No large, well-designed study has established that COVID-19 vaccines cause dementia. One widely cited study from South Korea did report a short-term uptick in mild cognitive impairment diagnoses among vaccinated people, but the finding came with serious methodological caveats and was not replicated for Alzheimer’s disease or vascular dementia in broader analyses. Meanwhile, a growing body of evidence points in the other direction: COVID-19 infection itself is linked to a meaningful increase in dementia risk, and vaccination appears to reduce the cognitive fallout of infection.
The South Korean Study That Sparked the Concern
Much of the worry traces to a 2024 study published in QJM that analyzed health records from South Korea. The researchers compared vaccinated and unvaccinated groups and found that people who received mRNA COVID vaccines had higher rates of mild cognitive impairment (MCI) and Alzheimer’s disease diagnoses in the three months following vaccination. The odds of an MCI diagnosis were roughly 2.4 times higher in the mRNA group, and the odds of Alzheimer’s were about 1.2 times higher.1PubMed. A potential association between COVID-19 vaccination and development of Alzheimer’s disease Those numbers sound alarming in isolation, and they spread quickly on social media. But the study’s design has significant limitations that weaken any causal interpretation.
A three-month follow-up window is extraordinarily short in the context of dementia. Alzheimer’s disease develops over decades. A diagnosis appearing within weeks of a vaccine likely reflects a disease process already underway, not one the vaccine triggered. More importantly, people who get vaccinated tend to interact with the healthcare system more often. They see doctors, get screened, and receive diagnoses that an unvaccinated person avoiding medical care might not receive for months or years. This kind of detection bias is one of the most common pitfalls in observational vaccine studies.
When a large systematic review and meta-analysis examined the same question, the authors noted that this South Korean dataset was essentially the only large study addressing COVID-19 vaccination and dementia. After accounting for confounders, the review found that COVID vaccination was associated with roughly double the risk of MCI, but showed no statistically significant effect on Alzheimer’s disease or vascular dementia as distinct outcomes.2Age and Ageing. Association between vaccinations and risk of dementia: a systematic review and meta-analysis A single study, even a large one, does not constitute proof, especially when its follow-up period is too brief to capture the natural history of the disease it claims to detect.
Why Timing Does Not Equal Causation
The question of whether a medical event after vaccination was caused by the vaccine is one that epidemiologists have wrestled with for decades, long before COVID. A useful illustration comes from influenza vaccination. A large Danish study tracking over 14 million person-years of follow-up found that people who received flu shots had a 4% higher rate of dementia diagnoses compared to those who were never vaccinated. That looks suspicious until you notice the study’s own negative control outcomes: vaccinated people also had a 7% higher rate of cancer diagnoses and a 6% higher rate of hip fractures.3PubMed Central. The effect of influenza vaccination on the rate of dementia amongst older adults Nobody seriously argues that the flu vaccine causes hip fractures. The pattern strongly suggests that people who get vaccinated are simply more engaged with healthcare and therefore more likely to receive diagnoses of all kinds. This is a textbook example of what researchers call healthy-vaccinee bias flipping in the opposite direction: the vaccinated group looks sicker not because the vaccine harmed them, but because they are being monitored more closely.
The same logic applies to the South Korean COVID vaccine data. When an older adult visits a clinic to get a booster shot, that interaction creates an opportunity for a clinician to notice early cognitive symptoms and order testing. An unvaccinated person with the same symptoms, staying home, might not get that diagnosis for another year or two. In a three-month follow-up window, this kind of ascertainment bias can make a vaccine look dangerous even if it had zero biological effect on the brain.
COVID-19 Infection and Dementia Risk
While the evidence linking vaccines to dementia remains thin and contested, the evidence linking COVID-19 infection itself to increased dementia risk is substantially stronger and comes from multiple independent research groups. A 2025 systematic review and meta-analysis pooling data from multiple cohort studies found that people who had been infected with COVID-19 had roughly a 49% higher risk of developing new-onset dementia compared to uninfected controls. That elevated risk persisted for up to two years and applied to both men and women, with the strongest signal in people over 65.4PubMed. Risk of new-onset dementia following COVID-19 infection: a systematic review and meta-analysis
A separate study focusing on adults aged 50 and older found that prior COVID-19 infection was associated with a 41% increase in all-cause dementia and a 77% increase in vascular dementia specifically, though no significant link was found for Alzheimer’s disease alone.5npj Dementia. COVID-19 infection associated with increased risk of new-onset vascular dementia in adults ≥50 years The vascular dementia finding is consistent with what we know about how COVID affects blood vessels and the brain’s vascular system. Interestingly, the same study found that when COVID patients were compared to people who had other respiratory infections, the excess dementia risk disappeared. That suggests the dementia risk may not be unique to the SARS-CoV-2 virus but could be a consequence of serious respiratory illness in general.
This matters for the vaccine debate because the practical question is not whether the vaccine has zero theoretical risk but whether vaccination leaves a person better or worse off than infection. If COVID infection reliably increases dementia risk by 40-65% and vaccination prevents or reduces infection severity, then avoiding the vaccine could mean accepting a much larger cognitive threat.
Vaccination Appears to Protect the Brain After Infection
Some of the most relevant evidence comes from studies examining what happens to the brain when a vaccinated person does get infected. A population-based cohort study looking at people infected during the Omicron wave found that having a recent booster vaccination before infection substantially reduced both neurocognitive and cardiorespiratory symptoms that appeared three or more months after the acute illness.6Vaccine. Post-COVID symptoms after SARS-CoV-2 omicron infection and the effect of booster vaccination In other words, vaccinated people who caught COVID were less likely to develop the lingering brain fog and cognitive difficulties associated with long COVID.
This fits a straightforward biological logic. Vaccination primes the immune system to clear the virus faster and with a less intense inflammatory response. Since many of the brain-related complications of COVID appear to be driven by prolonged inflammation and vascular damage, a faster immune response means less time for those processes to do harm. You can think of it as the difference between a small kitchen fire that gets extinguished in minutes and one that burns for hours before the fire department arrives.
How the Virus Interacts With Brain Proteins
Researchers have been investigating the molecular mechanisms by which SARS-CoV-2 might accelerate neurodegeneration. Several studies have explored whether the virus’s spike protein can interact with the proteins that misfold and aggregate in Alzheimer’s and other neurodegenerative diseases. Computational modeling and laboratory experiments have shown that the spike protein’s S1 subunit binds with notable affinity to amyloid-beta, the protein that forms plaques in Alzheimer’s brains. It also interacts with tau, the protein that forms the tangles characteristic of the disease.7PubMed Central. SARS-CoV-2 spike protein interactions with amyloidogenic proteins: Potential clues to neurodegeneration
Lab experiments have added weight to the computational predictions. One study found that a specific SARS-CoV-2 enzyme (its main protease) could accelerate the aggregation of tau protein in test-tube conditions, though the same effect was not seen for two other proteins sometimes implicated in neurodegeneration.8PLoS ONE. Tau protein aggregation associated with SARS-CoV-2 main protease And a 2025 study using human retinal tissue and lab-grown retinal organoids showed that exposure to the spike protein induced amyloid-beta aggregation, a process that was reduced when a specific receptor (neuropilin-1) was pharmacologically blocked.9PubMed Central. SARS-CoV-2 induces Alzheimer’s disease-related amyloid-β pathology in ex vivo human retinal explants and retinal organoids
These findings are important for framing the vaccine conversation. The spike protein produced by mRNA vaccines is a modified version that gets cleared by the immune system relatively quickly. During actual infection, however, the virus replicates and can produce spike protein continuously for days or weeks, at levels and in tissues far exceeding what a vaccine produces. If the spike protein does contribute to protein aggregation in the brain, infection represents a much higher and more prolonged exposure than vaccination does.
The Lipid Nanoparticle Question
One concern that circulates online is whether the lipid nanoparticles (LNPs) used to deliver mRNA in COVID vaccines can carry their cargo into the brain, potentially causing damage there. This worry is understandable given that LNPs are a delivery technology, and people reasonably want to know where the delivery goes. But the concern conflates two very different things.
The LNPs in authorized COVID vaccines were designed to deliver their payload primarily at the injection site and to the liver. They were not engineered to cross the blood-brain barrier, which is one of the body’s most selective gatekeepers. Researchers working on brain diseases have spent years trying to build LNPs that can cross this barrier, and the difficulty of that task is itself informative. Recent papers have described experimental LNP formulations that achieve brain delivery in mice by incorporating specialized molecules, like a serotonin receptor ligand, into their design.10PubMed Central. Lipid nanoparticles for mRNA delivery in brain via systemic administration Other teams have used entirely novel ionizable lipid designs to achieve similar results in mice and primates.11PubMed Central. Overcoming the blood-brain barrier using central nervous system-accessing lipid nanoparticles for enhanced mRNA therapeutics The fact that achieving brain penetration requires purpose-built, highly specialized nanoparticle formulations underscores that standard vaccine LNPs were not designed to do this and do not do it efficiently.
That said, no biological barrier is absolute. Trace amounts of any circulating substance can sometimes reach the brain, and some biodistribution studies in animals have detected very small amounts of vaccine lipid in brain tissue. The question is whether those trace levels are biologically meaningful or capable of causing harm. The weight of available evidence suggests they are not, but this is an area where continued monitoring and research are reasonable.
Other Vaccines and Dementia Risk
The broader picture of vaccination and dementia is worth understanding because it provides important context. Far from being a risk factor, most vaccines studied in older adults appear to be associated with lower dementia rates. A 2025 systematic review and meta-analysis found that herpes zoster (shingles) vaccination was linked to roughly a 24% reduction in any dementia and a 47% reduction in Alzheimer’s disease specifically. Pneumococcal vaccination was associated with a 36% lower risk of Alzheimer’s, and the Tdap vaccine (tetanus, diphtheria, pertussis) with about a 33% lower risk of any dementia.2Age and Ageing. Association between vaccinations and risk of dementia: a systematic review and meta-analysis
The recombinant shingles vaccine (Shingrix) has received particular attention. A study published in Nature Medicine found that it was associated with lower dementia risk compared to both influenza and Tdap vaccines in the same older population.12PubMed Central. The recombinant shingles vaccine is associated with lower risk of dementia Influenza vaccination showed a more modest and population-dependent effect: a meta-analysis found no significant protection in the general population, but among people at high risk for dementia, repeated flu vaccination (four or more doses over time) was linked to a dramatically lower dementia risk.13Age and Ageing. Influenza vaccination and risk of dementia: a systematic review and meta-analysis
Why would vaccines protect against dementia? Several hypotheses exist. Some infections, particularly herpes viruses, may directly contribute to neurodegeneration over time, and preventing those infections could slow the process. Vaccination also reduces the systemic inflammation that accompanies infection, and chronic inflammation is increasingly recognized as a driver of cognitive decline. These are observational associations, not proven causal relationships, but they consistently point away from the idea that stimulating the immune system through vaccination harms the brain.
Rare Neurological Events in Context
It would be dishonest to suggest that COVID vaccines have zero neurological side effects. Rare cases of autoimmune encephalitis, a condition where the immune system attacks brain tissue, have been reported after COVID vaccination. A systematic review identified 14 such cases across all reported literature, with symptoms appearing a median of five days after the shot. Most patients responded to treatment with steroids and other immune-modulating therapies, though the majority had some residual effects at the time of reporting.14PubMed Central. AI-CoV Study: Autoimmune Encephalitis Associated With COVID-19 and Its Vaccines—A Systematic Review Fourteen cases across hundreds of millions of doses is vanishingly rare, but it is not zero.
There is also a single published case report of a 65-year-old man who experienced sudden memory loss after receiving the AstraZeneca COVID vaccine. The episode lasted one day and the patient fully recovered within 24 hours, likely due to a transient cerebrovascular event rather than any lasting neurodegeneration.15ScienceDirect (Brain, Behavior, & Immunity – Health). Cognitive deficits and memory impairments after COVID-19 (Covishield) vaccination Individual case reports like this are important for safety surveillance, but they describe isolated events and cannot establish a pattern of causation.
Similarly, a case report of a man who developed Creutzfeldt-Jakob disease (CJD, a rare and fatal prion disease) 14 days after a COVID-19 illness has been cited in anti-vaccine circles. The paper itself, however, describes the onset as following COVID infection, not vaccination, though the patient had been vaccinated previously. The authors specifically called for monitoring of cognitive symptoms after COVID illness, not after vaccination.16PubMed Central. Post-COVID Neurodegeneration: A Puerto Rican Case of Rapidly Fatal Creutzfeldt-Jakob Disease
The Pandemic’s Hidden Cognitive Toll
Any honest accounting of dementia trends during the COVID era has to reckon with something that has nothing to do with either the virus or the vaccine: social isolation. The public health measures adopted during the pandemic, particularly lockdowns and visitor restrictions at care facilities, had measurable effects on the cognitive health of older adults. A systematic review and meta-analysis found that the proportion of people with dementia who experienced worsening cognitive impairment or new behavioral symptoms during isolation was roughly double the rate seen in cognitively healthy older adults experiencing subjective cognitive decline during the same period.17PubMed Central. The impact of social isolation from COVID-19-related public health measures on cognitive function and mental health among older adults
Social engagement, physical activity, and cognitive stimulation are all protective against dementia progression. When those were stripped away during lockdowns, existing cognitive decline accelerated. For someone looking at dementia rates in 2021 or 2022 and trying to assign blame, this background deterioration can easily be misattributed to the vaccine, which was rolling out during the same period. The timing overlap is almost perfectly designed to create false causal narratives.
Disruptions to routine medical care also played a role. Many older adults delayed or skipped appointments during the pandemic. When they finally returned to care, accumulated cognitive decline that would normally have been tracked gradually appeared as a sudden drop. A clinician seeing that patient for the first time in 18 months might diagnose MCI or early dementia that had actually been developing throughout the gap in care, and if the patient recently received a vaccine at their re-entry appointment, the temporal association writes itself.
What to Make of the Ongoing Research
The science is not finished. Researchers continue to study both COVID-19 infection and vaccination for long-term neurological effects, and it would be premature to say that every possible concern has been resolved. What the current evidence supports, though, is a clear asymmetry. COVID-19 infection carries a well-documented and replicated increase in dementia risk, roughly 40-65% higher than in uninfected populations, sustained for at least two years. COVID-19 vaccination has one contested observational study suggesting a short-term increase in MCI diagnoses, no replicated finding for Alzheimer’s or vascular dementia, and evidence of a protective effect against the cognitive complications of infection. Multiple other vaccines studied over longer time horizons have been consistently linked to reduced, not increased, dementia risk.
For anyone weighing personal risk, the picture from the existing data is that the virus poses a far greater cognitive threat than the vaccine. That does not mean the question is illegitimate. It means the answer, based on the evidence currently available, does not support the claim that COVID vaccines cause dementia.