A new CDC analysis finds prenatal drinking climbed from roughly 13.5% to 15% between the 2018 2020 and 2021 2024 windows. The patients behind that line are excluded by default from the trials that would produce better guidance.
The CDC published a number on Thursday that should embarrass the field it describes. Self-reported past-30-day drinking among pregnant adults rose from roughly 13.5% in 2018-2020 to about 15% in 2021-2024, a jump of around 1.5 percentage points reported by STAT News, citing the new CDC analysis. Read narrowly, that is a small move on a survey line. Read honestly, it is a signal that a known teratogen has more exposed pregnancies than it did four years ago, and the medical infrastructure that could respond is not built to.
Ethanol is a well-characterized cause of fetal harm, and a leading driver of preventable intellectual disability in the United States. The CDC's own framing of the new data, as carried in STAT's reporting, is that alcohol consumption during pregnancy "remains a public health concern" and that both clinical and community interventions "might help." That is the careful register of a federal agency that knows what its surveillance system can and cannot say. The trend is real. The mechanisms behind it are not something the trend chart can fix.
The CDC's window, 2021-2024, is also the period in which the structural lockout of pregnant patients from the evidence base is most visible. Pregnant women are excluded by default from most clinical trials of medications, including the ones used to treat alcohol use disorder. That default means the safety, dosing, and efficacy data that providers would normally lean on either do not exist for this population or are borrowed from non-pregnant cohorts and applied off-label. The downstream effect is documented in the literature and in clinical practice: providers are hesitant to refer pregnant patients to substance use disorder treatment, and hesitant to prescribe the medications that could plausibly reduce drinking in a patient who is drinking. The same system that records the rise is the system that declines to generate the evidence or deliver the treatment that would push it back down.
A second fact in the CDC data sharpens where the gap is doing the most damage. The subgroup most affected is not a moral story. Unmarried pregnant adults and those reporting frequent mental distress reported binge and heavy drinking rates more than double the population average, according to the STAT summary of the CDC figures. That is a population with less access to consistent prenatal care, more contact with the parts of the system that screen poorly for alcohol use, and more reason to be using alcohol in the first place. They are also the population least likely to be enrolled in the kind of trial that would clarify what to do for them.
The cultural thread matters here, and it cuts against the trend. Some clinicians have, in recent years, adopted a more relaxed posture toward alcohol in pregnancy, treating occasional drinking as low-risk and declining to push screening or referral. That posture is itself a clinical choice, and it is the wrong one given what is known about ethanol and fetal development. The CDC's data is a quiet rebuke to that drift. A surveillance number does not change provider behavior. A professional environment that treats prenatal alcohol exposure as a routine counseling topic, and that has the evidence base to back up that conversation, can.
The watch items for the next cycle are concrete. The CDC will publish the underlying instrument and subgroup breakdowns. The relevant question is whether BRFSS, PRAMS, or another survey produced the line, and whether the post-2020 shift is robust to the pandemic-era changes in survey mode and response. If the trend holds in the next data release, the policy question shifts from whether to act to who pays for the trial infrastructure that would let clinicians act with evidence rather than instinct. The 1.5 percentage point move is the door. The room behind it is a known teratogen, a locked trial pipeline, a hesitant referral system, and a subgroup with the most need and the least access. The CDC counted the door. The clinical research system has not built the room.