No association between pre-diagnostic plasma C-reactive protein concentration and subsequent prostate cancer

BACKGROUND. We evaluated the association of pre-diagnostic plasma concentration of C-reactive protein, a sensitive, but non-specific indicator of inflammation, with subsequent risk of prostate cancer. METHODS. Included were 264 histologically confirmed prostate cancer cases and 264 age-matched controls who were participants in the CLUE II cohort of Washington County, MD. C-reactive protein was measured using a high-sensitivity immunoturbidimetric assay. Odds ratios (ORs) and 95% confidence intervals (CI) were estimated from conditional logistic regression models by fourth of the C-reactive protein distribution with cutpoints based on the controls. The median concentrations from the lowest to highest fourth were 0.41,1.03,1.90, and 4.53 mg/L. RESULTS. Geometric mean plasma concentrations did not differ between the cases (1.24 +/- 2.94 mg/L) and controls (1.41 +/- 2.97 mg/L; P = 0.16). Compared to the bottom fourth, the ORs (95% CI) of prostate cancer were 1.29 (0.80 - 2.08), 0.98 (0.61 - 1.58), and 0.95 (0.57 - 1.58) for the second, third, and highest fourths (P trend = 0.66). These findings were unchanged after adjusting for body mass index (BMI) and cigarette smoking status or after excluding men with markedly elevated C-reactive protein, cases diagnosed during the first 2 years of follow-up, or controls who never had a PSA test. These findings did not differ by stage or grade of prostate cancer. CONCLUSIONS. Pre-diagnostic plasma concentration of C-reactive protein was not associated with subsequent risk of prostate cancer. (C) 2003 Wiley-Liss, Inc.