Impact of age, body weight and metabolic risk factors on steroid reference intervals in men

OBJECTIVE: To evaluate the independent impact of age, obesity and metabolic risk factors on 13 circulating steroid levels; to generate reference intervals for adult men. DESIGN: Cross-sectional study. METHODS: 315 adult, drug-free and apparently healthy men underwent clinical and biochemical evaluation. 13 steroids were measured by LC-MS/MS and compared among men with increasing BMI. Moreover, the independent impact of age, BMI and metabolic parameters on steroid levels was estimated. Upper and lower reference limits were generated in steroid-specific reference sub-cohorts and compared with dysmetabolic sub-cohorts. RESULTS: We observed lowering steroid precursors and testosterone, and increasing estrone levels in men showing increasing BMI levels. By multivariate analysis, 17-hydroxyprogesterone and dihydrotestosterone decreased with BMI, while cortisol decreased with waist circumference. Estrone increased with BMI and systolic blood pressure. Testosterone decreased with worsening insulin resistance. 17-hydroxypregnenolone and corticosterone decreased with increasing total/HDL-cholesterol ratio. Age-related reference intervals were estimated for 17-hydroxypregnenolone, dehydroepiandrosterone, 17-hydroxyprogesterone, corticosterone, 11-deoxycortisol, cortisol and androstenedione, while age-independent reference intervals were estimated for progesterone, 11-deoxycorticosterone, testosterone, dihydrotestosterone, estrone and estradiol. Testosterone lower limit was 2.29nmol/L lower (P=0.007) in insulin resistant vs insulin sensitive men. Furthermore, the upper limits for dihydrotestosterone (-0.34nmol/L, P=0.045), cortisol (-87nmol/L, P=0.045-0.002) and corticosterone (-10.1nmol/L, P=0.048-0.016) were lower in overweight/obese, in abdominal obese and in dyslipidaemic subjects compared to reference sub-cohorts, respectively. CONCLUSIONS: Obesity and mild unmedicated metabolic risk factors alter the circulating steroid profile and bias the estimation of reference limits for testosterone, dihydrotestosterone, cortisol and corticosterone. Applying age-dependent reference intervals is mandatory for steroid precursors and corticosteroids.