Authors' reply.

We are grateful to Wiwanitkit S and Wiwanitkit V[1] for their interest in the findings of our study.[2] The study by Dahan MH et al.,[3] was based in Washington University Reproductive Centre and though the race has not been mentioned, it is likely that ethnic variations and inclusion of patients with variable body mass index in their study may be responsible for the different results observed between the studies. Moreover, the study of Dahan et al. had aimed to study serum sex-hormone binding globulin (SHBG) levels in relation to insulin resistance parameters. They themselves have stated that polycystic ovarian syndrome (PCOS) and lean controls demonstrated different glucose and insulin responses to 75 gram glucose load as has been seen in our study too. Despite this difference their study showed paradoxically similar GI ratios, a paradox which even the authors were unable to explain. Possibly the smaller sample size of 19 PCOS and 17 controls, adding to chance negative finding and uncontrolled confounding factors (since primary aim was SHBG levels) could possibly be the limitations of this study.

Moreover, we contest the notion that the positive finding in our work was by chance or due to the small number of studied subjects or uncontrolled confounding factors none of which have been supported with any facts by Wiwanitkit S et al. Our study which had a robust sample size (50 women with PCOS and 20 controls) showed statistically significant results with the groups having similar age, and no other apparent confounding factors.[2] In our study “2-hour post-glucose insulin levels” were elevated in 88% of PCOS individuals but were normal in all females not suffering from PCOS. These levels significantly correlated with AUC of glucose and insulin, and insulinogenic index and inversely correlated with 2-hour glucose to insulin ratio (r=0.827, 0.749 and -0.732, respectively). Previous studies too have found the two-hour post-load glucose insulin to be a one-step, single and reliable test for determining insulin resistance.[45]

Therefore, we feel that this is a simple and clinically useful parameter to evaluate insulin resistance in PCOS patients may be able to replace the more cumbersome and elaborate tests currently used.