Nutritional supplements, ranked

One of my favourite shows on TV right now is The Big Bang Theory. For those who haven't seen it: it's like Friends, except instead of New York yuppies, it's PhD physicists and engineers at CalTech. It's nice to see geeks and smart people be the focus (rather than the comic relief) of a sitcom. Also, the equations on the ubiquitous whiteboards on the sets are actually meaningful, instead of the usual scattering of random symbols. Here's a typical exchange from the show (Penny is, literally, the girl next door; Sheldon is a genius physicist with no sense of sarcasm):

(Penny walks to a nearby shelf of vitamins and supplements.)

Sheldon: Oh boy.

Penny: What now?

Sheldon: Well, there's some value to taking a multi-vitamin but the human body can only absorb so much. What you're buying here are the ingredients for very expensive urine.

Penny: (sarcastically) Well, maybe that's what I was going for.

Sheldon: (trying to be helpful) Well then you'll want some manganese.

I laughed at that exchange: I've been saying the same thing as Sheldon for years. And so I was interested to see this of the value of nutritional supplements from Information is Beautiful, even if the eye candy factor rather outweighs the elegance of the information design. (Who needs to use the X axis, anyway?) Fortunately, R blogger Tal Galili has taken the source data (made available as a Google Spreadsheet, so it can be read directly into R) and presented it as a simple bar chart, ranking the supplements by their efficacy score (0 means no evidence of efficacy; 6 means strong evidence of efficacy for the condition listed to the right):

Tal has helpfully made the R code to recreate this plot available in his blog post.

R-statistics blog: Nutritional supplements efficacy score

Comments

[alessandro]

psst...selenium is listed twice, for the same condition, with opposing results...

[Tal Galili]

Hi David,

Thanks for sharing this with your readers!

A small note: In my blog post I published (IMHO) a more suited plot to present the data (There might be a point in giving it instead of the barplot).

Cheers,
Tal

[Tal Galili]

alessandro - Please notice that when a supplement got a grade 3, it doesn't mean it is "not effective" only that the evidence for it's effectiveness is "so so". For better understanding, one will have to go into the google spreadsheet and read about the research that was done.

Secondly, in the new graph I offer (jump to the blog post to see it), I offer the highest (e.g: best) evidence available for that supplement/benefit - so that makes the plot much easier to interpret.

Cheers,
Tal

[Dan]

Caltech is preferred to the CamelCase CalTech.

I like the balloon plot a bit better; it's easier to look at the benefits for a given supplement.