Earlier this week, the homepage of my alma mater’s website (Western University) featured a story about a recently published paper by a Western cardiologist who is arguing that egg yolk consumption is nearly as detrimental to arterial health as smoking: Research finds egg yolk consumption almost as bad as smoking.
Now, for anyone who has followed nutritional science over the past 10-15 years, this headline sounds patently ridiculous.
In fact, the proposition that egg yolks are on par with smoking sounds so ludicrous that I figured Western news featured the story simply because it was a slow news period (being mid-August and the majority of the university is on holidays) and the the story would, otherwise, largely be ignored.!
How wrong I was.
Since Monday, a large number of Canadian news websites have picked up the story and I have been inundated with emails from my clients and colleagues asking whether I will be writing about this.
Originally, I hadn’t planned on doing so (due to the time involved with properly treating this issue); however, based on the coverage this story is getting, I think it’s only appropriate to present a critique.
The central thesis to this paper (Egg yolk consumption and carotid plaque) is that cholesterol from eating egg yolks is associated with higher risks for plaque formation (a risk factor for cardiovascular disease).
As alluded to earlier, this flies in the face of our current knowledge of cholesterol biochemistry1 and very much reeks of the out-dated lipid hypothesis that dominated nutritional science over ~30 years ago.
Cholesterol (and dietary fat in general) was the nutritional boogey man of the 1970’s, 80’s and early 90’s. However, by the turn of the century, the majority of cholesterol research showed that dietary cholesterol intake has little, to no, impact on the development of coronary vascular disease for healthy individuals1,2,3,4.
That dietary cholesterol intake hasn’t shown much relationship to cardiovascular disease shouldn’t come as too much of a surprise.
Daily cholesterol needs for humans range from 1000-2000 mg/day5 (the amount of cholesterol in 5-10 eggs/day) and the human liver is more than capable of ramping up or down-regulating its natural production of cholesterol in response to dietary cholesterol intake.
Therefore, if we eat more cholesterol in the diet, we simply produce less in the liver. And assuming we aren’t doing something bizarre like eating 2-3 dozen eggs each day, every day, there doesn’t appear to be a convincing mechanism linking dietary cholesterol to cardiovascular disease in healthy individuals.
A few caveats: familial hypercholesterolemia is totally different and there’s some emerging evidence that oxidized cholesterol may be more significant to this whole equation… but that’s a post post for another day.
Mind you, just because the current state of nutrition research concludes dietary cholesterol is benign doesn’t conclusively prove it is. As new discoveries are made, science must continue to evolve.
Unfortunately, what the popular media often forgets is that just because research gets published, it doesn’t automatically mean what the researchers found is correct. Scientists and clinicians are fallible (like all humans).
Not only do scientists and clinicians make mistakes, but researchers can be biased (sometimes intentionally) and incorrect conclusions are easily made. This can even happen with well-meaning individuals, if the results were improperly collected or analyzed.
Sadly, every time popular media covers a sensational new finding in nutrition research without actually assessing the validity of the research, all it does is contribute to the mass confusion most North Americans have surrounding the concept of “healthy eating”.
So after hearing about this study, like most nutrition professionals, my gut instinct was to say “this is bogus”, ignore it, and continue recommending my clients do the same things they were doing before this paper was published.
But as a scientist, simply dismissing a controversial finding as “biased” or incorrect is a faux pas. Therefore, I wanted to read the original research paper with as objective an eye as I can to see whether their results should reshape my own views about nutrition.
After a thorough reading of this paper, I was more than a little dismayed (but not surprised) to discover an astonishing number of significant methodological flaws which, in my opinion, render their conclusions worthless.
I’d go so far as to state that the methods employed to collect the nutritional data were so poor that frankly, I’m amazed the paper was even accepted for publication.
For those of you who don’t have access to the paper, here is a quick summary:
- Participants in the study were patients at a vascular prevention clinic (so typically older adults with one or more risk factors for CVD to begin with).
- Cartoid plaque was measured by ultrasound.
- A brief lifestyle questionnaire was given, which included questions on smoking and egg consumption but not exercise, alcohol consumption, total diet composition or medication use.
From the results of the questionnaire, pack-years of smoking was calculated as follows: number of packs of cigarettes per day times the number of years spent smoking; and egg-yolk years as: number of egg yolks per week times number of years consumed.
If you are thinking this is a bizarre way to quantify egg intake you aren’t alone… I’ll break this down in a minute.
Then, based on pack-years and egg-yolk years, the ~1200 participants were grouped as follows:
- Smoking years: 0 years, < 10 years, 10-20 years, 20-40 years, > 40 years
- Egg-yolk years: < 50 years, 50-110 years, 110-150 years, 150-200 years, > 200 years
And what the researchers determined was that both the “> 40 pack-years” and “> 200 egg-yolk years” groups were significantly correlated with increased carotid plaque area.
Sound the alarm and publish the data!
Only before hitting the panic button, they may have wanted to take a closer look at what their data actually allows them to conclude.
In the authors’ own words, the weaknesses of their study were:
So let me get this straight, this paper was published despite:
#1. Egg consumption not being objectively measured.
Question: How valid is taking last week’s or month’s diet and generalizing it over a lifetime of intake?
Answer: It’s not. No one has published any data on this validity of this technique because it is essentially just a made-up number.
It’s pretty well accepted that retrospective food history questionnaires (asking someone to remember what they ate in the day/week/month prior) is a notoriously poor way of collecting nutrition data6.
A better approach of trying to establish a link between egg intake and arterial plaque formation would have been to conduct a prospective study.
In that case, you would have analyzed someone’s diet today… then again 5 years from now… and again 10 years from now and mapped it onto plaque formation. Although this doesn’t tell you specifically what people eat day to day, at least you have a few somewhat objective measurements at those specific time points.
And lo and behold, there are actually a number of published prospective studies investigating egg consumption and cardiovascular disease risk.
What to guess what they’ve shown?
You got it: they all show the same result: eating eggs, even daily, presents no risk for heart disease in healthy individuals7,8!
There is one study that shows that diabetic males may be at increased risk from increasing egg consumption, but diabetes is quite a different metabolic state than healthy, normalcy.
Of course, even prospective studies have their weaknesses so an even better way to study nutrition is to actually control someone’s intake for a fixed period of times (i.e. 4-24 weeks) and measure changes to their biochemistry. Obviously this is crazy expensive and time consuming, but thankfully, a few research groups have already done it.
In one 6-week trial, daily egg consumption was shown to be non-detrimental to endothelial function in hyperlipidemic adults9. In another study published on young, healthy individuals (this time 12-weeks in length), daily egg consumption was shown to decrease the ratio of total cholesterol to HDL-c10, which is considered to be a beneficial change to blood chemistry!
At this point, a fair question for these researchers would be to ask why using a retrospective diet analysis involving a made-up measure like “egg-yolk years” is a better way to establish the risks of egg consumption than a randomized clinical trial or even a prospective study?
#2. No attempt was made to analyze total diet composition for other factors potentially contributing to plaque formation.
Brilliant. Let’s totally ignore the potential impact that simple sugars, saturated fats, alcohol or total fat/carbs/protein might have on blood chemistry.
Better yet, let’s acknowledge how little we understand about nutrition by blatantly ignoring how many total calories were even being consumed!
Go after the egg yolks I say: calories, trans fats, refined sugars: you are all safe.
And let’s not even try to pin cardiovascular disease on saturated fats in general. It’s the egg yolks that’ll kill you. Beef liver, shrimp and full fat dairy, you are also safe.
I seriously wish I was making this stuff up, but it just keeps getting worse.
#3. Totally ignored exercise.
I’m the first to admit, the effectiveness of exercise as a weight-loss tool has been overstated by the fitness industry, but most of the data on exercise and health is quite compelling.
In fact, exercise is seen as so beneficial for arterial health that the organization that partially funded this research, The Heart and Stroke Foundation of Ontario, has this to say about exercise on their website:
So, how exactly can totally neglecting to assess “minutes of physical activity” per day be considered “acceptable science”?
Is it really that hard to ask someone “how much do you exercise each day”?
If asking someone how many eggs they ate last week and projecting that out over their life time is considered valid, why not do the same with physical activity?
Why am I asking so many rhetorical questions?
#4. Used BMI, even though research suggests that even a simple waist-hip ratio is more clinically relevant to CVD11,12.
Why does anyone still use BMI? It is equally quick to assess someone’s waist and hip girth as it is to toss someone on a scale.
Again, what were the authors’ rationale for not taking this measurement, considering they acknowledge its importance?
What I also found strange about the decision to not report (or inquire) about exercise, waist girth, alcohol consumption or total calorie intake was that in 2007, the lead author of this current study published a paper showing that daily exercise, maintaining a fit weight, and moderate consumption of alcohol were all important lifestyle factors that could reduce the risk of cardiovascular events by up to 80%13.
If, 5 years ago, these factors were recognized as significant players in the development of atherosclerosis, what changed since then to encourage their omission from this current analysis?
You would think that the errors in analysis that the authors admitted to would be enough to sink this study, but sadly, there are quite a few more coming!
In part II of this article, we’ll take a look at some of the flaws that weren’t mentioned by the study authors.
- Fernandez ML and Calle M. Revisiting dietary cholesterol recommendations: does the evidence support a limit of 300 mg/d? Curr Atheroscler Rep. 2010 Nov;12(6):377-83.
- Baum SJ et al. Fatty acids in cardiovascular health and disease: a comprehensive update. J Clin Lipidol. 2012 May;6(3):216-34. Epub 2012 Apr 13.
- Fernandez ML. Rethinking Dietary Cholesterol. Curr Opin Clin Nutr Metab Care. 2012 Mar;15(2):117-21.
- Jones PJ. Dietary cholesterol and the risk of cardiovascular disease in patients: A review of the Harvard Egg Study and other data. Int J Clin Pract Suppl. 2009 Oct;(163):1-8, 28-36.
- Lecerf JM and de Lorgeril M. Dietary cholesterol: from physiology to cardiovascular risk. British Journal of Nutrition. 2011. July; 1: 6-14.
- Kristal AR, Peters U, and Potter JD. Is It Time to Abandon the Food Frequency Questionnaire? Cancer Epidemiol Biomarkers Prev 2005;14(12):2826 – 8
- Hu FB et al. A Prospective Study of Egg Consumption and Risk of Cardiovascular Disease in Men and Women. JAMA, April 21, 1999—Vol 281, No. 15
- Zazpe I et al. Egg consumption and risk of cardiovascular disease in the SUN Project. European Journal of Clinical Nutrition (2011) 65, 676–682
- Njike et al. Daily egg consumption in hyperlipidemic adults – Effects on endothelial function and cardiovascular risk. Nutrition Journal 2010, 9:28
- Mayurasakorn K, Srisura W, Sitphahul P, Hongto PO. High-density lipoprotein cholesterol changes after continuous egg consumption in healthy adults. J Med Assoc Thai. 2008 Mar;91(3):400-7.
- Canoy D. et al. Body fat distribution and risk of coronary heart disease in men and women in the European Prospective Investigation Into Cancer and Nutrition in Norfolk cohort: a population-based prospective study. Circulation. 2007 Dec 18;116(25):2933-43. Epub 2007 Dec 10.
- Ness-Abramof R and Apovian CM. Waist circumference measurement in clinical practice. Nutr Clin Pract. 2008 Aug-Sep;23(4):397-404.
- Spence JD. Intensive management of risk factors for accelerated atherosclerosis: the role of multiple interventions. Curr Neurol Neurosci Rep. 2007 Jan;7(1):42-8.