Vegetarians Less Healthy, Lower Quality Of Life Than Meat-Eaters
plosone.org
plosone.org
I've been vegetarian for 20 years, but the health aspects of the diet are only about 10% of why I chose to be a vegetarian. I bet that's true for a large number of vegetarians.
This means the study suffers from the problems that plague all "retrospective" studies, studies of populations who have already made their choices for unknown reasons. It is this lack of control that calls any meaningful conclusions into question.
A prospective study by contrast, obviously totally impractical, would empanel subjects who would be randomly made to eat either vegetables or meat, and the results would be carefully examined. That might be science, this is not.
Woah there.
Science n. the systematic study of the structure and behavior of the physical and natural world through _observation_ and experiment.
This paper does not cover a double-blind randomized controlled-trial. That does not mean that a retrospective matched sample studies are not science. The conclusions are actually pretty reasonable for this type of study they essentially say "we need to make sure the nutrition of vegetarians is adequate, in Austria," "we also need to study this more, in Austria."
The different in abundance for mental illness is pretty striking, and seems worth investigating. Previous studies (e.g. [1]) have also linked low-fat diets to change in mental state, the outstanding control-intervention literature combined with retrospective studies make a compelling argument for further research.
All that said, the link-bait HN title is pretty bad.
Wait -- you're going to rely on a dictionary definition of science? Dictionaries don't define words, they can only report how people use words. This is why, as just one example, literally and figuratively mean the same thing -- because that's how people choose to use those words.
> This paper does not cover a double-blind randomized controlled-trial. That does not mean that a retrospective matched sample studies are not science.
Yes, in point of fact, that is exactly what it means -- it cannot distinguish causes from effects, and it cannot produce reliable conclusions about nature, or lead to testable, falsifiable theories about nature.
> The conclusions are actually pretty reasonable for this type of study ...
Yes, that is absolutely correct -- they're "pretty reasonable for this type of study". No dispute.
In science -- real science, not the dictionary version -- observations must successfully isolate the thing being studied, and this is something that retrospective studies cannot do. The study links vegetarianism to certain health issues, just as puddles are successfully linked to rain.
Science is not description, science is explanation -- explanations that are tentative, and that are open to falsification. Without a theory about an observation (an explanation) there can be no science.
> Previous studies (e.g. [1]) have also linked ...
There's my favorite pop-science buzzword: "linked". When unscientific people hear "linked", they think a cause-effect relationship has been established. But that's not what "linked" means, it's what the public thinks it means. It's the same with literally and figuratively -- they are also regularly confused by an ignorant public.
> The different in abundance for mental illness is pretty striking, and seems worth investigating.
You're missing the point that mental health issues might be a cause of vegetarianism, not an effect, or both vegetarianism and poor mental health are the result of some tertiary cause this study doesn't address.
Until an explanation is offered and then tested, until causes and effects can be unraveled, it's the most sordid kind of data mining, and it's ... not ... science.
What exactly is your problem with it ? What is the problem with retrospective-scare-quoted studies ? I personally have more problems with non-retrospective studies for obvious reasons.
Like all academic papers there are loads of problems with the research and conclusions, mostly stemming from an inability to do real experiments on humans and limited budgets, and of course these do introduce caveats. But if you're going to start judging papers on sample size ... first you don't understand statistics, secondly these guys are using more people than most (because it's such a widespread phenomenon, it's easy to get a big sample size of course, but still).
And frankly, artificially limiting your intake of different foods ... results in some chemicals not being available to your body ... first of course results in weight loss (because a lot of these require a lot of chemical energy to manufacture, or can't be manufactured at all) ... and then result in bigger problems.
It is only outside of medical circles that the benefits of modern food are in question. Nobody in the medical establishment doubts that processed, meaty and fatty foods are pretty much the best nutrition possible for the human body. Just yesterday I read an article about what everyone thought was an epidemic of the plague, at the end of the last century in Chicago. It wasn't, it was contaminated grain that made it into bread, was effectively invisible, and quite deadly. For obvious reasons this grain was grown perfectly natural, if only because modern stuff didn't exist yet. Yes, it makes you fat. Yes, there are historical screwups (not nearly as much screwups as with "natural" food of course, but somehow the screwups with industrial food are judged infinitely more important).
And yes, there are some, including medical researchers and so on, that doubt the morality of processed foods. I find this distasteful. Medical research and all science should first and foremost be value-free. Morality should be shut out (with the potential exception of describing methods to do direct large scale damage, ie. I find it acceptable that the exact gene sequence and samples for say the bubonic plague does require you to sign a letter you won't use it for nefarious purposes, even if I would have major problems with similar requirements for the cure).
Are you referring to the 20th century, which is currently the last century?
As retrospective studies go, it's not bad. But retrospective studies have a well-deserved and terrible reputation for confusing causes and effects.
> But if you're going to start judging papers on sample size ...
I invite you to locate the sentence where I complained about the sample size.
> Nobody in the medical establishment doubts that processed, meaty and fatty foods are pretty much the best nutrition possible for the human body.
Apart from your obviously false claim, The "medical establishment" are not scientists and are not competent to make that call. They are not researchers, they are practitioners of methods developed by others.
> I find this distasteful.
That's your argument? Please -- in a discussion of science, one is expected to produce something resembling evidence, not emotion.
You need to learn why retrospective studies have the poor reputation they do. You need to learn that they cannot distinguish causes and effects, and why. You need to learn how science works.
So do the double-blind studies you appear to be advocating. Ever thought of what exactly the placebo effect is and what some causes might be ?
> Apart from your obviously false claim, The "medical establishment" are not scientists and are not competent to make that call. They are not researchers, they are practitioners of methods developed by others.
The medical establishment are people who treat the sick, mostly. They observe that they regularly have people who come in with nutritional deficiencies because they are vegetarians and didn't follow the diets properly. One look at such a diet tells you exactly why this happens. Second is that you observe that vegetarianism and eating disorders go hand in hand.
You know what contains every single substance your body needs on a daily basis, all vitamins, minerals, calcium, you name it and will prevent nearly every food disorder, and it's hypo-allergenic (meaning perfectly safe for very nearly every patient you have) ? A 150gr steak. The only thing it doesn't have is sufficient calories, so you should also have something starchy with it. (And to be pedantic, it's the fatty tissue in the steak that's the main source of it's good properties. And yes, we're talking a rare to medium steak here)
You know what vegetarian food contains every single substance your body needs on a daily basis ... None. What should a vegetarian eat daily ? * Nuts * Beans * Spinach or soy (or tofu, even if that's less ideal) * Sunlight (yes, really), or Vitamin D supplements * Vitamin B12 supplements
Needless to say, this list is very much not hypo-allergenic, there is a significant fraction of the population (not that many, granted, but you have to check) that will swell up and start convulsing if you feed this to them. There are obviously other vegetarian diets, but they have similar problems.
True, missing one of these for a day is very much unlikely to kill you, but prolonged shortages will. Also keep in mind that you should never raise a kid vegetarian, because that is much riskier than being vegetarian as adults. There have been cases of kids dying because of vegan diets.
What I find extremely convenient is the physical symptoms of low-level protein deficiency. First : it's really, really bad for you. But the initial symptoms are sleeping more, weight loss and tissue loss (ie. not just fatty tissue). That's imho why people do this : because it slims you down.
Incidentally, the one thing that is confirmed to slim you down by removing fatty tissue is cold. And if you don't overdo it (e.g. go swimming in a 24 degree water pool) it is quite safe. It's not about feeling cold, merely about increasing your body's heat loss.
http://www.nih.gov/researchmatters/june2013/06102013vegetari...
"Overall, our findings reveal that vegetarians report poorer health, follow medical treatment more frequently, have worse preventive health care practices, and have a lower quality of life."
Which is unfortunately too long for HN submission, but is the (in fact, verbatim) published summary. A revised headline might read, Without any substantial editorialization or modification of meaning:
"Vegetarians report poorer health...and...lower quality of life."
Aruably the word "report" is more neutral than "have", but is substantially similar to the submission. All that being said if you guys want to change it...it is absolutely your perogotive. Its just hard to avoid making certain compromises (imperfect title or imperfect article). This is unfortunately one of those corner cases IMHO.
It goes without saying that the published academic title ishopelessly vague for anyone not already familiar with the results. To wit:
"The Association between Eating Behavior and Various Health Parameters: A Matched Sample Study"
Has infinite degrees of freedom.
[1] An third party edited and published title highlight what it is 'newsworthy' (ie, this is not self edited). The original research was submitted (ie, mismatched) to increase the quality and availability of the information provided to readers.
Actually that's not what the research concludes. You left out "Austrian adults who eat a vegetarian diet". That's different than American vegetarians, who are a much larger group with a different and longer culture of being aware of issues like B12, etc.
On the substance of your point, however, I appreciate the comments. A discussion of the limitations of the study cohort and methodology is certainly a good discussion topic.
I would caution the notion "americans" are superiour at being vegetarian. Other cultures are more natively vegetarian. EG, The Indian Sub-continent and Bhuddists in Nepal, etc. And have genetic and cultural adaptations much longer in gestation as a result.
Also, context switching doesn't solve the key issue: non-vegetarians veg-native cultures also self-selct. Ethnic variation is also a vexing problem in this area because of phenotypic variation (race, lactose tolerance etc) linked with health.
A study of vegetarian-dominant cultures may show something entirely different. Yet also not be applicable to the US population.