HJAR Sep/Oct 2023

HEALTHCARE JOURNAL OF ARKANSAS I  SEP / OCT 2023 33 head impacts and CTE pathology exists, but that measurement error obscured this relationship. However, these difficulties in diagnosing concussions are neither new, nor resolved, so the present study demon- strates a potential means of estimating head impact burden that is robust to these issues. This work utilized a PEM to better char- acterize which components of RHI expo- sure from football play may be implicated in CTE pathology. Models incorporating impact intensity, in addition to duration of exposure and number of impacts, improved model fit and prediction of CTE pathology compared with models without impact intensity. These findings provide insight into the pathogenesis of CTE, implicate potential targets for CTE interventions, and indicate that future studies may benefit from using PEM-derived measures of RHI exposure. METHODS Brain donor recruitment All participants were former football players from the Understanding Neuro- logic Injury and Traumatic Encephalopa- thy (UNITE) and Framingham Heart Study (FHS) Brain Banks (Fig. 3), which have been previously described5. Donors to the UNITE Brain Bank must have had a his- tory of repetitive head impact (RHI) expo- sure through contact and collision sports (CCS), military service, or domestic violence. Recruitment to the UNITE Brain Bank began in 2008 and is ongoing; the present study included donors through 2020. Since track- ing began in 2014, recruitment occurred in the following ways. Next-of-kin contact- ing the brain bank near or at the time of death was the most common recruitment method (n = 433). Other forms of recruit- ment included individuals joining the CLF Brain Donation Registry during life (n = 18), a medical examiner contacting the brain bank (n = 33), a CLF representative contact- ing the next-of-kin after the time of death (n = 19), and a consult request (n = 1). No sex or gender analyses were carried out as all football players in this cohort were male. FHS is a prospective surveillance study tracking incident cardiovascular disease, stroke, and dementia. The study was estab- lished in 1948, enrolling a representative group from Framingham, Massachusetts and subsequently their children (Gen 2) and grandchildren (Gen 3), as well as ethnically diverse cohorts (Omni 1&2) to better capture changing demographics of Framingham. The FHS Brain Bank contains a voluntary subset of these cohorts, and began recruit- ing in 1997. Because the FHS is a commu- nity-based sample, players tended to be older and to have fewer years of American football play. Their addition resulted inmore total players without CTE and in a similar distribution of age-at-death between play- ers with and without CTE. For both brain banks, prospective donors were excluded if there was a postmortem interval for donation greater than 72 hours, or if fragments or less than a hemisphere of tissue were received. Consent from par- ticipant’s next-of-kin was required to par- ticipate. Institutional review board approval was obtained through Boston University Medical Campus and Bedford VAHospital. Clinical data For the UNITE Brain Bank, previously detailed methods for retrospective clinical data collection and comprehensive review of all relevant medical records were fol- lowed for all participants. Astructured clini- cal history was obtained from informants, including a timeline of cognitive, behavioral, mood and motor symptomatology. Clini- cians with expertise in neurodegenerative disease reviewed all cases to reach con- sensus on a dementia diagnosis based on DSM-IV-TR criteria. Clinicians and clinical research assistants were blinded to the neu- ropathological examination and findings 59 . For the FHS brain bank, clinical data were obtained prospectively as part of the FHS clinical assessment. Participants underwent cognitive assessment and those with sus- pected cognitive impairment were brought to a consensus meeting, during which it was Fig. 2 | Performance of Exposure Measures as Classifiers of CTE Pathology. ROC Curves for Exposure Measures as Predictors of CTE Status (A) and CTE Severity (B). Source data are provided as a Source Data file. AUC area under the ROC curve, CHII cumulative head impact index representing estimated number of head impacts per donor, CHII-G cumulative head impact index representing estimated cumulative g-force experienced by each donor, CHII-R cumulative head impact index estimated cumulative rotational force experienced by each donor, CTE chronic traumatic encephalopathy, ROC receiver operating characteristics.