Reliability of maximal strength testing in older adults

Objectives: To determine (1) the reliability of a maximal strength test (1 repetition maximum) [1-RM] in older adults and (2) the impact of differing periods of familiarization. Design: Within-subject, repeated trials of maximal strength. Setting: Community-based senior center. Participants: Forty-seven independently living men (n=16) and women (n=31), with a mean age of 75.4+/-4.7 years. Interventions: None. Main Outcome Measures: Systematic error (shift in mean) and random error (% coefficient of variation [%CV]) was assessed between consecutive pairs of 1-RM trials. Results: For the bench press, systematic error was virtually eliminated for men between trials 2 and 3 (0.7%; 95% confidence interval [CI], -2.7% to 4.3%). The CV was stable (4.7%-7.3%) across all trials in both genders. For the leg press, a significant but clinically small systematic error (3.6%, P<.05: 95% CI, 0.8-6.6) was evident for women between trials 2 to 3. The CV was reduced across trial pairs by 3.3% for men and 0.9% for women. Three versus 6 or more sessions of familiarization produced small clinical differences in systematic error (less than or equal to4.1 %) and CV (less than or equal to0.2%) between trials 2 and 3 for both lifts. Conclusions: Reliability is an indispensable requirement for valid test outcomes. Our results show that, in this group of older adults, 3 familiarization sessions and 2 to 3 test trials produced highly reliable 1-RM measures. Additional periods of familiarization added little to test reliability. Effective reliability testing for 1-RM is a practical and attainable goal for outcomes based practitioners.