C-reactive protein and cardiovascular risk in 454,203 individuals without ASCVD

B. Kurt (Aachen)¹, M. Reugels (Aachen)¹, K. M. Schneider (Aachen)², A. Milzi (Lugano)³, A. G. Antwerpen (Aachen)¹, K. Rex (Aachen)¹, K. L. Aygar (Aachen)¹, J. Bornemann (Aachen)¹, N. Ganesh (Aachen)¹, A. Giacin (Aachen)¹, S. Just (Aachen)¹, A. Kapoor (Aachen)¹, A. M. Mertens (Aachen)¹, K. Müser (Aachen)¹, M. Neuhaus (Aachen)¹, L. B. Quintana Selek (Aachen)¹, R. Salagundi (Aachen)¹, M. Sausen (Aachen)¹, N. Tabaza (Aachen)¹, A. Gombert (Aachen)⁴, M. Lehrke (Traunstein)⁵, N. Marx (Aachen)¹, C. V. Schneider (Aachen)², F. Kahles (Aachen)^1
1Uniklinik RWTH Aachen Med. Klinik I - Kardiologie, Angiologie und Internistische Intensivmedizin Aachen, Deutschland; ²Uniklinik RWTH Aachen Med. Klinik III - Gastroenterologie und Stoffwechselkrankheiten Aachen, Deutschland; ³Istituto Cardiocentro Ticino - EOC Lugano, Schweiz; ⁴Uniklinik RWTH Aachen Klinik für Gefäßchirurgie Aachen, Deutschland; ⁵Klinikum Traunstein Kardiologie Traunstein, Deutschland

Background: High sensitivity c-reactive protein (hsCRP) is a marker of systemic inflammation and predicts cardiovascular risk in individuals without known atherosclerotic cardiovascular disease (ASCVD). More information about its clinical relevance (risk discrimination and calibration) in larger databases is needed to establish hsCRP as a routine clinical biomarker as useful as LDL-C and blood pressure to identify patients at risk of ASCVD.

 

Methods: HsCRP was measured at baseline in 454,203 UK Biobank participants with no history of previous ASCVD. The UK Biobank is a population-based cohort study that was conducted in the UK from 2006 to 2010, which recruited 502,505 volunteers aged 37 - 73 years at baseline. The primary end point was CV death.

 

Results: Mean age of the participants at baseline was 56.6 years and 46% were male. 6,562 CV death events occurred during the median follow-up period of 13.4 years. Quartiles of increasing baseline levels of hsCRP predicted CV death. This association remained significant after adjustment for age, sex, BMI, type 2 diabetes, total cholesterol, HDL-C, systolic blood pressure and eGFR (Chi²: 5495.21, p<0.0001). Addition of hsCRP to a model containing age, sex, BMI, smoking and eGFR led to a higher incremental added value in risk prediction than addition of total cholesterol or systolic blood pressure (model + hsCRP: delta Chi²: 196.26; fraction of new information: 3.4%; model + total cholesterol: delta Chi²: 85.77; fraction of new information: 1.52%; model + systolic blood pressure: delta Chi²: 142.46; fraction of new information: 2.6%). Adjustment of the Framingham risk score or SCORE2 by hsCRP enables a more accurate prediction of 10-year risk of CV death (Framingham + hsCRP: delta Chi²: 168.69; fraction of new information: 9.64%; SCORE2 + hsCRP: delta Chi²: 125.16; fraction of new information: 7.4%).

 

Conclusion: In this large-scale biomedical database hsCRP proves to be a strong and independent predictor of CV death in individuals with no history of ASCVD. The association of hsCRP with CV death is as large if not larger than systolic blood pressure or total cholesterol. HsCRP has potential to become a routine clinical biomarker for CV risk assessment in patients without history of ASCVD.