Multimorbidity patterns and their related characteristics in European older adults: A longitudinal perspective.

Background: The concurrence of several chronic conditions is a rising concern that poses a serious burden on ageing populations. Analysing how these conditions appear together and how they change through time may provide useful information to design successful multimorbidity-management programs.

Objective: To identify multimorbidity patterns and their related characteristics from a longitudinal perspective.

Subjects: 25,931 older adults aged 50+ drawn from the Survey of Health, Ageing and Retirement in Europe (SHARE), a population-based longitudinal European study.

Methods: A sex-stratified Latent Transition Analysis was conducted to fit latent classes based on 15 self-reported chronic conditions across three time points. Health-related and socioeconomic variables were assessed as covariates of those patterns.

Results: We identified 4 time-constant latent classes for each sex. A "severely impaired" class (with a weighted prevalence percentage of 7.24% for females and 3.30% for males at the first time point), a "metabolic" class (26.15% and 23.82%) and a "healthy" class (50.92% and 54.32%). The fourth class was named "osteoarticular" for females (15.70%) and "articular-COPD-ulcer" for males (18.56%). Age, smoke, material deprivation and a high body mass index were associated with worse health patterns, whereas education, being employed and physical activity were related to less multimorbid classes. Few class changes were detected when modelling transitions.

Conclusions: We reported information of multimorbidity classes and their characteristics that may help to develop targeted health strategies. Within a time window of four years, the identified latent classes were consistent between time points.