Wednesday, 09 May 2018 15:51

Media and scientific evidence, what about moderation?

In April, a meta-analysisRisk thresholds for alcohol consumption: combined analysis of individual-participant data for 599 912 current drinkers in 83 prospective studies” with more than 100 authors, among them many renowned scientists, was published in the well-known journal The Lancet.

The Lancet issued a press release to communicate the results of the study: “Alcohol limits in many countries should be lowered, evidence suggests

Examining the original data more carefully, including those in the supplementary appendix of the publication, the established J-shaped risk curve and thus, the health benefits of light to moderate (wine) drinking were confirmed.

Furthermore, the importance of drinking patterns (i.e., moderately with the meals versus binge drinking) was emphasized.

Then why did the study authors ask for lower drinking guidelines?

  • In the main publication and abstract, the reference group did not include – as conventionally done – lifelong abstainers (people that never consume alcohol) but individuals who, according to their own account, consumed a little alcohol.
  •  It should be also noted that, in this type of study, where questions are asked about the individuals’ alcohol consumption, under-reporting is quite common (i.e., less alcohol consumption is reported than is actually consumed) and it can be assumed that many “moderate” drinkers were included in the reference group.

 However, in separate analyses which are available in the supplementary appendix of the publication, where lifelong abstainers as well as the former drinkers were included, different results were observed.

 „Cut-off“ J-curve. For decades, large epidemiological studies with millions of participants have shown a J-shaped risk association. The term describes the relationship between alcohol consumption and mortality, where individuals who drink moderately, live longer and have a lower risk of certain diseases (i.e., cardiovascular disease, type 2 diabetes and potentially cognitive decline) than those who drink too much but also compared with those who are abstainers. The biological plausibility of these associations has also been determined - in particular for cardiovascular disease and type 2 diabetes.

 If you take – as it was obviously done in this case – very moderate drinkers as baseline (where a “protective effect” should already be noticed), it is no surprise that the J-curve disappears, it is practically “cut-off”.

Figure e10; Suppl. S. 31

What about the drinking patterns?

When analysing the drinking frequency, it was observed that mortality risk increased (dose dependent) considerably more when the weekly alcohol consumption was concentrated only on one to two drinking occasions. In contrast, when the drinking was distributed evenly throughout the week, only at 250 g of alcohol/week (20 drinks) was an increased risk was found.

When examining the drinking pattern, an increased dose-dependent risk was seen for binge drinkers (excessive consumption, > 100 g per drinking occasion). However, without such a binge drinking pattern, the risk increased only at intake levels between 200 and 300 g of alcohol/week.

This confirms the evidence from earlier studies that with regular but moderate consumption, the total mortality risk was reduced.


What about the type of alcoholic beverage?

When the type of the preferred beverage consumed was taken into account, previous existing evidence was confirmed:

  • With wine consumption, the total mortality risk did not increase up to 14 drinks (175 g of alcohol) per week. A protective effect was observed on the heart attack risk.
  • In contrast, the same level of alcohol in the form of beer and spirits significantly increased the mortality risk.

The study authors assumed that these results may be due to the particular drinking and eating pattern of the wine drinkers. It is known from other studies that there is a noticeable difference in health outcomes when the alcoholic beverage is consumed with versus without a meal.

The Mediterranean dietary pattern, where wine is part of the meal, is considered the healthiest eating pattern. This is reflected in the generally lower disease and mortality rates.

Figure e17; Suppl. S. 38


Analysing ALL of the study results, there is no reason to doubt the well-established J-curve and consequently the established drinking guidelines. It is also clear that is not only about the amount consumed but how the wine is consumed, ie. it is preferable to consume wine regularly and moderately with the meals than drinking the same amount on 1-2 occasions (binge drinking).

Considering the above analysis and a continuous assessment of the scientific data, it appears that the drinking guidelines used in Wine in Moderation are compatible with the current scientific evidence and in line with the international average of drinking guidelines.


Wood A. et al., Emerging Risk Factors Collaboration/EPIC-CVD/UK Biobank Alcohol Study Group Lancet. 2018 Apr 14; 391(10129): 1513–1523. doi: 10.1016/S0140-6736(18)30134-X


The authors have taken reasonable care in ensuring the accuracy of the information herein at the time of publication and are not responsible for any errors or omissions. Read more on our disclaimer and Privacy Policy.