Views & Reviews Personal View

Guidelines on safe alcohol drinking are probably about right

BMJ 2015; 351 doi: (Published 24 September 2015) Cite this as: BMJ 2015;351:h5082
  1. Rahul (Tony) Rao, consultant, old age psychiatry, South London and Maudsley NHS Foundation Trust, London SE5 8RS, UK
  1. tony.rao{at}

As the UK looks again to its alcohol recommendations, we have a much clearer evidence base to inform guidelines, writes Rahul T Rao

How much is it safe to drink? A statement from the Department of Health confirms that “a group of independent experts tasked with developing alcohol guidelines for the UK chief medical officers are currently researching and developing a proposal on these guidelines and expecting to consult on them in autumn 2015” (Department of Health, personal communication, July 2015).

It is 20 years since the Department of Health added daily drinking guidelines of 3-4 units for men and 2-3 for women to the existing 1987 weekly recommendations of 21 units for men and 14 for women. A more recent recommendation of having at least two alcohol-free days a week has only added to the confusion over the complex relationship between alcohol and health, particularly in mixed age populations.1

Bearing in mind that “a drink” in the United States is equivalent to 1.5 UK units of alcohol, the US National Institute on Alcohol Abuse and Alcoholism has advised that women should drink no more than the equivalent of 4.5 UK units a day or 10.5 UK units a week and that men should drink no more than 6 UK units a day or 21 UK units a week.2 For over 65s the American Geriatrics Society has defined “high risk drinking” as more than 4.5 units a day or 10.5 a week, in men and women.3 Australian guidelines for this age group recommend no more than 2.5 units a day for men and women.4

Plucked out of the air

The lack of a robust evidence base led the Sunday Times in 2007 to report that the current UK limits were “really plucked out of the air . . . it was sort of an intelligent guess by a committee.” But we now have a credible evidence base for testing the limits by examining the link between alcohol and health. Although validity and reliability are concerns, measurable outcomes exist that can guide such recommendations.

While the lack of a valid control group means that controversy rages over a putative protective effect on cardiovascular disease from drinking at current recommended UK safe limits,5 evidence is more conclusive for an increased risk of death from accidents, cancer, and stroke from drinking more than 1.5 units a day, compared with non-drinkers.6 Evidence from a 20 year follow-up of older people has shown that, in older men and women, drinking more than 3 units a day or 11 units a week was associated with the development of drinking problems.7 Regarding all cause mortality in a mixed age population of drinkers, alcohol consumption and mortality are linearly associated, with an increased standardised mortality rate from drinking more than 1.5 units a day.8 A slightly higher threshold of 3 units a day increases the risk of gastrointestinal cancers, hypertension, liver cirrhosis, chronic pancreatitis, injuries, and violence.9

A more recent analysis of mortality attributable to alcohol from seven European countries produced perhaps the most startling conclusion of all. Using the concept of “voluntary” risk taking that includes lifestyle factors such as drinking, the authors concluded, “If drinking guidelines were to be based on the usual accepted lifetime mortality risk for voluntary behaviours, the maximum alcohol consumption would be one drink a day.”10 This equates to 1.25 units of alcohol a day.

Meta-analyses of other negative health outcomes in mixed age populations have found thresholds similar to or below the UK daily guidelines. These have included an increased risk of liver cirrhosis from drinking 3 units a day in women and 4.5 in men.11 The threshold for alcohol related fractures is somewhat lower, at 2 units a day.12

To underpin future guidelines we need further evidence across many outcomes, including those that measure more global attributes such as quality of life and mental disorders. We also need to consider whether to recommend abstinence from alcohol in populations such as older people taking multiple medicines, people with physical and mental comorbidity, and pregnant women.

From the existing evidence, the current UK recommendations for safe limits on daily alcohol consumption may not be far off the mark. If anything, we may need guidelines to be more conservative to reduce the clinical and public health burden of morbidity and mortality related to alcohol.


Cite this as: BMJ 2015;351:h5082


  • Competing interests: I have read and understood the BMJ policy on declaration of interests and have no relevant interests to declare.

  • Provenance and peer review: Not commissioned; not externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription