Letters

Association cannot be assumed to be causal

Neither study provides convincing evidence that alcohol protects against heart disease. Association cannot be assumed to be causal, but the authors of both articles make this assumption. Eric B Rimm and colleagues write that "a substantial portion of the benefit is from alcohol."² Hans Ole Hein and colleagues discuss attributable risk among men who abstained from alcohol; this term is appropriate only to a causal relation.³

There are reasons to doubt a causal relation. In Hein and colleagues' study non-drinkers were older than drinkers, which suggests a lower all cause mortality. Subsequent higher mortality might be attributable to age, which would have a non-linear relation with mortality and therefore be inadequately controlled for in the regression. The authors did not analyse data from the 1971 baseline, and mortality related to alcohol before 1986 may have biased the sample. The authors dismiss the possibility of "sick quitters" causing bias but do not consider a possible "sick non-starter" effect, whereby those prone to heart disease never started drinking. Given that 87% of the non-drinkers had never drunk, this was potentially a much larger source of bias.

Rimm and colleagues base their claim for a causal relation on inconsistent observational data. Of 12 ecological studies cited, seven show a significant beneficial effect of wine, two show a harmful effect of beer, and only one shows a beneficial effect of spirits and one a beneficial effect of beer. Of three case-control studies, only one shows all forms of alcohol to be protective, and one shows all forms of alcohol to be harmful. Of 10 cohort studies, only one shows a consistent significant benefit for all forms of alcohol. Ecological studies are least prone to confounding by sick non-drinkers and are the least supportive of the theory that alcohol may protect against heart disease.

We question the wisdom of publishing these papers without giving any form of public health advice. They will be perceived by the public as giving a signal to consume more alcohol.

Research statistician Senior lecturer Department of Epidemiology and Public Health, Imperial College School of Medicine at St Mary's, London W2 1PG

Luke Whitaker, Helen Ward 

1. Royal College of Physicians. The medical consequences of alcohol abuse: a great and growing evil. London: Tavistock, 1987.
2. Rimm EB, Klatsky A, Grobbee D, Stampfer MJ. Review of moderate alcohol consumption and reduced risk of coronary heart disease: is the effect due to beer, wine, or spirits? BMJ 1996;312:731-6. (23 March.)
3. Hein HO, Suadicani P, Gyntelberg F. Alcohol consumption, serum low density lipoprotein cholesterol concentration, and risk of ischaemic heart disease: six year follow up in the Copenhagen male study. BMJ 1996;312:736-41. (23 March.) [Abstract/Free Full Text]