JOURNAL OF AGING AND HEALTH / April 2005 Wolinsky et al. / REPRODUCIBILITY OF ASSESSMENTS Reproducibility of Physical Performance and Physiologic Assessments

Objective: We evaluate the test-retest stability of physical performance and physio-logic assessments used in epidemiologic research. Method: Eighty subjects aged 50 to 65 were randomly selected from a probability sample of African Americans for test-retest assessments 5 to 45 days after baseline. Physical performance assessments included grip strength, chair stands, gait speed, and four standing-balance measures. Physiologic assessments included systolic and diastolic blood pressure, height, weight, body fat, and peak expiratory flow. Results: Intraclass correlations coeffi-cients (ICCs) were.81 for grip strength,.72 for chair stands,.56 for gait speed,.60 for one-leg stand,.52 for semitandem stand,.58 for tandem stand with eyes closed, and.27 for tandem stand with eyes open. Except for blood pressure (ICCs of.51 and.55 for systolic and diastolic), the physiologic assessments had ICCs>.89. Discussion: Additional interviewer training may improve the reproducibility of the tandem stand with eyes open