Between-session reliability of dry-land and in-water tests to measure inter-limb asymmetries in swimmers.

ABSTRACT

The aims of the present study were to i) analyse the between-session reliability of dry-land and in-water swimming tests, and ii) investigate the prevalence of meaningful asymmetries in swimming athletes. Twenty-eight swimmers (21 males, 7 females) performed anthropometric, shoulder range of motion (ROM), countermovement jump, shoulder isokinetic torque, and 15-s tethered swimming tests two times, 1 week apart. Inter-limb asymmetries were calculated for each variable. Raw data reliability was determined using the intraclass coefficient correlation (ICC) and the typical error of measurement (TEM), and effect size (ES) was used to determine systematic bias between test sessions. At an individual level, inter-limb asymmetries were compared to the coefficient of variation (CV) to determine whether they were real. The between-session reliability was good to excellent (0.75 to 1.00) for most of the raw data, except for ROM. Between-session ES was predominately "trivial" or "small" for raw data and asymmetries, reinforcing that the values did not change significantly between the sessions. In addition, real asymmetries were seen in some tested metrics, depending on the test. In conclusion, the tested variables presented good levels of between-session reliability and were able to detect real and consistent asymmetries.