Tel : +39 02 39 01 4658
Alcohol drinking and pancreatic cancer risk: a meta-analysis of the dose-risk relation
Article first published online: 8 OCT 2009
Copyright © 2009 UICC
International Journal of Cancer
Volume 126, Issue 6, pages 1474–1486, 15 March 2010
How to Cite
Tramacere, I., Scotti, L., Jenab, M., Bagnardi, V., Bellocco, R., Rota, M., Corrao, G., Bravi, F., Boffetta, P. and La Vecchia, C. (2010), Alcohol drinking and pancreatic cancer risk: a meta-analysis of the dose-risk relation. Int. J. Cancer, 126: 1474–1486. doi: 10.1002/ijc.24936
- Issue published online: 18 JAN 2010
- Article first published online: 8 OCT 2009
- Accepted manuscript online: 8 OCT 2009 12:00AM EST
- Manuscript Accepted: 10 SEP 2009
- Manuscript Received: 30 JUL 2009
- Italian Association for Cancer Research (AIRC)
- alcohol drinking;
- dose-risk relation;
- pancreatic cancer;
- systematic review
In order to provide a more precise quantification of the association between alcohol consumption and pancreatic cancer risk, we performed a meta-analysis of relevant dose-risk results. We conducted a PubMed search of all case-control (N=21) and cohort (N=11) studies published up to March 2009. We computed summary relative risk (RR) estimates using either fixed- or, in the presence of heterogeneity, random-effects models. The pooled RR was 0.92 (95% confidence interval, 95% CI, 0.86–0.97) for <3 drinks/day and 1.22 (95% CI, 1.12–1.34) for ≥3 drinks/day. The increased risk for heavy drinking was similar in women and men, but apparently stronger in cohort studies (RR=1.29), in studies with high quality index (RR=1.30), and did not appear to be explained by residual confounding by either history of pancreatitis or tobacco smoking. This meta-analysis provides strong evidence for the absence of a role of moderate drinking in pancreatic carcinogenesis, coupled to an increased risk for heavy alcohol drinking. Given the moderate increase in risk and the low prevalence of heavy drinkers in most populations, alcohol appears to be responsible only for a small fraction of all pancreatic cancers.
Whether a causal association exists between alcohol drinking and pancreatic cancer is an open research question of notable scientific and public health importance. Most studies have found no association between alcohol drinking and pancreatic cancer risk,1, 2 although a modest and inconsistent association has been identified in some studies.3–5 A meta-analysis of 17 studies published up to 2000 including 2,524 cases resulted in relative risks (RR) of 1.05 for consumption of 50 grams/day and of 1.18 for 100 grams/day,3 but the risk estimates were not statistically significant owing to the small numbers of heavy drinkers. A recent pooled analysis of 14 cohort studies showed a RR of 1.22 (95% confidence interval, 95% CI, 1.03–1.45) for intake above 30 grams/day.4 Thus, it still remains to be determined whether a dose-response association exists between alcohol consumption and risk of pancreatic cancer, or whether any potential association may be restricted to heavy intake.
Any possible association of alcohol with pancreatic cancer risk may be confounded by other major risk factors of pancreatic cancer, particularly pancreatitis and smoking with which alcohol intake may also be correlated. Heavy alcohol consumption is the most common cause of chronic pancreatitis in high- and medium-income countries,6–9 which in turn is one of the few established risk factors of pancreatic cancer, with observed RRs in the range 5–15.10, 11 However, a history of pancreatitis is present only in a small proportion (generally less than 10%) of cases of pancreatic cancer,10, 12, 13 and is responsible for less than 5% of all pancreatic cancer cases.14 In most populations heavy alcohol drinking is also correlated with tobacco smoking, the major recognized risk factor for pancreatic cancer.10, 14 Whether tobacco or history of pancreatitis are confounders or effect modifiers of the possible association between heavy alcohol drinking and pancreatic cancer has not been investigated in detail, mainly due to the limited number of heavy alcohol drinkers.
In order to provide a detailed quantification of the association between alcohol consumption and pancreatic cancer risk, and to assess the role of tobacco smoking in this association, we conducted a meta-analysis of studies published up to March 2009, which provided risk estimates for three or more categories of alcohol drinking.
Material and Methods
Search strategy and study selection
We performed a MEDLINE search in PubMed of all case-control and cohort studies published as original articles in English up to March 2009 using the MESH terms “alcohol” and combinations of “pancreas” or “pancreatic” and “cancer” or “carcinoma” or “neoplasm”, following the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines.15 Two of the authors (IT and LS) retrieved and assessed potentially relevant articles, and checked the reference list of all papers of interest to obtain other pertinent publications. Figure 1 gives the flowchart for selection of articles.
When multiple reports were published on the same study population, we included in the meta-analysis only the most recent and informative one. For example, 4 studies were published both individually16–18 and within a pooled analysis.4 We included in the meta-analysis the 4 studies individually, since the separate reports provided more detailed information. In total, we identified 42 papers: 30 of case-control design19–48 and 12 of cohort design.4, 16–18, 49–56 Each publication identified in this process was reviewed and included in the analysis if the following criteria were met: case-control and cohort studies considering at least 3 levels of alcohol consumption and reporting the estimates of the odds ratio (OR) or relative risk (RR) and the corresponding confidence intervals (CI)—or information sufficient to calculate them—for each exposure level. In addition, studies that met the above criteria but reported only information on specific types of alcoholic beverages21, 40 were not included. Thus, the present analyses are based on 32 studies: 21 case-control19, 20, 22–25, 28, 30–32, 34–39, 41, 42, 45, 46, 48 and 11 cohort studies4, 16–18, 49–55 (Table 1).
We assessed the quality of each study, according to predefined criteria, which addressed study design (3 questions), assessment of alcohol drinking (4 questions) and analysis (3 questions) (Table 2). For each question, low quality was scored 0 and high quality was scored between 0.5 and 2. An overall quality score was obtained for each study by adding up the scores of individual questions, with a range between 0 and 10.
For each study, we extracted details on study design, country, number of subjects (cases, controls or cohort size), period of enrolment (case-control studies) or follow-up (cohort studies), sex of the population considered, confounders allowed for in the analysis, RR estimates for categories of alcohol consumption and the corresponding 95% CI, and, when available, the number of cases and non-cases for each level of alcohol consumption considered.
Since the various studies used different measurement units for alcohol consumption, we used the midpoint of each category for each study. When grams of ethanol were used as measure, we define 1 drink as 12.5 g of ethanol. We used non- plus occasional (primarily <0.5 drinks per day) drinkers as reference category. On the basis of cut off points of most published data, we defined moderate alcohol intake as <3 drinks per day and heavy intake as ≥3 drinks per day. For men only, we were also able to obtain data on ≥5 and ≥7 drinks per day.
The measure of association was the RR for cohort studies and the OR for case-control studies. When available, we used multivariate-adjusted risk estimates. When RRs were not available in the published articles, we computed unadjusted RRs from the exposure distributions of cases and controls as reported in the publications.
We computed summary estimates of the RR using fixed-effect models (i.e., as weighed averages using the inverse of the variance of the log(RR) as weight), and we assessed the heterogeneity among studies using the χ2 test.57 When significant heterogeneity (defined as a p-value for heterogeneity less than 0.10) was found, we used a random-effects model (i.e., as weighed averages using the inverse of the sum of the variance of the log(RR) and the moment estimator of the variance between studies as weight).58 We calculated summary estimates in strata of sex, study design, geographic area, quality score, and allowance for tobacco smoking.
We report forest plots, in which a square is plotted for each study, whose center projection on the underlying scale corresponds to the study-specific RR. The area of the square is proportional to the inverse of the variance of the natural logarithm of the RR, and thus gives a measure of the amount of information available from that particular result. A diamond is used to plot the summary RR, the center of which represents the RR and its extremes the 95% CI.
For dose-response analysis, we used a random-effect meta-regression model in a non-linear dose-response relationship framework, providing the best fitting two-term fractional-polynomial model. The statistical methods used for this analysis are described in detail elsewhere.59
The main characteristics of the 21 case-control19, 20, 22–25, 28, 30–32, 34–39, 41, 42, 45, 46, 48 and 11 cohort studies4, 16–18, 49–55 included in the present analyses are reported in Table 1. Most studies were conducted in the USA and Europe, and only few studies in Asia.
Table 2 reports the distribution of studies according to the quality scores. One of the scores (report of results for different drinking patterns) was not informative as it was not satisfied by any of the studies considered. Lack of information on the validity or on the reproducibility of the questionnaire (69%) and on the response rate (59%), and the use of self-administrated questionnaires (63%) were the main contributors to low overall quality scores. The median quality score was 5.5 (range: 1–8).
Figure 2 shows the RRs of pancreatic cancer for moderate alcohol drinkers (<3 drinks/day) as compared to non- or occasional drinkers from 12 case-control and cohort studies. The summary RR was 0.92 (95% CI, 0.86–0.97, p for heterogeneity 0.06). The summary RR of case-control studies was 0.90 (95% CI, 0.80–1.01, 17 studies, p for heterogeneity 0.03), and that of cohort studies was 0.91 (95% CI, 0.86–0.95, 7 studies, p for heterogeneity 0.36). Consideration of men and women separately provided summary estimates of 0.95 and 0.89, respectively. In sex-specific analyses, the summary RR were not heterogeneous between case-control and cohort studies.
Figure 3 shows the RRs of pancreatic cancer for heavy alcohol drinkers (defined as 3 or more drinks/day) as compared to non- or occasional drinkers. Among studies reporting estimates both for men and women, the summary RR was 1.22 (95% CI, 1.12–1.34) from all studies, 1.12 (95% CI, 0.97–1.29) from 8 case-control studies, and 1.30 (95% CI, 1.16–1.47) from 5 cohort studies (including a pooled analysis). No heterogeneity across studies was observed. Considering men and women separately, the RR from all studies combined was 1.16 for women and 1.19 for men, respectively. Considering as high consumption at least 5 drinks/day, the summary RR for men from 7 studies providing data22, 30, 36, 37, 39, 51, 52 was 1.17 (95% CI, 0.99–1.38). Considering consumption of at least 7 drinks/day, the summary RR from 5 studies providing data for men22, 30, 37, 39, 51 was 1.27 (95% CI, 1.03–1.56). For women, only two studies22, 39 reported information for such high levels of alcohol consumption, giving a summary RR of 0.79 (95% CI, 0.43–1.44) for a consumption of 5 or more drinks/day, and of 0.78 (95% CI, 0.40–1.51) for 7 or more drinks/day.
Table 3 shows the summary RR of pancreatic cancer for drinkers of 3 or more drinks/day as compared to non- or occasional drinkers from 12 case-control and 6 cohort studies providing information on heavy drinking, according to selected factors. No significant heterogeneity was found across strata of geographic area, neither comparing Europe and USA combined with Asia (p for heterogeneity 0.39). A significant association was found in cohort studies only (RR=1.29 in cohort and RR=1.10 in case-control studies, respectively; p=0.069), and in studies reporting estimates adjusted for tobacco smoking (the RR adjusted for tobacco smoking was 1.23 and the corresponding unadjusted estimate was 1.09, in the absence of significant heterogeneity, p=0.266). For quality index, an association was observed among studies with a high quality index (RR=1.30, 95% CI, 1.16–1.45), but not in studies with a low score (RR=1.07, 95% CI, 0.93–1.22, p-value of test of heterogeneity 0.030).
Figure 4 gives the relative risk function and the corresponding 95% CI for the dose-response relation between alcohol consumption and risk of pancreatic cancer. Among the two-terms fractional-polynomial models, the best-fitting dose-response relationship was log(RR)=dose2 + dose2*log (dose). Compared to non-drinkers, the random-effect model-based pooled estimates of the RR were 1.03 (95% CI, 0.99–1.07) for 25, 1.10 (95% CI, 1.02–1.20) for 50, 1.19 (95% CI, 1.06–1.35) for 75, 1.30 (95% CI, 1.08–1.55) for 100, and 1.40 (95% CI, 1.06–1.85) for 125 grams of alcohol per day, respectively.
The present meta-analysis shows an approximately 20% increase in the risk of pancreatic cancer among heavy alcohol drinkers (defined as 3 or more drinks/day), compared with non- or occasional drinkers. Thus, we provided strong quantitative evidence of the association between heavy alcohol consumption and the risk of pancreatic cancer. Such an association was moderate, and was restricted to heavy alcohol drinkers. The observed increase in risk with heavy drinking was similar in women and men, and appeared to be stronger in studies with the best methodologies (cohort studies, studies with high quality score, studies with strict allowance for confounding by tobacco smoking). The association of heavy drinking with pancreatic cancer was also highlighted by a dose-risk analysis which showed a significant increase in cancer risk for an alcohol intake of 40 or more grams/day. The findings of this meta-analysis also provides evidence that moderate alcohol consumption, defined as less than 3 drinks/day, does not increase the risk of pancreatic cancer, and indeed the upper confidence limit for a moderate consumption was 0.97.
History of pancreatitis could affect the association between heavy alcohol drinking and pancreatic cancer. In one study38 the RR for alcohol consumption was not appreciably modified after allowance for history of pancreatitis. However, that was the only study including adjustment for pancreatitis, and data on pancreatitis are scanty. Thus, the question of whether the association between heavy alcohol drinking and pancreatic cancer is completely or partly mediated by pancreatitis cannot be adequately assessed with published data.
Other possible mechanisms by which heavy alcohol use may increase human pancreatic cancer risk include a direct effect of acetaldehyde, the main metabolite of ethanol, which is an established experimental carcinogen.60 In addition, fatty acid esters, products of the interaction between ethanol and fatty acids, accumulate in the pancreas and could induce inflammatory responses, fibrosis and hence contribute to pancreatic carcinogenesis.61–66 Selected types of alcoholic beverages (i.e., fruit derived spirits) may also contain specific carcinogens.67, 68 The present data were however inadequate to investigate the separate effect of various types of alcohol beverages, i.e., beer, wine and spirits.
Tobacco smoking is the major recognized risk factor for pancreatic cancer, and it is correlated with heavy alcohol drinking. Available data are inadequate to investigate the potential modifying effect of tobacco smoking on the carcinogenic effect of alcohol drinking. Most risk estimates included in the present analyses were carefully adjusted for tobacco, and, if anything, this RR for heavy alcohol drinking was higher in the 12 studies which made allowance for tobacco (RR=1.23) than in the 6 studies which did not adjust for tobacco (RR=1.09). Likewise, the results of a large cohort study from the United States51 suggest that heavy alcohol consumption exerted a comparable effect on pancreatic cancer risk in both never smokers (RR=1.35), and ever smokers (RR=1.50), although relatively few never smokers were included in that analysis, and the risk estimate in this subgroup was not significant.
Residual confounding by other risk factors of pancreatic cancer, namely overweight and diabetes, remains a concern. Whenever possible, we used risk estimates adjusted for these and other (i.e., education and other social class indicators, coffee drinking, selected aspects of diet such as meat, vegetable, fruit consumption, or energy intake, personal history of gallbladder disease, and family history of pancreatic cancer) potential confounding factors, but none of these appeared to appreciably modify the association between heavy alcohol drinking and pancreatic cancer.
The weaker association observed in case-control studies as compared to cohort studies may be due to the fact that subjects with pancreatic disorders, and consequently possible digestive tract complaints, may reduce their alcohol consumption and/or drinking patterns prior to diagnosis.69, 70 Moreover, case-control studies are generally considered more prone to recall and selection biases than cohort studies.71 An important problem concerns the assessment of alcohol intake, based on patients' self-reporting in most investigations. In fact, it is possible that alcohol drinking was under-reported in several studies.72 RRs would be biased towards unity in case of non-differential under-reporting of cases and controls.3 However, in a few case-control studies73–75 as well as in several cohort studies,76–81 reproducibility and validity of alcohol drinking were satisfactory. Likewise, studies with a high quality score, which in our judgments were less likely to suffer from bias, were more prone to show a significant association between alcohol consumption and the risk of pancreatic cancer, thus supporting the notion that the association observed between heavy alcohol drinking and pancreatic cancer is likely to be real.
Thus, the available data provide strong support to the concept that heavy alcohol drinking is associated to increased risk of pancreatic cancer. Given the absence of apparent effect of moderate alcohol drinking and pancreatic cancer risk, and the moderate association observed with heavy drinking, alcohol would be responsible for a small fraction of all pancreatic cancer, i.e., 2 to 5% of all pancreatic cancers in a population with a prevalence of heavy drinkers of 10 to 15%.82
The authors thank Ms. I. Garimoldi for editorial assistance.
- 1Carcinogenicity of alcoholic beverages. Lancet Oncol 2007; 8: 292–3., , , , , , , .
- 2WCRF/AICR. World Cancer Research Fund/American Institute for Cancer Research. Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. Washington DC: AICR 2007.
- 3A meta-analysis of alcohol drinking and cancer risk. Br J Cancer 2001; 85: 1700–5., , , .
- 4Alcohol intake and pancreatic cancer risk: a pooled analysis of fourteen cohort studies. Cancer Epidemiol Biomarkers Prev 2009; 18: 765–76., , , , , , , , , , , , et al.
- 5Diet, alcohol, coffee and pancreatic cancer: final results from an Italian study. Eur J Cancer Prev 1998; 7: 455–60., , , , .
- 6The epidemiology of alcohol-induced pancreatitis. Pancreas 2003; 27: 286–90., .
- 7Chronic pancreatitis: challenges and advances in pathogenesis, genetics, diagnosis, and therapy. Gastroenterology 2007; 132: 1557–73., , , .
- 8Epidemiology, aetiology and outcome of acute and chronic pancreatitis: An update. Best Pract Res Clin Gastroenterol 2008; 22: 45–63., , .
- 9Alcohol and cause-specific mortality in Russia: a retrospective case-control study of 48,557 adult deaths. Lancet 2009; 373: 2201–14., , , , , , , , , , .
- 10Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol 2006; 20: 197–209., .
- 11Pancreatitis and the risk of pancreatic cancer. International Pancreatitis Study Group. N Engl J Med 1993; 328: 1433–7., , , , , , , , .
- 12Pancreatitis and pancreatic cancer: a population-based study. J Natl Cancer Inst 1994; 86: 625–7., , , , , ,
- 13Pancreatitis and the risk of pancreatic cancer. Pancreas 1995; 11: 185–9., , , , , .
- 14Attributable risks for pancreatic cancer in northern Italy. Cancer Epidemiol Biomarkers Prev 1996; 5: 23–7., , .
- 15Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000; 283: 2008–12., , , , , , , , , .
- 16Smoking, alcohol, coffee, and tea intake and incidence of cancer of the exocrine pancreas: the Iowa Women's Health Study. Cancer Epidemiol Biomarkers Prev 1997; 6: 1081–6., , , , , .
- 17Coffee and alcohol consumption and the risk of pancreatic cancer in two prospective United States cohorts. Cancer Epidemiol Biomarkers Prev 2001; 10: 429–37., , , , .
- 18Dietary and other methyl-group availability factors and pancreatic cancer risk in a cohort of male smokers. Am J Epidemiol 2001; 153: 680–7., , , , , .
- 19A case-control study of diet and cancer of the pancreas. Am J Epidemiol 1991; 134: 167–79., , , , , .
- 20Alcohol, beer and cancer of the pancreas. Int J Cancer 1990; 45: 842–6., , , , .
- 21Lifetime consumption of alcoholic beverages, tea and coffee and exocrine carcinoma of the pancreas: a population-based case-control study in The Netherlands. Int J Cancer 1992; 50: 514–22., , , , .
- 22Coffee, alcohol, smoking and cancer of the pancreas: a case-control study. Int J Cancer 1989; 43: 17–21., , , , .
- 23Pancreatic cancer, alcohol, diabetes mellitus and gall-bladder disease. Int J Cancer 1989; 43: 415–21., .
- 24Life-style risk factors for pancreatic cancer in Louisiana: a case-control study. Am J Epidemiol 1988; 128: 324–36., , , ,
- 25Risk of pancreatic cancer in relation to medical history and the use of tobacco, alcohol and coffee. Int J Cancer 1990; 45: 816–20., .
- 26Tobacco, alcohol, and coffee and cancer of the pancreas. A population-based, case-control study in Quebec, Canada. Cancer 1991; 67: 2664–70., , .
- 27Diet and other risk factors for cancer of the pancreas. Cancer 1985; 55: 460–7., , , , , , .
- 28Pancreatic cancer, blood glucose and beverage consumption. Int J Cancer 1988; 41: 794–7., , .
- 29Coffee and alcohol as determinants of risk of pancreas cancer: a case-control study from Toronto. Int J Cancer 1991; 47: 384–9., , , .
- 30Cigarette smoking and alcohol consumption and the risk of pancreatic cancer: a case-control study in Shanghai, China. Cancer Causes Control 1995; 6: 369–76., , , , , , ,
- 31Tobacco, ethanol, coffee, pancreatitis, diabetes mellitus, and cholelithiasis as risk factors for pancreatic carcinoma. Cancer Causes Control 1993; 4: 375–82., , , , , .
- 32Pancreas cancer and smoking, beverage consumption, and past medical history. J Natl Cancer Inst 1986; 76: 49–60., , , .
- 33A multi-institute case-control study on the risk factors of developing pancreatic cancer. Jpn J Clin Oncol 1992; 22: 286–91., , , , , , , , .
- 34A nested case-control study on alcohol drinking, tobacco smoking, and cancer. Cancer Detect Prev 1996; 20: 557–65., , , .
- 35Diet and pancreatic cancer: a case-control study. Am J Epidemiol 1986; 124: 894–902., , , , , , , .
- 36A case-control study of pancreatic cancer and cigarettes, alcohol, coffee and diet. Am J Public Health 1989; 79: 1016–9., , , , .
- 37Alcohol and pancreatic cancer in blacks and whites in the United States. Cancer Res 1995; 55: 4899–905., , , , , , , , , , et al.
- 38Alcohol consumption and risk of pancreatic cancer. Nutr Cancer 1997; 27: 157–61., , , .
- 39Epidemiology of coffee and pancreatic cancer. Cancer Res 1983; 43: 3900–6., , .
- 40Cigarette smoking, alcohol, tea and coffee consumption and pancreas cancer risk: a case-control study from Opole, Poland. Int J Cancer 1993; 53: 601–7., , , , , .
- 41Lifestyle, dietary, and medical history factors associated with pancreatic cancer risk in Ontario, Canada. Cancer Causes Control 2009; 20: 825–34., , .
- 42Helicobacter pylori infection and development of pancreatic cancer. Cancer Epidemiol Biomarkers Prev 2008; 17: 1188–94., , , , , , .
- 43Risk factors for pancreatic cancer: case-control study. Am J Gastroenterol 2007; 102: 2696–707., , , , , , , , , , , .
- 44Epidemiology of pancreatic cancer in Japan: a nested case-control study from the Hospital-based Epidemiologic Research Program at Aichi Cancer Center (HERPACC). Int J Epidemiol 2003; 32: 257–62., , , , , , .
- 45Alcohol and mortality: a cohort study of male Japanese physicians. Int J Epidemiol 1986; 15: 527–32., , , , .
- 46Alcohol drinking and one-carbon metabolism-related gene polymorphisms on pancreatic cancer risk. Cancer Epidemiol Biomarkers Prev 2008; 17: 2742–7., , , , , , , , , , .
- 47Alcohol and smoking as risk factors in chronic pancreatitis and pancreatic cancer. Dig Dis Sci 1999; 44: 1303–11., , , , , , , , , , , , et al.
- 48Alcohol, tobacco and coffee consumption and the risk of pancreatic cancer: results from the Canadian Enhanced Surveillance System case-control project. Canadian Cancer Registries Epidemiology Research Group. Eur J Cancer Prev 2000; 9: 49–58., , , .
- 49Predictors of pancreatic cancer mortality among a large cohort of United States adults. Cancer Causes Control 2000; 11: 915–23., , , .
- 50Alcohol consumption and risk of pancreatic cancer in the Netherlands cohort study. Am J Epidemiol 2009; 169: 1233–42., , , , .
- 51Alcohol use and risk of pancreatic cancer: the NIH-AARP Diet and Health Study. Am J Epidemiol 2009; 169: 1043–51., , , , , , , .
- 52Risk of pancreatic cancer in relation to alcohol drinking, coffee consumption and medical history: findings from the Japan collaborative cohort study for evaluation of cancer risk. Int J Cancer 2002; 99: 742–6., , , , , , , .
- 53Ethanol intake and the risk of pancreatic cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC). Cancer Causes Control 2009; 20: 785–94., , , , , , , , , , , , et al.
- 54A prospective study of pancreatic cancer in the elderly. Int J Cancer 1994; 58: 46–9., , , , .
- 55Factors associated with incident and fatal pancreatic cancer in a cohort of middle-aged women. Int J Cancer 2009; 124: 2400–5., , , , , , .
- 56A cohort study of smoking, alcohol consumption, and dietary factors for pancreatic cancer (United States). Cancer Causes Control 1993; 4: 477–82., , , , , , , , ,
- 57Quantitative methods in the review of epidemiologic literature. Epidemiol Rev 1987; 9: 1–30..
- 58Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88., .
- 59Flexible meta-regression functions for modeling aggregate dose-response data, with an application to alcohol and mortality. Am J Epidemiol 2004; 159: 1077–86., , , .
- 60IARC. Re-evaluation of some organic chemicals, hydrazine and hydrogen peroxide. IARC monographs on the evaluation on the carcinogenic risks to humans. Vol. 71. IARC, Lyon, 1999: 319–335.
- 61Acute ethanol administration induces oxidative changes in rat pancreatic tissue. Gut 1996; 38: 742–6., , , , .
- 62Battle-scarred pancreas: role of alcohol and pancreatic stellate cells in pancreatic fibrosis. J Gastroenterol Hepatol 2006; 21 Suppl 3: S97–S101., , .
- 63Ethanol toxicity in pancreatic acinar cells: mediation by nonoxidative fatty acid metabolites. Proc Natl Acad Sci USA 2004; 101: 10738–43., , , , , .
- 64Alcohol and pancreatic cancer. Alcohol 2005; 35: 205–11., , .
- 65Ethanol diet increases the sensitivity of rats to pancreatitis induced by cholecystokinin octapeptide. Gastroenterology 1999; 117: 706–16., , , , , , , , .
- 66Inflammation and cancer: back to Virchow? Lancet 2001; 357: 539–45., .
- 67Alcohol and cancer. Lancet Oncol 2006; 7: 149–56., .
- 68Trends in mortality from hepatocellular carcinoma in Europe, 1980–2004. Hepatology 2008; 48: 137–45., , , , , .
- 69Hepatitis viruses, alcohol, and tobacco in the etiology of hepatocellular carcinoma in Italy. Cancer Epidemiol Biomarkers Prev 2006; 15: 683–9., , , , , , , , , , , , et al.
- 70Risk factors for hepatocellular carcinoma in northern Italy. Int J Cancer 1988; 42: 872–6., , , , .
- 71Epidemiology principles and methods. 2nd ed. New York: Little, Brown and Company 1996., .
- 72Mortality in relation to consumption of alcohol: 13 years' observations on male British doctors. BMJ 1994; 309: 911–8., , , , .
- 73Reliability of information on cigarette smoking and beverage consumption provided by hospital controls. Epidemiology 1996; 7: 312–5., , , , .
- 74Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 1996; 25: 775–82., , , , , , , .
- 75[The quantitative alimentary questionnaire technique used by the I.N.S.E.R.M. nutrition department (author's transl)]. Rev Epidemiol Sante Publique 1980; 28: 367–72., .
- 76Validation of the American Cancer Society Cancer Prevention Study II Nutrition Survey Cohort Food Frequency Questionnaire. Epidemiology 2000; 11: 462–8., , , , .
- 77The assessment of alcohol consumption by a simple self-administered questionnaire. Am J Epidemiol 1991; 133: 810–7., , , , , , .
- 78Validation of a dietary questionnaire used in a large-scale prospective cohort study on diet and cancer. Eur J Clin Nutr 1994; 48: 253–65., , , , , , .
- 79Dietary assessment in the California Teachers Study: reproducibility and validity. Cancer Causes Control 2008; 19: 595–603., , , , , , , , , .
- 80Dietary assessment of older Iowa women with a food frequency questionnaire: nutrient intake, reproducibility, and comparison with 24-hour dietary recall interviews. Am J Epidemiol 1992; 136: 192–200., , , , .
- 81Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires: the Eating at America's Table Study. Am J Epidemiol 2001; 154: 1089–99., , , , , , , .
- 82Estimating the population attributable risk for multiple risk factors using case-control data. Am J Epidemiol 1985; 122: 904–14., , , , .