Identifying proteins and pathways associated with multimorbidity in 53,026 adults

Metabolism. 2024 Dec 29:156126. doi: 10.1016/j.metabol.2024.156126. Online ahead of print.

Abstract

Background and aims: Multimorbidity, the coexistence of multiple chronic diseases, is a rapidly expanding global health challenge, carrying profound implications for patients, caregivers, healthcare systems, and society. Investigating the determinants and drivers underlying multiple chronic diseases is a priority for disease management and prevention.

Method: This prospective cohort study analyzed data from the 53,026 participants in the UK Biobank from baseline (2006 to 2010) across 13.3 years of follow-up. Using Cox proportional hazards regression model, we characterized shared and unique associations across 38 incident outcomes (31 chronic diseases, 6 system mortality and all-cause mortality). Furthermore, ordinal regression models were used to assess the association between protein levels and multimorbidity (0-1, 2, 3-4, or ≥ 5 chronic diseases). Functional and tissue enrichment analysis were employed for multimorbidity-associated proteins. The upstream regulators of above proteins were identified.

Results: We demonstrated 972 (33.3 %) proteins were shared across at least two incident chronic diseases after Bonferroni correction (P < 3.42 × 10-7, 93.3 % of those had consistent effects directions), while 345 (11.8 %) proteins were uniquely linked to a single chronic disease. Remarkably, GDF15, PLAUR, WFDC2 and AREG were positively associated with 20-24 incident chronic diseases (hazards ratios: 1.21-3.77) and showed strong associations with multimorbidity (odds ratios: 1.33-1.89). We further identified that protein levels are explained by common risk factors, especially renal function, liver function, inflammation, and obesity, providing potential intervention targets. Pathway analysis has underscored the pivotal role of the immune response, with the top three transcription factors associated with proteomics being NFKB1, JUN and RELA.

Conclusions: Our results enhance the understanding of the biological basis underlying multimorbidity, offering biomarkers for disease identification and novel targets for therapeutic intervention.

Keywords: Chronic diseases; Multimorbidity; Plasma proteins.