Validation of accelerometer prediction equations in children with chronic disease
journal contributionposted on 08.04.2016 by Samantha K. Stephens, Tim Takken, Dale Esliger, Eleanor Pullenayegum, Joseph Beyene, Mark S. Tremblay, Jane Schneiderman, Doug Biggar, Pat Longmuir, Brian McCrindle, Audrey Abad, Dan Ignas, Janjaap Van Der Net, Brian Feldman
Any type of content formally published in an academic journal, usually following a peer-review process.
© 2016 Human Kinetics, Inc. The purpose of this study was to assess the criterion validity of existing accelerometer-based energy expenditure (EE) prediction equations among children with chronic conditions, and to develop new prediction equations. Children with congenital heart disease (CHD), cystic fibrosis (CF), dermatomyositis (JDM), juvenile arthritis (JA), inherited muscle disease (IMD), and hemophilia (HE) completed 7 tasks while EE was measured using indirect calorimetry with counts determined by accelerometer. Agreement between predicted EE and measured EE was assessed. Disease-specific equations and cut points were developed and cross-validated. In total, 196 subjects participated. One participant dropped out before testing due to time constraints, while 15 CHD, 32 CF, 31 JDM, 31 JA, 30 IMD, 28 HE, and 29 healthy controls completed the study. Agreement between predicted and measured EE varied across disease group and ranged from (ICC).13-.46. Disease-specific prediction equations exhibited a range of results (ICC.62-.88) (SE 0.45-0.78). In conclusion, poor agreement was demonstrated using current prediction equations in children with chronic conditions. Disease-specific equations and cut points were developed.
This study was funded with a grant from the Canadian Institute of Health Research (# 167391/CIHR)
- Sport, Exercise and Health Sciences