‘Health guidelines’ are about forcing a change in the law, not changing behaviour

According to the Daily Mail, the NHS intends to ‘slash safe drinking limits’. The article is light on evidence, but that is apt since the alcohol guidelines themselves were famously ‘plucked out of the air’. It does, however, quote the usual self-appointed ‘stakeholders’ from Alcohol Concern, the Alcohol Health Alliance and the Institute of Alcohol Studies (formerly known as the UK Temperance Alliance) saying that there is no safe level of alcohol and, by implication, that the guidelines should be set at zero.

It is true that the NHS is reevaluating the drinking guidelines and, given the ongoing hysteria about alcohol and health, they are more likely to go down than up. As Michael Fitzpatrick recounts in his wonderful book, The Tyranny of Health, the limits used to be considerably higher: ‘In 1979 the Royal College of Physicians first indicated that a weekly consumption of more than 56 units of alcohol was the “absolute upper limit”. In 1984 the Health Education Council suggested that weekly levels of between 21 and 36 units for men and 14 and 24 units for women, would be “unlikely to cause damage”. Then in the late 1980s a new consensus emerged from the royal college and other medical bodies, setting the upper limits at 21 for men and 14 for women.’ Authorities later became concerned that people were drinking their weekly allowance in one sitting and so converted it into daily chunks. The current guidance is for men to ‘not regularly drink more than 3-4 units of alcohol a day’. For women the daily limit is 2-3 units.

It somehow feels inevitable that a day will come when the guidance will be to drink nothing at all. As we saw last week in relation to pregnant women, zero is the public health lobby’s favourite number. Moderation is tricky to define whereas zero is neat and simple. It sends a clear message, and campaigners love a clear message.

Whether it is grounded in evidence is another matter altogether. The ‘no safe level’ argument is based on epidemiological research which finds a link between very moderate drinking and certain cancers. The relative risk of breast cancer, in particular, is said to rise with just one drink a day. In absolute terms the risk remains low and it is doubtful whether many women would be prepared to become lifelong teetotallers in order to avoid it, but the trade-off between pleasure and risk does not concern the public health lobby. If there is a risk at low doses, there can be no safe level and, by implication, the government must do something.

Even if the evidence that light drinking increases breast cancer risk was stronger, it does not lead to the conclusion that abstinence is the best advice. None of the crusaders quoted in the Daily Mail article mentioned the awkward fact that teetotallers tend to suffer higher rates of heart disease and have higher rates of overall mortality than moderate drinkers. Advising total abstinence when abstinence kills would seem to violate the Hippocratic Oath (‘First, do no harm’). The J-Curve between drinking and mortality (shown below, from this study) is so inconvenient to the hardliners that there have been a few attempts to undermine the evidence in recent years, but it remains solid. If there is no safe level of drinking, there is also no safe level of not drinking.

Screen Shot 2015-06-29 at 23.24.50

If we must have drinking guidelines, they should be based on the amount of alcohol that has to be consumed for a drinker’s mortality risk to exceed that of a teetotaller. As you can see in the graph above, that would mean increasing the limits. But since that would be out of step with the mood of the times, they will probably be reduced on the basis of epidemiological evidence about breast cancer which is much weaker than the epidemiological evidence for the J-Curve.

Will people change their drinking habits if the guidelines are changed? Probably not, but that is to misunderstand their purpose. From the perspective of the public health lobby, guidelines are designed for political, rather than educational, use. If the limits are lowered, millions more people will suddenly become hazardous drinkers. The flagging moral panic of Binge Britain can be resurrected.

Last year, after some nudging from the Government, thousands of shops stopped selling cans of super-strength beer and cider. The explicit rationale was that these products contain more units than a man is supposed to drink in a day and are therefore inherently dangerous. Leaving aside the fact that even the current guidelines allow people to drink such quantities so long as they do not do so regularly, it is easy to see how lower limits would open the door to similar rhetoric about almost any alcoholic drink.

We see the same thing happening with sugar. It is routinely claimed that the World Health Organisation has declared that people should not get more than five per cent of their energy intake from free sugars. This is a very small amount and is half what they previously recommended. In fact, their official guidance is still 10 per cent, they merely note that reducing it further may have additional benefits for dental health, although this recommendation is ‘conditional’ because the evidence is weak.

Either way, the advice is too complicated to be much use to the man on the Clapham omnibus, even if he was aware of it. Who knows what their daily energy intake is, let alone how much sugary food will provide 10 per cent of it? Since most people exceed the current recommendations, it is most unlikely that they will suddenly abide by a lower recommendation. The advice is therefore of questionable use to the public, but it is very useful to the likes of Action on Sugar who, by portraying five per cent is the upper limit, can claim that everyday products, such as fizzy drinks, contain an inherently dangerous amount of sugar.

As one anti-sugar campaigner said when the WHO guidelines were published last year: ‘If the sugar threshold is lowered, I think breakfast cereal is going to have a really hard time justifying its existence.’ This is what guidelines are really all about. They are not aimed at you because you will ignore them. They are aimed at campaigners who can use them to lobby politicians to take the decision out of your hands.


  • Curmudgeon

    An official zero guideline would obviously be scientific bollocks, but it would force the drinks industry (and organisations like CAMRA) to confront the reality of their position. At present they can hide behind a convenient figleaf of being able to say “drinking in moderation is not incompatible with a healthy lifestyle” and indeed at times make common cause to a limited extent with the public health lobby, such as brewers like Greene King supporting minimum pricing.

  • Tarek

    A little hasty to condemn all guidelines based on the alcohol ones. Furthermore the fact that teetotalers had more heart disease is irrelevant given that the epidemiological data that references this established correlation and not causality. Guidelines I agree, aren’t always about helping but often about career advancement or to force a social change but one shouldn’t dismiss them simply because they often furnish proof of the bias of the establishment this giving one a useful place to start when rebutting the ‘recommendations’. An excellent example of this are the new NICE guidelines on statin use

    • Mark Bailey

      “given that the epidemiological data that references this established correlation and not causality”

      Well, given that the same data was deemed good enough for the link between tobacco and lung cancer – which it was – I think we can accept its validity.

      • Tarek

        I disagree. Too many variables which cannot be ignored. ‘The same data’ is incorrect as a description. You’re talking about different studies even if methodology was similar. Furthermore the tobacco hypothesis was prices biochemically, pathologically and epidemiologically. Again different studies and methodologies.

        • Mark Bailey

          I’m not talking about different studies; I’m talking about the daddy of these studies – Doll.

          • Tarek

            Subsequent studies established the biochemical/ pathological basis for the hypothesis that Doll’study established this proving causality. The same wasn’t done for alcohol and heart disease hence it remains a correlation and not causation

            Either way I take the broader point ie that the biosciences will continue to entertain us with claim and counter claim for many more years, somethingbthat will