American Infant Mortality Just Hit a Record Low. The International Comparison Is Half an Accounting Trick. The Mississippi Gap Is Not.

In 1900, more than one in seven American babies did not live to see a first birthday. The line on that chart has bent down for a hundred and twenty-five years through clean water, antibiotics, neonatal intensive care, surfactant for preterm lungs, vaccines, and safe-sleep campaigns. Preliminary 2025 numbers from the CDC, released this week, put the U.S. infant mortality rate at 5.4 deaths per 1,000 live births, the lowest ever recorded. Roughly 19,350 American babies died before their first birthday last year, down from 5.5 per 1,000 in 2024 and 5.6 in 2022 and 2023. Progress.

Then the headlines arrived: “still trails other wealthy nations.” That frame has run on every U.S. infant-mortality story for three decades. Some of it is true, some of it is an accounting artifact, and the part that is true has almost nothing to do with the things the framing usually blames.

Start with the comparison itself. The United States defines a live birth as any birth in which the baby shows any sign of life: a breath, a heartbeat, a flicker of voluntary muscle, regardless of weight or gestational age. That is among the broadest definitions in the developed world. Several European countries apply stricter registration thresholds, counting only births that meet minimum weight or gestational-age criteria, so the most fragile infants, the ones least likely to survive a week, are not entered in the numerator. A peer-reviewed analysis archived at PMC found that variation in how clinicians and registrars classify borderline deliveries materially shifts the rate. When the OECD ranks countries, it is comparing rosters built to different rules, and the U.S. counts more of the hardest cases. That does not close the entire gap with Japan or Sweden, but it is not honest to pretend it does not exist.

Now the domestic picture, which the international frame conveniently buries. Mississippi declared a public health emergency last August after its 2024 infant mortality rate climbed to 9.7 per 1,000, the highest in more than a decade and nearly double the national average. New Hampshire’s rate is around three. That gap is not about accounting. It is about prenatal care, maternal health, poverty, smoking rates, and whether a hospital with a neonatal intensive care unit is reachable when something goes wrong. Black infants in the United States die at 10.9 per 1,000, more than twice the rate of white, Hispanic, and Asian American infants, and the disparity has barely moved in two decades. Roughly two-thirds of U.S. infant deaths happen in babies born preterm. The American infant-mortality problem, to the extent there is one beyond the rich-country baseline, is a preterm-birth problem, and the preterm-birth problem is a maternal-health problem.

That makes the explanations offered for the new record worth scrutinizing. The two factors credited in the press-release version of this story are newly available RSV-prevention products, the nirsevimab monoclonal antibody for infants and the maternal RSV vaccine given at 32 to 36 weeks, and renewed safe-sleep education. The RSV tools work. They are also new and unevenly distributed: nirsevimab carries a U.S. list price around $520 per dose and runs through commercial insurance and the Vaccines for Children program, with rollout still uneven across the counties whose infants are most likely to die. Safe-sleep messaging costs almost nothing and reaches every parent with a pediatrician. The plausible reading is that pharma got the press credit while the cheaper, broader public-health work probably did more of the lifting, and the underlying driver, preterm birth, has not budged.

There is a policy version of this story the establishment outlets will not write. Mississippi and the Deep South lead the country in infant deaths because they have the thinnest prenatal-care networks, the highest uninsured rates among reproductive-age women, and the most pregnancy-related Medicaid churn. New Hampshire has the best rate because it has the opposite. The lever that would close more of this gap is not another monoclonal antibody priced for the commercial insurance market. It is sustained prenatal care for every American woman who wants a baby, delivered through public-health systems that already exist and have been chronically underfunded relative to the latest pharmaceutical launch.

The trend is good. Watch what happens when the preliminary 2025 file is finalized next year, and watch whether the racial and state gaps narrow alongside the headline rate or whether the best states keep pulling further away while Mississippi sits where it has been since the 1990s. The shape of the next two years of data will tell you whether 5.4 is a public-health victory or a wealthier-and-whiter-states victory wearing one.

Sources

1. STAT – U.S. infant mortality rate fell to an all-time low (June 16, 2026)
2. MedPage Today – U.S. Infant Mortality Rate Falls to All-Time Low
3. The Hill – Mississippi declares public health emergency over infant mortality rate
4. HHS Office of Minority Health – Infant Mortality and Black/African Americans
5. CDC Grand Rounds: Public Health Approaches to Reducing U.S. Infant Mortality (PMC)
6. PMC – Variation in classification of live birth with newborn period death
7. Center for Community Solutions – Differences in reporting of live births and infant mortality comparability
8. OECD – Infant mortality rates indicator
9. Contemporary Pediatrics – Nirsevimab to be 5% more expensive