# Egg Yolk and Carotid Plaque: No Diet Association

In part I of this article, the numerous methodological flaws acknowledged by the study authors were addressed.

Now I’d like to tackle a few of the other major flaws that conveniently went unmentioned.

#### #1. The total plaque area

If you take a close look at the means (and standard deviations) between the people who consumed 2 or fewer eggs vs 3 or more eggs per week, you’d notice:

• 2 or fewer eggs per week (n = 388): 125 ± 129.62 mm2
• 3 or more eggs per week (n = 603): 132.26 ± 142.48 mm2

When the difference in means between the two groups is only 7 mm2 (with a large variance in plaque readings for these subjects of over 125 mm2), this tells you there is no physiologically relevant difference in mean plaque formation between groups.

That right.

After pooling all 991 egg eaters together, high weekly egg consumption vs. low weekly egg consumption appears to have no bearing on plaque formation.

Odd that no mention of this was made in the dicussion…

Or not.

#### #2. “People Can Come Up With Statistics To Prove Anything… 14% of People Know That”

One of the primary arguments made in this paper was that there was an exponential increase in plaque formation in the >200 egg-yolk year condition.

Right…

By calculating “egg-yolk years”, you introduce a confounding factor into the analysis.

When your egg metric is tied to “years of consumption”, you create the scenario where any individual in the > 200 egg-yolk year category is heavily biased towards being older overall!

This is a massive bias that cannot be ignored.

Particularly since even these researchers acknowledge in their paper that plaque accumulation tends to increase linearly with age.

When you look at the age data based on egg-yolk quintiles, this is what you see:

• < 50 egg-yolk years: 55.7 ± 17.03 years of age
• 50-100 egg-yolk years: 57.97 ± 16.32 years of age
• 100-150 egg-yolk years: 56.83 ± 12.35 years of age
• 150-200 egg-yolk years: 64.55 ± 12 years of age
• >200 egg-yolk years: 69.77 ± 11.38 years of age

As we can see, individuals in the >200 egg-yolk year quintile are ~12-14 years older than individuals in the bottom three quintiles.

Now just based on these age discrepancies, some valid questions would be:

• What is the typical rate of atherosclerotic plaque formation for someone between the ages of 56 –> 70 years?
• What impact was there to being born during WWII vs. in the 1950’s on fetal health?
• Where more medications being used in the older individuals compared to the younger ones at time of assessment?

I could go on but you start to get the picture.

So in reality, all this study shows is the following:

In individuals predisposed to CVD, there is greater plaque accumulation the older you get.

I’m pretty sure no health professional would debate that finding.

But I’m equally sure that making this claim doesn’t do a heck of a lot to advance our body of scientific knowledge and it’s definitely not worth the fanfare this paper has generated.

#### #3. Committing a cardinal sin of research: thinking correlation implies causation.

The last major error I’ll point out is that a fairly basic error was made in trying to attribute causality.

As any first year university student taking a research methods class could tell you, correlation DOES NOT imply causation.

In other words, just because you noticed Factor A and Factor B happen to increase at the same time, it doesn’t mean that Factor A caused Factor B.

For the current study, even had they properly conducted the nutrition research component with a more substantive and objective questionnaire that controlled for more variables, at best, these researchers would have been able to conclude that higher egg consumption is associated with increased plaque formation among the elderly.

But thinking that a retrospective analysis based on a simple questionnaire could ever “prove” egg consumption causes plaque formation?

Not on your life.

To help illustrate the potential mistakes we can make by overdramatizing correlations, I’d like to point you to a hilarious graphic posted on Business Week a short while ago:

Image originally published: http://www.businessweek.com/magazine/correlation-or-causation-12012011-gfx.html#

As you can see, if you are creative enough, you can “prove” anything with a good correlation.

So that’s a wrap for me. Although these researchers made some pretty bold and suggestive claims, a more thorough investigation of their work reveals too many methodological flaws in their data collection to take their results seriously.

I guess you can say they were left with egg on their faces  😉

Till next time, train hard and eat clean!

* Special thanks to Carter Schoffer of BodyTransformation.com for assisting me with collecting a number of the research papers that went in to writing up this critique. *