Cho E, Lee JE, Rimm EB, Fuchs CS, Giovannucci EL. Alcohol consumption and the risk of colon cancer by family history of colorectal cancer. Am J Clin Nutr 2012;95:413–419.
Background: Individuals with a family history of colorectal cancer may be more susceptible to adverse effects of alcohol consumption.
Objective: We investigated whether the association between alcohol consumption and colon cancer risk differed by family history of colorectal cancer.
Design: We conducted prospective studies in women and men in the Nurses’ Health Study and Health Professionals Follow-Up Study, respectively. Alcohol consumption was first assessed in 1980 in women and in 1986 in men.
Results: During a follow-up of 26 y among 87,861 women and 20 y among 47,290 men, we documented 1,801 cases of colon cancer (1,094 women and 707 men). Higher alcohol consumption was associated with an elevated risk of colon cancer, although the association was significant only for the highest intake category of ≥ 30 g/d, with no significant linear trend. The association between alcohol consumption and colon cancer risk differed by family history of colorectal cancer; in comparison with nondrinkers, the pooled multivariate RRs for alcohol consumption of ≥ 30 g/d were 1.23 (95% CI: 0.96, 1.57; NS) among those with no family history and 2.02 (95% CI: 1.30, 3.13) among those with a family history of colorectal cancer (P value test for difference = 0.05). In comparison with nondrinkers with no family history, the RR for colon cancer was 2.80 (95% CI: 2.00, 3.91) for individuals who consumed ≥ 30 g/d and who had a family history of colorectal cancer.
Conclusion: Reducing alcohol consumption may decrease the incidence of colon cancer, especially among those with a family history of colorectal cancer.
Background: Data are inconsistent as to whether alcohol consumption increases the risk of colon cancer and, if so, if there is a threshold level of drinking for such an effect. The present analysis combined data from two large cohort studies, seeking to determine the association between alcohol consumption and the risk of color cancer by whether or not there is a family history of such cancer. The study was based on more than 87,000 women and 47,000 men in the Nurses’ Health Study and the Health Professionals Follow-up Study, respectively. A total of 1,801 cases of colon cancer were documented among these cohorts during follow-up periods averaging more than 20 years.
Comments on the results of the study: Significant increases in risk of colon cancer were reported for the highest category of alcohol consumption, but the average intake among such subjects was not given. As for interpreting their conclusions, it is of some concern that there was not a dose-response relation between alcohol and colon cancer. In fact, the Nurses’ Health Study showed higher risks for the women in the lightest drinking group (0.1-<5 g/day) than for most other drinking categories except for the top category of 30 g/day or more of alcohol per day. This pattern was not seen in the Health Professionals’ Follow-up Study.
The highest alcohol groups of subjects reported a cumulative average of ≥ 30 grams of alcohol per day, which suggests that they were long-term rather heavy drinkers. Further, subjects in this group were shown to have the greatest intake of red meat, smoke the most, and have the lowest intake of folate. There is a question whether or not multivariate analyses adequately adjusted for these variables; with such large differences in other factors associated with cancer risk, it is difficult to determine the specific role of each factor. The investigators did not allow for a “lag period” between the time period of alcohol consumption and the diagnosis of colon cancer, which is often done to adjust for potential changes in alcohol intake in a specified period before the diagnosis is made.
Strengths of the study include the availability of frequent updates of alcohol consumption and good assessment of disease, with confirmation of the diagnosis of colon cancer. On the other hand, Forum reviewers were concerned that only limited data were available on the pattern of drinking. The authors state that “there was an insignificant increase in risk from 1-2 drinking days to 5 or more drinking days per week,” but the authors do not say if they controlled for total alcohol; also they do not mention whether or not subjects reported binge drinking.
As one Forum reviewer commented: “This paper supports the existence of a limit of intake below which an association of alcohol and colon cancer has not been shown, and confirms the dangers of the combination of a low folate intake and a high alcohol intake in relation to cancer risk.” The reviewer added: “The problems of residual confounding are always present in observational studies, but the present paper seems to have adjusted for multiple risk factors as well as it is possible to do.”
Other comments on the paper by Forum reviewers: One Forum reviewer commented: “It is useful to have data such as these on alcohol and colon cancer. At first thought it might seem obvious that a personal family history of cancer would increase the risk of consuming alcohol for that person. However, in a recent case-control study of 1,925 matched pairs of women who carry a BRCA1 or a BRCA2 mutation, a modest inverse association between reported current alcohol consumption and breast cancer was observed among women with a BRCA1 mutation (OR= 0.82, 95% CI 0.70 – 0.96), but not among women with a BRCA2 mutation (OR= 1.00; 95% CI 0.71 -1.41).1” Hence, differences in the association of alcohol and colon cancer according to a positive family history of such cancer are of particular interest.
Another Forum reviewer stated: “The groups with the highest alcohol intake had by far the highest percentage of current smokers, and had the lowest folate intake (which should indicate a lower intake of fruits and vegetables). Thus, these people have the most unhealthy lifestyles of the populations studied, and it is not strange that their colorectal cancer incidence is the highest. I do not think that any adjustment for these confounding factors is sufficient to conclude that higher alcohol consumption per se is so important in the pathogenesis of colorectal cancer, or that the authors conclusion that reducing alcohol consumption may have a large effect on decreasing the incidence of colorectal cancer.”
Relation of results of paper to current cancer screening guidelines: A Forum reviewer commented: “The study does not stratify subjects with a positive family history into Lynch 1 or Lynch 2 categories, which might be more helpful in determining a subset association. Further, the study does not take into account that endoscopy was less common in the early eighties and subjects in the at-risk group were probably not receiving consistent screenings in the decade of life (the forties) that we now know may be when they are at highest risk for diagnosis; the at-risk group appears to have been observed in their 50s. I would argue, using data based on current American Cancer Society guidelines for colon cancer screening in first degree relatives of index cases, that the results may not be the same if the associations were evaluated now, because many cancers might have been prevented.
“Thus, I think that it is very hard to create conclusions on data for colon cancer that is 30 years old, because the new recommendation for screening frequency and age at initiation of screening is so different now. At least some of these cases would probably have been prevented if managed according to current guidelines.”
Reference from Forum comments
1. Dennis J et al. Alcohol consumption and the risk of breast cancer among BRCA1 and BRCA2 mutation carriers. The Breast 2010;19:479-83.
This paper presents an analysis based on a large number of subjects being followed in the Nurses’ Health Study and the Health Professionals Follow-up Study, relating alcohol consumption to the risk of colon cancer according to whether or not the subjects had a positive family history of colon cancer. Their results indicate that subjects with a family history whose average alcohol intake was 30 or more grams per day (about 2 ½ typical drinks by US standards) have an increase in their risk of colon cancer; there was not a significant association between alcohol and colon cancer among subjects without a positive family history.
Forum reviewers were somewhat concerned that the pattern of drinking (regularly or binge drinking) was not assessed, and that there was not a clear dose-response curve between the level of alcohol consumption and the risk of cancer (i.e., there was not a consistent increase in risk of cancer with greater alcohol intake). Further, folate intake was found to modify the association, with the highest risk for subjects with a positive family history of colon cancer, low levels of folate, and in the highest category of alcohol consumption.
The present study provides some support for an association between higher levels of alcohol intake and the risk of colon cancer among subjects with a positive family history of such cancer. However, there have been changes in the guidelines for screening for cancer (by endoscopy, with removal of pre-malignant tumors) and other preventive measures for people with a positive family history of colon cancer. Such measures could modify the effects of all risk factors for colon cancer in future analyses.
* * *
Comments on this critique were provided by the following members of the International Scientific Forum on Alcohol Research:
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Lynn Gretkowski, MD, Obstetrics/Gynecology, Mountainview, CA, Stanford University, Stanford, CA, USA
Dee Blackhurst, PhD, Lipid Laboratory, University of Cape Town Health Sciences Faculty, Cape Town, South Africa
R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA
Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Gordon Troup, MSc, DSc, School of Physics, Monash University, Victoria, Australia