Abstract 3614: Are circulating testosterone and PSA levels associated in a nationally representative sample of men without a diagnosis of prostate cancer

• Published in: Cancer research (Chicago, Ill.) Ill.), 2013-04, Vol.73 (8_Supplement), p.3614-3614
• Main Authors: Peskoe, Sarah B., Nelson, William G., Joshu, Corinne E., Rohrmann, Sabine, McGlynn, Katherine A., Platz, Elizabeth A.
• Format: Article
• Language: English
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2013-3614

Background: PSA-based prostate cancer screening is controversial, in part, because of imperfect specificity, especially for aggressive disease. PSA production by prostate luminal epithelial cells is under androgenic regulation. However, the association between circulating androgen and PSA levels is unclear in general populations. Knowledge of this association may enhance clinical decision-making for an elevated screening serum PSA. Methods: Included were 378 men, 40-85 years who participated in the National Health and Nutrition Examination Survey in 2001-04, who did not have a prior prostate cancer diagnosis, and who had not had a recent biopsy, exam, or infection of the prostate. Serum levels of PSA (Hybritech), testosterone, androstanediol glucuronide (AAG), estradiol, and sex hormone binding globulin (SHBG; carries testosterone and estradiol in circulation) (immunoassay) were previously measured. Free testosterone was estimated by mass action. We calculated geometric mean PSA levels and 95% confidence intervals by hormone quintiles. We applied sampling weights, and adjusted for age and race/ethnicity, and also for BMI, waist circumference, smoking, diabetes, and mutually for the hormones. We stratified by age, race/ethnicity, and adiposity. We estimated the OR of a mildly elevated PSA (≥2.5 ng/mL) per hormone quintile by logistic regression. Results: Geometric mean PSA level increased across testosterone quintiles after age and race/ethnicity adjustment (Q1: 0.80, Q5: 1.14 ng/mL; p-trend=0.001) and after further multivariable adjustment (Q1: 0.80, Q5: 1.15 ng/mL; p-trend=0.022). The same patterns were observed for free testosterone and AAG. SHBG was not associated with PSA after age and race/ethnicity adjustment (p-trend=0.85), but was strongly inversely associated after multivariable adjustment (Q1: 1.30, Q5: 0.81 nmol/L; p-trend=0.007). Estradiol and PSA levels were not associated. While PSA levels differed by age, by race/ethnicity, and by adiposity, the same sized increasing association for testosterone and PSA was seen within strata of age, of race/ethnicity, and of adiposity. The OR of PSA ≥2.5 ng/mL per testosterone quintile was 1.52 (95% CI 1.16-1.98) after age and race/ethnicity adjustment, and 1.74 (95% CI 1.13-2.68) after multivariable adjustment; SHBG explained the OR shift. The multivariable-adjusted OR of PSA ≥2.5 ng/mL per SHBG quintile was 0.60 (95% CI 0.37-1.00). Conclusions: In this nationally representative sample, men with higher tes