Plasma proteomic signatures linking preserved ratio impaired spirometry to cardiometabolic diseases and multimorbidity: A longitudinal cohort study

Background
Preserved ratio impaired spirometry (PRISm), defined by reduced FEV1 with preserved FEV1/FVC ratio, has been linked to cardiometabolic diseases (CMDs), but underlying mechanisms remain unclear. This study aims to identify plasma proteomic signatures linking PRISm to type 2 diabetes (T2D), ischaemic heart disease (IHD), stroke, first cardiometabolic disease (FCMD), and cardiometabolic multimorbidity (CMM).

Methods
We conducted plasma proteomic profiling of 2 911 proteins in 32 689 UK Biobank participants with a median follow-up of 14.37 years. Machine learning algorithms were employed to identify protein signatures associated with PRISm. Causal mediation analyses were performed to quantify the mediating effects of identified proteins on the associations between PRISm and cardiometabolic outcomes. Combined protein-covariate models were developed to assess predictive performance.

Results
We identified a 29-protein signature associated with PRISm. Outcome-specific analyses revealed 29 proteins for T2D, 26 for IHD, 13 for stroke, 29 for FCMD, and 27 for CMM, with 14 proteins showing consistent associations across all outcomes. Combined protein-covariate models significantly improved prediction over covariates alone (AUCs: T2D, 0.83; IHD, 0.72; stroke, 0.75; FCMD, 0.76; CMM, 0.79). Mediation analyses demonstrated that IGSF3 mediated 19.5% of the PRISm-T2D association, 12.8% for FCMD, and 11.3% for CMM, while GDF15 mediated 12.1% for IHD and 13.1% for stroke. These proteins converged on inflammatory signalling, extracellular matrix remodelling, and metabolic pathways.

Conclusions
The identified protein signatures offer promising biomarkers for risk prediction, enabling early identification of high-risk individuals with PRISm and informing targeted preventive strategies for cardiometabolic complications.