A Catalog of Rules, Variables, and Definitions Applied to Accelerometer Data in the National Health and Nutrition Examination Survey, 2003–2006

Abstract

IntroductionThe National Health and Nutrition Examination Survey (NHANES) included accelerometry in the 2003–2006 data collection cycles. Researchers have used these data since their release in 2007, but the data have not been consistently treated, examined, or reported. The objective of this study was to aggregate data from studies using NHANES accelerometry data and to catalogue study decision rules, derived variables, and cut point definitions to facilitate a more uniform approach to these data.MethodsWe conducted a PubMed search of English-language articles published (or indicated as forthcoming) from January 2007 through December 2011. Our initial search yielded 74 articles, plus 1 article that was not indexed in PubMed. After excluding 21 articles, we extracted and tabulated details on 54 studies to permit comparison among studies.ResultsThe 54 articles represented various descriptive, methodological, and inferential analyses. Although some decision rules for treating data (eg, criteria for minimal wear-time) were consistently applied, cut point definitions used for accelerometer-derived variables (eg, time spent in various intensities of physical activity) were especially diverse.ConclusionUnique research questions may require equally unique analytical approaches; some inconsistency in approaches must be tolerated if scientific discovery is to be encouraged. This catalog provides a starting point for researchers to consider relevant and/or comparable accelerometer decision rules, derived variables, and cut point definitions for their own research questions.

Assessing sedentary behavior using wearable devices: An overview and future directions

“…In particular, a non-wear time algorithm is considerably more important because sedentary behavior is likely to overlap with non-wear periods, and may often be misclassified as non-wear. To date, many wear time algorithms have been proposed and tested 33) . The two commonly used non-wear time algorithms are Troiano's and Choi's algorithms 29,34) .…”

“…Choi's algorithm, on the other hand, defined a non-wear period as 1 minute time intervals with consecutive zero counts for at least a 90 minute time window, allowing short time intervals with non-zero counts lasting up to 2 minutes if no counts are detected either during 30 minutes (window 2) upstream or downstream from a given interval; any nonzero counts, except the allowed short interval, are considered as wear time 34) . Tudor-Locke et al systematically summarized these non-wear rules, as well as relatively minor algorithms 33) .…”

“…The environmental information could be barometric pressure 40) , geographical information system 41) , and wearable camera 42) . Another challenge in this field is the comparability of sedentary outputs from different wearable devices 33) . Our research team compared the sedentary outputs from ActiGraph GT3X+ and Active style Pro among 13 older women, and found that Active style Pro reported 15% less sedentary time than ActiGraph did 43) .…”

Unique contributions of ISCOLE to the advancement of accelerometry in large studies

“…As cataloged in Table 1, however, minimal wearing requirements to define a valid case included ⩾ 3 days (no consideration for type of day) with ⩾ 8 h per day of wear time, ⩾ 3 days including 1 weekend with ⩾ 10 h per day of wear time and ⩾ 3 weekdays ( ⩾ 10 h per day of wear time) plus ⩾ 1 weekend (⩾8 h per day of wear time). As discrepancy and disagreement in the selection and application of various accelerometer-based cut points are well known 5 and can also differ between analyses of the same data set, 8 these additional details are not presented in this simple comparative table.…”

Accelerometer thresholds: Accounting for body mass reduces discrepancies between measures of physical activity for individuals with overweight and obesity

“…Accelerometers capture changes in velocity over time (accelerations) which are known as activity counts (Gabriel et al 2010;Tudor-Locke et al 2012). Thresholds for activity counts per minute (CPM) (Troiano et al 2008;Tudor-Locke et al 2012) have been created to correspond to Metabolic Equivalents (MET) for moderate (3-6 MET), and vigorous (>6 MET) intensities of PA .…”