Preoperative Prediction of Postprandial Glycemia After Roux-en-Y Gastric Bypass
We thank Quilliot et al for their comment1 on our recent article2 and for providing complementary data on post-prandial hypoglycemia after Roux-en-Y Gastric Bypass (RYGB). These authors measured blood glucose during 240 minutes after an oral 75 g glucose load in 171 patients experiencing symptoms of hypoglycemia after RYGB, among a cohort of 811 operated patients (21%). Glycemia below 0.5 g/dL was observed at 180 minutes in 69% of these patients. On the basis of these data, the prevalence of proven hypoglycemia as defined by glycaemia below 2.8 mmol/L during the test for the entire cohort was 14.5% (118 of 811 RYGB patients). Furthermore, the median delay between glucose ingestion and hypoglycemia was 120 minutes (range 60 to 240 minutes). The authors claimed that the standard 2 hours oral glucose tolerance test (OGTT) used in our study2 may have underestimated the prevalence of postprandial hypoglycemia after RYGB, and raised questions about the validity of the proposed mechanisms and preoperative predictive factors.
The lack of a clear and widely accepted definition and glucose threshold for the diagnosis of postprandial hypoglycemia after RYGB is a clear limitation of all studies exploring this condition.3 In our longitudinal study,2 we chose to submit participants to a classical 2-hour OGTT, according to the American Diabetes Association standards4 aiming their stratification based on glucose tolerance. The strength of our study was the longitudinal evaluation in a large cohort of the occurrence of postprandial hyperinsulinemic hypoglycemia (PHH) by a standardized test, systematically performed in each patient, even in the absence of symptoms. Nevertheless, in insulin-resistant morbidly obese patients, the inhibition by insulin of its own secretion is blunted partially through an altered sensitivity to insulin.5 This alteration may mask an underlying predisposition for hypoglycemia explaining the very low prevalence of PHH at baseline in our study. In line with this, we have shown that PHH after RYGB occurs in patients when insulin sensitivity is restored in relation to weight loss.2
The differences of PHH prevalence after surgery (9.1% and 7.9% at 12 and 60 months following RYGB in our study vs 14.5% at 30 ± 21 months in the study by Quilliot et al1) may be related to the different technique used for evaluation, duration of follow-up,3 and variability in gastric emptying and glucose intestinal absorption after surgery.6,7 Moreover, the definition of hypoglycemia after RYGB in the study by Quilliot et al1 did not include the threshold for plasma insulin above 3 mU/L, which is one of the criterion generally, required to qualify PHH.8 Another important feature of that study1 is that the OGTT was performed only in patients experiencing hypoglycemia symptoms after RYGB, in whom PHH is more likely. Furthermore, the authors did not describe the patient characteristics before surgery in their study, especially the presence of type 2 diabetes (T2D) that significantly decreased the prevalence of PHH after RYGB in our study.2 In addition, specific technical features may affect OGTT outcomes, such as the positioning of the subject that may impact gastric emptying and postprandial glycemic profile,9 especially after RYGB.
In order to confirm the crucial influence of preoperative beta cell function proposed in our initial study,2 we analyzed a subgroup of 61 patients presenting with and without T2D before surgery (n = 40 and n = 21, respectively) from the same prospective cohort study. These patients were enrolled in an ancillary study in which blood glucose was also measured during 180 minutes after a standardized mixed meal, 3 months after RYGB.
First, we found a tight correlation between blood glucose at 120 and 180 minutes after the meal (Fig. 1A). As suggested by Quillot et al,1 the mean ± s.e.m. blood glucose was slightly lower at 180 minutes (5.8 ± 0.