The concurrent validity and test-retest reliability of a submaximal exercise test in adolescents with autism.

Abstract

There is a need for valid and reliable clinical assessment tools to assess cardiorespiratory fitness (CRF) levels in adolescents with autism. Therefore, this study aimed to examine the concurrent validity and test-retest reliability of the Astrand-Rhyming Test (ART) in this population. 45 adolescents with autism aged 12-18 years (<i>n</i> = 32 males, 14.47 ± 1.79 years) performed the ART twice (test-retest reliability) and completed a maximal cardiopulmonary exercise test (CPET) (concurrent validity). Reliability parameters included Pearson correlations, intraclass correlation coefficients (ICCs), standard error of measurements (SEM), minimal detectable changes (MDC), coefficients of variation, paired sample t-tests, linear regressions and Bland-Altman plots. The concurrent validity was evaluated with Pearson correlations, ICCs, paired sample t-tests, linear regressions and Bland-Altman plots. Strong test-retest reliability (<i>r</i> = 0.84-0.85, ICC = 0.84-0.85) was found for the ART, but the wide limits of agreement intervals suggest the presence of substantial variability. The large SEM (4.73-5.08 mL/kg/min) and MDC (13.20-14.07 mL/kg/min) values suggest lower absolute reliability. Moderate to strong levels of association (<i>r</i> = 0.74-0.75) and agreement (ICC = 0.59-0.66) were found between estimated (ART1) and measured (CPET) VO₂ max levels, but significant systematic differences (5.71-8.82 mL/kg/min) were observed. The ART is an accessible and promising method to monitor submaximal CRF levels over time but is less appropriate to estimate maximal CRF levels in this population.