1. ¹Research Center for Prevention and Health, Glostrup University Hospital, Copenhagen, Denmark
2. ²Department of Internal Medicine, Glostrup University Hospital, Copenhagen, Denmark
3. ³Department of Clinical Physiology and Nuclear Medicine, Glostrup University Hospital, Copenhagen, Denmark
4. ⁴Department of Cardiology, University Hospital of Copenhagen, Rigshospitalet, Copenhagen, Denmark
5. ⁵Department of Internal Medicine, Holbæk Hospital, Holbæk, Denmark
6. ⁶Department of Cardiology, Bispebjerg University Hospital, Copenhagen, Denmark
7. ⁷Department of Cardiology, Roskilde University Hospital, Roskilde, Denmark

Correspondence: Dr MH Olsen, Clinic of Cardiology, Department of Internal Medicine, Glostrup University Hospital, DK-2600 Glostrup, Denmark. E-mail: mho@dadlnet.dk

Received 25 February 2008; Revised 31 March 2008; Accepted 1 April 2008; Published online 5 June 2008.

Abstract

Although the metabolic syndrome (MetS) is positively associated with high-sensitivity C-reactive protein (hsCRP), negatively associated with N-terminal pro-brain natriuretic peptide (Nt-proBNP) and inconsequently related to urine albumin/creatinine ratio (UACR) they are all associated with cardiovascular events. Therefore, we wanted to determine the influence of MetS on the predictive values of UACR, hsCRP and Nt-proBNP. On the basis of the definition of MetS by the International Diabetes Federation, a Danish population sample of 1983 apparently healthy subjects was divided into three groups: 530 subjects without any elements of MetS, 1093 subjects with some elements of MetS and 360 subjects with MetS. During the following 9.5 years the composite end point of cardiovascular death, non-fatal myocardial infarction or stroke (composite cardiovascular end point, CEP) occurred in 204 subjects. In Cox-regression analyses adjusting for age, gender and smoking, all three cardiovascular risk markers predicted CEP independently of MetS. Despite no significant interaction with MetS, high log(hsCRP) was associated with CEP primarily in subjects without any elements of MetS (hazard ratio (HR)=4.5 (1.5–14.0), P<0.01), log(Nt-proBNP) primarily in subjects with some elements of MetS (HR=3.0 (1.6–5.6), P<0.01), and logUACR independently of elements of MetS. Pre-specified gender-adjusted (men/women) cutoff values of hsCRP6.0/7.3 mg l^-1 predicted CEP in subjects without elements of MetS with positive and predictive values of 11.5 and 98%, respectively. UACR0.73/1.06 mg mmol^-1 predicted CEP in subjects with MetS with positive and predictive values of 23.5 and 93%, respectively. In apparently healthy subjects, high hsCRP was associated with CEP primarily in subjects without MetS, high Nt-proBNP in subjects with elements of MetS and UACR independently of MetS.