On the Evidence of a Cardioprotective Effect of Alcohol Consumption

We thank Addiction for giving us the chance to continue this discussion, and Dr Stockwell for reiterating his view 1 that our meta-analysis on the cardioprotective association of average alcohol consumption and heart disease 2 is technically excellent. We also agree that the underlying literature, which we consider overall relatively rich rather than poor, has deficits. However, we believe that Dr Stockwell's conclusions based on these deficits are problematic. First, Dr Stockwell might have overlooked that we separated occasional drinkers from lifetime abstainers, former drinkers, and on-average low level drinkers (2.5-12 g/day). The effect estimates, although not all statistically significant, were consistent with a non-linear dose-response relationship with a beneficial effect at low levels of intake 2. A detailed meta-analysis on the effect of former drinkers in comparison with lifetime abstainers has been published 3. However, we do agree that age and drinking over the lifetime should be examined in more detail than they have been in the past. Second, because we have distinguished the drinking groups described above, a comparison with Naimi et al. 4 is problematic. Their paper is based on a comparison between current abstainers (last thirty days) and moderate drinkers and does not take into account former or occasional drinkers, as our study does. Third, we agree that measurement of average intake should ideally cover more than two weeks. We chose to use the two-week limit for our inclusion criteria because we wanted specifically to exclude studies that relied on the prior week's alcohol intake only or used forty-eight-hour intake diaries. Most studies reported average alcohol intake as usual frequency and amount per occasion, many with reference periods such as ‘last year’ or ‘last month’. Again, this might be problematic for other reasons, such as recall bias. That does not mean that the evidence is poor or that the evidence against some beneficial effect is strong. The evidence base for each side has expanded considerably over the last two decades. While we consider the epidemiological evidence for a beneficial effect as strong, we also consider the evidence against it to be weak. Coincidently, the two studies cited by Stockwell 5 as fulfilling his criteria 6, 7 fit well into the current theory regarding the effect of average alcohol intake and drinking pattern on ischaemic heart disease as described in our paper [2, see also 8]. However, many categories of drinking in those two studies had very low numbers of cases, so we wouldn't give them much weight. Both studies provide evidence that drinking patterns showed a substantial differentiation in ischaemic heart disease risk 6, 7, thus questioning the usefulness of measuring average intake alone. Fourth, Dr Stockwell cites several examples of a (small) beneficial effect of alcohol on disease outcomes where one would not expect such an effect. It should also be mentioned that, overall, such ‘protective effects’ are much smaller than the cardioprotective effect. And by pure chance, some observational epidemiological studies indicate the opposite of causality, just as one can find the odd study showing a protective effect of smoking on heart disease. However, because there is no biological pathway indicating that alcohol should be beneficial on, for example, cancer or liver cirrhosis, we consider such effects as not substantiated and would not include them in a comparative risk assessment for alcohol. The relationships mentioned by Dr Stockwell fail to fulfil one of the most important conditions for causality proposed by Hill, which is biological plausibility 9. Epidemiology is only one of several pillars of evidence. Basic research on pathways in animals and humans constitute the other pillars. Together, the meta-analyses of both the epidemiology 2 and the two most important biological causal pathways 10, which can be examined using randomized controlled trials (more rigorous evidence), coincide with the conclusions that regular and moderate alcohol consumption is associated with a lower risk of ischaemic disease [see also 11]. Given the above points and further points made in 1, it seems that some researchers in the field may be using different standards in assessing the cardioprotective effect of alcohol vs. its detrimental effect. Consider two examples. One is the effects of alcohol on colon cancer 12. Would the same arguments used to judge the relationship between alcohol and ischaemic heart disease not hold for this relation as well? The other example is the more than 200 other risk relations between alcohol and disease and injury outcomes 13. Of course, this is not a good argument against scrutinizing the cardioprotective effect of alcohol, but we sense a desire by some in the field to apply tough standards on protective effects and more lenient standards on other effects, where sometimes the responses to very simple survey questions such as ‘Did your partner's alcohol consumption contribute to your marriage problems?’ are accepted as causal evidence. JR received several unrestricted grants from the pharmaceutical industry over the past five years (Schering- Plough, Lundbeck). MR has no potential interests to declare.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.