The prevalence of late-onset hipogonadism (LOH) in men is not accurately determined and may be overestimated. Androgen deficiency is causally linked to insulin resistance, regardless of age. Clinical signs of androgen deficiency are not uniquely specific for LOH. The aim of this study was to assess the correlations between androgen serum levels and clinical symptoms of andropenia depending on the age. We evaluated the relationships between the frequency of LOH and metabolic syndrome (MS) and compared different methods of androgen assessment in the context of their clinical relevance.
Material and methods: In 153 randomly selected men aged 40 to 70 years adrenal (DHEA and DHEAS) and gonadal (total-TT and free testosterone-FT) androgens were evaluated and compared with free (c-FT) and bioavailable testosterone (c-BAT) calculated with a mathematical algorithm. European Male Aging Study criteria for the diagnosis of LOH were used.
Results: TT concentration (r = -0.074, p = 0.363) and FT (r = -0.054, p = 0.505) did not correlate with age, in contrast to the c-FT (r = -0.280, p = 0.0005 ), c-BAT (r = -0.297, p = 0.0002) and both adrenal androgens: DHEA (r = -0.318, p = 0.00001) and DHEAS (r = -0.506, p < 0.00001). The clinical signs suggesting andropenia were found in 31.4% of participants and androgen deficiency in 28.8% of men. The incidence of these signs increased significantly with the age of respondents. In contrast, the criteria of LOH in the entire group were fulfilled by only 6.5% of men. Men diagnosed with MS showed significantly lower TT plasma levels: 5.05 (3.70-6.45) vs. 6.30 (5.40-8.30), p < 0.0001--contrary to the FT, c-FT, c-BAT as well as DHEA/DHEAS serum levels, regardless of age. Men meeting the full criteria for LOH (clinical symptoms and laboratory-confirmed hipoandrogenemia) revealed more than a five-fold higher risk of metabolic syndrome in relation to others (OR = 5.533, 95% CI: 1.134-27.+ 01). Total, free, calculated-free and calculated-bioavailable testosterone serum concentrations were strongly correlated with each other. Similarly, the plasma concentrations of both adrenal androgens were mutually positively correlated (r = 0.417, p < 0.0001), but any correlation between the gonadal androgens and DHEA/DHEAS were revealed.
Conclusions: The prevalence of isolated hipoandrogenemia and clinical symptoms suggestive andropenia is clearly greater than the actual frequency of LOH. Fulfilling the criteria for LOH significantly increases the risk of MS.