A new study condemning eggs has the dietary world all atwitter. Unfortunately, I am in the midst of a “career change” so I don’t have the time to get the full paper and do a proper critique. I’ll explain what the career change entails a bit later, after the dust settles on the mechanics of departing a large organization where one has worked for over 22 years. In the meantime, my somewhat desultory blogging habits will likely be even more so, at least until I reach my new equilibrium, probably a few months from now.
This newest egg study is from the group at U of Toronto where the glycemic index was invented by Jenkins who is a co-author on this paper. Jenkins has collaborated in the past with Dr Neil Barnard, the PETA-affiliated vegan evangelist, who runs the Physicians Committee for Responsible Medicine, a thinly disguised vegan advocacy group. Dr Barnard has the distinction of having been sanctioned by the American Medical Association more than once for the use of guerrilla tactics to advance the animal rights, vegan agenda. I am beginning to view these UofT folks as cut from the same cloth as T. Colin Campbell whose “China Study” has served as a paean to the virtues of veganism and validation for everyone who has that inclination while actually relying on science of the flimsiest kind to build the case (see Denise Minger’s excellent critique at http://rawfoodsos.com/the-china-study/).
Okay, enough ad hominem character assassination, what about the merits of the study? Just from the abstract, here are some major problems. Firstly, it is an observational study. As you know, you can establish correlations from this kind of data but not causation. Secondly, it relies on food recall questionnaires. These are notoriously flawed and may return accurate information at a rate as low as 15%. The findings have huge overlapping confidence intervals and yet achieve statistical significance. How does that happen? (recall what Dr Mike Eades said: if you torture the data enough you can get it to tell you anything you want) And finally, the data showed an increase in plaque area of only 5.6%. In an observational study, you don’t have to overlook much of a confounder to get that kind of variation. Recall the hormone replacement therapy cock-up where the observational data showed a 50% reduction in CVD risk and when the clinical trial was done the result was actually a 30% increase, instead. That’s an 80% spread because of overlooked confounders. A 5.6% effect, given the inherent problems with this type of data – why is this even being published?
For a critique from somebody who has actually read the study, I recommend you have a look at what Zoe Harcombe has to say: http://www.zoeharcombe.com/blog/
There has been, predictably, much coverage of this study in the mainstream media, as well. This is the only article that I have seen that puts the issue into proper perspective: http://news.nationalpost.com/2012/08/15/yolks-as-bad-as-smokes-we-try-to-unscramble-the-truth-about-eggs/
What does it all mean? Well, as I write this I am finishing my usual breakfast of frittata (with eggs, cream, mushrooms, onion, cheese and, oh the humanity – bacon!). And I will continue to tell anyone who will listen that an egg is a perfect food since it contains all the ingredients needed to make a baby chicken.