Comparative effectiveness of bariatric

surgery and nonsurgical therapy

in adults with type 2 diabetes mellitus
and body mass index <35 kg/m2
Federico J. Serrot, MD,a Robert B. Dorman, MD, PhD,a Christopher J. Miller, BA,b
Bridget Slusarek, RN,a Barbara Sampson, RN,a Brian T. Sick, MD,c Daniel B. Leslie, MD,a
Henry Buchwald, MD, PhD,a and Sayeed Ikramuddin, MD,a Minneapolis, MN

Background. Outcomes of bariatric surgery in patients with a body mass index (BMI) <35 kg/m2 have
been an active area of investigation. We examined the comparative effectiveness of Roux-en-Y gastric
bypass (RYGB) to routine medical management (nonsurgical controls; NSCs) in achieving appropriate
targets defined by the American Diabetes Association for type 2 diabetes mellitus (T2DM) in patients
with class I obesity (BMI 30.0–34.9 kg/m2) T2DM at 1 year.
Methods. We identified patients undergoing RYGB (N = 17) with both class I obesity and T2DM and
compared them to similar NSC (N = 17) treated in the Primary Care Center. Data were collected at
baseline and 1 year for systolic blood pressure (SBP), as well as blood levels for low-density lipoprotein
(LDL) cholesterol and hemoglobin A1c (HbA1c).
Results. After RYGB, BMI decreased from 34.6 ± 0.8 kg/m2 to 25.8 ± 2.5 kg/m2 (P < .001) and
HbA1c decreased from 8.2 ± 2.0% to 6.1 ± 2.7% (P < .001). The NSC cohort had no significant
change in either BMI or HbA1c. SBP and LDL did not significantly change in either group. The RYGB
group had a decrease in medication use compared to the NSC group (P < .001). The RYGB group ceased
the use of antihypertensive and antihyperlipidemia medications by 1 year despite abnormal values.
Conclusion. RYGB can be performed in patients with both a BMI <35 kg/m2 and T2DM with better
weight loss, glycemic control, and fewer antihyperglycemic medications than NSC. Inappropriate cessa-
tion of medications may partially explain the persistent increase in both SBP and LDL after RYGB.
(Surgery 2011;150:684-91.)

From the Department of Surgery,a Division of Biostatistics,b and Department of General Internal Medicine,c
University of Minnesota, Minneapolis, MN

OBESITY AND ITS COMORBID DISEASES continue to be a sub-
stantial burden in health care in the United States.1,2
Chief among the comorbid diseases of obesity is type
2 diabetes mellitus (T2DM). A product of inadequate
beta cell production in the face of increased insulin
resistance, T2DM is a remarkably difficult disease
to manage with long-term complications and
cardiovascular events related to poor control.3-5 Be-
yond improvement of hyperglycemia, successful
therapy of T2DM is measured by achievement of a
composite endpoint of glycosylated hemoglobin
Accepted for publication July 22, 2011.
Reprint requests: Sayeed Ikramuddin, MD, Department of Sur-
gery, University of Minnesota, 420 Delaware Street SE, MMC
290, Minneapolis, MN 55455. E-mail: ikram001@umn.edu.
0039-6060/$ - see front matter
Ó 2011 Mosby, Inc. All rights reserved.
doi:10.1016/j.surg.2011.07.069
A1c (HbA1c) <7%, low density lipoprotein (LDL)
cholesterol <100 mg/dl, and systolic blood pressure
(SBP) <130 mm Hg. In general, 10% of patients will
reach these goals, with success least likely in the
uninsured, young, and minority populations.6
Weight loss is the first step in therapy. In patients
starting with a mean body mass index (BMI) of 36
kg/m2, weight loss achieved with lifestyle modifica-
tion can result in an increase of 12% of patients
achieving this composite goal.7 In addition to pro-
ducing sustained weight loss, bariatric surgery in
morbidly obese patients has been shown to produce
durable and long-lasting effects in improvement of
hyperglycemia and other comorbid diseases.8-10
Current use of surgery is limited to patients with a
BMI >35 kg/m2.11,12 The mean onset of T2DM,
however, occurs at a BMI of 31 kg/m2. More re-
cently, numerous case series prescribed bariatric
surgery, principally the Roux-en-Y gastric bypass
(RYGB), for patients with a BMI <35 kg/m2.11,13-17

684 SURGERY

Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Surgery
Volume 150, Number 4
Data from US sites, however, are sparse, with an
absence of controlled data. The current National
Institutes of Health (NIH) guidelines set a BMI
>35 kg/m2 with comorbidities as an indication for
bariatric surgery; these criteria were established 2
decades ago.18 In response, both the NIH and the
Agency for Healthcare Research and Quality have
called for pilot studies to understand the effect
and mechanisms of bariatric surgery. Pilot data are
essential to design large-scale multicenter studies
that can help understand the role of bariatric
surgery in the reduction of cardiovascular events
in patients with T2DM.
The primary purpose of this paper is to examine
comparative effectiveness in 1-year follow-up data
of RYGB to routine medical management (non-
surgical controls; NSCs) in achieving appropriate
American Diabetes Association (ADA) therapeutic
targets for T2DM in patients with class I obesity
(BMI 30–34.9 kg/m2). We report clinical out-
comes, surgical readmissions, and complications.
METHODS
Patients. We retrospectively reviewed our Uni-
versity of Minnesota bariatric surgery database for
patients who had undergone bariatric surgery with a
BMI <35 kg/m2 and with T2DM between 2001 and
2009. All patients met current NIH criteria for bari-
atric surgery at their initial evaluation. For various
reasons, however, including mandatory preopera-
tive weight loss, the BMI fell below 35 kg/m2 in
this cohort of patients at the time of surgery. In
addition, we reviewed the University of Minnesota
primary care database, which tracks patients under-
going routine medical management, for patients
with a diagnosis of T2DM between 2002 and 2009
to identify NSCs based on BMI matched to patients
undergoing RYGB.
Patients’ heights and weights were obtained in the
bariatric surgery clinic, the primary care center at the
University of Minnesota Medical Center, or by their
primary physician. Laboratory assessments were sim-
ilarly obtained. Medications were reviewed with each
patient at each clinic visit. All patients were followed
for at least 1 year, and time points were compared at
both baseline (time of operation for surgical group)
and at 1 year. All study protocols were approved by
the University of Minnesota Internal Review Board.
Gastric bypass operation. The open and laparo-
scopic RYGB were performed as described previ-
ously.19 Briefly, a 15- to 30-mL gastric pouch is
constructed with a 10- to 12-mm gastrojejunal anas-
tomosis. The Roux limb measures 75 to 150 cm in
length and the biliopancreatic limbs measure 75 to
100 cm in length.
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Serrot et al 685
After bariatric surgery, patients are followed for
complications and dietary and lifestyle counseling.
The defined follow-up intervals for all postbariatric
patients during the first year were 1 week and 1, 3,
6, 9, and 12 months.
Medical management. All NSC patients were
followed in the University of Minnesota primary
care center. The principal goal in medical manage-
ment patients with T2DM is control of hyperglyce-
mia. However, as part of their treatment, patients
receive counseling regarding nutrition, exercise,
and weight management. Optimal outcomes of
diabetes care are based on the latest position
statement of the ADA for the treatment of diabe-
tes.20 Patient visits are based on HbA1c proximity to
a target of <7%. Patients with HbA1c >8% are mon-
itored monthly, those with HbA1c from 6.9% to
7.9% are followed every 3 months, and those at
goal (HbA1c <7%) are monitored every 6 months.
Medications are adjusted by a dedicated phar-
macist. A diabetes nurse specialist meets with the
patients at each visit. Patients are given the oppor-
tunity to attend monthly diabetic support groups,
and every newly diagnosed patient with T2DM is
encourage to attend diabetes education classes.
Weight loss and excess weight loss. Percent
weight loss (%WL) was calculated with the follow-
ing formula: 100 3 (weight at baseline–weight at
1 year)/weight at baseline. Percent excess weight
loss (%EWL) was calculated with the following
formula: 100 3 (weight at baseline–weight at
1 year)/(weight at baseline–ideal weight). Ideal
weight was obtained from the Metropolitan tables
based on height and sex.21
Medication use. Medication use for hyperglyce-
mia control was measured using a medication score.
Patients using no medications to control hypergly-
cemia were classified as 0. Patients using metformin
were classified as a 1. An additional point was
assigned for use of each additional oral agent,
such as a sulfonylurea or a thiazolidinedione. Two
points were added for patients requiring insulin,
and a final point was assigned to patients requiring
exenatide. The maximum medication score was 5.
Medication use for dyslipidemia and hyperten-
sion treatment was determined as ‘‘yes’’ or ‘‘no,’’ if
they were or were not using drugs.
Statistical analysis. Because of the relatively
small sample size of each group, we used nonpara-
metric and exact methods for all data analysis.
Differences between the control and surgical
groups at baseline were assessed with Wilcoxon
rank-sum tests for continuous variables and the
Fisher exact test for medication scores. We indexed
1-year improvement in BMI, HbA1c, LDL, and SBP
686 Serrot et al
using within-person difference scores between year
1 and baseline, and analyzed the difference be-
tween groups using rank-sum tests. We analyzed
medication scores by testing the difference in the
proportion of individuals whose medication scores
were decreased at 1 year. All statistics are reported
as median ± interquartile range (IQR), and
P values are for differences in change scores or dif-
ferences in proportions, unless otherwise noted.
We conducted post-hoc nonparametric power
analyses to determine the type II error rate (‘‘false
negative’’ rate) given the differences we observed.
All statistical analyses were conducted with R soft-
ware (version 2.12.1; R Development Core Team)
and power analyses were conducted with PASS
software (version 11; NCSS, Kaysville, UT).22
RESULTS
Between January 2001 and July 2009, 2,934
consecutive patients underwent RYGB; 605
(20.0%) of these patients had a diagnosis of
T2DM, excluding patients with impaired fasting
glucose (IFG) and impaired glucose tolerance
(IGT), with a mean BMI of 48.4 kg/m2 and a
mean age of 48.7 years. Seventeen of the 605 pa-
tients had a BMI between 30 and 34.9 kg/m2 at
the time of RYGB. In the NSC database, we identi-
fied 250 patients undergoing management of
T2DM with a BMI of 30.1 to 62.3 kg/m2. From
this group, we identified a total of 17 patients
with a matching BMI to serve as controls for the
patients undergoing RYGB. Equivalence of the
sample size for the 2 groups was coincidental.
Demographic and baseline information for age,
sex, weight, BMI, HbA1c, LDL, and SBP for each
group are displayed in the Table. The RYGB and
NSC groups were overall comparable at baseline,
although significant differences in baseline values
were observed for age (P = .05), sex (P = .04),
and HbA1c (P = .04). The difference in sex is con-
sistent with the bariatric surgical population, which
is predominantly female. The difference observed
in HbA1c proved to be inconsequential, because
the RYGB group, with a HbA1c significantly
greater at baseline, had values significantly less
than the NSC at 1 year.
Weight loss. Significant differences were found
with regard to weight loss between both groups.
BMI decreased from 34.6 ± 0.8 kg/m2 to 25.8 ±
2.5 kg/m2 in the RYGB group and did not change
perceptively in the NSC group (34.0 ± 1.0 kg/m2 to
34.3 ± 2.1 kg/m2; P < .001), Figure 1. Weight loss
followed a similar pattern, decreasing from 214 ±
25 to 157 ± 20 pounds in the RYGB group with no
meaningful change in the NSC group (237 ± 40 to
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Surgery
October 2011
233 ± 38 pounds; P < .001). Significant differences
were also observed when %WL and %EWL were
compared; %WL was 25% ± 6% in the RYGB group
and
1% ± 4% in the NSC group (P < .001), and %
EWL was 70% ± 21% in the RYGB group and
4% ±
10% in the NSC group (P < .001). Negative results
of %WL and %EWL in the NSC indicate that this
group actually gained weight.
Glycemic control, cholesterol, and blood pres-
sure. HbA1c decreased in the RYGB group from
8.2 ± 2.0% to 6.1 ± 2.7% at 1 year after RYGB but
did not change in the NSC group (7.0 ± 0.7% to
7.1 ± 1.8%; P < .001), Figure 2.
One year after intervention, SBP did not change
in either the RYGB or the medically managed
group. SBP increased nonsignificantly from 126 ±
28 mm Hg to 132 ± 27 mm Hg in the RYGB group
and decreased nonsignificantly from 126 ± 30 mm
Hg to 124 ± 26 mm Hg in the NSC group (P = .99).
LDL cholesterol was not significantly changed in
either group. LDL increased nonsignificantly
from 86 ± 44 mg/dL to 92 ± 62 mg/dL in the
RYGB group and 95 ± 47 mg/dL to 100 ± 66
mg/dL in the NSC group (P = .25). Eleven of
the 17 patients undergoing RYGB no longer met
the criteria for T2DM based on obtaining a
HbA1c <6.5%, and of those, 4 no longer met the
diagnosis of prediabetes based on a HbA1c <5.7.20
Post-hoc power analyses determined that the
current study was well powered to analyze the
difference in HbA1c (93% power). The differences
observed in LDL and SBP were not clinically rele-
vant, because they were unchanged in each group.
As a result, the current study was underpowered to
analyze the observed differences if they were em-
pirically true, because the observed power was only
12% and 5% for LDL and SBP, respectively.
Medication scores. The RYGB and NSC groups
were comparable with regard to their antihypergly-
cemia, antihypertensive, and antihyperlipidemia
medications at baseline (Table). After 1 year,
the RYGB group was taking fewer medications for
glycemia control than the NSC group, with 71%
of patients taking fewer medications at 1 year com-
pared to 6% in the NSC group (P < .001). Impor-
tantly, 3 of the 9 patients no longer taking
medications had an HbA1c >7%. More surpris-
ingly, 41% of the RYGB group had ceased the
use of antihypertensive medications despite the
fact that their SBP had not improved, compared
to 6% of the NSC group (P = .04). Similarly, 35%
of the RYGB group were no longer taking antihy-
perlipidemia medications 1 year after RYGB, even
though their LDL cholesterol levels had not im-
proved; none of the NSC group, on the other
Surgery Serrot et al 687
Volume 150, Number 4

Table. Demographics of RYGB and matched NSC

RYGB NSC
Median (IQR) or N (%) Median (IQR) or N (%) P value
N 17 17
Age (yrs) 56.0 (7.0) 62.0 (12.0) .05
Sex (female) 13 (76.5%) 6 (35.5%) .04
BMI (kg/m2) 34.6 (0.8) 34.0 (1.0) .46
Weight (lbs) 214 (24.9) 237 (40.1) .13
HbA1C (%) 8.2 (2.0) 7.0 (0.7) .04
LDL (mg/dL) 95 (44) 86 (47) .64
SBP (mm Hg) 126 (28) 126 (30) .43
Diabetes medications* 14 (82.3%) 9 (52.9%) .25
Hypertensive medicationsy 15 (88.2%) 12 (70.6%) .40
Hyperlipidemia medicationsy 12 (70.6%) 11 (64.7%) 1.00
*Summary statistics reflect proportion with score $2, P value reflects the Fisher exact test for difference in distribution of medication score.
yProportions for hypertensive and hyperlipidemia medications reflect the proportion of patients taking medications at baseline.
BMI, Body mass index; HbA1c, glycosylated hemoglobin A1c; IQR, interquartile range; LDL, low-density lipoprotein cholesterol; NSC, nonsurgical control;
RYGB, Roux-en-Y gastric bypass; SBP, systolic blood pressure.

Fig 1. Box Plot of BMI by treatment group at baseline and
1-year follow up. BMI decreased from 34.6 ± 0.8 kg/m2 to
25.8 ± 2.5 kg/m2 in the RYGB group and did not change in
the NSC group from 34.0 ± 1.0 kg/m2 to 34.3 ± 2.1 kg/m2
(P < .001). BMI, Body mass index; RYGB, Roux-en-Y gas-
tric bypass; NSC, non-surgical control.
hand, stopped taking medications (P = .02). All
analyses for the proportion of individuals ceasing
medications were generally well powered.
Readmissions and complications. Total readmis-
sion rate for the RYGB group was 18%. Within a
year, 4 complications were observed in the RYGB
cohort of patients; 2 incisional hernias that re-
quired operative repair and 2 marginal ulcers that
were successfully medically managed. No episodes
of hypoglycemia were reported in the RYGB co-
hort. There were no mortalities in either group.
Fig 2. Box Plot of HbA1C by treatment group at baseline
and 1-year follow up. HbA1C decreased in the RYGB
group from 8.2 ± 2.0% to 6.1 ± 2.7% at one year after sur-
gery but did not change in the NSC group from 7.0 ±
0.7% to 7.1 ± 1.8 (P < .001). HbA1C, Glycosylated
hemoglobin A1C; RYGB, Roux-en-Y gastric bypass;
NSC, non-surgical control.
DISCUSSION
This is the first comparison of RYGB patients to
medical managed patients with a BMI <35 kg/m2,
and it allows for a number of interesting observa-
tions. These data indicate that the RYGB in this
patient population can be performed safely in stan-
dard fashion. On average, the BMI of RYGB patients
in our study decreased to 25.8 ± 2.5 kg/m2. All pa-
tients were considered normal to slightly overweight

Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
688 Serrot et al
at 1 year; in comparison, although patients under-
going medical management received counseling
regarding nutrition and physical activity as part of
their treatment, no substantial weight loss occurred
at 1 year and they remained obese.
In a randomized controlled trial, Lee at al11
revealed that patients with a BMI of 25.0 to 34.5
kg/m2 who underwent RYGB were more likely to
achieve remission of T2DM when compared to pa-
tients randomized to sleeve gastrectomy without
duodenal exclusion. Patients assigned to gastric by-
pass lost more weight and had both lesser blood
glucose levels and HbA1c values than the sleeve
gastrectomy group at 1 year.11 In an uncontrolled
case series study by Cohen et al,14 patients with a
BMI <35 kg/m2 and severe comorbidities
benefited from laparoscopic RYGB with dramatic
improvement of metabolic syndrome. All patients
had normal fasting glucose and HbA1c <6%, and
%EWL was 72% at 1 year.14 A randomized trial of
laparoscopic adjustable gastric banding vs inten-
sive medical management had the greatest efficacy
in patients with mild T2DM relative to weight
loss.16 Not surprisingly, the glycemic improvements
of our patients were consistent with all studies in
patients undergoing RYGB, with a significant
decrease in HbA1c from 8.2% to 6.1%. This de-
creases places them well below within the specified
parameters set forth by the ADA.20
Interestingly, the 3 patients who had a HbA1c >7%
1 year after RYGB were no longer taking antihyper-
glycemic medications. This situation---likely inappro-
priate considering that these patients were
approaching the nadir of their weight loss and the
peak impact of weight decrease on glycemic control---
should promote caution defining remission of
T2DM as being the cessation of medications in the
literature. At baseline, HbA1c levels were signifi-
cantly less in the NSC group; however, the medical
group did not have any change in HbA1c. The NSC
patients on average had diabetic values that re-
mained outside of the ADA goals for treatment of
T2DM. With regard to hypertension and dyslipide-
mia, patients in either group had clinical decreases in
these variables; simultaneously, RYGB patients were
not consistent in taking their medications. The
importance of these findings are more striking in
light of the observation of the increased mortality
associated with intensive medical management to
achieve HbA1c <6%.23-25
A concern has been raised for post–gastric bypass
hypoglycemia.26,27 None of our patients had this
phenomenon, which is reported rarely in patients
with significant T2DM who are likely not to have a
marked beta cell reserve. In addition, these patients
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Surgery
October 2011
also have insulin resistance that is not completely
reversed by the time maximal weight loss occurs
(typically 1–2 years after following gastric bypass
surgery).28 Hypoglycemic events in the medical
patients were not collected in our study.
The incidence of readmissions was 18% within
the first year, and this is comparable to other
published 1-year data for patients with BMI >35
kg/m2.29 Thirty-day readmission rates range from
3% to 10% in patients with BMI >35 kg/m2 under-
going RYGB.30
While we believe that this study provides impor-
tant data, limitations do exist. First, the study was
limited by the relatively small number of patients
with low BMI and T2DM who underwent bariatric
surgery in our institution. Also, while HbA1c was
significantly greater in the RYGB cohort, the dis-
parity favored the medical cohort at baseline.
There was also a difference in sex between the 2
groups. In the United States, women comprise
approximately 80% of the bariatric surgery popu-
lation; inherent sex differences exist between the
bariatric patient population and the general pop-
ulation of patients with T2DM. In addition, while
the intensity of medical management is standard-
ized within our Endocrinology Clinic as noted in
our methods, poor patient compliance could cer-
tainly attenuate optimal outcomes in the medically
managed patients. While physicians intervene to
achieve specific therapeutic goals, poor patient
compliance can hinder this effort. This factor
could create a bias toward the patients managed
by RYGB having a better outcome.
We propose that the next step is to identify and
establish clinical studies to identify the impact of
bariatric surgery, particularly the RYGB, in achiev-
ing a decrease in cardiovascular risk in patients
who meet traditional NIH criteria for bariatric
surgery and also to expand studies to include class
I obesity. This study addresses the latter, and there
are number of studies underway that will comment
on the broader impact of bariatric surgery on
surrogate cardiovascular markers and glycemia.
REFERENCES
1. Kral JG. Morbidity of severe obesity. Surg Clin North Am
2001;81:1039-61.
2. Sivalingam SK, Ashraf J, Vallurupalli N, Friderici J, Cook J,
Rothberg MB. Ethnic differences in the self-recognition of
obesity and obesity-related comorbidities: a cross-sectional
analysis. J Gen Intern Med 2011;26:616-20.
3. Hu FB, Manson JE, Stampfer MJ, Colditz G, Liu S, Solomon
CG, et al. Diet, lifestyle, and the risk of type 2 diabetes mel-
litus in women. N Engl J Med 2001;345:790-7.
4. Colditz GA, Willett WC, Rotnitzky A, Manson JE. Weight
gain as a risk factor for clinical diabetes mellitus in women.
Ann Intern Med 1995;122:481-6.
Surgery
Volume 150, Number 4
5. Hogen P, Dall T, Nikolov P. Economic costs of diabetes in
the US in 2002. Diabetes Care 2003;26:917-32.
6. Saydeh SH, Fradkin J, Cowie CC. Poor control of risk factors
for vascular disease among adults with previously diagnosed
diabetes. JAMA 2004;291:335-42.
7. Redmon JB, Bertoni AG, Connelly S, Feeney PA, Glasser SP,
Glick H, et al. Effect of the look AHEAD study intervention
on medication use and related cost to treat cardiovascular
disease risk factors in individuals with type 2 diabetes. Dia-
betes Care 2010;33:1153-8.
8. Pories WJ, Swanson MS, MacDonald KG, Long SB, Morris
PG, Brown BM, et al. Who would have thought it? An oper-
ation proves to be the most effective therapy for adult-onset
diabetes mellitus. Ann Surg 1995;222:339-50.
9. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W,
Fahrbach K, et al. Bariatric surgery a systemic review and
meta-analysis. JAMA 2004;292:1724-37.
10. Buchwald H, Estok R, Fahrbach K, Banel D, Jensen MD, Po-
ries WJ, et al. Weight and type 2 diabetes after bariatric sur-
gery: systematic review and meta-analysis. Am J Med 2009;
122:248-56.
11. Lee WJ, Chong K, Ser KH, Lee YC, Chen SC, Chen JC, et al.
Gastric bypass vs sleeve gastrectomy for type 2 diabetes melli-
tus: a randomized controlled trial. Arch Surg 2011;146:143-8.
12. Hofsø D, Jenssen T, Bollerslev J, Ueland T, Godang K, Stum-
voll M, et al. Beta cell function after weight loss: a clinical
trial comparing gastric bypass surgery and intensive lifestyle
intervention. Eur J Endocrinol 2011;164:231-8.
13. Lee WJ, Wang W, Lee YC, Huang MT, Ser KH, Chen JC. Effect
of laparoscopic mini-gastric bypass for type 2 diabetes melli-
tus: comparison of BMI>35 and <35 kg/m2. J Gastrointest
Surg 2008;12:945-52.
14. Cohen R, Pinheiro JS, Correa JL, Schiavon CA. Laparo-
scopic Roux-en-Y gastric bypass for BMI <35 kg/m(2): a tai-
lored approach. Surg Obes Relat Dis 2006;2:401-4.
15. Angrisani L, Favretti F, Furbetta F, Iuppa A, Doldi SB, Paga-
nelli M, et al. Italian Group for Lap-Band System: results of
multicenter study on patients with BMI < or =35 kg/m2.
Obes Surg 2004;14:415-8.
16. O’Brien PE, Dixon JB, Laurie C, Skinner S, Proietto J,
McNeil J, et al. Treatment of mild to moderate obesity
with laparoscopic adjustable gastric banding or an intensive
medical program. Ann Intern Med 2006;144:625-33.
17. Fobi M, Lee H, Igwe D, Felahy B, James E, Stanczyk M, et al.
Gastric bypass in patients with BMI <40 but >32 without life-
threatening co-morbidities: preliminary report. Obes Surg
2002;12:52-6.
18. Gastrointestinal surgery for severe obesity. Consens State-
ment 1991;9:1-20.
19. Ikramuddin S, Kendrick ML, Kellogg TA, Sarr MG. Open
and laparoscopic Roux-en-Y gastric bypass: our techniques.
J Gastrointest Surg 2007;11:217-28.
20. American Diabetes Association. Standards of medical care
in diabetes---2011. Diabetes Care 2011;34(Suppl 1):S11-61.
21. Build Study, 1979. Chicago: Society of Actuaries and Associ-
ation of Life Insurance Medical Directors of America; 1980.
22. Hintze J. PASS 11. Kaysville, UT: NCSS, LLC; 2011.
23. Riddle MC. Effects of intensive glucose lowering in the
management of patients with type 2 diabetes mellitus in
the Action to Control Cardiovascular Risk in Diabetes (AC-
CORD) trial. Circulation 2010;122:844-6.
24. Siegel D, Swislocki AL. The ACCORD Study: the devil is in
the details. Metab Syndr Relat Disord 2011;9:81-4.
25. Klein R. Intensive treatment of hyperglycaemia: ACCORD.
Lancet 2010;376:391-2.
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Serrot et al 689
26. Kellogg TA, Bantle JP, Leslie DB, Redmond JB, Slusarek B,
Swan T, et al. Postgastric bypass hyperinsulinemic hypogly-
cemia syndrome: characterization and response to a modi-
fied diet. Surg Obes Relat Dis 2008;4:492-9.
27. Bantle JP, Ikramuddin S, Kellogg TA, Buchwald H. Hyperin-
sulinemic hypoglycemia developing late after gastric bypass.
Obes Surg 2007;17:592-4.
28. Muscelli E, Mingrone G, Camastra S, Manco M, Pereira JA,
Pareja JC, et al. Differential effect of weight loss on insulin
resistance in surgically treated obese patients. Am J Med
2005;118:51-7.
29. Saunders J, Ballantyne GH, Belsley S, Stephens DJ, Trivedi
A, Ewing DR, et al. One-year readmission rates at a high vol-
ume bariatric surgery center: laparoscopic adjustable gastric
banding, laparoscopic gastric bypass, and vertical banded
gastroplasty-Roux-en-Y gastric bypass. Obes Surg 2008;18:
1233-40.
30. Saunders JK, Ballantyne GH, Belsley S, Stephens D, Trivedi
A, Ewing DR, et al. 30-day readmission rates at a high vol-
ume bariatric surgery center: laparoscopic adjustable gastric
banding, laparoscopic gastric bypass, and vertical banded
gastroplasty-Roux-en-Y gastric bypass. Obes Surg 2007;17:
1171-7.
DISCUSSION
Dr Armour Forse (Omaha, NE): Our group, as well as
several others, including those at the University of Min-
nesota, have shown that bariatric surgery, despite being
performed in very morbid patients, is safe surgery. In ad-
dition, studies have shown that bariatric surgery, specifi-
cally a gastric bypass, is associated with significant GI
hormonal changes which favor the cure of type 2 diabe-
tes. Hence, it is very appropriate to investigate the use of
this surgery in patients with diabetes at a low BMI below
the NIH-accepted BMI of 35. The authors found, in a se-
lect group of type 2 diabetic patients with class 1 obesity,
that gastric bypass surgery resulted in a significant de-
crease in hemoglobin A1c and a decrease in the use of
diabetic medications. Neither was achieved in a medi-
cally treated control group. While these data are exciting
and promising, we do need to be very thoughtful about
the study before a broader application of these findings.
In the first place, the study is retrospective and the low
BMI surgical patients had been dieting in preparation for
bariatric surgery, thus not representing really, de novo,
low---BMI patients. The control group of medically treated
diabetic patients is not well matched to the surgical group.
I have a few questions to help clarify the study.
Why was such a variation in surgery used and how was
the surgery selected for each patient? Did the lower BMI
patients get the 75---cm Roux and biliopancreatic limbs
vs the 150---cm limbs?
While safe, the gastric bypass is known for complica-
tions such as marginal ulcers and incisional hernias. Two
out of the 17 patients, or 12%, with a marginal ulcer and
2, or 12%, with incisional hernias in 1 year seems high. Is
that a concern and something unique to this group of
patients, particularly the diabetic patient?
The other problem associated with gastric bypass is the
requirement for vitamin and mineral supplementation,
in some studies, as high as 50% of the patients. Did any of
690 Serrot et al
these patients develop such problems or was 1 year a little
too early to have developed these problems?
Finally, most of us see maximal weight loss until up to
about 18 months post–gastric bypass. Does this early
reporting of the weight loss perhaps mislead us to the
potential weight loss problems that might occur in low---
BMI patients?
Dr Federico Serrot (Minneapolis, MN): Regarding
the first question, I haven’t mentioned this here, but
we didn’t actually change the surgery depending on
the BMI of the patients. Actually, at the University of
Minnesota, Dr Buchwald does the Roux-en-Y with a
Roux limb of 75 cm. And Dr Ikramuddin performs a
Roux limb of 150 cm. So we didn’t change any; this
was depending on the surgery. And also, Dr Buchwald
performs open, comparing to Dr Ikramuddin, who
performs laparoscopic surgery.
But I didn’t actually look at if there were any differ-
ences depending on the length of the Roux limb or
the BP limb in these patients.
Regarding your second question, the rate of compli-
cations or readmissions, I think we are mostly familiar
with the rates of remissions in 30-day readmission rates,
and those are usually between 6% and 10%. But in some
of the few published data, they have shown that at 1 year,
the overall readmission rate for bariatric surgery is
around 18%.
Your question regarding the vitamins and mineral
supplementation, I haven’t looked at that in this study.
That’s a great comment. Finally, regarding the weight
loss and excessive weight loss, I think you are right,
too. In the patients that we usually perform gastric by-
pass, the weight loss achievement is usually greatest
weight loss is 18 months. But in this group of patients,
at least, I don’t have the data. But it seems that weight
loss seems to be a little slower and the weight loss seems
to decrease at one year compared to patients with
greater BMI.
Dr Scott Melvin (Columbus, OH): There’s no doubt
that obesity is at epidemic proportions in the United
States and the sequelae of type 2 diabetes is important.
It’s also recognized that, as you demonstrated very nicely,
that optimal medical management is in fact not optimal,
and that the guidelines set forth for management of type
2 diabetes is not being met by your primary care medical
management team. Surgery is an option for that. So I
appreciate and we should applaud the efforts to try to
further identify ways to better medically, better surgically
manage type 2 diabetes. With that in mind, I have a cou-
ple of comments, and specific questions.
You describe your comparison to surgery vs optimal
medical management. Your surgery program there is
very well regimented, despite the differences and minor
technical differences in the surgery. I’m sure there’s a
very discrete algorithm of the way your patients are man-
aged, with dietary counseling, exercise, and nutrition.
Are those same plans used in the medical management?
And are they as regimented? Obviously, they are not do-
ing a good job at controlling the hemoglobin A1c.
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Surgery
October 2011
For your follow-up period, did you perform a cost
analysis? Or, at some point in time, can you do some cost
modeling that would demonstrate better control of type
2 diabetes and the subsequent sequelae of that as
compared to the control in medical management and
include the cost of the medications that, as you have
demonstrated, have been markedly decreased in the
patients who have undergone surgery?
In your patients with the 17% readmission rate at
1 year, what was the admission rate for the patients who
were being medically managed? I think over time that
that may make a difference, and perhaps 1 year is too
short of a framework to identify those changes.
Three of your patients were not off medications for
type 2 diabetes. Three out of 17 seems a little bit high for
patients for type 2 diabetes to still have poor hemoglobin
A1c control. So I think, very happily, you have described
in your manuscript that type 2 diabetes is not cured but
it remains in remission. So if you wanted to comment
about that, that would be great.
Lastly, I think it’s reasonable to discuss your findings
in the setting that we all see in the paper every day of the
expanded indications for laparoscopic banding or gas-
tric banding for patients with low BMI rates. These
data, combined with the data that were submitted for
the increased expansion of the role of banding, should
we now consider this as standard therapy for patients
with lower BMI rates?
Dr Federico Serrot (Minneapolis, MN): Thank you
for your comments and questions. Regarding the first
question, yes; I don’t have those data, exactly how the
primary care physicians follow their patients compared
to how we do with our patients. So I am not exactly
sure if the visits and the follow-up is exactly the same
that they use between bariatric surgery and primary
care physicians.
Regarding your second question, in the cost analysis,
I think that’s a great comment and something that we
should do. Especially, as you said, regarding the cost of
medication in the treatment of diabetes, as we show
70% of these operated patients were taking fewer
medications.
Actually, regarding your other question, how many
patients were actually not taking any medications for
treatment of type 2 diabetes, at 1 year, from the 17
patients, actually 9 patients were not taking any medica-
tions for treatment of diabetes in the surgical group.
And actually, from those patients, as I said, 3 were not
a goal. So I think that’s something that we need to look
at when we are managing these patients. And we haven’t
looked at the readmission rate in medical patients. So I
think that should be something that we need to look at,
how many of these patients were readmitted, especially
for hypoglycemic episodes or any other complications
actually related to type 2 diabetes or treatment for type
2 diabetes.
Regarding the use of lap band and Roux-en-Y bypass,
I think, now that the FDA has approved the use of a
laparoscopic gastric band on patients with BMI <35, I
Surgery
Volume 150, Number 4
am not sure how this is going to or how much the
healthcare insurance are going to accept. And also,
because we know that the lap band does not affect the
type 2 diabetes in the same way that Roux-en-Y gastric
bypass does. I think that’s a great comment. I don’t know
how easy this is going to be, especially for the NIH to
accept this.
Dr David Dexter (Erie, PA): I did enjoy your presen-
tation. And I noticed that your mean age, I believe, was
55 years of age. But we are noticing an epidemic of obe-
sity in younger and younger patient populations.
So my question is, how do you feel we can apply this
to the age range of 20 to 30? And then, once we get to
that age, how do we start applying this to the high school
age population? Will lap band apply more at that age?
I personally see more and more patients that are
obese between the 20 and 30 age range, so I am curious
how you would apply this to that.
Dr Federico Serrot (Minneapolis, MN): We do also
see at the University of Minnesota a great number of ad-
olescents coming in with obesity and type 2 diabetes.
And I think, in these patients, the mean age was higher
than what you were mentioning. I think it’s something
that would be interesting to see if we can start applying
these type of treatments to these, especially in order to
prevent further complications from type 2 diabetes and
obesity.
Dr Gerald Larson (Louisville, KY): Very interesting
concept. And there’s growing evidence that it’s the by-
pass of the duodenum; that is, nutrition and food are
not going through the duodenum, and therefore certain
critical mediators, incretins and whatnot, which control
blood sugar levels are being bypassed.
And that probably is the mechanism because many
obese patients, will have normalization or improvement
of their insulin requirement within the first month of
the surgery, when weight loss has only been maybe 5 to
25 pounds.
My question is, in your medically treated group, do
you have a few patients who actually did lose weight and
go with a BMI maybe from 32 down to 26, and then
Downloaded for Anonymous User (n/a) at University at Buffalo The State University of New York from ClinicalKey.com by Elsevier on September 05, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Serrot et al 691
looked at their hemoglobin A1c levels? You have clearly
shown that medical treatment is not effective for morbid
obesity or severe obesity, but did you have a subset of
patients who really did lose weight? And what happened
to their other parameters?
Dr Federico Serrot (Minneapolis, MN): We did have a
subset of patients that did lose weight; but actually, their
hemoglobin A1c did not actually significantly change in
those patients. So as you said, it might be related to the
fact that the bypass on these patients are having the in-
cretin effect, probably because of GOP1 secretion. But
again, the weight loss was not, I think, significant at
1 year in these patients, although they had some weight
loss, and the A1c changes were not significant either in
the medical management.
Dr Henry Buchwald (Minneapolis, MN): I rise just to
add some additional data to the entire discussion in
terms of questions that have been asked.
There are several excellent papers in the literature
now, notably one by Cremier, another one by Cristo, that
show that the cost of taking care of patients after
bariatric metabolic surgery is lower than medical care
of these patients after about 2 to 4 years. That is, the care
of a medically treated patient for their comorbidities is
higher after a certain period of time than treating the
patient with surgery.
The second interesting piece of information is a study
being done by Scopanaro in Genoa, Italy, who is treating
patients with diabetes with a BMI between 25 and 30.
And he’s using his biliopancreatic diversion, which every-
body will say, including Scopanaro, is the most weight-
losing operation that we do. And he has a common
channel of about 50 cm. And yet, these patients lose
their diabetes, but they don’t shrink away. They do not
go below a BMI of 24 or 25.
So it’s important to point out that the operations that
we first used for bariatric surgery can now possibly be
used for metabolic surgery. And in individuals who are
close to normal weight, we are not going to get a tremen-
dous weight loss. But we may be able to achieve a resolu-
tion of metabolic problems such as diabetes.