I read with interest the article of Olson et al. (1) detailing the performance of A1C as a diagnostic test in comparison with the “gold standard” oral glucose tolerance test (OGTT) and its ensuing press coverage. The fundamental challenge regarding the diagnosis of diabetes is how to identify a population at risk for complications. It is a fallacy that the OGTT is a gold standard. In fact, for most of the several thousand year history of diabetes, it has not been. De facto, if you define one test as a gold standard, all comparators will be inferior.
On the other hand, the A1C test is at least as good as both the fasting plasma glucose and the OGTT in identifying people with both microvascular and macrovascular disease. That ability to define a population at risk of complications is what we need a screening/diagnostic test for diabetes to do. Coupled with the fact that A1C testing is easier for patients and providers to screen opportunistically when health care interactions occur, even if the patient is ill or not fasting, it is absolutely clear that A1C is a superior technique to identifying people at risk for complications who might benefit from validated therapies. A more robust treatment of these issues is available (2).
I am afraid that the article of Olsen et al. and the press coverage that ensued do not address the fundamental issue remaining—how do various screening/diagnostic strategies compare with regards to costs and outcomes in populations. I suspect that many will be confused by the implication that A1C testing is inferior, which is simply not addressable with the study design employed by Olsen et al. A1C testing is certainly different from the OGTT, but its relative worth can only be addressed through community- based intervention studies.
Acknowledgments
No potential conflicts of interest relevant to this article were reported.
