"Our results show that the safest level of drinking is none. " -- GBD 2016 Alcohol Collaborators

Thanks! I had no idea the ‘drinking is good for you’ thing was so entrenched in medical practice. I thought it was internet garbage-medicine, the kind of thing you read and you think ‘In five years this will be discovered to be utter nonsense.’

I used to work in an office of about 50 people. In the course of one year, 7 of them had an arterial stent (or two) implanted.

5 of them had an actual handwritten script for one glass of red wine every day.

The research supporting it is substantial.

Which one (and not based on animal data or very small human trials) ?

It’s not one…there’ve been a lot of them.

I don’t bookmark studies when I read them. Your Google skills are as good as mine.

I am not understand your question Are you asking if there is any study which shows alcohol to have adverse effect? This is known.

No - I was looking for serious studies which link positive effects of one glass a day of red wine to positive health effects as Sunshine842 claims - there are a number of animal studies and small human studies (but with little statistical power (or poorly designed) - and so they have little meaning but unfortunately are often used by the press (and doctors) - but it looks like Sunshine842 doesn’t have those studies

One benefit of alcohol is it does reduce clotting of blood which is a cardiovascular benefit and may also decrease inflammation

Sunshine doesn’t make up random bullshit, but she also doesn’t bookmark the copious number of articles she reads. She also doesn’t read halfbaked theories, so leave any notion of Facebook clickbait at the door.

Again…you’ve got Google foo. Knock yourself out. If I can find it so can you.

1 Like

jfood hijack your account? :smiley:


Maybe. What makes you say that?:rofl:


Watching this on BBC Breakfast right now.
It’s not mentioned here but I’ve read high alcohol consumption may increase chance of breast cancer middle aged women.

A new campaign is urging people between the ages of 45 and 65 to have regular “drink-free” days.

Middle-aged drinkers are more likely than other age group to drink more than the recommended 14 units a week.

A YouGov poll also shows that they find cutting back on alcohol far harder than eating healthily or exercising.

Doctors say “drink-free” days will improve sleep, help with weight loss and reduce the risk of high blood pressure and cancer.

Dr Julia Verne, a spokeswoman on liver disease for Public Health England said: "Having a day off drinking gives you a chance to clean your system and give your liver a rest. It also has an immediate impact on your sleep and calorie consumption.

“People have also told us that the idea of a ‘drink-free’ day is much easier to manage than cutting down, say, from one large glass of wine to a small glass of wine.”

The benefits of a day off drink

The campaign, Drink Free Days is a partnership between Public Health England and the alcohol education charity Drinkaware.

The YouGov poll - by PHE and Drinkaware - surveyed nearly 9,000 adults aged 18 to 85 during May and June this year.

It found that one in five were drinking more than the government’s 14 unit-a-week guidelines.

And two-thirds said they would find cutting down on their drinking harder to do than improving their diet, exercising more or reducing their smoking.

Image copyright NHS digital

Dr Verne said: "Most middle-aged people are not drinking to become drunk. They see it as a social activity, or as a reward for success or compensation for a hard day at work. It’s become a habit and part of their lives.

"But the more you drink, the more you increase your risk of high blood pressure, heart and liver disease and cancer.

“Ultimately you are more likely to cut down if you have some days off drinking,” she said.

She also pointed out that many people in this demographic were struggling with their weight, and that they didn’t realise how many calories were contained in alcohol.

The campaign is part of a growing awareness of the health risks of drinking.

Recently a large global study by the Lancet showed that there is no safe level of alcohol consumption, even though the risks associated with one glass a day were small.

In 2016, the government cut the alcohol limits it recommended for men and women to no more than 14 units a week - equivalent to six pints of average-strength beer or seven glasses of wine.


I don’t remember the last time I consumed 14 units in a week…! (And not because of the alcohol)

1 Like

I don’t know you and your background but know from my job that many studies linked to such correlation you mentioned are based on small numbers of individuals (<200-300) or animal data. If you go through databases like SciFinder you find hardly any study which has reasonable statistical power to support the the correlation (and Google isn’t really the best way to find serious scientific data)

So a week later you’re still trying to stir an argument?

Try another forum. I’m not interested.

I am traveling now.

Two things to consider. First, it is true that there are/were medical professionals (doctors) recommend alcohol as a healthy diet. This publication is gear toward this.
Second, it is difficult to control or conduct human studies. It is cleaner to see an effect from an animal study where you can limit other variability factors and even do a cross-over study. A 12-rats study can be more definitive than a 120 human study. The only problem of an animal study is the animal-human translation. 200-300 individuals is not the issue here. Otherwise, there will never be enough people to develop a small rare/orphan diseases.

While the number of subjects/volunteers are important in a study, it is far from the only things matter. The selection of the subjects is probably more important. This is the same reason why well respected pollster organizations like Gallup can use a smaller set of individuals in their polls.

Lastly, keep in mind why some clinical trials are very large, and not lose focus for the reasoning. As far as I know, the very large size usually is not to prove efficacy. Large clinical trial usually is about safety, especially long term and subtle safety issue. In short, for anything very obvious, you don’t need a large study samples. Large study samples is only needed if the effect is very subtle.

Romania, eek

I have to do with animal studies and clinical trials on a daily basis due to my work and what you write in your response is often wrong on so many different levels and it would take a very long reply to react on all of it. But to say a 12 rats study can be more definitevr than a 120 human study is just laughable (if that would be the case the FDA (or EMA) wouldn’t require human studies for every drug in clinical trials. Even 5 human study is scientifically more relevant than any large scale animal study. If PDX animal models show low transferability to human clinical studies and that covers a large variety of therapeutic areas.

The power of your study is the most important factor for your study. You can have the best patient selection but it would be meaningless for the results if the statistical significancy is not achieved (one of the reasons clinical studies have failed)

That is just plain wrong - the size of your clinical study is correlated to your expected effect window. You want the size of your clinical trial large enough that you have a chance to meet your primary and secondary clinical endpoints in a statistical meaningful way. Safety readouts are part of any clinical trial and are independent of trial size. A large clinical trial has nothing to do that you are looking for long term or subtle safety issues as your main reason.

This is actually one of the main problems and risks with a number of rare diseases that the clinical trial size is so small due to the small patient population that there are drugs approved with otherwise insufficient safety and efficacy data.

1 Like

Calm down. I said an animal study by itself is typically more definitive than a human study by itself. I also specially wrote the animal-human translation. What I wrote is that a typical animal study is more definitive about the effect on the animals. You usually don’t need some >300 rats to make a point about the animals.

Actually, a large size has a lot to do with safety readout. Efficacy alone usually is easy to show. If you are developing a drug, the drug better be clearly efficacious. It better not be a drug that is barely efficacious that a large sample is needed to get the statistical power. If you do, then you have a bigger problem. The drug simply don’t sell.

A challenge of selling a drug is about safety and/or efficacy data from a comparator. You are trying to demonstrate your drug is either safer and/or more efficacious than a comparator, and this is often a small difference which is the reason for a large sample size. This is what I mean by subtle. In a typical Phase II study, you should already get a decent idea about the efficacy. The reason for a larger Phase III study (>300) is to try to meet an endpoint against a comparator (or standard of care).

Let me give you a very simple example. If I want to demonstrate the safety difference between crashing a car at 30 mph vs 0 mph (not crashing), I don’t need large sample size. Now, if I want to show the difference between crashing at 30 mph vs 27 mph, I will need a larger sample size. Why is that?

Without all these excitement talks, let’s get back to the alcohol effect. You don’t need a large pop size if it is statistically significant. To insist one must have >200 people to show an efficacy/safety readout is to ignore the selection of subjects and the difference of the effect…etc Bottomline: you don’t need a large data set if the difference is large and the spread is small. You will need a large data set if the difference is small and the spread is large (subtle).

Britain’s oldest person drinks Famous Grouse every day. Hit the first or second link for photos.
Enjoy yourselves! Cheers!


It is obvious that you have very little experience with statistics and drug discovery/clinical trials in general and it is useless trying to discuss it with you when you have don’t understand basics of it, e.g. most drugs in today’s trials are often just incremental better than previous drugs and so it is necessary to run these large trials and efficacy compared to standard of care is often very hard to show and one of the main reasons drugs fail in clinical studies. In the end it all comes down to the p-value and large sample sizes are necessary in order to ensure the results are valid. (Here is one example from a beta blocker trial with “delay of death” as an endpoint: "Dr. Devereaux cited a beta blocker trial that randomized 112 patients. In that study, 11 patients died— in the beta blocker group and 9 in the control group. The p value was .02. “It looks like there is a big effect and beta blockers are truly worthwhile,” he said. Based on the study’s results, guidelines were written advising that beta blockers be given to patients having myocardial infarction surgery. Dr. Devereaux and his colleagues
conducted their own perioperative beta-blocker trial but with a larger sample size. They randomized 8,351patients in 23 countries and 190 centers. Their results were the opposite of the study with 112 patients. Of 226 deaths, 129 were in the beta blocker group and 97 were in the placebo control
group. The p value was .03 "
Same also with animal tests - it doesn’t matter if you use humans or animals the p-value is relevant - if your effect will be small in animal studies you still need a large number of animals to run the study. In addition, the effects from animal studies on alcohol are known to be problematic to be translated to human studies. And so similar to the beta-blocer study you would need a large patient number for an alcohol related study to get a good p-value.