Are Eggs Heart Healthy? A Response to The Nutrition Source for the Harvard School of Public Health


A patient recently shared with me this article from the Harvard School of Public Health, which I found very unsettling.

The article ostensibly seeks to resolve the debate about whether or not eggs are heart healthy. Nutritionists and physicians do not seem to agree on this topic, but the article from the HSPH itself references emerging evidence that dietary cholesterol (cholesterol that is found in the foods we eat) does not have a strong association to overall blood serum cholesterol (the amount of cholesterol present in our bloodstream), specifically citing this article from the Current Opinion in Clinical Nutrition and Metabolic Care. This is because the liver, which produces blood serum cholesterol, operates on a negative feedback loop, such that it will produce less cholesterol when it senses that we take in more. In addition, total cholesterol does not seem to be as indicative of Coronary Heart Disease (CHD) as is the ratio of HDL to LDL (and science suggests that even this ratio is an oversimplification).

The rest of the article provides conflicting conclusions. It cites evidence showing that dietary cholesterol is not harmful as it once was said to be; however, the article concludes with a recommendation to reduce our intake of eggs to no greater than one per day, as greater intake of eggs is associated with greater mortality rate by way of heart disease. This recommendation was not based on a biologically plausible explanation, but on the author’s misinterpretation of an already flawed article that was published in JAMA in 1999. Furthermore, the author does not explain or justify how he or she determined that qconsumption of specifically three eggs daily is more harmful than one (no clinical trials are referenced in this study which even evaluated consumption of greater than one egg per day).

Nutritional studies in general are profoundly vulnerable to confounding, meaning that it is hard to pinpoint exactly what food or habit correlates to what result. This study is no exception. Based on this, I looked into into the study design of the original 1999 paper that this article references. A number of items were troubling to me, and indicated that the author of the article from the HSPH has no understanding of confounding, nor does he or she have a grasp of how to read even the abstract of a scientific paper. Or perhaps the author did read the study, but chose to conveniently omit it from his or her article.

The abstract of the paper says:

  • “We documented 866 incident cases of CHD and 258 incident cases of stroke in men during 8 years of follow-up and 939 incident cases of CHD and 563 incident cases of stroke in women during 14 years of follow-up. After adjustment for age, smoking, and other potential CHD risk factors, we found no evidence of an overall significant association between egg consumption and risk of CHD or stroke in either men or women. The relative risks (RRs) of CHD across categories of intake were less than 1 per week (1.0), 1 per week (1.06), 2 to 4 per week (1.12), 5 to 6 per week (0.90), and > or =1 per day (1.08) (P for trend = .75) for men; and less than 1 per week (1.0), 1 per week (0.82), 2 to 4 per week (0.99), 5 to 6 per week (0.95), and > or =1 per day (0.82) (P for trend = .95) for women. In subgroup analyses, higher egg consumption appeared to be associated with increased risk of CHD only among diabetic subjects (RR of CHD comparing more than 1 egg per day with less than 1 egg per week among diabetic men, 2.02 [95% confidence interval, 1.05-3.87; P for trend = .04], and among diabetic women, 1.49 [0.88-2.52; P for trend = .008]).”

Essentially, there was no association between higher egg consumption CHD unless the participant also had diabetes; it is problematic to declare that there is an association with CHD and the entire population, when study only found this association to be true among diabetics. I would imagine a high level of confounding among diabetics, who due to their conditions and dietary restrictions are not representative of the greater population, and also receive various diet recommendations from their doctors.
From the “methods” sections of the paper it was explained that:
“We also computed intake of eggs included in other foods such as cakes, cookies, pancakes, muffins, sweet rolls, and donuts.”
This means that those who eat a lot of baked goods were also more prone to be placed in the “high egg consumer” group based on their dietary questionnaires. Therefore, one could rewrite this article linking the consumption of baked goods to CHD and find a strong association between the two.

The second of the three studies, published in 2013 in the American Journal of Clinical Nutrition, also examined egg consumption and incidence of CHD. The researchers of this study also concluded, based on their meta-analysis, “that egg consumption is not associated with the risk of CVD and cardiac mortality in the general population”.

The third study that the author cited, published in 2008 in Circulation, was the only of the three studies that does seem to match what the author’s of the HSPH article had said.However, they mentioned in the article, limitations to their study:

  • “Frequent consumption of eggs was associated with older age; higher body mass index; higher proportion of current drinkers and smokers and of physical inactivity; higher prevalence of diabetes mellitus and hypertension; and lower prevalence of hypercholesterolemia… additional lifestyle/dietary factors associated with frequent egg consumption could be responsible for the observed relationship. Given the observational nature of our design, we cannot exclude chance or residual confounding by measured and unmeasured factors as a possible explanation of our findings.”

The typical person who eats just one egg for breakfast might be more likely to be restricting food or caloric intake; therefore, this person might be healthier in terms of diet and exercise than the average participant. As the general population doesn’t seem to follow a “Transform-esque” diet, it would follow that those who regularly have a large amount of eggs for breakfast might also have a greater propensity to eat larger portions in general; these likely would be the same people who regularly consume large standard American breakfast, complete with bacon, home fries and white bread toast. In addition, it is possible that the extra eggs are actually those found in baked goods, as with the first study. Therefore, we would say that this association is correlative, rather than causative. To consider something to be causative, we would need to isolate the specific variables we are examining, and also determine an underlying biological explanation; this is to say, if dietary cholesterol does not strongly correlate to heart disease, what could be present in an egg that could increase the likelihood of one developing CHD? The authors of the article from Circulation conducted their research operating under the assumption that there is, in fact, a link between dietary cholesterol and CHD. Though the author of the HSPH article admitted that this is not the case, he or she did not inquire into why that could be.

The following statement, which is hidden away in the HSPH article as a brief side note, provides a much better picture of the underlying problem:

  • “You also need to pay attention to the “trimmings” that come with your eggs. To your cardiovascular system, scrambled eggs, salsa, and a 100% whole-wheat English muffin is a far different meal than scrambled eggs with cheese, sausages, home fries, and white toast.

We cannot base our recommendations on just one nutritional study, especially one that conflicts with biology and other studies. Nor can we definitively blame one food for causing a certain condition without eliminating all potential confounders.
As a public health major, this is one of the most disappointing and misleading articles I have ever seen from the Harvard School of Public Health. However, I do see it as a lesson that to this day we cannot trust nutritional information simply because the name of a prestigious entity such as JAMA or the Harvard School of Public Health is attached to it. I am reminded of other flagrantly erroneous studies, such as the one referenced in this article. In general, when reading scientific studies or articles, we should always keep in mind these questions:

  • Who funded/conducted the research?
    Were there any ties to specific lobbyists or industries?
    Were there confounding factors?
    Were the findings causative or correlative?