Multimorbidity and physical activity profiles: results of the SMILE community cohort study

Abstract

Ilse Mesters, Jan Schepers, Sarah Dorenkamp, Rob de Bie, Marjan van den Akker, Rein Vos

Background: Physical Activity has emerged as the key treatment option for a wide range of chronic diseases. Many adults aged 55 and up suffer from multiple chronic diseases. Behavioral change is driven by stimulating specific physical activity behaviors, such as walking. Knowing which behaviors are common among which patients may inform interventions. This study describes the types of physical activity performed in a representative, community-based Dutch sample of adults with multimorbidity. We hypothesize that chronic disease clusters exist within this large population and that these meaningful classes may encompass a range of specific physical activity behaviors.

Patients and Methods: Cross-sectional data was used from 3,386 patients, who were 55 years and up, were registered at one of nine primary healthcare facilities that participated in the SMILE cohort study and completed a lifestyle and a chronic illness questionnaire. A model for the association between multimorbidity and physical activity behavior was developed using latent class analysis.

Results: The best fit was a 4-class solution with: (1) a group with the lowest prevalence of cancer, rheumatoid arthritis, and cardiovascular disease and very active in walking, light and intensive household activities; (2) a relatively healthy group in which liver disease or cirrhosis and cycling, gardening, odd jobs and sports were prevalent; (3) a group with the lowest prevalence of migraine, inflammatory joint disease and lowest activity levels in all types of physical activities except for light household chores; and (4) a group with the highest prevalence of all chronic diseases except for liver disease or cirrhosis with moderate physical activity levels.

Conclusion: Multimorbidity did not seem to be the primary obstacle to engaging in physical activity. There was a difference between the classes in terms of the probability to engage in each form of activity, even though the pattern of physical activity behavior for all seven categories of activities appeared equivalent among classes.

PDF