Further evaluation of observational and mechanical measures of physical activity

Abstract

Recent behavioral research has used a variety of methods to quantify physical activity, including direct observation using the Observational System for Recording Physical Activity (OSRAC), and automated devices such as heart rate (HR) monitors, pedometers, and Fitbit Accelerometers. The current study evaluated the concurrent validity of these measures, as well as the reliability of the pedometers and Fitbits. Four children engaged in 15 activities listed in the OSRAC coding system while their HRs and steps taken were measured. The results indicated that for most activities, HR covaried with the OSRAC activity levels, although some exceptions were noted. Steps per minute as measured by the Fitbit also covaried with the OSRAC activity levels and Fitbit data were reliable when activities did not involve subjects losing contact with the ground.Pedometers produced similar but less reliable steps per minute data.

Using heart rate as a physical activity metric

“…For example, average HRs ranged from 122 to 136 and 174 to 185 BPM during walking and jogging, respectively. Although these data do not demonstrate as wide a range as previous studies (e.g., 152 to 195 BPM while jogging for participants in Larson et al, 2011 andBerth, 2018), as a whole, these individual differences highlight the importance of taking individualized measures of HR into account when trying to determine the efficacy of any one exercise. This assessment allowed us to determine individualized HR zones for moderate and vigorous PA, to which other less-structured activities could be compared in the subsequent activity assessment.…”

“…For example, it appears, based both on visual inspection of graphs and on data extrapolated using Graphclick (Arizona Software, Inc., ), that running or jogging (Level 5) were associated with a relatively wide range of HR across four participants (152, 167, 180, and 199 BPM, respectively). Van Camp and Berth () replicated and extended this study by including four additional exercise behaviors at each of the five OSRAC activity levels. Similar individual differences were observed; for example, running or jogging was associated with HRs of 168, 171, 183, and 195 BPM across four participants, who were 13, 8, 6, and 9 years old, respectively.…”

“…In the studies noted above, physical activity was measured using either strictly observational measures (i.e., the OSRAC) or mechanical measures (i.e., pedometers, accelerometers). In some cases, children may produce similar step counts, but idiosyncratic HRs (Larson et al ; Van Camp & Berth, ). Steps as a primary measure of PA may also be problematic when children engage in behaviors that do not involve translocation (Van Camp & Berth, ).…”

“…In some cases, children may produce similar step counts, but idiosyncratic HRs (Larson et al ; Van Camp & Berth, ). Steps as a primary measure of PA may also be problematic when children engage in behaviors that do not involve translocation (Van Camp & Berth, ). Including a more sensitive, noncontext‐specific measure such as HR might enhance the validity of the assessments of physical activity described above.…”

Physical activity in children: An evaluation of an individualized heart rate assessment.

“…Few studies have evaluated submaximal tests in children, although some exceptions have focused on comparing HR while children engaged in structured activities associated with the CARS and OSRAC 5-level observation system (Larson et al, 2011; Puhl et al, 1990; Van Camp & Berth, 2018). In the original CARS study, observation codes were calibrated by having 25 preschool children engage in representative activities (controlled walking indoors on a treadmill at various speeds and inclines) while HR was measured (Puhl et al, 1990).…”