Pre-injury migraines might be a risk factor for prolonged recovery after sport-related concussion (SRC). We sought to examine whether a pre-injury history of migraines is associated with worse recovery following SRC in collegiate athletes. Data were collected through a prospective concussion surveillance system in 11 National Collegiate Athletic Association (NCAA) Division III college athletic programs between September 2014 and March 2020. Our primary independent variable, pre-injury migraines, were self-reported by the athletes. Between those with and without migraines, the outcomes of days to return-to-learn (RTL) without academic accommodations and return-to-play (RTP) were compared using Mann-Whitney U tests. Each athlete's RTL and RTP status was dichotomized (i.e., returned vs. not returned) at various time points for RTL (i.e., 7/14/21/28 days) and RTP (i.e., 14/21/28/56 days). Chi-squared tests were performed to compare the proportions of RTL and RTP status between groups. Multivariable regressions analyzed potential predictors of RTL and RTP adjusting for age, gender, prior concussions, other health conditions, and symptom severity. Of 1409 athletes with an SRC, 111 (7.9%) had a pre-injury history of migraines. Compared with those without migraines, those with migraines had longer median (interquartile range [IQR]) days to RTL (migraines = 7.0 [3.0-12.3] vs. no migraines = 5.0 [2.0-10.0], U = 53,590.5, p = 0.022). No differences were found in RTP between the two groups (migraines = 16.0 [10.0-33.0] vs. nχo migraines 15.0 [11.0-23.0], U = 38,545.0, p = 0.408). Regarding RTL, significantly lower proportions of athletes in the migraine group had fully RTL, without accommodations, at ≤14 days (77.5% vs. 85.2%, χ2 = 4.33, p = 0.037), ≤21 days (85.3% vs. 93.0%, χ2 = 7.99, p = 0.005), and ≤28 days (88.2% vs. 95.6%, χ2 = 10.60, p = 0.001). Regarding RTP, a significantly lower proportion of athletes in the migraine group RTP at ≤28 days (72.0% vs. 82.7%, χ2 = 5.40, p = 0.020) and ≤56 days (84.0% vs. 93.0%, χ2 = 8.19, p = 0.004). In a multivariable model predicting RTL that was adjusted for age, gender, acute concussion symptoms, and other health variables (e.g., attention-deficit/hyperactivity disorder [ADHD], history of mental health difficulties), pre-injury history of migraine was associated with longer RTL (β = 0.06, p = 0.030). In a multivariable model predicting RTP, pre-injury history of migraine was not associated with RTP (β = 0.04, p = 0.192). In collegiate athletes, pre-injury migraine history was independently associated with longer RTL but not RTP. When comparing the proportions of those with successful RTP by days, significantly lower proportions of those with migraines showed successful RTP at ≤28 days and ≤56 days. Futures studies should study the generalizability of our findings in other school levels.
Read full abstract