Related Experiment Video
Updated: Sep 25, 2026
Vision Training Methods for Sports Concussion Mitigation and Management
Published on: May 5, 2015
Duration-dependent acute effects of foam rolling on dominant-hand visual reaction time in male youth basketball
Koray Samet Altınçekiç1, Şevval Çelen2, Osman Coşkun3
1Department of Coaching Education, Faculty of Sports Sciences, Istanbul Gedik University, Istanbul, Türkiye.
Objective:
This study examined the acute effects of different foam rolling (FR) durations on dominant-hand visual reaction time in male youth basketball players.
Methods:
In this parallel-group randomized trial, 102 U16-U18 players were allocated equally to FR-A (30 seconds per region), FR-B (60 seconds per region), or passive rest. Five valid trials were recorded before and after the intervention. The lowest value was the primary best-trial score; the mean and median were sensitivity measures. Baseline-adjusted post-test differences were assessed using ANCOVA. Because the standard best-trial ANCOVA violated slope-homogeneity and linearity assumptions, an extended nonlinear baseline-interaction model was evaluated using 5,000 bias-corrected and accelerated bootstrap replicates. Mixed-design ANOVAs examined pre-to-post trajectories.
Results:
FR-A had lower adjusted post-test reaction time than FR-B under best-trial (difference = -21.86 ms, 95% CI: -30.31 to -13.41; adjusted p < .01), mean-based (-12.36 ms, adjusted p < .01), and median-based scoring (-10.94 ms, adjusted p = .03). The FR-A-control difference was significant in the standard best-trial ANCOVA (-15.44 ms, 95% CI: -24.12 to -6.76; adjusted p < .01) and extended bootstrap comparison (-14.79 ms, adjusted p < .01), but not in mean-based (adjusted p = .33) or median-based analyses (adjusted p = .39). Extended bootstrap estimates were also higher in FR-B than in the control group (9.52 ms, adjusted p = .01). Because this model contained baseline interactions, these contrasts represented average rather than constant differences across baseline values. The time × group interaction was significant only for best-trial scoring (p < .01).
Conclusion:
FR-A had consistently lower adjusted reaction time than FR-B across all scoring methods. However, the FR-A-control comparison depended on baseline reaction time and scoring method. Thus, the findings do not establish a scoring-independent benefit of the 30-second protocol over passive rest or support a firm practical recommendation.
