First Systematic Review and Meta-analysis of the Validity and Test-Retest Reliability of Physical Activity Monitors for Estimating Energy Expenditure During Walking in Individuals With Stroke

To evaluate the validity and test-retest reliability of physical activity trackers (accelerometer, multisensor, smartphone, pedometer) for estimating energy expenditure during walking in individuals with stroke. Webline, MEDLINE, Scopus, ScienceDirect, Bielefeld Academic Search Engine, and Wiley Online Library databases from 1980 to November 2020. The inclusion criteria were studies that examined the validity of portable physical activity trackers for estimating energy expenditure in individuals with stroke during walking activities compared to indirect calorimetry. This systematic review was reported according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, and the methodological quality of the included studies was determined with the Quality Assessment of Diagnostic Accuracy Studies. The study selection was made by 2 blind observers. We screened 3677 articles; 3647 were excluded after duplicate removal and title and abstract review. Thirty articles were included for full-text analysis. Eight articles met the inclusion criteria (184 individuals with stroke) and were included in the data synthesis and meta-analysis. For all monitors, activities, and placements, the overall level of correlation with indirect calorimetry was 0.34 (95% confidence interval [CI], 0.23-0.44). After subgroups analysis, we showed that type and placement have no effect on the level of validity. Test-retest reliability was high, with intraclass correlation equal to 0.89 (95% CI, 0.76-0.95). Portable physical activity monitors provided a low correlation with indirect calorimetry during walking in individuals with stroke. It seems essential to pursue studies to improve their validity in this population.