The global obesity pandemic affects over 1 billion people, with studies suggesting a strong link between serum testosterone (T) levels and excess body weight in men. The relationship between body composition and the hypothalamic-pituitary-gonadal (HPG) axis is complex, involving multiple mechanisms. Obesity, particularly central obesity and steatosis, inhibits sex hormone-binding globulin (SHBG) production, leading to lower total T but only minimally affecting free testosterone (FT). Increased aromatization of T to estradiol (E2) in adipose tissue contributes to lower serum T and higher E2, potentially increasing negative feedback at the hypothalamic-pituitary level. Obesity-associated chronic inflammation and adipokine signaling may also impair gonadotropin secretion, Leydig cell function, and T production. Testosterone treatment is generally not recommended for obesity-induced low serum T, making differentiation from overt T deficiency a diagnostic challenge. Dynamic testing of the HPG axis, using gonadotropin-releasing hormone (GnRH) and human chorionic gonadotropin (hCG) tests, can aid in diagnosing T deficiency. However, the impact of body weight on HPG response to stimulation remains largely unknown. This study aimed to investigate the association between body mass index (BMI) and dynamic HPG axis responses to stimulation in healthy men.

Several studies have shown an association between serum testosterone (T) concentrations and excess body weight in men. Obesity, particularly central obesity, inhibits the production of sex hormone binding globulin (SHBG), resulting in lower circulating total T. Increased aromatization of testosterone to estradiol in adipose tissue can also contribute to lower serum testosterone. Additionally, obesity-associated chronic inflammation and adipokine signaling may impair gonadotropin secretion and testosterone production. While testosterone treatment is generally not recommended for obesity-induced low testosterone, distinguishing this from overt testosterone deficiency remains a diagnostic challenge. Dynamic testing of the HPG axis using GnRH and hCG tests has been used to evaluate testosterone deficiency, but the influence of body weight on the HPG response to stimulation has not been well studied.

This single-center cross-sectional study included 112 healthy Danish men (2012-2014) participating in a larger study of testicular function. Participants underwent a complete biochemical assessment of the pituitary-testis axis, including dynamic stimulatory tests: a GnRH test (100 µg GnRH intravenously, LH measured after 30 min) and an hCG test (5000 IU hCG intramuscularly, T measured after 72 h). A subset (n=78) also underwent DXA scans for body composition assessment. Exclusion criteria included chronic diseases, testicular surgery/trauma, and anabolic steroid use. Serum LH and T were measured using time-resolved fluoroimmunoassay, E2 using radioimmunoassay, and SHBG using time-resolved chemiluminescent immunoassay. Free testosterone (FT) was calculated using the Vermeulen formula. Statistical analyses included linear regression, ANOVA, and mediation analysis to examine the associations between BMI, body composition (DXA-derived fat and lean mass), and basal and stimulated hormone levels. Participants were categorized into normal weight (BMI <25 kg/m²), overweight (25-29.9 kg/m²), and obese (BMI ≥30 kg/m²) groups.

A higher BMI was significantly associated with lower basal serum LH (βu = −0.44, p = 0.04) and this association was partly mediated by serum E2 (14%). GnRH-stimulated LH increase was not significantly associated with BMI. BMI was strongly associated with lower basal testosterone (βu = −0.02, p < 0.001) and free testosterone (βu = −15.0, p < 0.001). Men with overweight and obesity had significantly lower testosterone (9% and 24%, respectively) and free testosterone (25% and 50%, respectively) compared to normal-weight men. The HCG-stimulated testosterone increase showed less dependence on BMI compared to basal testosterone (p = 0.04 for interaction). While men with obesity showed a trend towards lower HCG-stimulated testosterone, this was not statistically significant. Trunk fat, but not limb fat, was significantly associated with lower basal LH, total T, and FT. Neither trunk nor limb fat was significantly associated with stimulated LH or T. The GnRH-stimulated LH increase and HCG-stimulated T increase did not significantly differ between men with low vs. normal FT levels. Analysis of mediation effect showed that SHBG partially mediated the effect of BMI on total testosterone but not on the dynamic test results.

This study demonstrates that basal sex hormone levels are strongly associated with BMI, making the diagnosis of testosterone deficiency challenging in obese men. However, dynamic HPG axis responses to stimulation (GnRH and hCG tests) are less influenced by BMI. The relatively preserved responsiveness to stimulation in obese men suggests that the lower basal testosterone levels in these men are primarily due to alterations in the peripheral metabolism of testosterone, rather than a defect in the capacity of the pituitary-testicular axis itself. This supports the hypothesis that obesity-induced low testosterone may be largely due to reduced SHBG and increased aromatization of testosterone in adipose tissue rather than a primary impairment of the HPG axis. However, additional mechanisms such as inflammation and adipokine signaling are likely involved given that mediation analysis did not completely account for the effects of BMI on hormone levels. The findings suggest that dynamic HPG axis testing could be a valuable tool to differentiate between obesity-induced low testosterone and overt testosterone deficiency in obese men.

Basal testosterone levels are closely associated with BMI, complicating the diagnosis of testosterone deficiency in obese men. Dynamic testing of the HPG axis using GnRH and hCG stimulation, which shows less dependence on BMI, may serve as a valuable tool to aid in clinical decision-making for obese men with suspected testosterone deficiency. Future research should investigate the impact of different obesity phenotypes (metabolically healthy vs. unhealthy) on dynamic sex hormone responses.

This observational study cannot establish causality. The relatively small number of men with BMI ≥ 30 kg/m² limits the generalizability of findings to severely obese men. The rarity of low FT in normal-weight men prevented a comprehensive analysis of dynamic test responses in this group. The use of immunoassays for hormone measurements, and calculated FT instead of equilibrium dialysis, may affect the accuracy of the results. Finally, the lack of clomiphene testing prevented an assessment of hypothalamic-level responses to body weight.