Abstract
BACKGROUND: The estimation of physical activity (PA) and sedentary behaviour with accelerometers is typically based on the relationship between accelerometer output and metabolic equivalents (METs)-an index of PA intensity. But for adults with Down syndrome (DS), PA intensity may be better reflected in the percent oxygen uptake reserve (%VO(2Reserve)), as it accounts for their lower aerobic fitness. This study examined if the relationship between accelerometer output and METs or %VO(2Reserve) across various PAs and sedentary behaviours differs between adults with and without DS. METHODS: Forty-one adults with DS (age 35 ± 9 years; 18 women) and 41 adults without DS (age 24 ± 5 years; 18 women) performed 17 tasks of varying intensity. We estimated aerobic fitness with a submaximal treadmill test. We measured oxygen uptake with portable calorimetry and expressed it as METs and %VO(2Reserve). Output from triaxial accelerometers on the nondominant hip and wrist was determined as Vector Magnitude (VM). We used multilevel modelling to evaluate the relationships of METs or %VO(2Reserve) with VM, controlling for body mass index (BMI) and age. RESULTS: For the hip accelerometer, VM and the group-by-VM interaction significantly predicted METs (p < 0.001; conditional R(2) = 0.82), but between-group differences were small. For the wrist accelerometer, VM and age significantly predicted METs (p < 0.035; conditional R(2) = 0.76). For both the hip and the wrist accelerometer, VM, the group-by-VM interaction and BMI significantly predicted %VO(2Reserve) (p ≤ 0.047; conditional R(2) = 0.82 and 0.74, respectively). Between-group differences in the response to VM were larger for VO(2Reserve) than METs models. CONCLUSION: The relationship between accelerometer output and %VO(2Reserve) across PAs and sedentary behaviours is different between adults with and without DS. Calibrating accelerometer output against %VO(2Reserve) may be preferable to METs.