How do objectively measured physical activity behaviours relate to morbidity in older adults? A latent class analysis

Conference: 
SAPC ASM 2021 - virtual
Talk Code: 
1D.6
Presenter: 
Andrew O'Regan
Co-authors: 
Ailish Hannigan(1), Liam Glynn(1), Alan Donnelly(2), Enrique Garcia Bengoechea(2), Grainne Hayes(2), Catherine Woods(2)
Author institutions: 
(1)Health Research Institute and School of Medicine, University of Limerick; (2) Health Research Institute and Dept. of Physical Education and Sport Sciences, University of Limerick

Problem

Physical activity contributes to the prevention of chronic illness as well as promotion of physical and mental health, but most adults remain inactive. Chronic illness affects mainly middle aged and older adults, and very little objectively measured data on physical activity behaviours and associated health outcomes of this population is published.

The aims of this study are to: 1. Objectively measure physical behaviour outcomes of adults participating in the Move for Life study; 2. Develop distinct activity profiles based on six behaviour variables; 3. Investigate whether health outcomes differ across the activity profiles.

Approach

Participants were Irish adults aged 50 years and older. They had been recruited as part of the Move for Life feasibility cluster randomised control trial. This is a cross-sectional study, applying a latent class analysis to baseline data of the Move for Life trial. Using the activPAL, objectively measured data were collected on average daily: light physical activity (hours); moderate to vigorous physical activity (minutes); step count; time in bed (hours); standing time (hours); and waking sedentary time (hours). Data were obtained on chronic illness and health service utilisation. Validated questionnaires were used to collect data on wellbeing, loneliness and social isolation. Hierarchical cluster analysis using squared Euclidian distance was used to cluster behaviours based on similarity, using SPSS version 26. Regression models explored associations between health outcomes and activity profiles, adjusted for age and sex.

Findings

Data from 485 participants were analysed, and four activity profiles were identified: sedentary (n=50, 10.3% of total), low active (n= 295 ,60.8%), moderate active (111, 22.9%) and higher active (n=29, 6%). We will present the differences across the activity profiles for chronic illnesses, multi-morbidity, health service utilisation and validated health tools, comparing to data from the Irish Longitudinal Study on Ageing (TILDA) and the English Longitudinal Study on Ageing (ELSA).

 

Consequences

The use of physical activity behaviour clusters may identify people with multi-morbidity and higher utilisation of health services. These findings could be factored into the development of future targeted physical activity interventions.

 

Submitted by: 
Andrew O'Regan
Funding acknowledgement: 
The Move for Life research project is funded under the Health and Positive Aging Initiative (HaPAI), led by the Department of Health and is a joint national programme with the Health Service Executive and Atlantic Philanthropies.