Contribution
A state-of-the-science review for a special issue of the Journal of Athletic Training. It sorts the evidence on repetitive concussion and repetitive head-impact exposure into short-range effects, neuropsychiatric effects, cognitive effects and long-range neurodegeneration, and finds that the short term is well characterized while the long term is not studied by any design able to answer it. It ends with four research questions and a cohort design.
Key insight
The question moved in a decade from “how many concussions?” to “how much head impact?” and “what happens down the road?”, and the evidence did not move with it. What exists on the long term is retrospective: surveys of retired players, brain-bank autopsies, case series. None of it can separate exposure from selection, recall and comorbidity.
Assumptions
- The scope is sport-related concussion in contact sports, mostly American football and ice hockey, high-school to professional.
- The authors take “subconcussive” to be a poor term: it assumes every concussion is diagnosed, that the threshold is known, and that such impacts injure the brain.
- It is a narrative review. No search strategy, inclusion criteria or pooled estimates are given.
Key results
Figures are as the review reports them from its sources.
- Recurrence. Castile et al. (2012): 3.1 recurrent versus 22.2 new concussions per 100,000 athlete-exposures in high school, with recurrent ones worse acutely (more loss of consciousness, longer recovery).
- Exposure. About 520 (Schnebel, 16 players) and 652 (Broglio, 35 players; maximum 2,235) head impacts per high-school football season. Crisco (314 college players, 286,636 impacts): median per season 257–438 by team, peak linear acceleration 20.5 g at the 50th and 62.7 g at the 95th percentile, rotational 1,400 and 4,378 rad/s². Ice hockey: 287 per season for men, 170 for women.
- Short range. “The overwhelming majority of athletes achieve a complete recovery in self-reported symptoms and performance-based measures of recovery … by 1 to 2 weeks”. Season-long exposure studies (Broglio, Gysland, Miller, McAllister et al.) mostly find no cognitive effect of impact counts. Exception: 24% of contact athletes against 3.6% of non-contact athletes scored below prediction on the CVLT learning trials. Diffusion imaging (McAllister, 80 contact versus 79 non-contact athletes) found white-matter changes correlated with exposure and, for corpus-callosum MD, with verbal learning.
- Neuropsychiatric. Depressive symptoms rise after concussion, as they do after orthopaedic injury, hence the call for injured controls. Longer-term links (3.3-fold depression risk in a child survey; suicide rates 3–4 times the population’s after hospital-treated concussion in Ontario, with a weekend effect that suggests confounding) are “neither substantive nor mature enough” for conclusions.
- Cognitive. A meta-analysis of 8 studies found no overall effect of multiple mild TBIs; other studies disagree in method and result.
- Long range. Guskiewicz (2005): retired NFL players with three or more reported concussions had 5 times the prevalence of diagnosed MCI and 3 times that of memory complaints. CTE is a postmortem diagnosis. Bieniek: 21 of 66 contact-sport brains met CTE criteria against 0 of 198. Ling: 12% of 268 brains, with the same rate in elderly controls. Iverson et al. are quoted that the science “does not allow us to determine the extent to which repetitive neurotrauma uniquely causes, or partially contributes to” specific symptoms.
- The gap. “To date, no prospective, longitudinal studies of well-defined cohorts over longer periods … have been conducted”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Most athletes recover in symptoms and test performance within 1–2 weeks of a sport concussion | strong | several prospective studies, cited |
| C2 | Season-long head-impact exposure shows little measurable short-term cognitive effect, with a possible vulnerable subgroup | moderate | several small cohort studies, mixed |
| C3 | No prospective longitudinal study had measured long-term risk from repeated concussion or head impacts | strong | the authors’ survey of the literature, as of writing |
| C4 | Current evidence cannot say whether CTE pathology causes the clinical syndrome attributed to it, or whether it is a distinct disease | moderate | brain-bank and case-series studies, critical reviews, cited |
Concepts
- Sport-related concussion (SRC). A diagnosed mild TBI in sport.
- Head-impact exposure. Number and magnitude of head impacts, measured by helmet accelerometers, with or without diagnosed concussion.
- Chronic traumatic encephalopathy (CTE). A tauopathy with perivascular phosphorylated τ at the depths of cortical sulci, diagnosed only at autopsy; traumatic encephalopathy syndrome is the proposed clinical counterpart.
- Trauma controls. Patients with non-head injuries, used to separate the effects of injury in general from those of brain injury.
Connections
The review writes as part of the CARE consortium’s programme (named through www.careconsortium.net), which is the prospective design it calls for, and leans on Iverson et al.’s and Belanger et al.’s critical reviews. Nothing else in the record reads concussion.
Bearing on the record
- No THEORY, QUESTION or CLAIM cites this paper, and none should rest on it beyond its epistemic conclusion: as of its writing the long-term risk was unmeasured by any design able to measure it.
- The seed’s summary is right, with one sharpening: its “signals of vulnerability in subgroups (e.g., verbal memory)” are one finding (the CVLT 24% against 3.6%), not several.
- Dated by design. Published in March 2017, it predates the CARE consortium’s results that it anticipates. A reader wanting the current state needs those, not this.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- A narrative review by two authors who also wrote several of the studies it reviews (McAllister et al., refs 16 and 17).
- Its figures are second-hand; none was checked against the primary studies here.
- It does not weigh studies by quality or pool them.
Open questions
- Does repeated concussion or head-impact exposure raise late-life risk of cognitive decline or neurodegeneration, and for whom? The review’s answer is a prospective cohort with concussed, exposed-but-unconcussed and unexposed athletes, baseline characterization and long follow-up.