Hacker News Reader

Comment by Ovah | Hacker News Reader

Here are some actual numbers for risk increase. Heavy drinkers have a 5-fold increased risk of getting head and neck cancer, esofageal cancer, and 2-fold increase for liver cancer. The source also reviews colorectal, breast and head+neck cancer. My opinion as a medical professional is that these are significant, and not "very tiny amounts", of increased risk.

https://www.cancer.gov/about-cancer/causes-prevention/risk/a...

View on HN

Of course, it matters what the baseline frequencies of these cancers are. If these represent .1% of all cancers, even a tenfold relative risk only increases your absolute risk by 1%. That would be "significant" under some definition of the word significant. But it might not be decisive for someone considering whether they should drink.

On the other hand, if the baseline rate of those cancers is 8%, it would likely be enough to convince everyone to stop drinking.

The baseline risk of getting cancer['s whos risk is increased by alcohol] is somewhat low as it is. Simply refuted the above statement that alcohol doesn't affect cancer risk.

EDIT: Thank you asdfasgasdgasdg for the correction.

Cancer is the cause of ~22% of deaths in the US, not 1%?

https://www.cancer.org/latest-news/facts-and-figures-2019.ht....

Thank you, looks like I stand corrected. A very large part of that is prostate cancer which isn't really significantly affected by alcohol. But you're right, the total increased % mortality of cancer is higher than what I stated..

those are relative risks, you need to state them in absolute risk if you want people to be able to understand and make decisions. If the absolute baseline risk of getting head and neck cancer is very low, then having a 5x risk over that isn't really important- it's still absolutely small enough to say "don't care, compared to other risks".

And in alcohol, there's a long list of risks that will cause the majority of deaths compared to cancer.

> you need to state them in absolute risk if you want people to be able to understand and make decisions.

I refute this on the grounds that an alarming number of people seem not to care about their own health given a 1/1000 risk of death (or worse for older or at-risk populations) from SARS-CoV-2. If half the country doesn't care about those odds, what makes us think any measurable amount of people will care about... don't know, 1/50k odds?

But I see where you're coming from. In a perfect world, you'd be right.

The goal shouldn't be to get people to do what you want. It should be to give them sufficient information so they can make an informed decision on their own, even if you might disagree with it.

I'm not suggesting that personal liberty should trump SARS-Cov-2 precautions, but it's not really relevant to a discussion about whether risk should be framed in absolute terms or not. (Edited for clarity.)

I disagree for a simple reason: while liberty is a great goal, if the liberty comes at a cost to the commons, it needs to be restricted. For example, as much as I'd love to let people chose to smoke and die of cancer, smoking affects a lot more than just the smoker. Alcohol DUIs leads to many driving deaths.

The OP suggested that risk shouldn't be communicated in absolute terms because people can't be trusted to make sound decisions. Presumably it's therefore better to only share relative risk numbers, which are almost always more frightening.

That's what I was responding to. I probably agree with most of your views on personal liberty versus the common good, but I think it's tangential to the original point about risk communication.

> The OP suggested that risk shouldn't be communicated in absolute terms because people can't be trusted to make sound decisions. Presumably it's therefore better to only share relative risk numbers, which are almost always more frightening.

That's not actually what I said, but I love your reasoning, so I'll adopt it.

Why:

> The goal shouldn't be to get people to do what you want. It should be to give them sufficient information so they can make an informed decision on their own, even if you might disagree with it.

How much information do you figure that'll require? Because our perception of how much is probably the difference between your perspective and mine.

Society does not care about your personal liberties if you are endangering the lives and well-being of others. Case in point, most people won’t mind if you drink until your liver turns to dust, but they do care if you get in a car and endanger the lives of others. Similar to that, not taking precaution in public to avoid contracting and spreading covid is endangering the lives of others as everyone has to go out in public to survive. By not taking precautions someone would be effectively deciding for those around them that covid is nothing to worry about. Any actions a person takes that knowingly increases the risk of death and injury to persons other than themselves are generally criminal offenses.

I mean cancer which risk is affected by alcohol is somewhat low as it is. I simply refuted the above statement that alcohol barely affects cancer incidence.

Cardiovascular disease however is the #1 cause of death in the Western world. And here alcohol alcohol plays a much much more significant role in contributing to mortality.

Cancer causes <1% of deaths? Not sure where you got that belief, cancer is #2 in the US and it's more like 25-30% of all deaths. Different in the developing world since many more people die of communicable diseases.

You're going to run into walls of denial with people refuting what they're hearing because the lifestyle change is hard or, worse, because they're coping with addiction and don't see a way out.

I agree. Talking about cancer risk is not the way to go to motivate people to stop drinking. There are far more significant and clinically relevant risk from using alcohol than cancer, such as cardiovascular disease.

Reply on news.ycombinator.com