Stupid correlation tricks

We note with exasperation the latest dumbworm detritus from the CBC:

Here’s the startlingly precise claim from the researchers at the Harvard School of Public Health:

Eating one serving a day of unprocessed red meat such as beef, pork or lamb that was 86 grams in size, about the size of a deck of cards, was associated with 1.13 times increased risk of mortality, the researchers said.

For processed meats like one hot dog or two slices of bacon, the risk rose to 1.20 times for every extra serving.

Yes, it’s based on an epidemiological longitudinal study.  As such, only correlation is being measured, although the authors make some rather concrete claims:

“This study provides clear evidence that regular consumption of red meat, especially processed meat, contributes substantially to premature death,” Hu said in a release.

They also take a rather scattershot approach towards suggesting mechanisms for this “substantial contribution”:

Red meat, especially processed meat, contains ingredients that have been linked to increased risk of cancer and heart disease, the researchers noted. Those ingredients include heme iron, saturated fat, sodium, nitrites and carcinogens that are formed during cooking.

But of course this is only a correlational result, and buried in fourteenth paragraph is an important caveat:

In the study, men and women with higher intakes of red meat were less likely to be physically active and were more likely to smoke, drink alcohol and have a higher body mass. The researchers attempted to control for these variables, but it is difficult to separate out the effects of meat consumption from other lifestyle habits and a cause-and-effect relationship can’t be determined.

I rather suspect that what the researchers have actually shown is that making health-conscious lifestyle choices and believing that red meat is unhealthy (and therefore eating less of it) are correlated in the sample population — which, incidentally, consists of nurses and “health professionals”, folks whom you might expect to be steeped (and invested!) in the Conventional Wisdom that red meat is teh debil.  I wonder what kinds of outcomes they’d find if they included a sufficiently large sample of paleo practitioners in the data.


2 Responses to “Stupid correlation tricks”

  1. June 16, 2015 at 02:06

    If you want to take much from this post then you have to apply these techniques to your
    won website.

Leave a reply; use raw HTML for markup. Please blockquote quotations from the post or other comments.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

anarchocapitalist agitprop

Be advised

I say fuck a lot



Statistics FTW


%d bloggers like this: