BACKGROUNDThere have been several epidemiologic studies investigating the association between circulating levels of insulin‐like growth factor I (IGF‐I), insulin‐like growth factor binding protein 3 (IGFBP‐3), and insulin in relation to the risk of prostate carcinoma, with conflicting results. To examine this issue further, the authors conducted a nested case–control study within the Cardiovascular Health Study cohort.METHODSIn men who were diagnosed with prostate carcinoma (cases) between 1990 and 1999 (n = 174), the levels of IGF‐I, IGFBP‐3, and insulin were measured on blood samples that were obtained 1–9 years prior to diagnosis (mean, 3.4 years). Similar measurements were made on 174 male participants without prostate carcinoma (controls) who were matched to cases based on the year blood was drawn, survival until the date of diagnosis, race, and age.RESULTSRelative to the men with IGF‐I levels in the first (lowest) quartile of the distribution, the risk of prostate carcinoma for men in the second, third, and fourth (upper) quartiles were 0.77 (95% confidence interval [95% CI], 0.43–1.38), 0.73 (95% CI, 0.41–1.30), and 0.67 (95% CI, 0.37–1.25), respectively. The results were influenced little by adjustment for levels of IGFBP‐3 or, instead, by evaluating the molar IGF‐I/IGFBP‐3 ratio. An analysis that was restricted to men who had plasma prostate‐specific antigen levels < 4 ng/mL at the time of the blood draw yielded similar results. The corresponding relative risks for IGFBP‐3 were 0.91 (95% CI, 0.49–1.68), 0.47 (95% CI, 0.25–0.94), and 0.65 (95% CI, 0.35–1.20), respectively. The distribution of serum insulin levels in cases and controls were nearly identical.CONCLUSIONSThe IGF‐I level was not associated positively with the risk of prostate carcinoma; however, an increase in the IGFBP‐3 level was associated with a modest decrease in risk. Cancer 2005. © 2004 American Cancer Society.