I find most of this post persuasive, but it seems to me like there shouldn't be excessive certainty on either the yes or no column here. The author's claim is not "no, this is false" but "the study is not bad, for what it is, but is highly limited in the conclusions that can be drawn from it, and it has some serious limitations." It seems reasonable to remain agnostic for now and pursue further studies on the topic that attempt to more accurately correct for other confounding variables.
This also gives me pause: "It is certainly possible that some pesticide is causing adverse health effects at the currently used levels (although there is no current evidence for this)."
That claim seems to me to be false if we include studies on occupational use of pesticides [1]. It could definitely be the case that the residues that ordinary consumers encounter are entirely non-toxic, but again, it seems to me like the jury is still out on that.
The other correlation - that appeared strong was with women being vastly more likely to eat organic. Using similar logic to the headline, eating organic food makes you a woman.
Among my friends and coworkers, those who tend to eat more organic food also do a lot more physical activity/exercise than those who don't. This smells like correlation.
From the article: Participants in the French study also provided information about their general health status, their occupation, education, income and other details, like whether they smoked. Since people who eat organic food tend to be health-conscious and may benefit from other healthful behaviors, and also tend to have higher incomes and more years of education than those who don’t eat organic, the researchers made adjustments to account for differences in these characteristics, as well as such factors as physical activity, smoking, use of alcohol, a family history of cancer and weight.
Even after these adjustments, the most frequent consumers of organic food had 76 percent fewer lymphomas, with 86 percent fewer non-Hodgkin’s lymphomas, and a 34 percent reduction in breast cancers that develop after menopause.
Granted, it could be that these socio-economic factors weren't adequately corrected for, but they did at least attempt to.
The study controlled for income, other dietary factors, smoking, etc. Obviously there's only so much you can do along those lines in a self-reported study for a relatively rare phenomena (happily, not many people get cancer in a given year). But the effect seems pretty robust.
(interestingly, one thing they controlled for was ex-smokers, and not because they were less likely to eat organic but because they were more likely to do so, which I guess makes sense when you think about it, even if it seems counter-intuitive at first glance)
<Previous research has demonstrated other potential beneficial effects of eating an organic diet, such as a lowered level of pesticides in urine samples>
Well this can only be a good thing..
i find it unacceptable that pesticides should be found in urine from eating anything in a modern society.
People that drink exactly one glass of red wine per day are wealthier than average, don't over indulge, and live longer. Ergo, red wine makes you live longer.
A thing that really annoys me on HN (though it's certainly not specific to this website but more general to any place where the users self style as 'intellectuals' or 'rationally minded' or whatever) is the amount of smart-sounding comments that betray a very cursory reading of what is being posted in the first place. For instance:
-Yes, we know correlation isn't causation. Don't you think the authors aren't aware of that, you know, being scientists and all? That doesn't mean they aren't useful and shouldn't be taken into consideration for making policies. It doesn't mean people shouldn't investigate correlations in the search of a potential causal link. For the record, many therapeutic drugs are approved way before the exact mechanism causing the therapeutic effect is known.
-Yes, they did control for all the obvious things such as socio-economic factors and so on. You'd know it if you actually read the paper, but searching for confounding factors is literally high school statistics. Don't you think the PhD-level authors wouldn't have thought about that?
-No, no one ever said that there even was a causal link in the first place. It's explicitly acknowledged in the paper. What makes it interesting, though, is: 1) the large sample size and observation period, and 2) the prospective aspect of it, i.e. they simply waited for possible cancer outcomes instead of actively going for them and hunting for correlations. This makes the findings significantly more solid.
Please, remember the guidelines: always take the most charitable interpretation of what someone says. Do assume that scientists are capable of doing their jobs, and whatever you've thought of in the few minutes of attention you gave this link, chances are the authors thought about it in the seven-odd years it took to publish it.
Not discounting the benefits or organic food, but this specific study isn't very useful: there's certainly a correlation between eating organics and lower cancer rates, but there are too many other plausible variables (ie people who buy organics are more healthy in general: exercise more, have more disposable income (therefore higher quality of life), are more mindful of other environmental problems, etc).
What would actually be interesting is a double blind study: a group of people get a weekly box of unbranded vegetables, either organic or regular, then track health/cancer rates/whatever.
As noted elsewhere, the study accounted for those variables: "Participants in the French study also provided information about their general health status, their occupation, education, income and other details, like whether they smoked. Since people who eat organic food tend to be health-conscious and may benefit from other healthful behaviors, and also tend to have higher incomes and more years of education than those who don’t eat organic, the researchers made adjustments to account for differences in these characteristics, as well as such factors as physical activity, smoking, use of alcohol, a family history of cancer and weight.
Even after these adjustments, the most frequent consumers of organic food had 76 percent fewer lymphomas, with 86 percent fewer non-Hodgkin’s lymphomas, and a 34 percent reduction in breast cancers that develop after menopause."
The study could have controlled for income, as it should for smoking, exercise, etc. Richer people also eat out more often (some, all the time) and usually are served the cheapest ingredients there, except for very expensive venues.
> Richer people also eat out more often (some, all the time) and usually are served the cheapest ingredients there, except for very expensive venues.
I'm not sure about that last part. It's not hard to find moderately priced restaurants using quality and/or organic ingredients. It won't be fast food prices, obviously, but it isn't going to break the bank, either.
IDK, there’s so many reports of rancid olive oil and the wrong type of fish being served. Ingredients are where restaurants can really save costs without anyone noticing.
Your question implies you believe that a placebo-controlled trial is the only kind of medical study that carries any weight.
This is not true. It's only one kind of study.
In this case, the study was looking for a correlation between diet and cancer diagnoses. All it needs to be valid is to conform to the scientific method. Placebo-controlled therapeutic trials are not the only kinds of studies that conform to the scientific method, and indeed not all of them do.
Please try to avoid reflexively invoking this trope to unfairly dismiss potentially interesting studies.
I don't think that's uncommon. You have something which you think reduces problem X, and you give it to half of your population, and a placebo to the other half. Over a long period of time, you observe the rate of problem X across both populations. I believe we have done this with many different kinds of problems, including cancer, heart disease, arthritis, etc.
Your reply is unnecessarily combative and short. Assume good faith in grandparent's reply, and preferably try to find out yourself and report back your findings.
>> Design, Setting, and Participants In this population-based prospective cohort study among French adult volunteers, data were included from participants with available information on organic food consumption frequency and dietary intake. For 16 products, participants reported their consumption frequency of labeled organic foods (never, occasionally, or most of the time).
https://theness.com/neurologicablog/index.php/problems-with-...