The Gastrointestinal Symptom Rating Scale has a Good Test-Retest Reliability in Well-Trained Athletes With and Without Previously Self-Identified Gastrointestinal Complaints.

Abstract

Background: Athletes often report gastrointestinal (GI) complaints. Standardized validated tests validated in athletes are lacking.

Objective: The objective of the current study was to investigate the test-retest reliability of the gastrointestinal symptoms rating scale (GSRS), a disease-specific instrument of 15 items to quantify the severity of various GI symptoms.

Methods: For this purpose, a 3-week repeated measurements design was used. The mean difference (Wilcoxon signed rank test), associations (Spearman correlations), and systematic difference using Bland-Altman calculations for repeated measurements, as well as its internal consistency (Cronbach's alpha) on testing day 1 and day 2 were analyzed, with significance set at p ≤ 0.05. A total of n = 70 well-trained athletes (26.1 ± 5.4 years, of which 40% were female) were included.

Results: A high Cronbach's α for GSRS was found on testing day 1 (0.825), and day 2 (0.823), suggesting a good and comparable internal consistency of the questionnaire. When assessing the multilevel temporal stability for total GSRS scores (28.0, IQR 22.0-36.3 vs 26.5, IQR 18.0-35.0), there was a small but significant difference (Z =  - 2.489, and p = 0.013), but a fair correlation between day scores (r = 0.68, p < 0.001), and a Bland-Altman reporting difference between questionnaires within 10% of the total GSRS score, without significant reporting bias (p = 0.38). In most cases, except for hunger, burping, and loose stools, the individual GSRS items were reported in line with total scores and similar for sex.

Conclusion: In conclusion, the GSRS is reliable when used with athletes, with good internal consistency for most symptoms independently of sex, except for hunger, burping, and loose stools.