Diet, Drugs and Diabetes - One Hundred Years of Missed Opportunities

Dr. John McDougall presents "Diet, Drugs and Diabetes - One Hundred Years of Missed Opportunities" at the Advanced Study Weekend, September, 2011, in Santa Rosa, California.

Category:

Education

Tags:

• diet
• diabetes
• john
• mcdougall
• heart
• starch
• obesity
• weight
• healthy
• vegetarian
• lifestyle

[JackFook]

@Mentat1231

I had fun debating with you. I wish you the best as well.

[Mentat1231]

@JackFook

Also, just in my own defense, if you look back I never said that self-selection bias wasn't relevant in general. I said it wasn't relevant when they are reporting what they eat, and their arteries are being directly examined. That is a very fair statement, since we use the same assumptions in many animal studies, when examining particular organs directly.

In any case, it's been very engaging and interesting discussing this with you. I wish you the best.

[Mentat1231]

@JackFook

I hope you had a good weekend :-)

Honestly, I think we'll just go round and round at this point. As I've said, there are serious problems with short-term studies of Atkins (since ketosis is a survival mechanism with some short-term survival benefits), and the only long-term study (Shai, et al) showed lots of problems and unimpressive weight loss and compliance. Also the natural pathways toward disease on a high-fat diet are well documented, and none of your studies refutes this.

[JackFook]

@Mentat1231

I think we have a serious difference in the quality of evidence that we consider valid to support your arguments. If you believe that self-selection bias is irrelevant, then by all means. However you must realized that your credibility will suffer, in my eyes, if not yours.

[Mentat1231]

@JackFook

It's the size of the study that you questioned. That's it. You are only now combining it with the (completely unrelated complaint) of the self-selected diet. These are two separate objections. My point is that, to avoid bias, if you believe that a small number of subjects invalidates a study, then you must apply that across the board, and reject a few of your 16.

On the subject of self-selected diets: as long as the data collected is accurate, what difference does it make?

[Mentat1231]

@JackFook

I never contested that they gave out food questionnaires and collected the answers... No offense, but maybe you should re-read my post. That was never a point I contested.

[JackFook]

@Mentat1231

I may change my mind, but I doubt it. The Angiology only administered the intervention to the low fat group.

"The 16 people (treatment group/TG) studied modified their dietary intake as instructed."

The other study of 13 subjects also gave instructions, but to all subjects, not just a fraction. 16 or 13, it's about the same quality. But comparing a group of 16 with intervention to a group of 10 without intervention is not proper study design.

[JackFook]

@Mentat1231

From the abstract of the Angiology study, I don't see where it says the researchers administered the diets to the subjects, nor were present when the subjects ate. If they did, they would've printed it proudly because that's a critical design parameter. So I'm left with the conclusion that they simply gave out food questionnaires and collected the answers.

[JackFook]

@JackFook

Observational studies are not inherently bad. They are inherently unreliable though. And when you combine all potential flaws I cited in the same study, it's really hard to produce reliable results. Maybe we can rely on its results anyway. Maybe it's all the other studies that are badly designed. But I doubt it.

[JackFook]

@Mentat1231

I cited _potential_ flaws. The Angiology contains all three I cited. I didn't say that's why it didn't replicate the results of the bulk of low carb studies. But it's a good indication.