European Journal of Cancer 52 (2016) 129e137

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.ejcancer.com

Original Research

Survival impact of postoperative body mass index in

gastric cancer patients undergoing gastrectomy

Han Hee Lee a, Jae Myung Park a,*, Kyo Young Song b,
Myung-Gyu Choi a, Cho Hyun Park b

a
Division of Gastroenterology, Department of Internal Medicine, College of Medicine, The Catholic University of Korea,
Seoul, Republic of Korea
b
Division of Gastrointestinal Surgery, Department of Surgery, College of Medicine, The Catholic University of Korea, Seoul,
Republic of Korea

Received 25 June 2015; received in revised form 17 October 2015; accepted 24 October 2015
Available online 10 December 2015

KEYWORDS Abstract Background: The relationship between preoperative body mass index (BMI) and the
Stomach neoplasms; survival of postoperative gastric cancer patients is not clear. Furthermore, the survival impact
Body mass index; with postoperative BMI is not known, even though weight loss is inevitable after gastrectomy.
Prognosis; Methods: Patients who underwent gastrectomy for gastric cancer between 2000 and 2008 were
Survival; included in the study (n Z 1909). Patients were divided into three groups based on their BMIs:
Gastrectomy low (<18.5 kg/m2), normal (18.5e24.9 kg/m2), and high BMI (
25.0 kg/m2). Patient survival
was compared according to BMI at two time points: baseline and 1 year after surgery.
Results: Regarding BMI 1 year after surgery, overall survival, disease-specific survival, and
recurrence-free survival were longer in the high BMI group than the low and normal BMI
groups. In a Cox proportional hazards model, adjusting for the patient’s age, sex, type of sur-
gery, tumour stage, histology, curative resection, and BMI at baseline, a high BMI 1 year after
surgery was associated with lower overall mortality compared to normal BMI (hazard ratio
0.51; 95% confidence interval, 0.26e0.98). However, BMI at baseline was not an independent
prognostic factor.
Conclusion: BMI 1 year after surgery significantly predicted the long-term survival of patients
with gastric cancer compared with the preoperative BMI.
ª 2015 Elsevier Ltd. All rights reserved.

* Corresponding author: Division of Gastroenterology, Department of Internal Medicine, Seoul St. Mary’s Hospital, College of Medicine, The
Catholic University of Korea, 222 Banpo-daero, Seocho-gu, Seoul 137-701, Republic of Korea. Tel.: þ82 2 2258 6022; fax: þ82 2 2258 2089.
E-mail address: parkjerry@catholic.ac.kr (J.M. Park).

http://dx.doi.org/10.1016/j.ejca.2015.10.061
0959-8049/ª 2015 Elsevier Ltd. All rights reserved.
130 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137
1. Introduction
The increase in numbers of overweight and obese
people is a major public health concern. Excess body
weight can cause several lifestyle diseases, such as dia-
betes, hypertension, hyperlipidaemia and cardiovascular
disease. Overweight and obese statuses are associated
with an increased risk of death.[1] Body mass index
(BMI) is the most widely used method to estimate
obesity. It is very easy to measure and correlates better
with total body fat than with body weight alone.[2] BMI
is classified as underweight (<18.50 kg/m2), normal
(18.50e24.99 kg/m2), overweight (
25.00 kg/m2), or
obese (
30.00 kg/m2).[3] These current World Health
Organisation cut-offs provide an adequate basis for
taking action against risks related to being overweight
and obese in many populations.
Contrary to conventional belief, overweight and
obese patients may paradoxically have favourable out-
comes compared with patients of normal weight. This
has been supported by several recent studies on critically
and chronically ill patients, such as those suffering acute
lung injury, heart failure, chronic renal failure, chronic
obstructive pulmonary disease and rheumatoid
arthritis.[4e8] Although several studies have been con-
ducted in this population, conflicting results have been
reported about the effect of excess body weight on long-
term surgical outcomes.[9e11] These studies analysed
patient outcomes based on their preoperative BMI.
However, unlike other surgeries, gastrectomy is always
followed by weight loss. Therefore, the preoperative
BMI of the patients undergoing gastrectomy will differ
from their postoperative BMI. Prominent losses in body
weight and abdominal fat are observed during the first 3
months after gastrectomy in patients with early gastric
cancer. Body weights are then maintained without any
significant changes until 12 months after surgery.[12]
The aims of this study were to evaluate the relation-
ship between BMI and long-term survival of gastric
cancer patients undergoing gastrectomy using preoper-
ative and postoperative BMIs.
2. Material and methods
2.1. Patients
From January 2000 to December 2008, 1989 patients
with gastric cancer underwent gastrectomy at the Seoul
St. Mary’s Hospital, Korea. Gastric cancer patients who
underwent gastric wedge resection (n Z 6), Whipple’s
pancreaticoduodenectomy (n Z 1), palliative gastro-
jejunostomy (n Z 32), and open and closure (n Z 41)
procedures or had no BMI data (n Z 4) were excluded.
An electronic database of the remaining 1905 patients
was reviewed. All relevant data, including clinical,
surgical, and pathological records, were collected
retrospectively from the database. Cancers were staged
in accordance with the tumour node metastasis
classification system used by the Union Internacional
Contra la Cancrum, 6th edition.[13] The Institutional
Review Board of our institution approved this study
(KC14RISI0716).
Patients were divided into three groups on the basis
of their BMIs. The height and weight of all patients were
measured on the admission day for gastrectomy, and the
weight 1 year after gastrectomy was gauged on the scale
in outpatient clinic. Patients with BMIs <18.50 kg/m2,
18.50e24.99 kg/m2, or
25.00 kg/m2 were classified as
having a low, normal, or high BMI, respectively. This
grouping was applied to BMIs both at baseline and 1
year after gastrectomy.
2.2. Measurements
The percentage weight loss was calculated as the dif-
ference between the two BMIs divided by the BMI at
baseline. The histology of gastric cancer was divided
into two groups, differentiated and undifferentiated,
based on the Japanese classification of gastric carci-
noma.[14] Three patients had neuroendocrine carci-
nomas and two patients were missing data regarding
histology and were thus excluded from the multivariable
analysis. R0 resection was pathologically defined as
complete surgical resection without residual tumour. R1
resection was defined as no residual tumour except for a
positive resection margin. R2 resection was defined as
palliative resection with visible residual tumour.
2.3. Outcomes
All included patients were registered by the Compre-
hensive Cancer Institute of Seoul St. Mary’s Hospital.
The cause of death was determined by physician review
of the medical records in the institution and by mortality
data from the National Statistical Office of Korea. The
diagnostic codes of the International Classification of
Diseases, 10th edition, were used to classify death due to
cancer or other causes. Person-years (PY) were counted
from enrolment to the date of death or the end of
follow-up. Patient survival was categorised as overall
survival (OS), disease-specific survival or recurrence-free
survival.
2.4. Statistical analysis
Pearson’s c2 test for categorical variables and the Stu-
dent t test for unpaired data for continuous variables
were performed to compare clinicopathological charac-
teristics among the three groups. A P value of <0.05 was
regarded as significant. Survival rates were calculated by
the Kaplan-Meier method, with the date of gastrectomy
as the starting point. Patients who were alive were
 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137 131

censored at the time of the last follow-up. Differences in Table 1

survival were examined by the log-rank test. Multivari- Comparison of baseline characteristics of patients with BMI infor-
mation only at baseline and those with BMI information at both
able analysis was performed using a Cox proportional
baseline and 1 year after surgery.
hazards model with a backward stepwise selection pro-
Measures Total patients BMI at baseline and
cedure. All analyses were performed by SAS for Win- (N Z 1905) 1 year after surgery
dows software (version 8.02, SAS Institute, Cary, NC, (n Z 1418)
United States of America). Age (years)
Mean  SD 58.3  11.8 57.6  11.5
3. Results Range 20e88 23e88
Male 1294 (67.9%) 952 (67.1%)
Type of surgery
3.1. Characteristics of the study population Total gastrectomy 550 (28.9%) 392 (27.6%)
Distal subtotal gastrectomy 1354 (71.1%) 1026 (72.4%)
In total, 1905 patients were followed for 11,085.3 PYs. Proximal gastrectomy 1 (0.1%) 0 (0.0%)
The mean follow-up period was 5.8  3.2 years. BMI T stage
data 1 year after surgery could only be identified in 1418 T0 5 (0.3%) 4 (0.3%)
T1 951 (49.9%) 744 (52.5%)
patients (74.4%). The remaining patients had died within T2 562 (29.5%) 425 (30.0%)
1 year after surgery (n Z 124) or there was no record of T3 337 (17.7%) 218 (15.4%)
follow-up body weight (n Z 363). The percentages of T4 50 (2.6%) 27 (1.9%)
overweight and obese patients at the time of surgery in N stage
our study were 21.3% and 2.2%, respectively. One year N0 1137 (59.7%) 884 (62.3%)
N1 417 (21.9%) 315 (22.2%)
after gastrectomy, the proportion of patients with BMI N2 187 (9.8%) 134 (9.4%)

25 was 6.9%, and only two of these patients were N3 164 (8.6%) 85 (6.0%)
obese. Pathological stage
The baseline characteristics of the enrolled patients I 1175 (61.7%) 918 (64.7%)
are shown in Table 1. The mean age was 58.3 years II 274 (14.4%) 204 (14.4%)
III 239 (12.5%) 183 (12.9%)
(range, 20e88 years). Males comprised 67.9% IV 217 (11.4%) 113 (8.0%)
(n Z 1294) of the patients. Distal subtotal gastrectomy Histology
was the most commonly performed procedure (71.1%, Papillary adenocarcinoma 13 (0.7%) 6 (0.4%)
n Z 1354). Stage I tumours comprised 61.7% Tubular adenocarcinoma
(n Z 1175). When the patients with BMI information Well differentiated 247 (13.0%) 193 (13.6%)
Moderately differentiated 577 (30.3%) 428 (30.2%)
for both time points of this study were compared with Poorly differentiated 665 (34.9%) 478 (33.7%)
the total patients, there was no significant difference Mucinous adenocarcinoma 51 (2.7%) 36 (2.5%)
between these groups. Postoperative complications Signet ring cell carcinoma 347 (18.2%) 275 (19.4%)
developed in 167 cases (8.8%) among all included pa- NA 5 (0.3%) 2 (0.1%)
tients (Supplementary Table 1). Mean postoperative Preoperative BMI (kg/m2)
Low (<18.5) 121 (6.4%) 88 (6.2%)
hospital stay and the rate of postoperative complica- Normal (18.5e24.9) 1339 (70.3%) 997 (70.3%)
tions and in-hospital death were not significantly High (
25) 445 (23.4%) 333 (23.5%)
different among the three groups. Where appropriate, data are shown as the mean  SD.
Table 2 shows the clinicopathological features of the BMI, body mass index; NA, not applicable; SD, standard deviation.
three groups of patients with BMI information recorded
1 year after surgery. The high BMI group tended to have
high proportions of subtotal gastrectomy, R0 resection, rates of the low, normal, and high BMI groups were
early stage disease and differentiated histology. 67.5%, 83.6%, and 93.6%, respectively.
Fig. 1C and 1D show postoperative disease-specific
3.2. Long-term survival of the three BMI groups survival curves for the three BMI groups. In the
Kaplan-Meier curve for BMI at baseline, the 5-year
Fig. 1A and 1B show postoperative OS curves for the disease-specific survival rate was significantly greater in
three BMI groups. In the Kaplan-Meier curve for BMI at the high BMI group than the low (87.7% versus 74.6%;
baseline, the 5-year OS rate was significantly greater in P Z 0.001) and normal BMI groups (87.7% versus
the high BMI group than the low (84.7% versus 69.1%; 78.5%; P < 0.001) (Fig. 1C). In the Kaplan-Meier curve
P < 0.001) and normal BMI groups (84.7% versus 74.2%; for BMI 1 year after surgery, the high BMI group
P < 0.001) (Fig. 1A). There was no statistical difference showed a greater 5-year disease-specific survival than
in the OS rate between the normal and low BMI groups that of the low BMI group (93.6% versus 72.0%;
(P Z 0.155). In the Kaplan-Meier curve for BMI 1 year P < 0.001) (Fig. 1D). However, no significant difference
after surgery, there were significant differences in the OS was observed between the high and normal BMI groups
rate among the three groups (Fig. 1B). The 5-year OS (93.6% versus 86.7%; P Z 0.072).
132 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137

Table 2 BMI 1 year after surgery. Compared with patients with

Clinicopathological features of low (BMI <18.50 kg/m2), normal (BMI a normal BMI, the overall mortality risk was lower in
18.5e24.9 kg/m2), and high BMI (BMI
25.0 kg/m2) patients 1 year
the high BMI group (adjusted hazard ratio Z 0.51;
after surgery.
95% confidence interval Z 0.26e0.98).
BMI at 1 year Low Normal High P
after surgery BMI BMI BMI The BMI 1 year after surgery was not independently
associated with either disease-specific survival or
Number of 303 (21.4%) 1017 (71.7%) 98 (6.9%)
patients (%) recurrence-free survival after adjusting for patient age,
Age (mean  SD) 57.9  13.3 57.5  11.0 57.1  10.5 0.811 sex, type of surgery, tumour stage, histology, curative
Male gender 166 (54.8%) 719 (70.7%) 67 (68.4%) <0.001 resection, and BMI at baseline (data not shown).
BMI at baseline 20.1  2.1 23.4  2.4 28.1  2.9 <0.001
(mean  SD) 3.4. Survival analysis according to other criteria
Type of surgery
Total 123 (40.6%) 259 (25.5%) 10 (10.2%) <0.001
gastrectomy Fig. 2 shows the postoperative OS curves of patients
Subtotal 180 (59.4%) 758 (74.5%) 88 (89.8%) grouped by percentage weight loss after gastrectomy.
gastrectomy Among the patients who survived more than 1 year after
Curative resection gastrectomy, patients with >10% weight loss had the
R0 265 (87.5%) 970 (95.4%) 96 (98.0%) <0.001
R1 or R2 38 (12.5%) 47 (4.6%) 2 (2.0%)
worst survival compared with the other patients.
Pathological In a subgroup analysis, focussing on the patients with
stage a high-normal BMI (23.00e24.99 kg/m2) at baseline, a
I 149 (49.2%) 692 (68.0%) 77 (78.6%) <0.001 higher BMI 1 year after surgery was associated with a
II 54 (17.8%) 137 (13.5%) 13 (13.3%) longer survival. However, there was no statistically sig-
III 54 (17.8%) 125 (12.3%) 4 (4.1%)
IV 46 (15.2%) 63 (6.2%) 4 (4.1%)
nificant difference between the high and normal BMI
Histology groups (Fig. 3).
Papillary 1 (0.3%) 5 (0.5%) 0 (0.0%) 0.003 To examine the relationship between the change in
adenocarcinoma BMI status and OS, we made another grouping
Tubular including increased, stable, and decreased BMI group
adenocarcinoma
Well 26 (8.6%) 150 (14.7%) 17 (17.3%)
(Fig. 4). The increased BMI group shows the best sur-
differentiated vival compared to the other patients.
Moderately 72 (23.8%) 319 (31.4%) 37 (37.8%)
differentiated 4. Discussion
Poorly 131 (43.2%) 320 (31.5%) 27 (27.6%)
differentiated
Mucinous 11 (3.6%) 22 (2.2%) 3 (3.1%) This study was designed to evaluate the relationship
adenocarcinoma between BMI and long-term survival of gastric cancer
Signet ring cell 62 (20.5%) 199 (19.6%) 14 (14.3%) patients after gastrectomy. We compared the survival
carcinoma impact of BMI level at two different time points: pre-
NA 0 (0.0%) 2 (0.2%) 0 (0.0%)
operative and 1 year after surgery. Based on the retro-
Where appropriate, data are shown as the mean  SD. spective analysis of 9 years of experience at a single
BMI, body mass index; NA, not applicable; SD, standard deviation.
tertiary institution, we demonstrated a positive associ-
ation between BMI and survival following gastrectomy
Regarding recurrence-free survival, as with the OS for gastric cancer. BMI 1 year after gastrectomy was an
and disease-specific survival, the postoperative high independent prognostic factor after adjustment for age,
BMI group showed the best recurrence-free survival sex, tumour histology, curative resection, and cancer
compared with the other two groups, although there stage. However, the preoperative BMI was not a sig-
were no statistical differences among the three BMI nificant factor. Overweight and obese patients 1 year
groups at baseline (Fig. 1E and 1F). after gastrectomy had improved survival compared with
underweight and even normal weight patients. To the
3.3. Multivariate analysis best of our knowledge, this is the first study to investi-
gate the association between postoperative BMI and
In a Cox proportional hazards model, the BMI 1 year long-term survival of gastric cancer patients undergoing
after surgery was significantly associated with gastrectomy.
improved OS after adjusting for patient age, sex, type Although increased BMI is associated with increased
of surgery, tumour stage, histology, curative resection, risk of various cancers,[15] few studies have addressed
and BMI at baseline (Table 3). Hazard ratio estimates the relationship between overweight status and survival
greater than one indicate that higher mortality was following cancer diagnosis. In women with breast can-
associated with patients aged 65 and over, increasing cer, obesity at cancer diagnosis is related to poor
tumour stage, non-curative resection, and decreasing outcome, with increasing evidence of weight gain, even
 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137 133

Fig. 1. Survival data of the low (BMI <18.5 kg/m2), normal (BMI, 18.5e24.9 kg/m2), and high BMI (BMI
25.0 kg/m2) groups in gastric
cancer patients undergoing gastrectomy. Overall survival of the patients grouped by (A) BMI at baseline and (B) BMI 1 year after surgery.
Disease-specific survival of the patients grouped by (C) BMI at baseline and (D) BMI 1 year after surgery. Recurrence-free survival of the
patients grouped by (E) BMI at baseline and (F) BMI 1 year after surgery. BMI, body mass index.
134 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137

Table 3 after diagnosis, exacerbating the risk of cancer recur-

Multivariate analysis showing independence of the effect on overall rence and mortality.[16] In the case of prostate cancer,
survival
obese men have higher rates of recurrence after prosta-
Hazard ratio 95% CI P
tectomy and death from prostate cancer, whereas
Age (years) obesity is not associated with overall incidence.[17]
<65 1
Among gastrointestinal malignancies, several studies

65 1.73 1.37e2.17 <0.001
Gender have focused on the effect of obesity on colon cancer
Male 1 survival, but the results are inconclusive.[18,19]
Female 0.92 0.71e1.18 0.501 The relationship between obesity and gastric cancer is
Type of surgery also uncertain. From the meta analysis of several pro-
Total gastrectomy 1
spective observational studies, increased BMI is not
Subtotal gastrectomy 0.97 0.77e1.23 0.804
Pathological stage associated with an increased incidence of gastric cancer
I 1 in men or women.[15] However, some studies investi-
II 3.21 2.28e4.53 <0.001 gating the relationship between obesity and prognosis in
III 7.98 5.92e10.77 <0.001 gastric cancer patients undergoing gastrectomy showed
IV 12.06 8.03e18.11 <0.001
mixed results.[9e11]
Histology
Differentiated 1 The current study showed that a preoperative
Undifferentiated 1.13 0.89e1.44 0.329 BMI > 25 kg/m2 was not an independent prognostic
Curative resection factor, which is consistent with the findings presented in
R0 1 two recent reports [10,11]. However, a BMI > 25 kg/m2
R1 or R2 2.10 1.44e3.07 <0.001
1 year after surgery was an independent favourable
BMI at baseline
Low 1.01 0.72e1.40 0.974 prognostic factor after adjustment for well-known
Normal 1 prognostic factors, such as patient age or tumour
High 0.64 0.41e1.02 0.058 stage. This result is supported by the finding that the
BMI at 1 year after surgery greater the weight loss (
10%), the poorer the OS was in
Low 1.38 1.09e1.76 0.008
the present study. In general, a body weight loss of
Normal 1
High 0.51 0.26e0.98 0.046 5e15% was found after gastrectomy.[20,21] In addition,
the tendency that being overweight is associated with
BMI, body mass index; CI, confidence interval.
good prognosis was observed even among subgroups
with the same preoperative BMI.
The mechanism of the beneficial effect of high BMI is
uncertain. One possible explanation was suggested by a

Fig. 2. Overall survival of patients grouped by percentage weight Fig. 3. Overall survival of patients with high-normal BMI
loss 1 year after surgery. (23.0e24.9 kg/m2) at baseline. BMI, body mass index.
 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137
Fig. 4. Overall survival of patients grouped by change in BMI status. (A) Definition and number of patients of groups by change in BMI
status. (B) Kaplan-Meier curve showing the overall survival of three groups by change in BMI status. BMI, body mass index.
previous study based on health check-ups of the popu-
lation.[22] In this study, the optimal BMI tended to be
higher for the elderly, which possibly indicates that older
adults need nutritional reserves for the protective effects
of being overweight. Similarly, gastric cancer patients
have more nutritional needs because of cachexia from
cancer itself and the impact of gastrectomy.
Hormonal changes after gastrectomy can also explain
the current observation. The stomach produces an
important orexigenic hormone, ghrelin,[23] which se-
cretes growth hormones, increases food intake, and
causes weight gain.[24,25] Patients undergoing total or
partial gastrectomy commonly lose body weight,
accompanying impaired ghrelin production.[26,27]
Furthermore, administration of synthetic ghrelin less-
ened postoperative body weight loss and improved
appetite and food intake after total gastrectomy in pa-
tients with gastric cancer.[28] These results suggest that
ghrelin replacement can lead to weight gain in patients
with gastric cancer undergoing gastrectomy and may
lengthen survival.
Nutritional improvement by dietary counselling is
one intervention used for achieving weight gain and
prolonging survival. Previous studies have revealed that
various methods of nutritional support are effective for
improving dietary intake and outcomes in patients who
lost weight after receiving chemotherapy or radio-
therapy.[29,30] In a prospective randomised controlled
trial for gastric cancer patients undergoing gastrectomy,
less of a decrease in BMI occurred in the group with
patient participation-based nutritional interven-
tion.[31,32] In patients undergoing gastrectomy,
decreased caloric intake is quite frequent due to
135
oesophagitis or dumping syndrome. In addition, early
satiety and limited absorption of essential nutrients like
iron, calcium, and some vitamins are observed. There-
fore, nutritional intervention may be necessary to
improve the prognosis of these patients. Various inter-
vention methods including dietary counselling, nutri-
tional advice, and the prescription of an oral nutritional
supplement could be tried to achieve nutritional sup-
port.[33] In patients undergoing gastrectomy, there is
broad acceptance that a low baseline BMI and total
gastrectomy are indicative of serious weight loss in the
following months. As shown in Fig. 4, the increased or
stable BMI after surgery conferred better patient sur-
vival than that in the decreased group. Therefore, the
post-gastrectomy patients with weight loss should be
supported with nutritional supplementation. Dietary
advice and nutritional supplements in the management
of weight loss can be beneficial to weight gain, with in-
terventions providing as little as an additional 250e300
calories per day.[34] Although the optimal time to start
nutritional intervention is not known, early intervention
after surgery would be advisable.
This study has several strengths. First, not only OS
but also disease-specific survival and recurrence-free
survival were analysed in a large population. In addi-
tion, postoperative BMI was analysed to determine its
relationship with survival of gastric cancer patients
undergoing gastrectomy. Moreover, laboratory pa-
rameters related to nutritional status were evaluated
within a certain time period. A limitation of this study
is that it was based on an Asian population whose
obesity rate is much lower than that of Western
populations.
136 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137
In conclusion, BMI is a prognostic factor for long-
term OS in patients with gastric cancer undergoing
gastrectomy. High BMI 1 year after gastrectomy is a
more favourable prognostic factor than is preoperative
BMI, which provides additional prognostic information
due to its independence from established prognostic
factors, such as age or tumour stage. Aggressive nutri-
tional support should be considered in these patients to
improve their outcomes.
Conflicts of interest statement
None.
Acknowledgements
This research was supported by program of Global
Research and Development Center through the Na-
tional Research Foundation of Korea (NRF) funded
by the Ministry of Science, ICT and Future Planning
(NRF-2011-0031644), by Basic Science Research
Program through the NRF funded by the Ministry
of Education, Science and Technology (NRF-
2013R1A1A2007985), and by the Catholic Medical
Center Research Foundation made in the program
year of 2014.
Appendix A. Supplementary data
Supplementary data related to this article can be found
at http://dx.doi.org/10.1016/j.ejca.2015.10.061.
References
[1] Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint AJ,
Hannan L, MacInnis RJ, et al. Body-mass index and mortality
among 1.46 million white adults. N Engl J Med 2010;363(23):
2211e9.
[2] Janssen I, Heymsfield SB, Allison DB, Kotler DP, Ross R. Body
mass index and waist circumference independently contribute to
the prediction of nonabdominal, abdominal subcutaneous, and
visceral fat. Am J Clin Nutr 2002;75(4):683e8.
[3] Appropriate body-mass index for Asian populations and its im-
plications for policy and intervention strategies. Lancet 2004;
363(9403):157e63.
[4] O’Brien Jr JM, Phillips GS, Ali NA, Lucarelli M, Marsh CB,
Lemeshow S. Body mass index is independently associated with
hospital mortality in mechanically ventilated adults with acute
lung injury. Crit Care Med 2006;34(3):738e44.
[5] Fonarow GC, Srikanthan P, Costanzo MR, Cintron GB,
Lopatin M. An obesity paradox in acute heart failure: analysis of
body mass index and inhospital mortality for 108,927 patients in
the acute decompensated heart failure national registry. Am Heart
J 2007;153(1):74e81.
[6] Kalantar-Zadeh K, Abbott KC, Salahudeen AK, Kilpatrick RD,
Horwich TB. Survival advantages of obesity in dialysis patients.
Am J Clin Nutr 2005;81(3):543e54.
[7] Guenette JA, Jensen D, O’Donnell DE. Respiratory function and
the obesity paradox. Curr Opin Clin Nutr Metab Care 2010;13(6):
618e24.
[8] Kalantar-Zadeh K, Horwich TB, Oreopoulos A, Kovesdy CP,
Younessi H, Anker SD, et al. Risk factor paradox in wasting
diseases. Curr Opin Clin Nutr Metab Care 2007;10(4):433e42.
[9] Tokunaga M, Hiki N, Fukunaga T, Ohyama S, Yamaguchi T,
Nakajima T. Better 5-year survival rate following curative gas-
trectomy in overweight patients. Ann Surg Oncol 2009;16(12):
3245e51.
[10] Kulig J, Sierzega M, Kolodziejczyk P, Dadan J, Drews M,
Fraczek M, et al. Implications of overweight in gastric cancer: a
multicenter study in a western patient population. Eur J Surg
Oncol 2010;36(10):969e76.
[11] Bickenbach KA, Denton B, Gonen M, Brennan MF, Coit DG,
Strong VE. Impact of obesity on perioperative complications and
long-term survival of patients with gastric cancer. Ann Surg
Oncol 2013;20(3):780e7.
[12] Lee YH, Han SJ, Kim HC, Hyung WJ, Lim JS, Lee K, et al.
Gastrectomy for early gastric cancer is associated with decreased
cardiovascular mortality in association with postsurgical meta-
bolic changes. Ann Surg Oncol 2013;20(4):1250e7.
[13] Sobin LH, Wittekind CH. International union against cancer
(UICC). TNM classification of malignant tumors. 5th ed. New
York: Wiley; 1997. p. 59e62.
[14] Japanese Gastric Cancer A. Japanese classification of gastric
carcinoma - 2nd English edition. Gastric Cancer 1998;1(1):10e24.
[15] Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M. Body-
mass index and incidence of cancer: a systematic review and meta-
analysis of prospective observational studies. The Lancet 2008;
371(9612):569e78.
[16] Kroenke CH, Chen WY, Rosner B, Holmes MD. Weight, weight
gain, and survival after breast cancer diagnosis. J Clin Oncol
2005;23(7):1370e8.
[17] Freedland SJ, Platz EA. Obesity and prostate cancer: making
sense out of apparently conflicting data. Epidemiol Rev 2007;29:
88e97.
[18] Meyerhardt JA, Catalano PJ, Haller DG, Mayer RJ,
Benson 3rd AB, Macdonald JS, et al. Influence of body mass
index on outcomes and treatment-related toxicity in patients with
colon carcinoma. Cancer 2003;98(3):484e95.
[19] Sinicrope FA, Foster NR, Sargent DJ, O’Connell MJ, Rankin C.
Obesity is an independent prognostic variable in colon cancer
survivors. Clin Cancer Res 2010;16(6):1884e93.
[20] Fein M, Fuchs KH, Thalheimer A, Freys SM, Heimbucher J,
Thiede A. Long-term benefits of Roux-en-Y pouch reconstruction
after total gastrectomy: a randomized trial. Ann Surg 2008;247(5):
759e65.
[21] Kiyama T, Mizutani T, Okuda T, Fujita I, Tokunaga A, Tajiri T,
et al. Postoperative changes in body composition after gastrec-
tomy. J Gastrointest Surg 2005;9(3):313e9.
[22] Matsuo T, Sairenchi T, Iso H, Irie F, Tanaka K, Fukasawa N,
et al. Age- and gender-specific BMI in terms of the lowest mor-
tality in Japanese general population. Obes (Silver Spring) 2008;
16(10):2348e55.
[23] Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H,
Kangawa K. Ghrelin is a growth-hormone-releasing acylated
peptide from stomach. Nature 1999;402(6762):656e60.
[24] Takaya K, Ariyasu H, Kanamoto N, Iwakura H, Yoshimoto A,
Harada M, et al. Ghrelin strongly stimulates growth hormone
release in humans. J Clin Endocrinol Metab 2000;85(12):4908e11.
[25] Nakazato M, Murakami N, Date Y, Kojima M, Matsuo H,
Kangawa K, et al. A role for ghrelin in the central regulation of
feeding. Nature 2001;409(6817):194e8.
[26] Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K,
Akamizu T, et al. Stomach is a major source of circulating
ghrelin, and feeding state determines plasma ghrelin-like immu-
noreactivity levels in humans. J Clin Endocrinol Metab 2001;
86(10):4753e8.
[27] Karamanakos SN, Vagenas K, Kalfarentzos F, Alexandrides TK.
Weight loss, appetite suppression, and changes in fasting and
 H.H. Lee et al. / European Journal of Cancer 52 (2016) 129e137
postprandial ghrelin and peptide-YY levels after Roux-en-Y
gastric bypass and sleeve gastrectomy: a prospective, double
blind study. Ann Surg 2008;247(3):401e7.
[28] Adachi S, Takiguchi S, Okada K, Yamamoto K, Yamasaki M,
Miyata H, et al. Effects of ghrelin administration after total
gastrectomy: a prospective, randomized, placebo-controlled phase
II study. Gastroenterology 2010;138(4):1312e20.
[29] Bourdel-Marchasson I, Blanc-Bisson C, Doussau A, Germain C,
Blanc JF, Dauba J, et al. Nutritional advice in older patients at
risk of malnutrition during treatment for chemotherapy: a two-
year randomized controlled trial. PLoS One 2014;9(9):e108687.
[30] Ravasco P, Monteiro-Grillo I, Vidal PM, Camilo ME. Dietary
counseling improves patient outcomes: a prospective, random-
ized, controlled trial in colorectal cancer patients undergoing
radiotherapy. J Clin Oncol 2005;23(7):1431e8.
137
[31] Wu MH, Lin MT, Chen WJ. Effect of perioperative parenteral
nutritional support for gastric cancer patients undergoing gas-
trectomy. Hepatogastroenterology 2008;55(82e83):799e802.
[32] Kim H, Suh EE, Lee HJ, Yang HK. The effects of patient
participation-based dietary intervention on nutritional and func-
tional status for patients with gastrectomy: a randomized
controlled trial. Cancer Nurs 2014;37(2):E10e20.
[33] Baldwin C, Spiro A, Ahern R, Emery PW. Oral nutritional in-
terventions in malnourished patients with cancer: a systematic
review and meta-analysis. J Natl Cancer Inst 2012;104(5):371e85.
[34] Baldwin C, Weekes CE. Dietary advice for illness-related
malnutrition in adults. Cochrane Database Syst Rev 2008;1:
Cd002008.