Endocrine Abstracts

Introduction: Several studies point towards lower testosterone levels (TL) in patients with obesity and/or insulin resistance (IR).

Aim: To analyze the influence of anthropometric parameters and IR on TL in obese men and weight-dependent changes after a short-term weight loss program.

Methods: Data were collected from a group of male patients, aged 18 to 65, when they attended a first obesity appointment and at reassessment 6 months later. Patients with non-obesity-related hypogonadism or under androgen replacement therapy were excluded. Hypogonadism was considered if TL<2.6 ng/ml. The anthropometric parameters’ impact on TL was controlled for age and IR, as well as the IR impact was controlled for anthropometry and age. Insulin resistance was evaluated by homeostatic model assessment (HOMA-IR).

Results: We analyzed 156 patients characterized by mean age=41.8±11.7 years, BMI=45.9±8.6 Kg/m², waist circumference (Wc)=137.1±15.9 cm, TL=3.2±1.6 ng/ml and HOMA-IR=6.7±5.8. Hypogonadism was present in 66 (42.3%) males. TL was negatively correlated to age (r=−0.221; P=0.005), BMI (r=−0.344; P<0.001), Wc (r=−0.379; P<0.001) and HOMA-IR (r=−0.241; P=0.002). Controlling for age and IR, we observed a negative impact of BMI or Wc on TL (P<0.001, for both). Similarly, controlling for anthropometry and age, a negative impact of HOMA-IR on TL was verified (P=0.008). The 6 months’ follow-up appointment was attended by 111 patients. A significant change (Δ) towards decrease on BMI, Wc and IR (P<0.001 for all) and a trend (P=0.052) for TL increase between appointments were observed. An inverse correlation was identified for ΔTL and ΔBMI (r=−0.325; P<0.001), for ΔTL and ΔWc (r=−0.4; P<0.001) but not for ΔTL and ΔIR.

Conclusion: Hypogonadism is highly prevalent in obese men. Adiposity and IR independently contribute towards decreasing TL. Notwithstanding 6 months was not enough to significantly alter TL, the observed increase was dependent on the decrease verified on adiposity parameters but not on the decrease on IR.