# Prevalence of chronic traumatic encephalopathy at death in National Football League players: retrospective population based cohort study, 2008-21

- **DOI:** [10.1136/bmj-2026-100418](https://doi.org/10.1136/bmj-2026-100418)
- **Rating:** Sound (3.3 out of 5)
- **Verdict:** Enumerated NFL deaths establish a 24.5% minimum CTE prevalence in 2016–21, but symptom-driven brain donation limits interpretation of the dementia association.
- **Framework version:** v5.13
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-11-19
- **Evaluation page:** https://www.nabu.science/paper/10.1136%2Fbmj-2026-100418

## Key claims

1. At minimum, nearly a quarter of all former NFL players who died during 2016-21 had CTE neuropathology at death.
   - Support: Well-supported
   - Type: Primary finding
   - Basis: 215 confirmed cases among 878 enumerated deaths establish the stated lower bound without assuming representative donation
   - Location: Abstract, Conclusions
2. In [NFL player brain] donors, stage IV CTE was associated with study clinician diagnosed dementia (risk ratio 1.44, 95% confidence interval 1.16 to 1.78; P<0.001).
   - Support: Tentative
   - Type: Secondary
   - Basis: symptom-driven donation, retrospective dementia assessment, and unadjusted co-pathology leave material uncertainty despite donor-restricted wording
   - Location: Abstract, Results; Table 3
3. Only 40.6% of [NFL player brain] donors (n=82) with study clinician diagnosed dementia had neurodegenerative disease listed as the primary or secondary cause of death [on death certificates].
   - Support: Well-supported
   - Type: Secondary
   - Basis: direct comparison of 82 death-certificate records with 202 clinically adjudicated donor dementia cases
   - Location: Abstract, Results; Table 1
4. NFL players have a higher prevalence of CTE at death than has been previously shown in multiple community based studies.
   - Support: Tentative
   - Type: Secondary
   - Basis: selected community-donor proportions do not identify the source-population prevalence needed for this population comparison
   - Location: Conclusions

## Strengths

- **Enumerated deaths anchor a defensible minimum** The 215 confirmed CTE cases among 878 NFL deaths in 2016–21 establish a 24.5% lower bound without requiring brain donors to represent all decedents.
  - Dimension: contribution
  - Location: Results, Extrapolating the prevalence of CTE at death
- **Masked assessments and alternative analytical specifications** Neuropathologists were masked to clinical histories, and clinicians adjudicated dementia without neuropathological results. Table 3 compares weighted, age-adjusted, and unadjusted specifications.
  - Dimension: methodological rigour
  - Location: Methods, Collection of clinical data and Neuropathological assessment; Table 3
- **Death certificates compared with clinical adjudication** The donor-level comparison shows that only 82 of 202 clinically adjudicated dementia cases had neurodegenerative disease recorded as a primary or secondary cause of death.
  - Dimension: contribution
  - Location: Abstract, Results; Table 1

## Limitations

- **Donor association remains vulnerable to selection** Weighting cannot account directly for unobserved symptoms in non-donors, and the dementia regressions do not adjust for the co-occurring neurodegenerative pathologies concentrated in stage IV donors. These limitations leave the association's magnitude uncertain.
  - Dimension: methodological rigour
  - Location: Methods, Statistical analysis; Tables 2–3; Discussion, Strengths and limitations
- **Population comparison exceeds the available evidence** The Conclusions compare NFL prevalence with community studies whose reported figures are selected donor proportions. That contrast does not establish the corresponding difference between source-population prevalences.
  - Dimension: reporting
  - Location: Introduction; Conclusions
- **Recall limitation is acknowledged then dismissed** The limitations section explains how differential informant sensitivity could inflate the association, then argues that masking excludes recall as an explanation. Masking to the final diagnosis does not eliminate the differential suspicion and symptom ascertainment already described.
  - Dimension: positioning
  - Location: Discussion, Strengths and limitations, informant recall paragraph

## Reliability checks

Caveats (4 of 4 checks).

- **Results check:** Noted
  - "analyses without inverse probability weighting examining the relation between dementia and stage IV CTE (1.85, 1.77 to 1.94; P<0.001)" (Results, Dementia and clinical impairment underlying stage IV CTE among donors, unweighted RR) vs "94 of 104 (90.4%) of those with stage IV CTE" (Results, Dementia and clinical impairment underlying stage IV CTE among donors). Ignoring missing outcomes and imputation, 94/104 versus 108/234 gives RR approximately 1.96 and a conventional 95% interval approximately 1.68–2.28, so the fitted imputed interval cannot be reconstructed exactly. The effect on the result is unclear.
  - Minor reporting notes (5):
    - "the minimum prevalence of CTE at death in NFL players during the study period was 16.6%." (Results, Analyses without debut year restriction) vs "110 (31.7%) had stage IV CTE, and 23 (6.6%) did not have CTE." (Results, Analyses without debut year restriction). Subtracting the 23 non-CTE donors from 347 donors gives 324 cases, and 324/2079 = 15.6%, not 16.6%. If corrected, the size of the result changes materially.
    - "among all 1712 NFL players who died, the possible CTE prevalence at death ranged between 18.5% and 98.7%." (Abstract, Results) vs "CTE was the most common disease in this cohort (93.5%), often in the absence of other neurodegenerative disease findings (58.9%)." (Discussion, paragraph 1). The reported 315 cases give 315/1712 = 18.4% and 315/338 = 93.2%, whereas the printed 18.5% and 93.5% correspond to 316 cases after rounding. The result itself is unaffected.
    - "those with stage IV CTE were more likely to serve in the military (37.3% v 150.0%; P<0.001)" (Results, Analyses without debut year restriction, military service). The unrestricted cohort's 76 military donors minus approximately 41 stage IV military donors leaves 35/237 = 14.8%, while 150.0% is impossible. The result itself is unaffected.
    - "Of the 208 donors with CTE and no other neurodegenerative disease, 107 (53.0%) had dementia." (Results, Dementia and clinical impairment underlying stage IV CTE among donors) vs "CTE was often the only neurodegenerative disease found in donors with dementia (53.0%)" (Discussion, paragraph 2). The text's 107/208 = 51.4%, not 53.0%, and its counts conflict with Table 2's 199 isolated-CTE donors and 100 with dementia. The effect on the result is unclear.
    - "Not diagnosed(n=123) | Diagnosed(n=202) | All donors(n=338)" (Table 1, dementia-status column headings) vs "Clinical information, including dementia diagnosis, not available for nine donors." (Table 1, footnote). The displayed dementia-status groups total 123+202 = 325, leaving 13 of 338 donors rather than the nine stated in the footnote. The effect on the result is unclear.
- **Reporting and transparency:** Noted. The Discussion refers to an IPW prevalence point estimate not presented in the Results. Ethics approval, competing interests, and request-based data access are declared.
  - Conflict of interest: declared · Ethics approval: declared · Data availability: declared
- **Reference integrity:** Clean
  - 57 references in manuscript
  - 56 of 56 checkable references found in an index
  - 1 have no canonical index record — counted, but not index-checkable
  - 1 mistyped DOI matched to the cited work by its details
  - 1 reference confirmed by manual review
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Clean
- **Summary:** No retraction or expression of concern found in Crossref (includes Retraction Watch).
- **Last checked:** 2026-11-19
