Background: C-reactive protein (CRP), a nonspecific marker of the inflammatory status, is associated with cardiovascular disease risk factors and may be an important feature of the metabolic syndrome (MSX) in middle-aged subjects. Objectives: We assessed the relationship of CRP levels to specific components of MSX and other potential determinants in apparently healthy elderly subjects living in the South of France. Methods: In the framework of the population-based POLA (Pathologies Oculaires Li? ?'Age) Study, performed in 2,404 subjects aged 60 years or more, we measured the plasma CRP levels. All subjects with known systemic inflammatory diseases, such as chronic bronchitis, cardiovascular disease, and diabetes, and those who were on systemic steroid therapy as well as subjects with CRP levels >10 mg/l were excluded from the study, leaving 1,709 subjects for the statistical analyses. MSX was defined according to NCEP (National Cholesterol Education Program) criteria. Other potential determinants were assessed through interviewer-based questionnaire. Results: We grouped the subjects into three categories based on the 75th and 25th percentiles, corresponding to 3.05 and 0.82, respectively. We compared subjects in the highest quartile, i.e., with CRP >/=3.05 mg/l, with those in the two intermediate quartiles, i.e., with 0.82 < CRP < 3.05, and those in the lowest quartile, i.e., with CRP <0.82 mg/l according to gender. MSX, which had a prevalence of 31%, was significantly associated with elevated CRP levels. Among MSX components, the strongest positive association with the highest quartile of CRP was with waist circumference in males as well as in females (age-adjusted odds ratio OR 3.06 and 95% confidence interval CI 1.82-5.14; OR 7.04 and 95% CI 4.79-10.34, respectively). Each component of the MSX, such as abnormal fasting plasma glucose (OR 2.90, 95% CI 1.69-4.99), triglycerides (OR 1.96, 95% CI 1.30-2.96), high-density lipoprotein cholesterol (OR 2.31, 95% CI 1.61-3.30), and blood pressure (OR 1.66, 95% CI 1.12-2.45), was significantly associated with high CRP values in elderly women only. In men, only current smoking was significantly associated with high CRP levels (OR 1.52, 95% CI 1.04-2.2). In multivariate analysis, the waist circumference remained significantly associated with high CRP levels, with a graded effect of CRP quartile whatever the gender. In men, current and former smoking remained significantly associated with the CRP levels. In women, the association observed in univariate analysis with fasting glucose or hypertension did not reach statistical significance in the multivariate analysis, while only a weak association could be observed with lipid parameters such as triglycerides and high-density lipoprotein cholesterol. Conclusions: Abdominal adiposity adds to the variance in plasma CRP levels in elderly patients with MSX. This suggests that weight loss or other interventions targeted at adipocyte-related inflammation may represent an important means to prevent subclinical inflammation in the elderly, bearing a high risk of cardiovascular disease. Copyright (c) 2007 S. Karger AG, Basel.