New guidelines on alcohol published

[Gez_Blair]

nhs.uk/news/2016/01January/...

Nanny state nonsense or sensible, transparent advice? Answers on a postcard please (or a post)

[piglette]

I am none too sure about Dame Sally Davies, first she says obesity should be considered on a par with terrorism, now she says no alcohol. There have been numerous studies over the years that show that statistically drinkers live longer than abstainers. That is not to say that people should take up binge drinking!

[Zeno]

@piglette

Where did she say obesity should be considered on a par with terrorism?

@bigleg

But she doesn't say no alcohol, does she. And this is not Government limits on alcohol. And what commercial interests would suggest a lower level of consumption of alcohol anyway?

All misrepresents what the 'guidelines' say - as much of the media have. They are not advising people cut back and it's almost the opposite of the nanny state. The guidelines give information - based on the best available evidence - on what alcohol consumption increases your risk of early death by 1%. It doesn't tell anyone what to drink; it simply gives people information so they can decide for themselves whether to not to take that risk.

As usual, the media have a lot to answer for in misunderstanding or mis-representing science.

[imageh]

Good question by bigleg - about the link between alcohol and cancer. Is the link based purely on statistical correlation? Is it based on a mechanistic understanding?

If the former (and I suspect it is) then this is a difficult conclusion. Firstly correlation and causation are not the same thing. (There is a good straight line correlation between teenage literacy and waist dimensions - not due to fatter teenagers being more literate, just older ones.) So there are many confounding factors which are statistically very difficult to de-convolute. To go to the length of saying (as a Times headline did, yesterday) that one drink is one too many is almost certainly scientifically unjustifiable.

Given that cancers are so varied, and their causes also vary, if there was a mechanistic understanding, then we would have made something of a breakthrough. There may be pathological iinks - e,g, fatty livers more susceptible to cancer and fatty livers strongly correlated with alcohol consumption. That still doesn't justify such a headline.

Unless the public are given much more detailed justification, I think such pronouncements are actually counter-productive, as most people will ignore them. Obviously cutting down and being sensible about levels of consumption is to have common sense about one's health. Teetotalism - I'm not so sure.

[helvella]

Have a look here:

theregister.co.uk/2016/01/0...

What is the medical establishment going to do to help those whose bodies produce ethanol - right up to auto-brewery syndrome? Surely it would be dereliction of duty not to take them on-board as urgent cases?

[johnsmith]

How was the evidence gathered for the guidelines. I suspect lots of quality clinical data which is thoroughly unreliable.

Just because a sample of people with a particular diagnosis have been drinking alcohol does not mean that the alcohol caused the illness. Think "9 out of ten cats" advert.

[Zeno in reply to johnsmith]

Health risks from alcohol: new guidelines

gov.uk/government/consultat...

You need to look at the specific data here - blanket statements that clinical data are unreliable get us nowhere.

[johnsmith in reply to Zeno]

Yes it is a blanket statement. However, it does refer to the issue of clinical measurements which are accurate, but cannot be applied reliably to another person. For the simple reason that person has a particular genetic makeup ie has gene to effectively remove alcohol from system or has gene that causes very little removal of alcohol from his system.

Another issue is the effect of alcohol varies across the age ranges. So again you may not be able to apply the taken clinical data here.

There are occasions when alcohol is good for you because of certain conditions that you may have. So again accurate clinical data taken in one context cannot be applied to another.

Hope this shows my thinking.

[Zeno]

Two analyses of the guidelines. The first by David Spiegelhalter, Professor of the Public Understanding of Risk at the University of Cambridge:

Medicine, poison, poison, poison, ……: understandinguncertainty.or...

And a more detailed analysis by statistician, Adam Jacobs:

NEW ALCOHOL GUIDELINES: statsguy.co.uk/new-alcohol-...

[johnsmith in reply to Zeno]

Thanks for the references. They were most useful.

Daily mail says:

dailymail.co.uk/health/arti...

It adds a few references that throws doubt on the new alcohol guidelines.

I now wait to see what the New Scientist has to say.

[herdwick]

I cannot understand how a larger mass higher metabolic rate male can end up with the same alcohol intake recommendation as a lighter female.

Or is this reflecting the fact that we're all heading to be over 16 stone regardless of gender.

[OldBiker]

There is a lot of naiivity around this guidelines report. It gives no evidence, but states that "evidence" was accepted from the "experts group". When you think about how the "evidence" was gathered it can only be based on what people said they drank. As Kinsey found with his 60's sex study, there is no way any researcher can actually monitor/watch human sexual activity in its natural setting. Neither can it with drinking, over a day, a week, a month. The researchers will have recognised this and in some cases sought supporting evidence or checked peoples' Gamma GT levels, then compensated for some "under-reporting" of intake But not enough. Secondly, there will be much more evidence from patients who present with an illness than from healthy drinkers (probably from only a few control groups?). Think about it: a patient presents with cancer; how much do you drink? a day? a week? last month? "Er... Er.... I can't recall... Oh about one or two glasses... maybe occasionally another?" Yeah, right. Double that and double it again, then you've got more realistic numbers, but it's still not "evidence" - you'll never get that.

Soooooo if you can say you only drink 14 units a week, and drink no more than 4x14 = 42, you should be OK and in the lower risk area.

I also agree with a lot of what has been said above.