Science and Technology Committee HC 1536 Alcohol GuidelinesSupplementary written evidence submitted by the Institute of Alcohol Studies (AG 24a)

I have appended the additional evidence that was promised to the Committee which should help to further clarify some of the points raised.

Correction (in italic) to Professor Heather’s oral evidence on 12 October:

Q8: Professor Heather: [...] In relation to the cardio-protective effect, when the “Sensible Drinking” report was written in 1995, which was the last time that the Government addressed this problem, there was much more confidence in the cardio-protective effect, which is reflected in the report by the committee of which Dr Harding was a member, and also in the report of the three royal colleges that came out at roughly the same time (correction—were published some eight to 10 years earlier). That consensus has now largely disappeared, which is the result of more careful research. [...]

Dr Marsha MorganInstitute of Alcohol Studies

Attachment 1

EXPRESSING RISK IN RELATIVE OR ABSOLUTE TERMS

The Australian alcohol guidelines (Rehm et al, Int. J. Methods Psychiatr Res 2008; 17: 141–151) are based on an assessment of the absolute risk of alcohol consumption while the Canadian guidelines are allegedly based on an assessment of relative risk (Stockwell et al, Drug and Alcohol Review 2012)

When discussing the implementation of research evidence into clinical practice:

1.Relative risks (RRs) associated with an exposure (eg alcohol) should be lifted from credible research studies.

2.A particular individual’s baseline risk should be calculated, or roughly estimated.

3.The RR can be multiplied by the individual’s baseline risk to calculate the risk the individual would experience if exposed to alcohol.

4.The difference between baseline risk and risk if exposed to alcohol equals the absolute risk difference.

The Australians focused on the absolute risk difference and set the lifetime risk standard at one in 100. They argued that if the risk for alcohol-related chronic disease mortality fell below 1% for a given level of alcohol consumption, that level of consumption would be safe, but once it exceeds 1%, it is no longer safe. As a result of their analysis they deemed that both men and women should not drink more than two drinks a day (equivalent to 2.5 standard UK units) and for occasional drinking three or four drinks seem tolerable. They set a weekly maximum of 17.5 units for both sexes.

However, Stockwell et al argued that this threshold is arbitrary. However, the Australian team felt that in general the public would most likely accept a risk of 1:100. Stockwell et al claimed that their “relative risk approach” in which the risks for alcohol-related chronic diseases were compared with the risk in abstainers was better geared towards individuals’ own decisions. However, in presenting their data on relative risk they also included data on the frequency of the specific cause of death (Table 1). In a sense, this latter piece of information is analogous to baseline risk. So in reality Stockwell’s approach is not based on relative risks alone. The authors make the point, for example, that a 19% relative risk reduction in ischaemic heart disease associated with two drinks per day may more than wipe out the alarming 43% increase in oesophageal cancer, since the former is 10 times more common as a cause of death than the latter.

Table 1

% CHANGE IN RELATIVE RISK OF DEATH FROM ILLNESSES LINKED TO DRINKING FOR MEN AND WOMEN AGED BELOW 70 YEARS BY AVERAGE STANDARD DRINKS PER DAY*

*Estimates provided by Rehm and colleagues as part of a specially commissioned report.

Thus, Stockwell’s recommendation of levels where the net benefit equals zero must involve multiplying relative risks by absolute risks for the Canadian population. Supposedly Stockwell allows that individuals will have their own baseline risks, so that the balance may vary for individuals. Thus Stockwell’s assertion that they are using a relative risk approach is incorrect as they are in reality also considering absolute risks. As a result they recommended that, for a net zero risk, daily intake should not exceed two standard Canadian drinks for women (equivalent to 3.4 standard UK units) and three standard Canadian drinks in men (equivalent to 5. standard UK units). However in recommending weekly limits it is clear that they have included two drink free days as the totals for women equate to17.0 standard UK units in women and 25.5 standard UK units in men.

Attachment 2

A REVIEW OF HUMAN CARCINOGENS—PART E: TOBACCO, ARECA NUT, ALCOHOL, COAL SMOKE, AND SALTED FISH:

Béatrice Secretan, Kurt Straif, Robert Baan, Yann Grosse, Fatiha El Ghissassi, Véronique Bouvard, Lamia Benbrahim-Tallaa, Neela Guha, Crystal Freeman, Laurent Galichet, Vincent Cogliano, on behalf of the WHO International Agency for Research on Cancer Monograph Working Group, International Agency for Research on Cancer, Lyon, France.

Attachment 3

ALCOHOL AS A RISK FACTOR FOR LIVER CIRRHOSIS: A SYSTEMATIC REVIEW AND META-ANALYSIS:

Jürgen Rehm, Benjamin Taylor, Satya Mohapatra, Hyacinth Irving, Dolly Baliunas, Jayadeep Patra & Michael Roerecke.

Attachment 4

OUR INVISIBLE ADDICTS: FIRST REPORT OF THE OLDER PERSONS’ SUBSTANCE MISUSE WORKING GROUP OF THE ROYAL COLLEGE OF PSYCHIATRISTS, COLLEGE REPORT CR165, JUNE 2011

This report can be criticised on many accounts but only the main ones will be highlighted:

1.The Report at first seems ambivalent about lowering the safe limits: On pp 35–36 it states: “Screening instruments may not be appropriate for an older population, and criteria such as safe limits’ (see pp. 35–36) for alcohol consumption may be set too high to be of valid use with elderly people”.

   Later, however, a definite recommendation is made that “safe limits” for those aged 65 and older should be set lower than those for the general population at 1.5 units /day or 11 units/week. The US National Institutes of Health (NIH) is quoted (page 35) as the source for the recommended intake limit of 1.5 units per day. However the authors clearly have not appreciated that a US drink contains 13 g whereas a UK drink contains 8 g. They have not taken this difference into account and they have miscalculated the figures.

   The NIH recommendations state:

For healthy men up to age 65:

no more than four drinks in a day (UK equivalent 6.5 units); and

no more than 14 drinks in a week (UK equivalent 23.0 units).

For healthy women (and healthy men over age 65):

no more than three drinks in a day (UK equivalent 4.9 units); and

no more than seven drinks in a week (UK equivalent 11.4 units).

   See: http://pubs.niaaa.nih.gov/publications/practitioner/cliniciansguide2005/clinicians_guide5_help_p.htm

   Thus if we followed the NIH lead, as the authors of the report suggest, we would recommend intakes of no more than 5 units /day which is greater than the current recommended UK levels for the general population ---in other words the daily allowance for the >65’s would INCREASE not DECREASE.

2.It is correct that for various physiological reasons, a dose of alcohol that might not have caused unsteadiness or altered behaviour at age 50, may affect the individual more at the age of 80. It is also true that blood alcohol levels may be slightly higher level in an older person than in a younger person of the same body weight (p 23). These seems to be the main reasons for “lowering safe limits” for the over 65’s. However, the authors provide no evidence to support this view. Thus, they so not show data on hospital admission rates for accidents in the home, falls or injury, by age, nor do they review the data on alcohol and falls in the literature, particularly in relation to age.

3.They do not attempt to review the relationship between cognitive impairment and dementia and alcohol. The literature is extensive and somewhat complicated because of confounders such as sociability, diet, lifestyle and income. However there is some evidence to suggest that individuals who consume one to four drinks a day may be at lower risk of developing dementia.

4.The authors imply that many people aged over 65 are on medications and are frail. This is clearly a substantial generalisation. Most medications widely used in the elderly—eg statins, calcium and vitamin D supplements and even hypoglycaemic agents have no interaction with alcohol. In addition many people in this age group live happy, useful and active lives.

October 2011