Quantifying research.

March 22, 2010 12:12 PM

Is Vitamin C worth taking or not? Does Echinacea kill colds? Am I missing out not drinking litres of Goji juice, wheatgrass extract and flaxseed oil every day? A generative data-visualisation of all the scientific evidence for popular health supplements by David McCandless and Andy Perkins. (Still Image) (data) [via]

Note: Not perfect.

posted by Orange Pamplemousse (78 comments total)

well there is scurvy:
Scurvy is a disease resulting from a deficiency of vitamin C, which is required for the synthesis of collagen in humans.
http://en.wikipedia.org/wiki/Scurvy
posted by robbyrobs at 12:18 PM on March 22, 2010

Whoa. There's a supplement called "Devil's Claw", huh? That's kind of badass. And it's nice to see that it shows "promising" levels of evidence for the treatment of Slayer, even if it's below the "worth it" line for treatment of Judas Priest. (For the latter, most researchers recommend St. Halford's Wort)
posted by Greg Nog at 12:22 PM on March 22, 2010 [8 favorites]

well there is scurvy:

The operative word would be "supplement(s)".

That visualization is interesting. How can beta glucan have low evidence but the bubble sits so high?
posted by P.o.B. at 12:23 PM on March 22, 2010

This chart is not very helpful for one obvious reason: it charts separately the effectiveness of supplements for various conditions without giving me the capacity to segregate the data.

If there were a button that would hide all of the supplements that are not related to a particular condition (or vice-versa), this would be invaluable.
posted by jefficator at 12:25 PM on March 22, 2010

Nice to have my own bubble. Could this be why I have no urinary infections?
posted by Cranberry at 12:27 PM on March 22, 2010 [3 favorites]

jefficator, look at the right hand side. Hover over the "show me" text. Isn't that what you want?
posted by oddman at 12:27 PM on March 22, 2010 [1 favorite]

I don't pay super close attention to this stuff because my opinion is that worrying too much about all the things that might kill you is what actually kills you in the end. That said, this frontline show on alternative medicines is a really informative documentary.
posted by mcstayinskool at 12:29 PM on March 22, 2010 [3 favorites]

I have to question this chart/methodology because of how high it places St. Johns Wort. A number of recent studies have concluded that St. John's wort is ineffective in treating depression. I'd like this chart more if it weren't one person's views of effectiveness but rather was his visual representation of some more expert, group assessment of these various supplements.
posted by twsf at 12:36 PM on March 22, 2010

I'd like this chart more if it weren't one person's views of effectiveness but rather was his visual representation of some more expert, group assessment of these various supplements.

He links to the data the chart is based on. Unless I'm misreading his use of the data, it has a lot of sources (which he links to) and isn't one person's view.
posted by juv3nal at 12:40 PM on March 22, 2010 [1 favorite]

IIRC, that Frontline documentary pretty much skewers the purported efficacy of St. John's Wort for treatment of depression, and pays special attention to the problem of how test results are used to promote medicines even when they don't help the condition. Alternative medicines are not the only target for this criticism. See also: OTC cough suppressants.

Pretty data viz though :)
posted by mcstayinskool at 12:42 PM on March 22, 2010

show me... BONER PROBLEMS!
posted by nathancaswell at 12:43 PM on March 22, 2010 [1 favorite]

Well, twsf, he does share the data his graphic is based on. See the link in the OP.

My primary irritation with the graphic as it is is that the language he uses to label symptoms/diseases/systems effected etc is messy; I'd like to see it drawn on a controlled vocabulary, rather than lumping "sex," "skin," and "cancer".

I'd prefer to see adverse effects added but I realize that you have to control your scope.
posted by chesty_a_arthur at 12:43 PM on March 22, 2010

P.o.b How can beta glucan have low evidence but the bubble sits so high?

The size of the bubble is to do with popularity, based on google hits. So the small bubble just means it isn't being talked about much. The vertical axis is based on the amount of evidence. So beta-glucan isn't talked about much, but has good evidence for effectiveness.

Which also implies a load of snake-oil salesmen should get on the beta-glucan bandwagon pronto
posted by memebake at 12:47 PM on March 22, 2010

BONER PROBLEMS!

That's easy. You could go down to the supplement shop and get some Yohimbe or Horny Goat Weed right now. Anecdotally, I've heard they work great.

So beta-glucan isn't talked about much, but has good evidence for effectiveness.

But it's discolored, meaning that it has low evidence. Maybe the scant amount of evidence it has is solid enough to place it that high.
posted by P.o.B. at 12:53 PM on March 22, 2010

jefficator, look at the right hand side. Hover over the "show me" text. Isn't that what you want?
posted by oddman at 2:27 PM on March 22 [+] [!]

You are utterly correct and I retract my statement. The chart is now helpful.
posted by jefficator at 12:57 PM on March 22, 2010

Yohimbe, for instance, is available both as a vaguely constituted health food store formulation, or by prescription as yohimbe hydrochloride in reliable concentrations (typically much higher doses than the non-prescription versions.)
posted by StickyCarpet at 1:01 PM on March 22, 2010

P.o.B.: But it's discolored, meaning that it has low evidence.

I don't think the colours mean anything, except for brown which means 'few studies'. Are we looking at the same bubble? The one in the top left?
posted by memebake at 1:02 PM on March 22, 2010

I recently learned that vitamin C helps dissolve gallstones.
posted by telstar at 1:03 PM on March 22, 2010

Yeah, I'm just surprised he didn't include impotence on the "show me" list... seems like a glaring omission considering this is the internet, after all.
posted by nathancaswell at 1:07 PM on March 22, 2010

Yeah, I'm just surprised he didn't include impotence on the "show me" list.

If you're looking at yohimbe, be aware that it does have a high prevalence of an unwanted side-effect: euphoria.
posted by StickyCarpet at 1:09 PM on March 22, 2010

I assumed that "sex" on the "show me" list meant impotence, or at least included it.
posted by gingerbeer at 1:13 PM on March 22, 2010

When you buy any supplement, ymmv. I mean that in all seriousness. It may be the exact same product sitting next to each other on the shelf, possibly made by the same company, and the results received could be drastically different. If you need to get an extract, make sure it's an extract. Make sure there is the proper standardization if there is supposed to be one. Before you buy, you should check on which companies have a good reputation. You should also even check on where they harvested the supplement. And for the sake of skeptics everywhere, don't buy homeopathic.

The one in the top left?

Yeah. If you look at the top right, just above the chart, it has the bright orange/brown bubble and lists it as "One to watch (low evidence promising results)"
posted by P.o.B. at 1:13 PM on March 22, 2010

St. Halford's Wort

Surely you meant St. halford's Wart?
posted by ZenMasterThis at 1:13 PM on March 22, 2010

And, on closer look, it actually means "sexually-transmitted disease." Not the same thing at all!
posted by gingerbeer at 1:14 PM on March 22, 2010

Also, my original post was intended to be read aloud Family Feud style. Also, there's nothing wrong with my boner.
posted by nathancaswell at 1:19 PM on March 22, 2010

I have no scientific evidence. I never believed that Vitamin C prevented colds. It sounds ludicrous.

Then I had kids. Kids are disease machines. In 2004, I can't think of a single week where I didn't have a new cold for six months.

Then my wife put us all on Vitamin C supplements, every day. And it has to be every day, not just when you have a cold. And since then I've had maybe three colds. My kids get colds, they're gone in a day.

Do I have proof that vitamin C helped? No, but you'll have to pry it from my cold, dead hands.
posted by fungible at 1:28 PM on March 22, 2010 [1 favorite]

Why do they use the term "generative"? Does the graph build pieces of itself?
posted by Blazecock Pileon at 1:31 PM on March 22, 2010

Surely you meant St. halford's Wart?

I think you mean St. Lemmy's Wart.
posted by electroboy at 1:31 PM on March 22, 2010 [3 favorites]

I've taken so many of these snake oil health food herbs and vitamins and shit I must be the most gullible person on Earth. But only ONE definitely helped me: glucosamine/chondroitin. Years of gradually increasing joint pain as I passed the half-century mark stopped entirely once I started taking the stuff. Now the rest of it, I dunno...I take multi-vitamins and fish oil. Couldn't hurt. Anyway, nice chart.
posted by kozad at 1:38 PM on March 22, 2010 [1 favorite]

This piece was doggedly researched by myself, and researchers Pearl Doughty-White and Alexia Wdowski. We looked at the abstracts of over 1500 studies on PubMed (run by US National Library Of Medicine) and Cochrane.org (which hosts meta-studies of scientific research). It took us several months to seek out the evidence – or lack of.

Wait, whaaa? The basis is reading 1500 abstracts? I appreciate trying to limit to large, human, radomized placebo-controlled trials, but that doesn't mean the entire trial is valid or that the data supports the abstract conclusion. Where is the quality control in reading the study itself to determine if the meager abstract statement is warranted?

This kind of summary can propagate the kind of misleading stuff that makes people think vaccinations cause autism. I'm sorry folks, as much as the public wants an easy one-liner, medicine and drug efficacy just isn't that simple. If it were, we wouldn't need an FDA.
posted by Muddler at 1:43 PM on March 22, 2010 [4 favorites]

So many things wrong with this chart. I keep seeing it and people like to spread it around because it looks nice, but it's not right.
posted by SouthCNorthNY at 1:50 PM on March 22, 2010 [2 favorites]

I've taken so many of these snake oil health food herbs and vitamins and shit I must be the most gullible person on Earth. But only ONE definitely helped me: glucosamine/chondroitin. Years of gradually increasing joint pain as I passed the half-century mark stopped entirely once I started taking the stuff.
posted by kozad at 4:38 PM on March 22 [+] [!]

Kozad, that stuff allowed my parents' dog to walk and lie down without crying in pain the last few years of his life--and I doubt dogs benefit from the placebo effect. So it's something I'm comfortable endorsing for both canines and people.
posted by availablelight at 1:59 PM on March 22, 2010 [3 favorites]

Fish and flax oils keep my skin from drying out, and the omega 3 may help with ADD, valarian definitely works for sleep, as does melatonin, 5-htp works for serotonin, peppermint, fennel and ginger help with digestion, and I also take B complex with C and a mineral supplement - these last two may or may not help. But these are just from my own experiences.

The chart is confusing.
posted by krinklyfig at 2:02 PM on March 22, 2010

Oh, and if you have cold sores that reappear when you get stressed you can use lysine on a regular basis to keep that under control.
posted by krinklyfig at 2:07 PM on March 22, 2010

telstar: I recently learned that vitamin C helps dissolve gallstones.

Please, for the love of God, check your sources. Their 'research' includes a site recommending coke as a way to dissolve gallstones (hint: if they say it will come out in your poop they're a fucking moron - the biliary ducts are tiny enough that stones can be lodged and cause serious issues while still being too small to be seen on x-ray/ultrasound).
posted by geek anachronism at 2:15 PM on March 22, 2010 [2 favorites]

Vitamin C just as a daily thing, not necessarily a supplement, and then glucosamine chondroitin are the two I can personally endorse. The other stuff, I can take or leave.
posted by misha at 2:21 PM on March 22, 2010

I wish he'd separated out osteoarthritis from other varieties, but that's a nitpick. (And on the glucosamine & chondroitin front, my mother swears by the stuff but it does nothing for me but makes me violently sick. YMMV.)
posted by immlass at 2:29 PM on March 22, 2010

How interesting SouthCNorthNY. Perhaps you'd care to offer something substantive?

The visualization isn't without its flaws, but I think the spreadsheet alone would be worth a FPP.
posted by Orange Pamplemousse at 2:55 PM on March 22, 2010 [1 favorite]

So Fish Oil works but Omega 3 doesn't and Omega 6 does. That's very confusing.
posted by doctor_negative at 3:08 PM on March 22, 2010 [1 favorite]

Mere anecdata, of course, but more than once, I've developed cold symptoms... then slammed down a fistful of high-dosage Vitamin C tablets... then done it again, shortly thereafter... only to find my symptoms suddenly disappear.

Is there at least a degree of placebo effect involved? Sure. By this point, having done this successfully a number of times, I bloody well expect good results, and fast.
posted by darth_tedious at 3:43 PM on March 22, 2010

Mere anecdata, of course, but more than once, I've developed cold symptoms... then slammed down a fistful of high-dosage Vitamin C tablets... then done it again, shortly thereafter... only to find my symptoms suddenly disappear.

And I've done the same and still gotten a full-fledged cold.

That's why anecdotes do not necessarily equal data.
posted by muddgirl at 3:52 PM on March 22, 2010 [7 favorites]

Please, for the love of God, check your sources.

Not much incentive for me there.

But here's another source on the vitamin C-gallstone thingie.
posted by telstar at 4:02 PM on March 22, 2010

Vitamin C pills? Really? What's wrong with an orange, ffs?
posted by Sys Rq at 4:07 PM on March 22, 2010

There's nothing wrong with an orange, but if you're convinced that you want a 1000mg of Vitamin C, the pill looks a little easier than a dozen oranges. Personally, I think that's nature's little way of telling you that you are very, very unlikely to need 1000mg of C, but whatever.
posted by Wolfdog at 4:15 PM on March 22, 2010 [4 favorites]

Melatonin and valarian didn't do crap for me, though I know others have had luck with it. Also, some people's allergies can't tolerate citric acid in any form.
posted by Wuggie Norple at 4:18 PM on March 22, 2010

> And I've done the same and still gotten a full-fledged cold.

Sure. Of course, we can probably see the same effect with a great many treatments-- what works for you on Day X may not work for me on Day X, and what works for you on Day X may not work on Day Y, or may work to only a moderate degree on Day Z.

The things we classify as "medicine" are believed to have massive, reliable effects-- but in practice, not all medicines work on all people at all times. "Zylonex(tm) stopped working for you? Then maybe it's time you moved up to Sorbutal(tm)!"

More to the point, there are degrees of health-- I imagine many interventions may help a little... but not enough to affect the quality of some people's lives. If Person B is in worse shape-- or stricken with a stronger ailment-- at a given moment than Person A, the minor intervention that might perceptibly help A might not perceptibly help B.

When it comes to the illness of a particular person in a particular physical condition with a particular diet on a particular day with a particular treatment in a particular quantity-- and given the rich complexity of the human body-- there is no shortage of variables to consider...
posted by darth_tedious at 4:22 PM on March 22, 2010 [1 favorite]

> Vitamin C pills? Really? What's wrong with an orange, ffs?

Eating six oranges? Delicious, but... that's a lot of oranges.

Swallowing twelve high-dosage pills, in a matter of seconds? Easy.
posted by darth_tedious at 4:24 PM on March 22, 2010

When it comes to the illness of a particular person in a particular physical condition with a particular diet on a particular day with a particular treatment in a particular quantity-- and given the rich complexity of the human body-- there is no shortage of variables to consider...

If only humans had developed some method by which to evaluate the relevance of given variables to human health! If only!
posted by Justinian at 4:48 PM on March 22, 2010 [7 favorites]

"If only humans had developed some method by which to evaluate the relevance of given variables to human health! If only!"

They have, and it's deeply flawed. Not the principal of the evaluation, but the execution.
posted by digitalprimate at 5:06 PM on March 22, 2010

Saw this a couple of weeks back and immediately showed it to my wife, who, like me, is somewhat susceptible to being influenced by the latest bit of research (which frequently contradicts the previous bit of research) about various herbs and supplements.

If this does well what it claims to do and visualizes aggregrate research results on various supplements in terms of their efficacy for various ailments, well, it's just splendid.
posted by stavrosthewonderchicken at 5:15 PM on March 22, 2010

             _.-""""-._
           .'          `.
          /              \
         |                |
         |    Placebo     |
         |                |
          \              /
           `._        _.'
              `-....-'

posted by benzenedream at 5:47 PM on March 22, 2010 [6 favorites]

IIRC, that Frontline documentary pretty much skewers the purported efficacy of St. John's Wort for treatment of depression, and pays special attention to the problem of how test results are used to promote medicines even when they don't help the condition. Alternative medicines are not the only target for this criticism. See also: OTC cough suppressants.

Come again, mcstayinskool? I had bronchitis for well over a month last Fall, and Mucinex would cut my coughing down by 80-90%, easy. (Not endorsing that particular brand, as the ingredients - guaifenesin and pseudoephedrine - are in many other OTC cough suppressants.)
posted by IAmBroom at 5:56 PM on March 22, 2010

I have a feeling the recent "Cough suppressants don't really work!" hoopla had less to do with science than with the increasing awareness amongst youngsters of the fun effects of dextromethorphan.

You know, kind of like how marijuana and LSD were "proven" to have no medical potential.
posted by Sys Rq at 6:12 PM on March 22, 2010 [1 favorite]

Last I heard, the "cough suppressants don't work" research was on children's cough syrup only. But I could be out of touch.
posted by muddgirl at 6:26 PM on March 22, 2010

They have, and it's deeply flawed. Not the principal of the evaluation, but the execution.

That's why I prefer to rely on the opinion of people who sharpen their knives with a pyramid and refuse to pollute their kids' precious bodily fluids with vaccines.
posted by Justinian at 6:45 PM on March 22, 2010 [1 favorite]

Aggregate research results don't count for diddley if the underlying studies are crappy.

See this recent FPP for the terribly inadequate statistical underpinnings of many health and medical studies.

And this eye-opening quote from Stanley Young at the National Institute of Statistical Sciences.

In a survey of the recent literature, [Young] found that 95 percent of the results of observational studies on human health had failed replication when tested using a rigorous, double blind trial.

What we *really* need is this same webpage with the balloons vetted by statisticians.
posted by storybored at 6:48 PM on March 22, 2010 [1 favorite]

"That's why I prefer to rely on the opinion of people who sharpen their knives with a pyramid and refuse to pollute their kids' precious bodily fluids with vaccines."

You ever seen the sausage made, because I have. FDA statistical reviewers give it up like the boy or girl in high school with a bad rep in the back seat of a Volkswagen for an "honorarium" of $500.

...this while the drug/biologic is *still* in phase III.

You think for a minute those indications and labels are not bought and paid for you are ignorant of the process and sadly mistaken.
posted by digitalprimate at 7:05 PM on March 22, 2010

@Justinian

Or, let me break it down for you. There are around two dozen major individuals who are "trial designers" who will take your promising Phase II into Phase III and basically guarantee you'll get the label and indications you want. How can they do this? Because they've worked for the companies and they've worked for the CBER/CDER, and they know what works with the minimum amount of money spending. They will get your drug or biologic through all aspects of the process for around USD 120K a pop.

You honestly think the problems with, for example, the Cox-2s weren't apparent in the data? Well, they were, and they paid a lot of people to make sure that this Wasn't A Problem.

And that's why I take it all with a very, very large grain of salt.
posted by digitalprimate at 7:17 PM on March 22, 2010 [3 favorites]

Yohimbe or Horny Goat Weed - I believe these raise blood pressure. (thus your boner effects. Coffee used to be such and I knew a chap who liked his pseudoepherdrine for the same)

and the omega 3 may help with ADD, - look to the hormone known as Vit D-3.
http://www.grc.com/health/vitamin-d.htm has a audio CD you can listen to, talks about the UV light knocking loose hormones in the skin. (thus the health care tax on tanning beds may be an anti-health kinda thing)

glucosamine & chondroitin front, my mother swears by the stuff but it does nothing for me but makes me violently sick - Try the vegan version as you may have a shell fish allergy. But yes - it otherwise does seem to work for joints.
posted by rough ashlar at 8:13 PM on March 22, 2010

Dammit, where's red wine? I believe any study that says drinking alcohol is good for you and dismiss any that say it isn't.

Also, the evidence basis seems pretty loosey goosey for this thing. For example, he seems to ignore all the negative studies on St. John's wort and only gives two positive trials.
posted by Mental Wimp at 8:52 PM on March 22, 2010

I have to question this chart/methodology because of how high it places St. Johns Wort. A number of recent studies have concluded that St. John's wort is ineffective in treating depression.

The official NIH position is
Recently, controversy has been raised by two high-quality trials of St. John's wort for major depression that did not show any benefits. However, due to problems with the designs of these studies, they cannot be considered definitive. Overall, the scientific evidence supports the effectiveness of St. John's wort in mild-to-moderate major depression. The evidence in severe major depression remains unclear.

The dot in the chart for St. John's wort specifies, correctly, mild to moderate depression. Whether this is wise is obviously open to interpretation, but the chart does not deviate from conventional wisdom.

Personally, I do feel that it helped me, at least as a maintenance choice following a year on Prozac -- which at the time was not yet approved for maintenance dosage, now available as Lexapro. I may try out Lexapro now that I have state health insurance. Ultimately, I wish there were something OTC that someone could take for a week or two of mild blues, without all the stress and hassle of the shrink (if you can even get an appointment in that time frame). I'm personally a fan of the laxer environment for SSRIs, due to their greater safety, which has led to many people being prescribed by their GP.

I was glad to see green tea ranks high, at least for cholesterol, because I drink a lot of that. Cinnamon is also something I try to take a lot of as a diabetic.

I've wondered about Valerian. Since it's an ingredient in Feliway, an anti-anxiety treatment for cats that I can see produce fairly instant results, I've wondered whether anyone has studied humans in Feliway-infused air environments. There's a lot of stress around me and it doesn't particularly seem to help.

Probiotics are the only thing that seem to help my elderly, demented dad's gas (which can be severely noxious)

I also come from a family that was big on Vitamin C. I still take it but not daily. I have diabetic-related gum disease and it does seem to help how sensitive they are, so that falls into one of the benefits not listed. That is, it isn't technically an alternative therapy. I think it's important to keep in mind that some of these might be good at other things not listed.
posted by dhartung at 8:53 PM on March 22, 2010

http://www.grc.com/health/vitamin-d.htm has a audio CD you can listen to, talks about the UV light knocking loose hormones in the skin. (thus the health care tax on tanning beds may be an anti-health kinda thing)

What? You know what else UV light knocks loose? Your DNA. And then you get CANCER.
posted by delmoi at 8:57 PM on March 22, 2010 [2 favorites]

If I thought the spread sheet was good I would agree with you, but I don't think it's any where near complete enough to be useful to the public at large. I'm not going to say that this is a harmful chart and I think more people should take supplements, but I don't think its useful or worthwhile.
If there was a chart like this out there that I thought was really good, I would be the first person to front page it.
posted by SouthCNorthNY at 9:33 PM on March 22, 2010

I didn't see any mention of 5 HTP. Is that implicitly covered under 'tryptophan'?
posted by Bartonius at 10:17 PM on March 22, 2010

I'm surprised that zinc isn't ranked higher -- FWIW I swear by zinc to fight off colds and flu. At the first sign of any symptoms I start taking zinc immediately. At the very least my symptoms are reduced in severity and duration -- if I catch it early enough, symptoms are eliminated completely. The trick is to keep taking it for a week or so, otherwise the symptoms return. Again, it's important to start taking zinc as soon as you feel symptoms (eg. sore through, achy-painy, etc.) -- the later you start, the less effective it is.
posted by e-man at 10:55 PM on March 22, 2010

digitalprimate: You ever seen the sausage made, because I have. FDA statistical reviewers give it up like the boy or girl in high school with a bad rep in the back seat of a Volkswagen for an "honorarium" of $500.
...
They will get your drug or biologic through all aspects of the process for around USD 120K a pop.

Hi digitalprimate - can you elaborate on how this works? Since a phase III trial costs at minimum ~10 million, why wouldn't all companies do this since it's chump change compared to the cost of the trial itself?

This article claims "In non-small-cell lung cancer alone, between 1990 and 2005, a total of 1,631 new drugs were studied in phase II. Only seven of these new agents gained FDA approval."

If it's as easy as handing out payola, why do any new drugs fail in phase III?
posted by benzenedream at 11:57 PM on March 22, 2010 [2 favorites]

Ultimately, I wish there were something OTC that someone could take for a week or two of mild blues, without all the stress and hassle of the shrink (if you can even get an appointment in that time frame).

Since this is a thread about snake oil, you could always try something like passion flower tea.
posted by robtf3 at 12:15 AM on March 23, 2010

telstar: Not much incentive for me there.

But here's another source on the vitamin C-gallstone thingie.

Sorry to harp, but yet again we've got something claiming to flush gallstones out without any danger of blocking the very tiny biliary ducts with stones. Unless it's dissolving them, the sheer action whereby something too big to fit down a tube is somehow forced into the tube then out (all without agonising pain worse than childbirth) is highly suspicious. Yes low fat diets help. Yes prevention is good. But once you've got them removal isn't as simple as pooping, no matter how much vitamin C you guzzle. Or how much you pay a quack to send you his Super! Special! Tincture!
posted by geek anachronism at 12:26 AM on March 23, 2010 [1 favorite]

In 2004, I can't think of a single week where I didn't have a new cold for six months.

Then my wife put us all on Vitamin C supplements, every day.

I'm the control for your experiment. When my kids first entered school, I had cold after cold after cold too. But we did NOT go on C supplements. My colds have also gradually decreased to the point of 1 or maybe 2 per winter.

My theory is that when they entered the school system they got exposed to (and exposed us to) a lot of new cold viruses that we had to become immune to.
posted by DU at 6:00 AM on March 23, 2010 [2 favorites]

hmmmm.. don't see feverfew there. Absolutely takes care of my migraines if I put a sprig in my cheek at first warning. Oliver Sachs and others have mentioned it, I'll have to find some research on this.
posted by benjonson at 8:07 AM on March 23, 2010 [1 favorite]

You ever seen the sausage made, because I have. FDA statistical reviewers give it up like the boy or girl in high school with a bad rep in the back seat of a Volkswagen for an "honorarium" of $500.

I'm not sure how this is relevant, since we're talking about supplements here. Last time I checked, they were not subject to FDA approval, or any other regulation.
posted by lexicakes at 8:17 AM on March 23, 2010 [1 favorite]

FWIW I swear by zinc to fight off colds and flu.

You do realize that what would have happened if you hadn't taken zinc is unknowable, right? Because that's the relevant information.
posted by Mental Wimp at 10:45 AM on March 23, 2010 [1 favorite]

I think this chart is good in theory, but imperfect in execution. I was a little OMGWTFBBQ about magnesium getting such poor ratings, since I've had three (good) doctors recommend it to me for different conditions. I did a pub med search and without looking very hard, found a number of studies and reviews that indicate good results in treating/preventing migraine, hypertension, dyspepsia, pre-eclampsia, and PMDD, and as a supplement to antidepressant therapy. In contrast, only two magnesium papers are cited in the Google Doc.

I agree that part of the problem is that they are drawing conclusions from abstracts, and I wonder if the results are skewed towards papers with free full-text access. I can look up most of the full papers (yay VPN!) if anyone is curious about any other results.

Anyway, it's clear that the lower part of the graphic can either indicate "no proof of effectiveness for this use," OR "we just haven't gotten around to thoroughly researching this supplement." That doesn't make the infographic useless, but it's not very trustworthy.

Also, based on no background in graphic design, I wish the size of the bubbles indicated the number of studies cited, while color could indicate google popularity, and "promising" supplements could be a different shape.
posted by granted at 12:33 PM on March 23, 2010

More specifically, I think the image is skewed towards giving supplements low ratings, so anything above the Worth It line is probably more trustworthy than anything below. But again, that's hard to say without knowing how many/which studies they're citing for each supplement. You can of course find that out by going to the Google Doc, but that sort of defeats the whole purpose of the infographic.
posted by granted at 12:35 PM on March 23, 2010

I like the flavor of Cranberry Emergen-C. There, I said it.
posted by Pollomacho at 12:38 PM on March 23, 2010 [1 favorite]

@benzenedream

There are lots of reasons drugs fail in III, but the most common is that data from the relatively small scale IIs don't realistically show the side effect profile or efficacy (kind of self evident).

Why do companies then take them to III? Sometimes its because the II was so promising. Other times it's to scare other companies. Drug and biologics companies think in terms of pipelines, what kind of portfolio they will have in ten years. And sometimes making your competitors spend even more money than you on failed IIIs is the way to go.
posted by digitalprimate at 6:41 PM on March 23, 2010

"Wait, whaaa? The basis is reading 1500 abstracts? I appreciate trying to limit to large, human, radomized placebo-controlled trials, but that doesn't mean the entire trial is valid or that the data supports the abstract conclusion. Where is the quality control in reading the study itself to determine if the meager abstract statement is warranted?

This kind of summary can propagate the kind of misleading stuff that makes people think vaccinations cause autism. I'm sorry folks, as much as the public wants an easy one-liner, medicine and drug efficacy just isn't that simple. If it were, we wouldn't need an FDA."

Quite a few of these sourced to Cochrane metastudies, which might include 100 studies. So really, you might need to read not 1500 papers, but 15000 (although to be honest, probably nowhere near that many). Even 1500 is unfeasible, which is one of the reasons for the Cochrane papers. Part of the job of making a Cochrane metastudy involves reading all of the studies, and deciding whether the abstract is justified, and how strong the evidence is both for and against.

Anybody who might need to have any breadth of knowledge needs, at a certain point, to assign weight to studies based on authority and reputation of authors and reviewing bodies. I, and many others, consider the Cochrane studies among the most reputable.

NIH studies, which make up many of the remainder, are similarly reputable in my mind. They are, generally, well-funded, well-designed, and trustworthy. (There are exceptions.)

Random abstracts found on pubmed are another matter though, and there seems to be quite a few of those.
posted by nathan v at 12:52 PM on March 24, 2010

supplements here. Last time I checked, they were not subject to FDA approval, or any other regulation.

Yes they are. You can't say supplements "cure" anything.

Example:
Stomach ulcers used to be thought to be the result of stress. Turns out a bacteria does the job and just taking Bismuth does 'em in. Yet - pepto bismol can't advertise that.

The hormone commony known as Vit D may very well have many positive health effects. Yet, you can't pimp Vit-D3 as curing anything.
posted by rough ashlar at 10:16 AM on March 28, 2010