At least 24.5 percent of former NFL players who died between 2016 and 2021 had chronic traumatic encephalopathy at the time of death, according to a peer-reviewed study published in The BMJ, and the researchers are explicit that the figure is a floor rather than a best estimate.
The study, from Mass General Brigham, Boston University, and the Concussion and CTE Foundation, was published in The BMJ and takes an unusual approach to a persistent measurement problem. CTE can only be confirmed by examining brain tissue after death, so almost every prevalence figure ever published has come from donated brains, and families who donate are more likely to have watched a relative struggle with memory, mood, or behavior.
The consequence for readers is a number that is both alarming and genuinely uncertain. The researchers calculated a possible prevalence range of 24.5 percent to 97.7 percent, and they do not consider either end of the range to be the true rate.
The Floor Estimate, and the Reason Researchers Cannot Give a Midpoint
The design starts with a denominator rather than a sample. Researchers built a list of every player who appeared in at least one NFL game since 1949, then obtained death certificates through the CDC to identify 1,712 former players who died between 2008 and 2021.
They narrowed the window to 2016 through 2021, the period when brain donation was most frequent. In that six-year span, 878 former players died. Of those, 235 brains were donated and examined, and 215 of them had confirmed CTE.
The 24.5 percent figure is calculated by dividing the 215 confirmed cases by the 878 deaths. In other words, it assumes every single one of the 643 players whose brains were not examined had no CTE at all. That assumption is almost certainly false, which is precisely why the number functions as a minimum. The upper bound works the same way in reverse: 97.7 percent assumes that every unexamined player had the disease, except for the 20 whose brains were definitively cleared.
"We've known that this disease process can be caused by, in part, repeated traumatic brain injury. But we didn't know how common it was," lead author Dr. Daniel Daneshvar of Mass General Brigham told NPR. He has said the true rate sits somewhere in the middle of the range, and that 24.5 percent is on the low end.
Ninety-Three Percent of Donated Brains Is Not the Number That Matters
The study also reports that across the full 2008 to 2021 period, 315 of 338 donated brains, or 93.2 percent, showed CTE. That figure will travel widely and will be misread.
It does not mean 93 percent of former NFL players have CTE. It describes only the players whose families chose donation, a group selected in part by the presence of symptoms during life. Treating it as a population rate is the single most common error in coverage of brain bank research.
The contrast with the general population is still meaningful. Daneshvar has noted that community brain bank studies find CTE in less than 1 percent of people. Even the conservative floor of roughly one in four sits far above that.
Dr. Jesse Mez, a codirector of clinical research at the Boston University CTE Center, said the minimum prevalence "is higher than expected," and added that because risk appears to rise with the amount of play, higher prevalence might also be expected among college and high school players, according to Boston University.
Dementia Symptoms Tracked Closely with the Most Advanced Disease
The clinical finding underneath the headline number concerns what CTE looked like in life.
Among the 338 donors, 202, or about 60 percent, met a study clinician's criteria for dementia, and 104 had stage IV disease, the most advanced form. More than 90 percent of stage IV samples were also diagnosed with dementia. After adjusting for differences between donors and non-donors, stage IV CTE was associated with a 44 percent higher prevalence of clinician-diagnosed dementia.
The link was much weaker at lower stages. About 40 percent of donors whose brains showed no CTE at all were still diagnosed with dementia, which is a reminder that the symptoms are not specific to this disease. Separately, researchers found that death certificates for players who did not donate their brains listed a neurodegenerative disease as a cause of death in another 215 cases, suggesting the burden extends well past the donated sample.
The BMJ's own summary of the paper stressed that because the analysis relied on post-mortem brain donation, causal interpretation requires caution.
Two limitations deserve equal weight. This is an observational study of people who have already died, so it cannot establish that football caused CTE in any individual, cannot predict any living player's outcome, and does not change any clinical guideline. And the mean age at death among donors was about 66, meaning most of these careers took place before 2005, when CTE was first diagnosed in a former NFL player and before the league changed rules on kickoffs and helmet-to-helmet contact. Outside experts have cautioned that the cohort may not represent the NFL as it is played now.
Football Families Face a Dose Question, Not a Verdict
For parents of youth players, the practical implication is about exposure rather than a decision already made. The apparent relationship between years of play, cumulative head impacts, and risk is why researchers focus on reducing total exposure, including contact in practice, rather than on any diagnostic test. CTE still cannot be diagnosed in a living person.
Former players and their families face a different question, and the more useful one. Conditions common among retired players and tied to prior head injury, including sleep apnea, high blood pressure and low testosterone, can produce symptoms that look like CTE and are treatable. Researchers who study former player health have warned that fatalism about CTE may keep people from raising symptoms with a doctor at all. Memory complaints, mood changes, and impulse control problems in a former contact-sport athlete are worth a medical evaluation rather than acceptance.
The NFL said it continues working to make the game safer, including by reducing concussions and head impacts, and pointed to resources available to former players who are concerned about their health.
MedicalDaily previously reported on an earlier CTE brain bank study that found the disease in 110 of 111 examined NFL brains, a figure that carried the same donation bias this new work was designed to address. The floor estimate is the advance. What remains unknown is the actual midpoint, and answering that will require brain donation from families whose relatives had no symptoms at all.
Key Questions Answered
What did the study find? At least 24.5 percent of the 878 former NFL players who died between 2016 and 2021 had confirmed CTE at death. The researchers calculated a possible range of 24.5% to 97.7%.
Why is the lower number described as a minimum? Because it assumes all 643 players whose brains were not examined had no CTE. That assumption is almost certainly wrong, so the real figure is higher.
Does this mean 93 percent of NFL players get CTE? No. The 93.2 percent figure applies only to donated brains, and families are more likely to donate when a relative had symptoms. It is not a population rate.
Can CTE be diagnosed while someone is alive? No. Confirmation currently requires examination of brain tissue after death, which is the central limitation of all CTE prevalence research.
What symptoms are associated with CTE? Memory loss, impaired thinking, mood changes including anxiety and depression, and problems with impulse control. None of these confirms CTE, and all have other possible causes, several of them treatable.
Should parents pull children out of football? The study does not make that recommendation. Researchers point to reducing total head impacts, since risk appears to rise with cumulative exposure.
Does this change any medical guideline? No. This is observational research on people who have died. It establishes a prevalence floor and does not alter clinical guidance or provide a diagnostic tool.
