Determinants of testosterone recovery after bariatric surgery: is it only a matter of reduction of body mass index?

OBJECTIVE: To explore the correlation models between body mass index (BMI) and sex hormones constructed from a male cross-sectional survey and evaluate the effects of surgery-induced weight loss on sex hormones in morbidly obese subjects that are not predicted by the constructed BMI correlation models.
DESIGN: Cross-sectional population and longitudinal studies.
SETTING: Bariatric surgery center in a university hospital. PATIENT(S): A cross-sectional survey of a male general population of 161 patients (BMI median [interquartile range] = 29.2 [24.8-41.9] kg/m(2)) in addition to 24 morbidly obese subjects (BMI = 43.9 [40.8-53.8] kg/m(2)) who were undergoing bariatric surgery were prospectively studied for 6 and 12 months. INTERVENTION(S): Bariatric surgery on 24 morbidly obese men. MAIN OUTCOME MEASURE(S): Cross-sectional population: construction of the best-fitting models describing the relationship between baseline BMI with total (TT) and calculated free (cFT) testosterone, E2, sex hormone-binding globulin (SHBG), FSH, and LH levels. Longitudinal study deviation between the observed sex hormone levels at 6- and 12-month follow-up and those expected on BMI bases. RESULT(S): The correlation of BMI with sex hormones was not univocally linear (E2), but the best-fitting model was exponential for TT, cFT, FSH, LH, and TT/E2 and power for SHBG. In addition to the significant improvement of all parameters observed after surgery in the longitudinal cohort, the increase in TT and SHBG, but not in cFT, was significantly higher than expected from the corresponding weight loss at 6 months from surgery (14.80 [12.30-19.00] nM vs. 12.77 [10.92-13.64] nM and 40.0 [28.9-54.5] nM vs. 24.7 [22.5-25.8] nM for TT and SHBG, respectively), remaining rather stable at 12 months. CONCLUSION(S): The increase in TT and SHBG, but not the increase in cFT, after bariatric surgery is greater than expected based on weight loss.