Should You Eat Eggs or Not?

May 16 2019

This recent, controversial study was published in JAMA and was all over the news and social media for weeks. Some folks outright dismissed the study, some were more confused, while a few took a moderate stance. So which approach is right? Here I am going to use an evidence-based approach to answer this very question. So here we go:

What was the study about?

Association: The study was looking at the relation or association of eggs on cardiovascular disease and death in US adults. For this the study combined data from 6 studies from 1985 to 2016 that resulted in almost 30,0000 participants who were followed for a median duration of 17 years.

Should you eat eggs

So why just association and not causal? The reason is that in an observational study, the researcher just “observes” people, for example, in this case, people who eat eggs and those who don’t, and after many years see how many people end up with heart disease or die. As expected, the problem with this type of a study design is that the people who consume eggs might also have other unhealthy habits such as eating red meat, smoking, eating junk food, being sedentary and so forth. That means, people who eat eggs and those who don’t might be different in a number of ways. Consequently, we don’t really know what exactly caused the heart disease; is it the egg or some other unhealthy habit(s) they have. To overcome this problem, researchers commonly adjust for these differences using statistical techniques, but nobody is sure how well we have adjusted for these differences. Hence the reason why it is said that observation studies can only show association and not cause and effect.

Adverse effects/harms: On a passing note, observational studies cannot show cause and effect, but it’s important to note that adverse effects or harms can only be really studied using an observational type of study design (ethical, financial, & logistical reasons), So observational studies assessing harms shouldn’t be just casually dismissed unless you have a better way to study harms/adverse effects. Do you?

So what were the results of the study?

Relative Risk: Each additional half egg eaten per day (3-4 eggs/week) was associated with 6 percent higher risk of cardiovascular disease and 8 percent higher risk of any cause of death. That’s great, but what does 6 & 8 percent mean? Do 8 out of 100 people die or you are 8 % dead? A better way communicate risk to play people is through numbers need to harm or NNH.
Numbers Needed to Harm (NNH): Using absolute risk, if 90 people have 3-4 eggs per week, one additional person will develop heart disease in 17 years (and his could be as high 312 people!upper limit of interval). For the outcome of mortality, this means 50 people have to eat egg to have one death in 17 years (and this could be as high as 90!). Now how does that sound? As you guessed it, the risk appears much less threatening. However, from a population perspective we are talking about thousands of people getting heart disease which is a serious public health concern. From what I have seen, none of the major articles have managed to communicated the risk properly. Understanding the risk well is is an important aspect of an evidence based approach.
Any relation with Cholesterol: It should be noted that the study observed a similar increased risk with an increase in cholesterol intake. Apparently, the association between eggs and heart disease and mortality disappeared when they adjusted for cholesterol, showing that it could be the cholesterol in eggs that may be the problem.

What about the quality of the evidence?

Or, in other words, how certain are we about this relationship. One of the critical aspects in evidence-based approach is to grade the quality of evidence. There are number of grading schemes out there; one commonly used approach is to grade the evidence into: Very Low, Low, Medium, & High. And the higher the grade assigned, the greater the trust we can place on the evidence.

Observational: Of course, being observational is one of the major limitations of the study. So that means, we don’t really know for sure if the increase in heart disease is really due to eggs. In an evidence-based method, observational studies are graded always as low-quality evidence.

That beeing said, the study do have some major strengths:

Dose-response: One of the major strengths of this study is the dose-response effect. The results showed that the more eggs you eat, the higher the risk of heart disease and death. So if you had 2 eggs per day, your risk went up by 27% for heart disease and 34% for mortality. Mind you, a large reason why cigarette smoking and lung cancer is considered to be causal is because of the large effect on cancer and also the dose-response effect observed. Here, we have a dose-response effect, but the magnitude of effect on heart disease and mortality is very low.

Individual participant data: Also, this study used individual participant data (IPD) rather than average of each studies. So it is akin to one large study of six studies with data available from each person. In turn, this makes the study and statistical adjustments that we previously mentioned highly reliable than previous studies on this topic. Almost all articles have missed this important point. Thus, this study is the “best available evidence” we have on this topic.

Quality: Taking into account the dose-response effect and the observational nature, we can say that the study is of low to moderate quality. Or we only have a low to moderate confidence in the evidence, at best.

What about the benefits of eating eggs?

Benefits: Eggs as many are aware are nutritious, high-quality source of protein and B vitamins, and relatively inexpensive. That been said, there is little evidence to show that eggs have any critical benefits in terms of longevity or disease prevention. On the other hand, the risk is a small increase in heart disease and mortality that we are not so confident of.

So what recommendation can we make?: The bottom line is that we can only offer a weak recommendation against eating eggs. Or any recommendations made against eating eggs are just ‘suggestions’ and ‘contingent’ on the individual circumstances and preferences. On the other hand, strong recommendations are given out when we have very high confidence in a large risk vs benefit ratio.

Eggs Evidence based approach

So what is the controversy regarding cholesterol?

Cholesterol Cut-off: The 2010-2015 Dietary Guidelines of US recommended a cut-off of 300 mg. However, in the 2015-2020 updated guidelines, they removed this cut-off and instead stated “as minimum as possible”. So although they took off the quantitative limit, cholesterol is still a nutrient of concern.

What made everyone a bit confused was that they wrote “No appreciable relationship between dietary cholesterol and serum cholesterol or clinical cardiovascular events in general populations” and at the same time paradoxically wrote “Individuals should eat as little dietary cholesterol as possible”. I do think this will be changed in the upcoming guidelines. Just keep in mind that one egg yolk has around 186 mg of cholesterol.

So should I stop eating eggs or what?

Taking into account, the small risk, low to moderate quality of evidence, we can only offer weak recommendations or suggestions against eating eggs. Some of my ‘suggestions’ for people who eat eggs:

If you have a high cholesterol levels, it may help cutting back on egg yolks. However, if your cholesterol is low, eat them.
If you are living in another country, like India or UK, do remember that the average cholesterol levels are much lower than in USA and these results may not apply.
If you eat 2-3 eggs per day and a lot of meat, it may be better to cut back on eggs or maybe have egg whites instead.

Clearly, the recommendations vary depending on the individual circumstances and preferences. Finally, as you can see, evidence-based approach is more than just reading the study results and jumping to recommendations. In contrast, EBA involves finding the best-available evidence, grading the quality of the evidence, understanding and communicating the risk/benefits clearly, and then conveying the strength of your recommendation. Thank you for reading.