Chronic kidney disease-stratified plasma proteomic signatures of probable sarcopenia: A UK biobank cohort study

Chronic kidney disease (CKD) is frequently accompanied by reduced muscle strength, metabolic dysregulation, and chronic inflammation. However, the plasma proteomic signatures and underlying biological mechanisms of probable sarcopenia in individuals with CKD remain poorly characterized. This study aimed to systematically characterize the plasma proteomic signatures associated with probable sarcopenia in CKD and explore their potential biological significance.

We analysed 51,771 eligible participants with Olink measurements across 2920 proteins. The primary outcome was baseline probable sarcopenia, defined by European Working Group on Sarcopenia in Older People 2 (EWGSOP2) low-grip-strength thresholds. Chronic kidney disease (CKD) was defined broadly by estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2, urine albumin-to-creatinine ratio (UACR) ≥3 mg/mmol (approximately equivalent to ≥30 mg/g) or kidney-disease hospital International Classification of Diseases, 10th Revision (ICD-10) codes, with strict chronic-code sensitivity definitions. Analyses included CKD-stratified logistic discovery, enhanced-covariate sensitivity analyses, ICD-10 M62 cause-specific prospective validation, dual-outcome cis-protein quantitative trait locus (cis-pQTL) Mendelian randomisation (MR), fold-internal prediction modelling and nuclear magnetic resonance (NMR) metabolomic contextualisation.

Probable sarcopenia was more prevalent among UK Biobank participants with CKD than among UK Biobank participants without CKD (10.5% vs. 4.4%). CKD-stratified proteomic analysis identifies 574 plasma proteins significantly associated with probable sarcopenia (false discovery rate [FDR] < 0.05), led by COL6A3 (odds ratio [OR] = 1.64, 95% confidence interval [CI] 1.47-1.84) and GDF15 (OR = 1.53, 95% CI 1.38-1.69), alongside COLEC12, TNFRSF1B, CD300E, ADM and CST3. These associations remain largely robust under a strict CKD definition (521/574) and extended covariate adjustment (573/574). Prospective analysis links 52 proteins to incident muscle-related outcomes, with GDF15 emerging as a representative signal (hazard ratio [HR] = 1.18, 95% CI 1.10-1.27). Genetic prioritization provides supportive evidence for 170 proteins across both muscle-related and CKD-related outcomes. Incorporating least absolute shrinkage and selection operator (LASSO)-selected proteins into five-fold cross-validated penalized regression models improves discrimination beyond clinical covariates alone (area under the receiver-operating-characteristic curve [AUC] from 0.674 to 0.693-0.694). Metabolomic profiling reveals systematic alterations in lipid-related traits (low-density lipoprotein [LDL], intermediate-density lipoprotein [IDL] and very-low-density lipoprotein [VLDL]) among individuals with probable sarcopenia, and integrative analysis identifies over 7000 significant protein-metabolite associations (FDR <0.05) that converge on a lipid-transport-centred network anchored by COL6A3, IGFBP4 and CST3.

Probable sarcopenia in CKD is associated with a broad plasma protein signature linking extracellular matrix remodelling, inflammation, renal dysfunction and metabolic stress. These findings remain discovery-stage and require external CKD validation, direct nutrition assessment and mechanistic testing before clinical translation.