Monday, March 4, 2013

It's in your scenes, not your genes

The Big Health Story of last week was the Mediterranean diet is good for you story, that we blogged about at the time.  Of course, it didn't take but a couple of days before the Epidemiology Establishment--facing NIH funding cuts--decided to proclaim the need for 'more research' to see what it is in almonds or wine or olive oil that does the miracles, er, the modest reductions in heart disease--Gina Kolata broke the Big Story in the NYT, and also the follow-up.

Of course, in basic terms we've known for years or even decades that the Mediterranean diet has benefits.  The basic dietery differences between lower and higher risk components and habits are clear, even if the details down to how many ounces of McFood you can eat per year per tenth of a percent reduction in some disease risks (in some people) are not known.  The main findings were dramatically demonstrated years ago in a study in Finland, and in many others, as back issues of the NY Times will show, and components have been incorporated into a host of popular diets, and even the US government's dietary recommendations.  So, the story isn't really new, no matter how loudly it's been tooted now.

And, the new study didn't add anything at all new to the methodological toolkit of nutritional studies or epidemiological studies in general, and studies of diet and nutrition are notoriously unreliable.  Even this supposed ground-breaker did a poor job with its control group, and it's perfectly possible that the reason the experimental groups, who added nuts or olive oil to their diets, were as compliant with the diet as they were was because they were given weekly allotments.  So, how that could be translated to the general public changing their diets long-term is hard to figure, as implementing health education is a difficult task.  So why all this new research won't just be more of the same, giving us conflicting, hard-to-interpret results that are hard to talk people into anyway (well, maybe the daily glass of red wine wouldn't be so hard) is not clear.

What is afoot, rather than a massive effort to help Americans alter their diets and gain better health, is of course lobbying for 'more research'.  If you are naive (very naive) you may not realize that what every epidemiologist from here to Timbuktu will propose will be very large, very long, very comprehensive follow-up studies to determine the tenths-of-percent effects of dietary components that we absolutely need to document (and, it is not just cynical to say so, to keep each of their labs running handsomely for many years as a not-so-incidental by-product).

The nature of all of this, from the New England Journal's hoopla release of the study, to the modest findings and questionable data of the study itself, to the amnesia over the long-standing knowledge of the same basic findings, all shows the irresponsibility of so much that is wrong with NIH (and its reciprocal copy-cats in the EU science establishments).

Again, the study's findings seem almost patently plausible given what we know.  There is nothing serious to be gained by pouring huge amounts of money to fine-tune the results.  The complexity of lifestyle risk effects and their great difficulty in being accurately ascertained, are huge relative to the bottom line we already basically know.  The funds that schools of public health will suck up, relative to what smaller, but seriously focused studies, could do for countless junior, perhaps more deeply thinking investigators, nor what real changes in health-habits could do to make enormous reductions in disease and medical care burdens in western countries are a symptom of the sickness of our health establishment.  How to convince people to change their health habits--health education--is still a fundamental question; perhaps they money that's going to be sunk into figuring out just how many almonds and tablespoons of olive oil we should be eating everyday, should be diverted to figuring out how to do effective health education instead.

Environment vs genetics
It is slightly refreshing, we must admit, to see a study that doesn't slyly say we need to look for genetic susceptibility to the bad effects of not eating almonds or olive oil.  Clearly most instances of the diseases that are being GWASed and subject to other similar types of study are not in our genes but in our scenes--how we live and what we eat.

Of course, the gold standard in science is the ability to use what one thinks one understands to make predictions.  Predictions borne out tend to confirm one's understanding.  Anybody want to take bets as to how long it is before the geneticists, seeing the threat of lots of funding shifting to these epidemiological boondoggles, jump in, asserting that the key is to do the genetics of response or risk related to these diets?

We rant regularly on this here, because there is so much waste, and minor incremental research at high cost (and this, a time of budgetary restriction and lack of science jobs), that somebody needs to at least point it out again and again.  We don't have a horse in this race, and nothing to lose or gain by expressing our views.  The likelihood of change is very small, but the need for it is great.


Javier Blanco said...

Another one from Gina: "5 Disorders Share Genetic Risk Factors, Study Finds" (The New York Times, February 28, 2013).

Anne Buchanan said...

Yes, she's hard to keep up with. And she's one of the best science reporters at the Times. But still, one wonders, is it a requirement for science journalists that they not question anything?