Attention-Deficit Hyperactivity Disorder (ADHD)

Kytja K. S. Voeller, MD


J Child Neurol. 2004;19(10):798-814. 

In This Article

Diagnostic Criteria

ADHD is a relatively common brain disorder, affecting as many as 10% of school-aged children. Currently, there is no "gold standard" or laboratory test to confirm the diagnosis of ADHD. Diagnosis is currently based on criteria from the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), which lists nine behavioral characteristics (essentially "word pictures") for the inattentive type and nine behavioral characteristics for the hyperactive-impulsive type.[1] To be diagnosed with ADHD, at least six of the nine inattentive and/or hyperactive/impulsive criteria should be observed, both at home and at school. To meet the diagnostic criteria, the symptoms should significantly impair performance, be inconsistent with the child's developmental level, and be present for at least 6 months, and some symptoms should be apparent before age 7 years (although this last criterion has been the subject of considerable controversy). However, determining whether a patient meets a specific behavioral criterion (eg, "often has difficulty organizing tasks and activities") depends on the observer's interpretation of terms such as "often," "difficulty," and "organizing." Parents (especially those who are divorced or separated) might disagree, with one parent seeing the child as active, impulsive, and disorganized and the other parent seeing the child as "just having lots of energy." Sometimes parents see the child as having significant attentional and organizational problems, whereas the teacher perceives the child as "not trying" or vice versa. Nonetheless, there is often surprisingly good agreement between observers.[2]

In the last two decades, there has been a tremendous expansion of our knowledge about the neurobiology and natural history of ADHD. In the DSM-IV , ADHD is conceptualized as falling into one of three categories: predominantly inattentive, predominantly hyperactive or impulsive, or combined. There is a continuing debate as to whether the diagnosis should be categorical (ie, making tight distinctions between those children who meet criteria for specific numbers of symptoms) versus dimensional (ie, viewing ADHD as falling on a continuum). For example, in the categorical DSM-IV approach, a child who met the criteria for six of nine symptoms would be classified as having the disorder but would not be so classified if he or she had only five of nine symptoms. It is likely that many of these debates will be resolved once there is greater understanding of the neurobiology and molecular genetics of this disorder.

Some children with ADHD seem underaroused, have difficulty initiating, are drowsy, and daydream, with variable attentiveness, and have been described as the "sluggish cognitive tempo" type.[3] Although these symptoms had very high positive predictive power in the DSM-IV field trials, they did not have high negative predictive power (ie, although the behaviors were strongly associated with inattention, their absence did not predict the absence of inattention). As a result, the sluggish cognitive tempo type of ADHD was not included in the DSM-IV diagnostic criteria.[4] Since the publication of the DSM-IV in 1994, information from neurophysiologic and genetic studies has provided support for this subtype.

Based on a considerable body of research, the DSM-IV subtypes have been shown to differ in terms of age at onset, ratio of boys to girls, pattern of comorbid psychiatric conditions, and neuropsychologic learning disability profiles.

Based on the DSM-IV field trials, the symptoms of ADHD are apparent in most cases before the age of 7 years.[5] Although the median age at onset of the first symptom was 1 year, the median age at impairment did not occur until 3.5 years. The hyperactive or impulsive type emerges earlier than the combined type and significantly earlier than the inattentive type. The mean age at onset of the hyperactive or impulsive type was 4.21 years and that of the combined type was 4.88 years. Most children with the hyperactive or impulsive and combined types of ADHD were considered impaired before age 7 years (98% and 82%, respectively). The inattentive type followed a different trajectory; the mean age at onset did not occur until 6.13 years, and only 57% became symptomatic before age 7 years. Only 85% of children with the inattentive-type ADHD manifested at least one symptom before they were 7 years old in comparison with 100% with the hyperactive or impulsive type and 96% with the combined type. Stated somewhat differently, 43% of the inattentive type met the criteria for impairment after age 7 years. In a large epidemiologic sample (the Great Smoky Mountains Study), a similar pattern was noted, with 25% of the youths with inattentive symptoms reporting onset after age 7 years, in contrast to 13% of the combined subtype, and 8% of the hyperactive or impulsive subtype reporting onset before age 7 years.[6] In our clinic, we have encountered adults who had not been diagnosed with ADHD until they were well into their forties, after having struggled with their symptoms for much of their lives. Many sought care when their own children were diagnosed. Thus, although ADHD is a childhood-onset disorder, requiring a precise age at onset, particularly for the inattentive type, is probably not warranted.[7]

It is apparent from the above that ADHD symptoms become apparent in the preschool child. In fact, given the genetic nature of the disorder (discussed below), it is not surprising that manifestations of ADHD might be apparent in infancy. Infants with the dopamine receptor 4 (D4) polymorphism that is associated in older children and adults with ADHD are reported to have disorganized attachment behavior.[8]

Older children and adults with ADHD have been shown to have executive function deficits (discussed below and in the article by Powell and Voeller in this issue[9]). Preschool children with ADHD, when appropriately tested, have also been shown to have executive function deficits. In a study of 156 children ranging in age from 3 to 5½ years, a cluster of executive function deficits—working memory, planning, set shifting and difficulty tolerating a delay in gratification—together accounted for 70 to 80% of the variance. ADHD symptoms were correlated in a linear fashion with both of these factors, which were also correlated with age, IQ, and conduct problems.[10] Goldman et al. described an attentional task for 2-year-old children, which is correlated with heightened right frontal evoked potential responses and has good interobserver reliability.[11] Using standard clinical techniques, the diagnosis of ADHD, particularly the hyperactive or impulsive or combined type, can be reliably made in many children by the time they are in preschool.[12] It is important to identify and treat children with ADHD as early as possible because the parents of preschool children with ADHD find them very difficult to manage and extremely stressful.[13] Early treatment should also ameliorate later impairment in social behaviors, as well as poor school performance.

Some researchers have suggested that ADHD follows a developmental pathway in which the earliest manifestations are hyperactivity or impulsivity, which then develops into ADHD of the combined type during elementary school, and then evolves into the inattentive type of ADHD after the hyperactivity or impulsivity symptoms fade and academic and social demands for autonomous functioning increase.[12,14,15]

There are more boys than girls diagnosed with ADHD. In surveys dealing with children referred to clinics, the ratio of boys to girls varies from 6:1 to 12:1. In epidemiologic samples, the male-to-female prevalence ratio is much lower, 3:1, suggesting that ADHD in girls tends to be underdiagnosed. On the one hand, girls do not manifest disruptive behaviors to the extent seen in boys; girls with ADHD have half of the rates of conduct disorder and oppositional defiant disorder but are much more likely to have significant social problems.[16] Compared with boys with ADHD, they manifest more emotional distress, have higher rates of depression and anxiety, are highly vulnerable to stress, and have poor self-esteem and a limited sense of control. However, girls show an equivalent degree of prefrontal executive function impairment.[17] On the other hand, compared with unaffected girls, girls with ADHD are significantly impaired on the Global Assessment of Functioning Scale, as well as in cognitive functioning and academic performance. They have higher rates of disruptive behavior disorders, are vulnerable to alcohol and drug dependence, and are at risk of academic failure.[18,19]

It is rare to encounter a child with "pure" ADHD without other emotional or learning problems because ADHD is associated with an extremely high rate of comorbid psychiatric disorders and is usually accompanied by a learning disability. Relatives of children with ADHD[20,21] are also at higher risk of neuropsychiatric disorders than relatives in the control families.[20,22] Conduct disorder, oppositional defiant disorder, major affective disorder (depression or bipolar disorder), anxiety disorder, including obsessive-compulsive disorder, and Tourette syndrome are all such comorbidities. Teenagers with ADHD, particularly untreated ADHD, are at risk for drug and alcohol abuse. In addition, many individuals with mental retardation and autistic spectrum disorders (ie, pervasive developmental disorder, autistic disorder, Asperger's syndrome, and nonverbal learning disability) also often have associated ADHD. Language disorders are frequently associated with ADHD.[23] In one report, 45% of children with ADHD had at least one element of language impairment, and children with both specific language impairment and ADHD appeared to have greater difficulty with verbal short-term memory.[24] (Also see the article by Sundheim and Voeller in this issue for further discussion of this association.[25]) Language impairment in ADHD has been considered by some to reflect a common underlying prefrontal executive function deficit.[26]

Learning disabilities (dyslexia and dyscalculia in particular) are frequently associated with ADHD.[27,28] In the Remediation of Dyslexia study, we observed that of the 60 children selected because of severe phonologic awareness deficits, 80% met the criteria for ADHD.[29,30] Boys with ADHD and learning disability tended to have more serious executive function deficits than boys who were not learning disabled.[31]

Motor incoordination is often associated with ADHD[32] and can be an early and prominent feature in the preschool child who will develop ADHD symptoms. Kadesjo and Gillberg in Scandinavia pointed out the syndromic nature of this combination—deficits in attention, motor control, and perception (DAMP).[33]

Neuropsychologic studies on children with ADHD have revealed a pattern of cognitive deficits consistent with prefrontal executive function deficits: inattention, difficulty with self-regulation, response inhibition deficits (impulsivity), restlessness or hyperactivity, or apathy in some cases.[34] However, depending on a number of methodologic variables, including subject and control group selection and the presence of certain comorbid disorders, the results have been somewhat variable.

In a study drawn from a community sample, children were examined to find out if the inattentive and hyperactive or impulsive dimensions and subtypes based on the DSM-IV criteria were associated with different neuropsychologic profiles. The researchers found that the inattention dimension, not the hyperactive or impulsive dimension, was associated with significant neuropsychologic impairment. They suggested that both the inattentive and hyperactive or impulsive symptoms can relate to a latent trait, "disinhibition," with the inattentive symptoms referring more to the cognitive aspects and the hyperactive or inattentive symptoms relating to behavioral aspects.[35] The inattentive and hyperactive or impulsive types also differ in heritability. In a study of twins with ADHD, a high level of inattention was heritable regardless of the degree of hyperactivity and impulsivity. On the other hand, high levels of hyperactivity and impulsivity were tightly linked to the number of inattention symptoms manifested by the twin with ADHD.[36] This suggests the intriguing possibility that the essential dysfunction in ADHD involves inattention and disorganization rather than hyperactivity or impulsivity, as has been believed, and that hyperactivity or impulsivity might result from some other factor.

Behaviorally, ADHD is a disorder of self-regulation, which implicates some sort of dysfunction of the frontal-subcortical system.[34,37,38,39] Many magnetic resonance imaging (MRI) morphometric studies (ie, studies involving measurements of various brain regions) have been conducted using different techniques and different populations (including subjects from different regions of the globe). These studies have identified relatively consistent differences in the brains of children with ADHD compared with those of normal controls. A large, well-designed longitudinal study involving 544 MRIs from children with ADHD and age- and sex-matched controls has provided evidence that ADHD is associated with an atypical pattern of brain development that appears in early childhood.[40] The major findings of these studies are summarized as follows:

  1. Total cerebral volume is smaller in individuals with ADHD and in controls. There is a small but significant reduction (on the order of 5%) in mean total cerebral volume or intracranial volume.[41,42,43,44,45,46,47] In one study comparing boys with ADHD, their unaffected male siblings, and matched controls, the subjects with ADHD had a significant (4%) reduction in intracranial volume. Their unaffected siblings had a 3.4% reduction compared with controls (a statistical trend). Cortical right prefrontal gray matter and left occipital gray and white matter were reduced in the subjects with ADHD and their siblings.[42] This suggests that changes in cerebral volume need to reach a certain crucial level before they become obviously symptomatic. Moreover, this study strongly supports the genetic basis of ADHD.

  2. Frontal lobe volume is smaller in persons with ADHD. Brain regions involved in self-regulation (executive function) show differences from those of controls. In most studies, the frontal lobes or subregions of the frontal lobes were found to be smaller in subjects with ADHD than in controls.[40,41,45,47,48,49] In one study, the inferior portions of dorsal prefrontal cortices and anterior temporal cortices, bilaterally, were reduced in subjects with ADHD.[49]

  3. Various regions of the basal ganglia, particularly the caudate nucleus, have been reported to be smaller in children with ADHD compared with controls.[40,41,43,50] Studies on normal individuals have shown that the caudate decreases in size as the child matures (a manifestation of the normal "pruning" of neurons seen in many parts of the brain during development). Children with ADHD start out with smaller caudate nuclei than controls, and with maturation, there is a further decrease in size. As a result, any difference in size between children with ADHD and controls becomes less apparent with increasing age.[40,41,51] (This might explain the variability in size observed in different studies because the age of the subjects varied considerably across these studies.) Other regions of the basal ganglia have also been reported to be reduced in volume in subjects with ADHD relative to controls.[52,53]

  4. Right hemisphere structures are affected more than left hemisphere structures. In normal child and adult populations, the right frontal area is larger than the left frontal area. Given the important role that the right hemisphere plays in regulating attention and the deficits seen in ADHD, it would not be surprising to observe reduction in right frontal lobe volume. This was not a consistent finding, but it was noted in a number of studies.[40,41,42,43,44,54] In some studies, a decrease in the right frontal gray-matter volume was noted, or changes in volumes of certain subcortical structures were more prominent on the right.[42,53] It is possible that the reduction in size is due to the reduction of global brain volume, as suggested by Castellanos et al.[40] In normal adults, the right caudate is larger than the left caudate.[55,56,57] However, based on the large National Institute of Mental Health study of children with ADHD, the right caudate nucleus is smaller than the left caudate nucleus.[58] This asymmetry was not necessarily observed in all studies, but they generally involved many fewer children and were not longitudinal.[43,46,49]

  5. There is a relative decrease in the size of the cerebellum. The cerebellum also participates in the regulation of executive function as a result of its reciprocal connections to the prefrontal cortex.[59] The decreased size of the cerebellum in children with ADHD was initially described by Castellanos et al.[41] and has been corroborated in a number of other studies.[40,43,44,60,61]

  6. A number of studies reported a reduction in the area of the anterior[42,44,62,63,64] or posterior corpus callosum.[65] However, in the large National Institute of Mental Health study, this was not confirmed.[41]

It is worth noting that treatment with psychostimulants was not responsible for the reduction in various brain areas because these findings were also noted in children who were drug naive. Interestingly, children on psychostimulant treatment actually had somewhat greater white-matter volumes than those who had not been treated. [40]

In summary, there is now much research suggesting that, when carefully examined, groups of children with ADHD have small but significant reductions in total brain volume and in the various regions of the brain that are involved in the regulation of attention and impulsivity. This would suggest that the behaviors seen in children with ADHD are not simply the result of environmental factors or some sort of distortion of perception on the part of parents and teachers, but rather a very real brain dysfunction.

Functional neuroimaging studies have been used to study individuals with ADHD and controls and are consistent with the morphometric findings. These studies also revealed dysfunction of the prefrontal-subcortical system, often with greater involvement of areas of the right hemisphere.[66,67] [ Note: Functional neuroimaging studies involve a number of different techniques: positron emission tomographic (PET) scans, functional magnetic resonance imaging (fMRI) scans, single photon emission computerized tomography (SPECT), and magnetic resonance spectroscopy (MRS). The common feature of these studies is that they enable us to examine the brain while it is performing various cognitive or behavioral tasks and provides a remarkable window in the understanding of brain function.] A study of adolescents with ADHD using positron emission tomography (PET) with [[18] F]-fluorodeoxyglucose revealed that global cerebral glucose metabolism in the adolescent girls with ADHD was 15% lower than in the control girls and 19.6% lower than in boys with ADHD.[68] Brain regions in which this lower activity was observed were the right frontal premotor cortex and right temporal cortex. Activity was decreased bilaterally in the posterior putamen and middle cingulate cortex. These findings were not confirmed on a subsequent study, but it was noted that the degree of sexual maturation was probably a variable that had not been taken into consideration.[69] A study of adults (18.1 to 50.8 years of age) with ADHD by the same investigators demonstrated that global cerebral glucose metabolism was reduced in women with ADHD but not in men with ADHD or in control men or women. Women with ADHD demonstrated better performance on the auditory attention task with increasing age. These findings suggest a complicated interaction between gender, age, and hormonal effects.[70] In another PET study of adolescents in the age range of 13 to 14 years, in which the effects of sexual maturation were controlled, subjects with ADHD had a higher accumulation of dopa decarboxylase (indicating a high level of dopamine synthesis) in the right midbrain than controls. Although this study was not without methodologic problems (it involved a small number of subjects, the "controls" were siblings of the ADHD subjects, and the findings were significant only when computed without adjustments for multiple statistical comparisons), it still suggests that the dopaminergic system is dysfunctional in persons with ADHD.

When children are asked to perform a task that places demands on the frontal executive system, those with ADHD have atypical patterns of activation. In one study, children with ADHD and controls were studied using functional MRI during a go/no-go task. [ Note: Go/no-go tests involve establishing a pattern of response to a specific "go" signal and then inhibiting the response when a "no-go" signal is presented. It is one test of executive function in that it requires the ability to inhibit an established pattern of behavior. One example involves making two taps with a hand when the examiner makes one, and then when the examiner makes one tap, the subject makes none. Children have greater difficulty with these tasks than adults, but the child with ADHD has much greater difficulty than other children of the same age.] In general, functional MRIs on children performing tasks that demand executive function control have somewhat different patterns of activation than are seen in those of adults.[71] However, children with ADHD do not activate frontostriatal networks to the same extent seen in the children without ADHD but, rather, manifest a more diffuse activation pattern than was seen in controls, suggesting that the development of frontostriatal circuits was delayed in children with ADHD.[72] In another functional MRI study involving children with ADHD and controls performing two somewhat different types of go/no-go tasks, children with ADHD made more errors than controls.[73] In one task, children with ADHD activated frontal areas to a greater extent than controls. Although this is not consistent with the findings in other studies, it is possible that the task required the subjects with ADHD to exert more effort than controls. (In this study, other brain regions were not examined so that there was no opportunity to see the diffuse activation pattern described in the Durston et al. study.[72]) After receiving methylphenidate (Ritalin), both groups of children made fewer errors, with a highly significant improvement in the ADHD group. Methylphenidate increased frontal activation in both groups and increased striatal activation in the children with ADHD but decreased it in the controls.[73]

Some SPECT studies have identified decreased activity involving the temporal lobe and cerebellum in some children with ADHD.[74] This would support the observation that the dysfunction in ADHD involves not only the frontal-subcortical circuits but also the integration of temporal lobe and cerebellar function in emotion, cognition, and motor planning.[75]

A decision-making gambling task developed for patients with prefrontal deficits was administered to adults with ADHD. Subjects were required to choose between immediate rewards with the risk of high long-term losses and lower immediate gains with lower long-term losses.[76] The ability of adults with ADHD to tolerate delays in gratification was studied using PET while they performed the gambling task. A control task was also employed. Adults with ADHD did not activate the prefrontal cortex during the decision-making process to the same extent as did the controls and did not activate the anterior cingulate and hippocampus, which are involved in emotional arousal and memory. However, the subjects with ADHD activated the posterior right anterior cingulate more than the controls.[77]

In summary, neuroimaging studies reveal that children and adults with ADHD activate frontal subcortical structures to a lesser extent than control subjects. Although the pattern of activation in children with ADHD is somewhat more diffuse, they, like adults with ADHD, do not activate areas involved with emotion and memory to the same extent that controls do. This is consistent with the observed difficulty that these individuals have in motivation and arousal.

Electroencephalographic (EEG) studies of children with ADHD reveal an excess of slow-wave (theta) activity consistent with decreased alertness and underarousal. [ Note: EEGs are "brain wave" studies that record electrical activity of the brain by means of electrodes to the scalp. Different patterns are noted in sleep and wakefulness. Increased slowing during wakefulness is consistent with a lower level of alertness or disturbance of cerebral function.] These EEG patterns correctly classify over 90% of children with ADHD and normal controls.[78] However, although there are clear-cut differences between the EEG patterns of children with ADHD and those of controls, there is enough heterogeneity in the ADHD group to limit the diagnostic efficacy of this technology.[79] Comparing the EEG patterns of children with ADHD and controls, group differences were found in the mean frequency of the total EEG, as well as the specific amounts of activities of different frequencies (theta, alpha, and beta), the ratios of these different frequencies, and the coherence patterns across the three groups.[79,80] These EEG patterns suggest reduced cortical differentiation and specialization in ADHD, more prominently in children with the hyperactive or impulsive type than in those with the inattentive type. Moreover, children with the inattentive type of EEG were found to have two different EEG patterns, one consistent with hypoarousal (reminiscent of the sluggish cognitive tempo type described by Lahey et al.[3]) and one consistent with a maturational lag.[81]

In summary, quantitative EEG patterns appear to demonstrate differences in children with ADHD and those without ADHD. However, diagnostic accuracy is not much better than the clinical assessment and requires specialized equipment and technical expertise. Given the possibility that there can be different EEG patterns seen in children with ADHD, this technique appears to have limited application at this time; however, it might be useful in determining the response to medication[82] and possibly will be more meaningful once ADHD genetics are unraveled.

In summary, electrophysiologic studies of children with ADHD reveal atypical brain wave patterns, which suggest dysregulation of arousal and attention.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.