Sunday, June 14, 2009

Bob Michell on meta-analysis

Quote of the century, from Bob Michell (taught me applied physiology many moons ago at the RVC and is a seriously bright guy).

"Meta-analysis of dross remains dross"

Ahhhhh that's good.


Edit: It's possibly as good as Malcolm Kendrick's definition of meta-analysis: one, two, skip a few, 99, one hundred.


Ken said...

Meta–analysis of all trials showed that use of vitamin D supplements decreased the risk for all-cause mortality.

Scientific enthusiasm for vitamin D supplements matches that for antioxidants a few years ago.

The questionable association of vitamin E supplementation and mortality--inconsistent results of different meta-analytic approaches.

The claims for vitamin D are destined to be quietly withdrawn.

Mad dogs and ....

Thinking Aloud said...

Meta Analysis of dross is more than just normal dross... it becomes Meta dross.

Anonymous said...

This is really off the subject but I was wondering if you had any info on the effects of oral contraceptive use and if a high-fat, low carb/optimal diet can help to protect against the adverse effects the these oh so horrible but necessary pills.

Brad Reid said...


Too good not to share as this relates blood sugar to self control.

I wondered immediately if they have a correct hypothesis that low blood sugar means lowered self control and all the negatives that implies, this, if they presume and/or project that low carb (low glucose) diets would then be bad by definition for self control.

My "take" would be that a low carb/sugar diet would create neither high nor low blood sugar readings, but sort of a stable one without the hard and sudden dips of a sugar eater.

Try holding back a donut from someone whose blood sugar has just crashed from a previous sugar eating episode, and one might see a crime unfold.

Cheers! Brad

Peter said...

Hi Brad,

Huge article and they're definitely trying. They are probably on to something in a sugar based metabolism. It adds to a post I've roughed out on anorexia and mood. Might get it posted one day...

Hi Shadapooda,

That's a hard one, I guess the answer is probably... I've always thought that the effects of progesterone derivatives were probably mediated via insulin resistance. The industrial doses of medroxyprogesterone acetate we vets used through the last quarter of the last century would occasionally precipitated type 2 diabetes in cats and obesity in all species we used them in. In cats where steroids are needed (we don't use MPA nowadays, sins of the past) and diabetes is a risk, I tend to LC the cats prophylactically. No idea how much good this does but it does help to keep the animal living within the drug induced level of insulin resistance without hyperglycaemia.

Are humans the same? That's a question.

Ah Nostril, Meta dross. Unfightable as the meta analysis includes many papers of dross which you could not go back and trash one by one. Worse than reviews.


Peter said...

Hi Ken,

The discussion is very interesting


Anonymous said...

I was just reading something on progesterone in BC pills and how it effects insulin sensitivity. I think it def increases one's risk of developing diabetes. Needless to say I hate BC and have opted out. I was getting all these pains in my chest and killing myself to stay in shape running. I'm simply not willing to sacrifice my health for this! I have decided instead to further educate myself and start charting my natural cycles using the Fertility Awareness Method (FAM). If there are any women out there looking for a better way to stay in control of their fertility without sacrificing their health I highly recommend this method. Check out for more info and free charts. I am confident that if I keep eating high fat low carb that I will soon recover to full health again... Can't wait!