Wednesday, October 12, 2011

Take your vitamins, dear!

Well, score another big one for science....big bust, that is.  It now turns out, a news story about a new paper reports, that vitamins can shorten your life! 
Experts have suspected for some time that supplements may only be beneficial if a person is deficient in a nutrient.
And excess may even harm, as the study in Archives of Internal Medicine finds.
The study was based on recall, a big caveat, but the idea is that enhancing a balanced diet with supplements, particularly iron, can be harmful.  The conclusion is that supplements should be taken only by people deficient in some nutrient.  Unless it's calcium -- according to this study, supplementing with calcium may extend your life.  The authors don't know why. 

So much for the antioxidants, beta carotenes, bone strengthener, flu-resisting, pep and vim guaranteeing One-A-Days.  Watch what you put into your body, apparently, if you want to have that body for as long as possible.

Now, this study is about 'older women', but why should we assume that's the only impact of vitamin o'ding?  If we can't trust the legions of studies (expensive, by the way, on your tab) that have pushed us towards vitamin-doping, why trust this one?  If alphabet popping (A's, B's, C's, D's, and E's) is not so good for older women, why would this not also be so for younger ones, or for men, even?  After all, a lot of things may change in a lifetime, but not genotypes, and here 'older' includes women just in their 50's.

How is it that something that would seem to be rather straightforward is still poorly understood?  Vitamins are easy to identify, specific molecules, and pill dosages are also clear.  If we can't actually estimate exposure because we don't know how often respondents actually pop their V's, or how much they get the old fashioned way (by eating food), then perhaps we can't totally trust the studies or the studies can't adequately evaluate things.  They have to do this by statistical methods--more vitamins leading to statistically significant lower lifespans (significance doesn't mean importance, by the way, as we have often said, and the significance level by which one claims to have made a finding is inherently a subjective judgment).

The problem is not a criticism of the authors of this report--they do or don't deserve that based on the paper itself, nor of the journals (though, don't they always deserve it for over-dramatising results?).  The problem is in the nature of this kind of science, or the kinds of questions being asked.  The fault if there is one, is in assuming that this kind of knowledge is more sound and stable than it really is.  For example, suppose the study is as solid as the rock of Gibraltar.  Effects of B6 on lifespan may be affected by all sorts of other lifestyle factors.  The same may not be found in, say, Burkina Fos0 today, or say, Norway twenty years from now (when you will have to worry about them!).

Evolution and many other aspects of the life and health sciences share these problems, but none of these fields (yet?) have the level of serious, basic reflection it may take to do better or to ask better questions.  Until this happens, we will be stuck with the question of whether or not to be Tocopherol topers.


Anne Buchanan said...

And just this morning there's a story about vitamin E supplements raising the risk of death from prostate cancer -- and that from a study that set out to show that supplementation protected against prostate cancer, as previous results had suggested.

Predictably enough, "Duffy MacKay of the Council for Responsible Nutrition, a supplement makers' trade group, said the study shouldn't be interpreted as questioning the benefits of vitamin E as an essential nutrient, and he said there is evidence that many Americans don't get enough."

Holly Dunsworth said...

It's a conundrum because all the stuff we buy and ingest because we're told we "need it" and we believe it... all that stuff is made by and peddled to us by people who need those jobs to buy that stuff.

And it's the astronauts-of-the-future who are the real victims of this pill problem.

Anne Buchanan said...

I agree, that's a fundamental conundrum. Complicated by the fact that, if science can't help us make these decisions, and we know that people who make and sell the stuff can't tell us not to take it, and can't give us an unbiased account, on what basis do we decide what to do?

Vitamins are a pretty small dilemma (and the increase in mortality risk reported is pretty small), but what about mammograms or PSAs or policy issues like whether to change the way health care is delivered in the US, where the consequences of doing or not doing it can be large? If we can't rely on biology, or epidemiology, or economics, to make these policy decisions, we're pretty much left with politics....

Or the old truism, moderation in all things.

Holly Dunsworth said...

This makes me think about human longevity and how weird it is. What if the multitude of increasingly confounding mortality parameters as we age far beyond other mammals makes statistical tools powerless?

Anne Buchanan said...

If the confounding thing about human longevity is the unique accumulation of environmental exposures we each experience x our unique genomes x years of exposure to both, then it is definitely hard to know how our statistical tools can tease out causation. Yep.

Ken Weiss said...

Human longevity seems unrelated to any evolutionary genetic 'mandates' about our life histories. Our environment is now so very different from environments in the past, that what has become most important, perhaps, is how we wear out, not what 'gets' us.

Wearing out by old age was certainly not what happened in the past, though it may sometimes have been that older people (say, 50s) were slowed down enough, or had reduced enough vision, etc. as to be more likely to die off. But the evidence is more that what got us was on the whole much earlier and not related to just wearing out.

So vitamins, a minor aspect of our environmental exposures in the west, can do almost anything at the low level they do (or don't), and be very hard to detect or understand.

In any evolutionary perspective, it really seems like much ado about very little--even if the details of risk to your own or my own personal survival and life-experience seem so important.

Holly Dunsworth said...

I don't think that having a comparative perspective on human longevity requires that we first assume that humans evolved to live long lives. I think the fact that we do live a long time means that we probably face more complex and confounding mortality pressures and competing causes of death (if from nothing else than wearing out) than other animals which must make searching for the fountain of youth with mouse models that much more frustrating.

Ken Weiss said...

There has been an enormous amount of work on longevity and its evolution and determinants. While lots of hypotheses and tidy explanations have been offered (like telomere shortening), they have not really added much to the overall picture. Indeed, explanations that invoke human exceptionalism have been favorites among anthropologists, but with scant (if any) scientifically sound support.

Basically, the typical lifespan of mammals is correlated with their body size. The correlation is not perfect, but it's a clear pattern. Humans are basically on the curve with other mammals. No particular special explanations are needed.

It is difficult to know how the complexities of our modern environments relate to this, except we know that longevity is the result of patterns of survival of a host of independent and biologically different kinds of disorder (the competing causes you refer to).

I think the evidence is that other animals face the same kinds of complex causation (after all, they are biologically very similar to us), but lack of health care and predators do away with them before cancer or diabetes does.

One might think (as many or most in the field do) that mice or dogs will be a good model, but work has been frustrating as you say. To me, the key question is how a mouse with essentially the same genome as a human in so many ways, lives 2 years (in a protected lab environment) while a human lives 82.

The correlation with body size has long been known but has not yet been explained, but must be if we are to learn all that we might from other animals.