The association between hyperandrogenemia and the metabolic syndrome in morbidly obese women

BACKGROUND: Female abdominal obesity is associated with hyperandrogenemia (HA), but few studies have addressed the possible association between HA and metabolic syndrome (MetS) among obese women. Some studies indicate that insulin resistance may cause HA through different mechanisms. On the other hand, a bidirectional relationship between HA and insulin resistance has been suggested. Thus, we aimed to investigate if morbidly obese women with HA had higher odds of MetS and its components than those without HA (controls), independent of polycystic ovarian syndrome (PCOS) status. METHODS: This cross-sectional study comprised 1900 consecutive treatment seeking morbidly obese women <50 years. Free testosterone index (FTI) >0.6 defined HA. Women with previously diagnosed PCOS and those with oligo- / anovulation combined with clinical or biochemical hyperandrogenism were defined as having PCOS. Multiadjusted associations between HA and MetS were assessed by logistic regression analysis. RESULTS: Out of 1900 morbidly obese women, 1089 (57 %), 846 (45 %) and 312 (16 %) had MetS, HA and PCOS, respectively. Compared with controls (without HA), women with HA were younger (34 [1] years vs. 39 [2], p < 0.001) had a higher prevalence of MetS (62 % vs. 53 %, p < 0.001), type 2 diabetes (18 % vs. 15 %, p = 0.045), low HDL-cholesterol (65 % vs. 48 %, p < 0.001) and hypertriglyceridemia (48 % vs. 41 %, p = 0.004), but a lower prevalence of raised blood pressure (53 % vs. 59 %, p = 0.014). Multivariable analyses showed that HA was associated with increased odds of MetS (OR 1.61 [95 % CI 1.27, 2.02]), dysglycemia (1.65 [1.28, 2.11]), low HDL-cholesterol (1.58 [1.27, 1.97]), and hypertriglyceridemia (1.43 [1.15, 1.79]). After stratification for the presence of PCOS, the results remained largely unchanged in women without PCOS; MetS (1.52 [1.18, 1.96), dysglycemia (1.71 [1.30, 2.25]), low HDL-cholesterol (1.55 [1.22, 1.98]) and hypertriglyceridemia (1.36 [1.06, 1.74]). CONCLUSION: Morbidly obese women with HA had an approximately 1.5-fold increased odds of having MetS even in the absence of PCOS. Randomized controlled clinical trials, including therapeutic strategies to lower free testosterone levels, are however necessary to explore any cause-and-effect relationship.