Article Text

Download PDFPDF

Attention deficit hyperactivity disorder in adulthood
  1. B Toone
  1. Correspondence to:
 Dr B Toone
 King’s College London, Denmark Hill, London SE5 9RS, UK;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Attention deficit disorder may persist into adult life and have a genetic basis

A behavioural syndrome of childhood characterised by overactivity, excitability, and explosive behaviour has been recognised since the turn of the last century and is now thought to be, at certain ages, the most common disorder of childhood.1 The disorder is now recognised in the terminology of ICD-102 as “hyperkinetic disorder”, and in DSM-IV3 as “attention deficit hyperactivity disorder” (AD/HD). The cardinal features of overactivity and impaired attention should be pervasive—that is, evident in more than one situation, and observable before the age of six (ICD-10) or seven years (DSM-IV), though they are usually evident well before that. DSM-IV subclassifies AD/HD into predominantly inattentive, predominantly hyperactive–impulsive, and combined subtypes; ICD-10 insists on the presence of both inattention and hyperactivity for a diagnosis of hyperkinetic disorder.


The prevalence varies according to which set of diagnostic criteria is used, DSM-IV or the more restrictive ICD-10. Using the latter, prevalence in the United Kingdom is 1.5% among seven year old boys in British inner cities4; in the USA using DSM-IV criteria, this may rise as high as 9.5%.5 Sex differences are far more consistently reported, and boys regularly exceed girls in a ratio of 2.5:1. There is a substantial comorbidity with conduct disorder and oppositional deviant disorder, although the two conditions are not readily separated, and with reading disorder, obsessive compulsive disorder, multiple tic disorder, and Tourette’s syndrome.

The hyperactive child syndrome as it was originally conceived was thought to be a condition of childhood. Experienced clinicians could write: “In later years this syndrome tends to wane spontaneously and disappear. We have not seen it persist in those patients we have followed to adult life.”6 This view was widespread and was formally acknowledged in DSM-II (1968). This perception was altered by the outcomes of prospective studies from Montreal7 and New York.8,9 Weiss et al followed up 61 of an original cohort of 104 hyperactive children first seen between the ages of 6 and 12.7 At an average age of 25.1 years, 61% still reported features of restlessness, poor concentration, and impulsivity, compared with 7% of a matched control group. Twenty three per cent received a DSM-III diagnosis of antisocial personality disorder, against 2.5% in the control group.

Gittelman et al studied 101 white male children aged between 6 and 12 years who were deemed hyperactive.9 Subjects with comorbid conduct disorder were excluded. Nine years later 31% met DSM-III criteria for a diagnosis of attention deficit disorder with hyperactivity (a forerunner of AD/HD)—a diagnosis shared by only 3% of the control group. Twenty seven per cent and 16% of the original cohort, respectively, received diagnoses of antisocial personality disorder and substance abuse, compared with 8% and 3% of the controls, but the excess was largely accounted for by those in whom the AD/HD symptoms had persisted. Drug abuse was accounted for almost exclusively by those with antisocial personality disorder.

In an extension of the cohort follow up period to 16 years, and thus into the middle of the third decade,9 the proportion meeting criteria for DSM-IIIR diagnosis of AD/HD had fallen to 8%, against 1% in the control group. The prevalence of antisocial personality disorder, though still excessive, had fallen and was no longer linked to the retention of the AD/HD diagnosis. The probands had completed fewer years of schooling and were of a lower socioeconomic status.

A community based study,10 which included some subjects who were comorbid for conduct disorder, was continued into late adolescence. The subjects who were comorbid for hyperactivity and conduct disorder had a higher prevalence of persistent AD/HD (29%) than those who had hyperactivity (20%) or conduct disorder (11%) alone; but if ICD-10 rather than DSM-IIIR criteria were used, the prevalence figures were approximately halved. Data from these prospective US and UK cohort series suggest that AD/HD may persist into the third decade and probably beyond. The reported prevalence will vary according to selection criteria used—in particular, whether the study is community or clinic based, whether comorbid conduct disorder is included, and which classification system is used.

Relatively unbiased cross sectional data have only recently been reported.11 Applicants for driving licence renewal were invited to participate, and diagnosis was based on self report. Persistent AD/HD symptoms were reported in 4.7%, declining with age. Variables that might predict persistence of symptoms have received surprisingly little attention. Impulsivity/hyperactivity rather than inattention, and comorbidity for conduct disorder, each predict persistence, as does the presence of a family history of AD/HD.


There are many pitfalls to the diagnostic process. There are no confirmatory laboratory based investigations. Clinical observations may be quantified by standardised instruments but ultimately diagnosis is a matter for clinical judgement. Even a positive response to treatment cannot be taken to provide unqualified support for the diagnosis, as cerebral stimulants may enhance cognitive functions, particularly those mediated by the prefrontal cortex, in normal volunteers.12 The diagnosis of childhood AD/HD rests on relatively secure foundations. The child is constantly supervised, at home and in the school, and behavioural observations may be highly structured and detailed. The young adult frequently lives alone and is unemployed. Psychiatric illness in adults usually presents as a change in mental state, from normal to abnormal; the assessment of enduring characterological traits that have a dimensional rather than a categorical distribution is more challenging. The assessment of the adult who presents for the first time with possible AD/HD features is considerably more difficult than the continuing evaluation of the patient in whom the diagnosis was made during childhood. The concept of adult AD/HD has received considerable media attention in the United Kingdom, but few clinicians feel competent to diagnose it. Ninety per cent of referrals to a specialist clinic were initiated by the patient or family, with the medical referer playing a passive rather than a gate keeping role.13 The DSM-IV criteria are detailed and explicit and can be employed as a framework for a structured interview, using the AD/HD portion of the adult interview14 in which the original wording has been adapted to adult usage. This is not without its limitations. The cardinal features of AD/HD are ubiquitous in adult psychiatric practice, and the issue of symptom overlap has never been satisfactorily resolved.15 If consideration is given to the presentation and development of symptoms over time, a distinction can be more readily made. In a recent history of patients referred to a UK clinic,13 the majority sought a diagnosis of AD/HD and were dissatisfied not to get it. The risk of self rater bias should not be underestimated. Disruptive childhood behaviour may have other explanations.

AD/HD is a condition where the features are apparent, though not necessarily detected, during early childhood. It is essential when considering the diagnosis for the first time in adult life to remain cognisant of this and to seek confirmation, usually through a parental account of childhood behaviour, where possible supplemented by contemporaneous evidence—for example, school reports, educational psychology reports, and so on. Several rating scales have been devised in attempts to quantify childhood AD/HD behavioural traits, such as Conner’s abbreviated rating scale.16 These may be completed by the patient, but parental report is more reliable and should be sought wherever possible. Although quantitative psychometric testing may be used in the assessment of childhood AD/HD, its contribution as an aid to diagnosis has yet to be established.17 The position with adults is even less secure. Tasks such as the continuous performance test and the matching familiar figures test purportedly measure, respectively, sustained attention and impulsivity, but there is only limited and inconsistent evidence that they discriminate between AD/HD subjects and normal controls, even less so other psychiatric disorders. Improvements in the continuous performance test following the introduction of methylphenidate18 and the Stroop test after the introduction of tomoxetine, a noradrenergic receptor inhibitor,19 have been reported, but most attempts to demonstrate treatment effects in psychometric performance have been unsuccessful. The tests that are currently used in the assessment of adult AD/HD were not designed with that aim in mind; they are used because they are standardised tests that psychometrists are familiar with—hence their relatively poor discriminatory value.


The most effective element in the management of AD/HD is pharmacological. The recently published results of the multimodel treatment study of children with AD/HD20 emphasise the essentially adjunctive role of psychosocial treatment. The treatment of the adult does not differ from that of the child other than in details such as the dose adjustment for greater body weight.

The class of drugs described as cerebral stimulants remains the cornerstone of drug treatment. Drugs of this class, of which methylphenidate and dexamphetamine are the most widely prescribed, act principally by blocking reuptake of monoamines, particularly dopamine and noradrenaline. Over a dozen controlled trials have now been conducted in adults and all but one showed a significant benefit from the active drug.21 The mean response rate of 60% across trials is less than in children (60–70%) or adolescents (75%). This may reflect the uncertainties of diagnosis, comorbidity, non-compliance, or poorly developed measures of responsiveness. It may also be a consequence of inadequate dosage: in one study a dose of 1 mg/kg body weight was adopted and a response rate of 78% reported.22

There are few absolute contraindications to the use of cerebral stimulants in adults. These include the presence of psychotic symptoms or a history of substance abuse. It is best to treat comorbid conditions, particularly affective disorders, first, before proceeding to address AD/HD. Cerebral stimulants are well tolerated in adults. Sustained release preparations have been developed commercially and Concerta—the sustained release form of methylphenidate—is available in the United Kingdom, thus obviating the need for twice and thrice daily dosing.

The commonest symptoms in order of frequency of occurrence are: insomnia, edginess, diminished appetite, dysphoria, and headache.21 Most remit with time or a reduction in dosage. Psychosis has been reported in children but not adults. The potential for iatrogenic substance abuse exists, but has not been reported. Those patients who fail to respond to cerebral stimulants may respond to tricyclic antidepressants such as desipramine,23 tomoxetine,15 or venlafaxine.24 Other drug treatments, for example the α2a agonists such as clonidine and the β blockers, have been used but their value is yet to be established. The selective serotonin reuptake inhibitors (SSRIs) have not been systematically evaluated in adults or in children, but there appears to be a consensual view based on clinical experience that they are ineffectual.25 The more sedating neuroleptics may have a calming effect, but their side effects render them unsuitable for any role in the management of AD/HD.

It would seem that several different classes of drug may modify AD/HD symptoms. It is not yet possible to locate a common site of action with any precision, but the more effective treatments act on monoaminergic pathways, notably those involving dopamine and noradrenaline. Behavioural, cognitive, and psycho-educational treatments play an important, if subsidiary, role in the management of childhood AD/HD.20 A combination of drug treatment and behavioural strategies was superior to drug treatment alone. It would be surprising if the same were not also true of adults with AD/HD but very little has been published on this.


Although a precise understanding of the aetiology of AD/HD remains elusive, the directions along which the inquiry should proceed have been clearly indicated. AD/HD is known to aggregate within families, and twin studies that examine the relative importance of genetic and environmental influences have consistently shown an important genetic influence in childhood, with estimates of additive genetic variance of the order of 60–90%. The risk to first degree relatives of probands with uncomplicated childhood AD/HD is low (3.2% for siblings; 5.6% for parents) but rises when AD/HD is combined with conduct disorder or when it is persistent. In the presence of each additional risk factor the risk increases to 25.9% and 22%, respectively. Studies in molecular genetics have adopted the candidate gene strategy, relying upon a specific gene hypothesis suggested by the known pathophysiology of AD/HD, rather than positional cloning, a hypothesis-free approach depending on genome scanning. Drug response outcomes and neuroimaging findings suggest abnormalities in the genetic regulation of monoamine neurotransmission, and molecular research has focused on this area. Two genetic associations (DRD4 and DAT1) replicate across multiple studies; in others (the serotonin 1B receptor, the dopamine D5 receptor, and the synaptosomal associated25) the current evidence is less certain. The genetic contribution is likely to be one of multiple small effects: for example, one of the studies that confirm the significance of the DAT1 allele estimated that it explained only 3.6% of the variance in hyperactivity/impulsivity symptoms and 1.1% in inattention symptoms.

How is the genetic contribution to AD/HD mediated? A neurobiological model, albeit still somewhat hazy, is beginning to take shape. Much adult ADH/HD symptomatology—for example, impaired organisational skills, weak impulse control—suggests executive dysfunction. The results of neuropsychological investigations are broadly consistent, but no more, with these clinical observations. Attentional measures such as the continuous performance test and the Stroop test reliably differentiate between AD/HD subjects and normal controls, but not between AD/HD subjects and psychiatric controls.26 The clear superiority of the cerebral stimulants over other classes of drugs implicates the monoaminergic system, particularly dopamine and noradrenaline. Structural imaging, more specifically cranial magnetic resonance imaging studies, have been consistent in demonstrating reduced area or volume of prefrontal, basal ganglia, and cerebellar vermis structures. (In childhood an AD/HD syndrome may appear following severe closed head injury,27 particularly if the right putamen is involved.28) The outcomes of functional neuroimaging studies are less easily summarised, not least because of the multiplicity of imaging techniques and diverse methodology. Those studies that have used inhibitory paradigms—for example, the Gonogo29 and Stop30 tasks—find reduced or abnormal activation patterns in prefrontal and striatal structures.18 Fluorodopa was used to label catecholamine terminals to demonstrate a reduction in prefrontal areas in AD/HD adults31 and increase in midbrain in adolescents.32 These results, while not easy to piece together, lend support to the concept that catecholamine dysregulation is central to AD/HD pathophysiology. There is certainly evidence to suggest that the therapeutic actions of the cerebral stimulants are located in the prefrontal cortex, and a prevailing view is that many AD/HD symptoms may arise from insufficient catecholamine receptor stimulation in that region. This could arise in several ways—for example, reduced dopamine or noradrenaline prefrontal cortex innervation, presynaptic receptors that are overly responsive to dopamine, noradrenaline stimulation, and so on. The demonstration of increased binding of selected dopamine transporter ligands in AD/HD adults33,34 and the reduction in binding following treatment with methylphenidate,34 taken in conjunction with molecular genetic studies, provides an example of how catecholamine dysregulation can act, though it is highly unlikely that this occurs as a result of one dysfunctional locus. The probability is that, over time and perhaps through comprehensive genome mapping, multiple sites will be identified.

Attention deficit disorder may persist into adult life and have a genetic basis