A number of years ago, I came across an article about vitamin B12 in the journal Blood, which is a specialty journal for hematologists. The following notation appeared in small print at the bottom of the first page: “The publication costs of this article were defrayed in part by page charge payments. This article must therefore be marked ‘advertisement’ in accordance with 18 U.S.C. section 1734 solely to indicate this fact.”¹ The law to which this notation referred stated, “Whoever, being an editor or publisher, prints in a publication entered as second class mail, editorial or other reading matter for which he has been paid or promised a valuable consideration, without plainly marking the same ‘advertisement’ shall be fined under this title.”

The US Postal Service no longer has a category of second class mail, so the regulation listed in 18 U.S.C. section 1734 no longer applies. However, the practice of researchers or their sponsors paying journals to publish their papers (typically in open-access journals or in supplements to journals) has become widespread. Which raises the question, to what extent should these papers be considered promotional pieces (i.e., advertisements), as opposed to educational or scientific articles?

Supplements to medical journals are typically paid for by organizations or companies that have a vested interest in promoting a specific point of view. The pharmaceutical industry and their affiliates are the most frequent sponsors, but other vested interests are also involved in some cases. Because of this apparent conflict of interest, the information in supplements to journals is perceived by many to be less credible than the information in the parent journal. Nevertheless, articles in supplements to journals are frequently cited by others, possibly even more often than articles in the parent journals,² and can be influential in guiding clinical practice.

I recently came across a review article titled, “The Role of Eggs in Healthy Diets,” published in a supplement to the Journal of Family Practice.³ The article was written by ML Fernandez, a researcher with impressive academic credentials, who has published a number of studies regarding the biochemical and physiological effects of eating eggs. The review article was “supported by funding from the American Egg Board,” which I assume means that the American Egg Board paid Dr. Fernandez to write the article. The American Egg Board is a farmer-funded organization that is “dedicated to increasing demand for all US eggs and egg products.”⁴ Some of Fernandez’s research has been funded by the Egg Nutrition Center, which is the “science and nutrition education division of the American Egg Board.”⁵

As an academic exercise, I decided to undertake an analysis of this article, to determine whether there was evidence of bias. As a point of reference, I believe there is good evidence that eating eggs can have a number of health benefits, and that the evidence that egg consumption increases the risk for cardiovascular disease is weak. In addition to containing high-quality protein, eggs are a good source of lutein (which may help prevent age-related macular degeneration) and choline (which may enhance brain function).

In my review of the review article, most of the information presented appeared to be accurate, but I found two instances in which there appeared to be bias or misrepresentation of the data. One statement in the article was, “A recent epidemiological analysis conducted by Harvard investigators reported no correlation between egg intake and risk of diabetes.” However, the study cited to support that statement⁶ presented a different picture. The study actually had two parts. The first part was a pooled analysis of three large prospective cohort studies of US health professionals (the Nurses’ Health Study [NHS], NHS II, and the Health Professionals Follow-up Study). In the pooled analysis, after adjustment for potential confounding variables, for each one-egg-per-day increase there was a statistically significant 14% increase in the risk of developing type 2 diabetes. The review article did not mention that finding. The second part of the study was a meta-analysis of 16 prospective cohort studies (including a total of 589,559 participants). In the meta-analysis, for each one egg per day consumed, there was a 7% increase in the risk of developing type 2 diabetes, which was of borderline statistical significance. The association between egg consumption and diabetes differed significantly in different regions of the world. Among the studies conducted in the US, each 1 egg per day consumed was associated with a statistically significant 18% increase in risk of type 2 diabetes. None of this information was mentioned in the review article.

The abstract that accompanied the meta-analysis did state that there was “no overall association between moderate egg consumption and risk of [type 2 diabetes].” Even though that statement is of questionable validity (because the failure to demonstrate that a correlation is statistically significant is not the same as demonstrating the absence of a correlation), one might grant the author of the review article some leeway for simply quoting the other paper. However, the failure of the review article to mention the three large prospective cohort studies or the US portion of the data from the meta-analysis seems to indicate bias.

Another statement made in the review article was that “eggs reduce inflammatory markers including IL-6, CRP [C-reactive protein], serum amyloid alpha, TNF-alpha, and liver enzymes in patients with metabolic syndrome or those with diabetes.” The review article cited three references to support that statement. One of these references presented data on lipoprotein profiles and insulin sensitivity but did not discuss inflammation.⁷ The other 2 studies that were cited^8,9 presented a mixed picture. In one of these studies, consumption of eggs significantly decreased C-reactive protein levels, but in the other study the concentration of C-reactive protein was nonsignificantly higher in subjects assigned to consume eggs than in the control group. One of the studies found that another measure of inflammation (monocyte chemoattractant protein-1; MCP-1) fell substantially more in subjects consuming egg substitute than in those given whole eggs. One of the studies found that egg consumption decreased TNF-alpha levels, but the other study found that TNF-alpha levels fell slightly (nonsignificantly) more with egg substitute than with whole egg. Thus, it seems like a major stretch to imply that eating eggs is a good way to decrease inflammation.

Open-access journals and supplements to journals are not the only place one might find bias in medical writing. However, because of the financial incentives associated with these types of publications, the risk of bias is increased. Journals should improve their peer-review process, and we as readers should maintain a healthy degree of skepticism when evaluating educational materials that might be influenced by financial conflicts of interest.