Identifying clusters of multiple long term conditions and their associations with quality of life.

Conference:

SAPC ASM 2023 - Brighton

Talk Code:

5C.5

Presenter:

Bhautesh D. Jani

Co-authors:

Stefanie J. Krauth, Sayem Ahmed, Grace O. Dibben, Peter Hanlon, Jim Lewsey, Barbara Nicholl, Emma McIntosh, Rod S. Taylor, Sally J. Singh, Frances S. Mair, Lewis Steell

Author institutions:

School of Health and Wellbeing - University of Glasgow, Department of Respiratory Science - University of Leicester

Problem

Effectively managing the complex needs of people living with multiple long-term conditions (MLTCs) is a key healthcare priority. People living with MLTCs encompass heterogeneous disease profiles, yet multimorbidity classification has focused primarily on counting number of co-existing LTCs, rather than combinations or clusters of LTCs. Identifying LTC clusters and their associations with health and healthcare outcomes may facilitate development of targeted interventions and services.

Objectives

1. Identify age-stratified clusters of MLTCs and investigate their associations with health-related quality of life (HRQoL) in two population-based UK cohorts.

2. Compare the association between clusters of MLTCs and HRQoL to the association between LTC count and HRQOL.

Approach

Latent class analysis (LCA) was applied to baseline data from the UK Biobank and Understanding Society (US) datasets, to identify clusters of MLTCs in four age-strata: young (18 – 36 years [US only]), middle-aged (37 – 54 years), older (55 – 73 years) and elderly (74+ years [US only]) adults. People who self-reported ≥2LTCs were included in LCA. Optimal number of latent classes determined using model parsimony and clinical interpretation. Associations between LTC clusters and counts with HRQoL (EQ-5D Index scores) at approx. 5-year (US) and 10-year (UK Biobank) follow up were investigated using tobit regression models, adjusted for sociodemographic covariates/baseline HRQoL. People with no multimorbidity (zero/one LTC) were the reference group.

Findings

Composition of LTC clusters differed across age strata. Depression was highly prevalent across clusters in young/middle-aged adults. Painful conditions , arthritis, and hypertension were prominent in clusters identified across middle-aged/older/elderly adults. All identified LTC clusters were associated with lower HRQoL compared to those with no multimorbidity. In young/middle-aged adults, three clusters with depression as an anchoring LTC (i.e. >50% prevalence) were associated with large deficits in HRQoL (beta coefficients: -0.134 to -0.101 ). High prevalence of painful conditions and arthritis were associated with lower HRQoL across several LTC clusters from middle-age onwards. In US only, clusters with high prevalence of coronary heart disease were identified in middle-aged/older/elderly adults and were associated with the worst HRQoL scores at follow up (beta coefficients: -0.294, -0.143 & -0.104, respectively). Associations between LTC counts and HRQoL revealed poorer HRQoL scores in all age-categories as number of LTCs increased. For middle-aged/older adults, having ≥4 LTCs was associated with greater deficit in HRQoL than membership of any LTC cluster. In young adults, similar associations were found for having ≥3 LTCs.

Consequences

The magnitude of negative associations between MLTCs and HRQoL differs according to the cluster of LTCs and by age. However, having a high number of co-existing LTCs may detriment HRQoL more than the LTC cluster. Taking LTC clusters into consideration could improve the development of more targeted interventions for people living with MLTCs.

Submitted by:

Lewis Steell

Funding acknowledgement:

This project is funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research Programme (Grant Reference Number NIHR202020). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.