Leveraging American football helmet accelerometer data to estimate associations between cumulative repetitive head impact exposure and chronic traumatic encephalopathy

Abstract

AbstractBackgroundChronic traumatic encephalopathy (CTE) is a neurodegenerative tauopathy associated with repetitive head impact (RHI) exposure. We previously showed that duration of American football play is associated with risk and severity of CTE pathology. Helmet accelerometers have been used previously to examine frequency, linear acceleration, and rotational acceleration of hits sustained across youth, high school, and college football. Here we projected this data onto former American football playing brain donors to examine the relationship between cumulative frequency of hits, linear acceleration, and rotational acceleration during players’ athletic careers and CTE pathology.Method656 former American football playing brain donors from the Veterans Affairs‐Boston University‐Concussion Legacy Foundation Brain Bank were examined for CTE pathology and severity. Years and position of play at each level (youth, high school, college, professional) were acquired through retrospective clinical interviews with brain donors next‐of‐kin. From a literature search of studies using helmet accelerometers, we calculated the mean frequency of hits, linear acceleration (g‐force), and rotational acceleration (rad/sec2) for one year of play at each level‐position combination. These values were projected onto brain donors’ career exposure to derive cumulative frequency of head impacts (CHII), cumulative linear acceleration (CHII‐G), and cumulative rotational acceleration (CHII‐R). Separate logistic regression models examined the relationship of years of play, CHII, CHII‐G, and CHII‐R on CTE status and severity (low vs. high), adjusting for age at death.ResultThe mean (SD) age at death was 59.7 (20.1) and 451 (68.8%) had CTE pathology. Each cumulative measure was significantly associated with presence of CTE (p’s<.001) and severity of CTE (p’s<.001). Based on ROC analyses, CHII‐R (AUC=0.765, p<.001) and CHII‐G (AUC= 0.758, p<.001) performed significantly better than years of play (AUC=0.716) in classifying CTE pathology, and there was no difference between years of play and CHII (AUC=0.698, p=0.25). Similar relationships were observed for CTE severity. Model fit and cross‐validation statistics were also consistent (Table 2).ConclusionAmong former American football playing brain donors, cumulative linear and rotational acceleration were better predictors of CTE pathology than duration of play or cumulative hits. The findings suggest head impact intensity is an important factor in developing CTE pathology.