Accelerometer Metrics to Assess Movement Behaviours
For over a decade there have been calls to move away from cutpoint-based estimates of physical activity and to develop alternative methods to monitor physical activity and other movement behaviours using devices worn by research participants. The reasons for this proposed shift in methods are quite well rehearsed, with the main points are summarised here and here.
Since 2018 a number of alternative accelerometer-derived methods have been introduced to take full advantage of the high frequency and full intensity range of data that are captured from prolonged accelerometer wear. The average acceleration and intensity spectrum were developed and tested in this paper by Alex Rowlands and colleagues as metrics to represent activity volume and intensity. This study was followed up by Alex’s development of ‘MX’ translational metrics to represent the minimum acceleration accumulated over the most active number of minutes (where ‘X’ is the number of minutes).
I have taken a strong interest in this area and working with colleagues have applied these metrics in my own research. We looked at how average acceleration and intensity gradient were associated with a range of health and wellbeing outcomes in children in this study. We found that intensity gradient was generally associated with physical health outcomes independently of average acceleration. In contrast, the association between average acceleration and wellbeing was independent of intensity gradient. We also applied the MX metrics as an alternative physical activity guideline. Focusing on the school day where there is an expectation for children to accumulate at least 30 minutes of daily MVPA, we used the M30 metric (i.e., the minimum acceleration for the most active accumulated 30 minutes during the school day). We reported M30 thresholds for adiposity and CRF and found high agreement between M30 and MVPA tertiles. The study provided further support for M30 (and MX metrics more generally) as an accelerometer-specific metric for generating physical activity guidelines related to health indicators and easily understood forms of activity such as brisk walking.
We are currently interrogating our pooled accelerometer data set (see Highlighted Publications page) to further this avenue of investigation.