Association between PSA and leptin, adiponectin, HbA1c, or C-peptide among African-American and Caucasian men

A33

African-American (AA) men are more likely to be diagnosed with advanced prostate cancer and have a poorer prognosis following treatment compared to Caucasian (CA) men. Multiple factors are probably involved, with racial differences in prostate cancer screening and detection practices contributing to observed disparities. Interestingly, obese men are also more likely to be diagnosed with late-stage prostate cancer, and several prior studies report a higher body mass index (BMI) is associated with lower PSA levels. However the mechanisms by which obesity may affect PSA levels among AA or CA men remain unclear. Thus, the purpose of this study is to investigate the association between PSA and biomarkers of obesity, including blood HbA1c, C-peptide, leptin, and adiponectin levels. Participants included a random sample of 121 AA and 121 CA men recruited as a part of the Southern Community Cohort Study. Eligible men were between 40 and 79 years of age, without a prior cancer or diabetes diagnosis, and not currently using finasteride or other medication known to affect PSA. Furthermore, AA and CA participants were frequency matched to have a similar BMI distribution. PSA values were natural log transformed prior to analysis, and age-adjusted geometric mean PSA values within quartile categories of each biomarker are reported. Among AA men, PSA levels were approximately 50% lower among men with higher C-peptide levels (PSA=0.99, 0.93, 0.75, 0.53 ng/ml across increasing quartiles of C-peptide, respectively, p_trend =0.005). This inverse trend among AA men was evident across BMI categories, although this association was statistically significant only among participants with a BMI of 30 or more. Among CA men, PSA levels were approximately 50% lower among CA men with the highest HbA1c level (PSA=0.84, 0.73, 0.77, 0.45 ng/ml across quartiles of HbA1c, respectively, p_trend =0.005). PSA was not significantly associated with leptin or adiponectin levels within either race group. To further evaluate the independence of these associations, we included HbA1c, leptin, C-peptide, and adiponectin in a single model that also controlled for age, time since last meal, and BMI. Among CA men, HbA1c was the only biomarker significantly associated with log transformed PSA levels (b=-0.27, 95% CI(-0.45, -0.08), p=0.006). Among AA men, C-peptide was the only biomarker significantly associated with log transformed PSA (b=-0.15, 95% CI(-0.25, -0.05), p=0.003). These results suggest that metabolic disturbances related to the metabolic syndrome, insulin activity, or diabetes affect the ability to detect early-stage prostate cancer. Further investigations of the race-specific relationships between steroid hormone levels, insulin activity, obesity, and PSA are warranted.