Postprandial Hypoglycemia After Roux-en-Y Gastric Bypass
We read with great interest the study by Raverdy et al1 in the November 2016 issue of Annals of Surgery, suggesting that the incidence of postprandial hypoglycemia (PPH) significantly increased after Roux-en-Y Gastric Bypass (RYGB), but remained stable between 1 and 5 years, and that the estimation of beta-cell function with an oral glucose tolerance test (OGTT) before surgery could identify post-RYGB PPH risk patients. Although that data is potentially interesting, we also believe it to be questionable, and we would like to warn readers of the over-interpretation risk of this study's conclusions.
We observed that the prevalence of PPH after OGTT in this study was very low (0.5% at baseline, 9.1% and 7.9% at 12 and 60 months after RYGB). In fact, these postoperative values could be considered as similar or even lower than data observed in the general population without bariatric surgery. A systematic OGTT in healthy population reported a PPH prevalence of up to 10%, without symptom2 or any electroencephographic sign.3 Other studies have also reported a PPH prevalence of up to 50% in symptomatic or asymptomatic patients depending on glucose threshold (45–60 mg/dL).4–6 One of the main potential explanations for this discrepancy could be that the mean delay between glucose oral ingestion and hypoglycemia is likely far superior to 120 minutes after gastric bypass or total gastrectomy. In a cohort of 811 patients after gastric bypass, we observed 127 symptomatic patients with sympathetic symptoms of hypoglycemia (15.6%) and 44 patients with neuroglycopenia symptoms (5.4%). All these patients had a 4-hour OGTT. The mean delay between gastric bypass and hypoglycemia symptom appearance was 29.5 ± 20.7 months. Four-hour OGTT hypoglycemia (glycemia below 0.5 g/dL) was observed in 69% of these patients, and in 90% if the threshold was set at 0.6 g/dL. The minimal glycemia was 0.44 (±0.13). More importantly, the mean delay between glucose oral ingestion (OGTT) and hypoglycemia was 135 ± 36 minutes, and the median delay was 120.0 minutes (range 60–240 minutes).
In consequence, we believe that Raverdy et al may have ignored up to 50% of hypoglycemia episodes. This point is important because the underestimation of PPH prevalence may lead to biased conclusions and questions the validity of the predictive factors proposed in this study. To evaluate their utility, we believe they should be validated with a 4-hour OGTT. Furthermore, Emous et al 7 consider that the aim of OGTT is to confirm the diagnosis of symptomatic postgastric bypass hypoglycemia by combining identification of hypoglycemic complaints together with a provocation test, and the use of glucose below 60 mg/dL as a sensitive cut-off value for meal-induced hypoglycemia during OGTT is also more relevant. With the rising incidence of bariatric surgeries, we believe, as Malik et al,8 that the most clinically relevant point is to be able to recognize risk factors for neuroglycopenia in post-RYGB patients, and to investigate and manage it accordingly to avoid traumatisms and mortality risks. Most patients with symptomatic hypoglycemia respond well to dietary modifications with a low carbohydrate diet. Although this study's results are interesting, we believe that OGTT (or mixed meal tolerance test) is necessary in all patients with hypoglycemic symptoms, and these results need to be carefully evaluated to avoid misdiagnosis, to detect inappropriately elevated levels of insulin, and provide optimal management of patients after gastric bypass.