Model Accelerometer Paradata From The International Study Of Childhood Obesity, Lifestyle And The Environment (ISCOLE)

Background: Accelerometry paradata (administrative data related to collection/management/treatment) are inconsistently reported or limited to accounts of valid days and average wear time.

Purpose: To present a model for reporting accelerometry paradata collected from children (mean age 10 years) at the Baton Rouge, USA site of the International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE).

Methods: ISCOLE employed a 24-hr waist worn 7-day protocol using the ActiGraph GT3X+. Checklists, flow charts, and systematic data queries captured accelerometry paradata including, but not limited to, the number of instruments distributed, the number of lost/broken instruments, and the number of cases culled at each stage of data cleaning, leading to the final locked data set

Results: Of the 665 children enrolled, 651 were eligible to participate (3 withdrew and 11 were outside the 9-11 year old targeted age range), 648 received accelerometers on the scheduled day of distribution, and 629 returned them (19 lost). Four devices malfunctioned and 86 children provided inadequate data evaluated using the manufacturer’s ActiLife software. Of these, 17 children wore an accelerometer a second time (a case-by-case decision), resulting in 16 files reconsidered as adequate. Following additional quality control checks and data reduction decisions (e.g., cleaning extreme values, duplicate files, evident protocol deviations, etc.), a locked accelerometry data set representing 556 children was produced. After accounting for sleep period time and applying a more stringent non-wear definition than the manufacturer’s deprecated algorithm, 491 children provided valid data (defined as ≥ 4 days and ≥ 1 weekend day, each with ≥ 10 hrs/day of wear). Among children with valid data, the mean number of valid days and valid weekend days was 6.4 and 1.8, respectively, while mean wear time was 884.5 min/day (14.7 hrs/day).

Conclusions: We present a model of reporting accelerometry paradata to standardize communication, facilitate study management, improve the representative qualities of surveys, track study endpoint attainment, and ultimately anticipate and control costs.

Supported by The Coca-Cola Company