Repeated concussions associated with worsened working memory, attention and processing speed, but preserved delayed memory: A PROTECT study

Abstract

AbstractBackgroundApproximately 5% of all dementia diagnoses may be attributable to traumatic brain injuries (TBI), but the precise pathway and required thresholds for increased dementia risk remain unclear. TBI can cause disparate degrees of damage to parts of the brain and thus may affect different subdomains of cognitive function differently. This study examined the strength of associations between traumatic brain injuries and cognitive subdomains in a healthy cohort of older adults.MethodCognitive and demographic data was collected from individuals aged 50 ‐ 93 (n = 12,041) participating in the PROTECT study, a large online cohort study. Self reported, cross sectional data on head injury were collected through the Brain Injury Screening Questionnaire (BISQ). Cognitive tests were standardised and grouped into four domain scores using the FLAME composite (Working Memory, Delayed Memory, Processing Speed and Attention) based on a principle components analysis (KMO test = 0.724). Regression analyses were performed, examining the association between numbers of head injuries, episodes of loss of consciousness and the cognitive domain scores.Result62.8% of participants reported a previous head injury and 33% of those had at least one episode of subsequent loss of consciousness (21% overall). There was a significant association between the self‐reported historical number of head injuries and poorer working memory, attention and processing speed (B = ‐0.005, ‐0.007, ‐0.014 and p=0.004, <0.001, <0.001 respectively). Self‐reported historical episodes of head injuries followed by loss of consciousness (LOC) demonstrated similar, albeit stronger, associations with working memory, attention and processing speed (B = ‐0.019, ‐0.024, ‐0.024 and p = 0.002, <0.001, <0.001). Neither numbers of head injuries or LOC were associated with poorer delayed memory performance.ConclusionThe PROTECT study is one of the largest, healthy older adult cohorts to examine the link between TBI and cognitive outcomes. Although it is cross sectional the increasing effect sizes seen for increasing severity of head injury indicates a dose‐response relationship suggestive of causation. Our results suggest that TBI may differentially impact particular brain areas and functions and that cognitive impairment in those with a history of TBI has a distinct pathological and clinical phenotype.