Feasibility, criterion validity and retest reliability of exercise testing using the Astrand-rhyming test protocol with an adaptive ergometer in stroke patients

Abstract

Purpose: (i) to establish feasibility of the Astrand-Rhyming submaximal protocol using the adaptive ergometer in healthy participants, (ii) to test feasibility of this protocol in stroke patients, (iii) to identify clinical characteristics of stroke patients which limit testing, (iv) to establish criterion validity of V·O(2 steady state) ergometry data against the gold standard open-circuit spirometry measure and, lastly, (v) to test-retest reliability of the data generated by the test protocol.

Methods: A descriptive, cross sectional study design was implemented. Ten healthy participants (5 men, mean age 32 ± 6.6 y) and 20 chronic stroke patients (14 men, mean age 61.0 ± 11.6 y; mean months since stroke, 65.6 ± 34.4 y) took part in this study. Independent t-tests, Mann Whitney U tests and chi-square test identified characteristics of those who failed to successfully complete the protocol. Interclass correlation coefficients (ICCs) tested criterion validity and test-retest reliability.

Results: No adverse events occurred. All healthy participants (100%) complied with the test protocol and estimation of V·O(2max). 90% of stroke patients completed the test, with V·O(2max) estimated in 35% (7 / 20) stroke patients. Age and beta-blocker medication demonstrated statistical significance as barriers to the test protocol at p = 0.04 and p = 0.03, respectively. Criterion validity established by comparison with breath-by-breath gas analysis was excellent (ICC 0.93). Retest reliability of V·O(2max), steady-state V·O(2) and oxygen pulse had ICCs of 0.94, 0.96 and 0.97, respectively.

Conclusions: The Astrand-Rhyming test protocol and MOTOmed Viva2 adaptive ergometer were feasible in healthy participants. The protocol was not feasible for V·O(2max) estimation for the majority of stroke patients based on heart rate response. The adapted steady state VO2 measures used however were feasible and had strong relationship to actual VO2 consumption. Criterion validity and retest reliability of test data were excellent.