Cai S, Li Y, Ding Y, Chen K, Jin M. Alcohol drinking and the risk of colorectal cancer death: a meta-analysis. European Journal of Cancer Prevention 2014;23:532–539.
A causal link between alcohol consumption and colorectal cancer (CRC) was established only recently by the International Agency for Research on Cancer. However, the quantitative association between alcohol drinking and CRC mortality is still an open question. We performed a systemic review and meta-analysis on epidemiological studies to quantify the risk for CRC mortality at different levels of alcohol consumption. A literature search was carried out in PubMed and Web of Science to identify all relevant studies published from January 1966 to June 2013. The pooled relative risk (RR) and the corresponding 95% confidence interval (CI) were estimated by categorical meta-analysis. A dose–risk relation was also analyzed. Nine cohort studies exploring the association between CRC mortality and alcohol drinking were identified.
Compared with non/occasional drinkers, the pooled RR was 1.03 (95% CI, 0.93–1.15) for any, 0.97 (95% CI, 0.86–1.10) for light (≤12.5 g/day of ethanol), 1.04 (95% CI, 0.94–1.16) for moderate (12.6–49.9 g/day of ethanol), and 1.21 (1.01–1.46) for heavy drinkers (≥50 g/day of ethanol). For heavy drinkers, the pooled estimate was apparently higher for men (RR=1.28; 95% CI, 1.13–1.46) than for women (RR=0.79; 95% CI, 0.40–1.54; Pheterogeneity= 0.007). The dose–response analysis showed a J-shaped relationship between alcohol consumption and CRC mortality. The present meta-analysis provides the evidence for an association between heavy alcohol drinking (≥50 g/day of ethanol) and CRC mortality.
Data from prospective cohort studies on a possible association between alcohol consumption and the occurrence of colorectal cancer (CRC) are conflicting, with some suggesting an increase in risk while others failing to show such an effect. Given that CRC is such a common type of cancer, it would be important to determine if there is, or is not an effect, and especially the extent to which the amount of alcohol consumed relates to the association. As for the association of alcohol consumption to mortality related to CRC, there are very little data.
The present study is based on a meta-analysis of data from nine cohort studies (with more than two million subjects) to judge how the level of alcohol intake relates to CRC mortality. A total of almost 4,000 deaths from CRC were recorded. The conclusions of the authors are that the consumption of ≥ 50 grams of alcohol per day (about 4 typical drinks or more) increases the risk of death from CRC modestly [RR 1.21 (95% CI 1.01, 1.46)], but “light” drinking (≤ 12.5 g/day) or “moderate” drinking (12.6-49.9 g/day) do not increase the risk of CRC death.
Specific comments on the present study: This study provides considerable evidence that heavy drinking may increase the risk of death from a specific cancer, CRC, but that light to moderate alcohol intake does not increase the risk. The authors suggest that their results indicate a J-shaped curve for the association of alcohol with CRC mortality.
Reviewer Finkel commented: “Colorectal cancer is, in fact, a subtly diverse set of diseases, the distribution of which in different populations may help explain differences noted in the present study between Asians and North Americans. Unfortunately, from this meta-analysis we can learn nothing about differential effects related to the type of beverage or the pattern of drinking. Further, we learn nothing about the relation of the state of folate repletion of the subjects to their risk of disease. The sex difference is puzzling. In my opinion, the data are inadequate to clearly establish a J-shaped relation.”
Reviewer Stockley stated: “A recent comprehensive review of more than 7,000 peer-reviewed papers on the association of lifestyle factors and cancer undertaken by the World Cancer Research Fund in cooperation with the American Institute for Cancer Research (2007) reports that there is a threshold effect for alcohol for colorectal cancer. That review reports: ‘Increased risk is only apparent above a threshold of 30 g/day of ethanol for both sexes.’ The results of the present meta-analysis support such a threshold effect of alcohol in relation to death from colorectal cancer.”
Reviewer Van Velden believed that the present analysis adds little to what we already know about an increased risk of cancer from heavy drinking. The fact that light-to-moderate intake may ‘protect’ against certain diseases has repeatedly been demonstrated, often resulting in a “J” shaped curve. He stated: “It is important that we realize that alcohol can be a promoter of cancer (specifically in breast cancer) only when there is a folate deficiency. It is also dependant on the genetic blueprint of the individual. The cause of cancer is multifactorial, and cannot be found in meta analyses of studies based on epidemiological observations only.”
Reviewer Di Gaetano agreed that this was a good study and that “It supports the common finding that alcohol consumption shows a “J-shaped” relation with many diseases – this is unlike the relation with cigarette smoking, in which the effect tends to be a linear increase in disease with use.” He also agreed with comments from Forum member Ursini, in that “There is a great need to understand the mechanisms of the apparent protection against cancer deaths and total mortality from moderate drinking – they are not yet clear. The role of inflammation could certainly be a factor.”
Forum member Lanzmann-Petithory noted that the present study could not distinguish potential differences according to the type of beverage consumed; wine may have different effects from those of beer and spirits. She added: “In our studies in the Nancy COLOR cohort, we are finding a strong dose-dependent relation between total alcohol and colon cancer mortality, but an inverse association between wine intake and colorectal cancer. The mechanisms might be found in the interaction between mucosa proteins and some polyphenols, probably in the upper digestive tract.”
Differences in effects of alcohol on cancer incidence and mortality: Forum member Skovenborg had some interesting comments on this paper: “A number of studies have related alcohol to CRC incidence. The recent meta-analysis by Fedirko et al includes 31 cohort and 34 case-control studies; combined, the 53 studies included more than 207,00 CRC cases. The present meta-analysis is the first to study the association between alcohol consumption and CRC deaths; it includes data from 9 cohort studies with a total of 3,976 CRC deaths. A comparison of the meta-analyses of CRC incidence and CRC mortality illustrate some similarities and some differences. In both meta-analyses, the association between alcohol consumption and CRC incidence and mortality is stronger in men and in Asian populations. Several plausible explanations for those findings are suggested.
“However, the most important difference between cancer incidence and cancer mortality studies may be the finding of a linear dose-response association between alcohol drinking of >1 drink and CRC incidence and a J-shaped relationship between alcohol consumption and CRC mortality. This study shows that the consumption of ≥ 50 grams of alcohol per day increases the risk of death from CRC modestly, but neither “light” drinking nor “moderate” drinking increases the risk of CRC death.
“The explanation for the different effects on cancer incidence and mortality is not obvious; however, similar results have been reported regarding the association between breast cancer incidence and mortality A recent meta-analyses of 29,239 cases of breast cancer found little evidence that pre- or post-diagnosis alcohol consumption is associated with breast cancer–specific mortality for women with ER-positive disease (Ali et al). There was weak evidence that moderate post-diagnosis alcohol intake is associated with a small reduction in breast cancer–specific mortality in ER-negative disease.” Reviewer Ellison added: “Such a decrease in risk of total mortality could relate to the protective effects of alcohol on cardiovascular diseases, which are such common causes of death, even among subjects with cancer.”
References from Forum comments
Ali AM, Schmidt MK, Bolla MK, et al. Alcohol consumption and survival after a breast cancer diagnosis: a literature-based meta-analysis and collaborative analysis of data for 29,239 cases. Cancer Epidemiol Biomarkers Prev 2014;23:934-945. doi: 10.1158/1055-9965.EPI-13-0901.
Fedirko V, Tramacere I, Bagnardi V, Rota M, Scotti L, Islami F, et al. Alcohol drinking and colorectal cancer risk: an overall and dose–response meta-analysis of published studies. Annals of Oncology 2011;22:1958–1972. doi:10.1093/annonc/mdq653.
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.
Data from prospective cohort studies on the association between alcohol consumption and the occurrence of colorectal cancer (CRC) are conflicting, with some suggesting an increase in risk while others failing to show such an effect. There are little data on the effects of alcohol consumption on the risk of mortality from CRC. The present study is based on a meta-analysis of data from nine cohort studies (with a total of more than two million subjects) to judge how the level of alcohol intake relates to CRC mortality. A total of almost 4,000 deaths from CRC were recorded. The conclusions of the authors are that the consumption of ≥ 50 grams of alcohol (about 4 typical drinks or more) per day increases the risk of death from CRC modestly [RR 1.21 (95% CI 1.01, 1.46)], but “light” drinking (≤ 12.5 g/day) or “moderate” drinking (12.6-49.9 g/day) do not increase the risk of CRC death. In fact, they state that their data support a “J-shaped” relation between alcohol intake and CRC mortality (i.e., a slight decrease in mortality associated with light drinking but an increased risk with heavier drinking).
Forum members considered this to be a well-done analysis. They noted the inability of the authors to evaluate differences in effect according to type of beverage consumed, the pattern of drinking, or the underlying folate levels of subject, all of which probably modify such a relation. The results are in line with earlier reports on alcohol and breast cancer, where alcohol appears to increase the incidence of the disease but does not increase mortality. For most diseases, including colorectal cancer, there may a J-shaped effect on mortality: a reduction in risk for light-to-moderate drinking but an increase with heavier drinking.
Overall, the present meta-analysis supports a finding of increased risk of death from colorectal cancer from heavy drinking. However, it shows rather convincingly that light to moderate amounts of alcohol do not increase the risk of death from this disease, probably because of the protective effects of moderate drinking on cardiovascular disease, a more common cause of mortality.
* * *
Comments on this paper have been provided by the following members of the International Scientific Forum on Alcohol Research:
David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa
Fulvio Ursini, MD, Dept. of Biological Chemistry, University of Padova, Padova, Italy
Dag S. Thelle, MD, PhD, Senior Professor of Cardiovascular Epidemiology and Prevention, University of Gothenburg, Sweden; Senior Professor of Quantitative Medicine at the University of Oslo, Norway
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Creina Stockley, PhD, MBA, Clinical Pharmacology, Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia
Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner,
Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy
R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA