The influence of body composition on the response to dynamic stimulation of the endocrine pituitary-testis axis

The prevalence of obesity is rapidly increasing worldwide and is associated with lower serum testosterone (T) concentrations in men. The interplay between body composition and the hypothalamic-pituitary-gonadal (HPG) axis is complex, involving reduced sex hormone-binding globulin (SHBG) production (particularly in central obesity and steatosis), increased aromatization of testosterone to estradiol (E2) within adipose tissue, and obesity-associated inflammation and adipokine signaling that may impair gonadotropin secretion and Leydig cell function. Testosterone treatment is generally not recommended for obesity-induced low serum T, making it clinically important to distinguish this condition from overt testosterone deficiency where treatment is indicated. Dynamic testing can assess pituitary (GnRH test) and Leydig cell (hCG test) reserve, but the effect of body composition on these responses is largely unknown. The study aimed to investigate associations between body mass index (BMI) and dynamic responses of the HPG axis upon stimulation in healthy men.

Prior studies demonstrate strong associations between excess body weight and lower total testosterone, partly via reduced SHBG and increased aromatization to estradiol, with variable effects on free testosterone. Obesity-related inflammation and adipokine signaling (e.g., leptin, adiponectin) may impair HPG signaling. Severe obesity may suppress pituitary gonadotropin secretion, yet dynamic reserve can remain. Clinical guidelines generally advise against T therapy for obesity-induced low T and recommend reserving treatment for symptomatic men with low free T. The impact of body composition, particularly regional adiposity (visceral vs limb fat), on dynamic HPG axis testing outcomes had not been well defined prior to this study.

Design and setting: Single-center cross-sectional study conducted at the Department of Growth and Reproduction, Copenhagen University Hospital, Rigshospitalet, Denmark, examining men from the general Danish population (2012–2014). Ethics approval by the Ethical Committee of the Capital Region (H-KF-289428); written and verbal informed consent obtained. Participants: 112 healthy men; exclusions included chronic diseases, prior testicular surgery or trauma, and current anabolic steroid use. A subset (N=78) underwent full-body DXA scanning. Assessments: Baseline reproductive hormones included LH, total testosterone (T), estradiol (E2), and SHBG. Assays: LH and T by time-resolved fluoroimmunoassay (Delfia); E2 by radioimmunoassay (Biotech-IGG, Pantex); SHBG by time-resolved chemiluminescent immunoassay (Access, Beckman Coulter). Inter-/intra-assay CVs: LH 2%/3%, SHBG 5%/4%, T 10%/6%, E2 15%/8%. Free testosterone (FT) calculated via Vermeulen formula using fixed albumin (43.8 g/L). Reference thresholds: low total T <10 nmol/L, low FT <200 pmol/L. Dynamic HPG axis tests initiated between 08:00–12:00: - GnRH test: 100 µg GnRH intravenously (Relefact); LH measured at baseline and 30 min; LH increase calculated as stimulated minus basal. - hCG test: initiated immediately after GnRH test; 5000 IU hCG intramuscularly (Pregnyl); serum T measured at baseline and at 72 h (±1 h); T increase calculated as stimulated minus basal. Body composition: DXA (Lunar Prodigy Advance; enCORE 2004 v8.8) to estimate regional and total fat and fat-free mass; CV ~2% for total and regional fat. Statistical analysis: Continuous variables reported as medians (IQR). Non-normal data log10-transformed; distribution assessed via histograms and Q–Q plots. Linear regression modeled associations between body composition and basal or stimulated hormones; model assumptions checked. Interaction analyses tested whether associations of BMI with basal hormone levels differed from those with stimulated responses. Participants categorized by BMI: <25, 25–29.9, ≥30 kg/m². One-way ANOVA with Bonferroni post hoc compared hormone levels across BMI groups. Two-way ANOVA examined effects of BMI category (<25 vs ≥25 kg/m²) and FT category (<200 vs ≥200 pmol/L) on dynamic responses. Single-mediation analyses performed using PROCESS v4.2. p<0.05 considered significant. Analyses in IBM SPSS v28.

- Cohort characteristics: N=112; median age 30.5 years (IQR 19.0–35.5); median BMI 24.0 kg/m² (range 17.3–59.1); BMI groups: <25 (n=65), 25–29.9 (n=29), ≥30 (n=18). Median T 16.4 nmol/L, FT 369 pmol/L, LH 3.4 IU/L. Twenty-one men had T<10 nmol/L; 19 had FT<200 pmol/L. - BMI and LH: Higher BMI associated with lower basal LH (βu = −0.44, 95% CI −0.88 to −0.01, p=0.04), explaining 4% of variance; 14% (95% CI 1–36%) of the negative BMI–LH association mediated by E2. Basal LH did not differ across BMI groups (p=0.16). GnRH-stimulated LH increase not associated with BMI (βu = −0.10, 95% CI −0.72 to 0.51, p=0.74) and did not differ across BMI groups (p=0.70). - BMI and testosterone: Higher BMI strongly associated with lower basal total T (βu = −0.02, 95% CI −0.03 to −0.02, p<0.001; 35% variance explained), with 17% (95% CI 5–33%) of the association mediated by SHBG. Overweight and obesity had 9% (95% CI 5–14%, p=0.003) and 24% (95% CI 19–30%, p<0.001) lower total T vs normal weight. FT also decreased with BMI (βu = −15.0 pmol/L per BMI unit, 95% CI −19.9 to −10.0, p<0.001; 25% variance explained). Overweight and obesity had 25% (95% CI 16–34%, p=0.006) and 50% (95% CI 39–62%, p<0.001) lower FT vs normal weight. - hCG-stimulated T response: BMI explained 5% of the variance in hCG-stimulated T increase (βu = −0.36, 95% CI −0.69 to −0.03, p=0.04). However, the negative BMI association was significantly weaker for hCG-stimulated T compared to basal T (interaction p=0.04). hCG-stimulated T increase did not differ significantly in overweight (0%, 95% CI −15% to 17%, p=0.99) or obesity (−21%, 95% CI −57% to 14%, p=0.26) vs normal weight. SHBG did not mediate the BMI–hCG-stimulated T relationship (20%, 95% CI −12% to 71%). - Low vs normal FT and dynamic responses: GnRH-stimulated LH and hCG-stimulated T increases did not differ significantly between men with low vs normal basal FT (LH: p=0.41 for FT effect, interaction with BMI p=0.96; T: p=0.21; limited by missing data for normal-weight with low FT). - Regional adiposity (DXA subset): Greater trunk fat mass associated with lower basal LH (B = −0.02, 95% CI −0.03 to 0.00, p=0.03), total T (B = −0.02, 95% CI −0.03 to 0.00, p=0.02), and FT (B = −17.0, 95% CI −30.5 to −3.4, p=0.02). Trunk fat not significantly associated with GnRH-stimulated LH (B = −0.77, 95% CI −1.69 to 0.14, p=0.10) or hCG-stimulated T increase (B = −0.45, 95% CI −1.33 to 0.43, p=0.31). Limb fat showed no significant associations with basal or stimulated hormones.

The study addressed whether excess body weight alters dynamic pituitary and Leydig cell responses, which could inform differentiation between obesity-related low testosterone and overt hypogonadism. Findings show that while higher BMI is linked to lower basal total and free testosterone and modestly lower basal LH, the stimulated responses to GnRH and hCG are relatively preserved and less dependent on BMI. This suggests substantial pituitary and Leydig cell reserve despite obesity-related suppression of basal hormone levels. Trunk (visceral) adiposity correlated with lower basal LH, T, and FT, supporting a particular role for central fat in HPG axis inhibition beyond effects on SHBG and aromatization. Mediation analyses indicated only partial roles for E2 (on LH) and SHBG (on T), implying additional mechanisms, such as inflammation and adipokines, contribute to adiposity–HPG interactions. Clinically, these results support the potential use of dynamic stimulation tests to help distinguish obesity-induced functional suppression from true testosterone deficiency requiring treatment: a normal GnRH and hCG response in an overweight/obese man with low basal T may indicate obesity-related suppression rather than primary or secondary hypogonadism. Variability within the obese subgroup, and trends toward lower hCG responses in some men with obesity, align with heterogeneity in metabolic health among individuals with obesity, with metabolically unhealthy obesity possibly exerting greater inhibitory effects.

Basal sex hormone concentrations, particularly total and free testosterone, are closely and negatively associated with BMI and central adiposity, complicating the diagnosis of testosterone deficiency in men with overweight/obesity. In contrast, dynamic responses to GnRH and hCG stimulation are less influenced by BMI, suggesting these tests can aid clinical decision-making to differentiate obesity-related low T from overt hypogonadism. Future work should evaluate these approaches in symptomatic patients, include more individuals with higher classes of obesity, assess hypothalamic-level function (e.g., clomiphene testing), and utilize gold-standard hormone assays to refine diagnostic thresholds.

- Observational cross-sectional design precludes causal inference. - Limited number of participants with obesity (n=18) and very severe obesity (only two with BMI ≥40 kg/m²), reducing precision for these subgroups. - Few normal-weight men with low FT limited interaction analyses and conclusions for that subgroup. - Hormone measurements used immunoassays, with less accurate E2 assessment compared to mass spectrometry; total T within normal range also subject to some inaccuracy. - Free testosterone was calculated (Vermeulen formula) rather than measured by equilibrium dialysis (gold standard). - No clomiphene testing; hypothalamic contribution to dynamic responses not assessed. - Study cohort comprised healthy men without symptoms of T deficiency; findings may not generalize directly to clinical populations with overt hypogonadism.