LAUSR

Response by Chang et al to Letter Regarding Article, “Prostate-Specific Antigen Within the Reference Range, Subclinical Coronary Atherosclerosis, and Cardiovascular Mortality” In Response: We thank Cheng et al for their interest in our study and for addressing important issues in interpreting our results. To your first point, we investigated the association between prostate-specific antigen (PSA) within the reference range and cardiovascular disease (CVD) using 2 different study designs: (1) a cross-sectional study to test the association between PSA and coronary artery calcification (CAC), an established marker of subclinical atherosclerosis and (2) a cohort study to test the association between PSA and cardiovascular mortality. We found that the PSA level within the reference range was inversely associated with both CAC and cardiovascular mortality in a dose-response manner. As Cheng et al pointed out, cross-sectional studies can be limited by ambiguous temporality or reverse causation between exposure and outcome. However, we studied healthy, asymptomatic, young men free of clinically manifested CVD; thus, our findings are unlikely to be affected by reverse causation. Furthermore, the association we observed between PSA level and CAC was consistent with the findings from our additional cohort study of the association between PSA level and cardiovascular mortality. However, we agree that future cohort studies are needed to determine the temporal relationship between serum PSA level and subclinical atherosclerosis. To your second point, obesity, hypertension, diabetes mellitus, and metabolic parameters, which are all risk factors of CVD, increased as PSA level decreased, and those associations between PSA and cardiovascular risk factors could confound the association between PSA level and cardiovascular mortality. However, after adjustment for the potential confounders and cardiovascular risk factors (hypertension, diabetes mellitus, body mass index, and metabolic parameters), the inverse association between serum PSA level and cardiovascular mortality remained significant, indicating that those CVD risk factors did not fully explain the association between PSA and cardiovascular mortality. To your third point, there were typos regarding the P value for trend in multivariable-adjusted model 3 from the multinomial logistic regression analyses, shown in Table 3. All P values for trend should have been <0.001, and that mistake has been corrected. Specifically, the multivariable-adjusted prevalence ratios comparing PSA quartiles 2, 3, and 4 to the lowest quartile were 0.95 (0.90–1.01), 0.88 (0.82–0.93), and 0.84 (0.79–0.89), respectively, for a CAC score range from 1 point to 100 points and 1.00 (0.87–1.14), 0.87 (0.76–0.99), and 0.82 (0.72–0.93) for a CAC score range >100 points (P for trend <0.001). Finally, we agree that multiple biomarkers reflecting different pathophysiological aspects of CVD, including inflammation and thrombosis, could help to elucidate the mechanisms underlying the association between PSA level and CVD.2,3 Additionally, further research involving actual measurement of testosterone and other sex hormones in relation to serum PSA concentration is needed to elucidate the mechanisms underlying the relationships between PSA, androgen activity, and CVD.