The hormonal phenotype of nonclassic 3β-hydroxysteroid dehydrogenase (HSD3B) deficiency in hyperandrogenic females is associated with insulin-resistant polycystic ovary syndrome and is not a variant of inherited HSD3B2 deficiency

To test our hypothesis that the hormonal phenotype of mild 3beta-hydroxysteroid dehydrogenase (HSD3B) deficiency in hyperandrogenic females (HF) is related to insulin-resistant polycystic ovary syndrome (PCOS), we compared insulin sensitivity and gonadotropin secretion in HF with compromised (2) adrenal HSD3B phenotype despite normal HSD3B2 genes (n = 6) to those in HF with classic PCOS (n = 9) of similar ages (14-36 yr). The same was examined in premature pubarche (PP) girls with (n = 4) and without the down arrow HSD3B phenotype (n = 5). The down arrow HSD3B phenotype was defined by ACTH-stimulated Delta(5)-precursor steroid levels and Delta(5)-precursors to Delta(4)-product steroid ratios higher than those in normal females (n = 30 for adult, n = 12 for pubertal). Classic PCOS HF had elevated testosterone levels and normal ACTH-stimulated hormonal profiles. The insulin sensitivity index determined by the frequently sampled iv glucose-tolbutamide test (FSIVGTT) in all HF with down arrow HSD3B phenotype and in all HF with classic PCOS, regardless of body mass index (BMI), was lower than in all eight normal BMI and five high BMI normal females. Integrated incremental insulin determined by FSIVGTT, the area under the curve for insulin, and fasting and 2 h glucose load insulin levels determined by an oral glucose tolerance test in both HF groups were higher (P < 0.01-0.0001) than those in normal females with normal or high BMI. LHRH-stimulated LH levels and LH/FSH ratios in both HF groups were higher (P < 0.01) than those in normal females. No statistical differences were found in the insulin sensitivity and gonadotropin parameter between the two PP girl groups. The insulin sensitivity index in each half of PP girls with the down arrow HSD3B phenotype was lower than or similar to that in control PP girls with a similar weight length index. The fasting glucose to insulin ratio in three of four PP girls with the down arrow HSD3B phenotype was lower than that in control PP girls, but one of four with the down arrow HSD3B phenotype had a higher fasting glucose to insulin ratio than the control PP girls. The findings of insulin sensitivity and gonadotropin data in both HF with the down arrow HSD3B phenotype and classic PCOS indicate significant insulin resistance and LH hypersecretion in both. These suggest that the down arrow HSD3B phenotype in HF is associated with a variant of insulin-resistant PCOS. The variable insulin sensitivity parameter in the small number of PP girls with the down arrow HSD3B phenotype warrants a further large scale study to examine this phenotype association with childhood insulin resistance.