The new alcohol guidelines explained

The UK Chief Medical Officers have published new guidelines for low-risk levels of alcohol consumption.  This brief guide will explain what they mean.


The revised guidelines recommend that:

Men and women should not regularly drink more than 14 units of alcohol a week. Ideally, this should be spread evenly over three days or more.

Drinkers should limit the amount they consume on single occasions, and intersperse drinking alcohol with eating food and drinking water.

Pregnant women should avoid drinking altogether.

The full guidelines, alongside a consultation document on their wording, are available here.


The following is general guide – actual levels will vary depending on the strength of the drink and the size of the serving:


6 standard glasses of wine at around 14% ABV

Pint Beer Glass on WhitePint Beer Glass on WhitePint Beer Glass on WhitePint Beer Glass on WhitePint Beer Glass on WhitePint Beer Glass on White

6 pints of beer at around 4.5% ABV


7 440ml cans of lager at around 4.8%ABV

Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35Screen Shot 2016-01-07 at 09.23.35
7 double shots of spirits



The new guidelines recognise that while long-term alcohol-health risks are generally higher for women than men, men face much higher risks of acute harm (e.g. injury) on single drinking occasions.  Therefore, it has been decided to set the lower-risk level at the same amount for both.


The previous guidelines recommended that women avoid drinking when pregnant, but if drinking consumed no more than one or two units once or twice a week.  This was based on the current evidence that while heavy drinking in pregnancy creates serious risks of foetal alcohol disorders, drinking at very low levels during pregnancy has not been shown to be harmful (Science and Technology Committee, 2011: para 33). That evidence has not substantially changed, but the new guidelines adopt a more precautionary approach, with the the view that a simple no-drinking message will avoid confusion and that a zero-risk approach to drinking in pregnancy is appropriate.


Following a recommendation from the Science and Technology Committee, the 2012 Alcohol Strategy included a commitment to reconsider the existing guidelines on safe levels of drinking, which were last revised in 1995. The subsequent review considered a wide range of new research on alcohol-related health risks, commissioned new analyses of existing data on risk, and reflected on ways that guidelines have been developed internationally.

The revised guidelines reflect recent research that has identified associations between alcohol and some forms of cancer, even at low levels of consumption.  It also takes account of greater scepticism in the research community towards claims regarding the protective effects of low levels of alcohol consumption.


Drinking guidelines have a long history. In 1870, the British doctor Francis Anstie suggested that people shouldn’t drink more than 1 1/2 ounces of pure alcohol per day.  ‘Anstie’s Limit’ became very influential in the early 20th century, but later fell into disuse. The development of guidelines re-emerged in the 1980s, following publication of a Government report entitled Prevention and Health: Drinking Sensibly in 1981.

Between 1986 and 1987, the Royal Colleges of Physicians, Psychiatrists and General Practitioners all proposed recommended limits of 21 units a week for men and 14 units a week for women, a figure that was formally adopted by the Government in 1987. In 1995 the UK guidelines were revised to daily limits of 3 units for women and 4 units for men, partly to reflect research on links between regular moderate drinking and heart disease. However, there was concern that daily limits implied daily drinking was advisable, and in the light of new research on health risks at lower levels of consumption it was felt the guidelines needed reconsideration (Science and Technology Committee, 2011).


There is no research evidence to demonstrate that safe drinking guidelines change behaviours (Room and Rehm, 2012: 136; Science and Technology Committee: 28). There is the risk that when guidelines are too low, they may simply be ignored as impractical or not credible.  Indeed, recent research among young people suggests generic guidelines are widely ignored as they bear little relationship to way people drink, their attitudes to drinking, or how they calculate risk (Lovatt et al., 2015). It is possible that a greater degree of public knowledge around units and guidelines has contributed to the decline in consumption across the UK since the mid-2000s, but it has not been demonstrated by any social research – and recent studies suggest that while the majority of us are aware of units and guidelines, only a small minority actively use them to monitor their drinking (de Visser and Birch, 2012). The limited effectiveness of guidelines has led some researchers to call for them to be abandoned altogether (Casswell, 2012).

Another problem with implementing units-based guidelines is that most of us simply don’t know how many units are in the drinks we consume, especially when we pour them for ourselves. This has been demonstrated in a number of research studies (de Visser and Birch, 2012; Boniface et al. 2013). If the new guidelines are to be effective, more work will need to be done to help us better understand how much alcohol is in the drinks we consume.


The question of whether moderate alcohol consumption has a protective effect on the heart remains the subject of dispute. Large scale studies suggest that moderate drinkers are less likely to suffer ischaemic heart disease (and, indeed, have lower overall mortality) than either heavier drinkers or abstainers. However, the reasons for this remain unclear (see e.g. Fernandez-Sola, 2015; Hall, 2012). It may be partly due to a biological benefit of alcohol; partly because many non-drinkers have existing health problems that skew the figures; or partly because people who drink moderately tend, on average, to have healthier lifestyles than either heavier drinkers or abstainers (non-drinking is associated, for example, with economic deprivation).

In any respect, the protective effects of alcohol cease to be a factor where people drink more heavily, even occasionally. In other words, if you often drink more than one or two pints at a time then any benefits of moderate drinking are likely to be outweighed.

The protective effects of alcohol also only apply to older drinkers. The new guidelines suggest this effect may, in fact, be limited to women over the age of 55 – though there is not a consensus on this in the research literature.


Yes, it does. How alcohol affects individuals depends on a wide range of factors including genetic make-up, diet, general health, age and so forth.  One issue not explicitly addressed in the new guidelines is the so-called ‘harm paradox‘: the fact that alcohol-related deaths are much higher in deprived communities than affluent ones – despite levels of consumption being, on average, similar. In principle, a more complex set of guidelines could be developed that more fully reflect social diversity; however, this could make advice harder to communicate. The new guidelines opt for a broader approach, based on research evidence for minimum levels of risk, which has the advantage of allowing simple health messages to be developed.


The revised guidelines are a recommendation for reducing the risk of suffering alcohol-related health harms. They reflect the fact that, for some conditions, risks start to increase at even low levels of consumption. However, the absolute risk of suffering an alcohol-related disease when drinking at these levels is small: the new guidelines are partly based on estimated levels of consumption that would create ‘around or just below’ a 1 in 100 chance of dying from an alcohol-related cause, compared to not drinking.  The issue, therefore, is what level of risk is considered acceptable.

The risk calculations in the guidelines are complex, and news reporting is often poor at describing risk well.  There are some helpful resources on understanding absolute and relative risks, and how to navigate media coverage, by Sir David Spiegelhalter here.

The revised guidelines provide a useful benchmark against which to make informed choices regarding the level of risk one wishes to accept in relation to alcohol. How well they are understood by the public at large, and what affect they have on behaviours, remains to be seen.