Anti-Mullerian Hormone Protein-Altering Variants Are Associated With Hypertriglyceridemia in the General Population

We found 17 rare PCOS-specific functional protein-altering variants (PAVs) or mutations in the gene, AMH, that decrease the biologic activity of the encoded protein, anti-Müllerian hormone (AMH), in the heterozygous state. Approximately 3% of European ancestry PCOS cases in our cohort of ~700 were affected. Our preliminary studies found evidence for a metabolic phenotype in both PCOS as well as their male first-degree relatives who were heterozygous carriers of these AMH PAVs. We performed this study to test the hypothesis that AMH mutations are associated with metabolic abnormalities in the general population. The Mount Sinai BioMe Biobank is an electronic health record (EHR)-linked biobank, containing anonymized whole exome sequences from 30,813 participants of diverse ancestries. We interrogated the sequence data to identify individuals with PCOS-related AMH PAVs. IRB-approval was obtained to review the linked EHR. Outcomes were the presence of obesity (BMI ≥ 30 kg/m(2)), type 2 diabetes (hemoglobin A1C ≥ 6.5%), prediabetes (hemoglobin A1C 5.7% - 6.4%), elevated cholesterol (total cholesterol ≥ 200 mg/dL), hypertriglyceridemia (TG ≥ 150 mg/dL), and hypertension (≥2 blood pressure values ≥140/90, or administration of antihypertensive medications). Control subjects were obtained from the National Health and Nutrition Examination Survey using propensity score matching (for sex, age, and BMI) with a 1:4 case:control ratio. A total of 292 individuals with AMH PAVs were identified, resulting in a combined 0.95% prevalence of AMH PAVs in an unselected population (1.07% in Europeans, 0.28% in African Americans, 0.54% in Hispanics, and 0.07% in Asians). After adjusting for age, BMI, and race/ethnicity, there was a statistically significant increased prevalence of hypertriglyceridemia in both women (OR 7.29, 95% CI 3.77-14.00) and men (OR 10.15, 95% CI 4.68-22.00) with AMH PAVs compared to sex-, age- and BMI-matched controls. There was also a statistically significant increased prevalence of elevated cholesterol in men with AMH PAVs compared to controls (OR 2.48, 95% CI 1.15-5.34). There were no significant differences between individuals with AMH PAVs and matched controls with respect to the other outcomes. These findings suggest that decreased bioactivity of AMH is causally related to dyslipidemia in the general population. TG levels are elevated in both sexes, whereas increases in cholesterol are only seen in men. The mechanisms by which decreases in AMH bioactivity alter circulating lipid levels are of considerable interest since AMH has no known metabolic actions. It is possible that the putative metabolic effects of AMH are mediated by increases in circulating testosterone (T) levels that in turn alter lipid metabolism. Since T levels are not commonly available in EHR, future studies will be needed to investigate this hypothesis as well as to explore metabolic actions of AMH.