In Reply Muratli et al propose several important ideas that warrant consideration. Similar to the International Expert Committee,1 they point out that the hemoglobin A1c percentage is a more reliable index of hyperglycemia than fasting glucose. We agree this is an important consideration and our more recent studies now routinely collect hemoglobin A1c. Collecting fasting glucose at multiple points is also a fair consideration. However, regardless of how type 2 diabetes mellitus is diagnosed, we point out that covarying type 2 diabetes mellitus status or hyperglycemia (blood glucose level ≥100 mg/dL; to convert to millimoles per liter, multiply by 0.0555) yields nearly identical results in several studies, such as in our study,2 which used the homeostatic model assessment of insulin resistance. It is likely the case in this study and other reports that the degree of hyperinsulinemia, rather than hyperglycemia, is driving associations with brain outcomes.
Willette AA, Bendlin BB. Confounders Regarding the Association of Insulin Resistance and Alzheimer Disease—Reply. JAMA Neurol. 2016;73(2):240-241. doi:10.1001/jamaneurol.2015.3986