Visual reaction time cannot detect expertise in soccer goalkeepers versus controls

INTRODUCTION:
Visual reaction time (VRT) has been shown to be an expertise factor in some sports, compared to controls without prior sports training. [1] At the same time, it seems to play a relevant role in injury prevention, and return-to-play contexts.[2] While VRT evaluation is commonly assessed with computer-based tools, there is a need to develop new methods that better recreate the sport reality. [2] We investigated the reliability of a VRT test in athletes with high reaction time demands (soccer goalkeepers, GK) and a physically active population; and compared the differences between them.
METHODS:
A group of 21 male semi-professional GKs and 19 male students (STUD) practicing sports without a strong VRT component, participated in the study. A subset of 17 GKs and another group of 18 STUD, participated in two different sessions, within a two-week period, for the reliability analysis.
Field VRT Test: Two pairs of two FitLight sensors were placed 160cm apart at 90 & 170 cm height. Participants had to stand in the middle and “deactivate” a series of 12 visual stimuli (3 per sensor) as fast as possible by passing the side-matching hand <30cm in front of the sensor. An additional sensor was placed in front of the participant to pace the two-second pauses between stimuli. The mean of all responses was used for analysis.
Computer-based tests: Simple Response Time Task (SRT); Four Choice Visuospatial Response Task (4Cho); and Go/No-Go Response Inhibition Test (GoNoGo). Mean response time and number of errors were used for analysis.
We used Two-Way Mixed ICC for test-retest reliability; Wilcoxon and T-test for group comparisons (for p <.05 and >.05 Shapiro Tested normality, respectively). Significance was set at p<.05
RESULTS:
The reliability of the proposed VRT test was “moderate” (ICC=.74). GKs were not significantly faster than STUD (657±96 ms vs 680±69 ms, p=.400). GKs were also non-significantly faster on SRT (317±37 ms vs 329±42 ms, p=.531), 4Cho (455±98 ms vs 460±122 ms, p=.925), GoNoGo (438±42 ms vs 440±3 ms, p=.896); and had fewer errors on GoNoGo (9.6±7 vs 11.6±6, p=.273).
CONCLUSION:
Non-significant group differences suggest that VRT is not an expertise factor for GKs. Practitioners interested in performance assessment or talent identification are advised to prioritize the evaluation of other parameters. A multifactor use of VRT for these purposes remains to be explored.