Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6149
Address all correspondence and requests for reprints to: Peter J. Snyder, M.D., 415 Curie Boulevard, Philadelphia, Pennsylvania 19104-6149
As men age, testicular function declines gradually and moderately. The similarities between the consequences of hypogonadism due to known disease and those of hypogonadism due to aging alone suggest that the decline of testicular function with aging does have consequences.
Decline in testicular function with age
Testosterone and gonadotropins. Several studies demonstrate that the serum testosterone concentration declines with increasing age. In one cross-sectional study of 83 healthy men, the free testosterone concentration declined by age 80 to approximately 40% of that at age 20, but because the serum concentration of sex hormone binding globulin (SHBG) increased during this age span, the serum concentration of total testosterone declined only to approximately 80% of that at age 20 (1). In another cross-sectional study of 71 healthy men, aged 2690 yr, living in similar circumstances, the 33 younger men had a mean free testosterone concentration of 10.7 ± 3.4 ng/dL (SD) and the 38 older men had a mean free testosterone concentration of 5.8 ng/dL (P < 0.01) (2). Total serum testosterone concentrations in the two groups were 659 ± 201 ng/dL and 490 ± 160 ng/dL, respectively (P < 0.01). In a third cross-sectional study in 810 healthy men mostly between 50 and 84 yr old, the serum bioavailable testosterone concentration declined by approximately 5560%, but the serum total testosterone concentration did not decline at all (3).
In one longitudinal study, 39 healthy men whose mean serum total
testosterone concentration was 633 ± 22 ng/dL (SE)
when they enrolled in a study of aging in 1980 had a mean value of
464 ± 33 ng/dL in 1994 (4). In another longitudinal
study of 890 men, both total testosterone and the ratio of testosterone
to SHBG decreased with age (Fig. 1)
(5).
|
Spermatogenesis. Some studies show a preservation of spermatogenesis with increasing age, and others show a decline. In one study, spermatogenesis, as judged by ejaculated sperm, in 23 men 6088 yr old who had fathered children earlier in life was compared to 20 men 2437 yr old who had fathered children (8). The older men had higher sperm density than the younger men and equal results to the younger men in the hamster ova penetration test. However, the period of abstinence before ejaculation was much longer in the older men than in the younger men. Similarly, in a study of 189 elderly men, 51 middle aged men, and 50 young men, the serum concentration of inhibin B, a marker of Sertoli cell function and spermatogenesis, correlated negatively with age, but only weakly (Spearman correlation coefficient, -0.17) (9). However, in a histologic study of spermatogenesis in 30 men who died of trauma or myocardial infarction, spermatogenesis, as defined by spermatid nuclei, decreased with increasing age (10).
Possible consequences of the decline in testicular function
The decline in serum testosterone concentration with aging might have several consequences, based on a comparison with the consequences that occur in response to frank hypogonadism due to known disease. Men who become hypogonadal due to known disease experience a decrease in libido, energy, muscle mass, muscle strength, and bone density, and testosterone treatment reverses these effects. Because elderly men also experience the same effects, it is possible that the decrease in testosterone is a cause.
Libido. A study of 220 men ages 4193 yr old demonstrated a clear association between age and several parameters of sexual function, including sexual thoughts, enjoyment, and morning erections (11). The same investigators also demonstrated that testosterone treatment of frankly hypogonadal men with testosterone markedly improved the same parameters of sexual function (12). A more recent study confirmed that physiologic testosterone treatment markedly improves sexual function in frankly hypogonadal men within 3 months and maintains the effect for 3 yr (13).
Sarcopenia. Both cross-sectional (14) and longitudinal (15) studies document the loss of muscle mass, and concomitant increase in fat mass, as a part of the normal aging process. In the New Mexico Elder Health Survey, the prevalence of sarcopenia, defined as appendicular skeletal mass less than 2 SD below the mean in a young reference group, increased from 13.5% in non-Hispanic white men less than 70 yr of age to 52.6% in those more than 80 yr old (16). The prevalence of sarcopenia in the elderly of Olmstead County, MN, was similar (17). Hypogonadism has the same effect, and testosterone treatment reverses it. In a study of six healthy young men, treatment with two doses of leuprolide 3 weeks apart lowered the serum testosterone concentration from 535 ± 141 ng/dL to 31.0 ± 9.0 ng/mL, decreased fat-free mass from 56.5 ± 2.9 kg to 54.4 ± 2.5 kg (P < 0.005), and increased fat mass from 15.8 ± 1.9 kg to 16.9 ± 2.0 kg (P < 0.05) (18). Treatment of men who had hypogonadism due to known disease increased fat-free mass in two studies (13, 19).
Muscle strength. Muscle strength in men also decreases with increasing age. A recent study of 346 men aged 2093 in the Baltimore Longitudinal Study of Aging showed a decrease in the strength of knee extensors with increasing age (20). Another recent study showed that muscle power similarly decreases with increasing age (21). An effect of hypogonadism on muscle strength has been demonstrated less consistently. Decreasing the serum testosterone concentration drastically by treating normal young men with leuprolide (18), as described above, did decrease muscle strength, and increasing the serum testosterone concentration to normal in men with acquired immunodeficiency syndrome wasting and hypogonadism increased muscle strength (22). However, treating men with testosterone who were hypogonadal due to other etiologies did not increase their muscle strength (13).
Osteoporosis. Bone mineral density decreases gradually with
increasing age in men (Fig. 2)
(23). However, cross-sectional studies in elderly men do
not show a correlation between the serum testosterone concentration and
bone mineral density (24) or bone fractures
(25), but do show a direct correlation between the serum
estradiol concentration and these two parameters (24, 25).
However, lowering the serum testosterone concentration by castration,
which also lowers the estradiol, does lead to a decrease in bone
mineral density (26). Moreover, testosterone treatment of
men who are hypogonadal due to known disease is accompanied by a clear
increase in bone mineral density. In one study, 29 hypogonadal men who
were treated with 100 mg testosterone enanthate once a week for 18
months experienced an increase in bone mineral density of the lumbar
spine by 5% (19). In another study, 18 men who were
treated with testosterone transdermally for 3 yr experienced a 7.7%
increase in bone mineral density of the spine and a 4.0% increase in
that of the femoral trochanter (13).
|
To test the hypothesis that increasing the serum testosterone
concentration of elderly men to that of young men would reverse some of
the effects described above, 108 men more than 65 yr old whose serum
testosterone concentrations were less than mean - SD
for young men were randomized to receive either testosterone or placebo
transdermally for 3 yr. Men in the testosterone-treated group had a
mean serum testosterone concentration of 625 ± 249
(SD) ng/dL during treatment, and those in the
placebo-treated group had a value of 369 ± 75 ng/dL
(27). Bone mineral density of the lumbar spine increased
significantly (P < 0.001) in both groups, 2.5 ±
0.6% in the placebo-treated group and 4.2 ± 0.8% in the
testosterone-treated group. Although the difference between the 2
groups was not statistically significant overall, linear regression
analysis demonstrated a significant (P = 0.02) negative
relationship between the pretreatment serum testosterone concentration
and the testosterone treatment effect on lumbar spine bone mineral
density (Fig. 3). That is, the lower the
pretreatment serum testosterone, the greater the testosterone treatment
effect on bone mineral density.
|
|
Several parameters were monitored because they reflect potential deleterious effects of testosterone (27). Men treated with testosterone experienced 16 prostate events (such as a persistent increase in prostate-specific antigen, prostate cancer, urinary obstruction) during the 3 yr of treatment, compared with 11 experienced by the men treated with placebo. This difference was not statistically significant, but the lack of statistical significance needs to be interpreted cautiously, because the study was not designed to have adequate statistical power to assess this parameter. Prostate-specific antigen increased significantly, but within the normal range, in the testosterone-treated group but not in the placebo-treated group. Hematocrit also increased significantly, but within the normal range, in the testosterone-treated group. There was no difference between the two treatment groups in the respiratory distress index or any lipid or apolipoprotein parameter.
These results suggest that physiologic testosterone replacement of elderly men does increase bone mineral density of the spine, but only in men who are hypogonadal by the same standards used for young men, and does increase fat-free mass and decrease fat mass. It is still unclear whether testosterone replacement will increase muscle strength. Furthermore, it is unknown at this time whether testosterone replacement of elderly, hypogonadal men will increase the frequency of testosterone-dependent diseases, such as prostate cancer and benign prostatic hyperplasia.
Conclusions
Even though testosterone treatment of elderly men who are hypogonadal appears to improve bone density and body composition, the risks of this treatment are not yet known. Consequently, physicians should be cautious in treating such men with testosterone. At this time, treatment should be reserved for men who are severely hypogonadal, and men who are treated should be monitored carefully for potential deleterious effects.
Footnotes
1 Supported by grants from the NIH (AR41425 and AG10954 to P.J.S. and
RR040 to the General Clinical Research Center of the University of
Pennsylvania).
Received February 16, 2001.
Revised March 18, 2001.
Accepted March 19, 2001.
References