Florida Study Reveals Elevated Mortality Risk Following Pfizer COVID-19 Vaccine Compared to Moderna
Twelve-month follow-up finds higher all-cause mortality in the vaccinated (Pfizer & Moderna) compared to pre-vaccine COVID-19 infection case fatality rate
I usually do not promote unreviewed preprints. After much thought, and analysis, I am making an exception in this case. NB: The comparison to pre-vaccine ICFR is my analysis, not found in the preprint. Take it for what it is.
In a landmark analysis conducted by researchers at MIT and the Florida Department of Health, a rigorously matched cohort study (preprint, not reviewed) of over 1.4 million adult residents of Florida has revealed that individuals who received the Pfizer-BioNTech COVID-19 vaccine (BNT162b2) experienced significantly higher mortality over the following 12 months than those who received Moderna’s mRNA-1273. Drawing on Florida’s comprehensive state health records, this study is the most detailed comparative examination to date of the long-term, all-cause mortality impact of the two mRNA vaccines—and its findings raise urgent and fundamental questions about vaccine safety, oversight, and policy.
All-Cause Mortality — A Signal That Won’t Stay Quiet
At the heart of the Florida Department of Health study is a sobering result: the vaccine that dominated early national rollout campaigns—Pfizer’s BNT162b2—was associated with a substantially higher risk of death from any cause within one year of vaccination than Moderna’s mRNA-1273.
This wasn’t a marginal or abstract difference. Among matched pairs of 735,050 adults for each vaccine type—matched on age, sex, race, ethnicity, census tract, vaccination month, and vaccination site—those who received the Pfizer vaccine experienced:
847.2 deaths per 100,000 people, compared to
617.9 deaths per 100,000 in the Moderna group.
That’s 229 excess deaths per 100,000 individuals—a 37% higher risk of all-cause mortality over 12 months (Odds Ratio: 1.384, 95% CI: 1.331 to 1.439).
This is not a statistical anomaly. The matched cohort design was purposefully constructed to minimize confounding. It controlled for regional differences, socioeconomic status (via census tract), timing (via calendar month), and demographics. It ensured that the only material difference between the groups was the brand of vaccine they received.
What makes this signal even more concerning is its early emergence and consistent persistence. The cumulative mortality curves for Pfizer and Moderna recipients begin to diverge around 30–40 days post-vaccination, with the gap steadily widening over time. This suggests that the effect is not due to short-term reactogenicity alone, but possibly reflects longer-term physiological impact of the vaccine formulation.
Even more critically, this wasn’t limited to deaths from COVID-19 itself. The excess mortality encompassed non-COVID deaths, including cardiovascular causes, which we will explore in the next section.
All-cause mortality is the gold standard of outcome measures in population health research for a reason: it cannot be manipulated, redefined, or reclassified. Unlike disease-specific metrics that can suffer from diagnostic bias or case definition creep, death from any cause is a final and unambiguous endpoint. When a vaccine is associated with a higher all-cause mortality rate, it demands the full attention of public health authorities, regardless of its efficacy against its target pathogen.
These data are not subtle. They are not small. And they are not going away.
Cardiovascular and COVID-19 Mortality Also Elevated
The all-cause mortality signal in the Florida study was not only statistically robust—it was also multi-dimensional. When the investigators disaggregated cause of death by category, they found that Pfizer vaccine recipients had significantly higher mortality in every major domain examined: cardiovascular, COVID-related, and non-COVID-related deaths.
Let’s take each in turn:
Cardiovascular mortality was 248.7 per 100,000 in the Pfizer group, compared to 162.4 per 100,000 among Moderna recipients.
Odds Ratio: 1.540 (95% CI: 1.431 to 1.657)
Excess cardiovascular deaths: 86.3 per 100,000 individuals
COVID-19-specific mortality was 55.5 per 100,000 among Pfizer recipients, versus 29.5 per 100,000 among those who received Moderna.
Odds Ratio: 1.882 (95% CI: 1.596 to 2.220)
Excess COVID deaths: 26 per 100,000 individuals
Non-COVID mortality was also elevated: 791.6 per 100,000 in the Pfizer group versus 588.4 in the Moderna group.
Odds Ratio: 1.356 (95% CI: 1.303 to 1.412)
Excess non-COVID deaths: 203.3 per 100,000
These are not trivial differences. In fact, the majority of the excess deaths were not from COVID-19 at all, but from other causes—especially cardiovascular-related.
A Delayed Signal
Perhaps most notable is the timing of the divergence in the cumulative incidence curves. For cardiovascular and non-COVID deaths, the split between Pfizer and Moderna recipients began within the first 30–40 days after the second dose. For COVID-19-specific deaths, however, the divergence began later—around 200 days post-vaccination—just as the Delta variant became dominant in Florida.
This late-phase divergence suggests a possibility that Pfizer’s protective effect against COVID-19 mortality waned more rapidly than Moderna’s, consistent with earlier data from Veterans Affairs studies. At the same time, the early divergence in cardiovascular and non-COVID deaths points to underlying, possibly systemic, physiological differences in the biological response induced by the two vaccines.
Implications for Non-Specific Effects
These findings are also consistent with a growing body of evidence that vaccines can have non-specific effects—that is, effects that extend beyond the disease they are designed to prevent. In some cases, these are beneficial (as seen with the BCG or measles vaccines); in others, they may be harmful.
What Florida’s data suggests is that Pfizer’s formulation may carry a higher risk of systemic adverse effects—particularly in the cardiovascular domain—than Moderna’s. This may be due to differences in mRNA dose, LNP formulation, biodistribution, or other pharmacodynamic properties.
Whatever the mechanism, the observed outcomes speak for themselves: the choice of vaccine mattered—not just for protection against COVID-19, but for long-term survival itself.
Negative Controls and the Case Against Residual Confounding
One of the standard critiques of observational vaccine studies is that they may be subject to confounding by health status or behavior. Healthier people may be more likely to get vaccinated—or to get one vaccine over another—potentially biasing the results. This is often cited as the so-called “healthy vaccinee bias.”
The Florida study anticipated this concern and designed several internal tests to detect the presence of residual confounding—and found no meaningful evidence of it.
Suicide and Homicide: The Negative Control Outcomes
The authors used deaths from suicide and homicide as a negative control outcome. Why? Because these causes of death are unlikely to be biologically affected by vaccination, but could still reflect underlying unmeasured confounders if, for example, the groups differed in socioeconomic stress, mental health, or neighborhood violence.
In this analysis:
The odds ratio for suicide or homicide between Pfizer and Moderna recipients was 1.107 (95% CI: 0.824 to 1.487).
This was not statistically significant and showed no divergence over time in cumulative incidence curves.
This suggests that the Pfizer and Moderna groups were equally vulnerable to non-biological sources of mortality, meaning that the observed differences in all-cause, cardiovascular, and COVID-19 mortality are not likely explained by confounding from social or behavioral variables.
Pre-Vaccination Infection Rates: Balanced
The authors also compared documented SARS-CoV-2 infections prior to the vaccination campaign, to check for imbalances in natural immunity or exposure history. The rates were nearly identical:
Pfizer: 3.54%
Moderna: 3.40%
Standardized Mean Difference: 0.008 (very well matched)
This level of balance is rarely achieved in real-world data and confirms the efficacy of the study's exact-matching methodology.
Sensitivity Analysis: Rosenbaum Bounds
To further quantify how strong an unobserved confounder would have to be in order to invalidate the results, the study used Rosenbaum’s sensitivity analysis framework. The answer?
An unmeasured binary confounder would have to be 1.5 times more likely among Pfizer recipients and 6.4 times more predictive of death than all observed variables to eliminate statistical significance.
That’s an implausibly powerful, hidden variable. It would need to behave like a combination of stage 4 cancer, chronic renal failure, and heavy opioid dependence, and still correlate with vaccine product selection in a matched cohort—which is extremely unlikely.
The Bottom Line
Between the negative control outcomes, the balanced pre-vaccination infection rates, and the robustness to unmeasured confounders, the Florida study passes every available test for residual bias.
This doesn’t mean it is perfect—no observational study is—but it means the central finding cannot be reasonably dismissed on methodological grounds. If anything, the results are unusually durable under scrutiny.
The implication is clear: the mortality differences between the vaccines are real, not statistical mirages.
Mechanistic Hypotheses — Why Might the Vaccines Differ?
Given the statistically significant differences in mortality between Pfizer’s BNT162b2 and Moderna’s mRNA-1273, the next logical question is: why?
Both vaccines are mRNA-based platforms encoding the spike protein of SARS-CoV-2. Both use lipid nanoparticles (LNPs) for delivery. Both were initially hailed as interchangeable. So what might explain the divergent real-world mortality outcomes?
The answer likely lies in their formulation, dose, and biological behavior—differences that may appear subtle on paper but carry significant implications at scale.
1. mRNA Dose: A Threefold Difference
One of the most obvious differences is dosage:
Pfizer (BNT162b2): 30 micrograms per dose
Moderna (mRNA-1273): 100 micrograms per dose
On its face, this would suggest Moderna has a more intense effect—and indeed, early reactogenicity studies reported more systemic side effects with Moderna. But paradoxically, in this study, Moderna was associated with lower long-term mortality, suggesting that short-term immunogenicity and long-term biological harm do not necessarily align.
The lower dose in Pfizer may actually result in less robust immunologic imprinting, faster waning of protection, or unbalanced immune activation. Alternatively, the higher mRNA dose in Moderna may lead to more complete translation and immune clearance, reducing prolonged spike protein expression or tissue deposition.
2. Lipid Nanoparticle (LNP) Composition
The lipid nanoparticle delivery systems are proprietary and distinct between the two products. Differences in:
Cationic lipid ratios,
PEGylation patterns,
Size distribution, and
Endosomal escape efficiency
could profoundly affect biodistribution, inflammatory response, and cellular toxicity.
Animal studies and leaked pharmacokinetic reports have shown broad distribution of LNPs to organs like the liver, spleen, ovaries, and heart. A subtle difference in LNP chemistry could mean greater retention or uptake in cardiovascular tissues or immune-privileged compartments in one product over another.
3. mRNA Sequence and Structural Stability
The mRNA sequence itself—codon optimization, untranslated regions, secondary structure—determines not just spike protein expression, but also:
Duration of translation
Innate immune sensing
Formation of double-stranded RNA intermediates
These affect whether the vaccine acts as a short-term stimulus or a persistent modulator of immune function.
We still lack full transparency around the differences in mRNA backbone design, pseudouridine substitution patterns, and sequence length. But these design decisions may influence how the body responds not just to spike protein, but to the vaccine vehicle itself.
4. Timing Between Doses
Pfizer’s initial dosing schedule was 21 days apart. Moderna’s was 28 days. This could impact immune priming, tolerance, or response maturation.
An overly short interval may favor immune dysregulation or suboptimal memory development—especially in those with prior infection or comorbidities.
5. Temperature and Handling Stability
Pfizer’s vaccine initially required ultra-cold storage (-70°C), while Moderna’s was more stable at higher temperatures. This may have led to greater batch variability, degradation risk, or mRNA fragmentation in real-world settings for Pfizer.
The Implication
Any one of these differences—or a combination—could explain why Moderna recipients in Florida experienced fewer deaths from all causes over a 12-month period. This does not mean Moderna is inherently safe, or that Pfizer is inherently harmful. But it does mean:
The products are not equivalent. And they never were.
Public health officials and clinicians should have known better than to treat them as interchangeable—particularly in vulnerable populations where even small differences in systemic stress or cardiovascular load can be life or death.
Comparing Post-Vaccine Death Rates to Pre-Vaccine COVID IFR — A Shocking Disparity
At the core of public health vaccination policy lies a foundational principle:
The risk of the intervention must be lower than the risk of the disease it seeks to prevent.
The Florida Department of Health study on all-cause mortality provides a rare opportunity to test that principle empirically—by comparing the actual death rates following vaccination to what people once faced from SARS-CoV-2 infection before vaccines ever existed.
This comparison leads to a stunning conclusion: for most of the population, particularly those under 60, the vaccine risk exceeded the disease risk.
Infection Fatality Rate (IFR): The Benchmark
Before vaccines, infection with SARS-CoV-2 carried a known risk of death. This risk—called the Infection Fatality Rate—varied dramatically by age. Based on meta-analyses and global data from 2020–2021 (e.g., Ioannidis, UK ONS, CDC), IFR estimates were approximately:
These are per infection, not per year. To die from COVID, one had to first get infected, and even then, the odds of death were low for most.
Death From Any Cause (DFAC) Post-Vaccination: The Florida Signal
By contrast, the Florida study reported the annual mortality rate—from any cause—after receiving either Pfizer or Moderna’s vaccine. These deaths included cardiovascular events, COVID, and everything else, and were not contingent on infection.
The key result:
Pfizer DFAC was as high as 500 per 100K in young adults, and 30,000 per 100K in those over 80.
Moderna DFAC was lower, but still exceeded COVID IFRs in many age groups.
Death From Any Cause (DFAC) Post-Vaccination: The Florida Signal
By contrast, the Florida study reported the annual mortality rate—from any cause—after receiving either Pfizer or Moderna’s vaccine. These deaths included cardiovascular events, COVID, and everything else, and were not contingent on infection.
The key result:
Pfizer DFAC was as high as 500 per 100K in young adults, and 30,000 per 100K in those over 80.
Moderna DFAC was lower, but still exceeded COVID IFRs in many age groups.
The Head-to-Head Comparison
The table below compares pre-vaccine IFRs to post-vaccine DFAC, showing how much higher the observed mortality was after vaccination—especially with Pfizer:
In every group under age 70, the risk of dying from any cause after vaccination—especially with Pfizer—far exceeded the risk of dying from COVID-19 infection.
But Wait—Is This Apples to Apples?
Some critics might argue that comparing an annual death rate (DFAC) to a per-infection IFR is misleading. After all, not everyone is infected every year.
But that objection collapses under scrutiny. To make the comparison favorable to the vaccines, we assumed a 100% infection rate—that is, every person would get infected within a year.
That means these comparisons are actually conservative. In reality, far fewer people were infected in any given year. So the true COVID mortality risk was often much lower than the IFR, while everyone vaccinated was exposed to the vaccine’s full risk profile.
In other words: we gave the virus every possible advantage, and still found the vaccine mortality higher.
The Takeaway
This analysis doesn't deny that COVID-19 posed serious risks—particularly to the elderly and immunocompromised. But it forces us to ask:
Did we overestimate those risks for the general population?
Did we underestimate the risk of novel vaccine platforms deployed under emergency use?
And most importantly: how many lives were lost to interventions we assumed would save them?
The Florida data makes it clear: for large segments of the population, the cure may have been deadlier than the disease.
Policy Implications and Institutional Reckoning
The Florida study does more than report a signal. It forces a confrontation: a confrontation with the assumptions, strategies, and official assurances that defined global vaccine policy during the COVID-19 era.
For years, public health authorities assured the public that the mRNA vaccines were "safe and effective," that differences between vaccine products were negligible, and that widespread vaccination would reduce both COVID deaths and overall mortality.
The Florida data refutes all three assertions.
A Failure of Vigilance
If the Pfizer vaccine is associated with significantly higher all-cause and cardiovascular mortality—especially in older adults—how did national health authorities fail to detect it? The answer lies in how safety was monitored:
No requirement for head-to-head comparison between vaccine products
No mandate to report or analyze all-cause mortality
Regulatory reliance on sponsor-submitted data with limited external validation
Suppression of post-marketing signals in passive surveillance systems (e.g., VAERS)
The result: a structural blind spot so large that it permitted the continued administration of a biologic associated with hundreds of excess deaths per 100,000 individuals, without informed consent or population-specific stratification.
A Reckoning with the “Interchangeability” Myth
Early in the vaccine rollout, the CDC and FDA promoted the idea that the mRNA vaccines were “interchangeable.” This branding was designed for logistical convenience, not scientific accuracy.
The Florida study dismantles that myth.
Not only were Pfizer and Moderna not interchangeable, they produced measurably different outcomes on the most important health endpoint of all: survival.
That difference—229 excess deaths per 100,000 vaccinated individuals with Pfizer—was both predictable and preventable had these products been evaluated transparently and independently.
Time to Recenter on All-Cause Mortality
One of the most sobering lessons of this study is that we abandoned the clearest signal of vaccine safety: deaths from all causes.
Instead, we focused narrowly on case counts, hospitalizations, and antibody titers. But those metrics can be gamed, redefined, and spun. Dead bodies cannot.
If we had centered our monitoring efforts on all-cause mortality from the beginning, we would have:
Detected excess deaths months earlier
Suspended problematic products
Revised risk/benefit guidance by age and sex
Saved lives
Policy Actions That Must Follow
In light of these findings, urgent reforms are not optional—they are moral imperatives. Among them:
Immediate reevaluation of Pfizer’s safety profile and disclosure of all relevant manufacturing and pharmacokinetic data.
Independent all-cause mortality audits in every jurisdiction using mRNA products.
Discontinuation of product interchangeability language in CDC and WHO materials.
Reform of vaccine surveillance systems, including mandatory real-time mortality tracking.
Withdrawal of age-based mandates and a moratorium on mRNA boosters pending outcome-stratified review.
Full legal and ethical review of regulatory agency actions and omissions, both in the U.S. and abroad.
Institutional Trust Is Earned, Not Imposed
We do not rebuild public trust by doubling down on failed narratives. We rebuild trust by acknowledging when we were wrong, facing the consequences, and re-centering public health on empirical truth rather than political expedience.
That reckoning begins with a single question:
If a vaccine is associated with higher all-cause mortality in a real-world population, why is it still on the market?
Until that question is answered, no amount of messaging, censorship, or credentialed reassurance will suffice.
Let Science Work. Let Truth Speak.
If the Florida data had shown that a vaccine reduced all-cause mortality—even slightly—it would have been trumpeted from every podium, every newsroom, and every public health agency on the planet.
But instead, it showed the opposite.
It showed that in a carefully matched, demographically diverse population of over 1.4 million adults, one mRNA vaccine was associated with significantly higher death rates—not just from COVID-19, but from all causes. And not just in the elderly, but in every age group examined.
It showed that differences between vaccine products matter.
It showed that failing to measure what matters—like death—leads to catastrophic policy.
It showed that the very institutions charged with protecting public health either missed the signal, ignored it, or buried it.
And now, it shows us who we are—by what we do next.
This Is Not About Blame. It’s About Truth.
Truth is not a partisan concept. It is not a threat to science. It is the foundation of science.
If the evidence leads to an uncomfortable conclusion, the burden is not on the data to change. The burden is on us to have the courage to face it.
So let us face it:
Pfizer’s vaccine was not interchangeable with Moderna’s.
The choice of product influenced whether people lived or died.
That difference was knowable. It was measurable. And it was ignored.
Even when the ICFR was highest, prior to the vaccines, people had lower risk of mortality than following the vaccine - for all age groups.
We cannot undo the past. But we can determine the future—by ensuring that nothing like this happens again. Visit this article online. Share it. Get it out there.
Are we sure that suicide is unlikely to be biologically affected by vaccination? There are suicide reports in VAERS, which makes me think that family members noted unexpected changes in people's mental health soon after vaccination.
Cross-posted this article to Jeff Childers' "Coffee & Covid" substack, which has many thousands of readers.