“Orr was crazy and could be grounded. All he had to do was ask; and as soon as he did, he would no longer be crazy and would have to fly more missions. Orr would be crazy to fly more missions and sane if he didn't, but if he was sane he had to fly them. If he flew them he was crazy and didn't have to; but if he didn't want to he was sane and had to.”

Catch-22, Joseph Heller(1)

Davidson (2) argues that the plasma glucose (PG) obtained 2 h after a 75-g oral glucose load cannot be an appropriate criterion for the diagnosis of diabetes, because subjects with elevated 2-h PG levels may have normal HbAlc levels, and the diagnosis of diabetes is untenable in people with normal HbAlclevels. This line of reasoning presents an impasse, a “catch-22,”yet it is not consistent with the way we have traditionally defined disease states.

Population studies have demonstrated both normal bell-shaped curve distributions and bimodal distributions of physiologic measures. When physiologic measures are normally distributed, values greater than or less than the mean plus 2 SDs have been used to define abnormal glucose tolerance. When physiologic measures are bimodally distributed, one distribution may be used to define normal and the other to define abnormal glucose tolerance. Numerous population studies have demonstrated that measures of glycemia are bimodally distributed (3). In such instances, the lower distribution defines normal glucose tolerance, and the upper distribution defines diabetes. When there are alternative measures that can be used to define normal and abnormal glucose tolerance, the optimal measure for diagnosis is that which provides the most clear-cut distinction between the two populations. We and others(3,4)have demonstrated that compared with fasting plasma glucose (FPG) and HbAlc, the 2-h PG provides the best differentiation between normal and abnormal glucose tolerance.

We agree that diabetes reflects more than abnormal glucose levels. Diabetes confers a predisposition to unique microvascular and neuropathic complications. Clearly, it is reasonable to consider these risks when defining diabetes. Nevertheless, the argument that a small minority of patients with incident or prevalent retinopathy determines the diagnostic cut point for diabetes ignores the fact that both the presence and absence of complications determine sensitivity and specificity. Data from the entire population are informative. In addition, complications arise as a function of the degree and duration of hyperglycemia and as a result of individual susceptibility. Not everyone will develop complications. Consequently, the analysis of population distributions of glycemia, rather than the prediction of complications, may be the best approach to defining diabetes.

Finally, we believe that issues of diagnosis and treatment must be clearly distinguished. The first task is to make a diagnosis, and the second task is to contemplate therapy. Glucose homeostasis is the result of a complex feedback system involving a hormone, insulin, and its regulatory substrate,glucose. In endocrine systems, if autonomous hyperfunction is suspected, the diagnostic approach is to physiologically suppress the system. If hypofunction is suspected, the preferred diagnostic approach is to physiologically stimulate the system. To the extent that diabetes is a condition of absolute or relative insulin deficiency, a stimulated (2-h PG) rather than a basal measure (FPG or HbAlc) is most likely to unmask the abnormality.

We are not persuaded by the argument that there is no value in labeling an abnormality if it is not severe enough to require treatment. Because treatment goals change over time, this approach is problematic. In addition, we disagree with the assertion that diabetes with normal HbAlc levels does not require treatment. The Diabetes Control and Complications Trial found no evidence of any threshold below which a lower HbAlc level was not associated with a lower risk of the development or progression of complications (5). These data suggest that a more aggressive treatment is beneficial.

In summary, we believe that Davidson's apparent catch-22 arises from the belief that “a diagnosis of diabetes is untenable in a person with a normal HbAlc level”(2). Basing diagnostic cut points for diabetes on the 2-h PG level is supported by general principles of disease definition and endocrine testing, and the impasse is prevented if issues of diagnosis and treatment are distinguished.

Heller J: In
. New Laurel ed. New York, Dell Publishing,
, p.
Davidson MB: Catch-22 (Letter).
Diabetes Care
Engelgau MM, Thompson TJ, Herman WH, Boyle JP, Aubert RE, Kenny SJ,Badran A, Sous ES, Ali MA: Comparison of fasting and 2-hour glucose and HbAlc levels for diagnosing diabetes.
Diabetes Care
McCance DR, Hanson RL, Charles M-A, Jacobsson LTH, Pettitt DJ,Bennett PH, Knowler WC: Which test for diagnosing diabetes?
Diabetes Care
Diabetes Control and Complications Trial Research Group: The absence of a glycemic threshold for the development of long-term complications: the perspective of the Diabetes Control and Complications Trial.