To the best of our knowledge our analysis is the first population-based study that shows an association between circulating concentrations of androgens and renal function in men. The most consistent correlations were revealed between DHEA-S and renal function in lean subjects and between total testosterone and glomerular filtration estimate in non-lean subjects. Most importantly, the independent associations between sex steroids and renal function were demonstrated in a young, apparently healthy male population largely unaffected by hypertension, without any history of overt cardiovascular disease.
Our data provide a clinical context to the reported experimental investigations suggestive of the role of androgens in renal injury. Indeed, key enzymatic regulators of androgen biosynthesis and signaling are expressed in the renal tissue.26,27
Furthermore, data from rodent models directly demonstrate the negative impact of testosterone on renal structure and function in male rats26
and significant improvement in renal parameters after testosterone depletion (castration).28
Mechanistically, the adverse effects of androgens on renal outcomes is most likely mediated via activation of the renin–angiotensin–aldosterone system, oxidative stress, proapoptotic and proinflammatory processes,26,27
all of which are well-known drivers of cardiovascular and metabolic disorders. The inverse association between testosterone and renal function apparent only in nonlean men in our study suggests a permissive contribution of increased body weight to the androgen–renal relationship. Overweight and obesity are indeed increasingly recognized as risk factors for kidney diseases.29
In addition, increased body weight correlates with oxidative stress30
and low-grade inflammation30
—the two most likely intermediate phenotypes on the androgens–renal axis.26
Whether interactions between overweight and testosterone influence renal function via promotion of prooxidative and proinflammatory phenotype in men remains to be elucidated.
Our study also shows an association between renal function and DHEA-S—the most abundant circulating androgen31
that has been linked to atherosclerosis,32
and endothelial dysfunction.34
Given that renal function declines with aging and that endothelial injury is a well-recognized factor underlying kidney damage,35
association between DHEA-S and creatinine clearance is biologically plausible. Nonetheless, the direction of the association between DHEA-S and cardiovascular and renal phenotypes remains controversial. The majority,36
but not all,37
data suggest a protective effect of endogenous DHEA-S on the well-documented risk factors of renal injury (such as BP and lipids) and a possible negative impact of low DHEA-S on cardiovascular mortality.36
Low circulating concentrations of DHEA-S were also implicated as a risk factor for progression of glomerular injury in diabetic men.38
However, evaluation of relationships between DHEA-S and renal function was not examined directly in most of these studies. In addition, most of the reported associations36
come from cohorts of middle-aged or elderly subjects and as such may be confounded by coexisting disorders or in fact aging—recognized inverse correlates of DHEA-S plasma levels. Given that circulating concentrations of DHEA-S peak at 20–24 years in men and only thereafter decline with age39
our analysis based on young, apparently healthy subjects is unlikely to be confounded by strong inverse relationships between DHEA-S and aging or cardiovascular diseases.
Our study has a number of limitations. First, the analysis is based on calculated glomerular filtration rather than a direct measurement of renal hemodynamics and as such represents only a surrogate of genuine filtration rates. Second, lack of available sex hormone–binding globulin and albumin levels does not permit us to calculate free and bioavailable fractions of testosterone. Therefore, we cannot exclude that using free rather than total hormonal fraction of testosterone would provide even better estimate of its association with renal function. Third, due to unavailability of urine samples in the YMCA study we could not evaluate associations between androgens and urinary excretion of albumin (a well-recognized marker of renal injury) or verify the extent to which renal clearance of sex steroids contributes to the detected associations. We should also stress that although our data do not support the role of dihydrotestosterone as a mediator of negative effects of androgens on renal function in men, they should be interpreted with caution given that circulating concentrations of dihydrotestosterone may not reflect accurately its bioavailability within peripheral tissue, including the kidney.40
Future studies should focus on paracrine functions of dihydrotestosterone in the kidney bringing us closer to the full dissection of associations between androgens and renal function. It should be also acknowledged that our analysis was not corrected for multiple testing. However, consistency of associations between renal function and both androgens (in different strata of BMI distribution) in the crude, adjusted and factor analysis suggests a low probability of type 1 error. Finally, lack of prospective observation does not permit to assign a causal context to the demonstrated findings and provide an immediate mechanistic explanation for the observed associations. Further investigations are warranted to elucidate fully the physiological background of relationships between androgens and renal function in men.
Within the interpretational limitations discussed above our data may suggest that androgens are inversely associated with male renal function early in life. Future prospective studies are needed to confirm whether these associations may help to explain the well-documented epidemiological trends in higher incidence and progression of nondiabetic kidney diseases among men than women.