lundi 24 septembre 2012

Of Organic Foods and Meta Analyses


There has been much bruiting in the conventional media and the blogosphere, specifically amongst foodies, about a recent study published in the Annals of Internal Medicine called “Are Organic Foods Safer or Healthier Than Conventional Alternatives?: A Systematic Review.” I’m not a huge fan of organic foods (or a big detractor for that matter), and didn’t pay a great deal of attention to these stories when they came out (see, for example, one via NPR headlined “Why Organic Food May Not Be Healthier For You”), but while perusing the Green Grok blog, which clarifies in great detail what the study actually demonstrates, I was motivated to get a copy of the study and see what it says, rather than what the media said it said.
There has in fact been very little research that has demonstrated conclusively that organic is healthier (but in a moment, as the Green Grok’s Bill Chameides points out, it depends what we mean by “healthier”).
It’s important to stress that the current study is what is known as a “meta-analysis,” or an extensive roundup of previous studies done on a topic that uses statistical techniques to tease out where studies confirm each other, where they refute each other, and basically identify other relationships among the studies. The essential point, though, is that this was not a brand-new study that looked directly at the health benefits of organic foods.
So, the authors did a search of relevant studies and:
Two hundred thirty-seven studies met inclusion criteria: 17 evaluated health outcomes among human populations consuming organic and conventional foods (54 –70); 223 compared organic and conventional fruits, vegetables, grains, meats, poultry, milk, or eggs directly (50 –53, 57, 65, 69, 71–286) (3 reported on both human and food outcomes).
So really on 7% of the studies they analyzed directly looked at people consuming organic  vs. conventional food. That’s a pretty scanty base upon which to be making overarching claims about health.
And as for those 17 studies:
Seventeen articles describing 14 unique populations (13 806 participants) met inclusion criteria (Supplement 3, available at www.annals.org). Study designs varied: 6 randomized, controlled trials (56, 57, 62, 65, 66, 69), 2 prospective cohort studies (54, 61), 3 cross-sectional studies (55, 64, 68), 4 crossover studies (describing 2 populations) (59, 60, 63, 67), and 1 case– control study (70). Only 3 studies (61, 64, 70) examined clinical outcomes (for example, wheezing, allergic symptoms, or reported Campylobacter infections), and the remaining studies examined health markers (for example, serum lipid or vitamin levels).
So even among those relevant studies there was a great deal of heterogeneity.
In a nutshell, in order to answer the basic question, “is organic food healthier than conventional foods?” you would need to conduct a long-term study that followed one group of subjects who ate exclusively organic and another group that ate exclusively conventional, and then report their health outcomes after some period of time. Such studies may actually be underway somewhere, I don’t know, but at present the kinds of data required to make conclusive statements about organic food and human health is simply not available.
Now, my intent is not to debunk the study (for what it is—a meta-analysis—it’s quite good at synthesizing what research has been done in this area, although some statistical legerdemain may have crept in), nor to rise up in defense of organic food (as I said, I am fairly blasé on the subject, quite frankly), but rather to point out that sometimes studies can be misrepresented in the media—and sometimes by a study’s own authors. The paper’s conclusion says that:
The published literature lacks strong evidence that organic foods are significantly more nutritious than conventional foods. Consumption of organic foods may reduce exposure to pesticide residues and antibiotic-resistant bacteria.
And this is true: in terms of basic nutrition, there is scant evidence that organic foods contain significantly greater amounts of nutrients than conventional foods. But human health is not just about nutrient quantity and/or quality. Avoiding pesticide residues may be a good thing as far as human health is concerned, as might avoiding antibiotic-resistant bacteria. So if we define health in that way, perhaps organic is better than conventional.
I have nothing against the organic food movement, and if you want to support organic food as a political gesture, or as being environmentally beneficial, or even that it tastes better, great. But we need to be careful that we don’t ascribe benefits to something when the evidence is still too slim to conclusively demonstrate.

Source: WhatTheyThink (http://goo.gl/7MrPp)

Aucun commentaire:

Enregistrer un commentaire