“The numbers make it look like an epidemic. Well, it’s not. It’s preposterous. This is a concoction to justify the giving out of medication at unprecedented and unjustifiable levels.” (New York Times, Dec. 14, 2013). This penetrating declaration was not said about the coronavirus pandemic. It was said about the soaring prevalence of Attention Deficit/Hyperactivity Disorder (ADHD) by none other than the “father of ADHD,” the famous American psychologist Keith Conners.
Soon after the Centers for Disease Control and Prevention (the CDC) published its 2012 estimate—according to which 11% of children ages 4-17 have been diagnosed with ADHD—Dr. Conners, who had set the initial standards for the diagnosis and treatment of ADHD, could keep silent no longer. He had devoted his entire life to the study of the disturbance, and by his own appraisal, 2%-3% of children, at the most, actually suffer from it. He warned from the podium at a convention of ADHD experts in Washington that the exaggerated number of diagnoses was “a national disaster of dangerous proportions.” A few years before his death, with a rare level of integrity, he hoped to vanquish the diagnostic “epidemic” that had been spread, among other causes, because of the many research studies that he had conducted.
But the virtual epidemic has only gathered steam, and today, in 2020, the prevalence of the diagnosis in certain places (e.g., Israel) exceeds 20%! This unbelievable prevalence has been observed in a very recent study (July 22, 2020), which relied on reliable medical records by a large care service in Israel named Leumit.
How could this be? Even if Dr. Conners was right, and we are dealing with a concoction to justify the giving out of medication, how could such a thing possibly be invented as a matter of science?
In this article we will reveal the answer.
The ADHD Article
A month ago, a particularly disturbing research study was published in the Journal of Attention Disorders—ironically, the journal that Conners had founded—according to which, “ADHD may be a risk factor for COVID-19 infection, independently of other risk factors.” Children and adults who are diagnosed with ADHD are more active and less obedient than others, so it is no wonder that they are infected more frequently by the coronavirus. But, not to worry, there is a medicine. According to Dr. Merzon and his colleagues (the authors of the study), psychostimulant medications for ADHD such as Ritalin can lower the increased risk of being infected by the virus.
The study, which was conducted by a group of Israeli researchers in 2020 during the initial months of the COVID-19 pandemic (February 1 through April 30), was based on the medical records of 14,022 patients insured by Leumit, the large care service mentioned above. It immediately entered the list of the most trending articles in the journal and received a large amount of media attention—not surprising in view of the fact that its conclusions concern today’s hottest topic, COVID-19.
On the one hand, we are dealing with sensational data, given that they imply that it is possible to significantly and easily reduce the risk of falling ill with the coronavirus among a very large segment of the population. And indeed, the researchers recommend diligence in applying medicinal treatment for ADHD, because it can “mitigate the spreading of the virus,” a virus that has disrupted life across the entire globe.
On the other hand, if the study’s conclusions are ill-founded, then the researchers’ recommendations are ethically problematic. The medications for ADHD work much like other psychostimulant substances. And in spite of the wide belief that they are absolutely safe, many studies have demonstrated that they are liable to cause physiological addiction and severe side-effects. Moreover, the recommendation would not be met happily by many parents, who prefer not to administer medications to their children at times when there is no school, whether because of lockdown or thanks to summer vacation.
In light of the importance of the subject, particularly in these complex and sensitive times, we have decided to examine the study in depth from a scientific viewpoint. This examination, we believe, will assist parents in deciding how to care for their children who have been diagnosed with ADHD, and will simultaneously expose how an “objective” research study is liable to create misrepresentations, to stir up fears, and perhaps even to cause an increase in diagnoses and prescriptions of medications for attention disorders.
After two requests to each of two first authors of the study (made through a third party uninvolved in controversy over ADHD), we have failed to obtain the raw data set on which the research was based. Our examination is therefore based on an analysis of the article itself, including its comparison with data that were presented in two additional studies, published just one day later, by the same first author along with one or two other authors.
We chose to refer to these two additional studies because both were based on medical records of the same care service (Leumit), and because each of these two studies exposed one side of the issue at hand. One study dealt with ADHD, and one study dealt with COVID-19. The study dealing with ADHD was published in the same journal, Journal of Attention Disorders, and examined the relationship between the disorder and shigellosis, a bacterial disease that primarily strikes children; the study that dealt with COVID-19 examined the relation between vitamin D and coronavirus symptoms. In fact, this last study was carried out on the very same set of 14,022 medical records that was used in the study on ADHD and COVID-19.
In the process of our examination, we could not believe what we discovered. The study on ADHD and COVID-19 suffers from a series of manipulations and spins that are inappropriate in scientific research that aspires to objectivity and that aims to reveal truths. From our consternation came our decision to write a full-length article, which could be judged by our colleagues in the scientific community. We uploaded this article to a designated preprint server for psychological science (the full article can be found at Psyarxiv) to receive feedback from fellow scientists and, in parallel, we submitted it for publication in the Journal of Attention Disorders (the same journal in which the study was originally published).
“What was the reaction of the journal?” you probably wonder. We will reveal this information soon, but first, let’s see what was so problematic in the study on ADHD and COVID-19. The study, as can be seen in our full-length paper, is based on a series of distortions or falsifications, which make it difficult not to conclude that they were meant, whether intentionally or unintentionally, to achieve desirable outcomes, and perhaps also to obscure results that did not match the study’s conclusions.
Our examination has exposed seven practices that can be divided into two main types: (1) manipulation of the way in which the researchers chose to measure the phenomena being studied, i.e., of the operational definitions of variables in the study; and (2) spin in the choice and presentation of the data and the results. The concept “spin” might perhaps strike you as pertinent only to the political/public sphere, but in recent years, with the increase of distortions in the medical sciences, it has begun to be used in leading scientific journals to describe practices that are intended to achieve desired results by questionable means.
Manipulation of Variables
For example, there was an astonishing manipulation of the central independent variable of the study—ADHD. How did the researchers choose to define “a person with ADHD”? They included in this category everyone who had ever been medically diagnosed as such. That might seem to be a reasonable decision; however, in the article on the disease of shigellosis, which, recall, was written by the same first author, a person with ADHD was categorized as such only if, besides the diagnosis, there were at least three documented prescriptions obtained for ADHD. Why did the researchers choose a different definition for each study? After all, both studies were based on the same trustworthy source of data (Leumit medical records); and both were published more or less at the same time in the very same scientific journal.
In order to answer this question, it is worth reading closely what the authors write in the shigellosis study: The diagnosis of ADHD, they explain, is “given temporally with referral of the patient to ADHD clinic.” In other words, the preliminary diagnosis by itself, according to the researchers themselves, is an operational classification which is unreliable as an indicator of the existence of ADHD. Why, then, do the very same researchers use this problematic definition in their study on ADHD and COVID-19?
It is important to understand that we are not engaging in a petty focus on details. From the moment when the researchers chose to define ADHD, huge and unexplained gaps were created between the rates of the diagnoses. While the shigellosis study reported rates of diagnosis between 8.6% and 10.6% among children aged 5-18, the COVID-19 study reported rates of 18.85% to 28.14% among those aged 5-20 (weighted average = 20.25%!).
We are dealing with magnitudes that are anomalous at an Olympic level. Not only are these estimates double those that were reported by the same first author and another co-author, but they are light years distant from the data documented in the literature on ADHD. They are far from the CDC data, and they are certainly far from the data of the DSM (Diagnostic Statistical Manual of the American Psychiatric Association), which states that 5% of children suffer from ADHD.
In practice, when such an anomalous number is obtained, it should become the prominent result in the study and turn into an urgent subject of discussion in the scientific community. Either way, if the basic variable of the research is erroneous, then the study is entirely problematic, and if the basic variable is accurate, then that means that we are at the height of a great epidemic of ADHD, much greater than the coronavirus epidemic.
Another disturbing example of manipulation of the definition of variables can be seen in the way that the researchers defined the ages of the participants of the study. In order to verify that the risk for COVID-19 infection is indeed related to ADHD and not to other variables (for example, gender or socioeconomic status), the researchers must ascertain that the increased risk of infection is maintained even when they control statistically for socio-demographic variables.
In order to control for the age of the participants, the researchers divided them into four groups: youth up to age 20, 21-40 year-olds, 41-60 year-olds, and seniors over age 60. This division, however, does not fit the study framework because there is no rationale for including children, teens, and youth in the same basket, particularly in a study that deals with COVID-19 and ADHD. ADHD is far more common among children, according to both the scientific literature and the researchers’ own data.
Young age (i.e., up to age 20) was found in the study as the most dominant risk factor for catching COVID-19, a risk that was 50% higher than the one that was reported for ADHD. If, therefore, we wish to know what the real source of infection risk is, we must examine the differences within the first, young age group. Otherwise, we have no way of knowing what actually raised the risk, the diagnosis or the age. Is the source of the behaviors that contributed to the infection (e.g., often runs about, often leaves seat) in the ADHD or in the young age of the person diagnosed?
Beyond the fact that we have here an erroneous methodological decision, the fact that the researchers had access to the actual age of the subjects raises the suspicion of inappropriate manipulation, because they could easily have controlled for this variable as a continuous variable, as is appropriate statistically, or at least created a more appropriate division, such as kindergarten age, elementary school age, middle school, high school, and youth age 18 and up.
Spin in Reporting Results
An example of spin in the report of data and results is especially disturbing. Towards the end of the article, the researchers note “a major weakness” of the study and admit that “data regarding the presenting symptoms and severity of COVID-19 infection, as well as adverse clinical outcomes (hospitalization and mechanical ventilation) were not assessed.”
This choice of words (“not assessed”) is, in fact, quite problematic. After all, the authors had access to the medical records, which included, according to their own words, “medical visits, laboratory tests, hospitalizations, and medication prescriptions.” Furthermore, in the study that dealt with vitamin D and COVID-19, which was based precisely on the same data set of 14,022 medical records, and which had three of the same authors, the researchers did assess the symptoms and severity of COVID-19 infection.
Is ADHD related to complications of COVID-19? Why did the researchers choose not to assess these complications? They had the data. The decision to ignore the complications is disturbing for two reasons. First, at this stage of the epidemic, we know very well that the coronavirus is especially dangerous for the elderly, while children seldom develop severe symptoms. The significance of this is that if the researchers had considered the complications of COVID, they would have been liable to discover that the group of individuals with ADHD was the group with the lowest risk for complications, simply because they were significantly younger than the rest of the sample (here is yet another example of the importance of statistical control for the age variable).
It is certainly reasonable to assume that precisely the group of individuals with ADHD, even if they had a valid diagnosis, did not suffer from severe, if any, symptoms of COVID-19, and certainly they had no need for hospitalization or ventilation. The omission of such a result, one not suiting the general narrative of the article, is the very definition of spin.
Second, without the data on COVID-19 complications being shown, we cannot evaluate the safety of the medications for ADHD. Is it proper to prescribe ADHD medications during a global infectious epidemic? Even the supporters of medical treatment for ADHD recently launched, in the medical journal Lancet Child and Adolescent Health, special guidelines for “starting ADHD medications during the COVID-19 pandemic.”
According to the new guidelines, in the time of COVID-19, medicinal treatments should not be started if the patient with ADHD has a background of breathing problems and/or heart problems. Not only that, but there is evidence from studies on psychostimulants (including methamphetamines, which are similar in chemical constitution and in their effects to the stimulant medications for ADHD), that they are liable to harm the blood-brain barrier and to increase the risk of the penetration of dangerous bacteria and viruses into the brain, such as HIV.
These three examples are only some of the manipulations and spins which we discovered in the study at hand, although even these alone, as we have shown, already refute the researchers’ conclusion that makes it seem that there is evidence for ADHD being a risk factor for infection by the coronavirus. The editor of the journal was exposed to four more distortions and yet, he chose to reject our manuscript.
The formal justification for the rejection was the fact that we uploaded the manuscript to another publication venue (the Psyarxiv preprint service). This is of course a legitimate editorial decision. Yet, it is important to know that uploading manuscripts to preprint services is a conventional scientific practice. Preprint services were constructed to enhance scientific transparency and to allow researchers to receive responses from the scientific community as early as possible. These services are especially important in cases where the uploaded articles have time-sensitive implications, such as the articles that address the current pandemic.
Approximately, 6,000 COVID-19 related scientific articles were hosted by preprint servers within a period of 4 months since the beginning of the pandemic, thus emphasizing the unprecedented role of preprints in the dissemination of COVID-19 science.
In our response letter, we therefore reminded the editor of the journal of these facts (the common practice and the unique role of preprint services) and mentioned again the unfounded, biased, and even dangerous conclusion that was made by the authors of the study on ADHD and COVID-19: “adherence to anti-ADHD treatment should be encouraged in an attempt to reduce the spread of COVID-19 infection.”
We asked the editor to reconsider his decision not to accept our article and suggested that even if he remains with this decision, that he will objectively consider the severe manipulations and spins that we have documented when he decides whether to keep or retract the study on ADHD and COVID-19 from the journal.
The study on ADHD and COVID-19 did not prove that medications can lower the risk for COVID-19 infections. Even if these distortions were not deliberate, we are convinced, given the data they had, that the researchers’ recommendation to be diligent in applying medicinal treatment for ADHD during the COVID-19 pandemic is not ethical.
Besides the futility of trying to get many parents to give their children medicinal treatment at times when there is no school, the medications for ADHD, like all medications, come with side-effects. Even though it is customary to consider them as safe, an up-to-date literature review of 260 studies of medications for ADHD, conducted in 2018 by Cochrane,1 revealed that most of the children who use medications for ADHD experience at least one side-effect, including, for example, dizziness, anxiety, nervousness, nightmares, restlessness, sadness, and tics.
Further, around one percent of diagnosed children experience severe side-effects, including cardiovascular problems and psychotic or manic episodes. One percent might sound like a small number, but as we saw at the beginning of this article we are dealing with a great number of children, the millions of children across the world who are diagnosed with ADHD and are using medications for it.
Now we can understand what Dr. Keith Conners referred to. Overdiagnosis and overtreatment can be born of distorted and biased research studies. It is indeed possible that the study of COVID-19 and ADHD is anomalous in that so many manipulations and spins have clustered in it, but, as we wrote at the beginning, such distortions are not new to the scientific community, especially in the bio-medical field.
The bio-medical field is replete with biases, the most famous of which are publication bias—selective publication of studies that support the effectiveness of a medication and hiding of studies that do not support it; outcome/data reporting bias—exaggeration of positive results, and concealment of negative results, such as side-effects or limited effectiveness; selective analysis and presentation, and more.
These biases are not always intentional, but it seems that they may often happen because of unhealthy connections between scientists and pharmaceutical companies. These connections, as many physicians, investigators, and scientists warn, pollute the medical sciences and endanger public health. They are fundamentally immoral—all the more so in the case of medications for ADHD that are given to millions of children across the world.
The authors wish to thank psychologist Gideon Weisz for his major contribution to the translation and improvement of the current article.
- Cochrane is an international scientific organization that conducts high quality unbiased systematic meta-analyses of medical science topics. ↩
Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.