Obesity Surgery

, Volume 24, Issue 11, pp 1835–1842

Long-Term Effects of Laparoscopic Roux-en-Y Gastric Bypass on Diabetes Mellitus, Hypertension and Dyslipidaemia in Morbidly Obese Patients

Authors

    • Department of SurgeryRijnstate Hospital
  • B. Betzel
    • Department of SurgeryRijnstate Hospital
  • J. Homan
    • Department of SurgeryRijnstate Hospital
  • E. O. Aarts
    • Department of SurgeryRijnstate Hospital
  • N. Ploeger
    • Department of SurgeryRijnstate Hospital
  • H. de Boer
    • Internal MedicineRijnstate Hospital
  • Th. J. Aufenacker
    • Department of SurgeryRijnstate Hospital
  • C. J. H. M. van Laarhoven
    • Department of SurgeryRadboud University Medical Centre
  • I. M. C. Janssen
    • Department of SurgeryRijnstate Hospital
  • F. J. Berends
    • Department of SurgeryRijnstate Hospital
Original Contributions

DOI: 10.1007/s11695-014-1310-2

Cite this article as:
Dogan, K., Betzel, B., Homan, J. et al. OBES SURG (2014) 24: 1835. doi:10.1007/s11695-014-1310-2

Abstract

Background

Severely obese patients have an increased risk for developing metabolic complications such as type 2 diabetes mellitus (T2DM), dyslipidaemia (DL) and hypertension (HT). The aim of the present study is to research the effect of a primary laparoscopic Roux-en-Y gastric bypass (LRYGB) on T2DM, HT and DL in the long-term.

Methods

Fifty-two out of 89 (58 %) adult severely obese patients with T2DM who had received a LRYGB between January 2000 and December 2008 were evaluated. Primary outcome of evaluation was remission of T2DM according to the definition of 2009 consensus statement. Complete remission was defined as achievement fasting plasma glucose (FPG) of <5.6 mmol/l (<100.8 mg/dL) and HbA1c <42 mmol/mol (<6.0 %)) without glucose-lowering medication for at least 1 year. Partial remission was defined as a FPG of 5.6–6.9 mmol/l (100.8–124.2 mg/dL) and HbA1c 42–48 mmol/mol (6.0–6.5 %), without glucose-lowering medication for at least 1 year. Remission of T2DM was considered if the patient met the criteria for complete or partial remission. Secondary outcomes were remission of HT, DL and changes in medication use.

Results

Patients had a mean age of 47.5 ± 9.6 years, body mass index of 46.6 ± 6.4 kg/m2 and a mean duration of T2DM of 6.1 ± 5.4 years at the time of surgery. The mean post-operative follow-up period was 6.9 ± 2.3 years. At the end of the follow-up, mean weight loss was 60 ± 24 % excess weight loss (EWL) and 26 ± 10 % total body weight loss (TBWL). Mean HbA1c level had significantly decreased from 64.8 ± 19.7 mmol/mol to 46.4 ± 12.9 mmol/l (p < 0.0001). Overall medication use was reduced from 85 % to 37 % of the patients (p < 0.0001), while the number of insulin users was reduced from 40 % to 6 % (p < 0.0001). Nineteen percent of the patients had a relapse of T2DM during follow-up. Pre-operative HbA1ac level (odds ratio 0.911, p = 0.020) and duration of T2DM (odds ratio 0.637, p = 0.010) were independent risk factors for failed remission of T2DM. The number of patients with HT was significantly reduced from 73 % to 54 % (p = 0.042), and number of patients with DL was non-significantly decreased from 71 % to 54 % (p = 0.068).

Conclusions

The laparoscopic RYGB operation results in a sustained EWL of 60 % (26 % TBWL) with 52 % long-term remission of T2DM. However, 19 % of the patients had a relapse of their T2DM. Furthermore, HT and DL improved markedly.

Keywords

Morbid obesityBariatric surgeryRoux-en-Y gastric bypassDiabetes mellitus type 2HypertensionDyslipidaemia

Introduction

Type 2 diabetes mellitus (T2DM) is an epidemic disease with an estimated 285 million people affected worldwide and closely related to obesity [1, 2]. In addition to T2DM and elevated glucose levels, obese patients often have dyslipidaemia (DL) and hypertension (HT) which further adds to an increased risk for micro- and macrovascular long-term complications that significantly decrease the quality of life and life expectancy of these patients [3].

Weight loss is the most important therapeutic intervention in obese patients with T2DM because it restores insulin sensitivity and improves β-cell function [48]. Weight-reducing surgery, i.e. bariatric surgery, is far more effective in achieving sustained weight loss than conventional (non-surgical) treatment [9, 10]. The Roux-en-Y gastric bypass (RYGB) procedure is currently the most performed bariatric procedure and has been shown to reduce the metabolic abnormalities of severely obese T2DM patients with a body mass index (BMI) >35 kg/m2 [9, 1114]. Therefore, the Diabetes Surgery Summit Consensus Conference promotes bariatric surgery as an alternative therapy for the treatment of obesity-related (BMI > 35 kg/m2) T2DM [15, 16].

However, long-term post-operative data on severely obese T2DM patients are limited. The aim of the present study was to evaluate the long-term effectiveness of laparoscopic RYGB (LRYGB) on the remission of T2DM, DL and HT.

Patients and Methods

Patient Selection and Outcome

All adult obese patients (age > 18 years, BMI > 35 kg/m2) with T2DM who underwent a primary LRYGB between January 2000 and December 2008 were invited to our outpatient clinic for a long-term follow-up visit. First invitations were sent by letter, followed by contact by phone if the patient had not responded. The local board of ethics of the Rijnstate hospital waved the necessity for informed consent since the follow-up of patients after surgery is already part of the postoperative protocol. All patients gave their permission for conducting this study.

Patients were excluded when they had type 1 diabetes mellitus or if the LRYGB had been a revisional bariatric procedure. All patients underwent a pre-operative multidisciplinary screening by a multidisciplinary team, including a surgeon, endocrinologist, dietician, psychologist and physiotherapist. All patients met the criteria for bariatric surgery according to the National Institutes of Health Consensus Conference [17].

The patients were selected from our prospective collected (since 2000) database. Operation type and date, medical history, weight (kilograms), height (centimetres), co-morbidities, medication use and follow-up were registered in this database. The following parameters were documented at the follow-up: medical history, height (centimetres), weight (kilograms), BMI (kilograms per square meter), weight loss expressed in percentage excess weight loss (percent EWL) and percentage total body weight loss (percent TBWL), presence of co-morbidities, medication use and laboratory tests after an overnight fasting, including fasting plasma glucose (FPG), HbA1c, insulin, C-peptide, total cholesterol (TC), HDL-cholesterol (HDL-C), LDL-cholesterol (LDL-C) and triglycerides (TG).

Primary endpoint was remission of T2DM according to the definition of 2009 consensus statement [18]. Patients were classified into three groups according to their remission rate. Complete remission was defined as achievement of ‘normal’ glucose parameters (FPG of <5.6 mmol/l (<100.8 mg/dL) and HbA1c <42 mmol/mol (<6.0 %)) without glucose-lowering medication for at least 1 year [19]. Partial remission was defined as a FPG of 5.6–6.9 mmol/l (100.8–124.2 mg/dL) and HbA1c 42-48 mmol/mol (6.0–6.5 %), without glucose-lowering medication for at least 1 year. Remission of T2DM was considered if the patient met the criteria for complete or partial remission.

Secondary outcome measures were remission of other metabolic co-morbidities such as HT and DL. Patients were considered to have hypertensive control if they had an acceptable blood pressure (systolic of <160 mmHg and diastolic of <90 mmHg measured on the left arm) without antihypertensive medication. The DL treatment goal in patients with T2DM is to lower LDL-C level (<2.6 mmol/l (<100.5 mg/dL)), raise HDL-C level (>1.02 mmol/l (>39.4 mg/dL)) and lower TG level (<1.7 mmol/l (<150.7 mg/dL)) [20]. Dosage of medication use was expressed in daily defined dose (DDD) as recommended by the World Health Organization (WHO) [21]. The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults. DDDs provide a fixed unit of measurement independent of price, currencies, package size and strength enabling to assess trends in drug consumption and to perform comparisons between population groups. For example, 150 mg of metoprolol (beta-blocker) is equal to 1 DDD, while 75 mg of atenolol (beta-blocker) is equal to 1 DDD.

Surgical Procedure

All patients underwent a LRYGB procedure in a single bariatric centre. Antecolic antegastric bypass was performed with a small gastric pouch of 30–50 mL and fixed biliopancreatic limb (BPL) of 50 cm. Patients with BMI < 50 kg/m2 had a Roux limb of approximately 100 cm (range, 75–125 cm) and patients with BMI > 50 kg/m2 a Roux limb of 150 cm (range, 125–175 cm) [22].

Statistical Analysis

Results are presented as mean values ± standard deviation (SD), unless otherwise specified. Variance between groups was analysed using unpaired t test or Mann–Whitney U test (in cases of not normally distributed data) for continuous data and Chi-square test or Fisher exact test for categorical data. Paired t test was used for variance within the group. Parameters were selected for multivariate regression analysis for detection of independent risk factors for remission of T2DM if there was a difference of p < 0.05 for individual parameters. Data were analysed using IBM® SPSS® (version 20.0 for Windows) statistical software, and all figures were created using Graph Path Prism® (version 5.01). A p < 0.05 was considered significant.

Results

Eighty-nine patients with T2DM out of a total of 310 severely obese patients underwent a primary LRYGB and were eligible for long-term follow-up. Two patients died post-operatively after 1 (alcohol intoxication) and 7 years (sudden death, known with M. non-Hodgkin, splenectomy, myocardial infarction and nephropathy). Twenty-five patients were lost to follow-up, and 12 patients refused follow-up for personal reasons. In total, 52 (58.4 %) patients completed the long-term follow-up and were included in the analysis. Their mean age at the time of operation was 47.5 ± 9.6 years, with a BMI of 46.6 ± 6.4 kg/m2. Mean T2DM duration at surgery was 6.1 ± 5.4 years, and 21 (40 %) patients were on insulin therapy. Table 1 shows the baseline characteristics.
Table 1

Baseline characteristics

 

Total group (n = 52)

Age (years)

47.5

±9.6

Female gender

31

(59.6 %)

Weight (kg)

140.6

±21.8

BMI (kg/m2)

46.6

±6.4

EW (kg)

64.8

±18.9

T2DM

  

Duration (years)

6.1

±5.4

Medication use (n, %)

44

(84.6 %)

Insulin treatment (n, %)

21

(40.4 %)

Data are mean (±standard deviation), unless otherwise specified

BMI body mass index, EW excess weight, T2DM type 2 diabetes mellitus

Mean post-operative follow-up period was 82.2 ± 27.6 months (6.9 ± 2.3 years). At the end of the follow-up, mean weight and BMI were 103.7 ± 21.1 kg (−36.9 ± 15.6 kg) and 34.3 ± 6.6 kg/m2 (−12.3 ± 5.0 kg/m2), respectively, with an overall %EWL of 59.5 ± 24.4 % and %TBWL of 26.3 ± 10.1 %.

Glucose Metabolism

The changes in parameters of glucose metabolism and that of glucose-lowering medication are summarised in Table 2. In total, 27 (51.9 %) patients had remission of their T2DM at their last follow-up visit (p < 0.0001). In the overall population, both mean FPG and HBA1c levels had decreased significantly from 9.3 ± 3.7 mmol/l to 7.2 ± 2.8 mmol/l and from 64.8 ± 19.7 mmol/mol to 46.4 ± 12.9 mmol/l, respectively (p < 0.0001). Medication use was reduced from 44 (85 %) patients to 19 (37 %) (p < 0.0001). All types of diabetes medication were reduced in numbers and dosages compared with baseline. The number of insulin users was reduced from 21 (40 %) to 3 (6 %) (p < 0.0001), and the mean insulin dose was reduced from 2.85 ± 2.24 DDD to 0.20 ± 0.51 DDD (p < 0.0001). In all three groups (remission, temporary remission and no remission), metabolic improvement was observed. In the group with long-term remission, FPG and HbA1c levels had decreased significantly from 7.7 ± 2.4 to 5.5 ± 0.8 (p < 0.0001) and 51.8 ± 12.8 to 38.6 ± 4 mmol/mol (p < 0.0001), respectively. In patients who had a temporary remission, FPG and HbA1c levels had decreased by −2.2 ± 4.0 mmol/l to 8.7 ± 3.1 mmol/l (p = 0.137) and by −27.4 ± 21 mmol/mol to 53.6 ± 9.2 mmol/mol (p = 0.004), respectively. Even patients without remission had a lowering of FPG and HbA1c levels at their last follow-up visit; FPG had decreased from 11.1 ± 4.2 to 8.8 ± 3.3 mmol/l (p = 0.028) and HbA1c level from 79.5 ± 15.6 to 54.5 ± to13.6 mmol/mol (p < 0.001).
Table 2

Follow-up glucose metabolism

 

Before surgery (n = 52)

After surgery (n = 52)

P value

T2DM (n, %)

52 (100)

25(48)

<0.0001

Medication use (n, %)

44 (85)

19 (37)

<0.0001

Biguanide (n, %)

39 (75)

19 (37)

<0.0001

DDD

0.76 ± 0.41

0.37 ± 0.47

<0.0001

SU-derivate (n, %)

16 (31)

6 (12)

0.016

DDD

1.85 ± 1.34

0.56 ± 0.97

0.002

Thiazolidinedione (n, %)

1 (2)

0 (0)

1.000

DDD

0.67

0.00

Insulin (total) (n, %)

21 (40)

3 (6)

<0.0001

DDD

2.85 ± 2.24

0.20 ± 0.51

<0.0001

Insulin (mix) (n, %)

8 (15)

3 (6)

0.201

DDD

1.70 ± 1.33

0.44 ± 0.61

0.004

Insulin (long-acting) (n, %)

12 (23)

1 (2)

0.002

DDD

1.75 ± 1.19

0.046 ± 1.16

<0.0001

Insulin (fast-acting) (n, %)

11 (21)

0 (0)

0.001

DDD

2.21 ± 1.28

0.00

<0.0001

Fasting glucose (mmol/l)

9.3 ± 3.7

7.2 ± 2.8

<0.0001

HbA1c (mmol/mol)

64.8 ± 19.7

46.4 ± 12.9

<0.0001

Fasting C-peptide (nmol/l)

1.89 ± 1.38

0.92 ± 0.39

0.001

Data were mean (±SD), unless otherwise specified

T2DM type 2 diabetes mellitus, DDD daily defined dose

Baseline characteristics of patients reaching remission of T2DM were compared with those who failed to reach remission (Table 3). Patients with remission of T2DM were younger (44.1 vs. 51.3 years, p = 0.006), had a shorter duration of T2DM (2.4 vs. 10.1 years, p < 0.0001), less users of insulin (22 % vs. 60 %, p = 0.006) and had lower glucose (7.7 vs. 11.1 mmol/l, p = 0.001) and HbA1c levels (51.7 vs. 79.5 mmol/mol, p < 0.0001) compared with patients who had no remission. Logistic regression analysis showed that pre-operative HbA1c level (odd ratio 0.911, p = 0.020) and duration of T2DM (odds ratio 0.637, p = 0.010) were independent risk factors for the persistence of T2DM after LRYGB. Age (p = 0.937), insulin usage (p = 0.905) and glucose level (p = 0.632) were not identified as risk factors (Table 4).
Table 3

Difference between remission and no remission (long-term follow-up)

 

Remission (n = 27)

No remission (n = 25)

P value

Demographic

   

 Age (years)

44.1 ± 9.3

51.3 ± 8.7

0.004

 Female (n, %)

17 (63)

14(56)

0.609

 Long Roux limb (n, %)

8 (30)

4(16)

0.329

Weight

   

 Weight pre-operative (kg)

144.9 ± 20.9

136.0 ± 22.1

0.141

 Weight loss (kg)

−39.4 ± 17.1

−34.3 ± 13.6

0.238

 BMI (kg/m2)

47.5 ± 6.6

45.5 ± 6.1

0.274

 BMI loss (kg/m2)

−13.0 ± 5.6

−11.4 ± 4.3

0.259

 EW (kg)

61.1 ± 17.9

68.3 ± 19.4

0.166

 %EWL

60.7 ± 28.1

58.2 ± 20.2

0.712

Glucose metabolism

   

 Diabetes duration (years)

2.4 ± 2.8

10.1 ± 4.7

<0.0001

 Medication use (n, %)

20(74)

24(96)

0.051

 Insulin treatment (n, %)

6(22)

15(60)

0.006

 Glucose preoperative (mmol/l)

7.7 ± 2.4

11.1 ± 4.2

0.003

 HbA1c preoperative (mmol/mol)

51.7 ± 12.6

79.5 ± 15.6

<0.0001

 C-peptide preoperative (nmol/l)

2.24 ± 1.7

1.57 ± 0.9

0.207

Data were mean (±SD), unless otherwise specified

BMI body mass index, EW excess weight, %EWL percentage excess weight loss

Table 4

Multiple regression analysis for remission of T2DM

Individual variable

P value

Odds ratio (95 % CI)

Age (years)

0.937

1.005 (0.879–1.150)

Pre-operative FPG

0.632

0.909 (0.616–1.343)

Pre-operative HbA1c

0.020

0.911 (0.841–0.985)

Duration of T2DM

0.010

0.637 (0.452–0.898)

Insulin usage

0.905

0.851 (0.061–11.958)

T2DM type 2 diabetes mellitus, FPG fasting plasma glucose

Ten out of 25 patients (40 %) who had no remission at their last follow-up had temporarily met the criteria for remission during the preceding period (within 2 years after surgery), but later experienced a relapse of T2DM. Compared with patients in sustained remission, these patients with relapse had a longer duration of T2DM (8.9 ± 4.7 vs. 2.4 ± 2.8 years; p < 0.0001), higher pre-operative FPG levels (10.9 ± 3.8 vs. 7.7 ± 2.4 mmol/l; p = 0.034) and higher HbA1c levels (81 ± 16.6 vs. 51.7 ± 12.6 mmol/mol; p < 0.0001). Table 5 demonstrates the parameters for patients who had relapse, no remission and remission of T2DM during follow-up.
Table 5

Comparison of patients with no remission, relapse and sustained remission of T2DM during long-term follow-up

 

No remission (n = 15)

Relapse (n = 10)

Sustained remission (n = 27)

pa value

pb value

Demographic

     

 Age (years)

51.9 ± 8.5

50.3 ± 9.5

44.1 ± 9.3

0.723

0.108

 Female (n, %)

10(67)

4(40)

17(63)

0.241

0.274

 Long Roux limb (n, %)

3(20)

1(10)

8(30)

0.626

0.393

Weight

     

 Weight preoperative (kg)

130.8 ± 13.6

143.8 ± 30.0

144.9 ± 20.9

0.233

0.901

 BMI preoperative (kg/m2)

45.1 ± 5.4

46.3 ± 7.2

47.5 ± 6.6

0.641

0.623

 EW (kg)

57.6 ± 12.2

66.3 ± 23.9

68.3 ± 19.4

0.309

0.789

 %EWL

58.6 ± 21.7

57.6 ± 18.8

60.7 ± 28.1

0.911

0.749

Glucose metabolism

     

 Diabetes duration (years)

10.9 ± 4.7

8.9 ± 4.7

2.4 ± 2.8

0.297

<0.0001

 Medication use (n, %)

14(93)

10(100)

20(74)

1.000

0.155

 Insulin treatment (n, %)

10(67)

5(50)

6(22)

0.442

0.101

 Pre-op glucose (mmol/l)

11.2 ± 4.5

10.9 ± 3.8

7.7 ± 2.4

0.861

0.034

 Pre-op HbA1c (mmol/mol)

78.5 ± 15.5

81.0 ± 16.6

51.7 ± 12.6

0.688

<0.0001

 Pre-op C-peptide (nmol/l)

1.12 ± 0.41

2.16 ± 1.0

2.2 ± 1.7

0.017

0.913

Data were mean (±SD), unless otherwise specified

BMI body mass index, EW excess weight, %EWL percentage excess weight loss, IDDM insulin-dependent diabetes mellitus

aRelapse versus No remission

bRelapse versus Remission

Hypertension and Dyslipidaemia

The number of patients with HT was significantly reduced from 38 (73 %) to 28 (54 %) (p = 0.042). The total number of medication users was non-significantly reduced from 34 (65 %) to 25 (48 %) patients (p = 0.075). Nearly all types of anti-hypertensive medication were reduced in both the number of users and dosages compared with baseline; however, statistical significance for single medication types was not reached (Table 6).
Table 6

Prevalence, medication use and laboratory tests of hypertension and dyslipidaemia

 

Before surgery (n = 52)

After surgery (n = 52)

P value

Hypertension (n, %)

38(73)

28(54)

0.042

Medication use (n, %)

34(65)

25(48)

0.075

Ace–inhibitor (n, %)

16(31)

11(21)

0.263

DDD

1.31 ± 1.09

0.71 ± 0.85

0.077

Angiotensin-II–inhibitor (n, %)

16(31)

9(17)

0.108

DDD

1.33 ± 0.80

0.74 ± 0.87

0.064

Dihydropyridin derivate (n, %)

9(17)

7(13)

0.587

DDD

1.63 ± 1.06

1.13 ± 1.20

0.273

β–Blocker (n, %)

9(17)

8(15)

0.791

DDD

0.60 ± 0.56

0.39 ± 0.41

0.329

Sulphonamides (n, %)

6(12)

3(6)

0.488

DDD

1.67 ± 0.41

0.67 ± 0.82

0.076

Potassium-sparing agent (n, %)

2(4)

2(4)

1.000

DDD

0.58 ± 0.69

0.25 ± 0.32

0.546

Thiazide (n, %)

10(27)

7(13)

0.426

DDD

0.71 ± 0.69

0.33 ± 0.39

0.222

Anti-adrenergic agent (n, %)

1(2)

1(2)

1.000

DDD

1.00 ± 1.41

0.50 ± 0.71

0.795

Calcium antagonist (n, %)

0(0)

1(2)

1.000

DDD

0.00

0.83

Dyslipidaemia (n, %)

37(71)

28(54)

0.068

Medication use (n, %)

18(35)

16(31)

0.676

HMG CoA reductase inhibitor (n, %)

16(31)

16(31)

1.000

DDD

1.17 ± 1.01

0.78 ± 0.57

0.147

Fibrate (n, %)

2(4)

0(0)

0.495

DDD

1.0 ± 0.00

0.00

Total cholesterol (mmol/l)

4.78 ± 0.92

4.35 ± 0.81

0.028

HDL-cholesterol (mmol/l)

1.17 ± 0.31

1.42 ± 0.40

<0.0001

LDL-cholesterol (mmol/l)

2.47 ± 0.83

2.31 ± 0.55

0.307

Triglycerides (mmol/l)

2.79 ± 2.22

1.42 ± 0.82

<0.0001

Data were mean (±SD), unless otherwise specified

DDD daily defined dose

Overall lipid profile improved after surgery compared with baseline. Mean serum TC and TG decreased, while HDL-C increased significantly compared with baseline. Number of patients with DL decreased non-significantly from 37 (71.2 %) to 28 (53.8 %) (p = 0.068). Medication use was slightly decreased after surgery, from 18 (35 %) patients to 16 (31 %) (p = 0.676).

Discussion

In the present study, the long-term metabolic effects of LRYGB were analysed in severely obese patients with T2DM. After a mean post-operative follow-up period of 82 months, there was a reduction in T2DM (52 %), HT (26 %) and DL (24 %). Furthermore, there was a marked reduction of medication use for T2DM and HT.

Although the LRYGB is a proven effective treatment for T2DM and other cardiovascular risk factors in the short-term (<2 years), in patients with obesity (BMI < 35 kg/m2) [14, 23] and severe obesity (BMI > 35 kg/m2) [13, 14, 24, 25], with remission rates of T2DM of 64–81 % after RYGB in the mid-long-term (3–5 years) [2629], some professionals argue the durability of these positive effects in the long-term and advocate for better predictors for success [30]. In a long-term follow-up of 10 years after primary LRYGB, Higa et al. [31] showed that T2DM had resolved or improved in 67 % of patients; however, this study was based on questionnaires which are prone to misdiagnosis. In contrast, a large non-randomised, prospective Swedish Obese Subjects (SOS) study showed a low recovery rate of T2DM of only 36 % 10 years after bariatric surgery. It must be mentioned that only 5 % of patients in the SOS had undergone an RYGB procedure [32]. In our experience, 52 % of the patients had sustained remission of T2DM after a mean period of approximately 7 years. This is slightly lower compared with other studies within the same timeline [33, 34]. Nevertheless, the published remission rates in the long-term are still lower compared with data on the short- and mid-long-term, probably due to the recurrence of T2DM over time. In this study, 71 % of the patients reached remission of T2DM during follow-up; however, 19 % had a relapse of their diabetes over time.

Predictors for remission and/or recurrence of T2DM after bariatric surgery, i.e., RYGB, can help to estimate the success rate in the long-term. Furthermore, it could be helpful in the selection of pre-operative patients and could provide more realistic expectations on the chance of diabetes resolution for the patients. In the present study, pre-operative HbA1c level (OR 0.911) and duration of T2DM prior to surgery (OR 0.637) were independent predictors for remission of T2DM. Previous studies postulated different independent predictors for glycaemic control after RYGB, including age [29, 35], female gender [34], BMI [36, 37], C-peptide level [3739], oral medication and/ or insulin use [29, 34], duration of T2DM prior to surgery [37, 40] and post-operative weight loss and/or weight regain [29, 35, 36].

Previous studies, including one of our own study group, demonstrated that pre-operative fasting C-peptide, a marker for beta-cell function of the pancreas, lower than 1.0 nmol/l is associated with a reduced chance of complete remission of T2DM after RYGB surgery [3739]. In the present study, a non-significant difference was found for pre-operative C-peptide levels between patients who had remission (2.24 ± 1.7 nmol/l) and no-remission (1.57 ± 0.9 nmol/l) of their T2DM at the last follow-up visit. Relapse of T2DM has not been described in earlier studies with a short-term follow-up [3739]. When eliminating patients who had a relapse of their T2DM, the C-peptide level was significantly lower for patients who did not have remission (1.12 ± 0.4 nmol/l) during the follow-up compared with patients who remained in remission (2.2 ± 1.7 nmol/l; p = 0.030). Patients who had a relapse of T2DM had similar pre-operative C-peptide levels (p = 0.913) compared with patients who remained in remission. These findings implicate that a low pre-operative C-peptide level is associated with no remission of T2DM in the short- and long-term; however, high C-peptide levels do not predict the chance of sustained remission of T2DM in the long-term.

Low C-peptide level, use of insulin and longer duration of T2DM prior to operation are determinants of severity of disease. It can be stated that higher severity of T2DM prior to surgery results in less remission of T2DM after RYGB. Due to the progressive nature of T2DM, long-term remission after surgery is hard to achieve in many patients; however, almost all patients profit from surgery in the long-term.

Several studies have also demonstrated a long-term reduction of other obesity related co-morbidities, such as HT (42–73 %) [31, 33, 4143] and DL (62–80 %) [41, 42] after RYGB. In the present study, there was a significant reduction in the prevalence of hypertension and a reduction of anti-hypertensive medication use after surgery. Furthermore, lipid profile significantly improved and prevalence of DL decreased. However, the reduction of HT (26 %) and DL (24 %) were lower compared with these previous studies.

Before interpreting the findings of the present study, the limitations of this study should be considered. Eighty-five percent of the eligible population had completed the long-term follow-up visits. This fact could result in over- or underestimation of the metabolic effects, since patients with good results following their operation could deem a control visit unnecessary, while patients with disappointing results may reject visits due to a loss of motivation. Therefore, we had a relatively small study population; however, all patients were seen in the outpatient clinics, which provided us with reliable data. In contrast, some other published long-term studies were based on telephone contact. Additionally, this study corroborates the data which is available from long-term cohort studies, such as the SOS study, with biochemical data.

Conclusion

The results of this study showed that LRYGB, in addition to the short- and mid-term follow-up data from the literature, affords long-term remission of T2DM as well as a decrease in HT and DL. All patients seem to benefit in terms of better glucose regulation and lower medication use. Individual predictors for the remission of T2DM were lower pre-operative HbA1c level and shorter duration of T2DM in the medical history. The limited rate of follow-up in this study may be responsible for either an over- or underestimation of this long-term effect and needs to be researched further.

Disclosure

There are no conflicts of interest, and no funding was provided.

Copyright information

© Springer Science+Business Media New York 2014