Critique 090: Effects of stopping alcohol consumption on subsequent risk of esophageal cancer — 13 September 2012

Jarl J, Gerdtham Ulf-G.  Time pattern of reduction in risk of oesophageal cancer following alcohol cessation—a meta-analysis.  Addiction, 107, 1234–1243. doi:10.1111/j.1360-0443.2011.03772.x

Authors’ Abstract

Aim  To establish the current level of knowledge of the effect of drinking cessation on the risk of developing oesophageal cancer.

Method  A meta-analysis was conducted based on relevant studies identified through a systematic literature review.  A generalized least squares model for trend estimation of summarized dose–response data were utilized in order to estimate the effect of years since drinking cessation on risk of oesophageal cancer.

Result  Seventeen studies that estimate the risk reduction after quantified drinking cessation were identified in the systematic literature review.  Nine of these were appropriate for inclusion in the meta-analysis . A large degree of heterogeneity existed between the studies, but this was explainable and the increased risk of oesophageal cancer caused by alcohol consumption was found to be reversible, with a common trend between studies.  A required time-period of 16.5 years (95% confidence interval 12.7–23.7) was estimated until no risk from former drinking remained, although this might have been an overestimation due to sample characteristics.  The dose–response relationship was found to have an exponential decay.  This means that about half the reduction in alcohol-related risk occurred after just a third of the time-period required to eliminate the additional risk.

Conclusion  The alcohol-related increased risk of oesophageal cancer is reversible following drinking cessation.  It is most likely that about 16 years are required until all elevated risk has disappeared.  Due to lack of research and data, more research is urgently required to increase the robustness of the estimates and to approach study limitations.

Forum Comments 

Background:  Alcohol consumption, especially heavy drinking in conjunction with smoking, is related to an increased risk of upper-aerodigestive tract (UADT) cancers, including esophageal cancer.  Several previous studies have suggested that the risk of esophageal cancer decreases among heavy drinkers who stop drinking, with estimates of the number of years required to return to the risk of a non-drinker ranging between 14 and 23 years.1,2

In one study from Hong Kong by Cheung et al,3 based on 400 consecutive admissions of patients with histologically confirmed diagnosis of oesophageal cancer, the authors concluded that “Current light drinking (<200 g ethanol/week) was not associated with significant increase in risk.  With cessation of drinking, risk fell more quickly in moderate (200-599 g/week) than heavy (>/=600 g/week) drinkers.  Even among heavy drinkers, however, risk had dropped substantially after five to nine years of not drinking. The results of that study estimated that the time taken for risk to return to that of subjects who are non-drinkers was 10-14 years for moderate drinkers and 15 years or more, if ever, for heavy drinkers.”3 

A follow-up study from Denmark4 found that an increase in alcohol intake > 14 drinks/week among subjects in a prospective cohort study was associated with significantly elevated risk (hazard ratio = 2.5; 95% confidence interval, 1.1-5.3) of UADT cancers.  Complete cessation of drinking may not be necessary, however, as these investigators report a “suggestive evidence that a decrease in risk occurred among persons lowering alcohol intake > 7 drinks/week (0.5; 0.1-2.5); the trend test was highly significant (p < 0.0001).”4

The present meta-analysis was based on nine studies that provided data to estimate the time between a drinker stops drinking and the time when the individual’s risk of esophageal cancer is lowered to that of a non-drinker.  

Comments on present paper:  Forum reviewers considered the analytic techniques in this study to be appropriate.  Unlike results of some individual studies included (e.g., Rehm et al5), the results of this meta-analysis did not suggest an increase in esophageal cancer risk in the earlier years after quitting drinking.  The overall results of the authors indicate a consistent decrease in risk over time with the cessation of drinking.

Forum reviewer Skovenborg finds it difficult to understand what kind of biological mechanism is behind a 16-year lag time of alcohol cessation and return to normal of oesophagus cancer risk.  (However, other reviewers point out that the decrease in risk appears to begin very shortly after stopping drinking, and the authors estimate that about one-half of the reduction in risk of cancer may occur within about 4 or 5 years.)  Dr. Skovenborg adds: “In Denmark there are 400 new cases of cancer of the oesophagus per year.  If the Population Attributable Risk of alcohol (without smoking) is 0.4%, then a message of stopping drinking to heavy drinkers might eventually save 1-2 persons from that cancer per year.”  Other reviewers also noted that the study did not describe the net health effects of stopping drinking (e.g., effects on cardiovascular disease, diabetes, etc.)

Forum reviewer Finkel, an oncologist, states: “Both my decades of clinical experience and the review of numerous published studies lead me to expect conclusions very much like those of this paper.  I was also immediately reminded of the similar time course of risk following cessation of smoking in the case of lung cancer and other tobacco-related adversities.”  Reviewer Djoussé thought that “It seems peculiar that in this study the risk of cancer after quitting for 16+ years fell below the risk in never drinkers; this makes me wonder whether quitters were adopting other healthful behaviors or whether this is due to the fact that authors used ‘current non-drinkers’ at baseline as reference and not ‘never drinkers’ (the latter may have had even lower risk).”

Interaction of alcohol consumption and smoking on risk of UADT cancers.  While details of baseline drinking levels and the risk of esophageal cancer was not the focus of this paper, many studies suggest little increase in risk at low levels of drinking.  However, it is the combination of smoking and drinking that has the largest effect on the risk of UADT cancers.  As an example, a recent review paper of the effects of smoking and alcohol use on UADT cancers by Szymańska et al6 concluded: “In this largest study on UADT cancer in Latin America, we have shown for the first time that a prevailing majority of UADT cancer cases is due to a combined effect of alcohol and tobacco use and could be prevented by quitting the use of either of these two agents.” Similarly, Anantharaman et al7 concluded that tobacco use alone explained 28.7%, the combination of smoking and drinking 43.9%, and alcohol use alone only 0.4% of the population attributable risk of UADT cancers.  

While concurrent smoking is usually found to be the major risk factor for upper aero-digestive cancers, the authors of the present paper state that they were unable to judge precisely the specific effects of each risk factor as smoking and drinking were so closely related.  They report that smoking “captured mainly the same effect as the definition of former drinkers (and studies with a required time-period before defining a person as a former drinker also generally controlled for smoking).”

An analogy of stopping drinking with studies of health effects of stopping smoking:  Similar analyses as those used in the present paper have been carried out in the past to estimate the health effects of stopping smoking.  The decline in risk of heart disease is much faster than it is for cancer after cessation of smoking.  As summarized by Johns Hopkins Medicine,8 there are almost immediate effects on cardiovascular risk factors from stopping smoking.  This source concludes that “In one year, the added excess risk of heart disease is half that of a smoker,” and some studies suggest the risk of heart disease is even faster, reaching the risk of a non-smoker within one or two years.  However, the time required for a smoker’s risk of certain cancers to return to that of a non-smoker is much longer, and is frequently found to be 20 years or more.


1.  De Stefani E, Munoz N, Esteve J, Vasallo A, Victoria C, Teuchmann S.  Mate drinking, alcohol, tobacco, diet and esophageal cancer in Uruguay.  Cancer Res 1990;50: 426–431.

2.  Castellsague X, Munoz N, De Stefani E, Victora C, Quintana M J, Castelletto R, et al.  Smoking and drinking cessation and risk of esophageal cancer (Spain).  Cancer Causes Control 2000;11:813–818.

3.  Cheng K., Day N., Lam T., Chung S., Badrinath P.  Stopping drinking and risk of oesophageal cancer. BMJ 1995; 310 doi: 10.1136/bmj.310.6987.1094.

4.  Thygesen LC, Keiding N, Johansen C, Groenbaek M.  Changes in alcohol intake and risk of upper digestive tract cancer.  Acta Oncol 2007;46:1085-1089.

5.  Rehm J., Patra J., Popova S. Alcohol drinking cessation and its effect on esophageal and head and neck cancers: a pooled analysis. Int J Cancer 2007;121:1132–1137.

6.  Szymańska K, Hung RJ, Wűnsch-Filho V, Eluf-Neto J, Curado MP, Koifman S, Matos E, Menezes A, Fernandez L, Daudt AW, Boffetta P, Brennan P.  Alcohol and tobacco, and the risk of cancers of the upper aerodigestive tract in Latin America: a case–control study.  Cancer Causes Control 2011;22:1037–1046.  DOI 10.1007/s10552-011-9779-7.

7.  Anantharaman D, Marron M, Lagiou P, Samoli E, Ahrens W, Pohlabeln H, et al. Population attributable risk of tobacco and alcohol for upper aerodigestive tract cancer.  Oral Oncology 2011;47:725–731.

8.  Johns Hopkins Medicine comments, at

Forum Summary

This paper provides an evaluation of the time following cessation of alcohol consumption that the risk of esophageal cancer returns to that of non-drinkers; it is based on 17 studies providing such information, 9 of which provided data for a meta-analysis.  The authors conclude that an alcohol-related increased risk of esophageal cancer is reversible following drinking cessation, most likely requiring up to 16 years.  The authors estimate that about one-half of the reduction in risk of cancer may occur within in a much shorter time, perhaps within about 4 or 5 years.

Forum reviewers considered this to be a well-done analysis.  Forum members emphasized, as did the authors, a number of limitations of the study.  Adjustments for smoking may not have been adequate: most upper aero-digestive cancers show a strong interaction between smoking and alcohol consumption in relation to cancer risk.  (For many “alcohol-related” cancers, there is little effect of alcohol consumption among non-smokers.)

Further, large differences in the alcohol-cancer association were shown in this study for different geographical regions (some associations being much higher in Asia than in Europe or North America), but such differences were not adjusted for in the main analyses.  The fact that the authors of this paper did not have data permitting the separation of ex-drinkers and never drinkers (both groups being included in the “non-drinker” category), and their inability to judge the effects of the baseline pattern of drinking (regular versus binge drinking), may also be limitations to the interpretation of their results.  Adjustment for such factors may have influenced the effects of stopping drinking on subsequent cancer risk, and markedly changed the calculated effects on the numbers of cancers prevented worldwide.

In any case, the fact that cessation of drinking may reduce the risk of esophageal cancer is of importance.  Other studies suggest further that just reducing the amount of alcohol consumed, rather than the complete cessation of drinking, may be associated with lowering of cancer risk, and low-level alcohol intake has been shown to have beneficial health effects on cardiovascular disease, diabetes, and other medical conditions. 

                                                                       *                      *                      *

Contributions to this critique by the International Scientific Forum on Alcohol Research were provided by the following members:

Luc Djoussé, MD, DSc, Dept. of Medicine, Division of Aging, Brigham & Women’s Hospital and Harvard Medical School, Boston, MA, USA

Gordon Troup, MSc, DSc, School of Physics, Monash University, Victoria, Australia

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway

David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa

Andrew L. Waterhouse, PhD, Marvin Sands Professor, Department of Viticulture and Enology, University of California, Davis; Davis, CA, USA

Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark

R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA