Objective To assess long-term neurocognitive function changes in former collegiate American football players in early midlife and examine associations between history of head trauma and changes in neurocognitive function.

Design Longitudinal-cohort.

Setting University laboratory.

Participants Former collegiate American football players (n=31; aged=38.4±1.3 years at follow-up; 18.2±0.8 years between assessments).

Independent Variables Self-reported concussion history (SR-CHx) and Head Impact Exposure Estimate (HIEE).

Outcome Measures Neurocognitive test performance at two timepoints (during college [T1]; long-term follow-up [T2]) and T1-to-T2 change scores (T2-T1): Hopkins Verbal Learning Test-Revised Immediate and Delayed Recall; Verbal fluency (FAS); Symbol-Digit Modalities Test; and Trail-Making Test-B (TMTB). Paired-samples t-tests assessed performance differences between timepoints. Coefficients of variation (CoV) were computed for change scores. Multivariable linear regression models were fit to test associations of SR-CHx and HIEE with neurocognitive test change scores. Standardized beta-values (β) and p-values were computed.

Main Results FAS (mean difference[sd]=4.36[11.5]; p=0.043; Cohen’s d[95%CI]=0.379[0.011–0.741]) and TMTB (mean difference[sd]=-8.16[19.9]; p=0.030; Cohen’s d[95%CI]=0.410[0.040,0.774]) performance improved between timepoints. CoV-values (243%-560%) indicated high variability in individual change scores. Neither SR-CHx (|β-values|=0.048 to 0.273; p-values>0.14) nor HIEE (|β-values|=0.027 to 0.281; p-values>0.16) were significantly associated with raw-score changes for any neurocognitive test.

Conclusions Performance improved across 18-years on measures of verbal fluency and executive function, with high inter-individual variability across all measures. Neurocognitive performance changes were not associated with head trauma history. Further prospective investigations should evaluate factors associated with the course of neurocognitive changes through mid-to-late life.