2016 - Cigarette Smoking and Gasstric CA in The Stomach CA Pooling Project
2016 - Cigarette Smoking and Gasstric CA in The Stomach CA Pooling Project
2016 - Cigarette Smoking and Gasstric CA in The Stomach CA Pooling Project
There may be differences between this version and the published version. You are
advised to consult the publisher’s version if you wish to cite from it.
http://eprints.gla.ac.uk/123919/
Delphine Praud 1, Matteo Rota 2, Claudio Pelucchi 2, Paola Bertuccio 1, Tiziana Rosso 1,
Carlotta Galeone 2, Zuo-Feng Zhang 3, Keitaro Matsuo 4, Hidemi Ito 5, Jinfu Hu 6, Kenneth C.
Zaridze 15, Dmitry Maximovitch 15, Nuria Aragonés 16,17,18, Gemma Castaño-Vinyals 19,20,21,17,
17,22 17,22 13,23,24
Jesus Vioque , Eva M. Navarrete-Muñoz , Mohammadreza Pakseresht , Farhad
Pourfarzi 13,25, Alicja Wolk 26, Nicola Orsini 26, Andrea Bellavia 26, Niclas Håkansson 26, Lina
Mu 27, Roberta Pastorino 28, Robert C. Kurtz 29, Mohammad H. Derakhshan 13,30, Areti Lagiou
31
, Pagona Lagiou 32,33, Paolo Boffetta 34, Stefania Boccia 35, Eva Negri 2, Carlo La Vecchia 1
Italy.
Milan, Italy.
4. Division of Molecular Medicine, Aichi Cancer Center Research Institute, Nagoya, Japan.
Nagoya, Japan.
Ontario, Canada
9. Molecular and Nutritional Epidemiology Unit, Cancer Research and Prevention Institute-
1
10. Penn State College of Medicine, Hershey, PA, USA
11. EPIUnit - Institute of Public Health, University of Porto (ISPUP), Porto, Portugal.
12. Department of Clinical Epidemiology, Predictive Medicine and Public Health, University
Sweden
15. Department of Epidemiology and Prevention, Russian N.N. Blokhin Cancer Research
18. Cancer Epidemiology Research Group, Oncology and Hematology Area, IIS Puerta de
Spain
20. IMIM (Hospital del Mar Medical Research Institute), Barcelona, Spain
22. Department of Public Health, Miguel Hernandez University, Campus San Juan, Alicante,
Spain
23. Nutritional Epidemiology Group, Centre for Epidemiology and Biostatistics, University
of Leeds, Leeds, UK
Edmonton, Canada
2
25. Department of Community Medicine, Ardabil University of Medical Sciences, Ardabil,
Iran
27. Department of Social and Preventive Medicine, School of Public Health and Health
28. Section of Hygiene - Institute of Public Health; Università Cattolica del Sacro Cuore,
Rome, Italy.
29. Department of Medicine, Memorial Sloan Kettering Cancer Centre, New York, NY, USA
31. Department of Public Health and Community Health, School of Health Professions,
33. Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA,
USA
34. The Tisch Cancer Institute, Mount Sinai School of Medicine, New York, NY, USA.
35. Section of Hygiene - Institute of Public Health; Università Cattolica del Sacro Cuore,
Department of Epidemiology
Funding: This project was supported by the Italian Ministry of Health (Young Researchers,
by the Fondazione Italiana per la Ricerca sul Cancro (FIRC) and by the Associazione Italiana
per la Ricerca sul Cancro, Project no. 16715 (Investigator Grant). MR and DP were supported
by a fellowship from the FIRC. BP was supported by an individual grant from the “Fundação
4
Abstract
Tobacco smoking is a known cause of gastric cancer, but several aspects of the association
remain imprecisely quantified. We examined the relation between cigarette smoking and
gastric cancer risk using a uniquely large dataset of 23 epidemiological studies within the
“Stomach cancer Pooling (StoP) Project”, including 10,290 cases and 26,145 controls. We
estimated summary odds-ratios (ORs) and the corresponding 95% confidence intervals (CIs)
by pooling study-specific ORs using random-effects models. Compared to never smokers, the
ORs were 1.20 (95% CI 1.09-1.32) for ever, 1.12 (95% CI 0.99-1.27) for former and 1.25
(95% CI 1.11-1.40) for current cigarette smokers. Among current smokers, the risk increased
with number of cigarettes per day to reach an OR of 1.32 (95% CI 1.10-1.58) for smokers of
more than 20 cigarettes per day. The risk increased with duration of smoking, to reach an OR
of 1.33 (95% CI 1.14-1.54) for more than 40 years of smoking, and decreased with increasing
time since stopping cigarette smoking (p for trend<0.01) and became similar to that of never
smokers 10 years after stopping. Risks were somewhat higher for cardia than non-cardia
gastric cancer. Risks were similar when considering only studies with information on
Helicobacter pylori (HP) infection, and comparing all cases to HP+ controls only. This study
provides the most precise estimate of the detrimental effect of cigarette smoking on gastric
cancer risk based on individual data, including the relationship with dose and duration, and
5
Introduction
Tobacco smoking has been causally linked to gastric cancer (IARC Working Group,
(Ladeiras-Lopes et al., 2008) and 46 case-control studies (La Torre et al., 2009) showed a
significant increase in risk of gastric cancer among smokers, though quantification of dose-
Risk of gastric cancer was reported to increase with increasing dose and duration of
cigarette smoking in some studies (Gonzalez et al., 2003; Ladeiras-Lopes et al., 2008;
Tramacere et al., 2011; Nomura et al., 2012). Furthermore, some studies showed a stronger
association between cigarette smoking and gastric cardia rather than non-cardia cancer
(Gonzalez et al., 2003; Freedman et al., 2007; Nomura et al., 2012), but others did not
Lower risks have been generally found in former compared to current smokers, and the
risk seems to decrease with increasing years since stopping smoking, although such
relationship was not significant in several studies (Gonzalez et al., 2003; Koizumi et al.,
2004; Freedman et al., 2007; Kim et al., 2007; Zendehdel et al., 2008; IARC Working Group,
2012).
To better define and quantify the association between cigarette smoking and gastric
6
Methods
This analysis is based on data from 23 case-control studies included in the first release
of the “StoP Project” dataset, including 10,290 cases (6804 men, 3486 women) and 26,145
controls (15,600 men, 10,545 women) from Greece (Lagiou et al., 2004), Italy (4 studies)
(Buiatti et al., 1989; La Vecchia et al., 1995; Lucenteforte et al., 2008; De Feo et al., 2012),
Portugal (Lunet et al., 2007), Russia (Zaridze et al., 2000), Spain (2 studies) (Santibanez et
al., 2012; Castano-Vinyals et al., 2015), Sweden (3 studies, 2 of which were nested in cohort
studies) (Ye et al., 1999; Harris et al., 2013), China (4 studies) (Setiawan et al., 2000; Mu et
al., 2005; Setiawan et al., 2005; Deandrea et al., 2010), Iran (3 studies) (Derakhshan et al.,
2008; Pourfarzi et al., 2009; Pakseresht et al., 2011), Japan (Matsuo et al., 2013), Canada
(Mao et al., 2002), and USA (2 studies) (Zhang et al., 1999). Detailed information on the
aims and methods of the “StoP Project” has been given elsewhere (Pelucchi et al., 2015).
about cigarette smoking status (never, former, and current smoker), number of cigarettes
smoked per day, duration of smoking, and time since stopping smoking, when applicable.
Data were harmonized according to a pre-specified format and completeness and consistency
between variables were carefully checked. For the present analysis, ever cigarette smokers
were defined as participants who had smoked at least 100 cigarettes in their lifetime or more
than one cigarette per day for at least 1 year, independently of the type of cigarette smoked
(e.g., cigarettes with or without filter, with blond or black tobacco, hand-rolled cigarettes,
etc.). Smoked forms of tobacco other than cigarettes, i.e. cigar and pipes, were considered in
a separate category. Two studies (Setiawan et al., 2000; Derakhshan et al., 2008) providing
information on lifetime cigarette status (ever, never) only were not considered in the dose-risk
7
For two cohort studies included in the StoP Project consortium, the Swedish
Mammography Cohort and the Cohort of Swedish Men (Harris et al., 2013), a nested case-
control design was used by selecting 4 controls for each case, matched on age.
To estimate the association between cigarette smoking and gastric cancer, we used a
two-stage modeling approach (Smith-Warner et al., 2006). In the first stage, we assessed the
association between cigarette smoking and gastric cancer for each study by estimating the
odds ratios (ORs) and the corresponding 95% CIs using multivariable unconditional logistic
regression models (for categorical variables). These models included, when available and
appropriate, terms for age (<40, 40-44, 45-49, 50–54, 55–59, 60–64, 65–69, 70–74, ≥75
(White, Hispanic/Latino, Black/African American, other), alcohol drinking (never, low ≤12
gr/day, intermediate >12-≤47 gr/day, high >47 gr/day), fruit and vegetables consumption
(study-specific tertiles) and study center (for multicentric studies). The list of study-specific
In the second stage, summary (pooled) effects estimates were computed using a
obtained in the first stage, using as weights the inverse of the sum of the study-specific
log(ORs) variances and the between-study variance components (DerSimonian and Laird,
1986).
For categorical variables, heterogeneity between studies was evaluated using the Q test
statistics and quantified using I2, i.e. the proportion of total variation contributed by between-
selected covariates, we conducted analyses stratified by age (<55, >55-<65, >65), sex,
8
geographic area (Europe, Asia, America), cancer site (cardia, non cardia), cancer histotype
controls; controls from 2 nested case-control studies were considered together with the latter).
We also evaluated the influence of H.pylori infection on the relation between cigarette
smoking and gastric cancer. We considered only studies with the information on H.pylori
infection and compared the pooled ORs obtained using all controls with that obtained using
only H.pylori positive controls (assuming all gastric cancer cases are positive for H.pylori
infection.
We tested for the significance of linear trends across levels of smoking intensity and
duration variables by estimating study-specific trends and using the Wald test P value
For continuous variables, we studied the functional form of the relation using one-
order and two-order fractional polynomial models. The method was based on a two-stage
procedure. In a first step, we fitted first-order and second-order fractional polynomial models
to each study adjusting for the aforementioned confounders. This family of models includes
the linear one. In the second step, the pooled dose-risk relation was estimated through a
bivariate random effects model (Rota et al., 2010). The best fitting model, i.e. the one
minimizing the model deviance, was selected when the best fitting model was non linear
Results
Table 1 shows the distribution of cases and controls by study, sex, age and major selected
potential confounding factors, for a descriptive purpose only. Cases were older than controls
9
(mean age ± SD: 63 ± 11 vs. 60 ± 13, respectively) and had a lower social class. Overall,
8.6% of cases reported a history of stomach cancer in first degree relatives and 11.2%
ethanol).
Figure 1 gives a forest plot of the study-specific and the pooled ORs for gastric cancer
risk on the basis of all 23 studies participating to the consortium. The pooled estimate was
The pooled ORs of gastric cancer according to cigarette smoking habits are given in
Table 2. Among 21 studies reporting information on former smoking status, the pooled ORs
were 1.19 (95% CI 1.09-1.31) for ever cigarette smokers, 1.12 (95% CI 0.99-1.27) for former
cigarette smokers and 1.25 (95% CI 1.11-1.40) for current smokers, compared with never
smokers. Among current smokers, the risk increased with number of cigarettes smoked per
day. Compared to never smokers, ORs were 1.08 (95% CI 0.91-1.28) for 1 to 10 cigarettes
per day, 1.30 (95% CI 1.16-1.45) for 11 to 20 cigarettes per day and 1.32 (95% CI 1.10-1.58)
for more than 20 cigarettes per day, with a significant trend in risk (p<0.01). The risk also
increased with increasing duration of smoking (p-value for trend<0.01) with ORs of 1.04
(95% CI 0.94-1.16) for up to 30 years of smoking, 1.32 (95% CI 1.17-1.49) for 31 to 40 years
of smoking and 1.33 (95% CI 1.14-1.54) for more than 40 years of smoking, as compared to
never smokers. A significant decreasing trend in risk was found with increasing time since
Study-specific and pooled ORs of gastric cancer for former and current smokers
studies was low to moderate (I2 between 19% and 56%) across categories of consumption.
10
Figure 3 gives the modeled relation between smoking intensity (Panel A) and duration
(Panel B) considered as continuous variables and gastric cancer risk. For both variables, the
best fitting model was the one with powers p1=-2 and p2=2, i.e. log(OR)= β1 X-2 + β2 X2. The
estimated regression coefficients were β1= -0.00002 and β2= 0.000023 for smoking intensity,
β1= -1.46E-06 and β2= 0.000136 for duration. Overall, duration appeared to have a somewhat
stronger effect on risk than intensity. These graphs suggest that the intensity- and duration-
Supplementary Table 2. The ORs for high intensity of cigarette smoking (i.e., >20 cigarettes
per day) were similar in men (OR=1.44, 95% CI 1.00-2.08) and women (OR=1.42, 95% CI
1.18-1.72), and higher in subjects aged 56-65 years (OR=1.64, 95% CI 1.33-2.03) and in
studies conducted in Asia (OR=1.71, 95% CI 1.08-2.71) than in the other groups. However,
none of these differences was statistically significant. Risks of gastric cardia cancer were
somewhat higher than those of non-cardia gastric cancer in former (ORs=1.30 and 1.05,
respectively), light (ORs=1.71 and 0.96, respectively) and moderate cigarette smokers
(ORs=1.55 and 1.21, respectively). For subjects reporting high intensity of cigarette smoking,
the ORs were 1.58 for cardia (95% CI 1.11-2.24) and 1.29 for non-cardia (95% CI 1.03-1.61)
gastric cancer, and when looking at histotype, 1.28 (95% CI 0.90-1.81) for diffuse-type and
1.57 (95% CI 1.12-2.20) for intestinal-type gastric cancer. Effects of cigarette smoking on
gastric cancer risk did not materially change when considering only studies with information
on H.pylori infection (>20 cigarettes per day, OR=1.40, 95% CI 0.91-2.15) even when the
analysis was restricted to H.pylori positive controls (>20 cigarettes per day, OR=1.48, 95% CI
0.93-2.36). Risks for smoking appeared to be somewhat higher in studies with hospital
controls (>20 cigarettes per day, OR=1.58, 95% CI 1.33-1.87) than in those with population
11
Stratified analyses on smoking duration according to sex, age, geographic area, cancer
site, cancer histotype, source of controls and H.pylori infection are given in Supplementary
Table 3. When considering long-term smokers (i.e., >40 years of smoking), risks appeared to
be higher in gastric cardia (OR=1.78, 95% CI 1.44-2.19) than in non-cardia (OR=1.19, 95%
CI 0.97-1.46) cancer cases, and in studies with hospital controls (OR=1.64, 95% CI 1.29-
Discussion
This global dataset confirms the association between cigarette smoking and gastric cancer
risk. This is the largest individual participant analysis on stomach neoplasms, including
comprehensive and uniform information on relevant covariates, and thus provides the most
precise and valid estimates of several quantitative aspects of the association. A 25% excess
risk of gastric cancer was found among current smokers. The risk significantly increased with
cigarette smoking intensity and duration, to reach 32% for smokers of more than 20 cigarettes
per day and 33% for smoking duration of more than 40 years, as compared to never smokers.
The increase in risk was steeper for duration than for dose. The risk declined with time since
stopping smoking and approached the level of never cigarette smokers about 10 years after
quitting.
These results are generally consistent with previous meta-analyses. Our OR estimates for
current cigarette smokers were, if anything, slightly lower than those based on published
studies, which found risks ranging between 1.5 and 1.7 (Tredaniel et al., 1997; Ladeiras-
Lopes et al., 2008; La Torre et al., 2009; Bonequi et al., 2013), resulting in an estimated
worldwide population attributable fraction of 19.5% in men and 3.0% in women (Peleteiro et
al., 2015). Publication bias may have led to an over estimation of the risk in the published
literature. In fact, only about one-third of the studies included in this pooled-analysis were
12
present in previous meta-analyses of case-control (La Torre et al., 2009) or cohort studies
Concerning the dose-risk relation for intensity and duration, previous results are limited.
risk with smoking intensity, with a relative risk (RR) varying from 1.3 for the lowest dose, to
1.7 for 30 cigarettes per day. A significant trend in gastric cancer risk with increasing duration
was reported in the European Investigation into Cancer and Nutrition (EPIC) (Gonzalez et al.,
2003) and in the Multiethnic Cohort (MEC) study (Nomura et al., 2012). Although the
association of gastric cancer with smoking has been long recognized, this study confirms that
the risk is modest even at high doses, in spite of the increasing dose-risk gradient. Although
the association with cigarette smoking is weaker for gastric cancer than for lung cancer, our
pooled analysis finds a stronger relation with duration than with dose, as it has been shown
An interesting finding of our study is the lower gastric cancer risk for former smokers as
compared to current smokers, in particular >10 years after stopping. This has been suggested
by previous studies (Gonzalez et al., 2003; Ladeiras-Lopes et al., 2008; Nomura et al., 2012)
and adds stomach cancer to the list of diseases whose risk is favorably affected by stopping
smoking.
This pooled analysis reported similar findings in men and women. With reference to the
site of cancer, two large cohort studies reported a higher risk of gastric cardia than non-cardia
cancer in current smokers (Gonzalez et al., 2003; Nomura et al., 2012). A recent meta-
analysis including 10 studies on gastric cardia adenocarcinoma reported an over two-fold risk
for smokers of more than 40 years compared to never smokers (Tramacere et al., 2011). A
differential role of smoking (besides H.pylori (Kamangar et al., 2006)) in the aetiology of
13
different gastric cancer subsites has also been hypothesized to contribute to the diverging
trends in risk of gastric cardia and non-cardia cancers over the last decades (Gonzalez et al.,
2003). In this investigation, though the ORs for high intensity cigarette smokers were not
statistically heterogeneous between subsites (i.e., 1.40 for gastric cardia and 1.29 for gastric
non-cardia cancer), the risk tended to be higher for gastric cardia and patterns tended to differ.
In fact, gastric cardia cancer risk was increased and higher than that for non-cardia subsite in
light to moderate smokers, in long-term smokers, and among former smokers. Thus, the dose-
risk and time-risk relations might be different between gastric cardia and non-cardia cancer,
Many mechanisms may explain the effect of tobacco smoking on gastric cancer
compounds, have been involved in the etiology of gastric cancer (Mirvish, 1995). In vitro, a
carcinogenic effect of tobacco smoke on the gastric mucosa has been reported (Tayler and
Piper, 1977). In a population-based gastroscopic screening study, smoking has also been
significantly associated with the development of precursor lesion of gastric cancer, i.e.
dysplasia, chronic atrophic gastritis and intestinal metaplasia, (Kneller et al., 1992). In gastric
cancer cases, levels of stable DNA adducts were significantly higher in the DNA of smokers
than in that of non smokers (Dyke et al., 1992). Some studies investigated gene interaction
between smoking and gastric cancer (Agudo et al., 2006; Lee et al., 2006; Boccia et al.,
gastric tumors from smokers rather than non-smokers (Poplawski et al., 2008).
Among the strengths of the study, the “StoP Project” included original and individual data
on smoking on over 10,000 cases and 26,000 controls, which provided a unique opportunity
14
to investigate and accurately quantify the dose- and time factors- risk relationships and,
among former smokers, the pattern of risk with time since stopping. The individual level
2013). For instance, we were able to investigate the confounding effect of H.pylori infection
by restricting the analysis to controls positive for H.pylori infection (all cases were supposed
to be H.pylori infected). This sensitivity analysis confirmed the results of the main analysis,
thus providing further evidence of a role of cigarette smoking independent from that of
H.pylori. Confounding from other specific factors, such as consumption of salted and smoked
foods, cannot entirely be ruled out. Also, we found a substantial heterogeneity between
studies that was not explained by age, sex and geographic area. Among potential explanations
for the reported heterogeneity is that the type of cigarettes commonly smoked (e.g., with or
without filter, with blond or black tobacco, hand-rolled, etc.) varies in different countries,
together with their variable tar and nicotine concentrations. We were not able to address the
role of such factors, however, since only a minority of studies had available information on
type of cigarette smoked. The association was, if anything, stronger in studies using hospital
controls. This is reassuring since it has been suggested that smokers may be over represented
among hospital controls given the higher hospitalization rates and longer hospital stays in
cigarette smoking on gastric cancer risk, provides the most valid and precise estimates of
dose-risk and duration-risk relations, and clearly shows a decrease in risk after stopping
Acknowledgments
15
The authors thank the European Cancer Prevention (ECP) Organization for providing support
16
References
metabolic genes related to tobacco smoke and the risk of gastric cancer in the
smoke and alcohol consumption and risk of gastric cancer: a case-control study in an
Bonequi P, Meneses-Gonzalez F, Correa P, Rabkin CS, Camargo MC (2013). Risk factors for
gastric cancer in Latin America: a meta-analysis. Cancer Causes Control 24: 217-
231.
study of gastric cancer and diet in Italy. Int J Cancer 44: 611-616.
effect modifier of the association between green tea intake and gastric cancer risk? Eur
17
Derakhshan MH, Malekzadeh R, Watabe H, Yazdanbod A, Fyfe V, Kazemi A, et al. (2008).
DerSimonian R, Laird N (1986). Meta-analysis in clinical trials. Control Clin Trials 7: 177-
188.
Dyke GW, Craven JL, Hall R, Garner RC (1992). Smoking-related DNA adducts in human
Freedman ND, Abnet CC, Leitzmann MF, Mouw T, Subar AF, Hollenbeck AR, et al. (2007).
A prospective study of tobacco, alcohol, and the risk of esophageal and gastric cancer
Gonzalez CA, Pera G, Agudo A, Palli D, Krogh V, Vineis P, et al. (2003). Smoking and the
risk of gastric cancer in the European Prospective Investigation Into Cancer and
Mammography Cohort and the Cohort of Swedish Men: Study design and
Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003). Measuring inconsistency in meta-
IARC Working Group (2012). Tobacco smoking. In: IARC Monographs on the Evaluation of
Carcinogenic Risks to Humans, vol. 100E: Personal Habits and indoor combustion.
Ioannidis JP, Schully SD, Lam TK, Khoury MJ (2013). Knowledge integration in cancer:
current landscape and future prospects. Cancer Epidemiol Biomarkers Prev 22: 3-10.
18
Kamangar F, Dawsey SM, Blaser MJ, Perez-Perez GI, Pietinen P, Newschaffer CJ, et al.
associated with Helicobacter pylori seropositivity. J Natl Cancer Inst 98: 1445-1452.
Kim Y, Shin A, Gwack J, Jun JK, Park SK, Kang D, et al. (2007). [Cigarette smoking and
gastric cancer risk in a community-based cohort study in Korea]. J Prev Med Public
Kneller RW, You WC, Chang YS, Liu WD, Zhang L, Zhao L, et al. (1992). Cigarette
smoking and other risk factors for progression of precancerous stomach lesions. J Natl
Cigarette smoking and the risk of gastric cancer: a pooled analysis of two prospective
al. (2008). Smoking and gastric cancer: systematic review and meta-analysis of cohort
vitamin C and adenocarcinoma of the stomach. Cancer Causes Control 15: 67-72.
Lee K, Caceres D, Varela N, Csendes DA, Rios RH, Quinones SL (2006). [Allelic variants of
19
polymorphisms and their association with smoking and alcohol consumption as gastric
Lindblad M, Rodriguez LA, Lagergren J (2005). Body mass, tobacco and alcohol and risk of
esophageal, gastric cardia, and gastric non-cardia adenocarcinoma among men and
and alcoholic beverages and the risk of stomach cancer: a case-control study in Italy.
Lunet N, Valbuena C, Vieira AL, Lopes C, David L, Carneiro F, et al. (2007). Fruit and
vegetable consumption and gastric cancer by location and histological type: case-
Mao Y, Hu J, Semenciw R, White K (2002). Active and passive smoking and the risk of
Matsuo K, Oze I, Hosono S, Ito H, Watanabe M, Ishioka K, et al. (2013). The aldehyde
Mu LN, Lu QY, Yu SZ, Jiang QW, Cao W, You NC, et al. (2005). Green tea drinking and
multigenetic index on the risk of stomach cancer in a Chinese population. Int J Cancer
116: 972-983.
Nomura AM, Wilkens LR, Henderson BE, Epplein M, Kolonel LN (2012). The association of
cigarette smoking with gastric cancer: the multiethnic cohort study. Cancer Causes
20
Pakseresht M, Forman D, Malekzadeh R, Yazdanbod A, West RM, Greenwood DC, et al.
(2011). Dietary habits and gastric cancer risk in north-west Iran. Cancer Causes
Cancer Attributable to Tobacco Smoking in 2012 and Predictions for 2020. Dig Dis
Pelucchi C, Lunet N, Boccia S, Zhang ZF, Praud D, Boffetta P, et al. (2015). The stomach
cancer pooling (StoP) project: study design and presentation. Eur J Cancer Prev 24:
16-23.
Pourfarzi F, Whelan A, Kaldor J, Malekzadeh R (2009). The role of diet and other
application to alcohol and esophageal squamous cell carcinoma. Stat Med 29: 2679-
2687.
21
Setiawan VW, Zhang ZF, Yu GP, Li YL, Lu ML, Tsai CJ, et al. (2000). GSTT1 and GSTM1
null genotypes and the risk of gastric cancer: a case-control study in a Chinese
Setiawan VW, Yu GP, Lu QY, Lu ML, Yu SZ, Mu L, et al. (2005). Allium vegetables and
Smith-Warner SA, Spiegelman D, Ritz J, Albanes D, Beeson WL, Bernstein L, et al. (2006).
Steevens J, Schouten LJ, Goldbohm RA, van den Brandt PA (2010). Alcohol consumption,
Tayler R, Piper DW (1977). The carcinogenic effect of cigarette smoke. The effect of
cigarette smoke on human gastric mucosal cells in organ culture. Cancer 39: 2520-
2523.
Tramacere I, La Vecchia C, Negri E (2011). Tobacco smoking and esophageal and gastric
Tredaniel J, Boffetta P, Buiatti E, Saracci R, Hirsch A (1997). Tobacco smoking and gastric
Ye W, Ekstrom AM, Hansson LE, Bergstrom R, Nyren O (1999). Tobacco, alcohol and the
risk of gastric cancer by sub-site and histologic type. Int J Cancer 83: 223-229.
smoking and risk of gastric cancer: case-control study from Moscow, Russia. Cancer
22
Zendehdel K, Nyren O, Luo J, Dickman PW, Boffetta P, Englund A, et al. (2008). Risk of
gastroesophageal cancer among smokers and users of Scandinavian moist snuff. Int J
Zhang ZF, Kurtz RC, Klimstra DS, Yu GP, Sun M, Harlap S, et al. (1999). Helicobacter
pylori infection on the risk of stomach cancer and chronic atrophic gastritis. Cancer
23
FIGURES LEGEND
Figure 1. Study-specific and pooled ORs and corresponding 95% confidence intervals of
gastric cancer risk for ever smokers compared to never smokers in the Stomach cancer
Figure 2. Study-specific and pooled ORs and corresponding 95% confidence intervals of
gastric cancer risk for former smokers (a), smokers of up to 10 cigarettes (b), smokers of more
than 10 to 20 cigarettes (c) and smokers of more than 20 cigarettes per day (d), compared to
Figure 3. Relation between intensity (a), duration of cigarette smoking (b) and risk of gastric
cancer according to the best fitting fractional polynomial model, i.e. the one with powers p1=-
Figure 4. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) for gastric cancer
according to intensity of cigarette smoking in strata of sex, age, geographic area, cancer site,
cancer histotype, Helicobacter Pylori infection and controls recruitment in the Stomach
24
Table 1. Distribution of 10,290 cases of gastric cancer and 26,145 controls according to study
center, sex, age, and other selected covariates in the Stomach cancer Pooling (StoP) Project
consortium.
Cases Controls
N % N %
Total 10290 26145
Study center (Reference)
Europe 5079 49.4 12664 48.4
Greece (Lagiou et al., 2004) 110 1.1 100 0.4
Italy 1 (La Vecchia et al., 1995) 769 7.5 2081 8.0
Italy 2 (Lucenteforte et al., 2008) 230 2.2 547 2.1
Italy 3 (De Feo et al., 2012) 160 1.6 444 1.7
Italy 4 (Buiatti et al., 1989) 1016 9.9 1159 4.4
Portugal (Lunet et al., 2007) 692 6.7 1667 6.4
Russia (Zaridze et al., 2000) 450 4.4 611 2.3
Spain 1 (Castano-Vinyals et al., 2015) 441 4.3 3440 13.2
Spain 2 (Santibanez et al., 2012) 401 3.9 455 1.7
Sweden 1 (Harris et al., 2013) 88 0.9 352 1.3
Sweden 2 (Harris et al., 2013) 161 1.6 644 2.5
Sweden 3 (Ye et al., 1999) 561 5.5 1164 4.5
Sex
Male 6804 66.1 15600 59.7
Female 3486 33.9 10545 40.3
Age
<40 355 3.4 1917 7.3
40-45 362 3.5 1542 5.9
45-50 608 5.9 2009 7.7
50-54 995 9.7 2700 10.3
25
55-59 1340 13.0 3128 12.0
60-64 1616 15.7 4079 15.6
65-69 1864 18.1 4240 16.2
70-75 1864 18.1 3857 14.8
≥75 1286 12.5 2673 10.2
Social class1
Low 5416 52.6 10625 40.6
Intermediate 2697 26.2 7857 30.1
High 1242 12.1 5422 20.7
Missing 935 9.1 2241 8.6
26
Table 2. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) for gastric cancer according
to cigarette and tobacco smoking habits in the Stomach cancer Pooling (StoP) Project consortium.
Cases Controls
OR (CI 95%)1
N % N %
Total 10290 26145
Cigarette smoking status
Never smoker 4261 41.4 11742 44.9 1
Ever cigarette smoker 5814 56.5 13910 53.2 1.20 (1.09-1.32)
Other than cigarette smoker 122 1.2 343 1.3 1.07 (0.76-1.51)
missing 93 0.9 150 0.6
27
10 to <20 515 6.7 1392 7.6 0.80 (0.70-0.92)
≥ 20 (median value: 27) 615 8.0 1728 9.5 0.79 (0.63-0.99)
Other than cigarette smoker 122 1.6 343 1.9
Missing 248 3.2 453 2.5
P value for trend 0.007
1
Pooled ORs were computed using random-effects models. Study-specific ORs were adjusted, when available, for sex,
age, race/ethnicity, social class, alcohol drinking, fruit and vegetable consumption and study center for multicentric
studies. 2Information on former/current smoking status was not available for studies China 4 (Setiawan et al., 2000)
and Iran 3 (Derakhshan et al., 2008).
3
Current smokers only.
4
Time since stopping cigarette smoking was not available for studies Greece (Lagiou et al., 2004), Canada (Mao et al.,
2002), China 1 (Deandrea et al., 2010), Iran 1 (Pourfarzi et al., 2009), Iran 2 (Pakseresht et al., 2011), USA 1 (Zhang
et al., 1999), Sweden 1 (Harris et al., 2013), and Sweden 2 (Harris et al., 2013).
28
Figure 1
29
Figure 2
30
Figure 3
31
Figure 4
32