A post moving through Telegram groups and X this month makes a specific, checkable claim about a real emergency. Nicolas Hulscher, an epidemiologist at the McCullough Foundation, wrote that "CDC data shows babies born are now dying at a >50% excess rate — YEARS after mass mRNA vaccination of young women," and that "Mississippi declared a PUBLIC HEALTH EMERGENCY over soaring infant deaths, now linked to transgenerational mRNA 'vaccine' fatal adverse events." A common response to posts like this points at the state's abortion ban instead. A third possibility — COVID-19 infection itself — gets raised least often and has the largest documented effect of the three. The emergency is real and Mississippi's numbers are genuinely alarming: 323 babies died before their first birthday in 2024, a rate of 9.7 per 1,000 live births, the state's worst in a decade and nearly double the national rate. We tested all three against Mississippi's own vital records, using the same standard for each, and we also say plainly at the end what none of this can rule out.
What Mississippi actually declared, and when
Dr. Daniel Edney, Mississippi's State Health Officer, declared the public health emergency on August 21, 2025 — roughly thirteen months before this post began circulating, which makes the word "just" in the post inaccurate. The declaration names four leading causes of infant death: "congenital malformations, preterm birth, low birth weight and Sudden Infant Death Syndrome (SIDS)." It proposes an obstetric system of care, prenatal services in counties with no obstetric provider, community health workers, hospital partnerships, and safe-sleep education. Vaccines are not mentioned in it, and neither is the state's abortion law. Both attributions are additions made by other people.
This is a state health department explaining an emergency that same department declared, so its account of causes is not an outside audit. That is why what follows leans on the underlying vital-records counts — births and deaths, tabulated by year and race — rather than on the department's framing of them.
The national number the claim rests on
The post says babies nationally are dying at a greater-than-50% excess rate, so the national series is the direct check. According to the National Center for Health Statistics, the US infant mortality rate was 5.52 deaths per 1,000 live births in 2024, against 5.61 in 2023 — a difference the agency describes as not statistically significant. Across the decade the rate ran 5.9 in 2015, 5.6 in 2019, 5.4 in 2020 and 2021, 5.6 in 2022 and 2023, and 5.5 in 2024. On the NCHS final files, 2021 — the year of mass mRNA vaccination, the rollout having begun on December 14, 2020 — came in at 5.44, the lowest figure in the modern series, and the rise from 2021 to 2022 was the first annual increase since 2002. The rate then came back down, and 2024 sits below 2019. A greater-than-50% excess does not exist anywhere in that series.
Where the ">50%" comes from
We looked for the analysis behind the figure. The ">50%" appears in Hulscher's own post; we could not locate a published methodology document for that specific number, and we say so rather than assume one exists. The nearest documented analysis is one Hulscher wrote up on the Substack publication The Focal Points, crediting an anonymous analyst who writes as "The Ethical Skeptic." That analysis claims a 77% increase above baseline in all-cause deaths for ages 0–4 — a different figure, a different age range — using CDC death certificate data from 1995 to 2023 and a technique the analyst calls "Deviation from Trend": pick a historical baseline, project it forward, treat the gap between projection and reality as excess. It has not been peer reviewed and by its own account establishes timing rather than causation. Everything in such a result depends on which baseline is chosen and how far it is projected, and the method itself does not distinguish the cause an analyst has in mind from anything else that changed in the same years. That is a limitation of the technique, not an accusation about the analyst.
Test one: the vaccine explanation, against Mississippi's timeline
Mississippi's infant mortality rate by year, from state vital records: 8.3 in 2020, 9.3 in 2021, 9.2 in 2022, 8.9 in 2023, and 9.7 in 2024. The mass vaccination of women of childbearing age happened in 2021. Mississippi's rate then declined for three consecutive years — 9.3, 9.2, 8.9 — before jumping in 2024, the fourth year after the rollout. An exposure that peaked in 2021 and fell away afterward does not comfortably explain a rate that improved through 2023 and then worsened in 2024.
The second problem is who the 2024 increase happened to. Between 2023 and 2024 the infant mortality rate for Black infants in Mississippi rose from 12.3 to 15.2 per 1,000 live births, while the rate for white infants fell from 6.3 to 5.8. In counts: 171 Black infant deaths in 2023 and 195 in 2024, against 110 white infant deaths in 2023 and 100 in 2024. The entire increase sits on one side of that line.
Here we have to be careful, because this is where it would be easy to overclaim, and where an earlier version of this article did. Nationally there was a very large racial gap in vaccination during pregnancy — the CDC's Vaccine Safety Datalink, covering 135,968 pregnancies across eight integrated health systems, recorded coverage of 6.0% among non-Hispanic Black women against 19.7% among non-Hispanic white women in the first months of the rollout. But Mississippi has no site in that system, and that national gap cannot be assumed to hold there. In Mississippi it demonstrably does not hold for the adult population: Black residents are 38% of the state and were receiving about 40% of weekly doses by May 2021, and by late 2022 roughly 52% of Black Mississippians were fully vaccinated against roughly 46% of white Mississippians — with state figures and CDC figures disagreeing about which group was slightly ahead. Mississippi was one of the few states where Black vaccination did not lag. We could find no Mississippi-specific figure for COVID-19 vaccination during pregnancy broken down by race. The place such a number would most likely live is PRAMS, the CDC survey that samples birth certificates and can produce state-level breakdowns, and Mississippi's health department does not publish vaccination by race on its own dashboard at all; we could not obtain a Mississippi pregnancy figure from either. So this is an absence of available data, not a demonstration that the gap was absent.
So the honest version of the argument is not that the least-vaccinated group had the most deaths. It is this: in Mississippi, vaccination was distributed roughly evenly across Black and white residents, and yet between 2023 and 2024 Black infant mortality rose by nearly a quarter while white infant mortality fell. A cause spread evenly across two groups does not produce a large change in one of them and an improvement in the other. Whatever drove the 2024 divergence tracks something that differs sharply between those two populations in Mississippi, and vaccination is not one of those things.
Test two: the abortion-ban explanation, against the same timeline
The competing explanation deserves the same scrutiny, and it does not survive it either.
The research is real and serious. Alison Gemmill, a perinatal epidemiologist at Johns Hopkins, and colleagues published a study in JAMA Pediatrics in August 2024 finding infant deaths in Texas rose 12.9% after Senate Bill 8 took effect in September 2021, with deaths from congenital anomalies up 22.9% there while falling 2.9% in the rest of the country. A companion study in JAMA in February 2025 extended the analysis to 14 states with complete or six-week bans, Mississippi among them, estimating 478 excess infant deaths, a 5.60% relative increase, concentrated among Black infants (10.98% above expected) and in congenital-anomaly deaths (10.87% above expected).
Now the timing. That study's data runs January 2012 through December 2023, and contains no 2024 data at all. For Texas the post-ban window analyzed covers 2022–2023; for the other thirteen states, Mississippi included, the analyzed exposure window is January through June 2023 only. Mississippi's ban took effect in July 2022, and the state's infant mortality rate in 2023 was 8.9, down from 9.2 the year before. The jump to 9.7 came in 2024, a year the study does not cover and a year no published analysis of abortion bans has yet reached. Citing that research to explain Mississippi's 2024 increase asks it to speak about a period it never examined, in a state where the year it did examine went the other way.
None of this makes the Gemmill findings wrong. Their result about 2023 stands on its own and their limitations are stated plainly by the authors, including suppressed cells requiring imputation, no gestational-age data, and provisional 2023 figures. It means the study cannot carry the weight of explaining 2024, and anyone reaching for it to answer this emergency is doing what the viral post does: attaching a preferred cause to a number that arrived outside its reach.
Test three: the infection itself, which we did not test the first time
The explanation that gets raised least often has the largest documented effect of the three, and an earlier version of this article omitted it entirely. That was our error, and correcting it matters more than the omission would suggest, because on the evidence this hypothesis outperforms both of the others for one of the two years in question.
COVID-19 infection during pregnancy is a documented, substantial risk to the baby. The CDC examined 1,249,634 delivery hospitalizations from March 2020 through September 2021 and found that women with COVID-19 at delivery had an adjusted relative risk of stillbirth of 1.90 compared with women without it. During the Delta period specifically that figure was 4.04, against 1.47 before Delta. In the infected group, stillbirth prevalence ran 3.2% against a 0.6% baseline, and preterm delivery ran about 70% above baseline. Preterm birth and low birthweight are the leading drivers of infant death in Mississippi, so this is a mechanism that connects directly to how babies there actually die.
Now the Mississippi timeline. In August 2021 the state's hospital system was, by its own health officials' account, five to ten days from failure. The University of Mississippi Medical Center built a field hospital in a parking garage. The state hit a record 5,023 daily cases and had the lowest vaccination rate in the country. And Mississippi's infant mortality rate rose from 8.3 per 1,000 in 2020 to 9.3 in 2021 — the Delta year.
That is a closer temporal fit than either of the other two explanations, attached to a mechanism with a large measured effect. It would be easy to stop there and call it the answer. We are not going to, for the same reason we did not let the abortion-ban study stand: no study has established that link for Mississippi. Our own deep search returned Mississippi-specific maternal COVID deaths but no analysis tying the state's 2021 infant mortality rise to the Delta wave. A temporal coincidence plus a plausible mechanism is a strong hypothesis, not a finding, and treating it as a finding is exactly the move this article is about.
It also does not explain 2024. As the pandemic receded, Mississippi's rate fell to 9.2 and then 8.9 — then rose to 9.7. Infection looks like a serious candidate for 2021 and not the answer to the year the emergency was declared over.
What none of this can rule out
A fair reader will ask whether an mRNA vaccine could cause harm that only shows up years later, and whether anyone has actually checked. The honest answer is that nobody can verify a ten-year horizon, because it has not happened yet. The first doses were given on December 14, 2020. That is five years and nine months of human experience. Anyone claiming long-term safety has been proven is overstating the record, and we are not going to make that claim on their behalf.
Here is the actual state of the evidence, with its limits attached. The longest follow-up we could find of children born to vaccinated mothers is the ASPIRE cohort, which assessed neurodevelopment at 12 and 18 months and found no signal. Eighteen months is the horizon, not a decade. The National Academies reviewed the fertility question formally and concluded that "the evidence favors rejection of a causal relationship" between the Pfizer and Moderna vaccines and female infertility — a real conclusion, but the second-strongest of their four categories rather than proof, and the committee noted in the same breath that female infertility "was not an outcome studied in the clinical trials submitted for authorization or approval" and that follow-up in the studies they reviewed was limited.
On persistence, a common reassurance is overstated in the other direction: the vaccine is not simply gone in days. Vaccine mRNA has been detected in plasma up to 15 days after injection, spike protein in axillary lymph node biopsies up to 60 days, and the ALC-0315 lipid persists in tissue for several weeks, concentrating in liver and spleen. Weeks to months is the accurate range — longer than the reassurance, far shorter than the claim. On genomic integration, the paper usually cited is Alden et al. 2022, which reported reverse transcription of vaccine mRNA in a liver cancer cell line in a dish. It has drawn a published critical comment and has not been replicated in a living organism; it also, contrary to what is sometimes said, has not been retracted. It is neither proof nor a debunked artifact. For a germline or epigenetic effect passed to a child conceived later, no mechanism is established and no study has looked. Untested is the correct word, not disproven.
So why does that not rescue the claim in the viral post? Because the test in this article does not depend on latency at all. Harm that arrives at year one, year three or year ten still has to travel through the people who were vaccinated. In Mississippi, vaccination was distributed roughly evenly across Black and white residents, and yet one group's infant mortality rose by nearly a quarter while the other's fell. A delayed effect would arrive in both groups. The argument survives any lag you care to propose — and its real weak point is not latency but the one we have already named: those are adult figures, and no Mississippi pregnancy-specific breakdown by race exists.
One more thing belongs here, in the interest of applying the same standard in both directions. The longest-horizon evidence that does exist points at the infection rather than the vaccine: a cohort of 18,124 live births followed for three years found that maternal SARS-CoV-2 infection in pregnancy was associated with a significantly increased risk of neurodevelopmental diagnoses. Three years of follow-up for the infection, eighteen months for the vaccine. Both horizons are shorter than the question deserves.
What the state actually counted
Mississippi's report is specific about the deaths themselves. Across 2022–2024, the three leading individual causes were extremely low birthweight or extreme immaturity (129 deaths), accidental suffocation and strangulation in bed (98 deaths), and Sudden Infant Death Syndrome (86 deaths). In 2024, 55% of neonatal deaths occurred within five hours of birth, and a third of all infant deaths were newborns who died in those first five hours. Three health districts — the Delta, East Central, and Southwest — carried neonatal mortality rates around 13 per 1,000. The report states that Black babies "are more likely to die soon after birth due to extreme prematurity, and they are more likely to die due to unsafe sleep conditions."
What nobody has established
The most accurate thing that can be said about Mississippi's 2024 increase is that its cause has not been established, by us or by anyone else we could find. The state has named the mechanisms of death — extreme prematurity and unsafe sleep — but naming how babies died is not the same as explaining why the number rose in that particular year, and Mississippi's report does not claim to have done the second thing. A single-year jump in a state with about 33,000 annual births can also move for reasons that are partly statistical noise; 2024's 323 deaths against 2023's 305 is a difference of eighteen babies.
That is an unsatisfying answer, and it is the one the evidence supports. Three explanations are in circulation and each was checked here against the same records, by the same standard. Infection is a serious candidate for the 2021 rise and none of the three accounts for 2024. What remains is a state with the worst infant mortality in the country, where babies die of being born too small and too soon and of sleeping somewhere unsafe, concentrated in its poorest counties, among mothers who in many cases cannot reach an obstetrician. Those are the deaths in the number. Attaching them to a preferred national storyline — from either direction — points attention away from the two things Mississippi's own count says are killing the most babies, and which are the two things a family, a hospital or a legislature could actually act on.
Sources, all checkable. Mississippi's declaration and infant mortality figures: the MSDH emergency declaration of August 21, 2025 and the MSDH 2025 Infant Mortality Report. National infant mortality: NCHS period linked birth/infant death file, 2024. Vaccination in pregnancy by race (national, not Mississippi): Razzaghi et al., MMWR volume 70, number 24. Mississippi vaccination by race: MSDH figures as reported by KFF Health News and Mississippi Public Broadcasting. Abortion-ban research: Gemmill et al., JAMA Pediatrics (2024) and JAMA (2025). Stillbirth risk with COVID-19 infection: CDC MMWR, 1,249,634 delivery hospitalizations, March 2020–September 2021. Long-term evidence: the ASPIRE cohort in JAMA Pediatrics (2024) for offspring neurodevelopment at 12 and 18 months, and the National Academies’ 2024 evidence review for the fertility conclusion. Where we could not find something — a published methodology for the ">50%" figure, and any Mississippi-specific figure for vaccination during pregnancy by race — we have said so rather than substituting a national number in its place.
