Sunday, September 21, 2014

Should idiots be allowed to write the methods section of any "scientific" paper?

Time to get back to the blog. We had a great summer and life has just kept on being more interesting than blogging. There are about 20 comments I need to read and approve which I'll do my best to get around to, but I thought it was time to hit the keyboard after the summer holidays.

I thought I would just post briefly on the struggles of trying to work out exactly what a given paper is describing in dietary research. I did set out a post a few weeks ago, being rather derogatory, about this paper on ALS. Here is the preamble:

I feel I should like the paper. Really. What with all this iced water being poured over people's heads in the name of ALS research etc. But it's hard.

OK. We're looking at a ketogenic diet for mice endowed with an engineered model of ALS which is quite similar to one of the familial forms of human ALS.

Being me, I go to the methods section first, to see what they fed the poor mice on. From the philosophical point of view I expect the methods section of a paper to allow me to duplicate a given research protocol. All I am told in this case is that the ketogenic diet is 60% fat, 20% carbs and 20% protein and that it was made by Research Diet, Inc. New Brunswick, NJ. That's it. Now, until RD are bought out by some other multinational company, they have a website and this tells me that they supply only one ketogenic diet, D12336, which is 11% protein 89% fat and zero carbohydrate, pretty much what you need to get a rodent in to mild ketosis. So this research group are using a custom diet, what goes in to it is anyone's guess.

My guess is medium chain triglycerides. I don't think you can get a mouse in to ketosis with protein at 20% of calories and carbohydrate at 20%. You'd have trouble getting a human in to ketosis with this, unless you used MCTs.

This is important because I'm interested in teasing out whether there is any point in the enormous effort and endless tedium of eating a low carbohydrate driven ketogenic diet with thyroid deficiency, lethargy, brain fog, glucose deficiency and auto immune disease predisposition as routine sequelae, not to mention the constipation and halitosis (is this me?), when merely popping down to my local Caribbean corner store for a bottle of coconut oil might do the job equally well.

What goes in to the diet matters. Coconut oil is not safflower oil, is not butter. What goes in to the methods matters. Research must be replicable.

End of preamble. I wasn't best impressed.

Before I go on to think about ALS and what help medium chain triglycerides may or may not provide in another post, I thought I would just like to revisit the lethal effects of a VLC keogenic diet paper on the outcome of induced ischaemia and reperfusion of the myocardium in some hapless rats.

It is fairly clear that using "vegetable shortening" as your primary ketogenic source of calories is likely to destroy your myocardium if you have an ischaemic episode. Your first heart attack might be your last. It took an email to Research Diets to get the information about the probable trans fat content of their diet and confirmed to me that the research group had written a methods section which put their paper, and probably the researchers, in to the garbage category.

The same appears to apply to the flip side. I had wondered what a non vegetable shortening based ketogenic diet might do under the same circumstances. Well I'd missed the study, which fully reinforces my pro-LC confirmation bias. Ketogenic diets are the bees-knees for surviving a period of cardiac ischaemia, Crisco excepted.

So, what does the miracle diet for surviving your next coronary look like? I don't know. You don't know. You can read the full text. You still won't know.

The diet is 60% protein by calories! And 10% carbohydrate. The remaining 30% is "oil". Now you know as much as in the paper. Can you replicate the study, based on the methods? No.

BTW: They didn't even check ketone levels! I think we have to assume MCTs again and assume some degree of ketosis.


The scrutineers also need to be up against the wall come the revolution.

For all three papers. Crap. Even though I like the results.

Hiya all!



fff said...

Welcome back Peter - missed your erudite acerbity :)


JasmineJohend said...

Yippee!! Welcome back, you're posts have been sorely missed :)))

Anonymous said...

Life is more interesting than blogging? You might be on to something there.

Welcome back.

As to idiots, it seems there is a worldwide effort to outlaw pretty much everything except idiocy. It's a plague
and is probably correlated with consumption of veggie oils.

Anonymous said...

Peter, if idiots weren't allowed to write scientific papers, Pubmed would be a very small database.

NKSL55 said...

I don't know -- doesn't seem like anything as straightforward as "idiocy" to me. Most scientific manuscripts read like genre prose to me. That is, ever watch a TV show where, without having seen it before, you pretty much know what is going to happen? The writers are just plugging into a genre construct or trope that has been previously constructed in the brains of many of their audience.

I don't think it is much different for many manuscripts. They tend towards a "by the numbers" recitation of previous work.

So, for whatever reason, including details of the diet of the rodents you are studying is not widely done. With even a minor amount of thought it is obvious that this is completely unacceptable. But that is part of the drive of story telling or singing. Structures in the brain reacting with pleasure to repetition.

It takes some effort and risk to break away from genre norms, I think.

Unknown said...

"thyroid deficiency, lethargy, brain fog, glucose deficiency and auto immune disease predisposition as routine sequelae, not to mention the constipation and halitosis"

Petro! No! Tell me it ain't so. Not you, too? How did you come to swallow these garbage ideas?

Or, to put it into a more positive light, would you write, or have you already written, a blog post arguing the above so that I can learn from it? :-)


Zooko (

NKSL55 said...

Sorry Peter, looks like sarcasm doesn't work on the internet. Maybe you could use some of those emoticon things to, well, "illuminate" any sentence containing that "dry wit" of yours?


Peter said...


"Tell me it ain't so"

It ain't so!


Scott Russell said...

Any reason you think the diet in the first study wasn't D12492? Its listed on the website and has the appropriate macronutrients. Ketosis on those macros is odd but not impossible.

Scott Russell said...

And as an aside, I still buy that low-carb diets would do potentially worse in a period of zero oxygen. Having anaerobic energy substrates strikes me as a good thing during the ischemia. But I also buy that increased mitochondrial biogenesis from a low-carb diet would aide in the recovery process.

Peter said...

IcedCoffee, I still can't see how you could get ketosis on D12492. If it is ketogenic then people keep very quiet about this and it appears to be very widely used.

Now, the Surwit diet might do it I guess...

Re ischaemia: It would be nice if someone would do the research with a smidge of honesty. That seems unlikely in the foreseeable future! The Kuwaities stacked the deck as much as the transfatters.


ItsTheWooo said...

Great and funny blog post as per usual

Unknown said...

Okay, I'm embarrassed. I failed to detect sarcasm. How do I internet?