The review by Britton and McKee(1) asserts a general causal relationship between episodes of high dose alcohol intake (binge drinking) and cardiovascular deaths, based on universal physiological mechanisms. They claim that this causal relationship explains the dramatic fluctuations in cardiovascular deaths in Russia and (less dramatic ones) in Poland. We could not agree more that an assessment of the pattern of drinking should be included in future studies, but we recommend caution in interpreting the evidence presented in their review.

First, the definitions of binge drinking used in the studies included in the review vary widely, and some of them probably do not represent a distinct population of binge drinkers, such as ‘unable to perform job properly' or ‘being registered as a heavy abuser'. Studies using such heterogenous definitions of exposure are not "combinable".(2) These unreliable definitions do not distinguish between binge drinkers and regular heavy drinkers or dependent moderate drinkers, all of which are treated as quite separate populations in their review of mechanisms for cardiac mortality.(3) Combining these studies cannot, in our view, answer the question whether binge drinking is related to cardiovascular deaths.

Second, it is far from clear to which extent the reviewed studies controlled for important confounding factors. It is very likely that people who were "unable to perform their job properly" or were registered as "heavy abusers" also differed from the reference group in many factors that confound the association with cardiovascular disease. Poor diet and malabsorption of nutrients, unsatisfactory employment and social exclusion, low educational achievement, chain smoking or inadequate housing could all co-vary with episodes of high dose alcohol intake. It is unlikely that all such factors could adequately be taken into account by adjusting for a limited set of variables. Third, the specificity of the association between binge drinking and sudden deaths is problematic. The reference cited to support this point in fact examined sustained heavy drinking, rather than binge drinking.(4) Sustained heavy drinking is rare in Russia.(5)

Given all these problems, the statement that the association is "strong and consistent" and "fulfills the criteria of causality" is misleading. There is simply not enough evidence for this.

The origins of the hypothesis that binge increases the risk of cardiovascular death are based on the mortality crisis in Russia. However, it ignores the important dissociation between the mortality trends in men and women and the gender gap in drinking. All data from Russia are consistent in one aspect: men drink much more than women (in fact women in Russia drink relatively little compared to the UK).(5) However, the magnitude of cardiovascular mortality fluctuations has been the same, proportionately, for men and women in Russia.(6) Given the consistently low consumption in women, it is difficult to see how the massive mortality changes could caused primarily by alcohol.(7) This is critical to the argument of Britton and McKee because the hypothesis they developed had the initial aim of explaining the massive mortality changes in Russia.

To date, we are aware of two prospective cohort studies on alcohol and mortality in Russia, neither of which provide any support for the hypothesis. In the Russian Lipid Research Clinic Study cohort, there was no suggestion of an increased mortality in the highest category of alcohol intake (which, given the drinking pattern common in Russia, would probably include binge drinking).(8) In our analysis of the Novosibirsk MONICA cohort, binge drinking (measured as the average alcohol intake on typical occasion) was related to increased risk of death from injuries and accidents but not from cardiovascular disease (submitted manuscript). In the light of the present evidence, it is not likely that binge drinking can explain the paradox of high cardiovascular mortality in Eastern Europe, certainly not on the scale proposed by the reviewers. Acute effects of alcohol may include some cardiovascular deaths but it is unlikely that binge drinking caused the several hundred thousand extra cardiovascular deaths that occurred above the long term rates in Russia between 1987 and 1994.(6)

No one would dispute the gravity of the health crisis in Eastern Europe or the harmful effects of drinking on many health outcomes. However, the current state of knowledge does not (yet?) justify the conclusion that binge drinking causes large numbers of deaths from cardiovascular disease.

Martin Bobak
Department of Epidemiology and Public Health
University College London
1-19 Torrington Place
London WC1E 6BT, UK

Woody Caan Directorate of Public Health
Barking and Havering Health Authority
London, UK

References

1 Britton A, McKee M. The relation between alcohol and cardiovascular disease in Eastern Europe: explaining the paradox. J Epidemiol Comm Health 2000;54:328-332.

2 Egger M, Smith GD, Phillips AN. Meta-analysis: principles and procedures. BMJ 1997;315:1533-1537.

3 McKee M, Britton A. The positive relationship between alcohol and heart disease in eastern Europe: potential physiological mechanisms. J R Soc Med 1998;91:402-407.

4 Wannamethee G, Shaper AG. Alcohol and sudden cardiac death. Heart 1992;68:443-448.

5 Bobak M, McKee M, Rose R, et al. Alcohol consumption in a national sample of the Russian population. Addiction 1999;94:857-866.

6 Leon DA, Chenet L, Shkolnikov V, et al. Huge variation in Russian mortality rates 1984-94: artefact, alcohol, or what? Lancet 1997;350:383-388.

7 Bobak M, Marmot M. Alcohol and mortality in Russia: is it different than elsewhere? Ann Epidemiol 1999;9:335-338.

8 Deev A, Shestov D, Abernathy J, et al. Association of alcohol consumption to mortality of middle aged US and Russian men and women. Ann Epidemiol 1998;8:147-153.