Miss a day, miss a lot. Subscribe to The Defender's Top News of the Day. It's free.

Last week, Robert F. Kennedy, Jr., chairman of the board of Children’s Health Defense (CHD) and Brian Hooker, Ph.D., Francis P. Owen Distinguished Professor of Biology at Simpson University and CHD board member, sent a letter to Piotr Romaniuk, Ph.D., associate professor at the Medical University of Silesia in Katowice, Poland, and editor of Frontiers in Public Health.

The letter details gross errors in “Considering Interim Interventions to Control COVID-19 Associated Morbidity and Mortality — Perspectives,” an original research paper by Mark Christopher Arokiaraj, published Sept. 22 in Frontiers in Public Health.

After reviewing the paper, Kennedy and Hooker concluded that “because of the nature of the error and the faulty conclusions made from the erroneous results, this paper should be retracted as soon as possible.”

Here’s the letter:

Dear Dr. Romaniuk,

The journal Frontiers in Public Health lists you as the editor for the recently published paper, “Considering Interim Interventions to Control COVID-19 Associated Morbidity and Mortality – Perspectives.”

The author of the paper, Mark Christopher Arokiaraj, states in the abstract, “There is a correlation between COVID-19 related mortality, morbidity, and case incidence and the status of influenza vaccination, which appear to be protective.”

Unfortunately, this paper is fatally flawed. Your journal should retract it without delay to avoid complicity in potentially dangerous policies that might flow from the paper’s faulty conclusions.

Arokiaraj demonstrates the supposed correlation between flu shots and lowered COVID-19 risks in Figures 1 through 3 and 5 through 8. Their graphs show COVID-19 mortality, morbidity and incidence data divided by influenza vaccine uptake percentage. Arokjaraj then plots this value against influenza vaccine uptake percentage.

Each of these graphics also includes a trendline, showing a downward trend of the y-variable with increasing vaccine uptake. This line, along with the equations Arokiaraj includes on each graph, gives the illusion that there is an inverse relationship between COVID-19 mortality, morbidity and incidence and vaccine uptake. The effect suggests a protective role of influenza vaccination against COVID-19.

However, this relationship is solely the artifact of the author’s deceptive gimmick of dividing the y-variable by vaccine uptake and plotting the resulting values directly against vaccine uptake. This sophomoric error in statistics is so fundamental that it has a name: “multicollinearity.” Statisticians employing this trick can show a correlation regardless of whether one exists or not. Specifically, multicollinearity occurs when the statistician correlates one variable to another variable containing redundant information.

For example, the Arokiaraj paper divides COVID-19 mortality per million persons by vaccine uptake percentage to yield a new variable, which he then plots against vaccine uptake percentage. He thereby creates an inherent inverse relationship. The erroneous inverse relationship appears in Figure 1 (COVID-19 incidence), Figure 2 (COVID-19 mortality), Figure 3 (deaths per cases), Figure 5 (deaths per cases multiplied by tests), Figure 6 (deaths per cases divided by tests), Figure 7 (critical cases) and Figure 8 (critical cases per total cases).

Below is the actual mortality data from Table 1 plotted against vaccine uptake. This in fact shows mortality increasing slightly in countries with higher uptake of the influenza vaccine, with a modest correlation coefficient.

The second graph (below) shows the inherent inverse relationship between 1/vaccine uptake percentage plotted against vaccine uptake percentage. The third graph shows the result when this is combined with the first graph. Now, because of the flaw of multicollinearity, the graphs demonstrate an erroneous downward trend in the data.

Thus, any protective benefit offered by the influenza vaccine as reported in this paper is an erroneous result of the fatally flawed data presentation.

Your journal should retract this paper as soon as possible before it causes harm in the form of damaging policies based on the paper’s unsupported conclusions.


Robert F. Kennedy, Jr. and Brian S. Hooker, Ph.D.