0.2%) plus micro-doses of a lipid-loving opioid (for example fentanyl 2 mcg/mL) into the lumbar epidural space. That space sits outside the dural sac, the covering that holds spinal fluid around the cord.
Think of the catheter as a local shutoff valve, not a whole-building flood. The medicine bathes nearby spinal nerve roots and dorsal root ganglia (the junctions where pain signals enter the cord) and blocks those signals at the source. Only a small amount enters maternal blood vessels. Transfer across the placenta stays in trace amounts, far below levels associated with fetal brain depression or nerve toxicity in clinical use.
Mothers and babies both belong in the risk conversation
This newborn-safety paper builds on earlier work from Dr. Rachel Kearns and colleagues. In a previous Scottish population study of 567,216 births, labor epidurals were associated with a 35% reduction in severe maternal morbidity (SMM: life-threatening complications around birth, using a standard U.S. Centers for Disease Control and Prevention definition). The reduction was larger among women with a medical indication for an epidural (about 50%) and among those giving birth preterm (about 47%).
Severe untreated labor pain can drive surges in catecholamines (stress hormones such as adrenaline), strain the heart and breathing, and complicate an unplanned cesarean if the team later needs a rapid general anesthetic. A working epidural can blunt that stress response and, if surgery becomes necessary, can often be dosed for an operative anesthetic without putting the mother to sleep.
When clinicians counsel families, the honest picture is two-sided: this newer study does not show added serious neurological harm for the baby, and the earlier study associates epidurals with fewer severe complications for the mother. Neither paper says an epidural is mandatory. Both support treating the choice as medical care, not a moral test.
Clinical Insight for ProvidersFever versus infection, and how to triage
Epidural-related maternal fever (ERMF) remains a frequent labor-and-delivery dilemma. It is a non-infectious inflammatory temperature rise seen in a subset of people, especially during long inductions. Because infectious chorioamnionitis (infection of the membranes and amniotic fluid) can look similar in real time, pediatric teams have historically responded with separation, cultures, and empiric antibiotics.
The BMJ cohort cannot settle every fever case. Administrative Scottish records did not capture continuous maternal temperature, blood pressure dips, or fetal heart-rate tracings, so investigators could not directly test ERMF against newborn outcomes inside this dataset. What they can say is more limited and still useful: after adjustment, neonatal sepsis, the neurological composite, low 5-minute Apgar scores, and 28-day mortality were not higher in the epidural-exposed group.
Use that distinction in counseling and triage. A low-grade temperature during a long epidural labor is not, by itself, proof of neonatal sepsis. Look for supporting infection signs before reflexively separating the dyad or starting a full septic workup: foul-smelling amniotic fluid, persistent maternal tachycardia, uterine fundal tenderness, or other markers of true chorioamnionitis.
Counsel with the actual interval, not a slogan. For the primary neurological composite, aRR 0.87 (95% CI 0.68 to 1.12) excludes a relative increase larger than 12% at the upper bound. Secondary point estimates for sepsis and other severe morbidity sit slightly above 1.0 with intervals that also include no difference. Name residual confounding: people who request epidurals often have longer, more complex labors, which would be expected to bias toward harm in the exposed group if anything. The null result is therefore more persuasive, not less.
Limitations to say out loud: this is observational administrative data, not a randomized trial. Socioeconomic status used area-level Scottish Index of Multiple Deprivation scores rather than household income. Maternal ethnicity was missing for a large share of older records (about 37.5%) and required imputation. The file also lacks indication for the epidural, total labor duration, and timing of placement. Those gaps argue for careful language, not for discarding the strongest population signal currently available.
In Arizona hospitals, a midwife does not leave when you choose an epidural
A common misconception is that choosing a Certified Nurse-Midwife means rejecting pain relief, or that choosing an epidural means your midwife will step away. In Arizona hospitals across Phoenix, Chandler, Gilbert, and Peoria, hospital-based CNMs practice on collaborative teams where comfort and safety stay in the same plan.
When you request an epidural, the midwife stays in the room. The anesthesia team places the catheter. The midwife and labor nurse then keep labor biomechanics active: a peanut ball between the legs, side-lying turns every 30 to 60 minutes, upright Fowler positioning, and laboring down (waiting for the baby to descend before hard pushing). Those steps keep pelvic diameters open, support rotation of the baby's head, and can reduce prolonged pushing, vacuum or forceps births, and severe perineal tears.
American College of Obstetricians and Gynecologists (ACOG), American Society of Anesthesiologists (ASA), and American College of Nurse-Midwives (ACNM) guidance agrees that maternal request alone justifies labor analgesia. There is no requirement to reach an arbitrary cervical dilation number first.
Name it, normalize it, then offer a real plan
If alarming videos left you anxious or guilty about considering an epidural, say that out loud at a prenatal visit. Clinical teams can use a three-step frame without talking down to you.
- Name it: the exact fear is that pain relief will cause brain damage, autism, developmental delay, or cerebral palsy, or that asking for it means you failed.
- Normalize it: that fear is a common response to online claims, not evidence that something is already wrong with you or your baby.
- Guide it: share that this half-million-birth analysis did not find added serious neurological harm, that maternal request is enough under ACOG, ASA, and ACNM guidance, and that your birth team can support both pain relief and continuous bedside care.
Questions worth asking before labor starts
Bring these operational questions to a prenatal visit so the plan is specific to your hospital, not to a generic internet thread.
- Anesthesia coverage: does the hospital keep dedicated obstetric anesthesia in house around the clock, so placement is available at night and on weekends?
- Flexible comfort: can you start with a tub, nitrous oxide, or walking, then request an epidural if you want one later?
- Movement after placement: how do midwives and nurses use peanut balls and position changes once the epidural is in?
- The hour after birth: do protocols protect delayed cord clamping and uninterrupted skin-to-skin, with or without pain medicine?
Questions this headline raises
This BMJ study of 495,695 hospital births did not find a higher risk of childhood cerebral palsy or other serious newborn neurological harm after a labor epidural. After researchers accounted for differences between groups, the main comparison (adjusted relative risk 0.87, 95% CI 0.68 to 1.12) was compatible with no statistical difference. Babies whose mothers had epidurals had comparable neurological outcomes to babies born without them.