A groundbreaking 2026 study published in Molecular Psychiatry has uncovered a significant association between maternal use of common medications during pregnancy and increased rates of autism spectrum disorder (ASD) in offspring. Led by researchers at the University of Nebraska Medical Center (UNMC), the analysis of over 6.1 million U.S. births from the Epic Cosmos database found that exposure to any of 15 sterol biosynthesis inhibiting medications (SBIMs) was linked to a 1.47-fold (47%) higher risk of ASD diagnosis after adjusting for confounders. Risk escalated in a dose-dependent manner: each additional SBIM raised the hazard ratio by 1.33 times, reaching 2.33-fold when four or more were prescribed concurrently. Cariprazine showed the highest individual risk at 2.59-fold.

The medications—sertraline, fluoxetine, bupropion, buspirone, aripiprazole, cariprazine, haloperidol, trazodone, metoprolol, propranolol, nebivolol, atorvastatin, simvastatin, rosuvastatin, and pravastatin—collectively accounted for roughly 400 million prescriptions annually. Usage among pregnant women in the dataset tripled from 4.6% in 2014 to 16.8% in 2023. Of the 3.8% of children diagnosed with ASD (approximately 235,000 cases), 15% had mothers prescribed at least one of these drugs. The association tracked closely with each drug's documented ability to elevate 7-dehydrocholesterol (7-DHC), a potentially toxic byproduct of disrupted cholesterol synthesis.

This finding illuminates a critical but under-discussed mechanism: cholesterol is indispensable for fetal brain development, forming cell membranes, myelin, and signaling molecules like neurosteroids. A fetus does not synthesize its own cholesterol until roughly mid-gestation, relying entirely on maternal supply in the first half of pregnancy. Disruptions here—whether genetic (as in Smith-Lemli-Opitz syndrome, which features high autism overlap) or pharmacological—can cascade into neurodevelopmental anomalies. The study's polypharmacy data and lab correlations strengthen the case for causality via the sterol pathway, a pattern mainstream coverage has been slow to emphasize amid rising ASD prevalence.

The research also indirectly challenges decades of aggressive cholesterol-lowering dogma. Statins carried pregnancy warnings until their 2021 removal, despite cholesterol comprising 20-25% of the brain's dry weight. As prescriptions of antidepressants, antipsychotics, and beta-blockers have soared, so too have neurodevelopmental disorder rates—an iatrogenic factor often sidelined in favor of purely genetic explanations that absolve pharmaceutical accountability. Earlier lab studies on 7-DHC toxicity and animal models of sterol inhibition align with these epidemiological patterns, suggesting a hidden driver in the autism epidemic that demands urgent reevaluation of prescribing practices during pregnancy. While observational, the scale, specificity, biological gradient, and mechanistic coherence warrant caution and further randomized or mechanistic follow-up. Clinicians are advised to weigh these risks carefully, especially given the sheer volume of affected pregnancies.

This heterodox lens reveals connections long missed: the lipid hypothesis's flaws, the brain's unique dependence on cholesterol, and how modern polypharmacy may be chemically recreating rare genetic disorders at population scale.