Science and Technology Committee HC 1536 Alcohol GuidelinesWritten evidence submitted by the Department of Health (AG 00)
The Evidence Base for Alcohol Guidelines
What evidence are Government’s guidelines on alcohol intake based on, and how regularly is the evidence base reviewed?
1. The Government’s lower risk drinking guidelines were published in the December 1995 report of an Inter-Departmental Working Group, “Sensible Drinking”. The authors of the 1995 report drew upon a wide range of research, including epidemiological evidence, and written and oral advice of experts, as set out in the report. The report considers the harmful effects of alcohol consumption to both health and mortality and considers the evidence for its potential benefits. Sources of evidence included relevant reports of the various Royal Colleges.
2. The 1995 report carefully described the scientific basis for its recommendations, which included review by the authors of the major published research evidence, review of written evidence submitted by a wide range of contributors, independent assessment and critique of the medical and scientific evidence by an external academic statistician, and receipt of oral evidence by invited key experts. The Government continues to monitor the emergence of any major new evidence on risks relating to alcohol consumption, to ensure its guidance remains consistent with current scientific knowledge.
3. In summary, the Government’s lower risk drinking message for the general public on consumption, that was based on the analysis in the 1995 report, is as follows:
men should not regularly drink more than three to four units a day;
women should not regularly drink more than two to three units a day; and
after an episode of heavy drinking, it is advisable to refrain from drinking for 48 hours to allow tissues to recover.
4. The report authors makes clear that they see value in setting population benchmarks to enable people to monitor their own drinking levels but also highlight the importance of appreciating that there will be variation in the impact of the amount of alcohol consumed on different individuals (eg with different body weights). The authors also warn against suggesting a rigid “critical” limit, which they considered would not reflect the scientific evidence.
5. The 1995 Inter-Departmental Working Group included representatives of the then Scottish Office, the Welsh Office, and the Northern Ireland Office. The report was agreed across Government and the Sensible Drinking Message has since been used by administrations across the United Kingdom.
6. The 1995 report included advice on alcohol and pregnancy. This was revised in 2006, when the UK Chief Medical Officers published revised guidelines, taking into account a report commissioned by the Department of Health.
pregnant women or women trying to conceive should avoid drinking alcohol; if they do choose to drink, to minimise the risk to the baby, they should not drink more than one to two units of alcohol once or twice a week and should not get drunk; and
NICE additionally advised that the risks of miscarriage in the first three months of pregnancy mean that it is particularly important for a woman not to drink alcohol at all during that period.
7. The Scottish Chief Medical Officer’s current advice on alcohol and pregnancy is that there is no “safe” time for drinking alcohol during pregnancy and there is no “safe” amount. This advice can be found at www.drinksmarter.org.
8. The 1995 report considered only very briefly the issue of alcohol consumption by children and young people, noting that recommendations for adults are not appropriate for people aged under 16.
9. In December 2009, the Chief Medical Officer for England published specific guidance on the consumption of alcohol by children and young people. The report provides a comprehensive review of the scientific evidence on the links between alcohol-related harm and children and young people. It details key studies from an epidemiological review of the harms associated with adolescent alcohol consumption upon which the guidance is based. It also draws on findings from a review of the associations between alcohol use and teenage pregnancy and consultation with the public, including parents and young people. The new advice was that:
An alcohol-free childhood is the healthiest and best option.
If children do drink alcohol, they should not do so until at least 15 years old.
If 15 to 17 year olds drink alcohol, it should be rarely, and never more than once a week. They should always be supervised by a parent or carer.
If 15 to 17 year olds drink alcohol, they should never exceed the recommended adult daily limits (3-4 units of alcohol for men and 2-3 units for women).
10. The evidence base behind the Government’s lower risk drinking message for the general public has not been the subject of formal review since 1995. The Department of Health’s officials monitor the evidence base and seek additional, independent advice on the developing evidence on health risks linked to alcohol consumption, when required. Such advice was sought, for example, during development of the Department’s 2008 “Know Your Limits” campaign. This included seeking an external expert summary of the relative risks of developing the wide range of key alcohol-related harms, at different levels of consumption, based on the extensive published epidemiological research evidence. This ensured our advice to the public on risks provided in the campaign was up-to-date and reflected the progressive increase in risk of harm seen in those drinking above the recommended lower risk limits.
Could the evidence base and sources of scientific advice to Government on alcohol be improved?
11. The 1995 report “Sensible Drinking” noted a number of areas of uncertainty in the evidence base. It noted a general point of an insufficiency of UK studies to be able to base recommendations for the UK population on these alone. While this remains the case, the 2009 analysis for the Australian Government’s guidelines on alcohol noted that the US and the UK were the main sources of relevant epidemiological research worldwide. In addition, the increased range of suitable studies worldwide since 1995 linking risk and consumption has filled a number of specific gaps in our knowledge and given us greater confidence in some areas.
12. The Government seeks scientific advice from a wide range of sources including academic institutions, national and international independent experts, and organisations such as the National Institute for Health and Clinical Excellence, and relevant Royal Colleges. We used a number of such sources for the 2006 revised guidelines on alcohol and pregnancy, for the 2008 Know Your Limits campaign, and for development of the 2009 guidance on alcohol and children and young people.
13. The 1995 report “Sensible Drinking” reviewed harmful and beneficial effects due to alcohol consumption for a range of conditions. Its guidance on levels of regular consumption was based primarily on an aggregation of what was then understood for health conditions that may be acquired, or avoided, over a long term—often known as “lifestyle-related” conditions. The report also discussed the evidence on mortality due to short term drinking episodes. It did not offer guidance on levels of consumption for individual drinking episodes beyond advice to reduce episodes of drunkenness, to have drink-free days after episodes of heavy drinking, and it recommended a range of situations when it is not appropriate to drink alcohol at all.
14. The report discussed issues around the “J-shaped curve” for all cause mortality due to alcohol consumption. A protective effect for men over 40 and post-menopausal women who drink alcohol regularly at low levels is estimated to reduce mortality for these groups compared to non-drinkers. As consumption rises, the protective effect is neutralised and then overtaken by an increasing risk of mortality from other conditions.
15. A number of studies have been published since 1995 on the protective effects of low level alcohol consumption. Some have suggested that the effect for coronary heart disease may have been over-estimated, and this issue is discussed in the 2009 Australian guidelines report. The 1995 report itself discussed one element of this issue under a heading the “sick quitter” hypothesis (paragraph 5.4) but concluded on the basis of expert testimony that the evidence for a protective effect was sound. We think it likely that the conclusion of the 1995 report that a risk reduction is likely from levels of regular consumption as low as one unit per day, with limited additional benefit at levels above that, is still correct. However, we have acknowledged in advice to the public that a similar reduction of risk may be achieved through other means such as improved diet and exercise.
16. The 1995 report suggested that a recommended upper level of alcohol consumption should be made by reference to where a judgement of the evidence suggests that there is a steady increase of risk rising “significantly” from the lowest all-cause mortality point on the J-shaped curve. The report discussed the difficulty of identifying the precise point of minimal all-cause mortality due to alcohol (paragraph 7.10). While an element of judgement is required, this has been misstated, in our view, in some public statements as meaning that there is no objective basis for establishing guidelines for lower risk drinking.
17. The fact that there is no single scientific definition of “significant” or acceptable increase in risk also requires an element of judgement. The 1995 guidelines considered the balance of the evidence on regular levels of consumption that would not accrue significant health risk. A different approach was taken for the recent 2009 Australian guidelines, in which case the authors choose to use a lifetime risk of one in 100 of death from alcohol-related disease or injury to inform their recommended drinking thresholds for acceptable risk. However, the authors state quite explicitly that this may be seen as too high or too low a risk by individual drinkers. As in the 1995 report, they also make clear that the actual risk to an individual may be affected by factors such as gender, age and body size.
18. An important development since 1995 has been the growth in understanding of the links between alcohol and cancer. Successive evaluations between 1988 and 2009 by the WHO International Agency for Research on Cancer (IARC) have shown that the occurrence of malignant tumours of the oral cavity, pharynx, larynx, oesophagus, liver, female breast and colo-rectum is causally related to the consumption of alcoholic beverages. The best current estimate is that 6% of cancer deaths in the UK are caused by alcohol,
19. Scientific developments since 1995 have not undermined our confidence in the current guidelines, although we continue to keep the science under review. The Department is not currently planning a formal review of the guidelines, but would be willing to consider this if it were felt to be useful, for example to help communicate health impacts better.
20. As already mentioned, the 1995 report did not offer guidance on levels of consumption for individual drinking episodes beyond advice to reduce episodes of drunkenness, to have drink-free days after heavy drinking and advice on situations when no alcohol use is recommended. We are aware that some governments do offer advice on levels of consumption for individual drinking episodes, in addition to advice for regular drinking. For example, the 2009 Australian Government’s guidelines, do include such advice. The guidance can be found at www.alcohol.gov.au The recommendations are based on statistical evidence of the lifetime risk of death from injury related to individual drinking episodes. While we do see some possible value in such a guideline, we have no plans at present to introduce this within the UK. We believe that this would require particular consideration of its likely impact and its real value in influencing the behaviour of individuals who currently choose to engage in “binge” drinking. The Australian Government’s guidelines (Guideline 2) do acknowledge issues with the quality of the evidence quantifying the links between alcohol and injury.
21. In contrast with the position in 1995, improved information on risks now makes it possible to estimate the overall burden of alcohol-related ill health in the UK, placing this in the context of other risks to health.
How well does the Government communicate its guidelines and the risks of alcohol intake to the public?
22. The 1995 report set out (10.29) what it saw as three elements for a coherent health education strategy—its advice on (i) benchmarks for regular drinking, (ii) reducing episodes of excessive drinking and intoxication, and (iii) specific messages for particular groups. It was concerned to reduce people’s underestimation of their consumption. It argued strongly that the recommendations on daily benchmarks should always be presented firmly in the context of the other advice and health risks and benefits.
23. The Government did not develop a strategy for education on alcohol in England until some time after publication in 2004 of the Alcohol Harm Reduction Strategy for England. In 2007, the Home Office commissioned a campaign targeted to young adult binge drinkers. In 2008, the Department of Health commissioned the “Know Your Limits” campaign, which focussed on raising understanding of alcohol units. In 2009, the “Alcohol Effects” campaign had a stronger focus on health risks and on influencing “increasing risk” and “higher risk” drinkers. In early 2010, the Department for Education funded the “Don’t Let Drink Decide” campaign aimed at children and their parents.
24. From 2008 the Department has adapted the terminology used to communicate differing levels of risk to practitioners and the public, focussing on three broad levels of consumption (lower risk, increasing risk and higher risk). It was consistent with the 1995 report
25. The new terminology was developed to aid communication of the known risks following research with members of the public commissioned by the Department from the Central Office of Information (COI) and after consultation with some 20 leading experts from the alcohol field. It focuses on three bands of regular consumption and links these to their known levels of risk for ill health and death. This helps individuals to consider their personal risks. It allows an emphasis on those at greatest risk (higher risk), with an honest, risk-based language shown to be better understood by that group. It reinforces the relatively low risks posed to the larger numbers drinking within the recommended lower risk limits and the progressive increases in risk faced by those in the intermediate category of increasing risk.
26. The results of the “Alcohol Effects” campaign were positive. In a post-campaign survey, increasing risk and higher risk drinkers claiming to have cut down their drinking rose from 21% to 26%. Nevertheless, public understanding of both unit measures (especially for wine) and the guidelines needs to improve. About a third of the population can say what the guidelines are with some accuracy, with others both under- and over-estimating.
How do the UK Government’s guidelines compare to those provided in other countries?
27. In late 2008, the Department commissioned an external expert, Peter Anderson, to carry out a survey of the advice on alcohol consumption provided by EU and EEA/EFTA Member States. Responses were received from 22 EU Member States and three other EEA/EFTA Member States covering:
whether there is a defined unit or standard drink size;
general guidelines for adults on “low risk”, “responsible” or “safe” consumption;
any guidelines for children or pregnant women;
how guidelines are communicated; and
inclusion of any advice on consumption on alcohol labels.
28. Broadly the findings were:
for those governments (13) which do issue such guidelines, those for “low risk” regular consumption by women were similar, ranging between 108g per week and 140g per week (UK = 112g); guidelines for men were more varied, ranging between 140g per week and 280g per week (UK = 168g); one UK unit = 8g alcohol;
six countries say they follow WHO recommendations—though understanding of these seems to vary; and
six countries do not issue such guidelines officially, though in some cases non-official guidelines are recognised.
29. The Australian Government’s guidelines include a 2007 chart from the International Centre for Alcohol Policy (page 19) comparing some international guidelines, although we note that this is not accurate for the UK.
September 2011