Associations of endogenous androgens and sex hormone-binding globulin with kidney function and chronic kidney disease

Abstract

Introduction: The role of endogenous androgens in kidney function and disease has not been extensively explored in men and women.

Research design and methods: We analyzed data from the observational KORA F4 study and its follow-up examination KORA FF4 (median follow-up time 6.5 years) including 1293 men and 650 peri- and postmenopausal women, not using exogenous sex hormones. We examined the associations between endogenous androgens (testosterone [T], dihydrotestosterone [DHT], free T [fT], free DHT [fDHT], and T/DHT), with estimated glomerular filtration rate (eGFR) at baseline and follow-up, prevalent, and incident chronic kidney disease (CKD) adjusting for common CKD risk factors.

Results: At baseline, 73 men (5.7%) and 54 women (8.4%) had prevalent CKD. Cross-sectionally, no significant associations between androgens and kidney function were observed among men. In women, elevated T (β=-1.305, [95% CI -2.290; -0.320]) and fT (β=-1.423, [95% CI -2.449; -0.397]) were associated with lower eGFR. Prospectively, 81 men (8.8%) and 60 women (15.2%) developed incident CKD. In women, a reverse J-shaped associations was observed between DHT and incident CKD (P_non-linear=0.029), while higher fDHT was associated with lower incident CKD risk (odds ratio per 1 standard deviation=0.613, [95% CI 0.369; 0.971]. Among men, T/DHT (β=-0.819, [95% CI -1.413; -0.226]) and SHBG (P_non-linear=0.011) were associated with eGFR at follow-up but not with incident CKD. Some associations appeared to be modified by type 2 diabetes (T2D).

Conclusion: Suggestive associations are observed of androgens and SHBG with kidney impairment among men and women. However, larger well-phenotyped prospective studies are required to further elucidate the potential of androgens, SHBG, and T2D as modifiable risk factors for kidney function and CKD.

Keywords: chronic kidney disease; dihydrotestosterone (DHT); kidney function; sex hormone-binding globulin (SHBG); testosterone (T); type 2 diabetes.

Figure 1

Exclusion flowchart.

Figure 2

Restricted cubic splines describing non-linear associations between androgens or SHBG with kidney function in men and women of the KORA F4/FF4 study. Non-linear associations between (A) ln(SHBG) and follow-up eGFR in men, plus (B) between DHT and incident CKD in women. Non-linearity was investigated by introducing a quadratic term to the models for all exposures. Interaction between diabetes and DHT or SHBG was assessed using a 3-way interaction (i.e. DHT or SHBG*diabetes + (DHT or SHBG)²*diabetes). Non-linear associations were not modified by diabetes (P_interaction: (A) 0.920 and (B) 0.948). Solid line represents the estimated spline function for follow-up eGFR and incident CKD, and the grey shaded area represents the respective 95% CI spline estimation. The histograms provide insight to the population density along ln(SHBG) and DHT values. Models were adjusted using Model 2, comprising of baseline eGFR, baseline age, waist circumference, systolic blood pressure, usage of antihypertensive medication, non-HDL cholesterol, prevalent cardiovascular disease, usage of lipid-lowering medication, smoking status, physical activity, alcohol consumption, diabetes status, and ln(CRP). CKD: Chronic kidney disease, CI: Confidence interval, CRP: C-reactive protein, DHT: Dihydrotestosterone, eGFR: Estimated glomerular filtration rate, HDL: High-density lipoprotein, OR: Odds ratio, SD: Standard deviation.