Restriction lift date: 2021-05-18
Activity profiles of adults aged 50 - 70 years: functional data analysis
Loading...
Files
Full Text E-thesis
Date
2019-10
Authors
Weedle, Richard
Journal Title
Journal ISSN
Volume Title
Publisher
University College Cork
Published Version
Abstract
Physical activity has a major impact on health. Questionnaires are the most
common method of physical activity assessment. While cost effective, these are
subjective and can correlate poorly with actual activity levels. Accelerometers
have gained popularity given their accuracy, objectivity and ability to capture
large amounts of data. Simple summary measures such as the total or average
activity over the day are often used. However, these fail to exploit the
longitudinal nature of the data and do not capture the variation in activity
levels throughout the day. This study intends to capitalise on this nature by
implementing a functional data analysis approach. Activity data was collected from a cohort of 475 people in Mitchelstown in 2011. The individuals wore wrist worn accelerometers in a free living environment for a week. This data was collapsed into 1 minute epochs and each epoch was then aggregated over the week to get an estimate of daily circadian activity. The discrete wavelet transform was chosen as the smoothing technique to reveal the underlying functional nature of the data. This allows every individual in the cohort to be represented by a smooth activity profile. This study aimed to
identify and characterise subgroups within a cohort based on these activity
profiles. Functional principal component analysis was applied to these activity profiles in
order to explore the dominant patterns within the data. Each individual’s
profile was approximated by a weighted sum of profiles and these weights were
then used to perform a cluster analysis. Five distinct subgroups were identified.
These differed from each other in both the magnitude of the activity and the
times at which the activity occured. A more simplified approach, based purely
on the distance between profiles, was also implemented. Two distinct clustering
methods identified the exact same 5 subgroups in the cohort. To ensure their
robustness, these results were subject to a sensitivity analysis with respect to
the epoch length, smoothing technique and number of functional components
utilised in the clustering. Other studies have clustered accelerometer data in terms of absolute activity volume, as in high or low activity groups. However, they do not place too much value in using the granularity of the data to determine what time of day people
are active. In addition to the high, moderate and low activity subgroups, our
analysis revealed two subgroups which have a propensity to be active in either
the morning or evening. It is suggested that these are indicative of an
individual’s biological rhythm or chronotype. The Mitchelstown cohort was
re-screened 5 years later in 2016, which presents an exciting opportunity to
examine changes in these profiles over time.
Description
Keywords
Functional data analysis , Accelerometer
Citation
Weedle, R. P. 2019. Activity profiles of adults aged 50 - 70 years: functional data analysis. MRes Thesis, University College Cork.