- Email: [email protected]
Remission of type 2 diabetes after Roux-en-Y gastric bypass is associated with greater weight loss
Surgery for Obesity and Related Diseases 5 (2009) 305–309
Original article
Remission of type 2 diabetes after Roux-en-Y gastric bypass is associated with greater weight loss Brian E. Kadera, B.S.a, Kristian Lum, M.S.b, John Grant, M.D.c, Aurora D. Pryor, M.D.c, Dana D. Portenier, M.D.c, Eric J. DeMaria, M.D.c,* a Duke University School of Medicine, Durham, North Carolina Department of Statistical Science, Duke University, Durham, North Carolina c Department of Surgery, Duke University Medical Center, Durham, North Carolina Received May 2, 2008; revised February 11, 2009; accepted February 15, 2009 b
Abstract
Background: Physiologic studies in rodents and preliminary human studies have suggested that Roux-en-Y gastric bypass (RYGB) improves type 2 diabetes mellitus (T2DM) by way of metabolic changes, long before the bariatric or weight loss effects occur, leading to the concept of “metabolic surgery.” To test this hypothesis, we studied patients with insulin-dependent T2DM who underwent RYGB to determine whether T2DM remission in this treatment-resistant subgroup occurred independent of weight loss. Methods: Of all the patients undergoing RYGB from 2000 to 2006 (n ⫽ 1546) with ⱖ12 months of follow-up, 318 had T2DM (21%), and 75 (24%) of these were insulin dependent. Of the 75 patients, 4 were found to have T1DM (5.3%) and were excluded, leaving a study population of 71 patients. The patients who achieved remission, defined as a cessation of diabetic medications with a hemoglobin A1c level of ⬍7%, were compared with those who did not achieve remission. Statistical significance was set at P ⬍ .05, using the Student t test, chi-square test, and logistic regression analysis, as appropriate. Results: After RYGB, all 71 patients with insulin-dependent T2DM had achieved a reduction in the dose and/or number of medications at 29.6 ⫾ 17.0 months. Of these 71 patients, 35 (49%) demonstrated a remission of T2DM. The preoperative body mass index, age, number of medications, and hemoglobin A1c level did not differentiate between those who attained remission and those who still required diabetic medication. From the multivariate analysis, the significant factors associated with remission were the preoperative insulin dose and the percentage of excess weight loss. The percentage of excess weight loss was greater in the remission patients as early as 3 months postoperatively (P ⫽ .04) and also at 6, 12, 18, and 24 months. Conclusion: RYGB uniformly improved the medication requirements of patients with insulindependent T2DM. Although physiologic mechanisms likely contributed, early rapid weight loss was associated with the remission of T2DM. (Surg Obes Relat Dis 2009;5:305–309.) © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Diabetes; Remission; Excess weight loss; Gastric bypass
*Reprint requests: Eric DeMaria, M.D., Department of Surgery, Duke Endosurgery Center, Duke University Medical Center, DUMC-3288, Durham, NC 27710. E-mail: [email protected]
In 2005, 20.8 million people in the United States were living with diabetes mellitus (DM) [1]. A recent study estimated that the total monetary cost of DM in the United States in 2007 was $174 billion [2]. As the rate of obesity increases, the number of persons with DM globally is projected to be ⬎366 million by 2030, doubling in just 30 years from 2000 [3]. This major health crisis, with all its associ-
1550-7289/09/$ – see front matter © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2009.02.007
306
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
ated morbidity and mortality, demands efforts toward prevention and effective treatment. Currently, the most effective treatment for type 2 DM (T2DM) is bariatric surgery [4,5]. We sought to better understand the mechanism by which this improvement or remission of DM occurs. Many patients with T2DM are able to discontinue oral hypoglycemics just days after undergoing Roux-en-Y gastric bypass (RYGB) [6,7]. This has been the basis for investigations into the hormonal or “metabolic” mechanisms of surgery. Although T2DM includes a broad disease spectrum, we chose to narrow in on a more uniform patient population, those that require insulin. Patients with insulindependent T2DM lose less weight after RYGB than do those with non–insulin-dependent T2DM [6,8,9]. This group is thus well-suited for a study of the mechanisms of remission beyond weight loss. In this retrospective case series, we assessed the variables that correlated with postoperative improvement of DM, with the hypothesis that if remission of T2DM occurred by way of a weight lossindependent mechanism, we would see no association between weight loss and remission. Methods The data from 1546 patients undergoing open or laparoscopic Roux-en-Y gastric bypass at our institution from January 1, 2000 to October 1, 2006 were prospectively collected and entered into an electronic database. The operative technique included a 20 –30-cm3 gastric pouch, 35– 50-cm biliopancreatic limb, and a 50 –95-cm Roux limb. All had had a minimum of 12 months of recorded follow-up before this retrospective review. According to the past medical record or preoperative screening findings, 318 patients with a diagnosis of DM were identified by a fasting glucose level of ⬎126 mg/dL, hemoglobin A1c (HbA1c) of ⬎7.0%, and/or diabetic medication use. These patients were then categorized by their clinical treatment: insulin dependent or oral medication. Insulin dependence was defined as the previous failure (HbA1c ⬎7.0%) of strictly oral medications and the requirement for insulin therapy at surgery. Of the patients with DM, 4 were found to have T1DM and were excluded, leaving an insulin-dependent T2DM population of 71 patients. The criterion for the discontinuation of DM medications was the same for all patients. The initial management included a titration down to a fewer number and/or lower doses, followed by a trial of no medication if the HbA1c level was ⬍7.0%. Remission was defined as an HbA1c level of ⬍7.0% after the patient had not been taking any diabetic medications. The remission group (n ⫽ 35) was compared with the patients who continued to require medication (n ⫽ 36) using the preoperative demographics of gender, age, body mass index, excess body weight, HbA1c level, years with a DM diagnosis, and diabetic medication requirements, including years of insulin therapy. The insulin requirements
were calculated as the sum of the scheduled insulin doses, disregarding sliding scale corrections. Long-acting and short-acting insulin medications were treated as equivalent. Postoperative follow-up was conducted at regular intervals ⱕ7 years and included weight loss, percentage of excess weight loss (%EWL), HbA1c, months since surgery, and duration of follow-up. The Student t test was used for continuous data, and the chi-square test for categorical data, to compare the 2 groups (remission versus nonremission) at different points. Multivariate logistic regression analysis included those variables that were significant on Student’s t test, years with DM diagnosis, preoperative insulin dose (U/d), and the earliest point at which the %EWL was significant (3 months). All numbers are reported as the mean ⫾ standard deviation, unless otherwise noted. All P values are 2-tailed, with statistical significance at P ⬍ .05. Results Postoperatively, all patients achieved some improvement in insulin resistance, as determined by a reduction in HbA1c and the requirement of fewer and/or lower doses of diabetic medications. Of the 71 patients with insulin-dependent T2DM, 35 (49%) achieved remission. For this group, the mean start of remission was 6.2 ⫾ 4.9 months, with 19 (54%) of 35 patients in remission by their 3-month postoperative visit. No significant difference was found in gender distribution, age, preoperative excess body weight, body mass index, HbA1c level, or number of oral hypoglycemic medications between the insulin-dependent patients who achieved remission and those who continued to require medication (Table 1). The number of years with a DM diagnosis was a significant negative predictor for the remission of DM (P ⫽ .037). However, when taking into account the %EWL and the years requiring insulin therapy, it did not meet statistical significance (Table 2). The preoperative Table 1 Preoperative profile stratified by patient group Variable Gender (n) Male Female Age (yr) Excess body weight (kg) BMI (kg/m2) HbA1c (%) Diabetes diagnosis (yr) Insulin therapy (yr) Diabetic medications (n) Insulin dose (U/d)
Remission (n ⫽ 35)
Improved (n ⫽ 36)
21 14 46.0 ⫾ 10.1 77.6 ⫾ 22.8 48.7 ⫾ 7.9 8.2 ⫾ 1.8 7.2 ⫾ 6.8* 2.5 ⫾ 4.8 3.1 ⫾ 1.0 55 ⫾ 45*
27 9 48.4 ⫾ 8.9 76.8 ⫾ 17.3 49.4 ⫾ 6.8 8.5 ⫾ 1.6 11.0 ⫾ 6.6* 5.3 ⫾ 6.3 3.4 ⫾ .8 97 ⫾ 67*
P value .177
BMI ⫽ body mass index; HbA1c ⫽ hemoglobin A1c. Data presented as mean ⫾ standard error, except for gender. * P ⬍ .05.
.358 .871 .660 .420 .037* .078 .135 .003*
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
307
Table 2 Predictors of ID-T2DM remission after RYGB from binary logistic regression analysis Variable
OR (95% CI)
Coefficient ⫾ SE
P value
%EWL at 3 mo Insulin dose (U/d) Diabetes diagnosis (yr)
7.7* (1.3–45.60) .98* (.96–.99) .88 (.77–1.00)
2.0 ⫾ .9 ⫺.02 ⫾ .009 ⫺.12 ⫾ .07
.025* .018* .058
ID-T2DM ⫽ insulin-dependent type 2 diabetes mellitus; RYGB ⫽ Roux-en-Y gastric bypass; OR ⫽ odds ratio; CI ⫽ confidence interval; SE ⫽ standard error; %EWL ⫽ percentage of excess weight loss. ORs expressed as probability (off)/probability (on). * P ⬍ .05.
insulin dose was also a significant negative predictor of remission, but with only a small effect (odds ratio .98). One patient in remission at 6 months postoperatively required DM medication at 3 years postoperatively when his HbA1c had increased to ⬎7%. This interval of reduced glycemic control was associated with a weight regain of 7 kg. Of the 36 patients with improvement, only 10 (28%) continued to require insulin therapy postoperatively. All 10 required lower doses, decreasing from 97 ⫾ 67 U/d to 25 ⫾ 10 U/d. Of the 36 patients with improvement, 24 (66%) were able to achieve a HbA1c value of ⬍7.0% during the follow-up interval; however, in each case, the HbA1c was ⬎7.0% when the medication was withdrawn. The HbA1c values in the remission group were significantly lower at 3, 6, and 12 months of follow-up, attributable to the greater glycemic control (Table 3). Although not different preoperatively, the body mass index was significantly lower in the remission group at 12, 18, and 24 months. No difference was found in the average follow-up period or months since surgery between the 2 groups. Table 3 Postoperative profile stratified by patient group Variable BMI (kg/m2) 3 mo 12 mo Weight loss (kg) 3 mo 12 mo %EWL 3 mo 12 mo HbA1c 3 mo 12 mo Interval since surgery (mo) Follow-up (mo)
Remission (n ⫽ 35)
Improved (n ⫽ 36)
P value
39.2 ⫾ 7.3 31.9 ⫾ 5.5*
41.1 ⫾ 6.0 35.2 ⫾ 5.2*
.320 .018*
26.3 ⫾ 6.7 50.5 ⫾ 13.6*
23.6 ⫾ 7.9 41.5 ⫾ 13.0*
.201 .009*
36.7 ⫾ 10.2* 65.7 ⫾ 13.7*
30.9 ⫾ 9.2* 53.8 ⫾ 12.8*
.043* ⬍.001*
6.3 ⫾ .5* 6.0 ⫾ .7* 51 ⫾ 15 29 ⫾ 18
6.9 ⫾ .8* 6.8 ⫾ .5* 48 ⫾ 19 31 ⫾ 17
.007* ⬍.001* .597 .388
%EWL ⫽ percentage of excess weight loss; other abbreviations as in Table 1. All data reported as mean ⫾ standard deviation. * P ⬍ .05.
Fig. 1. Postoperative %EWL for patients in remission of insulin-dependent T2DM versus patients with improvement (taking reduced medications). Data reported as mean ⫾ standard error of mean.
The mean %EWL at 1 year was 59.9%. Because the %EWL is determined from the ideal body weight, a comparison between the remission and improved groups would only be appropriate if the distribution of gender, age, height, and weight were similar. This was the case for this relatively homogenous study population. Of all the variables followed, the %EWL correlated most strongly with the remission of DM (odds ratio 7.7 at 3 months). Those patients with insulin-dependent T2DM who had achieved remission had a greater postoperative %EWL at 3, 6, 12, 18, and 24 months compared with the group who continued to require medication (Fig. 1). At 1 year of follow-up, the group with the greatest quartile of %EWL (range 72.5–101) had a remission rate of ⬎80% compared with ⬍20% for those with the lowest quartile (range 0 – 49.2; Fig. 2). Discussion All patients in this observational study had clinical resolution or improvement in insulin-dependent T2DM after RYGB, and about one half achieved remission. This retrospective case series has revealed that a greater %EWL as early as 3 months after RYGB is associated with the remission of DM. The present study had several limitations. Only 71 patients met our inclusion criteria. Thus, our study was underpowered to detect a significant difference in the %EWL at a point ⬍3 months. Using the years with DM, years requiring insulin, and insulin dose as surrogates for the severity of disease was an oversimplification. These variables might not accurately represent the true underlying pathologic features of beta cell death and/or the potential for recovery and can be complicated by patient compliance and practitioner preferences. It is possible that the group that lost
308
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
Fig. 2. Percentage of patients in remission of insulin-dependent T2DM at 12 months after RYGB, grouped by %EWL.
less weight did so simply because they did not adhere to their diet or required insulin postoperatively. We did not have data for sliding-scale correction dosing or patient ethnicity; thus, these analyses were not available. Finally, the intracorporeal measurements of biliopancreatic and Roux limb lengths were estimated, the variation of which might have had some effect. The proposal of a weight loss-independent mechanism of T2DM remission after RYGB has arisen from 2 main clinical observations. First, many patients have experienced clinical resolution of DM just days after surgery, before significant weight loss has been achieved, and second, that combined restrictive/bypass procedures are more effective at producing remission than purely restrictive procedures. In the present study, our null hypothesis was disproved: weight loss is associated with the remission of DM. However, a greater %EWL at 3 months could still be considered a “late” effect. Other more immediate mechanisms that could contribute to remission include caloric restriction and changes in the enteroinsular axis. Caloric restriction in the perioperative period certainly plays a role. Kelley et al. [10] demonstrated a substantial decrease in fasting plasma glucose in patients with non– insulin-dependent DM after 7 days of a calorie-restrictive diet (800 kcal/d). This diet regimen is less restrictive than the vertical gastric banding described by Deitel et al. [11], in which T2DM was improved in patients before discharge from the hospital. The incretins, gastric inhibitory peptide and glucagonlike peptide 1, have received much attention. These hormones exhibit antidiabetic effects by suppressing glucagon
secretion, stimulating glycogen synthesis, and enhancing beta cell proliferation and inhibiting apoptosis [12]. Several studies have attempted to correlate the increases in these hormones after RYGB with the clinical improvement of DM [13–16]. A recent study by Laferrère et al. [17] found that the incretin effect was enhanced in 5 of 9 patients who had undergone gastric bypass, but not in any of 10 matched controls after diet-induced weight loss. Although this supports a weight loss-independent metabolic mechanism, the high variation of glucagon-like peptide 1 and gastric inhibitory peptide within the RYGB group, along with differences in caloric intake, interval during which the weight loss was achieved, and possible medication effects limits the strength of these conclusions. Other studies have shown that diet-induced weight loss or even the simple restriction of dietary fat can increase postprandial glucagon-like peptide 1 levels [18,19]. The strength of our study was in the relative uniformity of the patient population, omitting patients with impaired glucose tolerance or those treated by diet or oral hypoglycemics. Although a subgroup analysis of ethnicity was not performed, previous studies have shown no differences in T2DM remission rates between blacks and whites [20,21], which constitute the vast majority of our population. The surgical technique has been well standardized at our institution since the beginning of the study period. Any operatordependent variation in limb length would thus be small. Furthermore, it has been shown by Pinheiro et al. [22] that a near doubling of the length of the Roux limb from 150 cm to 250 cm does not lead to a significant difference in weight loss. It is possible that the greater insulin dose of the improved group, along with the continued insulin therapy in 31% of these patients postoperatively, contributed to the reduced weight loss [9]. Despite this confounding factor, little correlation was found between the postoperative insulin dose and the %EWL (R2 ⫽ .01). As reported by Schauer et al. [6], the severity of disease is an important predictor of remission. The present study is the first to show that not just insulin dependence, but the total daily insulin dose, is a negative predictor of remission. This lends additional support to the notion that patients with T2DM need to be considered surgical candidates early in their disease course. The years with a DM diagnosis did not reach significance in our logistic regression model, likely because of our small sample size. However, our observations have implications in assessing the treatment potential of surgery. A greater insulin dose might indicate that a patient’s failing beta cells have passed a threshold of recovery, more so than the number of years with the disease. From our study, it appears that no minimal threshold of %EWL exists that results in remission. The difference was in the rate of weight loss, which was greater in the remission group up until 12 months postoperatively. Because the weight loss was effectively stable in both groups after 12 months, early rapid weight loss might be the most critical.
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
Future studies could be aimed at maximizing early weight loss with adjuvant treatment (medical or exercise therapy) to potentially improve the remission rates. Our study was underpowered to detect a difference in weight loss at a point earlier than 3 months. According to our power analysis, it would have required nearly 2800 patients to detect a significant difference at 3 weeks. Even if we were able to enroll this many patients in a future study, it is hard to believe that a kilogram difference in weight loss would be the root cause of the remission. Our strongest evidence for a weight loss-independent effect, or metabolic effect, is that we found no minimal threshold of %EWL that led to remission. If the effect were rate dependent, what drives the weight loss? Despite uncertainty in the hormones responsible, an underlying metabolic mechanism is likely and future studies should target the 0 –3-month interval. A randomized study of gastric bypass versus calorie-matched controls must be undertaken to properly study this perioperative period. It is quite possible that the metabolic mechanism that helps patients’ achieve remission of T2DM also promotes greater weight loss. Conclusion Understanding the underlying mechanism whereby T2DM is improved after RYGB is of vital importance to the proper management of this disease. From the present study of a highly insulin-resistant subgroup, it can be stated that the remission of T2DM does not occur independent of weight loss. Although underlying metabolic mechanisms likely contribute, early rapid weight loss after RYGB is an important factor in the remission of T2DM. Disclosures The authors claim no commercial associations that might be a conflict of interest in relation to this article. References [1] Centers for Disease Control and Prevention. National diabetes fact sheet: general information and national estimates on diabetes in the United States. Bethesda: Centers for Disease Control and Prevention; 2005. [2] American Diabetes Association. Economic costs of diabetes in the U.S. in 2007. Diabetes Care 2008;31:596 – 615. [3] Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047–53.
309
[4] O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to moderate obesity with laparoscopic adjustable gastric banding or an intensive medical program. Ann Intern Med 2006;144:625–33. [5] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37. [6] Schauer PR, Burguera B, Ikramuddin S, et al. Effect of laparoscopic Roux-en-Y gastric bypass on type 2 diabetes mellitus. Ann Surg 2003;238:467– 85. [7] Wickremesekera K, Miller G, Naotunne TD, Knowles G, Stubbs RS. Loss of insulin resistance after Roux-en-Y gastric bypass surgery: a time course study. Obes Surg 2005;15:474 – 81. [8] Torquati A, Lutfi R, Abumrad N, Richards WO. Is Roux-en-Y gastric bypass surgery the most effective treatment for type 2 diabetes mellitus in morbidly obese patients? J Gastrointest Surg 2005;9:1112– 8. [9] Carbonell AM, Wolfe LG, Meador JG, Sugerman HJ, Kellum JM, Maher JW. Does diabetes affect weight loss after gastric bypass? Surg Obes Relat Dis 2007;3:279 – 82. [10] Kelley DE, Wing R, Buonocore C, Sturis J, Polonsky K, Fitzsimmons M. Relative effects of calorie restriction and weight loss in noninsulin-dependent diabetes mellitus. J Clin Endrocrinol Metab 1993;77: 1287–93. [11] Deitel M, Sidhu PS, Stone E. Effect of vertical banded gastroplasty on diabetes in the morbidly obese. Obes Surg 1991;1:113– 4. [12] Mauck MA, Meier JJ, Creutzfeldt W. Incretins and their analogues as new antidiabetic drugs (review). Drug News Perspect 2003;16: 413–22. [13] Rubino F, Gagner M, Gentileschi P, et al. The early effect of the Roux-en-Y gastric bypass on hormones involved in body weight regulation and glucose metabolism. Ann Surg 2004;240:236 – 42. [14] Morinigo R, Lacy AM, Casamitjana R, Delgado S, Gmis R, Vidal J. GLP-1 and changes in glucose tolerance following gastric bypass surgery in morbidly obese subjects. Obes Surg 2006;16:1594 – 601. [15] Morinigo R, Moize V, Musri M, et al. Glucagon-like peptide-1, peptide YY, hunger, and satiety after gastric bypass surgery in morbidly obese subjects. J Clin Endocrinol Metab 2006;91:1735– 40. [16] Rodieux F, Giusti V, D’Alessio DA, Suter M, Tappy L. Effects of gastric bypass and gastric banding on glucose kinetics and gut hormone release. Obesity 2008;16:298 –305. [17] Laferrère B, Teixeira J, McGinty J, et al. Effect of weight loss by gastric bypass surgery versus hypocaloric diet on glucose and incretin levels in patients with type 2 diabetes. J Clin Endocrinol Metab 2008;93:2479 – 85. [18] Verdich C, Toubro S, Buemann B, Lysgard MJ, Juul HJ, Astrup A. The role of postprandial releases of insulin and incretin hormones in meal-induced satiety-effect of obesity and weight reduction. Int J Obes Relat Metab Disord 2001;25:1206 –14. [19] Damci T, Yalin S, Balci H, et al. Orlistat augments postprandial increases in glucagon-like peptide 1 in obese type 2 diabetic patients. Diabetes Care 2004;27:1077– 80. [20] Sugerman HJ, Wolfe LG, Sica DA, Clore JN. Diabetes and hypertension in severe obesity and effects of gastric bypass-induced weight loss. Ann Surg 2003;237:751– 8. [21] Madan AK, Whitfield JD, Fain JN, et al. Are African-Americans as successful as Caucasians after laparoscopic gastric bypass? Obes Surg 2007;17:460 – 4. [22] Pinheiro JS, Schiavon CA, Pereira PB, et al. Long-long limb Rouxen-Y gastric bypass is more efficacious in treatment of type 2 diabetes and lipid disorder in super-obese patients. Surg Obes Relat Dis 2008;4:521–5.
Original article
Remission of type 2 diabetes after Roux-en-Y gastric bypass is associated with greater weight loss Brian E. Kadera, B.S.a, Kristian Lum, M.S.b, John Grant, M.D.c, Aurora D. Pryor, M.D.c, Dana D. Portenier, M.D.c, Eric J. DeMaria, M.D.c,* a Duke University School of Medicine, Durham, North Carolina Department of Statistical Science, Duke University, Durham, North Carolina c Department of Surgery, Duke University Medical Center, Durham, North Carolina Received May 2, 2008; revised February 11, 2009; accepted February 15, 2009 b
Abstract
Background: Physiologic studies in rodents and preliminary human studies have suggested that Roux-en-Y gastric bypass (RYGB) improves type 2 diabetes mellitus (T2DM) by way of metabolic changes, long before the bariatric or weight loss effects occur, leading to the concept of “metabolic surgery.” To test this hypothesis, we studied patients with insulin-dependent T2DM who underwent RYGB to determine whether T2DM remission in this treatment-resistant subgroup occurred independent of weight loss. Methods: Of all the patients undergoing RYGB from 2000 to 2006 (n ⫽ 1546) with ⱖ12 months of follow-up, 318 had T2DM (21%), and 75 (24%) of these were insulin dependent. Of the 75 patients, 4 were found to have T1DM (5.3%) and were excluded, leaving a study population of 71 patients. The patients who achieved remission, defined as a cessation of diabetic medications with a hemoglobin A1c level of ⬍7%, were compared with those who did not achieve remission. Statistical significance was set at P ⬍ .05, using the Student t test, chi-square test, and logistic regression analysis, as appropriate. Results: After RYGB, all 71 patients with insulin-dependent T2DM had achieved a reduction in the dose and/or number of medications at 29.6 ⫾ 17.0 months. Of these 71 patients, 35 (49%) demonstrated a remission of T2DM. The preoperative body mass index, age, number of medications, and hemoglobin A1c level did not differentiate between those who attained remission and those who still required diabetic medication. From the multivariate analysis, the significant factors associated with remission were the preoperative insulin dose and the percentage of excess weight loss. The percentage of excess weight loss was greater in the remission patients as early as 3 months postoperatively (P ⫽ .04) and also at 6, 12, 18, and 24 months. Conclusion: RYGB uniformly improved the medication requirements of patients with insulindependent T2DM. Although physiologic mechanisms likely contributed, early rapid weight loss was associated with the remission of T2DM. (Surg Obes Relat Dis 2009;5:305–309.) © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Diabetes; Remission; Excess weight loss; Gastric bypass
*Reprint requests: Eric DeMaria, M.D., Department of Surgery, Duke Endosurgery Center, Duke University Medical Center, DUMC-3288, Durham, NC 27710. E-mail: [email protected]
In 2005, 20.8 million people in the United States were living with diabetes mellitus (DM) [1]. A recent study estimated that the total monetary cost of DM in the United States in 2007 was $174 billion [2]. As the rate of obesity increases, the number of persons with DM globally is projected to be ⬎366 million by 2030, doubling in just 30 years from 2000 [3]. This major health crisis, with all its associ-
1550-7289/09/$ – see front matter © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2009.02.007
306
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
ated morbidity and mortality, demands efforts toward prevention and effective treatment. Currently, the most effective treatment for type 2 DM (T2DM) is bariatric surgery [4,5]. We sought to better understand the mechanism by which this improvement or remission of DM occurs. Many patients with T2DM are able to discontinue oral hypoglycemics just days after undergoing Roux-en-Y gastric bypass (RYGB) [6,7]. This has been the basis for investigations into the hormonal or “metabolic” mechanisms of surgery. Although T2DM includes a broad disease spectrum, we chose to narrow in on a more uniform patient population, those that require insulin. Patients with insulindependent T2DM lose less weight after RYGB than do those with non–insulin-dependent T2DM [6,8,9]. This group is thus well-suited for a study of the mechanisms of remission beyond weight loss. In this retrospective case series, we assessed the variables that correlated with postoperative improvement of DM, with the hypothesis that if remission of T2DM occurred by way of a weight lossindependent mechanism, we would see no association between weight loss and remission. Methods The data from 1546 patients undergoing open or laparoscopic Roux-en-Y gastric bypass at our institution from January 1, 2000 to October 1, 2006 were prospectively collected and entered into an electronic database. The operative technique included a 20 –30-cm3 gastric pouch, 35– 50-cm biliopancreatic limb, and a 50 –95-cm Roux limb. All had had a minimum of 12 months of recorded follow-up before this retrospective review. According to the past medical record or preoperative screening findings, 318 patients with a diagnosis of DM were identified by a fasting glucose level of ⬎126 mg/dL, hemoglobin A1c (HbA1c) of ⬎7.0%, and/or diabetic medication use. These patients were then categorized by their clinical treatment: insulin dependent or oral medication. Insulin dependence was defined as the previous failure (HbA1c ⬎7.0%) of strictly oral medications and the requirement for insulin therapy at surgery. Of the patients with DM, 4 were found to have T1DM and were excluded, leaving an insulin-dependent T2DM population of 71 patients. The criterion for the discontinuation of DM medications was the same for all patients. The initial management included a titration down to a fewer number and/or lower doses, followed by a trial of no medication if the HbA1c level was ⬍7.0%. Remission was defined as an HbA1c level of ⬍7.0% after the patient had not been taking any diabetic medications. The remission group (n ⫽ 35) was compared with the patients who continued to require medication (n ⫽ 36) using the preoperative demographics of gender, age, body mass index, excess body weight, HbA1c level, years with a DM diagnosis, and diabetic medication requirements, including years of insulin therapy. The insulin requirements
were calculated as the sum of the scheduled insulin doses, disregarding sliding scale corrections. Long-acting and short-acting insulin medications were treated as equivalent. Postoperative follow-up was conducted at regular intervals ⱕ7 years and included weight loss, percentage of excess weight loss (%EWL), HbA1c, months since surgery, and duration of follow-up. The Student t test was used for continuous data, and the chi-square test for categorical data, to compare the 2 groups (remission versus nonremission) at different points. Multivariate logistic regression analysis included those variables that were significant on Student’s t test, years with DM diagnosis, preoperative insulin dose (U/d), and the earliest point at which the %EWL was significant (3 months). All numbers are reported as the mean ⫾ standard deviation, unless otherwise noted. All P values are 2-tailed, with statistical significance at P ⬍ .05. Results Postoperatively, all patients achieved some improvement in insulin resistance, as determined by a reduction in HbA1c and the requirement of fewer and/or lower doses of diabetic medications. Of the 71 patients with insulin-dependent T2DM, 35 (49%) achieved remission. For this group, the mean start of remission was 6.2 ⫾ 4.9 months, with 19 (54%) of 35 patients in remission by their 3-month postoperative visit. No significant difference was found in gender distribution, age, preoperative excess body weight, body mass index, HbA1c level, or number of oral hypoglycemic medications between the insulin-dependent patients who achieved remission and those who continued to require medication (Table 1). The number of years with a DM diagnosis was a significant negative predictor for the remission of DM (P ⫽ .037). However, when taking into account the %EWL and the years requiring insulin therapy, it did not meet statistical significance (Table 2). The preoperative Table 1 Preoperative profile stratified by patient group Variable Gender (n) Male Female Age (yr) Excess body weight (kg) BMI (kg/m2) HbA1c (%) Diabetes diagnosis (yr) Insulin therapy (yr) Diabetic medications (n) Insulin dose (U/d)
Remission (n ⫽ 35)
Improved (n ⫽ 36)
21 14 46.0 ⫾ 10.1 77.6 ⫾ 22.8 48.7 ⫾ 7.9 8.2 ⫾ 1.8 7.2 ⫾ 6.8* 2.5 ⫾ 4.8 3.1 ⫾ 1.0 55 ⫾ 45*
27 9 48.4 ⫾ 8.9 76.8 ⫾ 17.3 49.4 ⫾ 6.8 8.5 ⫾ 1.6 11.0 ⫾ 6.6* 5.3 ⫾ 6.3 3.4 ⫾ .8 97 ⫾ 67*
P value .177
BMI ⫽ body mass index; HbA1c ⫽ hemoglobin A1c. Data presented as mean ⫾ standard error, except for gender. * P ⬍ .05.
.358 .871 .660 .420 .037* .078 .135 .003*
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
307
Table 2 Predictors of ID-T2DM remission after RYGB from binary logistic regression analysis Variable
OR (95% CI)
Coefficient ⫾ SE
P value
%EWL at 3 mo Insulin dose (U/d) Diabetes diagnosis (yr)
7.7* (1.3–45.60) .98* (.96–.99) .88 (.77–1.00)
2.0 ⫾ .9 ⫺.02 ⫾ .009 ⫺.12 ⫾ .07
.025* .018* .058
ID-T2DM ⫽ insulin-dependent type 2 diabetes mellitus; RYGB ⫽ Roux-en-Y gastric bypass; OR ⫽ odds ratio; CI ⫽ confidence interval; SE ⫽ standard error; %EWL ⫽ percentage of excess weight loss. ORs expressed as probability (off)/probability (on). * P ⬍ .05.
insulin dose was also a significant negative predictor of remission, but with only a small effect (odds ratio .98). One patient in remission at 6 months postoperatively required DM medication at 3 years postoperatively when his HbA1c had increased to ⬎7%. This interval of reduced glycemic control was associated with a weight regain of 7 kg. Of the 36 patients with improvement, only 10 (28%) continued to require insulin therapy postoperatively. All 10 required lower doses, decreasing from 97 ⫾ 67 U/d to 25 ⫾ 10 U/d. Of the 36 patients with improvement, 24 (66%) were able to achieve a HbA1c value of ⬍7.0% during the follow-up interval; however, in each case, the HbA1c was ⬎7.0% when the medication was withdrawn. The HbA1c values in the remission group were significantly lower at 3, 6, and 12 months of follow-up, attributable to the greater glycemic control (Table 3). Although not different preoperatively, the body mass index was significantly lower in the remission group at 12, 18, and 24 months. No difference was found in the average follow-up period or months since surgery between the 2 groups. Table 3 Postoperative profile stratified by patient group Variable BMI (kg/m2) 3 mo 12 mo Weight loss (kg) 3 mo 12 mo %EWL 3 mo 12 mo HbA1c 3 mo 12 mo Interval since surgery (mo) Follow-up (mo)
Remission (n ⫽ 35)
Improved (n ⫽ 36)
P value
39.2 ⫾ 7.3 31.9 ⫾ 5.5*
41.1 ⫾ 6.0 35.2 ⫾ 5.2*
.320 .018*
26.3 ⫾ 6.7 50.5 ⫾ 13.6*
23.6 ⫾ 7.9 41.5 ⫾ 13.0*
.201 .009*
36.7 ⫾ 10.2* 65.7 ⫾ 13.7*
30.9 ⫾ 9.2* 53.8 ⫾ 12.8*
.043* ⬍.001*
6.3 ⫾ .5* 6.0 ⫾ .7* 51 ⫾ 15 29 ⫾ 18
6.9 ⫾ .8* 6.8 ⫾ .5* 48 ⫾ 19 31 ⫾ 17
.007* ⬍.001* .597 .388
%EWL ⫽ percentage of excess weight loss; other abbreviations as in Table 1. All data reported as mean ⫾ standard deviation. * P ⬍ .05.
Fig. 1. Postoperative %EWL for patients in remission of insulin-dependent T2DM versus patients with improvement (taking reduced medications). Data reported as mean ⫾ standard error of mean.
The mean %EWL at 1 year was 59.9%. Because the %EWL is determined from the ideal body weight, a comparison between the remission and improved groups would only be appropriate if the distribution of gender, age, height, and weight were similar. This was the case for this relatively homogenous study population. Of all the variables followed, the %EWL correlated most strongly with the remission of DM (odds ratio 7.7 at 3 months). Those patients with insulin-dependent T2DM who had achieved remission had a greater postoperative %EWL at 3, 6, 12, 18, and 24 months compared with the group who continued to require medication (Fig. 1). At 1 year of follow-up, the group with the greatest quartile of %EWL (range 72.5–101) had a remission rate of ⬎80% compared with ⬍20% for those with the lowest quartile (range 0 – 49.2; Fig. 2). Discussion All patients in this observational study had clinical resolution or improvement in insulin-dependent T2DM after RYGB, and about one half achieved remission. This retrospective case series has revealed that a greater %EWL as early as 3 months after RYGB is associated with the remission of DM. The present study had several limitations. Only 71 patients met our inclusion criteria. Thus, our study was underpowered to detect a significant difference in the %EWL at a point ⬍3 months. Using the years with DM, years requiring insulin, and insulin dose as surrogates for the severity of disease was an oversimplification. These variables might not accurately represent the true underlying pathologic features of beta cell death and/or the potential for recovery and can be complicated by patient compliance and practitioner preferences. It is possible that the group that lost
308
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
Fig. 2. Percentage of patients in remission of insulin-dependent T2DM at 12 months after RYGB, grouped by %EWL.
less weight did so simply because they did not adhere to their diet or required insulin postoperatively. We did not have data for sliding-scale correction dosing or patient ethnicity; thus, these analyses were not available. Finally, the intracorporeal measurements of biliopancreatic and Roux limb lengths were estimated, the variation of which might have had some effect. The proposal of a weight loss-independent mechanism of T2DM remission after RYGB has arisen from 2 main clinical observations. First, many patients have experienced clinical resolution of DM just days after surgery, before significant weight loss has been achieved, and second, that combined restrictive/bypass procedures are more effective at producing remission than purely restrictive procedures. In the present study, our null hypothesis was disproved: weight loss is associated with the remission of DM. However, a greater %EWL at 3 months could still be considered a “late” effect. Other more immediate mechanisms that could contribute to remission include caloric restriction and changes in the enteroinsular axis. Caloric restriction in the perioperative period certainly plays a role. Kelley et al. [10] demonstrated a substantial decrease in fasting plasma glucose in patients with non– insulin-dependent DM after 7 days of a calorie-restrictive diet (800 kcal/d). This diet regimen is less restrictive than the vertical gastric banding described by Deitel et al. [11], in which T2DM was improved in patients before discharge from the hospital. The incretins, gastric inhibitory peptide and glucagonlike peptide 1, have received much attention. These hormones exhibit antidiabetic effects by suppressing glucagon
secretion, stimulating glycogen synthesis, and enhancing beta cell proliferation and inhibiting apoptosis [12]. Several studies have attempted to correlate the increases in these hormones after RYGB with the clinical improvement of DM [13–16]. A recent study by Laferrère et al. [17] found that the incretin effect was enhanced in 5 of 9 patients who had undergone gastric bypass, but not in any of 10 matched controls after diet-induced weight loss. Although this supports a weight loss-independent metabolic mechanism, the high variation of glucagon-like peptide 1 and gastric inhibitory peptide within the RYGB group, along with differences in caloric intake, interval during which the weight loss was achieved, and possible medication effects limits the strength of these conclusions. Other studies have shown that diet-induced weight loss or even the simple restriction of dietary fat can increase postprandial glucagon-like peptide 1 levels [18,19]. The strength of our study was in the relative uniformity of the patient population, omitting patients with impaired glucose tolerance or those treated by diet or oral hypoglycemics. Although a subgroup analysis of ethnicity was not performed, previous studies have shown no differences in T2DM remission rates between blacks and whites [20,21], which constitute the vast majority of our population. The surgical technique has been well standardized at our institution since the beginning of the study period. Any operatordependent variation in limb length would thus be small. Furthermore, it has been shown by Pinheiro et al. [22] that a near doubling of the length of the Roux limb from 150 cm to 250 cm does not lead to a significant difference in weight loss. It is possible that the greater insulin dose of the improved group, along with the continued insulin therapy in 31% of these patients postoperatively, contributed to the reduced weight loss [9]. Despite this confounding factor, little correlation was found between the postoperative insulin dose and the %EWL (R2 ⫽ .01). As reported by Schauer et al. [6], the severity of disease is an important predictor of remission. The present study is the first to show that not just insulin dependence, but the total daily insulin dose, is a negative predictor of remission. This lends additional support to the notion that patients with T2DM need to be considered surgical candidates early in their disease course. The years with a DM diagnosis did not reach significance in our logistic regression model, likely because of our small sample size. However, our observations have implications in assessing the treatment potential of surgery. A greater insulin dose might indicate that a patient’s failing beta cells have passed a threshold of recovery, more so than the number of years with the disease. From our study, it appears that no minimal threshold of %EWL exists that results in remission. The difference was in the rate of weight loss, which was greater in the remission group up until 12 months postoperatively. Because the weight loss was effectively stable in both groups after 12 months, early rapid weight loss might be the most critical.
B. E. Kadera et al. / Surgery for Obesity and Related Diseases 5 (2009) 305–309
Future studies could be aimed at maximizing early weight loss with adjuvant treatment (medical or exercise therapy) to potentially improve the remission rates. Our study was underpowered to detect a difference in weight loss at a point earlier than 3 months. According to our power analysis, it would have required nearly 2800 patients to detect a significant difference at 3 weeks. Even if we were able to enroll this many patients in a future study, it is hard to believe that a kilogram difference in weight loss would be the root cause of the remission. Our strongest evidence for a weight loss-independent effect, or metabolic effect, is that we found no minimal threshold of %EWL that led to remission. If the effect were rate dependent, what drives the weight loss? Despite uncertainty in the hormones responsible, an underlying metabolic mechanism is likely and future studies should target the 0 –3-month interval. A randomized study of gastric bypass versus calorie-matched controls must be undertaken to properly study this perioperative period. It is quite possible that the metabolic mechanism that helps patients’ achieve remission of T2DM also promotes greater weight loss. Conclusion Understanding the underlying mechanism whereby T2DM is improved after RYGB is of vital importance to the proper management of this disease. From the present study of a highly insulin-resistant subgroup, it can be stated that the remission of T2DM does not occur independent of weight loss. Although underlying metabolic mechanisms likely contribute, early rapid weight loss after RYGB is an important factor in the remission of T2DM. Disclosures The authors claim no commercial associations that might be a conflict of interest in relation to this article. References [1] Centers for Disease Control and Prevention. National diabetes fact sheet: general information and national estimates on diabetes in the United States. Bethesda: Centers for Disease Control and Prevention; 2005. [2] American Diabetes Association. Economic costs of diabetes in the U.S. in 2007. Diabetes Care 2008;31:596 – 615. [3] Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047–53.
309
[4] O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to moderate obesity with laparoscopic adjustable gastric banding or an intensive medical program. Ann Intern Med 2006;144:625–33. [5] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37. [6] Schauer PR, Burguera B, Ikramuddin S, et al. Effect of laparoscopic Roux-en-Y gastric bypass on type 2 diabetes mellitus. Ann Surg 2003;238:467– 85. [7] Wickremesekera K, Miller G, Naotunne TD, Knowles G, Stubbs RS. Loss of insulin resistance after Roux-en-Y gastric bypass surgery: a time course study. Obes Surg 2005;15:474 – 81. [8] Torquati A, Lutfi R, Abumrad N, Richards WO. Is Roux-en-Y gastric bypass surgery the most effective treatment for type 2 diabetes mellitus in morbidly obese patients? J Gastrointest Surg 2005;9:1112– 8. [9] Carbonell AM, Wolfe LG, Meador JG, Sugerman HJ, Kellum JM, Maher JW. Does diabetes affect weight loss after gastric bypass? Surg Obes Relat Dis 2007;3:279 – 82. [10] Kelley DE, Wing R, Buonocore C, Sturis J, Polonsky K, Fitzsimmons M. Relative effects of calorie restriction and weight loss in noninsulin-dependent diabetes mellitus. J Clin Endrocrinol Metab 1993;77: 1287–93. [11] Deitel M, Sidhu PS, Stone E. Effect of vertical banded gastroplasty on diabetes in the morbidly obese. Obes Surg 1991;1:113– 4. [12] Mauck MA, Meier JJ, Creutzfeldt W. Incretins and their analogues as new antidiabetic drugs (review). Drug News Perspect 2003;16: 413–22. [13] Rubino F, Gagner M, Gentileschi P, et al. The early effect of the Roux-en-Y gastric bypass on hormones involved in body weight regulation and glucose metabolism. Ann Surg 2004;240:236 – 42. [14] Morinigo R, Lacy AM, Casamitjana R, Delgado S, Gmis R, Vidal J. GLP-1 and changes in glucose tolerance following gastric bypass surgery in morbidly obese subjects. Obes Surg 2006;16:1594 – 601. [15] Morinigo R, Moize V, Musri M, et al. Glucagon-like peptide-1, peptide YY, hunger, and satiety after gastric bypass surgery in morbidly obese subjects. J Clin Endocrinol Metab 2006;91:1735– 40. [16] Rodieux F, Giusti V, D’Alessio DA, Suter M, Tappy L. Effects of gastric bypass and gastric banding on glucose kinetics and gut hormone release. Obesity 2008;16:298 –305. [17] Laferrère B, Teixeira J, McGinty J, et al. Effect of weight loss by gastric bypass surgery versus hypocaloric diet on glucose and incretin levels in patients with type 2 diabetes. J Clin Endocrinol Metab 2008;93:2479 – 85. [18] Verdich C, Toubro S, Buemann B, Lysgard MJ, Juul HJ, Astrup A. The role of postprandial releases of insulin and incretin hormones in meal-induced satiety-effect of obesity and weight reduction. Int J Obes Relat Metab Disord 2001;25:1206 –14. [19] Damci T, Yalin S, Balci H, et al. Orlistat augments postprandial increases in glucagon-like peptide 1 in obese type 2 diabetic patients. Diabetes Care 2004;27:1077– 80. [20] Sugerman HJ, Wolfe LG, Sica DA, Clore JN. Diabetes and hypertension in severe obesity and effects of gastric bypass-induced weight loss. Ann Surg 2003;237:751– 8. [21] Madan AK, Whitfield JD, Fain JN, et al. Are African-Americans as successful as Caucasians after laparoscopic gastric bypass? Obes Surg 2007;17:460 – 4. [22] Pinheiro JS, Schiavon CA, Pereira PB, et al. Long-long limb Rouxen-Y gastric bypass is more efficacious in treatment of type 2 diabetes and lipid disorder in super-obese patients. Surg Obes Relat Dis 2008;4:521–5.