In March of 1902 Dr. George Still stood before the Royal College of Physicians, in London, and described some children he had observed—mostly boys—who seemed to him restless, passionate, and apt to get into trouble. The children were suffering, he declared, from "an abnormal defect of moral control."
Despite his invocation of morality, Still's lecture is often billed as the first recorded discussion of hyperactivity. Thousands of articles on the subject have been published in professional journals since then, the great majority of them within the past two decades. A good proportion of these papers begin by citing the pervasiveness of the disorder. An opening sentence such as "Hyperactivity is the single most prevalent childhood behavioral problem" is usually regarded as sufficient, because the readers of these journals are already convinced that, on average, at least one hyperactive child sits in every elementary school classroom in the United States.
The actual estimates vary, however, and not by a little. If a psychiatrist says that about three percent or 10 percent or between one and five percent of elementary school children are hyperactive, this is simply a rough average of studies whose findings differ dramatically. One series of papers estimated the rate of hyperactivity at 10 to 20 percent. A California survey put it at precisely 1.19 percent. A nationally recognized expert says without hesitation that it is six percent. The one thing researchers generally agree on is that among children labeled hyperactive, boys outnumber girls by at least four to one.
The disparities can be explained to some extent by the varying stringency of the criteria that are applied, and the assumptions guiding those applying them. According to the latest guidelines for diagnosing what is now officially called attention-deficit hyperactivity disorder (ADHD), the problem must have been noticed before age seven, must persist for at least six months, and must include any eight of fourteen symptoms, among which are the following: the child is easily distracted by extraneous stimuli; has difficulty sustaining attention, following through on instructions, or waiting his or her turn in games; and often does such things as talk excessively, fidget with hands or feet, squirm while sitting down, lose things, and fail to listen to what is being said to him or her.
Most experts emphasize the importance of teachers' observations, which are often quantified on a rating scale that was developed by Keith Conners, a psychologist, in the late 1960s. When that score—or any single judgment—is the sole basis for diagnosis, 10 percent or more of all elementary school children may be labeled hyperactive. But if the observations of others—parents or pediatricians, for example—are also taken into account, then the prevalence of the disorder can be as low as one percent.
The experience of two Canadian researchers, Nancy J. Cohen and Klaus Minde, is illustrative. They had the teachers of 2,900 kindergartners in one community submit the names of children thought to be hyperactive. The researchers expected, because of the estimate offered by a widely used textbook, to find that four to 10 percent would be referred. At first their procedure yielded sixty-three names. But when they looked more closely, they found that most of these children had altogether different psychological problems or else seemed to be suffering from poor nutrition or too little sleep. Only twenty-three children—less than eight tenths of one percent—were left in the hyperactive category after this more rigorous screening.
The wildly divergent estimates of prevalence are disturbing enough in themselves, given that each percentage point stands for hundreds of thousands of children. But they also underscore the fact that different criteria for diagnosis produce different conclusions about whether a particular child will carry the ADHD label and, as a consequence, be required to swallow a drug every day. Most unsettling is a flicker of doubt about the integrity of the diagnosis itself. Can we in fact be confident that any child has a disorder called hyperactivity, or ADHD?
Overwhelmingly, child psychiatrists and psychologists answer in the affirmative. Just because the prevalence of hyperactivity is difficult to pin down, or because we can't be sure a particular child is afflicted with it, doesn't mean the phenomenon isn't real, according to most people in the field. "If you'd ever seen a hyperactive kid, you'd know it," the psychologist Susan Campbell says. "Something's there." But my review of more than a hundred journal articles and book chapters, and also conversations I had with many of the leading researchers in the field, suggests that this assessment may be too sanguine.
First, whether such a distinctive disorder exists is open to question, because each of the symptoms that are supposed to lead to a diagnosis of hyperactivity—restlessness, impulsiveness, and difficulty paying attention—occurs at least as often in children who have entirely different problems, as Cohen and Minde discovered.
Second, the key symptoms often do not appear together. Douglas G. Ullman, a psychologist, and his colleagues have found that children said to be hyperactive do not always turn out to have difficulty paying attention, and vice versa. One's ability to predict that a child will be inattentive because he is restless—or the other way around—"is not much better than if one tossed a coin to decide the matter," the authors concluded.
Third, the procedure for deciding which behaviors belong on that list of fourteen, and also the decision that eight of them (rather than seven or ten) will suffice for a diagnosis of hyperactivity, are arbitrary. These decisions are "made by committee," as Dennis Cantwell, a leading researcher in the field who was himself on such committees, admits.
A score on the Conners Teacher Rating Scale, or any of the other scales used in diagnosis, gives the appearance of scientific precision, as though it were, say, a white-cell count. In reality, the score is nothing more than a numerical value that sums up a particular teacher's subjective judgments about whether a child bounces around too much.
Some theorists have argued that ADHD is actually "heterogeneous"—that it is characteristic not of a single population of hyperactive children but of several distinct subgroups. This has a professional ring to it and seems plausible on its face, but it simply sets the problem back a step. What are the disorders mistakenly collected under the ADHD umbrella? How do we know that they are valid diagnostic categories? One might say that using the word "heterogeneous" tells more about what we don't know than about what we do.
This history of the diagnosis does nothing to allay one's doubts. For many years children with symptoms identical to those that are now considered to add up to hyperactivity were said to have "minimal brain damage." When researchers eventually acknowledged that they had no proof these children's brains were actually damaged, the label was changed to "minimal brain dysfunction." This, in turn, gave way to the diagnosis of "hyperkinetic reaction," which became "attention deficit disorder with (or without) hyperactivity," which became "attention deficit hyperactivity disorder."
These changes—and the latest in the series will surely not be the last—reflect something more than quibbling over labels. They suggest a fundamental disagreement about what, if anything, is behind the labels. "The whole notion has gone through so many metamorphoses as to suggest a catastrophe in terms of conceptual integrity," says Gerald Coles, the author of The Learning Mystique, a critical analysis of what are commonly called learning disabilities. "Rather than moving toward ever greater precision, they're constantly sweeping over the disasters of last year's conception."
A new diagnosis has appeared in every successive revision of the mental-health clinician's bible, the Diagnostic and Statistical Manual of Mental Disorders, or DSM. Whereas DSM II talked about "hyperkinetic reaction," the third edition, published in 1980, switched to "attention deficit disorder": now difficulty paying attention seemed to be the core of the disorder, with excessive activity merely an optional by-product. In 1987, with the publication of the latest revision (DSM III-R), the definers changed their minds again, deciding that insufficient data existed to support the emphasis on attention, and that hyperactivity really was the center of the problem after all.