We agree with Lachin et al. (1) that intensive therapy should be started as early as possible after type 1 diabetes onset. The concept of cumulative glycemic exposure in no way alters that clinical message. The intention of our article (2) was to raise the argument that “metabolic memory” has been accepted without a full exploration of a simpler alternative explanation for the observed persistence of benefit in the intensive therapy arm after Diabetes Control and Complications Trial (DCCT) closeout. Importantly, we do not “purportedly disprove metabolic memory” (1) but merely suggest “There is no need [our emphasis] to invoke a ‘metabolic memory’ phenomenon to explain the persistence of a lower incidence of complications in the DCCT intensive therapy group… which can be fully explained by cumulative glycemic exposure.” (2). Notably, Lachin et al. do not challenge this assertion per se. Rather, they merely claim we have not studied “metabolic memory.” In an observational study, like the Pittsburgh Epidemiology of Diabetes Complications (EDC), people do not self-randomize into contrasting groups, as done experimentally in the DCCT, so a similar analysis is not possible. Nevertheless, we did demonstrate that the complication incidence curves observed in DCCT/EDIC (Epidemiology of Diabetes Interventions and Complications) by treatment group correspond to what would be expected given the glycemic exposure of the two treatment groups using actual DCCT/EDIC HbA1c values from the publicly available data sets.
We are also confused by the assertion by Lachin et al. that we “fail to recognize that cumulative glycemic exposure includes long-term effects, such as metabolic memory,” for we extensively described deleterious long-term effects of cumulative glycemic exposure. This included reviewing two potential leading pathogenic pathways. If “metabolic memory” is something more than differential prior cumulative glycemic exposure, surely the onus is on its proponents to hypothesize underlying pathways/mechanisms of action, especially as nearly 20 years have now elapsed since it was first proposed (3).
Finally, Lachin et al. reframe the “fundamental issue” as the “pattern of glucose exposure” rather than differential cumulative glycemic exposure. We agree that we did not examine patterns of glycemic exposure directly. Indeed, it would be impossible to replicate the DCCT/EDIC analyses given the EDC participants’ median 18 years’ diabetes duration at baseline compared with 4 years in DCCT. Thus, we cannot study “early” intervention in EDC and do not challenge their assertion of a “particularly pernicious effect of elevated levels of HbA1c early in the course of type 1 diabetes.” However, we did show that mean A1c months at complication development did not differ by diabetes duration, which, in EDC, ranged from 7 to 37 years at baseline. Thus, a similar absolute cumulative glycemic exposure was reached prior to complication development, within the durations of diabetes available for study in EDC. Finally, while we support the call by Lachin et al. for intensive therapy to be initiated as early as possible, we also suggest that it is never too late, for even after 20 years of diabetes, improved glycemic control is associated with lower subsequent coronary artery disease risk (4).
Article Information
Duality of Interest. T.J.O. is a consultant for Boehringer Ingelheim International GmbH. No other potential conflicts of interest relevant to this article were reported.
