Author(s): SCHAUN, G., RAIDL, P., WURZER, G., STRIZIK, M., MORHOVSKA, A., CSAPO, R., Institution: UNIVERSITY OF VIENNA, Country: AUSTRIA, Abstract-ID: 2450

INTRODUCTION:
Aging is associated with declines in muscle mass and strength, particularly in the lower extremities, highlighting the need for effective resistance training strategies. Velocity-based training, and specifically the use of velocity-loss thresholds (VLTs), has emerged as a promising approach to individualize training load and fatigue management. However, the reliability and practical equivalence of performance outcomes at different VLTs in older adults remain unclear. This study examined the test–retest reliability and equivalence of repetition- and velocity-based outcomes at three VLTs in older adults.
METHODS:
Eleven older adults (67.3 ± 3.5 years; 8 women) completed four sessions: familiarization, one-repetition maximum (1-RM) testing, and two identical experimental trials. In each trial, participants performed one set to failure at 70% 1-RM in the 45° leg press, with movement velocity recorded via a linear position transducer. The repetition number and mean concentric velocity (MCV) corresponding to 10%, 20%, and 40% velocity loss (VL10, VL20, VL40) were determined. Paired-sample equivalence testing (±2 repetitions; ±0.06 m/s) and reliability analyses (ICC2,1; CV; SEM; MDC) were performed.
RESULTS:
Maximal strength was 111.3 ± 50.5 kg. Reliability improved progressively with increasing VLT. At VL10, MCV reached equivalence (p = 0.030), whereas repetitions did not (p = 0.576), with greater variability for repetitions (ICC = 0.33 and 0.42; CV: 51.4% vs. 24.0%). At VL20, neither repetitions (p = 0.736) nor MCV (p = 0.052) met equivalence criteria, and reliability was poor for both outcomes (ICC = 0.18; CV: 44.8% vs. 29.9%). In contrast, VL40 demonstrated equivalence for MCV (p = 0.028) and improved reliability for repetitions and MCV (ICC = 0.81 and 0.78; CV = 14.6% and 22.6%).
CONCLUSION:
Although lower VLTs are commonly recommended to limit fatigue in older adults, the present findings indicate that these conservative thresholds may exhibit greater measurement variability when used for performance monitoring. This increased variability was evident across multiple reliability indices and may be partly amplified by the disproportionate impact of small absolute repetition differences at lower thresholds, which translate into larger relative fluctuations. Consistent with this interpretation, large MDC values for repetitions suggest that only substantial changes (> 5-13 repetitions) exceed measurement error. In contrast, velocity-based outcomes showed greater stability across thresholds, indicating relative robustness to inter-session performance fluctuations. While these findings provide preliminary insight, the modest sample size warrants confirmation in larger samples.