Skip to main navigation Skip to search Skip to main content

Multimorbidity patterns in an older population from North Europe

Student thesis: Doctoral thesis

Abstract

1.- Background Multimorbidity, known as the presence of two or more chronic diseases in one person, is a growing health condition related to aging. It is well known that people suffering multiple chronic diseases tend to cluster into homogenous groups, and chronic diseases tend to have a long duration and, generally, a slow progression. Therefore, the methods applied to estimate multimorbidity patterns should be flexible enough to identify those patterns and their evolution over time. 2.- Objective The aim of this thesis was to identify longitudinal multimorbidity patterns in a Swedish population-based cohort of older adults. The specific aims were 1) to estimate multimorbidity patterns and their sociodemographic, lifestyle, clinical and functional characteristics; 2) to trace the patterns’ evolution and detect clinical trajectories and mortality over time; and 3) to estimate the longitudinal evolution of older individuals and their permanence time as they move among patterns. 3.- Methods We conducted three studies to meet the aims of this thesis, using data from the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K), a population-based study including 3363 community-dwelling and institutionalized individuals aged 60 years and older. For Study 1 and Study 2, we used the fuzzy c-means cluster algorithm to cluster multimorbid participants. For Study 3, we stratified the overall sample into three ten-year age groups and applied Hidden Markov Models to track the temporal evolution of multimorbidity patterns and individuals’ transitions over 12 years of follow-up. 4.- Results In Study 1, the clusters showed significantly different sociodemographic, lifestyle, clinical and functional profiles. In Study 2, changes in cluster composition, participants’ transitions from one cluster to another and participant mortality over 12 years generated a dynamic but well-defined clinical picture. In Study 3, we identified four longitudinal multimorbidity patterns for each decade, observing that with increasing age, clinical stability, and the permanence time within a single multimorbidity pattern both decreased. 5.- Conclusions Multimorbidity patterns showed significant sociodemographic, lifestyle and functional differences. Clinical trajectories showed great dynamism and complexity but can be tracked over time. Different clusters were differentially associated with mortality. The dynamism among multimorbidity patterns was reflected by the varying permanence times across patterns. Fuzzy c-means and Hidden Markov Models captured the longitudinal nature of multimorbidity patterns. Our results may help to clarify the concept of multimorbidity and improve preventive health interventions.
Date of Award10 Feb 2023
Original languageEnglish
SupervisorConcepció Violan Fors (Director)

Cite this

'