Would you take the death test?

This is from the UK Daily Mail so it’s probably bulltish, but would you take the death test if were accurate?

Doctors can tell if I have cancer, heart disease or diabetes by simply taking sample fo my blood? I call BS.

Moderator Action

Since this is basically an informal poll of user opinions, let’s move this to IMHO.

Moving thread from General Questions to In My Humble Opinion.

Are you kidding? Because blood testing can quickly diagnose all kinds of stuff now.

This test surely can’t actually predict death: All it can do, at most, is identify risk factors. And ince you know what risk factors you have, you can take steps to mitigate them.

That said, though, 5 years isn’t really a particularly useful window. Most things that will kill you will start showing major symptoms earlier than 5 years. If you’re really that close to death, you’d probably know it anyway, even without the blood test.

Most people probably won’t I understand, even if they theoretically think they will. There is a “death test” for Huntington’s disease that lets potential victims know with fair accuracy how long they have, but most don’t want to know I’ve read.

I guess it depends on how you define predict. There are plenty of models that predict death using blood tests and other factors. If you mean - the test will be able to tell with 100% certainty that bob will die on June 4, 2016 - no - a test can’t do that.

However tests can and do provide predictions - such as you have a 12% chance of dying in 5 years. If you properly design the models - these tests will be accurate (but not able to take in knowledge of future changes - such as war - new poisons or environmental factors).

There are tests that if it says (to say 100 people) you have a 22% chance of dying in ten years - you will find around 22 of them dead ten years later - it might be 18 or 24, but it would be a very poor model that would come back with 2 or 50.

Sometimes the models aren’t THAT much better than best case educated guessing (using a life expectancy chart for example), but as long as they are better than that using rigorous techniques (done correctly) such as cross validation and statistical analysis - they do have predictive power.

I agree it is probably unlikely that most normally healthy people would find anything interesting in the test, but in theory it is possible. However - if it was THAT good - we’d probably be hearing more about it.

Can it tell me what model of bus I will be hit by? 'Cause that would be impressive.

Only if it’s just about to hit you.

We’re going to be hearing more about it. A lot more, whether it’s debunking it or supporting it. This was just the second study, done as a follow up for a study that had it’s own researchers scratching their heads and wondering if they did something wrong.

Thing is, at this point, it doesn’t tell you what you’re likely to die of, just that you’re likely to die within the next 5 years:

So…interesting and potentially terrifying, but I’m not sure how clinically useful it is at this point. (Neither are the researchers.) The best advice for anyone who tests high in these biomarkers is still pretty much the same advice for all of us: eat better, exercise, quit smoking. Perhaps it would be motivation to do those things, but it might also be demotivation.

To figure out what you’re likely to die *of *to more effectively monitor and test and treat for that thing still requires a doctor to look for heart disease and cancer and whatever else triggers these biomarkers.

Would I get the test at this point? No, I don’t think I would. But I have to admit it would be tempting to get my husband tested. Being older and in poor health, I’d be reassured if he tested low in these biomarkers, and perhaps buy a house or make other long term plans with him, which I’m not comfortable doing now. But I suspect since his health is already sketchy, it would only reinforce my fears that I don’t have much longer with him, and I’m not sure I could take that. I think this is one area where I prefer to remain ignorant.

Its nonsense.

When you find out about a risk factor, you can avoid the risk.
eg On finding out that you have a risk of developing type two diabetes, you can avoid being on the path to get it.

If there are specific markers, and the test can be validated across populations and age groups, yes, such predictions are possible.

I have my blood sliced and diced routinely, but do not recognize any of those markers as being among the numbers that get spit out.

Unless there are standard insurance-company-pays tests for these markers, validation is going to be slow.

Statistically, I have between 8 and 12 years left. Grandparents typically lived longer than parents - both parents died at 73 of lunf cancer from smoking. I quit after 23 years of smoking. Started late, so my lungs were relatively clean while growing.
I did predict “secondary smoke” as a child, in an enclosed car with 2 heavy smokers (vacation trips) “I’m going to get lung cancer by induction”. God, I was a smart kid…

Let’s not forget how statistics work: the math is good for large groups, but tells you nothing about any specific individual.
684 people died out of 17000. So after 683 deaths, there’s one more unlucky guy in the group. But you can’t tell which person of the 16317 is the next in line.