INCOHERENCE WITH STUDIES USING ACTIGRAPH MTI AMONG CHILDREN AGE 6-12 YR

Abstract

Dear Editor-in-Chief: Activity monitors, and especially the ActiGraph MTI (ActiGraph MTI, LLC, Fort Walton Beach, FL), have come into wide use as a measure of physical activity (PA) among children (6-12 yr). Sherar et al. (6), using this device, addressed the important issue of the influence of physical maturity on the gender inequalities in PA. We commend the efforts by the authors to address these important questions. However, the authors used an inappropriate MET-to-count prediction equation and found a daily moderate to vigorous PA (MVPA) > 100 and about 80 min·d−1 across chronological and biological ages, respectively. These findings are as challenging as they are amazing, because concomitantly published data revealed a geometric mean of about 16 min·d−1 among children of a similar age group (3). If it should be acknowledged that PA is likely influenced by sociocultural and environmental disparities, and even if the PA among children deals with important interindividual variations, such a discrepancy between studies is surprising. The gap between the two reports is mainly due to an incoherent use of MVPA cutoff points. Previously, we have reported an important bias (mean error of 113 min·d−1 in our sample) and some risks of misclassification when using an inadequate threshold to define PA among children (2). As opposed to earlier data (1,8), more recent studies (5,7) showed that the cut point of 3 METs underestimates the level of moderate intensity of PA behavior among children. This translates into an inflation of the time spent in MVPA, which is neither realistic regarding the growing prevalence of sedentary-related diseases (e.g., obesity and its comorbidities) nor reliable regarding the present PA recommendations for children. If, in a given sample, almost 100% of children meet the PA recommendations, one could ask oneself some questions about the true value of PA in combating sedentary-related diseases, and about the relevance of the recommendations themselves. Moreover, the MET-to-count cut-point used by Sherar et al. (6) may increase the risk of type I error when comparing boys with girls, because it is not discriminative enough. "Brisk walking" is often considered as a reference of moderate intensity of PA. When compared with adults, the energy cost of this type of activity may be higher among children because of some physiological and biomechanical factors. Thus, the required 3 METs for adults are not applicable to children, and some data suggested a value close to 6 METs as the lower boundary of moderate intensity of activity for children (4,7). Accordingly, it is difficult to have confidence in the conclusion of the authors about the effect of physical maturity (biological age) on the gender inequalities in PA. Because comparisons between studies need to be based on solid background, the data provided by Sherar et al. (6) need to be reconsidered. Finally, we suggest that the correct cutoff point to define MVPA should not be under approximately 3000 counts per minute for children. Comlavi B. Guinhouya, PhD Hervé Hubert, PhD Institute of Engineering in Health of Lille University of Lille 2 Lille, France