« Links With Your Coffee - Tuesday | Main | Decade in Review - 2009 »

Links With Your Coffee - Thursday




You can go after raw milk but please leave raw milk cheeses alone.

Have to agree. Raw cheese = yum.

Just don't go reading Big Daddy's stuff. I want that cheese!

I love cheese, too.

So it's 32 minutes into the new year here in LA and I find myself reading the raw milk thing and getting annoyed again at the HuffPo. Interesting article, love the scienceblogs. The comments are pretty interesting too.

Milk is bad for you, raw or otherwise.

Really? How? I'm not trying to be snarky here, I'm genuinely interested.

Cow's milk has evolved to serve the digestive apparatus of calves, not humans. Though cow's milk contains large quantities of calcium & vitamin D, as is suggested by dairy advertising, these nutrients are virtually inaccessible through digestion because humans cannot properly digest cow's milk. What's more, ingesting large quanitities of cow's milk produces curd-like deposits in the intestines, which the body scrubs away by using calcium extracted from our bones, leading many researchers to the conclusion that milk causes osteoporosis, rather than preventing it. Non-dairy consuming nations, such as China, have virtually no osteoporosis. Furthermore, milk has been linked to leukemia; children and dairy farmers are the two largest contractors of leukemia because they are the two segments of society most closely associtated with cow's milk. On top of all that, the recent implementation of Bovine Growth Hormone has produced utters so large that the cows frequently step on their own teats, causing infections called mastitis. Pus from these infections makes its way inevitably into the milk. This is so commonplace that the dairy industry has succeeded in convincing the FDA to increase the acceptable level of bacteria in milk. Indeed, Monsanto's data show's a 79% increase in mastitis, resulting in a 19% increase in somatic (pus & bacteria) cell counts. The warning label on Monsanto's Posilac (their brand of BGH) states, "Cows injected with POSILAC are at an increased risk for clinical mastitis (visibly abnormal milk)." The label goes on to say "use of POSILAC may result in an increase in digestive disorders such as indigestion, bloat, and diarrhea...." The increase in mastitis has led to an increased use of antibiotics in dairy animals, which in turn has led to drug resistant strains of bacteria that you may have heard of, such as untreatable staph infections.

Some helpful links:

Milk and leukemia

Milk and osteoporosis

Milk and rBGH

Interesting theories BDM. Thanks.

Good god - the hubristic, bargain-basement intellects of the "science-based medicine" site strike again!

I will once again, ad nauseum, say that I have no particular interest in the usefullness of "natural" remedies, but when I read gems like this, I really want to gag:

Usually such studies involve some random noise in the results, especially when several outcomes are measured. But with such a large study, random fluctuations should average out, and that is exactly what happened.

"...and that is exactly what happened." what?!?!

and this...

The lessons from this study and the lack of effect for Ginkgo biloba should learned and generalized.

...ummm...and why is that?

and this...

Preliminary, small, or poorly designed studies are unreliable, and often result in false positives. Only large definitive trials are reliable.

how convenient!

and this...

There is also likely confirmation bias and the file-drawer effect at work – favoring the publishing of interesting and positive studies.


Jesus - why don't these people just post a fucking link to the article (as they did in the 4th sentence) and leave their own hair-brained interpretations out of it?!

Everything you quoted seems reasonable and even obvious to me. What, you didn't understand the post?

Why shouldn't the lessons from a well-conducted large (in statistics this is very valuable) study not be learned, you reckon?

crap, double negative in the last sentence, but you know what I mean.

True, everything I quoted seems reasonable, but only in a superficial sense, which gives me the impression that either the author has only a superficial grasp of study design and statistical analysis (and, seemingly, research in general), or just wants to put his own spin on the article and expects that no one will notice. If you really want me to go through the quotes point-by-point, I am happy to do so.

As for your more specific question of why it's unreasonable to generalize the lessons of this study-- my reading of the author's statement is that since this large study's results were counter to the results of smaller, non-randomized studies, we should expect that the positive (i.e., those indicating efficacy) results of other small studies on different natural remedies are also invalid. To me, this type of logic is quite dangerous, but again, that's just my reading of the author's (quite vague) statement.

But the small, non-randomized (that's a BIG flaw!) studies are not valid. It doesn't prove either positive or negative. They're just irrelevant, cause lack of rigor. Why should they be taken seriously with such a glaring flaw? It seems it's the favorite type of study of alt-med proponents in general.

Well, I agree that non-randomized studies are never ideal (I'm an epidemiologist), but outside the world of phase 2 & 3 clinical trials (which are extraordinarily expensive and time-consuming, not to mention often logistically or ethically impossible), non-randomized studies are the norm. Now, that in itself doesn't mean they're valid, of course, but making a blanket statement that such studies are "irrelevant" seems kind of silly. Suggesting that the natural med folks choose non-randomized studies because they want to sway the outcome is baseless at best.

And, as for the size of these studies - this factor is rarely governed by anything less concrete than the number of available subjects. Hardly anyone even pretending to do science will choose to do a small study when a larger pool of data is (practically) accessible. [as a side note, I will say that it's a common misconception that increasing the sample size automatically makes your results more valid - in fact, this only addresses randomness in data and does not at all address bias. It's the bias, in fact, that randomization, and lacking that, smart study design, is meant to control.]

NY is local?


Support this site

Google Ads

Powered by Movable Type Pro

Copyright © 2002-2017 Norman Jenson


Commenting Policy

note: non-authenticated comments are moderated, you can avoid the delay by registering.

Random Quotation

Individual Archives

Monthly Archives