JMIR Publications

ABSTRACT

Background:

Vast numbers of physical activity monitor models are being released each year and are being used by everyday consumers and researchers as well. As more studies are carried out on children and adolescents in terms of sedentary behaviour (SB) assessment, knowledge improvement about valid monitoring methods is important.

Objective:

Main aim of the systematic review was to investigate and communicate the findings on the accuracy of consumer-grade physical activity monitors in terms of assessing SB in children and adolescents.

Methods:

Searches of PubMed (MEDLINE), Scopus, SPORTDiscus (with Full Text), ProQuest, Open Access Theses and Dissertations, DART (Europe E-theses Portal), and Networked Digital Library of Theses and Dissertations electronic databases were performed. All relevant studies that compared different types of consumer-grade monitors against a comparison method in the assessment of SB, published in European languages from 2015 onwards were considered for inclusion. Risk of bias was estimated by Consensus-Based Standards for the Selection of Health Status Measurement Instruments. To enable comparisons of accuracy measures within the studied outcome domain measurement accuracy interpretation was based on group mean or percentage error values as well as 90% confidence intervals (CI). Acceptable limits were predefined as ±10% percentage error in controlled and free-living settings. To determine the number of studies with group error percentages that fall within or outside one of the sides from previously defined acceptable limits, two one-sided tests (TOST) of equivalence were carried out with the direction of measurement error also examined.

Results:

A total number of 8 studies complied with the predefined inclusion criteria, and 3 studies provided acceptable data for quantitative analyses. In terms of presented accuracy comparisons, 14 were subsequently identified with 6 of those comparisons being acceptable in terms of quantitative analysis. Results of the Cochran Q test indicated that the included studies do not share a common effect size [Q (df = 5) = 82.86; p < 0.0001]. I2 as it represents the percentage of total variation across studies that is due to heterogeneity amounted to 94%. Summary effect size based on the random effects model was not statistically significant (ES = 14.36; SE = 12.04; 90%CI = -5.45, 34.17; p = 0.23). According to equivalence tests results, consumer-grade physical activity monitors did not generate equivalent estimates of SB in relation to comparison methods. Majority of the studies that reported MAPEs have reported values less than 30%.

Conclusions:

This is one of the only studies that focused solely on the accuracy of consumer-grade physical activity monitors when measuring SB in children and adolescents, representing a need for more research in the future. The presented results highlight heterogeneity in this area of research, of which utilized devices and cut-off points defining SB are emphasized. This study should inform the researchers, consumers, and clinicians on the measurement accuracy of widely attainable physical activity monitors when measuring SB in children and adolescents.