Towards Integrated Physical Activity Profiling

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Recently, there has been some discussion of whether it is possible to score highly in one dimension of physical activity behaviour (e.g., moderate intensity exercise) whilst also scoring poorly in another (e.g., sedentary time). Interestingly, direct empirical observations to support these proposals are lacking. New technologies now enable the capture of physical activity thermogenesis on a minute-by-minute basis and over a sustained period. We used one of the best available technologies to explore whether individuals can score differently in various physiologically-important physical activity dimensions. We determined minute-by-minute physical activity energy expenditure over 7 days in 100 men aged 28±9 years. We used combined accelerometry and heart rate with branched equation modelling to estimate energy expenditure and extracted data for key physical activity outcomes and descriptors. Although some physical activity outcomes were tightly correlated, the attainment of one threshold for a given physical activity dimension did not automatically predict how well an individual scored in another dimension (with bivariate correlations ranging from 0.05 to 0.96). In one illustrative example of this heterogeneity, although 41 men showed a relatively low Physical Activity Level (total energy expenditure/resting energy expenditure ≤1.75), only 17% (n = 7) of these men showed consistently low physical activity across other dimensions (moderate intensity activity, vigorous intensity activity, and sedentary time). Thus, physical activity is highly heterogeneous and there is no single outcome measure that captures all the relevant information about a given individual. We propose that future studies need to capture (rather than ignore) the different physiologically-important dimensions of physical activity via generation of integrated, multidimensional physical activity ‘profiles’.

This publication has 35 references indexed in Scilit: