Association of total and calculated free testosterone with androgen deficiency symptoms in patients with type 2 diabetes

Abstract

The Androgen Deficiency in the Aging Male (ADAM) questionnaire is commonly used to screen type 2 diabetes mellitus (T2DM) patients for androgen deficiency symptoms, but the association of low total (TT) and free testosterone (FT) levels with divergent responses to the ADAM questionnaire remains unclear. The aim of this study was to assess the predictive ability of TT and calculated FT (cFT) levels for accurately classifying ADAM status. We recruited 70 patients each with positive (group A) and negative (group B) responses to the ADAM questionnaire and 70 age-matched healthy controls (group C) (mean age, 50.5 years); serum levels of TT were estimated and cFT were estimated using Vermeulen equation. Hypogonadism was defined as the presence of symptoms (positive ADAM score) along with TT level < 300 ng/dL or calculated (cFT) <6.35 ng/dL. BMI was highest in group A (P < 0.05), followed by groups B and C. Group A had longer diabetes durations (P < 0.05) and higher fasting plasma glucose (FPG) and HbA1c levels than group B (P < 0.001). TT levels and cFT were significantly lower in Group A than in the other two groups. In group A, 51 (73%) men had low TT levels (<300 ng/dL) and 48 (69%) had low cFT (<6.35 ng/dL). TT levels < 300 ng/dL had higher sensitivity and specificity (73 and 96%, respectively) than cFT < 6.35 ng/dL (69 and 90%, respectively) for predicting ADAM status. Multivariate-adjusted logistic regression showed that diabetes duration, HbA1c level, and BMI predicted low TT levels, whereas diabetes duration and HbA1c and high-density lipoprotein levels were significant predictors of low cFT. TT levels were a better predictor of ADAM responses than cFT in male T2DM patients. Our results suggest that TT level is better than cFT for diagnosing hypogonadism in T2DM patients when equilibrium dialysis is not feasible.