[resoluteteeth]

If multivitamins are really improving things like sleep and strength why is that not showing up in mortality? Isn't there evidence that sleep quality and strength in later life are negatively correlated with mortality?

Feelings of well-being have also been shown to be negatively correlated with morality.

[zosima]

A correlating with B and B not correlating with C does not imply lack of correlation between A and C.

[resoluteteeth]

Things like sleep, sense of well-being, and strength are inversely correlated with mortality.

If this is a causal relationship, and vitamins cause these things to improve, then mortality should also improve. The fact that this is not the case therefore suggests that either 1) the relationship between these things and mortality isn't causal, 2) vitamins are not improving these things (or I suppose, 3) the relationship is causal and vitamins are decreasing mortality by improving these things but also increasing mortality by other effects in a way that balances out).

While it is theoretically possible that none of these things have a causal relationship with mortality and it would therefore be be possible to improve them without decreasing mortality, I'm not sure it's likely.

It just seems pretty improbable to me that multivitamins are having all these wide-ranging health benefits that people are claiming and none of these benefits are having even the slightest effect on mortality.

For this reason, I think evidence that vitamins don't decrease mortality is actually pretty strong evidence that they are not having these other effects.

[guerrilla]

You've made a major mistake. It may be that vitamins do improve all of those things and would improve mortality as a consequence but something else keeps mortality the same, for example heart disease and cancer.

[resoluteteeth]

> It may be that vitamins do improve all of those things and would improve mortality as a consequence but something else keeps mortality the same, for example heart disease and cancer.

Are you saying that vitamins have improved mortality over time but increasing heart disease and cancer rates have balanced that out over time? If so, that's not what's being talked about here. The study in the article is comparing people who take multivitamins with people who don't, not looking at the overall rate in multivitamin usage compared to mortality over time in the entire population.

Or are you saying that people who take vitamins are more likely to have heart disease and cancer? That can be an issue in this type of study but there's no reason to think that's true. If anything, I think there is a concern that the confounding factors go in the opposite direction, where people who take vitamins are more likely to do other things that improve their health like exercise more. However, the study in the article also controlled for various potential confounding factors like BMI and physical activity.

[guerrilla]

No, that's not what I'm saying.

I think I was clear, but I'll spell it out for you:

Imagine this possibility. Everyone gets cancer and dies at 65 years old. Everyone starts taking vitamins and therefore starts sleeping better, getting stronger, etc. They all still die at 65 years old though because of cancer. Suddenly, cancer is cured. All those people now live to 85, except the ones who still don't take vitamins. They only live to 66. Vitamins caused a 20 year increase in lifespan, but this was undetectable because cancer was killing everyone at 65. In all cases though, everyone slept better and were stronger after taking vitamins.

Mortality is a terrible measure of whether people are sleeping better and stronger after taking vitamins. It would be better to measure sleep and strength directly if that's what one wants to know about. I hope your mistake is clear to you now.

[resoluteteeth]

> Imagine this possibility. Everyone gets cancer and dies at 65 years old. Everyone starts taking vitamins and therefore starts sleeping better, getting stronger, etc. They all still die at 65 years old though because of cancer. Suddenly, cancer is cured. All those people now live to 85, except the ones who still don't take vitamins. They only live to 66. Vitamins caused a 20 year increase in lifespan, but this was undetectable because cancer was killing everyone at 65. In all cases though, everyone slept better and were stronger after taking vitamins.

Cancer isn't killing everyone at 65 years old. Not everyone dies of cancer at all. Even if you assume that vitamins have no effect on cancer rates, if they decrease mortality due to other causes, people who take vitamins should have lower overall morality than people people who don't take vitamins.

So again, you would have to have a situation where vitamins are improving sleep and strength, which are both inversely correlated with mortality, but somehow not decreasing mortality, as I said in my previous comment.