Obesity Surgery

, Volume 23, Issue 12, pp 2020–2025

Remission of Type 2 Diabetes Mellitus Should Not Be the Foremost Goal after Bariatric Surgery


  • Ana M. Ramos-Levi
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
  • Andres Sanchez-Pernaute
    • Department of Surgery, Facultad de MedicinaComplutense University
  • Lucio Cabrerizo
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
  • Pilar Matia
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
  • Ana Barabash
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
  • Carmen Hernandez
    • Department of Surgery, Facultad de MedicinaComplutense University
  • Alfonso L. Calle-Pascual
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
  • Antonio J. Torres
    • Department of Surgery, Facultad de MedicinaComplutense University
    • Department of Endocrinology and Nutrition, Facultad de MedicinaComplutense University
    • Hospital Clinico San Carlos. Instituto de Investigación Sanitaria San Carlos (IdISSC), Facultad de MedicinaComplutense University
Original Contributions

DOI: 10.1007/s11695-013-1032-x

Cite this article as:
Ramos-Levi, A.M., Sanchez-Pernaute, A., Cabrerizo, L. et al. OBES SURG (2013) 23: 2020. doi:10.1007/s11695-013-1032-x



Remission of type 2 diabetes (T2D) is a desired outcome after bariatric surgery (BS). Even if this goal is not achieved, individuals who do not strictly fulfill remission criteria experience an overall improvement. The aim of this study was to evaluate the metabolic control status in patients considered as diabetes “non-remitters.”


A retrospective study of 125 patients (59.2 % women) with preoperative diagnosis of T2D who underwent BS in a single center (2006–2011) was conducted. We collected anthropometric and metabolic parameters before surgery and at 1-year follow-up. T2D remission was defined according to the 2009 consensus statement: glycosylated hemoglobin (HbA1c) <6 %, fasting glucose (FG) <100 mg/dLs, and absence of pharmacologic treatment. We evaluated metabolic status of non-remitters, according to the American Diabetes Association's (ADA) target recommendations: HbA1c <7 %, LDL-c <100 mg/dL, triglycerides <150 mg/dL, and HDL-c >40 (male) or >50 mg/dL (female). Statistics: analysis of variance.


Baseline characteristics (mean ± SD): age 53.5 ± 9.7 years, BMI 43.5 ± 5.6 kg/m2, time since diagnosis of T2D 7.7 ± 7.9 years, FG 162.0 ± 56.3 mg/dL, HbA1c 7.7 ± 1.6 %. ADA's target recommendations were present in 12 patients (9.6 %) preoperatively, and in 45 (36.0 %) at 1-year follow-up (p <0.001). Sixty-two (49.6 %) patients did not achieve diabetes remission; 26 (41.9 %) had now diet treatment, 30 (48.4 %) oral medications, and 6 (9.7 %) required insulin. Of the non-remitters, 57 (91.9 %) had HbA1c <7 % and 18 (40.0 %) achieved ADA's target recommendations. There were no differences between remitters and non-remitters in the number of individuals reaching ADA's combined metabolic control.


Although almost 50 % of the patients may not be classified as diabetes remitters, their significant improvement in metabolic control should be regarded as a success, according to most scientific societies' target recommendations.


Diabetes MellitusType 2Diabetes RemissionBariatric SurgeryRemission CriteriaObesityMorbid ObesityCombined Metabolic ControlADA Recommendations


Remission of type 2 diabetes (T2D) after bariatric surgery (BS) has become a “hot topic” over the last years. Bariatric surgical series published up to now have reported rates of T2D reversal ranging from 37 to 95 %, depending on the criteria used for definition, the cohort of patients' studied, the type of BS performed, and the time of follow-up considered [16].

This rising concern and focus on diabetes remission has sometimes left behind attention to the overall improvement achieved by those patients who are not considered as cured from the disease. However, as time goes by and evidence grows, studies have documented an increasing number of patients who do not achieve “biochemical remission” of hyperglycemia, according to the definition proposed by Buse et al. [7], or who fail to maintain long-term diabetes remission [812]; but their metabolic status has not been endorsed. The aim of this study is to evaluate metabolic control status in patients who may not be classified as under diabetes remission following BS.

Materials and Methods

Study Population

A retrospective study of a cohort of 657 bariatric surgeries performed in a single center over the period from January 2006 to December 2011 was carried out. We identified those individuals with preoperative diagnosis of type 2 diabetes, according to current guidelines of the American Diabetes Association (ADA) [13] and we extracted data from their medical charts. Patients with type 1 diabetes and latent autoimmune diabetes in adults were excluded from the study, as well as those cases of T2D who were not with hypoglycemic medications. By considering only patients with preoperative active hypoglycemic pharmacologic treatment, we aimed to focus on those individuals in whom diabetes was objectively established (i.e., their starting point was a non-remission status, regardless of their metabolic control) and had an impact on their everyday life. We collated information regarding clinical, anthropometric, and laboratory parameters, before surgery, and at 1-year follow-up. These included: diabetes duration and record of current hypoglycemic treatment; age, height, weight, and body mass index (BMI) calculated as weight(kg)/height(m2); percentage weight loss (%WL) and percentage excess weight loss (%EWL) [14]; and fasting glucose (FG), glycosylated hemoglobin (HbA1c), total cholesterol and its fractions (high-density lipoprotein [HDL-c], low-density lipoprotein [LDL-c]), and triglycerides. All patients signed a written informed consent prior to surgery in which it was specified that clinical and analytical data collected before the bariatric procedure and during follow-up could be potentially used in an anonymous way for investigation and publication. This study was approved by the Ethics Committee of the Hospital Clinico San Carlos and was in compliance with the Helsinki Declaration.

Bariatric Surgery

Bariatric procedures were considered under three categories: Roux-en-Y gastric bypass (RYGB) (45 patients, 36 %), biliopancreatic diversion (BPD) (65 cases, 52 %), and sleeve gastrectomy (SG) (15 cases, 12 %). Eligibility for each of them was decided by the treating physician (endocrinologist and/or surgeon), according to expertise and consideration of patients' previous diabetes medical history and comorbidities. All surgeries were performed laparoscopically by the same team in a single center. RYGB consisted of the creation of a small vertical gastric pouch of less than 30 ml, which was anastomosed to the jejunum in a Roux-en-Y fashion, 75–100-cm distally to the Treitz's ligament, and restoration of bowel continuity was obtained by an entero–entero anastomosis between the excluded biliary limb and the alimentary limb, at 100–150 cm form the gastrojejunostomy. BPD included two types of procedures: classic duodenal switch (17 cases) and single-anastomosis duodeno–ileal bypass with sleeve gastrectomy (SADI-S) (48 patients). Classic duodenal switch involved a “sleeve” gastric resection, followed by a duodeno–ileal bypass with a 250-cm alimentary limb and a 75–100-cm common channel, as described by Hess and Hess [15]. In SADI-S, sleeve gastrectomy was performed over a 54 Fr (18 mm) bougie, and, after the duodenal division, the proximal duodenal stump was anastomosed in an end-to-side way to the ileum, at 250 cm form the ileocecal valve, thus creating a long biliopancreatic channel and a 250-cm common + alimentary limb [16]. Because SADI-S consists of a one-loop duodenal switch, which has proved to act in a similar way to classic duodenal switch, both techniques were considered together for statistical analyses. SG involved a vertical resection of the stomach, calibrated with a 42 Fr (14 mm) bougie, from 4 cm proximal to the pylorus, to the pericardial fat.

Definitions of Diabetes Remission and Optimal Metabolic Control

Definition of T2D remission was considered as according to Buse et al. [7]: HbA1c <6 % and FG <100 mg/dL, in the absence of active pharmacologic treatment. Combined metabolic status was regarded as optimal following current ADA's target recommendations for glucose and lipid control: HbA1c <7 %, c-LDL <100 mg/dL, triglycerides <150 mg/dL, and c-HDL >40 mg/dL (male) or >50 mg/dL (female) [17].

Statistical Analysis

IBM SPSS Statistics version 19.0 (Chicago, IL, USA) was used for statistical analysis. For descriptive data, results were expressed as mean ± standard deviation for continuous variables, and as frequency and percentages for categorical variables. Paired t test for comparison of variables before and after BS was performed. Possible existing associations of baseline characteristics with remission rates were evaluated using two-sided analysis of variance. Comparison between categorical groups was assessed with chi-square test. Significance was considered at p <0.05.


A total of 141 individuals were diagnosed with T2D prior to BS. Of these, full clinical, anthropometric and laboratory data at baseline and 1-year follow-up was available in 125 patients (59.2 % women).

Baseline characteristics were: age 53.6 ± 9.7 years, BMI 43.5 ± 5.6 (range 29.8–61.0) kg/m2, duration of T2D 7.7 ± 7.9 years, FG 162.0 ± 56.3 mg/dL, and HbA1c 7.7 ± 1.6 %. Before undergoing surgery, 54 (43.2 %) patients were on insulin therapy, 48 (38.4 %) had an HbA1c level <7 %, and only 12 (9.6 %) fulfilled the four ADA's target recommendations for glucose and lipid control.

One year after BS, T2D remission, as defined by the 2009 consensus statement [7], was not achieved in 62 (49.6 %) individuals. Table 1 displays baseline and 12-month follow-up patients' characteristics according to their diabetes remission status. BMI, FG, HbA1c, LDL-c, and triglycerides were significantly reduced in all patients 1 year after BS (p <0.001 in all cases), whilst HDL-c increased (p = 0.071). A longer duration of T2D and higher preoperative values of FG and HbA1c were associated to non-remission. %WL was greater in those who achieved remission in comparison to those who did not (p = 0.002). Previous treatment with insulin was more frequent in those who failed to obtain T2D remission at 1-year follow-up.
Table 1

Patients' baseline and 12-month follow-up characteristics according to remission status. Values show mean ± SD or number of patients and percentages (%)




No remission


Number of patients


63 (50.4)

62 (49.6)



74 (59.2)

31 (49.2)

43 (69.3)


Age (years)

53.6 ± 9.7

52.0 ± 10.0

55.1 ± 9.3


Preop-BMI (kg/m2)

43.5 ± 5.6

44.7 ± 5.8

42.2 ± 5.2


12 m–BMI (kg/m2)

29.1 ± 5.3

28.8 ± 5.3

29.4 ± 5.3 23.2



32.8 ± 9.3

35.4 ± 8.2

30.2 ± 9.7



70.4 ± 21.2

73.6 ± 18.6

67.1 ± 23.4


Preop–FG (mg/dL)

162.0 ± 56.3

150.4 ± 52.7

173.8 ± 57.7


12 m FG (mg/dL)

99.9 ± 23.6

85.9 ± 7.6

114.2 ± 25.8


Preop-HbA1c (%)

7.7 ± 1.6

7.3 ± 1.2

8.0 ± 1.8


12 m HbA1c (%)

5.5 ± 0.8

5.1 ± 0.6

6.0 ± 1.8


Preop-LDL-c (mg/dL)

102.1 ± 31.1

100.9 ± 33.4

103.5 ± 28.6


12 m LDL-c (mg/dL)

81.3 ± 32.4

78.6 ± 35.4

84.2 ± 29.0


Preop-HDL-c (mg/dL)

48.9 ± 11.3

47.3 ± 11.1

50.4 ± 11.4


12 m HDL-c (mg/dL)

50.8 ± 13.3

51.0 ± 12.3

50.5 ± 14.2


Preop-Tg (mg/dL)

195.7 ± 128.7

197.7 ± 140.8

193.7 ± 115.8


12 m Tg (mg/dL)

114.0 ± 54.1

97.3 ± 35.2

131.0 ± 64.2


Duration of diabetes (years)

7.7 ± 7.9

4.9 ± 4.0

10.7 ± 9.7


Previous insulin use

54 (43.2)

17 (27.0)

37 (59.7)


p Values for comparisons using chi-square analysis for categorical variables and analysis of variance for continuous variables

Preop preoperative, 12 m 12-month, BMI body mass index, %WL percentage body weight loss at 12 months, %EWL percentage excess body weight loss at 12 months, FG fasting glucose, HbA1c glycosylated hemoglobin, LDL low-density lipoprotein cholesterol, HDL-c high-density lipoprotein cholesterol, Tg triglycerides.

aRemission of diabetes defined according to Buse et al. [7]: HbA1c <6 %, FG <100 mg/dl and no pharmacologic treatment.

Table 2 shows patients' type of hypoglycemic treatment, according to T2D remission status, preoperatively, and 12 months after BS. Of the 62 patients who were classified as non-remission at this time of follow-up, 26 cases (41.9 %) were now on diet, 30 (48.4 %) received oral agents, and only 6 (9.7 %) individuals required insulin, in comparison to 37 patients (59.7 %) that required it before the surgical procedure (p <0.001).
Table 2

Preoperative and 12-month follow-up hypoglycemic treatment, according to remission status [6], number of patients, and percentage


Remission of T2D

No remission of T2D


12-month follow-up


12-month follow-up

No treatment

0 (0)

0 (0)

0 (0)

26 (41.9)

Oral agents

46 (23.0)

0 (0)

25 (40.3)

25 (48.4)


17 (27.0)

0 (0)

37 (59.7)

6 (9.7)

A total of 57 (91.9 %) patients of the non-remitters had a postoperative HbA1c level <7 % and 18 (40.0 %) fulfilled all four ADA's target recommendations for combined metabolic control. When patients were stratified according to their T2D remission status, there were no differences in the total number of individuals who reached these combined target recommendations (Fig. 1) 1 year after BS.
Fig. 1

Percentages of patients with HbA1c <7 % (a) and ADAs recommendations of combined metabolic control [HbA1c <7 %, LDL-c <100 mg/dl, triglycerides <150 mg/dl, and HDL-c >40 (male) or >50 (female) mg/dl] (b), preoperatively and at 12-month follow-up (preop and 12 m, respectively), according to remission status. t Test for paired samples showed differences between preoperative and postoperative values (*p <0.001). Chi-square analysis (Fisher's exact test) was used for comparison between patients in remission and non-remission of T2D in each group: for Preop HbA1c <7 %, p = 0.005 (#); for 12 m-HbA1c <7 %, p = 0.028 (#); for Preop combined control, p = 0.363; for 12 m-combined control, p = 0.137

Table 3 shows patients' characteristics according to the type of BS. Based on the treating physician's expertise, in general, patients with a longer T2D duration, insulin-users, and with a worse preoperative metabolic control were preferably assigned to BPD, except in some cases in which LSG was used as a first-step technique. One year after BS, we did not find differences in remission of T2D according to the type of procedure performed: 22 out of 45 patients (48.9 %) in those who underwent RYGB; 34 out of 65 cases (53.3 %) in the BPD group; and 7 out of 15 (46.7 %) in those with SG (p = 0.896). However, the number of patients with T2D resolution was greater in those who underwent BPD. ADA's target recommendations for combined metabolic control were best achieved after RYGB (22 out of 45 patients, 48.9 %) or BPD (21 out of 65 cases, 32.3 %) in comparison to SG (2 out of 15, 13.3 %) (p = 0.031).
Table 3

Patients' characteristics according to the type of BS performed


Type of bariatric surgery




p Value

Number of patients





Age (years)

53.4 ± 7.9

54.0 ± 10.0

52.1 ± 13.5


Previous insulin use

13 (28.9)

33 (50.8)

8 (53.3)


Diabetes duration (years)

5.0 ± 4.3

10.1 ± 9.4

5.6 ± 6.0


Preop-BMI (kg/m2)

42.9 ± 5.7

43.6 ± 5.5

44.8 ± 5.8


12 m BMI (kg/m2)

29.3 ± 5.1

28.0 ± 5.0

33.6 ± 4.7



31.7 ± 7.4

35.5 ± 9.3

24.5 ± 9.5



69.4 ± 19.3

75.8 ± 20.3

49.9 ± 18.6


Preop-HbA1c <7 %

23 (53.5)

20 (31.7)

5 (35.7)


12 m HbA1c <7 %

45 (100)

62 (95.4)

13 (86.7)


Preop-ADA objectives

6 (13.3)

4 (6.2)

2 (13.3)


12 m ADA objectives

22 (48.9)

21 (32.3)

2 (13.3)


Values show mean ± SD or number of patients and percentages of the column (%). p Values are shown for chi-square test (categorical values) and analysis of variance (continuous variables)

Preop preoperative, 12 m 12-month; BMI body mass index, %WL percentage body weight loss at 12 months, %EWL percentage excess body weight loss at 12 months, HbA1c glycosylated hemoglobin, ADA objectives ADA's target recommendations for combined metabolic control


In this study, almost 50 % of patients did not achieve T2D remission according to the criteria established by Buse et al. [7]. This rate is similar to recently published papers, in which it had been evidenced that the use of more stringent cut-off levels to define diabetes resolution would lower remission percentages [5, 6, 18, 19]. We think that this rate is more trustworthy and realistic than some previously reported [2], especially considering the wide variety of patients' preoperative characteristics such as BMI, diabetes duration and insulin treatment in diverse publications. It has been observed across various reports that these factors influence diabetes remission [20, 21].

Our study identified preoperative predictors of T2D remission in the same way as other authors: a shorter duration of the period since the diagnosis of T2D, the absence of previous insulin treatment, and a greater %WL were associated to higher rates of remission [5, 21, 22].

Since Pories et al. [23] described that resolving diabetes after an operation could be, in fact, achieved, growing interest and subsequent publications have evaluated in detail multiple mechanisms and patients' features related to remission [24]. It has even led to scientific societies to include bariatric surgery in treatment algorithms for T2D [17, 2527]. This has inevitably led to a subtle forgetfulness of those patients who were not able to achieve it. Some studies have evaluated why diabetes may not resolve in some patients after BS [28], but there are not so many reports that address the overall improvement of non-remitters regarding metabolic control, compare them to those under strict remission, and analyze differences between them, besides glycemic levels.

Indeed, it has been observed that an increasing number of patients fail to achieve biochemical remission of hyperglycemia, or are unable to sustain this effect despite a previous short-term postoperative success [11]. In our study, almost 50 % of individuals were not able to fulfill diabetes remission criteria 1 year after the bariatric procedure. Definition of remission has been highly controversial [7], and we believe that considering it solely in terms of biochemical normality may lead to underestimating the real global value of BS for individuals with long term and poorly controlled T2D.

For instance, FG, HbA1c and BMI at 12-month follow-up were significantly lower than preoperative ones; hence, an overall amelioration of obesity and diabetes is inferred. Moreover, the number of patients taking insulin went from 54 (43.2 %) preoperatively, to 6 (9.7 %) at 1-year follow-up. The vast majority of the non-remitters (90.3 %) were controlled with oral medications or with diet, which again, supports the observation that their glucose control had significantly improved.

If we evaluate glucose metabolism on its own and consider ADA's HbA1c cut-off value of <7 % [17], there were differences between remitters and non-remitters in the total number of patients achieving it, since the definition of remission includes the HbA1c level. However, if we compare the overall combined metabolic control of remitters and non-remitters, there were no significant differences. This should make us reflect about the fact that, although diabetes remission was not strictly achieved according to a stringent definition, either because of continued hypoglycemic treatment or FG/HbA1c levels slightly above reference values, the overall metabolic control was more than optimal. And this is especially desirable when patients' preoperative characteristics are less favorable for remission. If baseline features are less predictive of diabetes remission (for example, long duration of T2D or prior insulin treatment), BS should not be disregarded as a treatment option because significant amelioration is actually possible and patients may achieve an optimal glycemic control with minimum active hypoglycemic treatment, such as one metformin tablet per day.

The limitations of our study are mainly related to the retrospective analysis of data collected from medical records and the fact that follow-up was limited to 12 months. Also, due to the absence of aleatorization of patients to the different types of surgeries and the small number of SG performed, results regarding the differences between BS should be interpreted with caution. However, this was not the main objective of our study, and we consider that its influence in our main findings is not relevant, since what we aimed to analyze was if the overall improvement of patients endured despite not achieving strict T2D remission, at least in the short term, and if both groups were comparable besides glycemic control. Finally, we have not considered blood pressure measurements, which are also taken into account in current standards of medical care [17]. Nonetheless, once again, as the aim of our study was to evaluate overall metabolic control, blood pressure measurements could be considered a secondary target.

In conclusion, patients with a worse preoperative profile of T2D, who a priori have a reduced chance of complete resolution after BS according to specific criteria, still aim to achieve an improved combined metabolic control. This, in our opinion, should be viewed as a success, especially according to most scientific societies' target recommendations, even if the patient still requires a minimum amount of hypoglycemic treatment. Thus, strict T2D remission, although desirable and targeted, should not be the only and foremost goal after BS. Nevertheless, well-controlled prospective follow-up studies are still needed to evaluate long-term effects of BS on diabetes remission and overall metabolic control.


The authors would like to acknowledge grant support from the Fundación Mutua Madrileña de Investigación Biomédica AP 89592011.

Conflict of Interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer Science+Business Media New York 2013