The list of things pregnant women have been told to worry about isn’t getting any shorter.. Coffee, cold cuts, hot tubs, hair dye, stress, not enough stress, too much fish, not enough fish. To this catalog of anxieties, recent years have added a new entry: the humble antacid.
Studies had begun to suggest that the medications millions of women take to manage pregnancy heartburn might be linked to ADHD, intellectual disability, and other neurodevelopmental conditions in their children. While the findings were preliminary and the effect sizes small, the headlines were not.
Now, a large study from South Korea offers a different conclusion and, perhaps more importantly, a lesson in how to read the research that shapes so much of what we think we know about prenatal risk.
The study, published in JAMA on January 7, 2026, followed more than 2.7 million mother-child pairs and found no evidence that acid-suppressive medications increase the risk of neuropsychiatric disorders in children, even after the researchers accounted for factors that earlier studies had missed. The key was a methodological choice that is quietly reshaping how scientists study prenatal exposures: comparing siblings within the same family.
The Initial Signal
Gastroesophageal reflux is among the most common complaints of pregnancy. About two-thirds of expectant mothers experience heartburn at some point, and it’s often driven by hormonal changes that relax the lower esophageal sphincter. When lifestyle modifications and over-the-counter antacids prove insufficient, physicians often prescribe acid-suppressive medications: histamine H2 receptor antagonists like famotidine (Pepcid) or proton pump inhibitors like omeprazole (Prilosec) and esomeprazole (Nexium). Studies estimate that roughly a quarter of pregnant women in developed countries take some form of acid-suppressive medication.
Over the past decade, observational studies began reporting modest associations between these medications and childhood neuropsychiatric conditions. The effect sizes were small, typically in the range of 10 to 20 percent increased risk; but in samples of millions, they reached statistical significance. The findings generated headlines, clinical uncertainty, and a familiar kind of parental anxiety.
The new research, led by Seohyun Hong and colleagues at Kyung Hee University College of Medicine in Seoul, set out to test whether those associations would hold up under more rigorous scrutiny. Drawing on the South Korean National Health Insurance Service database, the researchers identified 2,777,119 mother-child pairs with births between January 2010 and December 2017, including 507,845 pairs in which the mother had been prescribed at least one acid-suppressive medication during pregnancy. Children were followed through December 2023, a mean of 10.3 years.
The Sibling Study
The researchers ran two parallel analyses, and the divergence between them tells the story.
In the first approach, using propensity-score overlap weighting to adjust for observable differences between exposed and unexposed groups, the researchers found what earlier studies had suggested. Children whose mothers took acid-suppressive medications showed modestly elevated rates of ADHD, intellectual disability, severe neuropsychiatric disorders, autism spectrum disorders, and obsessive-compulsive disorder. In absolute terms, the differences were small: 4.85 percent of exposed children were diagnosed with ADHD, compared to 4.25 percent of unexposed children.
Then the researchers tried something different. They identified families where the mother had taken acid-suppressive medication during one pregnancy but not another, and compared outcomes between the siblings. This design, known as sibling-control analysis, has a crucial advantage: Siblings share roughly half their genes and all of their family environment, including factors like parental health behaviors, socioeconomic status, and household exposures that are difficult or impossible to measure directly. If the medication itself were causing the increased risk, the effect should persist even when comparing siblings. If the apparent risk were actually due to unmeasured family characteristics, it should disappear.
It disappeared.
The sibling analysis included 157,069 exposed children and 164,669 unexposed siblings. The hazard ratio for ADHD dropped to 0.98, statistically indistinguishable from no effect. For intellectual disability, it was 1.02. For severe neuropsychiatric disorders, 1.00. For autism spectrum disorders, 0.98. For obsessive-compulsive disorder, 0.95. Every association that had appeared significant in the standard analysis was nullified.
“Exposure to acid-suppressive medication during pregnancy was not associated with children’s risk of ADHD, severe neuropsychiatric disorder, obsessive-compulsive disorder, intellectual disability, or ASD in sibling-control analyses,” the authors conclude. “Small associations were observed in overlap-weighted models; these may reflect confounding by shared familial factors.”
The Confounding Problem
In an accompanying JAMA Guide to Statistics and Methods article, Viktor H. Ahlqvist, Brian K. Lee, and Yu-Han Chiu explain why this matters. The challenge in observational research is that people who take a medication may differ from people who don’t in ways that also affect the outcome of interest. A mother who takes heartburn medication during pregnancy may have different dietary habits, different stress levels, different genetic predispositions, or different underlying health conditions than a mother who doesn’t. Any of these factors could independently affect her child’s neurodevelopment, and if researchers can’t measure and adjust for them, the medication will appear to cause effects that are actually attributable to something else.
Sibling comparisons address this problem by design. When you compare children born to the same mother, you automatically control for everything about that mother that stays constant across pregnancies: her genes, her long-term health conditions, her socioeconomic circumstances, her dietary patterns, her environment. What remains is the difference in medication exposure between pregnancies. In this study, that difference predicted nothing.
A Broader Pattern
The South Korean findings are not an isolated case. The same pattern (apparent associations in standard analyses, null results in sibling comparisons) has emerged repeatedly in recent research on prenatal exposures and childhood neuropsychiatric conditions.
The most prominent example involves acetaminophen (Tylenol/paracetamol), which generated considerable public controversy after claims by the U.S. government in September 2025 that prenatal exposure might contribute to neurodevelopmental disorders. In April 2024, a Swedish study published in JAMA examined 2,480,797 children and found that while standard analyses showed small associations between prenatal acetaminophen exposure and ADHD (HR 1.07), autism (HR 1.05), and intellectual disability (HR 1.05), sibling-control analyses showed no such links. The hazard ratios dropped to 0.98 for both ADHD and autism, and to 1.01 for intellectual disability.
A comprehensive meta-analysis published in The Lancet Obstetrics, Gynaecology, & Women’s Health on January 16, 2026, reinforced the point. Examining 43 studies on acetaminophen and neuropsychiatric outcomes, with 17 included in the primary meta-analysis, the researchers found that when analyses were restricted to sibling-comparison studies, there was no association with ADHD, autism, or intellectual disability. “Current evidence does not indicate a clinically important increase in the likelihood of autism spectrum disorder, ADHD, or intellectual disability in children of pregnant individuals who use paracetamol as directed,” the authors concluded.
Implications for Practice
For clinicians, the immediate takeaway is straightforward: the evidence does not support advising pregnant patients to avoid acid-suppressive medications out of concern for neurodevelopmental outcomes. Most proton pump inhibitors remain classified as FDA pregnancy category B, meaning animal studies have shown no risk. And the human data, properly analyzed, now point in the same direction.
For behavioral health providers who work with families of children with ADHD, intellectual disabilities, or other neuropsychiatric conditions, the research offers an opportunity to address one source of parental anxiety. Mothers who took heartburn medication during pregnancy need not wonder whether their choice contributed to their child’s diagnosis. The evidence, when properly analyzed, suggests it did not.
But the broader lesson may be more important than any single finding. Much of what clinicians and families have been told about prenatal risk comes from observational studies that are vulnerable to the same confounding issues exposed here. The pattern emerging from this research suggests that many proposed risk factors may, upon closer examination, prove innocent.
This is not a license to dismiss all safety concerns. Some medications are genuinely teratogenic, and the principle of minimizing unnecessary medication during pregnancy remains sound. But it is a reminder that correlation is not causation, and that observational studies showing modest associations should be interpreted with particular caution when they have not been tested against sibling-control designs.
For the families navigating diagnoses of ADHD, intellectual disability, or other neuropsychiatric conditions, that distinction matters. The search for causes has generated a long list of proposed culprits, from vaccines to prenatal ultrasounds to over-the-counter medications, most of which have failed to hold up under rigorous scrutiny. The new research significantly weakens the case for concerns about heartburn medication, and in doing so, lifts one small burden from parents who already carry enough.
