Polycystic ovary syndrome (PCOS) is characterised by reproductive, hyperandrogenic and dysmetabolic features (including insulin resistance). There remain major questions about the basis of the metabolic dysfunction in PCOS and about its genetic aetiology. Having genotyped samples from >460 PCOS cases and >1300 female controls, I recently successfully identified the first genome-sequence variant (the FTO gene) to be implicated in susceptibility to PCOS (Barber et al. 2008, Diabetologia, 51, 11531158), providing the first genetic corroboration of the link between PCOS and obesity. Within the same cohort, I have also shown that variants implicated in T2D-susceptibility acting through adverse effects on beta-cell function, do not associate with PCOS. This is not consistent with the notion of a primary role of the beta-cell in the establishment of hyperinsulinaemia in PCOS.
Established dogma dictates that body fat distribution (visceral adiposity) is implicated in the inherent, fat-mass independent insulin resistance that characterizes many women with PCOS. Using axial MRI images, I recently demonstrated that groups of PCOS cases and BMI/fat mass-matched control women are indistinguishable with respect to distribution of fat within visceral, abdominal and gluteo-femoral subcutaneous depots, despite significant differences in measures of insulin resistance between the groups (Barber 2008, JCEM, 93, 9991004). Therefore, insulin resistance in PCOS is less closely linked to eutopic fat distribution than has previously been thought, and ectopic fat may play an important role. To corroborate the notion of adipo-normality in PCOS, I have shown that adipokine levels (adiponectin and retinol-binding protein 4) are indistinguishable between PCOS cases and BMI/fat mass-matched control women (Barber 2008, JCEM, 93, 28592865).
In conclusion, adipose tissue (both distribution and function) appears to behave normally in women with PCOS. Although the link between obesity and PCOS (probably mediated via effects on insulin resistance) has been genetically corroborated for the first time recently, our understanding of this association remains incomplete.