OBJECTIVE: To evaluate serum sex steroid hormone concentrations and long-term risk of subsequent lower urinary tract symptoms (LUTS) in a cohort of community-dwelling older men. SUBJECTS AND METHODS: Between 1984 and 1987, serum sex hormone concentrations were measured in participants in the Rancho Bernardo Study, a prospective, community-based study. In 2006, the American Urological Association Symptom Index (AUA-SI) was mailed to surviving male participants. Logistic regression was used to examine associations of baseline hormone concentrations with AUA-SI. RESULTS: Among 158 surviving men with complete data and no history of prostate cancer, the mean (sd) age at serum sex steroid assessment was 58 (6.6) years with a mean (sd) follow-up of 20.3 (0.6) years. In age-adjusted logistic regression, there was a significant inverse association of testosterone : dihydrotestosterone (DHT) with LUTS (P = 0.05). Also, men with higher concentrations of bioavailable testosterone had a 56% decreased risk of LUTS compared with those with hypogonadal concentrations, although the association was not statistically significant (odds ratios 0.44, 95% confidence interval 0.14-1.40) or distributed evenly among quartiles. There were no significant associations of total testosterone, oestradiol (E(2)), testosterone : E(2), DHT, or dehydroepiandrosterone with LUTS or with any measured hormones and urinary bother. CONCLUSIONS: In this cohort, men with higher mid-life levels of testosterone : DHT and bioavailable testosterone had a decreased 20-year risk of LUTS. These data support other studies reporting inverse associations of serum testosterone with LUTS. Clinical trials of testosterone therapy should include LUTS and clinical benign prostatic hyperplasia as outcomes.
OBJECTIVE: To evaluate serum sex steroid hormone concentrations and long-term risk of subsequent lower urinary tract symptoms (LUTS) in a cohort of community-dwelling older men. SUBJECTS AND METHODS: Between 1984 and 1987, serum sex hormone concentrations were measured in participants in the Rancho Bernardo Study, a prospective, community-based study. In 2006, the American Urological Association Symptom Index (AUA-SI) was mailed to surviving male participants. Logistic regression was used to examine associations of baseline hormone concentrations with AUA-SI. RESULTS: Among 158 surviving men with complete data and no history of prostate cancer, the mean (sd) age at serum sex steroid assessment was 58 (6.6) years with a mean (sd) follow-up of 20.3 (0.6) years. In age-adjusted logistic regression, there was a significant inverse association of testosterone : dihydrotestosterone (DHT) with LUTS (P = 0.05). Also, men with higher concentrations of bioavailable testosterone had a 56% decreased risk of LUTS compared with those with hypogonadal concentrations, although the association was not statistically significant (odds ratios 0.44, 95% confidence interval 0.14-1.40) or distributed evenly among quartiles. There were no significant associations of total testosterone, oestradiol (E(2)), testosterone : E(2), DHT, or dehydroepiandrosterone with LUTS or with any measured hormones and urinary bother. CONCLUSIONS: In this cohort, men with higher mid-life levels of testosterone : DHT and bioavailable testosterone had a decreased 20-year risk of LUTS. These data support other studies reporting inverse associations of serum testosterone with LUTS. Clinical trials of testosterone therapy should include LUTS and clinical benign prostatic hyperplasia as outcomes.