The oral glucose tolerance test is frequently abnormal in patients with uncontrolled epilepsy

The clinical efficacy of the ketogenic diet as therapy for patients with difficult-to-treat epilepsy prompted us to investigate the glucose metabolism of these patients under an oral overload of glucose, that is, in the oral glucose tolerance test (OGTT).

Thirty patients (12 males, 18 females; age range: 17-59, mean: 35.1) with difficult-to-treat epilepsy, 23 patients with controlled epilepsy (11 males, 12 females; age range: 14-66, mean: 36.9), and 39 control subjects (18 males, 21 females; age range: 16-58, mean: 33.3) were evaluated with the OGTT. For patients with epilepsy, we also measured C-peptide and glycosylated hemoglobin in the fasting state. Glucose levels lower than 70 mg/dL at any point of the curve were considered to be abnormal.

All subjects in the control group and the group with controlled epilepsy had a normal OGTT. In contrast, all 30 patients with difficult-to-treat epilepsy had at least one point on the OGTT curve below the normal range (P<0.001), most often 180 and 240 minutes after the oral glucose load (P<0.001). C-peptide levels were significantly lower in the group with difficult-to-treat epilepsy as compared with the group with controlled epilepsy. Fasting glycohemoglobin and insulin levels did not differ between the two patient groups.

We suggest that undiagnosed metabolic disturbances in patients with difficult-to-treat epilepsy may somehow contribute to their refractoriness to conventional pharmacological therapy. We propose the hypothesis that calorie-restricted diets aimed at correcting OGTT curves may prove beneficial in treating patients with difficult-to-treat epilepsy. Our hypothesis generates a clear endpoint for the diet, and its demonstration would provide new standards for diet-based antiepileptic regimens. Accordingly, our results may help in understanding the positive consequences of ketogenic or calorie-restricted diets in persons with seizures.