Alcohol Guidelines - Science and Technology Committee Contents

Written evidence submitted by the Sheffield Addiction Research Group at the University of Sheffield (AG 11)

1.  What evidence are Government's guidelines on alcohol intake based on, and how regularly is the evidence base reviewed?

1.1  The evidence base for alcohol intake guidelines is primarily epidemiological studies relating alcohol consumption at different levels to the risk of various harms which have been shown to result from drinking. Such studies can be used to identify at what level of alcohol consumption the risk of harm begins to increase. The evidence base addresses both individual harms and "all cause mortality" (the total risk of premature death from any harm).

1.2  Different measures of risk

1.2.1  There are different ways of reporting the risk associated with levels of alcohol consumption. Amongst the most common ways are the use of absolute risk and relative risk measures. The absolute risk of experiencing harm is the simple probability that harm will be experienced at a given level of consumption. The relative risk is the probability of experiencing harm as a proportion of the risk associated with a reference category (usually abstention).

1.2.2  Both approaches have advantages and drawbacks. The absolute risk approach gives a clear indication of the level of risk associated with a particular behaviour (eg drinking x units a day is associated with a y% risk of premature mortality). However, it gives no comparison of how much greater this risk is than that experienced by non-drinkers. This can produce counterintuitive results whereby an individual belonging to a low risk group may perceive their high levels of consumption to be less harmful than if they were in a high risk group with lower consumption. Moreover, when setting a level of consumption which should not be exceeded, the threshold is typically based on a choice of absolute risk level which is essentially arbitrary (eg although the threshold could be set at a 1% risk of premature mortality, it could just as easily be 2%, 5% or 10%).

1.2.3  In contrast, the relative risk approach provides a measure of how much greater the risk of drinking at a particular level is compared to abstinence (eg drinking x units a day is associated with a y times greater risk of premature mortality than abstinence). Therefore, it focuses more clearly on the increase in risk which is actually associated with consumption. However, relative risk measures do not indicate whether the risk of higher levels of consumption is large in absolute terms or just large relative to abstinence.

1.2.4  As noted above, when using absolute risk measures guidelines for alcohol consumption can be derived by selecting a threshold level of risk and recommending not drinking at levels above those associated with that risk. When using relative risk measures there are two possible ways of deriving guidelines. Firstly, the threshold can be set at the consumption level above which the increase in risk of harm, relative to abstinence, becomes statistically significant. Secondly, and noting the protective effects of moderate consumption (see below), the threshold can be set at the consumption level above which the increase in risk of harm is statistically significantly higher than the greatest protective effect.

1.3  How up to date is the evidence base?

1.3.1  Studies of the relationship between alcohol and harm are frequently subjected to systematic reviews in order to identify those studies of the highest quality and to aggregate the effects into a more robust overall estimate of the relationship. Systematic reviews are carried out on a regular basis for individual harms[8],[9],[10] and for all-cause mortality.[11],[12] This evidence is also compiled in the WHO's work on the burden of total disease which is due to alcohol.[13] Therefore, the evidence on which guidelines are based can be considered as of the highest quality available and is reviewed and updated on a regular basis.

1.3.2  Recent examples of relative and absolute risk estimates for different levels of consumption taken from systematic reviews are provided in figures 1 and 2 below.

Figure 1


Figure 2


2.  Could the evidence base and sources of scientific advice to Government on alcohol be improved?

2.1  We identify several weaknesses in the current evidence base which should be taken into consideration when formulating drinking guidelines.

2.2  The type of harm experienced and its link to different types of drinking

2.2.1  Elevated levels of alcohol consumption are associated with increased risk of both chronic and acute health-related harms. Chronic harms are conditions such as liver cirrhosis, coronary heart disease and cancers and the risk of having such conditions has been shown to be higher for individuals with higher levels of average alcohol consumption over a given time period. Acute harms include injuries and alcohol poisoning and these harms have been shown to be more closely associated with higher levels of recent alcohol consumption (ie the level of intoxication). The evidence base for individual harms is therefore divided between the risks of harm from an individual's average alcohol consumption and also their risk from individual episodes of alcohol consumption.

2.2.2  As drinkers are more likely to be interested in their risk of mortality from any cause, rather than from a particular harm, guidelines tend to be based on studies of all cause mortality. This presents a problem as harms associated with average and episodic consumption are aggregated. Average consumption is therefore serving as a less accurate proxy measure of the level of episodic consumption.

2.2.3  Evidence should recognise that different types of harm are the results of different types of drinking. Where possible, evidence should link acute harms to episodic drinking and chronic harms to average consumption although the limitations of data do not always permit this. Guidelines should also reflect this distinction and recommend upper limits for both typical and episodic consumption. Using the current guidelines as an example, this would give the recommendation that "men should not regularly drink more than 3-4 units of alcohol a day and not more than 8 units in any single drinking session".

2.3  The possible protective effects of moderate drinking

2.3.1  Drinking guidelines must take into account the evidence of a protective effect from moderate alcohol consumption.[15] This is seen in all cause mortality studies and largely stems from an apparent reduced risk of coronary heart disease for moderate drinkers compared to non-drinkers. This protective effect is not seen at higher levels of alcohol consumption and, importantly, there is also no protective effect seen in moderate drinkers who also have regular heavy drinking occasions.[16]

2.3.2  Recent work has questioned the validity of the protective effect after a review found studies which pooled "never drinkers" and "former drinkers" into a single "non-drinker" category saw larger protective effects than those which separated never and former drinkers.[17] This suggests a proportion of the mortality risk associated with abstinence may be due to the inclusion of abstainers who are in fact former heavy drinkers who now suffer from alcohol-related coronary heart disease. If ex-drinkers are excluded, a small cardio-protective effect of moderate consumption remains for older age groups. However, age-differentiated drinking guidelines would likely make the message very complex and should probably be avoided.

2.4  The reliability of survey data on consumption

2.4.1  Many studies of the risk of alcohol consumption are based on survey data which ask respondents to report their consumption. Such reports are known to substantially underestimate the amount of alcohol believed to be consumed based on sales data, by between 40% and 60%. Although efforts have been made to explain and address this problem,[18] many estimates of the risk from alcohol consumption may be biased upwards by under-reporting of heavy consumption.

2.4.2  Further efforts are therefore required to account for the under-reporting and under-estimation of alcohol consumption and to minimise the extent of bias on estimates of the risk from alcohol consumption.

2.5  The influence of factors other than drinking which are not included in analyses

2.5.1  Most studies make some attempt to take account of factors other than alcohol consumption which may lead to harms or premature mortality. For example, studies will typically account for the fact that men are often at greater risk than women and that risk of mortality increases with age. However, other important influences are less consistently taken account of, such as socioeconomic status, broader health status and a range of lifestyle factors (eg diet, physical activity and working environment). As many of these factors are also related to alcohol consumption (eg those who drink more tend to be less active and have a poorer diet) failure to account for them may mean their effects on mortality or harm are incorrectly attributed to alcohol consumption. In particular, these factors have been identified as potentially leading to overestimation of the protective effects of alcohol at moderate levels of consumption.[19]

2.5.2  Again, further work is required to establish the extent to which estimates of the level of risk due to alcohol consumption are biased by factors not included in analyses.

2.6  Protection of data

2.6.1  We also note that the NHS Information Centre recently withdrew its contribution to the funding of the General Lifestyle Survey (GLS). This has meant no further data will be collected on alcohol and other health issues. The GLS (and its predecessor the General Household Survey) provided a valuable source of in-depth data on alcohol consumption in the UK stretching back over four decades. They are recognised as the best source of data on UK consumption in the UK and are the basis for key government reports on consumption and many academic studies. Failing to protect such data sources greatly weakens the ability of the scientific community to advise the Government on alcohol.

2.6.2  The GLS also collected longitudinal data on the alcohol consumption of a panel comprising of the same individuals at several time points; however, due to funding shortages, this data is not being made available to researchers. This is a wasted opportunity as no other panel study in the UK has collected such rich data on alcohol consumption and other health-related variables. Again, failure to fund the release of valuable consumption data is a major hindrance to the scientific community's efforts to provide authoritative advice to the Government.

3.  How well does the Government communicate its guidelines and the risks of alcohol to the public?

3.1  The NHS Information Centre reports on data collected in two surveys examining adults' awareness, recall and adherence to drinking guidelines.[20]

3.2  The findings show that awareness of units of alcohol is high with 90% of respondents in 2009 having heard of units and this has risen from 79% in 1997. Respondents were slightly less likely to have heard of drinking limits with 75% having done so, up from 54% in 1997. However, it is concerning that only 35% of men and 43% of women claimed to know what the recommendations which applied to them were. Furthermore, the NHS Information Centre report notes that those respondents who attempted to state the recommendations were more often wrong than right.

3.3  In addition to being aware of the guidelines, adherence to guidelines require drinkers to also monitor their consumption which means being aware of the number of units in drinks they consume and, as necessary, keeping a check on their daily or weekly consumption. Although knowledge of the number of units in preferred drinks was relatively high at around 60%-70%, just 13% of drinkers kept a check on their consumption in 2009. This may many individual's belief that their consumption is at a moderate level, however it also suggests conscious adherence to drinking guidelines is rare.

3.4  Heavier drinkers were more likely to have heard of units and drinking limits and frequent drinkers were also more likely to be able to correctly identify the unit content of drinks. This suggests information campaigns have been successful in ensuring those drinkers at the greatest risk are more aware of that risk. It may also suggest that including the unit content on labels eventually leads to retention of this information. However, other information was less successfully communicated as heavy drinkers were no more likely to recall what the recommended drinking guidelines were. Heavier drinkers were also only slightly more likely to monitor their alcohol consumption.

3.5  Overall, these findings suggests the Government has been increasingly successful in communicating the idea of drinking guidelines and some of the concepts around it, such as units and unit content of drinks. Success has also been highest in the target group of heavier drinkers. However, it can claim less success in communicating recommended levels of consumption, knowledge of which was low for all groups.

3.6  The use of information campaigns has led to greater public awareness of the risks associated with drinking and how these relate to the consumption of different types and quantities of alcoholic drinks. However, there is little evidence this knowledge has impacted on behaviour. Whilst it is important that information campaigns continue to inform the public about the risks associated with drinking, information alone may have limited utility in motivating drinkers to reduce their consumption. To achieve reductions in consumption, further action may be required both by making explicit reference to the harms which may be incurred if drinking guidelines are exceeded and taking additional action beyond information campaigns.

4.  How do the UK Government's guidelines compare to those provided in other countries?

4.1  The attached document produced by Alcohol in Moderation entitled "Sensible Drinking Guidelines" gives recommended drinking guidelines for a selection of developed nations.[21] As different nations define a unit of alcohol differently or base guidelines upon the notion of a "standard drink", it is easier to compare guidelines after converting recommended levels into pure alcohol consumption in grams. Some nations give guidelines as average daily thresholds whereas others use average weekly thresholds. For simplicity here, we only compare those with average daily thresholds.

4.2  The UK guidelines recommend not regularly drinking more than 24-32g of pure alcohol a day if you are a man and not more than 16-24g if you are a woman. These levels are similar to those used in many other nations such as Italy (24-36g and 12-24g), the USA (24g and 14g), France (30g and 20g), Germany (36g and 24g) and New Zealand (30g and 20g). Some nations do have slightly higher recommendations, particularly for men, such as The Netherlands and Spain (both 40g and 24g). Few nations have significantly lower guidelines and those that do include Denmark (21g and 14g), Poland (20g and 10g) and Slovenia (20g and 10g).

4.3  The UK drinking guidelines can be considered as in line with other developed nations and there appears no case to be made for altering them on the basis of international consensus.

5.  Acknowledgements

5.1  The Sheffield Addiction Research Group would like to thank Tim Stockwell of the University of Victoria, Canada for his assistance in producing this document.

6.  Declaration of interests

6.1  The Sheffield Addiction Research Group do not have any potential conflicts of interest and receive no funding from the alcohol industry.

September 2011

8   Corrao, G, Bagnardi, V, Zambon, A & Arico, S (1999). "Exploring the dose-response relationship between alcohol consumption and the risk of several alcohol-related conditions", Addiction, 94 (10) pp.1551-73 Back

9   Corrao, G, Bagnardi, V, Zambon, A & La Vecchia, C (2004). "A meta-analysis of alcohol consumption and the risk of 15 diseases", Preventive Medicine, 38 (5) pp.613-9 Back

10   Taylor, B, Irving, H M, Kanteres, F, Room, R, Borges, G, Cherpitel, C, Greenfield, T, & Rehm, J (2010). "The more you drink, the harder you fall: A systematic review and meta-analysis of how acute alcohol consumption and injury or collision risk increase together", Drug and Alcohol Dependence, 110 pp.108-16 Back

11   Holman, C D, English, D R, Milne, E & Winter, M G (1996). "Meta-analysis of alcohol and all-cause mortality: a validation of NHMRC recommendations", The Medical Journal of Australia, 164 (3) pp.141-5 Back

12   Di Castelnuovo, A, Costanzo, S, Bagnardi, V et al (2006). "Alcohol Dosing and Total Mortality in Men and Women: An Updated Meta-analysis of 34 Prospective Studies", Archives of Internal Medicine, 166 pp.2437-45 Back

13   Rehm, J, Baliunas, D, Guilherme, L G et al (2010). "The relation between different dimensions of alcohol consumption and burden of disease: an overview", Addiction, 105 pp.817-43 Back

14   NHMRC (National Health and Medical Research Council) (2006) "Australian Guidelines to Reduce Health Risks from Drinking Alcohol", available at accessed 12 September 2011 Back

15   Ronksley, P E, Brien, S E, Turner, B J, Mukamal, K J & Ghali, W A (2011). "Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis", British Medical Journal, 342, d671, DOI: 10.1136/bmj.d671 Back

16   Bagnardi, V, Zatonski, W, Scotti, L, La Vecchia, C & Corrao, G (2008). "Does drinking pattern modify the effect of alcohol on the risk of coronary heart disease? Evidence from a meta-analysis", Journal of Epidemiology and Community Health, 62 pp.615-9 Back

17   Fillmore, K M, Kerr, W C, Stockwell, T, Chikritzhs, T & Bostrom, T (2006). "Moderate alcohol use and reduced mortality risk: Systematic error in prospective studies", Addiction Research and Theory, 17, (5 Supplement) S16-23 Back

18   Meier, P S, Meng, Y, Holmes, J, et al (2011). "Alcohol policy appraisals: The effect of adjusting survey and per capita consumption estimates for known biases", paper presented at 37th Annual Alcohol Epidemiology Symposium of the Kettil Bruun Society, Melbourne, Australia, 11-15 April 2011 Back

19   Chikritzhs, T, Fillmore, K & Stockwell, T (2009). "A healthy dose of scepticism: four good reasons to think again about protective effects of alcohol on coronary heart disease", Drug and Alcohol Review, 28 pp.441-4 Back

20   NHS The Information Centre (2011) "Statistics on Alcohol: England, 2011", available at accessed 2nd September 2011 Back

21   Alcohol in Moderation (2011) "Sensible drinking guidelines" available at accessed 2 September 2011 Back

previous page contents next page

© Parliamentary copyright 2012
Prepared 9 January 2012