Athletes and Experimental Pain: A systematic review and meta-analysis

The evidence that athletes respond to and report indices of experimental pain differently to non-athlete populations was analysed. Databases screened were SPORTDiscus, PubMED, PsycArticles, the Cochrane Library (Cochrane Database of Systematic Reviews), Web of Science, Scopus and CINAHL. Studies that compared experimentally induced pain responses (threshold, tolerance, intensity, unpleasantness, bothersomeness and effect on performance) in athletes and controls were included. Meta-analyses were performed where appropriate and effects were described as standardised mean differences, pooled using random effects models. Thirty-six studies (2492 participants) met the inclusion criteria comprising 19 pain tolerance, 17 pain threshold, 21 pain intensity, five pain unpleasantness, two performance in pain and one bothersomeness study. Athletes demonstrated greater pain tolerance (g = .88 [95% confidence interval [CI] .65, .13]) and reported less pain intensity (g = −.80, [95% CI −1.13, −.47]) compared to controls; they also had higher pain threshold but with smaller effects (g = .41, [95% CI .08, .75]). Differences for unpleasantness did not reach statistical significance but the effects were large (g = −1.23 [95% CI −2.29, .18]). Two studies reported that performance in pain was better in contact athletes than non-athletes, and one concluded that athletes find pain less bothersome than controls. There were considerable inconsistencies in the methods employed that were reflected in the meta-analyses’ findings. Sub-group analyses of tolerance and intensity were conducted between endurance, contact, and other athlete groups, but were not significant. The data suggest that athletic participation is associated with altered pain responses, but mechanisms remain unclear and more transparent methods are recommended.This study was registered on the PROSPERO site in January 2019 (ref ID: CRD42019119611).