Physical activity assessment under free-living conditions using pattern recognition monitors

Extensive literature has documented the health benefits of physical activity. Valid, reliable and feasible physical activity assessment tools are necessary to assess the complexity and multidimensionality of physical activity behavior. Pattern-recognition activity monitors that integrate information from multiple sensors appear to be the most promising approach for assessing physical activity under free-living conditions. Previous studies have provided support to the validity of pattern-recognition monitors for assessing the energy cost of activity under-free living conditions in young adults. However, children and older adults present unique measurement challenges for the assessment of physical activity under free-living conditions. The series of studies in this dissertation extends previous research by assessing the accuracy of a pattern-recognition monitor (SenseWear Armband) in children and older adults under free-living conditions. Consistent with previous findings in young adults, results indicate that the SenseWear Armband monitors provide valid estimates of total energy expenditure and activity energy expenditure in older adults and children, under free-living conditions. Collectively, the findings of this research support the validity of the SenseWear Armband for assessing physical activity under free-living conditions in children and older adults