This is the second in a two-part series exploring Attention Deficit and Hyperactivity Disorder. In Part I, Mark Fineman argued that the current diagnosis and treatment of ADHD has not met the burden of evidence and reason to be considered a legitimate science, and in Part II Steven Novella will make the rebuttal argument that it has.
by Steven Novella, MD
In the last issue of the NEJS, Mark Fineman did an excellent job of defending the position that attention deficit and hyperactivity disorder (ADHD) is deserving of skepticism and may not be a real and legitimate medical phenomenon. In this follow up article, I will take the opposite position, that ADHD is a legitimate medical diagnosis and that most criticism of ADHD is based upon a misunderstanding of clinical science.
Skepticism or Denial
As a skeptic, I am often thrust into the position of arguing that a discipline is a pseudoscience rather than a legitimate science, or that a belief or claim is not warranted by logic or current evidence. There are a few topics, however, on which I find myself in the opposite position, that of defending a science as legitimate or a specific claim as probably true. Evolution, for example, deserves to be defended from creationists who falsely attack it as a pseudoscience.
Those who attack legitimate science or history are referred to as deniers. Holocaust deniers, for example, deny the reality of the Jewish Holocaust of World War II. Michael Shermer has advocated referring to creationism as “evolution denial,” arguing that the tactics used are similar to those used by the holocaust deniers (Shermer, 1998). Skeptics have also recently had to confront a group that I will refer to as “HIV deniers,” who argue that the human immunodeficiency virus (HIV), despite accepted scientific belief, does not cause the acquired immunodeficiency syndrome (AIDS).
The most controversial of the denial syndromes that I have personally confronted is what I believe to be “psychiatry denial.” Psychiatry deniers represent a broad group but have in common the denial of mental illness as a legitimate medical or biological phenomenon (Ryley 2001, for example). Extreme deniers even argue that the biology of the brain is not responsible for human thought and behavior. The most moderate deniers accept the possibility of mental illness but believe that the current state of psychiatry has yet to achieve sufficient scientific validity.
I will argue that the anti-ADHD position, as argued by Mark Fineman, falls into this moderate category of psychiatry denial. Our current state of scientific knowledge is sufficient to state that ADHD is a genuine disorder and that there is probably a biological cause in some, perhaps most, cases.
It is important for skeptics and scientists to have a working definition of science that includes all legitimate science and excludes all pseudoscience. This problem of definition, referred to as the demarcation problem, still challenges scientists and philosophers. However, there is general agreement that a key feature of legitimate science is that it is empirical – hypotheses are tested against reality in such a way that it is possible to prove the hypothesis wrong. A broad range of activities may fit under this definition.
One manner in which skepticism slips into denial is to use an arbitrarily narrow definition of science in order to exclude as nonscientific the discipline which is being denied. Evolution deniers, for example, use a laboratory or experimental definition of science as a definition of all science. Their intent is to exclude all historical sciences, including evolution, as not really science. UFO skeptics may become deniers if they refuse to accept intelligent inference as a proper scientific method, insisting that direct physical evidence is the only evidence that counts.
Psychiatry deniers often fall into this trap, by using a “basic science” definition and excluding the so-called “clinical” sciences. In order to understand psychiatry, one must view it in the proper context of a clinical, or applied, science, as is all of medicine. Practitioners engage in clinical decision making in order to arrive at the best intervention for a given patient (even not treating is a decision). Such decisions should be based upon the best scientific evidence available, and on a basic understanding of biology, but they are always made in the absence of definitive or absolute scientific certitude. Often practitioners are forced to make decisions based upon probability, or our current best guess. Also, because we deal with people, decisions must be individualized to some degree, and judgment is often involved.
Of the medical disciplines, psychiatry has the greatest proportion of judgment and the smallest proportion of hard clinical evidence. This is due to the fact that psychiatry deals primarily with human thought, mood and behavior, which is a chaotic and complex phenomenon. It is in fact an attempt by the human mind to understand itself. This presents a special challenge for achieving scientific objectivity.
Human thought, mood and behavior are best viewed as the final product of biology (the hard-wiring and physiology of the brain), psychology, and culture. It is very difficult to tease out exactly how and to what proportion each of these three factors play in any given behavior. For example, if a person is very aggressive, is that due to the hard-wiring of their brain, is it due to the fact that they were physically punished as a child, or is it due to the violent culture in which they were raised? Most likely it is a complex interaction of all three. To add more complexity, the brain is also a plastic organ— it changes with use. Therefore a lifetime of aggression may change the biochemistry of the brain, biologically reinforcing the aggressive behavior.
In general psychiatry deniers emphasize the psychological and cultural influences on behavior, and downplay or dismiss the biological influences without real justification. They seem annoyed at attempts to explain human behavior with biological arguments, as if that may somehow justify bad or criminal behavior, or call into question our comfortable assumption of free will. Certainly their arguments gain public support from the abuse of biological arguments in the courtroom, and here I believe there is room for legitimate skepticism. There are also those psychiatrists who overemphasize biological causes of behavior. This is, after all, the old “nature vs nurture” debate that will likely never be completely resolved. However, in general there is adequate recognition in the neurosciences (including psychiatry) of the complex interplay of all three factors in determining human thought, mood, and behavior.
Let us now turn to so-called “abnormal” behavior. Because of the broad spectrum of human behavior, especially across cultures, it is difficult to derive a clean operational definition of “normal” vs “abnormal” when it comes to human behavior (another demarcation problem). In fact psychiatry has moved away from these terms, preferring the less judgmental “healthy” vs “unhealthy.” In determining which behaviors are deserving of psychiatric attention, a purely practical approach has been taken. If someone presents to medical attention with a perceived problem of mood, thought, or behavior (by themselves or by others), then that is worthy of psychiatric attention (not necessarily a diagnosis of being mentally ill or treatment).
The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) is an attempt by the psychiatric community to create a standard of diagnosis for psychiatric problems. It is understood among neuroscientists that abnormal human behavior cannot be legitimately pigeon-holed, as any attempt at categorization might suggest. The DSM-IV, rather, is meant as a clinical framework, a way of discussing and approaching the spectrum of psychiatric illness. It is also used to provide for consistency between different clinical studies.
With all of the complexity I have outlined above, it is important to recognize that psychiatry often deals with individuals whose mood, thought, or behavior are so far outside of the common experience of most humans that it begs for some explanation. For example, individuals who cut their arms (often misinterpreted as suicide attempts) because (they claim) the physical pain helps them deal with their psychological pain, young women who starve themselves literally to death because of an all-consuming fear of being overweight, or individuals who are unalterably convinced that the CIA is monitoring their thoughts through the fillings in their teeth. I feel it is legitimate to characterize such behavior as unhealthy, or even abnormal, even though there is no way to objectively measure the degree of abnormality, or to provide a sharp demarcation between such behavior and “normal” behavior. Just as it is not necessary to be able to define exactly where the line between science and pseudoscience is in order to feel confident that homeopathy is indeed a pseudoscience.
Psychiatric and Medical Diagnosis
There is a diagnosis commonly made by physicians which is based entirely on the subjective symptoms reported by the patient. There is no method to objectively quantify the severity of these symptoms, or even to document their presence, and physicians must rely upon the patient’s subjective interpretation of their own symptoms. Although the disorder is felt to be a brain disease, there is no test which can demonstrate any brain abnormality. MRI scans reveal normal anatomy, EEG’s reveal normal electrical activity, blood tests are all normal, and even pathological investigation reveals no difference between the brains of alleged suffers and non-sufferers. There have been a few studies showing tantalizing biological differences between sufferers and non-sufferers, but the significance is unclear and the differences are purely statistical and not useful for diagnosis.
After reading Mark Fineman’s criticism of ADHD the reader might think that I am talking about this disorder; in fact I am referring to migraine headaches. But the paragraph above may just as well apply to ADHD. It might also apply to a host of other neurobehavioral or psychiatric disorders. That migraine is a “real” disorder is not controversial, yet virtually every criticism that has been leveled against the validity of ADHD can also be applied to migraine headaches. Migraines are also, for example, diagnosed when patients have a certain number of symptoms out of a list of possible symptoms, for a certain duration of time, just like ADHD. The reasons for the different popular reputations of migraine and ADHD are complex, but they are due in part to the fact that the principle symptom of migraine headaches is pain, which everyone can relate to and easily understand as a symptom of something being wrong. Healthy people should not feel physical pain in the absence of trauma.
Psychiatric symptoms, however, are not easily understood as being unhealthy. Admittedly, this is partly a matter of pure definition. Should we extend the definition of health to include behavior, thought, and mood? Is it possible for any person to suffer from what can reasonably be called a “disorder” which causes only certain behaviors? Or, should all behaviors, all mood states, and all patterns of thought be considered part of the normal human spectrum and chalked up to individuality and simple “human frailty?”
The medical community has increasingly over the last century moved toward the idea that some behaviors, moods, and thoughts are so far outside of the common experience of most humans that they should be considered a medical condition, and thus was born the specialty of psychiatry. I will again return to a few extreme examples to establish the point. Some individuals (mostly women) restrict their eating and increase their exercise to such a degree that they lose a physically dangerous amount of weight, often to the point of death. Regardless of the cause of such behavior, should the behavior itself be considered abnormal? Since different individuals exhibit roughly the same set of such eating behaviors, should we refer to them collectively by some name, and should we investigate possible causes of such behavior (again recognizing the interplay of biology, psychology, and culture in forming human behavior)? Since there is clearly established harm which results from such behavior, is it proper to use a term such as “disorder” to refer to this pattern of behavior? I maintain that the answer to all of these questions is a clear “yes.”
If one is unwilling to accept that anorexia nervosa is rightly considered a psychiatric disorder, worthy of study and treatment, then clearly all of psychiatry must be discarded, along with a good deal of neurology as well. Yet if we agree that a collection of recurring symptoms, even when dealing purely with mental function, can be recognized as a “syndrome” and if that syndrome is associated with clear harm and can therefore be further called a “disorder” then we must derive some meaningful way to define mental disorder. Such a definition cannot be restricted to demonstrable biological cause or pathophysiology, since anorexia nervosa (like migraines) displays no such biological change.
Is ADHD a Disorder?
ADHD is an emotional and controversial topic in current society. I believe this is largely due to the fact that the diagnosis is primarily applied to children and is being used to prescribe psychoactive drugs to children. The question of ADHD can be divided into three primary questions: is ADHD a valid disorder; what is the cause (biology, psychology, or culture) of ADHD; and what is the current state of diagnosis and treatment of ADHD?
Mark Fineman provides an adequate definition of disorder, quoted from Dr. Barkley, the leading advocate of ADHD as a biological illness. “It constitutes a failure or serious deficiency in a mental mechanism that is universal to humans (a psychological adaptation in the evolutionary sense), in this case response inhibition and self-regulation. And it produces harm.” Mark Fineman dismissed this definition as vague, relying on what I feel is a slippery slope logical fallacy, without offering an alternative definition.
Most children who are diagnosed by standard methods with ADHD are identified because they are performing poorly in school and are considered by their teachers and/or parents to be disruptive and distracted. Neuropsychological testing has determined that most such children have a diminished capacity to regulate their own behavior and to focus productively on mental tasks. In our society these deficits produce measurable harm in that such children do poorly in school, end up in a lower socioeconomic status than their peers, and have a higher rate of divorce. ADHD therefore meets Barkley’s criteria for a mental disorder. I would add also that ADHD is a recognizable syndrome, in that many individual children seem to be suffering from a similar set of deficits and subsequent harm.
Mark Fineman echoes many of the common criticisms of ADHD as a real disorder, but the criticisms are not valid. For example he argues that ADHD may not be due to biology but may rather be the result of poor parenting, or even poor teaching. The underlying cause of ADHD (or any disorder) is irrelevant to whether or not it is a real disorder. There is no reason, for example, why there cannot be purely psychological disorders (such as post-traumatic stress disorder). The definition of disorder above makes no mention of underlying cause.
Another criticism is that ADHD may simply represent one end of the normal spectrum of human ability and behavior. Again, even if true, this is not incompatible with ADHD being a “real” disorder. Let us look, for example, at dyslexia. Dyslexia, simply defined, is a diminished capacity for reading and writing. (It is not, as commonly believed, the tendency to reverse letters.) Dyslexics have great difficulty learning to read, and can never read quickly or efficiently. The current thinking is that dyslexics represent the lower end of the bell curve of human reading capability. However, it meets the criteria for a disorder given above – it is a diminished mental ability normally possessed by most humans, and it produces harm. The harm of dyslexia is similar to the harm of ADHD (although much narrower in scope), in that it results in poor school performance. In pre-literate societies dyslexia would not be recognized as a disorder. It is interesting to speculate that if there were a medication available which was demonstrated to improve the ability of dyslexic children to read, would there be a controversy surrounding the prescribing of this medication to dyslexics?
Another strategy to deny ADHD as a real disorder is to claim that the definition is too liberal. As Mark Fineman writes, “By this vague and liberal definition, which human frailty, which problem in life management is not a disorder?” This is somewhat of a slippery slope fallacy, however. There is admittedly no sharp line between “disorder” and just human frailty, but this does not exclude extreme enough behavior that recognition as a discrete syndrome is warranted. How much difficulty reading is necessary to diagnose dyslexia? How overweight does someone have to be to be considered unhealthy? Where should we mark the boundary lines of “normal” thyroid function from overactive or underactive? These are ultimately arbitrary, but does not invalidate that extreme difficulty reading, or overweight, or underactive thyroid should be treated as disorders.
So I maintain that the definition of a mental disorder as lacking a universal human mental capability, producing demonstrable harm, is a perfectly reasonable definition. The criticisms that disorders by this definition may lack demonstrable biological pathology, or may represent the extremes of human variation with no objective line of demarcation, are irrelevant. Further, no substitute or modified definition has been proposed. Therefore we are left with the only option (as extreme deniers wish) of eliminating the concept of mental disorder entirely.
The Diagnosis of ADHD
Perhaps the primary criticism, reiterated by Mark Fineman, of ADHD as a real disorder is that the methods used to diagnose it have no validity. Critics place great significance in the fact that the diagnosis of ADHD, like many psychiatric disorders, is based solely upon the symptoms that the patient reports and the behavior that is observed, and not on any kind of biological test. Some psychiatry deniers categorically deny that anything can be considered an illness or disorder unless it has a demonstrable biological cause. Admittedly, this is a matter solely of definition. I advocate a more liberal definition, as outlined above, in that all that is required is a demonstrable deficit and demonstrable harm, regardless of cause.
Here we see the selective use of criteria, which is an important hallmark of denial strategy. For example, there are non-controversial disorders which are also based entirely upon symptoms for diagnosis, with no biological test. Mark Fineman used Alzheimer’s disease as his example, but this was not a fair example, because, as he pointed out, Alzheimer’s disease is a pathological diagnosis that can be made on biopsy or autopsy. Better examples would include dyslexia and migraines, as I described above. By Mark Fineman’s criteria, migraines are not a real disorder.
The final level of criticism is that the diagnostic criteria for ADHD are too vague to be objective and reliable. They could, it is argued, apply to any child. If true, this would be a legitimate and devastating criticism. But is it true?
One way to test this assertion is to have different practitioners apply the criteria to the same set of children, without knowing what the other practitioners concluded, and see if the diagnoses match. The current method of diagnosing ADHD has passed such a test, which is called inter-rater reliability. Different people get the same result when the method is applied to the same children. Therefore the criteria for diagnosis are reliable.
Are they valid? Fineman argued no, offering as testimony the 1998 NIH report on ADHD, but he selectively quoted the report and misinterpreted their ultimate conclusions (NIH, 1998). The NIH was referring only to a biological test (which does not yet exist, which is what they meant), not the current diagnostic criteria based upon symptoms. Here is a more complete quotation:
“The diagnosis of ADHD can be made reliably using well-tested diagnostic interview methods. However, as of yet, there is no independent valid test for ADHD. Although research has suggested a central nervous system basis for ADHD, further research is necessary to firmly establish ADHD as a brain disorder. This is not unique to ADHD, but applies as well to most psychiatric disorders, including disabling diseases such as schizophrenia. Evidence supporting the validity of ADHD includes the long-term developmental course of ADHD over time, cross-national studies revealing similar risk factors, familial aggregation of ADHD (which may be genetic or environmental), and heritability.”
The report’s final word on diagnosis reads:
“In summary, there is validity in the diagnosis of ADHD as a disorder with broadly accepted symptoms and behavioral characteristics that define the disorder.”
This is a far cry from Mark Fineman’s characterization of the NIH report’s conclusions.
What validity really means is that the symptoms upon which the diagnosis is based relate to the disorder. Here we have a chicken and egg problem. We believe the disorder is real because we observe a set of symptoms occurring in different children, then we use that set of symptoms to diagnose the disorder in other children. This has internal validity but cannot have external validity without reference to an underlying cause. External validity would certainly help establish ADHD as a real disorder but is an unreasonable criteria for disorders where the cause is currently unknown.
Are they objective? Here there is legitimate room for criticism. The criteria are only partly objective, because a degree of reporting and interpretation is required. As described above, this is unavoidable when dealing with the clinical science of psychiatry. Someone dedicated to psychiatry denial can use this perfectly reasonable sounding criteria to argue that almost all of psychiatry is not a real science because of this lack of objectivity. However, a great deal of medicine would be thrown out too. Pain, for example, is a completely subjective symptom. We cannot measure pain, we can only infer its presence by what a patient reports and how they act. It would not be reasonable, however, to remove the concept of pain from all medical textbooks because it is a subjective criteria. This is the equivalent, however, to removing behavior from the criteria of all psychiatric diagnoses.
Let me put this concept of using subjective criteria into the broader context of clinical medicine. During a course in medical school on physical diagnosis, I had the privilege of being taught the art of auscultation by a world-famous cardiologist, Dr. Harvey. He taught that before we can learn to discern heart sounds which are abnormal, we must first learn the spectrum of what is normal. Once we have learned what is normal we will then recognize what is abnormal when we hear it.
This is a basic principle of clinical medicine and is true in psychiatry as well. Practitioners must become intimate with the spectrum of normal or healthy human behavior in order to recognize what is abnormal or unhealthy. All children have a certain degree of inattentiveness, for example. Does this mean, however, that any degree of inattentiveness is normal? Can an experienced practitioner make the judgment that a particular child has an abnormal degree of inattentiveness?
Having said all this, I must point out that, of course, objective criteria are better than those which require judgment or interpretation. Whenever we can base a diagnosis on a number, or the presence or absence of a hard finding, this is always preferable to having to make a clinical judgment. One of the ways in which the science of medicine advances is by finding objective ways to establish a diagnosis. When a diagnosis is based entirely on subjective complaints, there remains a little question mark in the mind of a good skeptical physician. This is an area that requires further study, and one in which we must remain open to questioning our beliefs. However, like migraine headaches, the reality of a disorder can be established on purely clinical grounds to a very high degree of certainty—high enough to warrant diagnosis and treatment.
Does ADHD Cause Harm?
The consensus of opinion, as reflected by the NIH report, is that ADHD is correlated with a host of measurable harm. The NIH report says of children with ADHD (NIH 1998):
“Their academic and social difficulties have far-reaching and long-term consequences. These children have higher injury rates. As they grow older, children with untreated ADHD in combination with conduct disorders experience drug abuse, antisocial behavior, and injuries of all sorts.”
Critics dismiss this high degree of correlation by stating that it is correlation only, with no proof of causation (I will address this style of argument in more detail later). However, when ADHD is treated, the associated harm improves, which strongly suggests that there is a causal connection. One recent study, for example, shows that when ADHD is treated the associated risk of substance abuse decreases as well (Biederman 1999).
Mark Fineman attempts to throw further doubt upon the reality of ADHD by arguing that there is a lack of consensus as to the prevalence of the disorder. I disagree with his characterization of the facts, however. The figure of 3-5% of the pediatric population is used as a conservative interpretation of the data, and is quoted in the NIH report. A thorough review of epidemiological studies reveals a broader range of 2-17%, although if only the better designed studies are used this range narrows to 5-10% (Scahill 2000). The American Academy of Pediatrics currently quotes a figure of 4-12% (AAP 2000). This degree of agreement is typical and perfectly acceptable for any medical diagnosis. Variability is the result of differences in the population studied, methods of diagnosis, and criteria used. Advocacy groups of any disorder will tend to quote the highest numbers they can find, regardless of validity, but it is not reasonable to use this as a basis for criticism of the disorder itself.
Fineman also argues that if ADHD simply represents one end of the normal spectrum of behavior, then we must also consider the other end of the spectrum abnormal as well; therefore all estimates for the prevalence of ADHD must be doubled in order to know the full prevalence of abnormal children with respect to activity and attention. This argument, however, is not logical. Remember, the working definition of a mental disorder is that it represents the lack of a universal and adaptive ability, and that it produces demonstrable harm – not simply that it represents two standard deviations from the mean of behavior.
Let us again take dyslexia as an example. We may consider the worst 5% of readers in our society to have a reading disorder, because they have difficulty in a literate society. This does not logically mean that we must consider the top 5% of readers as having a disorder. Having very above average reading skills is an advantage, not a disadvantage. It produces no harm, and does not represent the lack of an ability.
Likewise, children who have a well above average attention span, ability to modulate their behavior, and overall “executive function”, do not need to be considered as having a disorder simply because the functional lack of these traits is.
Again, to put this into a broader clinical perspective – there are different ways in which physical or mental traits statistically correspond to diseases or disorders. In the first type, normal is defined as the middle two standard deviations on a normal, or bell-shaped, curve, with the extreme right and left of the curve being considered abnormal. However, sometimes, as with ADHD, only one end or the other is considered abnormal, usually because it represents the lack of a function where an overabundance of that function produces no harm. However, not all traits fall on a normal curve. Some disorders represent a separate population from the normal population. In such cases there will be a separate curve, representing the diseased population, superimposed upon the normal curve representing the healthy population. This produces a bimodal, or two-humped, curve. The healthy or normal population blends into the diseased population. Therefore if we use this trait (it may be the result of a blood test or some physical measurement) as a criteria for diagnosis, we must decide where to place the line between “normal” and “abnormal,” keeping in mind that no matter where we draw the line we cannot have a perfect test.
There is as yet no consensus on whether or not ADHD represents one end of a normal curve, or a bimodal (two-humped) curve representing a disordered population superimposed over a “normal” population. Both situations, however, are consistent with ADHD being a real disorder, but this does have implications for its ultimate underlying cause.
Is ADHD a Biological Disease?
This is a very interesting question of critical importance to ongoing scientific advance in our thinking about ADHD. Before I consider the evidence for ADHD directly, it is important to discuss the nature of neurological disorders in general. One of the major points of criticism of psychiatry deniers in general is that psychiatric diagnoses often cannot be established on biological grounds. But is this a legitimate criticism, or another example of setting the bar arbitrarily high?
To answer this question we must first consider what is the biological basis of mood, thought, and behavior. It has been established beyond all reasonable scientific doubt that the brain is the organ of the mind. When we feel a feeling, think a thought, or take an action, this is directly caused by the firing of neurons in our brains. Neuroscientists have only recently started to unravel the ultimate question of the brain – how exactly is information stored in the brain, and how does the anatomy and physiology of the brain relate to our thoughts, mood, personality, beliefs, likes and dislikes, etc.
What we do know is that this level of information is coded for in the particular structure of connections that the different neurons (brain nerve cells) in our brains make with each other. This includes the type of neurotransmitters (chemical signals) that the neurons release and in what strength. Also, it has been discovered that astrocytes, cells that surround neurons and which were previously thought to be only necessary for physiological support of neurons, also modulate neuronal function.
When we look at the neurons of a brain under the microscope we can see these connections, but the pattern of connections has no meaning to us – we haven’t yet broken the code. Therefore, the brain of a brilliant and gentle scientist may look identical under the microscope to that of an aggressive predatory serial killer. The brain cells are not diseased or abnormal in a serial killer, but there must be something to the pattern of neuronal connections that corresponds to this behavior. This is true whether the pattern was caused by the killer’s upbringing, or was genetically predetermined. Our genetics, the physiological environment of our brains, the quirkiness of embryology, and all subsequent learning affect the pattern of neuronal connections.
Therefore, if a disorder were the result of the pattern of neuronal connections and activity, leading to an extreme and harmful behavior, we would not expect to be able to see the underlying biological cause with standard diagnostic techniques, even though we know from our basic understanding of neuroscience that a neuronal correlation must exist, regardless of cause.
Advanced diagnostic techniques allow us to image the distribution of activity of various neurotransmitters in the brain. This can be done with positive electron tomography (PET) scanning, where a radioactive label is attached to a certain chemical which is given to a patient, the chemical is then taken up by the brain and metabolized in accordance to the relative activity of certain types of neurons. We then get a picture of blobs of color (artificially computer added to aid in interpretation) representing different degrees of activity.
More recently functional MRI scan has allowed us to use MRI (magnetic resonance imaging) technology to directly view brain activity. Functional MRI and PET scanning have given us an important, yet still crude, window into brain activity. Even older methods, such as electroencephalogram (EEG) which looks at the electrical activity of the brain, allows us to view the relative activity of different areas of the brain. What we are finding is that some classical psychiatric illnesses, such as schizophrenia, sometimes display patterns of brain activity which are different from those of healthy controls.
Before I proceed to the biology of ADHD, I will discuss another important piece of background information. Neuroscientists have known for many years that the brain is modular, certain parts of the brain correspond to certain functions. We have learned what and where these functions are mainly by examining patients who have had brain damage. Also, one principle which has emerged from the last century of medical advancement is that for every identifiable biological function, there is a disease or disorder associated with the lack or malfunctioning of that function, unless such a lack would be immediately fatal. This general principle applies to the brain as well—for every identifiable neurological function there appears to be a neurological or psychiatric disease or disorder which is associated with the lack or underdevelopment of that function.
An important question for ADHD, therefore, is whether or not there is a part of the brain that correlates well with the functions that those who suffer from ADHD apparently lack. The answer to this question is yes. The frontal lobes of the brain are the most recently developed, from an evolutionary point of view, and provide what is now referred to as executive function. In other words, the frontal lobes are the masters of our brain – they allow us to modulate our own behavior. The frontal lobes provide for us a filter, enabling us to consider our deeds before we act, to consider our words before we speak, to act in socially appropriate ways, and to inhibit our desires in order to serve more abstract ideals, such as security and morality. Without this executive function of our frontal lobes we would literally say and do the first things that popped into our minds, with no filter.
What happens if someone sustains damage to their frontal lobes? Unfortunately, this is not an uncommon occurrence, as it is common during a head-on car collision to hit the front of the head against the windshield. Those who have suffered such damage have an immediate change in their personality. They become what is called disinhibited. They will often act and dress in a gregarious and inappropriate manner, they will swear and act out in public, they may take up smoking, spend their money without concern for their future financial needs, and in general do not take good care of themselves. They become different people.
It did not escape the notice of neuroscientists that many of these traits are similar to those displayed by people diagnosed with ADHD, leading to the hypothesis that ADHD is a disorder of the frontal lobes. In fact, given the principle stated above, if a disorder of executive function did not exist, one would have to wonder why. There are people who display an underdevelopment of every other type of neurological function, why not executive function, why is this part of the brain privileged?
Also, if there is a portion of the public whose frontal lobes are developed in such a way as to display poor ability in executive function, what would their brains look like? Again, they would look normal, unless we can find a way to image the particular pattern of neuronal connections and degree of neurotransmitter activity which corresponds to this function.
To the degree that we are currently able, we have looked at the function of the brains of people diagnosed with ADHD and compared them to normal controls. When we look with electroencephalograms, PET scanning, and functional MRI scan we find that the frontal lobes of those diagnosed with ADHD do in fact display less activity than normal controls, confirming what was predicted from a biological model of ADHD (Hale, 2000).
Mark Fineman and other critics have written off this emerging evidence as demonstrating correlation only, not cause and effect. We are reminded that correlation is not causation. Although this is a legitimate logical fallacy, it is also a frequently abused fallacy. Correlation does not prove causation, but sometimes correlation is due to causation. We must not assume causation, but we should not stop there and just dismiss correlation either. I think that it is instructive to point out that this same argument is liberally used by other deniers. I have been informed by creationists that simply because fossils may correlate to an evolutionary pattern does not mean that they were caused by evolution. HIV deniers also point out that the correlation of HIV to AIDS does not prove it causes AIDS. Another great example is the link between smoking tobacco and lung cancer. There are still tobacco industry apologists who argue that this well-established correlation does not prove causation.
We must take this argument one step further – what are all the possibilities and what is the best explanation? ADHD critics argue that perhaps the medication used to treat ADHD has caused these biological changes on functional imaging (deniers also use this argument to discount the same type of evidence for a biological cause for schizophrenia). This is certainly possible, but not very plausible. Why would stimulants cause a decrease in neuronal function? Also, what is the probability that stimulants would cause this decrease just in the frontal lobes – in the very part of the brain predicted by biological models of ADHD? I would say this explanation is unlikely, and not a good reason to dismiss the increasing amount of such biological evidence which is being accumulated. Further, more recent studies conducted on unmedicated subjects are showing the same correlations (Lazzaro 2001).
Critics also argue that the biological markers discovered so far cannot be used to diagnose ADHD. This is true, but this does not invalidate these markers, as critics suggest. In order for a test to be clinically useful it must be highly specific and sensitive. A test which is correct only 50% of the time is not very useful. However, a 50% correlation is highly significant in terms of establishing a biological connection.
Therefore, the current evidence strongly suggests that ADHD has an underlying biological cause. However, the full story is almost certainly more complex. Specifically, ADHD may not be a single disorder but two or more disorders which look similar clinically and are therefore being lumped together. Perhaps some children with the clinical syndrome of ADHD are the result of parenting style, while others are destined by their genes to have ADHD. Still others may have a genetic predisposition which needs to be activated by a certain childhood environment. There may even be other biological factors as yet not even considered which modulate the manifestation of ADHD.
Perhaps the most controversial aspect of ADHD is the most common method of treatment—stimulant medication. Multiple studies have demonstrated that several different specific stimulants (classically Ritalin, and most recently Aderall) are effective in reducing both the symptoms of ADHD and the associated harm. Medicine is a risk vs benefit game, and at present the best evidence suggests that the benefits of medication therapy for ADHD outweigh the risks. Counseling therapy is also helpful, but not as effective as medication.
Critics dismiss the evidence on a variety of grounds; for example they argue that the studies are not valid because the methods of diagnosis are not valid. Others suggest that excessive drug company influence calls into question the integrity of the scientific evidence. The first issue I dealt with earlier. The latter is somewhat beyond the scope of this article, but suffice to say there is no specific problem of integrity identified with the ADHD literature, so unless one wishes to dismiss the entirety of pharmaceutical research there is no reason to single out ADHD medication for criticism.
Others, such as Mark Fineman, point out that stimulants are potentially drugs of abuse, and are therefore dangerous. I find this argument to be irrelevant and inflammatory. Many prescription drugs can be abused, if used improperly or by a dedication recreational user. This in no way invalidates their usefulness as part of rational pharmacotherapy.
There are, however, very legitimate concerns regarding medication treatment for ADHD. Most of the studies of the safety and effectiveness of stimulants are short term. We therefore need more long term studies to more fully assess their impact. Of course, long term studies take a long time to complete, but such studies are under way. There are also concerns that some practitioners are prescribing stimulants to children who are too young, the effects of which have not been properly studied.
These are good questions that need to be asked and investigated. The strength of scientific medicine is that is continuously examines its own practices looking for flaws, gaps in knowledge, and areas that need improvement. However, similar concerns exist for many non-controversial medical disorders. It takes clinical judgment, and an unbiased look at risks and benefits, to determine if treatment as it presently exists should continue while we further examine questions which remain open. The consensus of opinion in the medical community, as reflected in the NIH report, is that careful and proper use of stimulants are safe and effective for ADHD.
The best scientific evidence available to date strongly suggests that ADHD is a legitimate medical disorder. Current methods of diagnosis are reasonably reliable, valid, and are semi-objective, but still dependent on clinical judgment. The evidence also suggests that ADHD may be a biological disease of the brain, which is a plausible hypothesis supported by the latest in neurobiological research. Current methods of treatment have also been demonstrated to be safe and effective.
Areas of future research should include further investigation into the possible biological cause of ADHD, and confirmation and identification of a possible genetic link. Related to this is the potential benefit of developing a biological marker that can be used to diagnose ADHD. Studies are required to examine the long term effects of stimulant medication and to continue to explore other options for treatment.
Criticisms of the legitimacy of ADHD are largely based upon a misunderstanding of the nature of clinical medicine, or upon an extreme reluctance to accept symptoms of mood, thought, or behavior as a basis for describing a disease or disorder. All such criticism can be applied to a broad spectrum of non-controversial medical and psychiatric diagnoses.
In summary, while there remains room for some reasonable skepticism, many critics of ADHD have slipped beyond the fuzzy border that separates skepticism from dedicated denial.
1) Shermer M. Why People Believe Weird Things. W. H. Freeman and Company, New York, 1997.
2) Ryley R. Facts and Falsehoods in Psychiatry and Mental Health. http://members.nbci.com/_XMCM/psych_books/errors_frame.htm [cited August 2001]
3) Diagnosis and Treatment of Attention Deficit Hyperactivity Disorder. NIH Consensus Statement Online 1998 Nov 16-18; [cited 2001, August 16]; 16(2):1-37
4) Biederman J. Wilens T. Mick E. Spencer T. Faraone SV. Pharmacotherapy of attention-deficit/hyperactivity disorder reduces risk for substance use disorder. Pediatrics 104(2):e20, 1999 Aug.
5) Clinical practice guideline: diagnosis and evaluation of the child with attention-deficit/hyperactivity disorder. American Academy of Pediatrics. Pediatrics 105(5):1158-70, 2000 May.
6) Scahill L., Schwab-Stone M. Epidemiology of ADHD in school-age children. Child & Adolescent Psychiatric Clinics of North America 9(3):541-55, vii, 2000 Jul.
7) T. Sigi Hale, Ahmad R. Hariri, James T. McCracken. Attention-deficit/hyperactivity disorder: Perspectives from neuroimaging. Mental Retardation and Developmental Disabilities Research Reviews, Volume: 6, Issue: 3, Date: 2000, Pages: 214-219
8) Lazzaro I. Gordon E. Whitmount S. Meares R. Clarke S. The modulation of late component event related potentials by pre-stimulus EEG theta activity in ADHD. International Journal of Neuroscience. 107(3-4):247-64, 2001 Apr.