Reliability and validity of clinical tests of cardiorespiratory fitness: A systematic review and meta-analysis

IntroductionInsufficient physical activity is a significant contributor to non-communicable disease amongst the global population. Insufficient physical activity is directly linked with reduced cardiorespiratory fitness (CRF). CRF is as strong a predictor of mortality as well-established risk-factors such as smoking, hypertension, dyslipidaemia, and type 2 diabetes mellitus, however, it remains the only major risk factor not routinely assessed in primary health care settings. The aim of this review was to assess the validity and reliability of existing submaximal tests of CRF which can be employed in a standard medical consultation for the estimation of CRF and physical function in adults. MethodsA systematic review of the scientific literature was undertaken to find all studies reporting the reliability and/or validity of submaximal tests of CRF and physical function. Studies published up to 12 January 2023 were included in the search of the Medline, Embase, Cinahl, SPORTdiscus, Cochrane library, Informit Health and Web of Science databases. Risk of bias was assessed using the JBI critical appraisal checklist for analytical cross-sectional studies. Data including reliability of the submaximal protocols as measured by test-retest Pearsons r (r) or Intraclass co-efficient (ICC); and validity as measured by the correlation between the submaximal protocol results and the graded exercise test results (r) was extracted. Meta-analyses were performed to determine the overall mean r of the correlation coefficients. ResultsIn total 1754 studies were identified. Following screening, 143 studies including 15,760 participants were included. All clinical tests included in meta-analysis demonstrated strong reliability. The Siconolfi step test (r=0.81), Incremental shuttle walk test (r=0.768) and 1- minute sit-to-stand test (r=0.65) demonstrated strongest validity following meta-analysis. ConclusionBased on the validity of the tests outlined, these can be used as an acceptable method of estimating VO2peak in a broad population, without the cost and access issues of formal GXT.