Journal Pre-proof

Factors Mediating Type 2 Diabetes Remission and Relapse after Gastric Bypass
Surgery

Bernardo M. Pessoa, MD, Matthew G. Browning, PhD, Guilherme S. Mazzini, MD

PhD, Luke Wolfe, MS, Amy Kaplan, Jad Khoraki, MD, Guilherme M. Campos, MD,
PhD, FASMBS, FACS
PII: S1072-7515(19)32115-5
DOI: https://doi.org/10.1016/j.jamcollsurg.2019.09.012
Reference: ACS 9649

To appear in: Journal of the American College of Surgeons

Received Date: 30 June 2019

Revised Date: 19 August 2019
Accepted Date: 16 September 2019

Please cite this article as: Pessoa BM, Browning MG, Mazzini GS, Wolfe L, Kaplan A, Khoraki J,
Campos GM, Factors Mediating Type 2 Diabetes Remission and Relapse after Gastric Bypass
Surgery, Journal of the American College of Surgeons (2019), doi: https://doi.org/10.1016/
j.jamcollsurg.2019.09.012.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Inc. on behalf of the American College of Surgeons.

Factors Mediating Type 2 Diabetes Remission and Relapse after Gastric Bypass Surgery
Bernardo M Pessoa, MDa, Matthew G Browning, PhDa, Guilherme S Mazzini, MD PhDa, Luke
Wolfe, MSa, Amy Kaplana, Jad Khoraki, MDa, Guilherme M Campos, MD, PhD, FASMBS,
FACSa
a
Virginia Commonwealth University, Department of Surgery, Division of Bariatric and
Gastrointestinal Surgery; Richmond, VA
[CE note: no degree for Kaplan]

Disclosure Information: Nothing to disclose.

Presented at the American College of Surgeons 105th Annual Clinical Congress, Scientific
Forum, San Francisco, CA, October 2019.

Correspondence Author/Address:
Guilherme M Campos, MD, PhD, FACS, FASMBS
Paul J Nutter Professor of Surgery
Chairman, Division of Bariatric and Gastrointestinal Surgery
Director, Bariatric Surgery Program. Director, Advanced GI and MIS Fellowship
Virginia Commonwealth University, Department of Surgery, Medical College of Virginia
1200 East Broad Street
PO Box 980519
Richmond, Virginia 23298
Phone: (804) 628-3626
guilherme.campos@vcuhealth.org

Brief Title: Diabetes Remission after Gastric Bypass

1
Abstract

Background: Defining factors associated with remission and relapse of type 2 diabetes (T2D)

after gastric bypass (RYGB) may allow targeting modifiable factors. We investigated factors

associated with T2D remission and relapse after RYGB.

Study Design: Retrospective review of consecutive patients with T2D who underwent RYGB

between 1993 and 2017. T2D remission was defined as medications discontinuation and/or

HbA1c<6.5%; relapse was defined as recurrent medication use and/or HbA1c≥6.5%.

Independent correlates of T2D remission and relapse were identified using logistic regression.

Results: 621 patients (aged 46.7±10.6 years; 30% on insulin; BMI 49.8±8.3 kg/m2) had at least

one-year follow-up. Median follow-up was 4.9 (range 1-23.6) years. Prevalence of T2D

remission was 74% at one year, 73% from one to three years, 63% between 3 and 10 years, and

47% beyond 10 years. 93% of remissions occurred within 3 years of RYGB, 25% relapsed. The

median (IQR) time to relapse was 5.3 (3-7.8) years after remission. Higher one-year percentage

total body weight (%TBW) loss, lack of preoperative insulin use, and younger age at surgery

were independently associated with T2D remission. Preoperative insulin use, lower %TBW loss

at one year, and greater %TBW gain after one year were independently associated with T2D

relapse.

Conclusion: This longitudinal retrospective analysis shows that preoperative insulin use and age,

one-year weight loss, and regain after that influence T2D remission and relapse following

RYGB. Referring patients at a younger age, before insulin is needed, and optimizing both weight

loss and weight maintenance after RYGB may improve the rates and durability of T2D

remission.

2
Abbreviations
BMI: Body Mass Index
BPD: Biliopancreatic Diversion
BW: Body Weight
EBW: Excess Body Weight
EMR: Electronic Medical Record
HbA1c: Hemoglobin A1C
IQR: Interquartile range
RYGB: Roux-en-Y Gastric Bypass
SD: Standard Deviation
SG: Sleeve Gastrectomy
T2D: Type 2 diabetes
TBW: Total Body Weight

3
 1. INTRODUCTION

The prevalence of Class II and III obesity (severe obesity), defined as a body mass index

(BMI) 35 to 39.9 kg/m2 and ≥40 kg/m2, respectively, has tripled over the last three decades in the

US. (1) One of three adults with severe obesity is diagnosed with type 2 diabetes (T2D), three-

fold higher than the national average. (2, 3) Of importance, 33% of subjects with T2D and severe

obesity are placed on insulin compared to less than 25% of those with Class I obesity (BMI 30-

34.9 kg/m2), and the likelihood of attaining the target hemoglobin A1c (HbA1c) value under 7%

progressively decreases across increasing degrees of obesity. (4, 5) It follows that interventions

centered on caloric restriction and increased physical activity, with or without anti-obesity drug

therapy, are cornerstones of T2D care in patients with obesity, and that 10% reductions in body

weight can significantly reduce HbA1c in these individuals. (6, 7) However, four-year follow-up

outcomes from the Look AHEAD trial suggest that over 80% of this patient population is unable

to maintain this magnitude of weight loss over time, (7) and other evidence suggests that even

those who are successful may still require intensification of the anti-diabetic drug regimen to

prevent worsening hyperglycemia. (8)

On the other hand, current bariatric surgery techniques, such as Roux-en-Y gastric bypass

(RYGB), are proven to yield substantial and durable reductions in body weight, HbA1c, and

anti-diabetic medications requirements in the majority of patients. Randomized controlled trials

have shown that over 50% of RYGB patients enter T2D remission compared to less than 5% of

those allocated to non-surgical treatment arms, including five-year remission rates that were six-

fold higher with RYGB versus conventional therapies. (9, 10) Nevertheless, a variable fraction

(15-40%) of patients do not achieve postoperative T2D remission, and about one-third of patients

who achieve remission have been shown to experience a relapse of T2D. (11-14) A better

4
understanding of factors that contribute to T2D remission and relapse with RYGB can allow for

better patient counseling about postoperative expectations with T2D outcomes, assist referring

providers in identifying patients with characteristics associated with greater chances for durable

remission and target modifiable variables to improve remission and reduce relapse rates of T2D

after RYGB.

Various combinations of preoperative patient characteristics and T2D duration and

severity have been proposed for estimating an individual’s likelihood of achieving T2D

remission at one year of follow-up. Importantly, in addition to older age and insulin use at the

time of surgery, a greater percentage of total body weight (%TBW) lost at one year after RYGB

was recently reported as an additional independent predictor of five-year T2D remission (13).

Given that T2D remission and relapse rates appear to plateau three years and nine years after

RYGB, respectively, (14, 15) shorter follow-up periods may be insufficient to capture a

considerable proportion of patients at risk for relapse of T2D. Thus, questions remain regarding

the potential contributory roles of individual patient preoperative demographic and clinical

variables, one-year weight loss and, importantly, weight regain to post-RYGB remission and

relapse rates over time (i.e., after the three-year plateau in remission rates and during the

subsequent period where relapse has a greater chance of occurring).

We addressed these knowledge gaps in the current study by evaluating associations in

between T2D remission and relapse with preoperative patient characteristics and weight loss

outcomes at one year, between one and three years, three and ten years, and from ten to twenty-

four years after RYGB.

2. METHODS

2.1. Study Design and Participants

5
 This is a retrospective observational study of patients with T2D who underwent primary

RYGB surgery at one quaternary academic medical center in the United States between January

1993 and December 2017. Patients were identified using our institution’s prospectively

maintained bariatric surgery database. Two independent researchers individually reviewed each

patient’s electronic medical record (EMR), including all bariatric and non-bariatric providers’

clinical notes and charts that had been scanned into the EMR system. Patients who had not

continued long-term follow-up were contacted by e-mail or telephone and asked to return for

routine clinical evaluation and laboratory testing. Patient-reported data were not included in our

analysis. All patients had a preoperative BMI ≥35 kg/m2 and met bariatric surgery eligibility

criteria set forth by our center and the National Institutes of Health. (16) Patients who underwent

RYGB at another center or had reoperative bariatric surgery (conversion, revisional or reversal)

were excluded from the analysis.

Length of follow-up was defined as the time (in years) between RYGB and the last follow-up

visit recorded. The percentage of completeness of follow-up after surgery within each period was

calculated. Loss of follow-up was determined when a patient had no current contact information,

was not successfully contacted, was unwilling to return to our center for a clinic visit or was

known to have become pregnant or died. We censored patients post-operatively at the time of

pregnancy to mitigate confounding, based on the fact that pregnancy exposes patients to changes

in body weight and glycemic control that cannot be attributable to RYGB alone and certainly are

not relevant to male participants or women of non-child-bearing age. This is consistent with

another highly-referenced paper on this topic. (14) All patients that had loss of follow-up were

searched by social security number to identify deceased patients. To do so, we used the Social

Security Administration death master file, a public online source

6
(http://ssdi.rootsweb.ancestry.com), considered reliable for providing survival information in

previous studies. (17, 18)

2.2. Surgical technique

RYGB technique evolved during the years based on the surgeon's preferences and the

introduction of the newer generation of commercially available surgical products and

laparoscopic techniques. Operations were either open or laparoscopic RYGB. In general, a

biliopancreatic limb of about 50 cm and an alimentary limb of about 50 to 150 cm were

measured, and a linear stapler was used to create a side-to-side jejunojejunostomy (JJ) that was

completed either hand-sewn or with another linear stapler firing. The mesenteric defect at the JJ

was routinely closed, and the Petersen’s defect was closed in the majority of cases, both with

nonabsorbable sutures. A lesser curvature-based gastric pouch measuring approximately 5 cm

was created using a linear stapler. The alimentary limb was approximated to the gastric pouch

either in an antecolic or retrocolic fashion, and a gastrojejunostomy (GJ) was created using either

a circular or linear stapler.

2.3. Preoperative evaluation and Perioperative care

All patients had comprehensive medical evaluations and underwent individual preoperative

counseling by a multidisciplinary team that included dieticians, nurses, psychologists, and

bariatric surgeons. They participated in preoperative educational classes with a minimum of four

to ten hours of educational sessions with the multidisciplinary team.

After surgery, patients were discharged at the discretion of the attending surgeons and seen at

two weeks; then at 3, 6, 12 months and annually after that. More frequent clinic visits were

arranged if needed. Routine nutritional supplements after surgery included multi-vitamins,

7
calcium supplements, cholecalciferol (vitamin D), as well as ursodeoxycholic acid and acid

suppression medication (for three months). Additional supplements were prescribed for

documented deficiencies. Weight was recorded at each of the follow-up visits.

2.4. Data and Study outcomes

Primary outcomes of interest were the prevalence of and factors associated with T2D

remission and relapse and time to T2D remission and relapse after RYGB. Time to relapse was

calculated in both years since remission and years since surgery. All patients had a pre-existing

physician diagnosis of T2D at the time of preoperative evaluation.

T2D remission was defined in patients with an HbA1c <6.5% for one year of medication‐free

follow-up. (19, 20) Patients were also deemed to be in remission if not taking anti-diabetic

medications, and no HbA1c laboratory testing was performed within a given year. (21)

Relapse of T2D was defined as HbA1c ≥6.5% or need for antidiabetic medication after any

period of remission. (19, 20) Sustained T2D remission was defined in patients who were in

remission across at least two-time points without having any documented relapse of T2D.

Baseline data were collected at the last visit before surgery and included age, sex, race

(patient-reported through a standard intake questionnaire with separate questions for race and

ethnicity), body weight to the nearest 0.1 kg, and the presence of obesity-related co-morbidities,

including T2D, hypertension, gastroesophageal reflux disease (GERD), obstructive sleep apnea

(OSA), and dyslipidemia. Comorbidities were defined using available notes from the bariatric

surgeons and referring physicians, the preoperative anesthesiology evaluations, in addition to

reviewing patient's medications.

BMI was calculated before surgery and at each follow-up visit by dividing the weight (in

kilograms) by the square of height in meters (m2). We calculated preoperative excess body

8
weight (EBW) as the difference between initial weight and weight equivalent to a BMI of 25

kg/m2. (20) Percentage of total body weight (%TBW) loss was calculated as (Postoperative

Weight - Initial Weight) / Initial Weight x 100. We also evaluated weight loss maintenance or

regain by calculating the %TBW change from one year using the formula: (Postoperative Weight

– Weight at One Year) / Weight at One Year x 100.

Medication/insulin use and remission of T2D were recorded throughout the follow-up period.

We evaluated sustained remission and relapse of T2D in patients who achieved initial T2D

remission and had at least one additional follow-up visit after that. Co-morbidity status and

laboratory values were collected at annual visits to our bariatric clinic or were available in the

EMR. All follow-up data available in January 2019 were included.

Postoperative outcomes data are presented within the following periods: at one year after

surgery, between one and three years, from three to ten years, and from ten to twenty-four years.

This stratification captures patient outcomes across discrete follow-up periods, accounting for the

lack of standardization of postoperative visits and present outcomes for patients who did not

have complete follow-up visits by January 2019. Importantly, this stratification system also

aligns with the plateau in T2D remission (within the first three years) and relapse (within the first

nine years) that have been reported by several previous large cohort studies that studied T2D

remission and relapse. (11, 14).

2.5. Statistical Analysis

Normally distributed continuous variables are presented as mean ± standard deviation (SD)

and were compared using independent samples T-tests or analysis of variance (ANOVA).

Variables deviating from the normal distribution or violating homogeneity of variance tests are

9
presented as median and interquartile range (IQR) and were compared between groups using

either Mann-Whitney U Tests or Kruskal Wallis Tests.

Separate logistic regression models were constructed to examine associations between either

T2D remission (versus non-remission) or relapse (versus sustained remission) and preoperative

demographics (age, sex, BMI, EBW), T2D medications, and comorbidities including

hypertension, OSA, dyslipidemia, and GERD, as well as postoperative %TBW changes. In the

multivariate analysis of factors distinguishing patients with remission from non-remission, one-

year %TBW loss was entered to account for differences in length of follow-up between groups.

Subsequent %TBW gain from one year was further considered as a potential predictor of T2D

relapse. Independent variables with a P-value <0.10 and variance inflation factor >5 were

retained in the final models. All analyses were performed using SPSS Software Version 25

(SPSS Inc., Chicago, IL, USA).

3. RESULTS

A total of 706 patients with T2D underwent RYGB during these 24 years, of whom 621

(88%) had available follow-up data of at least one year after surgery and comprised our final

study sample. Baseline demographics, anthropometrics, and clinical characteristics of all patients

and the 621 patients included in the subsequent analyses are summarized in Table 1.

The median length of follow-up of the 621 patients was 4.9 years (IQR: 1.9 to 8.7 years,

range: 1 to 23.6 years). Complete follow-up data were available in 85% of patients at one year

(n=530/621), 95% from one to three years (n=588/621), 64% from three to ten years

(n=380/591), and 23% after 10 years (n=113/484). Forty-five of the 113 patients with greater

than 10 years of follow-up information available were more than 15 years from RYGB. Most

patients were female (81%) and of the white race (72%). The mean±SD age at surgery was

10
46.7±10.6 years, preoperative BMI was 49.8±8.3 kg/m2, and 30% of the patients were on insulin.

At the time of surgery, 75% of patients had hypertension, 55% had GERD, 48% had

dyslipidemia, and 44% had OSA.

Relative to patients with at least one year of follow-up, a higher proportion of those without

one year of follow-up were male (35% versus 19%, P=0.001) and died during follow-up (19%

versus 11%, P=0.035), while a lower proportion had hypertension at the time of surgery (60%

versus 75%, P=0.004). There was no difference in insulin use between patients who did or did

not have one-year follow-up data.

3.1. Remission of T2D

Overall, 478 (77%) patients achieved T2D remission after RYGB. The median (IQR) time to

first recorded remission was one (1-1.2) year after surgery. Prevalence of T2D remission was

74% after one year, 73% from one to three years, 63% between three and ten years, and 47%

beyond ten years (Figure 1). Of those who achieved T2D remission, 85% did so within the first

year of surgery, and 93% did so within the first three years of surgery.

Demographic, clinical, and one-year weight loss measures are compared between patients

who did and did not achieve T2D remission in Table 2. The median follow-up period after

surgery was greater in those who achieved remission than those who did not (5.2 vs. 3.7,

P=0.013). Additionally, patients who remitted were younger (median age 46.4 vs. 50.6 years,

P<0.001), less likely to be on insulin (22% vs. 59%, P<0.001) or have hypertension (73% vs.

83%, P=0.019) at the time of surgery, while neither preoperative body weight nor BMI differed

between patients who did and did not achieve T2D remission.

Alternatively, T2D remission relative to non-remission was associated with significantly

greater BMI reduction and %TBW loss one year after RYGB. We considered the possibility that

11
patients who required insulin at the time of surgery may have had poorer weight loss outcomes

than those with insulin-independent T2D but found no significant differences in pre-operative

weight, EBW, BMI, or their post-operative changes at any time point (P>0.05 for all).

3.1.1. Multivariate Analysis – Remission versus non-Remission

Factors independently associated with remission of T2D after RYGB were younger age at

surgery, lack of preoperative insulin use, and greater percentage total body weight loss from

surgery to one-year follow-up (Table 3).

3.2. Sustained Remission and Relapse

Of the 478 patients who achieved T2D remission, the median (IQR) duration of remission

was 4.9 (2.1-7.8) years and ranged from 1 to 22.4 years.

Three hundred and fifty-six of those that remitted (74%), had at least one subsequent follow-

up visit, with a median (IQR) length of follow-up of 6.1 (3.3-9.2) years, and were further

evaluated for the sustainability of remission or relapse. Of these 356, 267 patients (75%) had

sustained remission, and 89 (25%) relapsed.

The median (IQR) duration of sustained remission was 6.0 (3.1-9.1) years and ranged from 1

to 22.4 years. The median (IQR) time to relapse was 5.3 (3-7.8) years after remission and 6.5 (4-

9.2) years after surgery. The proportion of patients who experienced a relapse of T2D

progressively increased from 5% of initial remitters within the first three years to 19% between

three and ten years and 34% after ten years (Figure 1). Of the 45 patients with at least 15 years of

follow-up (median 16.2 years), 23 (51%) were still in remission, while 15 (33%) had relapsed.

There were no differences in baseline comorbidities prevalence, time from surgery to

remission, and length of follow-up in between groups with sustained remission or relapse (Table

12
4). Relative to patients with sustained remission, those who relapsed were older, had a higher

BMI, and had more EBW at the time of surgery (P<0.05 for all). The proportion of patients with

sustained remission on insulin at baseline was less than half that of patients who relapsed (16%

versus 34%, P<0.001).

Sustained remission was also associated with significantly larger BMI reduction and greater

%TBW loss at the one-year follow-up and between three to ten years after RYGB when

compared to patients who relapsed. From three to ten years postoperatively, patients who

achieved sustained remission regained approximately 5% of the body weight lost over the first

postoperative year, while those who relapsed regained nearly four times this amount (P<0.05 for

all).

There were no differences in BMI or weight loss measures across groups with sustained

remission or relapse after ten years. Notably, when restricting the analyses to only the 45 patients

with 15 or more years of follow-up, those with sustained remission demonstrated significantly

greater %TBW loss after one year (5.7±10.9%, P=0.032) but not beyond ten years (4.0±15.3%,

P=0.271).

3.3. Multivariate Analysis – Sustained Remission, non-Remission, and Relapse

Factors independently associated with relapse (when compared to those with sustained

remission) were the need for insulin at baseline and a lesser %TBW loss from surgery to one

year, as well as larger %TBW gain between three and ten years after RYGB (Table 3).

4. DISCUSSION

To our knowledge, the current study details T2D remission and relapse prevalence after

RYGB across the longest follow-up period reported to date. The few prior groups that

13
investigated correlates of long-term T2D remission and relapse in large populations of RYGB

patients (n >100) did so in patients with up to 13 years of follow-up. (11, 14) We extended this

postoperative timeline by over a decade and found similar declines in remission prevalence over

time (i.e., 73% over the first 3 years, 63% between three and ten years, and 47% from 10 to 24

years). Furthermore, our sample population included 45 patients with data available over 15

years of follow-up, of whom 51% were still in remission, 33% had relapsed, and 16% never

remitted. The only other study of which we are aware of reporting T2D outcomes beyond 15

years included only 12 RYGB patients (4 still in remission). (22) Importantly, the present results

point to a strong interdependence of weight loss, weight regain, and T2D outcomes after RYGB.

We found that a higher %TBW by the end of the first post-operative year was independently

associated with remission and, together with an approximately four-fold lesser weight regain

after one year, were protective against relapse.

In our study, 85% of remission occurred by the one-year follow-up, similar to previously

reported one-year remission rates in the range of 50-80%. (13, 23-25) Furthermore, 93% of

recorded T2D remissions were within the first three years after RYGB, which is consistent with

prior evidence that less than 4% of patients may be expected to remit beyond the third post-

operative year. (15) On the other hand, 25% of our patients who achieved remission experienced

a subsequent relapse of T2D at a median time of 5.3 years after remission and 6.5 years after

surgery, which is within the range of relapse prevalence (17-40%) reported in the few prior

studies that evaluated T2D outcomes beyond five years after RYGB. (11, 12, 14, 26). Thus, our

observed time courses for remission and relapse are in overall agreement with those previously

described up to 13 years after RYGB (11, 14) and, altogether, underscore the need for follow-up

beyond five years when evaluating remission durability and factors associated with relapse.

14
 Of note, along with RYGB, data coming from two randomized controlled trials to support

that laparoscopic sleeve gastrectomy (SG) and biliopancreatic diversion (BPD) can also provide

for adequate shorter-term (5 years or less) T2D control. (27, 28) Schauer et al. (27) randomized

patients with a BMI ranging from 27 to 43 kg/m2 to either intensive medical and behavioral

therapy alone or with RYGB or SG and found that the TBW loss after both RYGB (-23 kg) and

SG (-19 kg) was significantly larger than after non-surgical therapy (-5 kg). This coincided with

five-year T2D remission prevalence of 22% after RYGB and 15% after SG (P=0.34 between

surgeries), whereas no patients receiving conventional therapy were in remission after five years

(P<0.01 versus RYGB and SG). (27) In another RCT, BPD resulted in greater five-year

reductions in BMI, fasting glucose, and HbA1c than RYGB. However, BPD utilization in the US

and worldwide is limited due to the higher risk of macro and micronutrient malnutrition in the

long term in many series. (29, 30) Thus, although several prior groups have reported marked

improvements in T2D within weeks or even days after RYGB surgery and often before massive

weight loss occurs, (31, 32) results from these RCTs and of cohort studies such as ours

collectively support the notion that greater weight loss is associated with a higher likelihood of

T2D remission.

In addition to the association between greater one-year weight loss and improved remission

rates, we further found that patients who achieved remission were less likely to relapse if they

avoided significant weight regain (Table 3). Therefore, strategies to optimize early postoperative

weight loss and mitigate weight regain are needed to enhance the durability of T2D remission.

Previous groups have suggested the control of problematic eating behavior as one intervention

for increasing early postoperative weight loss (33), and others have acknowledged poor

compliance with physical activity recommendations both before and after bariatric surgery as an

15
additional contributing factor to suboptimal weight loss outcomes. (34, 35) Still, weight

recidivism after RYGB is multifactorial and can also be influenced by operation-specific factors

including gastric pouch size, stoma size, alimentary limb length, gastro-gastric fistulas and other.

(36-39)

Along with greater %TBW loss at the one-year follow-up, we also found that patients with

younger age and without insulin use at the time of RYGB were more likely to achieve T2D

remission. These same factors have been identified as independently related to remission (versus

non-remission) at five years after RYGB. (13) We interpret these consistent results to have

important implications for patient referral practices and postoperative management. Specifically,

surgical therapy appears most successful in achieving T2D remission when offered at an earlier

timepoint in T2D progression and before pancreatic beta-cell exhaustion or destruction. While

weight loss with RYGB may help to offset the deterioration in pancreatic β-cell mass and/or

function with increasing age and T2D duration, it is a well known that the β-cells cannot

regenerate once destroyed and RYGB does not allow for recovery or promote hyperplasia of

pancreatic β-cell mass. (40, 41) Accordingly, Inge et al. (42) recently showed that five-year post-

operative T2D outcomes are superior in adolescents (aged 13-19 years at baseline) when

compared to adults who underwent bariatric surgery (71% RYGB). Despite a lack of between-

group differences in either baseline insulin use (20% in adolescents versus 22% in adults) or

%TBW loss after five years (26% in adolescents versus 29% in adults), adolescents were less

likely than adults to still be taking anti-diabetic medication (0% versus 26%, P<0.001) and more

likely to achieve durable T2D remission (86% versus 53%, P<0.05). (42) These observations are

consistent with our findings that likelihood of remitting decreases with age and, altogether, point

16
to earlier access to RYGB and other current bariatric surgeries as a possible approach to

improving T2D outcomes in appropriate candidates.

Also, the durability of enhancements to peripheral glucose uptake and insulin sensitivity after

RYGB seems to directly correlate with successful weight loss and weight loss maintenance. (43)

Similar to predictors of T2D remission versus non-remission, larger one-year weight loss, lesser

body weight gain up to ten years after RYGB, and preoperative insulin-independence, were also

independently related to sustained remission as opposed to relapse. It is plausible that weight

loss-related improvements in insulin sensitivity are the predominant drivers of long-term T2D

remission after RYGB. Still, the fact that preoperative insulin use was independently associated

with both non-remission and relapse is in line with prior propositions that β-cell mass and

function are also key determinants of postoperative glycemic control, but it is not a modifiable

factor once established. Thus, while there are still significant benefits of bariatric surgery for

patients who require insulin, as demonstrated by reductions in medication requirements, vascular

complications, and T2D-related mortality (22, 44), offering surgical therapy prior to insulin-

dependence and the irreversible β-cell destruction will allow patients to expect and enjoy greater

T2D remission and lower relapse rates with RYGB.

4.1. Limitations

Our study is limited by its retrospective nature and decreasing retention rates over time.

However, our retention rates throughout the first ten years are in line with those of past studies

on T2D remission and relapse rates after RYGB. (12, 14) We also point out that our total of 113

patients with follow-up information from 10 to 24 years after RYGB compares favorably to

sample sizes of other studies reporting 12-year (N=84) (26) or 15-year (N=12) (22) T2D

outcomes after surgery. Lastly, we did not have records of T2D diagnosis date and thus duration

17
of T2D in the majority of the current study’s patients; however, we were able to study insulin use

as another appropriate marker of T2D severity and pancreatic β-cell reserve. (41)

4.2. Conclusion

We found that RYGB provides for T2D remission in three-fourths of patients with Class II

and III obesity, and many will remain in remission for over fifteen years. However, one of four

patients who achieves remission may experience a relapse of T2D. Younger age and insulin

independence at the time of RYGB, together with greater one-year weight loss, were associated

with an increased likelihood of remission and together with weight loss maintenance were

protective against relapse. Strategies to promote referrals at a younger age and before insulin is

needed, as well as to optimize and maintain weight loss are expected to support rates and

durability of T2D remission with RYGB.

18
References

1. Gregg EW, Cadwell BL, Cheng YJ, et al. Trends in the prevalence and ratio of diagnosed

to undiagnosed diabetes according to obesity levels in the U.S. Diabetes Care. 2004

Dec;27(12):2806-12.

2. Nguyen NT, Nguyen XM, Lane J, Wang P. Relationship between obesity and diabetes in

a US adult population: findings from the National Health and Nutrition Examination Survey,

1999-2006. Obes Surg. 2011 Mar;21(3):351-5.

3. Ganz ML, Wintfeld N, Li Q, et al. The association of body mass index with the risk of

type 2 diabetes: a case-control study nested in an electronic health records system in the United

States. Diabetology & metabolic syndrome. 2014 Apr 03;6(1):50.

4. Johnston SS, Ammann EM, Kashyap SR, et al. Body mass index and insulin use as

identifiers of high-cost patients with type 2 diabetes: A retrospective analysis of electronic health

records linked to insurance claims data. Diabetes, obesity & metabolism. 2019 Feb 15.

5. Bertoni AG, Clark JM, Feeney P, et al. Suboptimal control of glycemia, blood pressure,

and LDL cholesterol in overweight adults with diabetes: the Look AHEAD Study. J Diabetes

Complications. 2008 Jan-Feb;22(1):1-9.

6. Rowe R, Cowx M, Poole C, et al. The effects of orlistat in patients with diabetes:

improvement in glycaemic control and weight loss. Current medical research and opinion. 2005

Nov;21(11):1885-90.

7. Wing RR, Espeland MA, Clark JM, et al. Association of Weight Loss Maintenance and

Weight Regain on 4-Year Changes in CVD Risk Factors: the Action for Health in Diabetes

(Look AHEAD) Clinical Trial. Diabetes Care. 2016 Aug;39(8):1345-55.

19
8. Paisey RB, Frost J, Harvey P, et al. Five year results of a prospective very low calorie

diet or conventional weight loss programme in type 2 diabetes. Journal of human nutrition and

dietetics : the official journal of the British Dietetic Association. 2002 Apr;15(2):121-7.

9. Khorgami Z, Shoar S, Saber AA, et al. Outcomes of Bariatric Surgery Versus Medical

Management for Type 2 Diabetes Mellitus: a Meta-Analysis of Randomized Controlled Trials.

Obes Surg. 2019 Mar;29(3):964-74.

10. Kodama S, Fujihara K, Horikawa C, et al. Network meta-analysis of the relative efficacy

of bariatric surgeries for diabetes remission. Obes Rev. 2018 Dec;19(12):1621-9.

11. de Oliveira VLP, Martins GP, Mottin CC, et al. Predictors of Long-Term Remission and

Relapse of Type 2 Diabetes Mellitus Following Gastric Bypass in Severely Obese Patients. Obes

Surg. 2018 Jan;28(1):195-203.

12. Brethauer SA, Aminian A, Romero-Talamas H, et al. Can diabetes be surgically cured?

Long-term metabolic effects of bariatric surgery in obese patients with type 2 diabetes mellitus.

Ann Surg. 2013 Oct;258(4):628-36; discussion 36-7.

13. Debedat J, Sokolovska N, Coupaye M, et al. Long-term Relapse of Type 2 Diabetes After

Roux-en-Y Gastric Bypass: Prediction and Clinical Relevance. Diabetes Care. 2018

Oct;41(10):2086-95.

14. Arterburn DE, Bogart A, Sherwood NE, et al. A multisite study of long-term remission

and relapse of type 2 diabetes mellitus following gastric bypass. Obes Surg. 2013 Jan;23(1):93-

102.

15. Isaman DJ, Rothberg AE, Herman WH. Reconciliation of Type 2 Diabetes Remission

Rates in Studies of Roux-en-Y Gastric Bypass. Diabetes Care. 2016 Dec;39(12):2247-53.

20
16. NIH conference. Gastrointestinal surgery for severe obesity. Consensus Development

Conference Panel. Ann Intern Med. 1991 Dec 15;115(12):956-61.

17. Wojcik NC, Huebner WW, Jorgensen G. Strategies for using the National Death Index

and the Social Security Administration for death ascertainment in large occupational cohort

mortality studies. Am J Epidemiol. 2010 Aug 15;172(4):469-77.

18. Huntington JT, Butterfield M, Fisher J, et al. The Social Security Death Index (SSDI)

most accurately reflects true survival for older oncology patients. American journal of cancer

research. 2013;3(5):518-22.

19. Buse JB, Caprio S, Cefalu WT, et al. How do we define cure of diabetes? Diabetes Care.

2009 Nov;32(11):2133-5.

20. Brethauer SA, Kim J, el Chaar M, et al. Standardized outcomes reporting in metabolic

and bariatric surgery. Surg Obes Relat Dis. 2015 May-Jun;11(3):489-506.

21. Madsen LR, Baggesen LM, Richelsen B, Thomsen RW. Effect of Roux-en-Y gastric

bypass surgery on diabetes remission and complications in individuals with type 2 diabetes: a

Danish population-based matched cohort study. Diabetologia. 2019 Apr;62(4):611-20.

22. Sjostrom L, Peltonen M, Jacobson P, et al. Association of bariatric surgery with long-

term remission of type 2 diabetes and with microvascular and macrovascular complications.

Jama. 2014 Jun 11;311(22):2297-304.

23. Kadera BE, Lum K, Grant J, et al. Remission of type 2 diabetes after Roux-en-Y gastric

bypass is associated with greater weight loss. Surg Obes Relat Dis. 2009 May-Jun;5(3):305-9.

24. Hariri K, Guevara D, Jayaram A, et al. Preoperative insulin therapy as a marker for type 2

diabetes remission in obese patients after bariatric surgery. Surg Obes Relat Dis. 2018

Mar;14(3):332-7.

21
25. Smith BR, Hinojosa MW, Reavis KM, Nguyen NT. Remission of diabetes after

laparoscopic gastric bypass. The American surgeon. 2008 Oct;74(10):948-52.

26. Adams TD, Davidson LE, Litwin SE, et al. Weight and Metabolic Outcomes 12 Years

after Gastric Bypass. N Engl J Med. 2017 Sep 21;377(12):1143-55.

27. Schauer PR, Bhatt DL, Kirwan JP, et al. Bariatric Surgery versus Intensive Medical

Therapy for Diabetes - 5-Year Outcomes. N Engl J Med. 2017 Feb 16;376(7):641-51.

28. Mingrone G, Panunzi S, De Gaetano A, et al. Bariatric-metabolic surgery versus

conventional medical treatment in obese patients with type 2 diabetes: 5 year follow-up of an

open-label, single-centre, randomised controlled trial. Lancet. 2015 Sep 5;386(9997):964-73.

29. Homan J, Boerboom A, Aarts E, et al. A Longer Biliopancreatic Limb in Roux-en-Y

Gastric Bypass Improves Weight Loss in the First Years After Surgery: Results of a Randomized

Controlled Trial. Obes Surg. 2018 Dec;28(12):3744-55.

30. Sethi M, Chau E, Youn A, et al. Long-term outcomes after biliopancreatic diversion with

and without duodenal switch: 2-, 5-, and 10-year data. Surg Obes Relat Dis. 2016

Nov;12(9):1697-705.

31. Berggren J, Lindqvist A, Hedenbro J, et al. Roux-en-Y gastric bypass versus calorie

restriction: support for surgery per se as the direct contributor to altered responses of insulin and

incretins to a mixed meal. Surg Obes Relat Dis. 2017 Feb;13(2):234-42.

32. Bojsen-Moller KN, Dirksen C, Jorgensen NB, et al. Increased hepatic insulin clearance

after Roux-en-Y gastric bypass. The Journal of clinical endocrinology and metabolism. 2013

Jun;98(6):E1066-71.

22
33. Conceicao EM, Mitchell JE, Pinto-Bastos A, et al. Stability of problematic eating

behaviors and weight loss trajectories after bariatric surgery: a longitudinal observational study.

Surg Obes Relat Dis. 2017 Jun;13(6):1063-70.

34. Browning MG, Baugh NG, Wolfe LG, et al. Evaluation of pre- and postoperative

physical activity participation in laparoscopic gastric banding patients. Obes Surg. 2014

Nov;24(11):1981-6.

35. Bond DS, Evans RK, Wolfe LG, et al. Impact of self-reported physical activity

participation on proportion of excess weight loss and BMI among gastric bypass surgery

patients. The American surgeon. 2004 Sep;70(9):811-4.

36. Stanford FC, Alfaris N, Gomez G, et al. The utility of weight loss medications after

bariatric surgery for weight regain or inadequate weight loss: A multi-center study. Surg Obes

Relat Dis. 2017 Mar;13(3):491-500.

37. Maleckas A, Gudaityte R, Petereit R, et al. Weight regain after gastric bypass: etiology

and treatment options. Gland surgery. 2016 Dec;5(6):617-24.

38. Campos GM, Rabl C, Mulligan K, et al. Factors associated with weight loss after gastric

bypass. Arch Surg. 2008 Sep;143(9):877-83; discussion 84.

39. Ciovica R, Takata M, Vittinghoff E, et al. The impact of roux limb length on weight loss

after gastric bypass. Obes Surg. 2008 Jan;18(1):5-10.

40. Meier JJ, Butler AE, Galasso R, Butler PC. Hyperinsulinemic hypoglycemia after gastric

bypass surgery is not accompanied by islet hyperplasia or increased beta-cell turnover. Diabetes

Care. 2006 Jul;29(7):1554-9.

23
41. Khanna V, Malin SK, Bena J, et al. Adults with long-duration type 2 diabetes have

blunted glycemic and beta-cell function improvements after bariatric surgery. Obesity (Silver

Spring). 2015 Mar;23(3):523-6.

42. Inge TH, Courcoulas AP, Jenkins TM, et al. Five-Year Outcomes of Gastric Bypass in

Adolescents as Compared with Adults. N Engl J Med. 2019 May 30;380(22):2136-45.

43. Campos GM, Rabl C, Peeva S, et al. Improvement in peripheral glucose uptake after

gastric bypass surgery is observed only after substantial weight loss has occurred and correlates

with the magnitude of weight lost. J Gastrointest Surg. 2010 Jan;14(1):15-23.

44. Ardestani A, Rhoads D, Tavakkoli A. Insulin cessation and diabetes remission after

bariatric surgery in adults with insulin-treated type 2 diabetes. Diabetes Care. 2015

Apr;38(4):659-64.

24
Tables

Table 1. Baseline Demographics, Anthropometrics and Clinical Characteristics of Patients with

Type 2 Diabetes Who Underwent Roux-en-Y Gastric Bypass Surgery from January of 1993
through December of 2017
Measure Total Total with No follow-up p Value
(n=706) follow-up data
(n=621) (n=85)
Age, y, mean ± SD 46.6 ± 10.6 46.7 ± 10.6 46.1 ± 10.1 0.612
Female sex, n (%) 555 (79) 500 (81) 55 (65) 0.001
White race, n (%) 510 (72) 444 (72) 66 (78) 0.235
Weight, kg, mean ± SD 139.2 ± 27.3 138.7 ± 27.3 143.0 ± 26.9 0.177
BMI in kg/m2, mean ± SD 49.9 ± 8.5 49.8 ± 8.3 50.3 ± 9.0 0.639
Insulin use, n (%) 211 (30) 189 (30) 22 (26) 0.390
Hypertension, n (%) 515 (73) 464 (75) 51 (60) 0.004
(available for 704 patients)
OSA, n (%) 306 (43) 271 (44) 35 (41) 0.667
Dyslipidemia, n (%) 186 (47) 169 (48) 17 (42) 0.465
(available for 397 patients)
GERD, n (%) 383 (54) 340 (55) 43 (51) 0.470
Known death during follow-up, n 84 (12) 68 (11) 16 (19) 0.035
(%)
Years to last follow-up, median 4.9 (1.9-8.7)
(IQR)
GERD, gastroesophageal reflux disease; OSA, obstructive sleep apnea

25
Table 2. Patient Characteristics and One-Year Weight Outcomes Associated with Type 2
Diabetes Remission after Roux-en-Y Gastric Bypass Surgery
Measure Any remission Non-remission p Value
(n=478) (n=143)
Age, y, median (IQR) 46.4 (38.2-54.1) 50.6 (43.8-55.4) <0.001
Female sex, n (%) 384 (80%) 116 (81%) 0.835
White race, n (%) 344 (72%) 100 (70%) 0.636
Weight, kg, mean ± SD 138.3 ± 27.6 139.9 ± 26.3 0.540
EBW, kg, mean ± SD 68.3 ± 24.0 70.6 ± 24.5 0.305
BMI, kg/m2, median (IQR) 48.2 (43.5-53.8) 49.5 (43.6-55.4) 0.162
Insulin use, n (%) 105 (22%) 84 (59%) <0.001
Hypertension, n (%) 346 (73%) 118 (83%) 0.019

OSA, n (%) 209 (44%) 62 (43%) 0.938

Dyslipidemia, n (%) 124 (47%) 45 (50%) 0.619

GERD, n (%) 255 (53%) 85 (59%) 0.199

Years to final follow-up, median 5.2 (2.0-8.9) 3.7 (1.3-7.8) 0.013
(IQR)
Weight loss outcomes 1-year after
RYGB
∆ BMI, kg/m2, mean ± SD -16.8 ± 5.1 -15.5 ± 5.1 0.015
%TBW loss, mean ± SD -34.0 ± 8.0 -30.2 ± 7.5 <0.001
EBW, excess body weight; GERD, gastroesophageal reflux disease; OSA, obstructive sleep
apnea; TBW, total body weight

26
 TABLE 3. Factors Related to Type 2 Diabetes Remission and Relapse after Roux-en-Y Gastric
Bypass
Independent variable Any remission Relapse
(vs non-remission) (vs sustained remission)
OR (95% CI) p Value OR (95% CI) p Value
Age (years) at baseline 0.97 (0.94-0.99) 0.008 -- --
Insulin use at baseline 0.14 (0.08-0.22) <0.001 2.19 (1.02-4.71) 0.046
Hypertension at baseline 0.54 (0.28-1.04) 0.065 -- --
%TBW loss from baseline to 1 year 1.06 (1.02-1.09) 0.001 0.92 (0.87 -0.96) <0.001
%TBW gain from 1 year to 3-10 years -- -- 1.04 (1.02-1.07) <0.001
%TBW, percentage of total body weight; OR, odds ratio

27
 Table 4. Preoperative Patient Characteristics and Postoperative Weight Loss Outcomes In
Relation to Type 2 Diabetes Remission, Relapse, and Non-Remission
Outcome measure Sustained remission Relapsed p Value
(n=267) (n=89)

Age, y, mean ± SD 45.9 ± 10.8 49.0 ± 10.0 0.017

Female sex, n (%) 219 (82) 74 (83) 0.810
White race, n (%) 187 (70) 65 (73) 0.590
2
BMI, kg/m , median (IQR) 49.2 (43.9-54.3) 46.7 (42.5-51.2) 0.011
EBW, kg, mean ± SD 69.8 ± 24.7 62.1 ± 20.9 0.009
Insulin use, n (%) 42 (16) 30 (34) <0.001
Hypertension, n (%) 191 (72) 70 (80) 0.167
OSA, n (%) 120 (45) 33 (37) 0.194
Dyslipidemia, n (%) 63 (45) 21 (47) 0.845
GERD, n (%) 151 (57) 41 (46) 0.086
Years to remission, median (IQR) 1.0 (1.0-1.1) 1.0 (1.0-1.3) 0.377
Years of follow-up, median (IQR) 6.0 (3.1-9.1) 6.5 (4.1-9.2) 0.279
Weight loss outcome 1 year after RYGB
Patients with follow-up, n 267 89 --
2
BMI in kg/m , median (IQR) 31.6 (28.0-36.4) 31.9 (28.7-35.8) 0.706
Change in BMI from baseline, mean ± SD -17.4 ± 4.9 -15.3 ± 4.7 0.002
%TBW loss from baseline, mean ± SD -34.8 ± 7.2 -31.8 ± 7.9 0.005
Weight loss outcome 3-10 years after RYGB
Patients with follow-up, n 210 80 --
BMI (kg/m2), median (IQR) 32.5 (28.6-38.2) 34.9 (30.7-38.7) 0.041
Change in BMI from baseline, mean ± SD -16.2 ± 7.1 -11.9 ± 6.2 <0.001
%TBW loss from baseline, mean ± SD -34.8 ± 7.7 -31.8 ± 8.3 0.002
%TBW change from 1 year, mean ± SD 3.6 ± 14.8 12.2 ± 24.4 0.001
Weight loss outcome >10 years after RYGB
Patients with follow-up, n 51 41 --
BMI in kg/m2, median (IQR) 32.1 (27.9-41.4) 32.3 (28.0-37.8) 0.903
Change in BMI from baseline, mean ± SD -16.3 ± 8.0 -14.3 ± 6.2 0.201
%TBW loss from baseline, mean ± SD -31.6 ± 12.7 -29.7 ± 12.5 0.242
%TBW change from 1 year, mean ± SD 9.4 ± 21.6 3.4 ± 18.8 0.530
EBW, excess body weight; GERD, gastroesophageal reflux disease; OSA, obstructive sleep
apnea; RYGB, Roux-en-Y gastric bypass; TBW, total body weight

28
Figure Legends

Figure 1. Type 2 diabetes (T2D) status before and after Roux-en-Y gastric bypass.

29
Precis

Longitudinal retrospective analysis of 621 consecutive patients with type 2 diabetes (T2D) who

underwent Roux-en-Y gastric bypass (RYGB) between 1993 and 2017. We investigated factors

associated with T2D remission and relapse after RYGB and found that insulin use, age, 1-year

weight loss, and weight regain mediate T2D remission and relapse after RYGB.

30