International children's accelerometry database (ICAD): design and methods.
BMC Public Health 2011 ; 11: 485.
Sherar LB, Griew P, Esliger DW, Cooper AR, Ekelund U, Judge K, and Riddoch C
DOI : 10.1186/1471-2458-11-485
PubMed ID : 21693008
PMCID : PMC3146860
Over the past decade, accelerometers have increased in popularity as an objective measure of physical activity in free-living individuals. Evidence suggests that objective measures, rather than subjective tools such as questionnaires, are more likely to detect associations between physical activity and health in children. To date, a number of studies of children and adolescents across diverse cultures around the globe have collected accelerometer measures of physical activity accompanied by a broad range of predictor variables and associated health outcomes. The International Children's Accelerometry Database (ICAD) project pooled and reduced raw accelerometer data using standardized methods to create comparable outcome variables across studies. Such data pooling has the potential to improve our knowledge regarding the strength of relationships between physical activity and health. This manuscript describes the contributing studies, outlines the standardized methods used to process the accelerometer data and provides the initial questions which will be addressed using this novel data repository.
Between September 2008 and May 2010 46,131 raw Actigraph data files and accompanying anthropometric, demographic and health data collected on children (aged 3-18 years) were obtained from 20 studies worldwide and data was reduced using standardized analytical methods.
When using ≥ 8, ≥ 10 and ≥ 12 hrs of wear per day as a criterion, 96%, 93.5% and 86.2% of the males, respectively, and 96.3%, 93.7% and 86% of the females, respectively, had at least one valid day of data.
Pooling raw accelerometer data and accompanying phenotypic data from a number of studies has the potential to: a) increase statistical power due to a large sample size, b) create a more heterogeneous and potentially more representative sample, c) standardize and optimize the analytical methods used in the generation of outcome variables, and d) provide a means to study the causes of inter-study variability in physical activity. Methodological challenges include inflated variability in accelerometry measurements and the wide variation in tools and methods used to collect non-accelerometer data.