At least 24.5% of former NFL players who died between 2016 and 2021 had chronic traumatic encephalopathy in their brain tissue at death, according to a retrospective population based cohort study published August 25, 2026 in The BMJ. The team, led by Dr. Daniel Daneshvar of Harvard Medical School and Massachusetts General Hospital, built its estimate from all 1,712 former NFL players who died between 2008 and 2021, not just the ones whose families donated a brain, which makes it the first prevalence figure for this group anchored to a real denominator.
Key takeaways
- Among all 1,712 NFL players who died from 2008 to 2021, the possible rate of CTE at death ran from 18.5% to 98.7%. For the 2016 to 2021 deaths, when donation was most common, the range narrowed to 24.5% to 97.7%.
- The widely quoted 93.2% comes from brain donors only, a group selected in part because families already suspected disease.
- Among donors, the most advanced form of CTE was linked to a 44% higher rate of dementia diagnosed from records before death (risk ratio 1.44).
What the study found
Researchers identified every player who appeared in at least one NFL game after 1949, the first year hard plastic helmets were mandatory, and used the National Death Index to obtain personal details and causes of death for everyone in that group who died during the study window. That produced a denominator of 1,712 deaths.
Of those, 338 players (19.7%) had their brains donated and examined, and 315 of them (93.2%) had CTE. Neuropathologists doing the readings were masked to clinical and playing history.
From there the authors set two bounds. The floor assumes every player who did not donate a brain was free of CTE. The ceiling assumes every non-donor had it. For the full 2008 to 2021 period, that gives 18.5% to 98.7%. For 2016 to 2021, it gives 24.5% to 97.7%.
Among the donors themselves, 104 (30.8%) had stage IV CTE, the most advanced form. Study clinicians, working blind to the autopsy results, diagnosed dementia in 202 donors (59.8%), with onset at an average age of 63.4 years and death at 73.1 years. Stage IV CTE was associated with about a 44% higher rate of that dementia diagnosis (risk ratio 1.44, 95% confidence interval 1.16 to 1.78), very likely somewhere between 16% and 78% higher, with less than a 1 in 1,000 chance the link is coincidence.
One number deserves more attention than it is getting: only 40.6% of the donors with a clinician diagnosed dementia (82 people) had any neurodegenerative disease listed as the primary or secondary cause of death on their certificate.
Dr. Kumar’s take
The 93.2% figure leading most of the coverage is an artefact of who ends up in a brain bank. Families donate when something was wrong. That is the whole reason a denominator study like this one was needed, and the number it reports is not 93.2% but a range, 24.5% to 97.7%.
The honest number here is 24.5%, and the important thing about it is that it is a floor, not an estimate. It was produced by counting every single player who did not donate a brain as CTE free, which nobody believes. The real figure sits above it. How far above is unknown, which is what the 97.7% ceiling is telling you: this study nailed down the bottom of the range and left the top wide open.
Two other cautions. The donors died at an average age of 66.5 years, so these brains come from men whose playing careers were long finished, not from anyone currently on a roster. And this is observational work built on death certificates and family interviews, so recall of symptoms is imperfect.
The death certificate finding cuts the other way, though. Fewer than half of donors with documented dementia had a neurodegenerative condition recorded anywhere on their certificate. Any prevalence estimate that leans on death records is undercounting.
What it means for you
If you or someone in your family played contact sports, the practical message is not fatalism. CTE cannot be diagnosed in the living and cannot be treated. But the symptoms people attribute to it, memory trouble, mood swings, poor concentration, are also produced by sleep apnea, uncontrolled blood pressure, and low testosterone, all of which are common after a career of head impacts and all of which respond to treatment. Assuming the worst and skipping the workup is the actual risk.
For everyone else, this is a lesson in reading numbers. When a prevalence figure comes from people who volunteered because they were already sick, it measures the volunteers, not the population. Ask what the denominator was. The habits that protect an aging brain remain the ones with real evidence behind them: controlling blood pressure, treating sleep disorders, staying active, and staying cognitively engaged.