Objective: To examine the relationship between different measures of testosterone and estradiol (E 2 ), muscle mass, muscle strength, and physical performance; and to test whether the association of sex hormone level with muscle strength and physical performance was independent of muscle mass. Design and methods: A cross-sectional survey on 1489 community-dwelling men older than 64 years of age. Serum levels of testosterone and E 2 were measured by mass spectrometry, and sex hormonebinding globulin (SHBG) levels were measured by immunoradioassay. Muscle mass was examined by dual-energy X-ray absorptiometry and physical performance was assessed by hand-grip strength, gait speed, step length and chair-stand test. Results: Appendicular skeletal mass (ASM) was positively associated with total testosterone (TT; P!0.001), free testosterone (FT; P!0.001), and total E 2 (P!0.001) but not with free E 2 (PZ0.102). After adjustment for age, serum SHBG and relative ASM, both TT and FT were significantly associated with grip strength, narrow-walk speed and the composite neuromuscular score. Higher total E 2 , but not free E 2 was associated with lower grip strength (P!0.05) after adjustment for age, FT, SHBG and relative ASM. Conclusions: Testosterone level was related to both muscle mass, strength and physical performance. Total E 2 level, though related to muscle mass positively, affected muscle strength adversely in older men.
European Journal of Endocrinology 164 811-817
IntroductionSeveral studies have examined the inter-relationship between various measures of testosterone level, muscle mass and muscle strength (1-6). In men aged 65 or older, Baumgartner et al.(1) reported that free testosterone index, but not total testosterone (TT) level, was modestly associated with muscle mass, whereas Iannuzzi-Sucich et al. (2) demonstrated that muscle mass was correlated to bioavailable testosterone (BioT), leg strength and physical performance. However, in another cohort of healthy older men, Van den Beld et al. (3) reported that testosterone level, as represented by TT and BioT, was related to muscle strength but not to muscle mass. On the other hand, Roy et al. (5) observed that TT and BioT were positively correlated with both the muscle mass and strength, yet after adjustment for muscle mass, they were no longer independently associated with muscle strength. Szulc et al. (4) had compared the muscle mass and physical performance of hypogonadal and eugonadal men and found that hypogonadal men exhibited impaired static and dynamic balance despite relatively preserved muscle mass. Similarly, Orwoll et al. (6) have observed that testosterone level was related to physical performance and fall risk but not to muscle mass. Although the effect of testosterone supplementation on muscle mass and strength was less controversial in interventional studies (7-14), whether endogenous testosterone was related to muscle mass alone, muscle strength alone, or both; and which measures of endogenous testosterone best related to muscle ...