Middle-aged men secrete less testosterone at night than young healthy men

Aging men largely maintain their testicular androgen production. Cross-sectional studies have demonstrated that after the age of 40 yr a 0.2-2% annual decline is observed in morning total testosterone. In elderly males, the coordinate release of LH and testosterone became asynchronous despite normal serum levels of these hormones. The aim of this study was to test the reproductive hormone rhythm at night in middle-aged men. We studied seven healthy middle-aged (46.6 +/- 6.7 yr) and six healthy young (23.9 +/- 2.4 yr) men by determining their serum levels of LH and testosterone levels every 15 min from 1900-0700 h with simultaneous sleep recordings. The nocturnal rise in testosterone occurred earlier in young men (2235 +/- 0022 h) and at 2331 +/- 0057 h in middle-aged men (P < 0.04). In young men, the mean testosterone level at night (5.0 +/- 1.3 ng/ml; 17.4 +/- 4.4 nmol/liter) and the integrated nocturnal secretion [area under the curve (AUC); 60.6 +/- 8.9 ng/ml.h; 210 +/- 31 nmol/liter.h] were significantly higher compared with the values (3.6 +/- 1.1 and 31.1 +/- 7.2 ng/ml.h; 12.6 +/- 3.8 and 108 +/- 24.8 nmol/liter.h, respectively) observed in middle-aged men (P < 0.04 and P < 0.01, respectively). The mean (3.5 +/- 0.3 mIU/ml; 3.5 +/- 0.3 IU/liter) and AUC (43.4 +/- 8.3 mIU/ml.h; 43.4 +/- 8.3 IU/liter.h) LH values in middle-aged men were significantly higher than the values observed in young men (2.0 +/- 0.7 and 30.8 +/- 6.1 mIU/ml.h; 2.0 +/- 0.7 and 30.8 +/- 6.1 IU/liter.h; P < 0.05 and P < 0.01, respectively). Young men had significantly more testosterone pulses at night (6.7 +/- 1.6/12 vs. 3.8 +/- 1.1/12 h in middle-aged men; P < 0.005) of shorter interpulse interval (88.5 +/- 23.6 vs. 137.4 +/- 46.4 min; P < 0.02). LH pulse characteristics and sleep quality were similar in both groups. However, the first rapid eye movement (REM) sleep episode occurred earlier in middle-aged men (2303 +/- 0034 h) vs. young men (0010 +/- 0054 h; P < 0.04). As a consequence, the testosterone rise antedated the first REM episode by 90 min in young men. The link between testosterone rise and REM sleep episode was not observed in middle-aged men. Linear regression analysis revealed that the LH AUC was significantly related to age (P < 0.02). Analysis of covariance revealed that the two groups differed significantly in testosterone AUC (P < 0.04). Comparison of LH and testosterone concentrations showed significant and positive cross-correlations between LH and testosterone only in young men, with the testosterone rise lagging 60 min after the rise in LH. Our findings suggest that in middle-aged men, less pulsatile testosterone and more LH are secreted at night than in young men, with disruption of the association between testosterone rhythm and REM sleep. The decline in nocturnal testosterone secretion appears to involve a combination of testicular and pituitary hypogonadism.