Longitudinal evaluation of high-sensitivity C-reactive protein levels for cardiovascular risk prediction in apparently healthy individuals

B. Kurt (Aachen)¹, M. Reugels (Aachen)¹, K. M. Schneider (Dresden)², F. Wenzl (Zürich)³, C. B. Fordyce (Vancouver)⁴, N. J. Pagidipati (Durham)⁵, V. Rocha (Sao Paulo)⁶, M. Fudim (Durham)⁷, A. Sharma (Montreal)⁸, J. Spießhöfer (Aachen)⁹, A. Milzi (Lugano)¹⁰, M. Lehrke (Traunstein)¹¹, H. Shimokawa (Sendai)¹², G. Liuzzo (Rome)¹³, L. Tokgozoglu (Ankara)¹⁴, F. Crea (Rome)¹⁵, T. F. Lüscher (Zürich)¹⁶, P. Libby (Boston)¹⁷, P. M. Ridker (Boston)¹⁸, N. Marx (Aachen)¹, C. V. Schneider (Aachen)¹⁹, F. Kahles (Aachen)^1
1Uniklinik RWTH Aachen Med. Klinik I - Kardiologie, Angiologie und Internistische Intensivmedizin Aachen, Deutschland; ²Technische Universität Dresden Medizinische Klinik und Poliklinik I Dresden, Deutschland; ³University of Zurich Center for Molecular Cardiology Zürich, Schweiz; ⁴Vancouver General Hospital, University of British Columbia Division of Cardiology, Department of Medicine and the Centre for Cardiovascular Innovation Vancouver, Kanada; ⁵Duke University School of Medicine Duke University Medical Center; Division of Cardiology, Department of Medicine Durham, USA; ⁶University of Sao Paulo Medical School Hospital Heart Institute (Instituto do Coracao (InCor)) Sao Paulo, Deutschland; ⁷Duke Cardiology Clinic - Clinic 2F/2G Durham, USA; ⁸McGill University Division of Cardiology, Research Institute of the McGill University Health Centre Montreal, Kanada; ⁹Uniklinik RWTH Aachen Med. Klinik V - Klinik für Pneumologie und Internistische Intensivmedizin Aachen, Deutschland; ¹⁰Istituto Cardiocentro Ticino - EOC Lugano, Schweiz; ¹¹Klinikum Traunstein Kardiologie Traunstein, Deutschland; ¹²Tohoku University Graduate School of Medicine Department of Cardiovascular Medicine Sendai, Japan; ¹³Fondazione Policlinico Universitario A. Gemelli – IRCCS and Catholic University of the Sacred Heart
 Department of Cardiovascular Sciences - CUORE Rome, Italien; ¹⁴Hacettepe University Medical Faculty Department of Cardiology Ankara, Türkei; ¹⁵Catholic University of the Sacred Heart Agostino Gemelli Hospital Rome, Italien; ¹⁶Stiftung für Herz- und Kreislaufforschung Zürich, Schweiz; ¹⁷MassGeneralBrigham Hospital, and Harvard Medical School Heart and Vascular Institute Boston, USA; ¹⁸Brigham and Women’s Hospital Center for Cardiovascular Disease Prevention, Division of Preventive Medicine Boston, USA; ¹⁹Uniklinik RWTH Aachen Med. Klinik III - Gastroenterologie, Stoffwechselerkrankungen und Internistische Intensivmedizin Aachen, Deutschland

Background: High-sensitivity C-reactive protein (hsCRP) is the guideline-recommended biomarker for assessment of residual inflammatory risk in atherosclerotic cardiovascular disease (ASCVD). Despite this, its adoption in clinical practice has lagged the evidence base and these recommendations, especially in individuals without established ASCVD. We therefore aimed to evaluate the intra-individual stability and prognostic value of hsCRP. By longitudinal assessment of hsCRP, the goal of this study was to inform the clinical utility of hsCRP assessments and to refine the interpretation of inflammatory risk for preventive strategies.

Methods: Longitudinal hsCRP levels for two time-points were available in 15 967 UK Biobank participants without known history of ASCVD. The UK Biobank is a population-based cohort study conducted in the UK between 2006 and 2010. Median time period between repeated measurement was 4.4 years. Categorization of hsCRP levels was performed using a primary cutoff at 2 mg/L and additionally at <1, 1–3 and >3 mg/L according to previous data and current guideline recommendations. The primary endpoint was all-cause death and its association with hsCRP was assessed using Kaplan-Meier and Cox proportional hazards analyses with restricted cubic spline modeling.

Results: In total, repeated hsCRP measurements were available in 15 967 participants, with comparable levels at T₀ (1.16 [0.59–2.37] mg/L) and T₁ (1.21 [0.61–2.42] mg/L). Among all participants, 884 deaths occurred during a follow-up of 15.3 years starting from T₀. At both time points, hsCRP levels above 2 mg/L were associated with higher all-cause mortality in Kaplan–Meier analyses (all log-rank p<0.001). Similar results were observed when subjects were stratified by hsCRP categories of <1, 1–3, and >3 mg/L (all log-rank p<0.001). In univariable Cox regression analyses, higher continuous hsCRP concentrations were associated with increased all-cause mortality at both time points (T₀: Hazard ratio (HR): 1.42; 95% confidence interval (CI): 1.30, 1.54; p<0.001; T₁: HR: 1.58; 95% CI: 1.45, 1.72). Dichotomization of hsCRP at 2 mg/L, yielded comparable associations. At the first measurement (T₀), individuals with hsCRP levels >2 mg/L had a 46% higher risk of death (HR: 1.46; 95% CI: 1.33, 1.60; p<0.001). At the second measurement (T₁), elevated hsCRP levels were associated with a 63% higher risk of mortality (HR: 1.63; 95% CI: 1.49, 1.78; p<0.001). All results remained significant after multivariable adjustment for age, sex, body mass index, type 2 diabetes mellitus, smoking, systolic blood pressure, low-density lipoprotein cholesterol and estimated glomerular filtration rate.

Conclusion: This large-scale cohort study affirmed hsCRP as a stable, strong, and independent predictor of mortality in individuals without ASCVD. Its predictive value remained stable across two independent time points separated by several years. These findings demonstrate the temporal stability and prognostic robustness of hsCRP and support its clinical use as a reliable marker of inflammatory risk in primary prevention.