877. The Compositional Isotemporal Substitution Model: A method for estimating changes in a health outcome for reallocation of time between sleep, sedentary behaviour, and physical activity.
Commentary by Prof Adrian Bauman, GlobalPAnet Executive, The University of Sydney, Australia.
Source: Statistical Methods in Medical Research
This paper by Dumuid and colleagues  describes a new statistical technique relevant to the analysis of physical activity data. This technique is described as compositional analysis, and takes account of the 24-hour boundary around our total moderate to vigorous activity time, light intensity time, sedentary behaviour time, and sleep. It uses innovative statistical methods to test the ratio between these domains of time use, as the 24 hour frame is “time bounded”, and can therefore model in more robust ways transitions between physical activity states.
What this means in practice is you can assess the benefits of increasing moderate to vigorous physical activity at the expense of light intensity time, or at the expense of sedentary time, and assess the benefits on some health outcomes. This is a relevant new method for physical activity research, and has been used to assess the relationships between changes in physical activity states and quality of life, obesity levels particularly in children, and other parameters. It complements but extends the methods previously developed as "isotemporal substitution analyses".
The purpose of mentioning it here is that it will increasingly be used in research, and will answer modelling questions of relevance to policymakers and practice. For example what change in body mass index could you expect if you reallocated half an hour of sitting time or standing time to moderate to vigorous physical activity? As 24-hour guidelines for physical activity in children already exist in Canada, and increasingly physical activity measurement has 24 hour objective assessment, this method of compositional data analysis is likely to become more widely used. It is technical, but like many new analytic techniques the research field will adapt to it fairly quickly. Communications here are to introduce the concept such that practitioners and policymakers are not lost in statistical language, as the concept is underpinning "data composition" is quite straightforward, facilitating better research across the 24-hour physical activity continuum. This may be more informative than focusing on the individual components (such as MVPA) in isolation.
Access to this article may depend on your institutional rights: Access the full article here.