The evidence-based pharmacological treatment of paediatric ADHD

Abstract

Attention deficit hyperactivity disorder (ADHD) is common in children, adolescents, and adults, with extensive research establishing it as a valid neurobiological disorder. Without intervention, ADHD can result in significant impairment throughout the lifespan for the individuals it afflicts. Fortunately, multiple evidence-based options are available for the treatment of ADHD, including several efficacious pharmacotherapies. The role of medication, including stimulants as well as non-stimulants, is well-documented by an extensive body of literature. Although there may be less enthusiasm for behavioural and other psychosocial interventions as stand-alone treatments for moderate to severe ADHD, they are recommended as first-line treatment for ADHD management in preschool-aged children, for those patients with mild symptoms, and as an adjunct to medication in patients with comorbid disorders or suboptimal responses to pharmacotherapy. When planning treatment for individuals with ADHD, the potential risks associated with the available interventions must be carefully balanced against the risks of not treating, or not treating adequately. The treatment plan must also include ongoing re-assessment of the effectiveness of and the need for continued therapy. Recent practice parameters provide further specific guidance for the evidence-based assessment and treatment of children and adolescents with ADHD.

Introduction

Attention deficit hyperactivity disorder (ADHD) is a neurobiological disorder, affecting significant numbers of children, adolescents, and adults worldwide. Research throughout the past century has established a strong scientific foundation for our current understanding of the aetiology, epidemiology, and treatment of ADHD. The American Medical Association's Council on Scientific Affairs in 1998 stated, ‘Overall, ADHD is one of the best-researched disorders in medicine, and the overall data on its validity are far more compelling than for many medical conditions’ (Goldman et al.1998). The American Academy of Child & Adolescent Psychiatry (AACAP), in their 2007 ADHD Practice Parameters concluded, ‘Although scientists and clinicians debate the best way to diagnose and treat ADHD, there is no debate among competent and well-informed healthcare professionals that ADHD is a valid neurobiological condition that causes significant impairment in those whom it afflicts’ (Pliszka, 2007).

Neuropsychological, neuroimaging, and genetic studies have demonstrated the biological underpinnings of ADHD. These studies have correlated deficits in executive functioning, response inhibition, delay aversion, vigilance, working memory, and planning with specific regions of the brain (Willcutt et al.2005). Structural imaging studies have demonstrated that children with ADHD have significantly smaller brain volumes, on average, than same-aged comparison children (Castellanos & Tannock, 2002; Durston et al.2004; Mostofsky et al.2002), with smaller cerebellar and total cerebral volumes noted (Castellanos et al.2002). In addition, functional imaging has revealed discrete variations in brain activation, specifically in the frontal-striatal cerebellar circuits (Krain & Castellanos, 2006). Family, twin, and more recently, genotyping studies provide further support for the biological basis of ADHD. There is considerable evidence that the principal cause of ADHD is genetic, with an estimated heritability of 76% (Faraone et al.2005). Parents of children with ADHD are 2–8 times more likely to have the disorder themselves, and the risk is similar for siblings of affected children (Faraone & Biederman, 2000).

ADHD prevalence has been conservatively estimated to occur in 3–7% of children (APA, 2000), with other estimates as high as 7–12% (CDC, 2005; Woodruff et al.2004). While most commonly diagnosed between ages 7 and 10 yr, symptom presentation and impairment can often be seen in children as young as age 3 yr (Lavigne et al.1996). Epidemiological studies have shown that 2–6% of preschool children meet diagnostic criteria for ADHD (Angold et al.2000; Lavigne et al.1996). Of those diagnosed with ADHD as children, 60–85% continue to meet criteria for the disorder as adolescents, and as many as 60% continue to experience symptoms as adults (Barkley et al.1990, 2002; Biederman et al.1996; Kessler et al.2005).

A comprehensive differential diagnosis is essential for an accurate evaluation. Behaviours which are characteristic of normal childhood development may be misinterpreted as ADHD if not considered in an age-appropriate context. In addition, developmental disabilities, learning disorders, mental retardation, and hearing or vision impairments, as well as general medical problems such as hyperthyroidism, partial complex seizures, or lead toxicity may mimic ADHD. Several aspects of the core symptoms of inattention, hyperactivity, and impulsivity, can also be indicative of depressive and anxiety disorders, substance abuse, or paediatric bipolar disorder.

The diagnostic criteria for ADHD require the presence of at least 6/9 inattentive symptoms, and/or 6/9 hyperactive-impulsive symptoms, with onset prior to age 7 yr. Symptoms must be developmentally inappropriate and result in clinically significant impairment in social, academic, and/or occupational functioning (APA, 2000). Even preschool children with ADHD are at high risk for academic, social, behavioural and family dysfunction due to the disorder (DuPaul et al.2001), and are more likely to be placed in specialized educational settings (Lahey et al.2004; Lahey et al.1998). These children also have increased rates of accidents and injuries (Lahey et al.1998), aggression (Connor et al.2003), and internalizing symptoms (Cunningham & Boyle, 2002). School-aged children with ADHD as a group have more difficulties with peer interactions, academic tasks, and conflicts with parents than do same-aged peers without ADHD. In addition to ongoing difficulties common to younger children, adolescents have elevated rates of substance use and abuse, motor vehicle accidents, academic and occupational impairments, teen pregnancy, and sexually transmitted diseases (Barkley, 2006).

Nearly two-thirds of children diagnosed with ADHD also have at least one co-occurring psychiatric condition. The Multimodal Treatment Study of Children with ADHD (MTA) consisted of one of the largest and best characterized ADHD populations to date (n=579 children aged 7–9.9 yr), and demonstrated that only 31% of participants had ADHD alone, while 40% also met criteria for oppositional defiant disorder, 38% for anxiety/mood disorders, 14% for conduct disorder, and 11% for tic disorders (MTA Cooperative Group, 1999).

The National Initiative for Children's Healthcare Quality (NICHQ) recommends that children with ADHD and their families receive individualized treatment with ongoing support and education (Bodenheimer et al.2002a, b). They recommend that an effective ADHD management plan for children should generally include parent training, behavioural modification and social-skills training, and school-based interventions. In preschool children, or those with mild symptoms, the AACAP (Pliszka, 2007) and American Academy of Pediatrics (AAP, 2001) recommend a trial of behavioural interventions prior to starting medication. Unfortunately, studies have shown that while behavioural therapies offer some benefit, they may have limited effectiveness as a monotherapy for treating moderate to severe ADHD. In the majority of cases, behavioural interventions may be only one component of a more extensive treatment plan.

The MTA study randomized participants to intensive behavioural therapy, pharmacotherapy with systematically delivered methylphenidate, a combination of the two, or standard community care. The pharmacotherapy and combined treatment groups demonstrated significant improvement, and both were superior to behavioural therapy alone. Interestingly, however, the combined treatment group's response was not significantly better than pharmacotherapy alone for the treatment of core ADHD symptoms. Medication, therefore, appears to have the most significant acute impact on the treatment of ADHD (MTA Cooperative Group, 1999). The addition of behavioural interventions to pharmacotherapy did, however, increase parent and teacher satisfaction with treatment, improved the child's interpersonal relationships, and on average, children receiving behavioural interventions required lower medication doses (MTA Cooperative Group, 1999). A later study of children aged 3–5.5 yr with moderate to severe ADHD, the Preschool ADHD Treatment Study (PATS), demonstrated limited response to behavioural therapy alone, resulting in the majority of children warranting the initiation of pharmacotherapy following treatment with only behavioural intervention (Greenhill et al.2006).

Practice parameters

The AACAP Practice Parameters for ADHD published in 2007 combine short- and long-term empirical evidence with expert opinion from paediatric mental health researchers and clinicians. They offer specific recommendations (Table 1) for a comprehensive treatment plan, potentially consisting of pharmacological and behavioural interventions, and that if pharmacotherapy is indicated, the initial agent selected should be one with Food and Drug Administration (FDA) approval for ADHD. The AACAP further states that if the response to an FDA-approved treatment is robust and normalizes the patient's functioning, medication alone may be sufficient (Pliszka, 2007).

In their 2001 clinical practice guideline for treating ADHD in children, the AAP recommended that the first intervention for the young child with ADHD be behavioural (AAP, 2001). The 2007 AACAP (Pliszka, 2007) parameters indicate that behavioural therapy alone may be appropriate in mild cases of ADHD and should be considered for young children. Additionally, Gleason and colleagues made specific recommendations regarding treatment algorithms for pharmacotherapy in preschool-aged children (Table 2) (Gleason et al.2007). Gleason and colleagues went on to specifically address treatment of preschool-aged patients with ADHD and referenced the PATS study when providing guidance for treating young children with a psychostimulant. The AACAP does note that subjects in PATS were only randomized to pharmacotherapy if they did not demonstrate significant or satisfactory improvement following 10 wk of parent training (Greenhill et al.2006).

What is the first-line treatment for ADHD?

The role of pharmacotherapy (Table 3) as a first-line treatment of ADHD is strongly supported in the literature (Biederman & Spencer, 2008). The stimulant medications have decades of efficacy data from hundreds of controlled trials, beginning as early as the 1930s, and were well-established as effective treatments for ADHD by the 1970s. The paediatric safety and efficacy database on acute and long-term use of these agents has continued to grow and includes data not only on school-aged children, but more recently has expanded into preschool children and adolescents (AAP, 2001; Biederman & Spencer, 2008; Brown et al.2005; Greenhill et al.2002; Pliszka et al.2007). There has also been a significant increase in data supporting the utility of non-stimulant agents for ADHD in the past 10 yr (AAP, 2001; Biederman & Spencer, 2008; Brown et al.2005; Greenhill et al.2002; Madaan et al.2006; Pliszka et al.2007). A meta-analysis of atomoxetine and stimulant studies revealed a robust effect size for atomoxetine and the stimulants, both of which are currently approved by the FDA for the treatment of ADHD. Atomoxetine demonstrated an effect size of 0.62, which would be considered a medium effect size, compared to 0.91 and 0.95, considered large effect sizes, for immediate- and extended-release stimulants, respectively (Faraone, 2003). A more recent FDA-approved agent, the α₂ agonist guanfacine XR, demonstrated effect sizes of 0.43–0.86 in two double-blind, placebo-controlled (DBPC) trials (Biederman et al.2008b; Sallee et al.2009b).

Stimulants

Stimulants have historically been considered a first-line treatment for ADHD, with approximately 75% of children responding to the first agent selected, and 80–90% eventually responding if two different stimulants are tried consecutively (Pliszka, 2003). Although the MTA study examined the use of immediate-release methylphenidate, extended-release preparations are now commonly used to improve adherence to the treatment schedule, thus providing less opportunity for gaps in coverage. A combination of immediate- and extended-release preparations, selected and titrated according to tolerability and response, may ultimately be required to optimally manage the child's individual pharmacotherapy needs. All stimulant medications currently approved for the treatment of ADHD are derivatives of either methylphenidate or amphetamine, both of which act by enhancing the neurotransmission of dopamine, and to a lesser extent, norepinephrine (Biederman & Spencer, 2008). DBPC studies in children, adolescents and adults have demonstrated that 65–75% of subjects typically respond to stimulant treatment, compared to 4–30% of those on placebo (Greenhill et al.2002; Pliszka, 2007). Recent research has focused on improving the delivery mechanisms of the stimulant medications in order to extend the duration of action. With multiple formulations of these medications (short-, intermediate- and long-acting) as well as a variety of administration options available (e.g. capsules, sprinkleable capsules, tablets, chewable tablets, oral solution, transdermal patches), treatment can be tailored to individual patient needs.

The MTA study demonstrated the tolerability and efficacy of t.i.d. immediate-release methylphenidate in a randomized trial of 579 children aged 7–9.9 yr with the combined subtype of ADHD. Dose titration was based on effect as reported by parent and teacher rating scales, and tolerability. Children in the manualized pharmacotherapy arm of the study had mean final doses of 32.1±15.4 mg/d, and those assigned to manualized pharmacotherapy plus behavioural intervention had mean final doses of 28.9±13.7 mg/d. The MTA study allowed children weighing <25 kg to have methylphenidate doses of up to 35 mg/d, and allowed doses up to 50 mg/d for children who weighed more. Average doses in the smaller children were 0.95±0.40 mg/d, and 1.13±0.55 mg/d in those that were heavier (MTA Cooperative Group, 1999).

Prior to the NIMH-funded PATS there were less than a dozen small placebo-controlled trials of psychostimulants in preschool children, and all utilized immediate-release methylphenidate (Kratochvil et al.2004). Doses in these studies did not exceed 0.6 mg/kg, a narrower range than the 0.3–1.0 mg/kg used in older children (Kratochvil et al. 2004), and were administered q.i.d. or b.i.d., rather than the t.i.d. schedule often required for optimal effect. Efficacy of methylphenidate in the preschool age group varies from older children (Connor, 2002), as does the adverse effect profile (Firestone et al.1998). PATS, which used a titration model similar to the MTA's, included 165 children aged 3.5–5 yr initially randomized to either placebo or immediate-release methylphenidate (1.25 mg, 2.5 mg, 5 mg or 7.5 mg t.i.d.). Subjects received a week of treatment with each dose during the double-blind cross-over titration phase. Twenty-two percent of subjects were identified as best responding to 7.5 mg t.i.d. The mean final best dose in PATS was 14.22±8.1 mg/d, or 0.7±0.4 mg/kg.d (Greenhill et al.2006).

When PATS data were compared to MTA data, it was noted that the younger children had lower optimal doses, by weight, of immediate-release methylphenidate (0.7 mg/kg.d compared to 1.0 mg/kg.d). Pharmacokinetic data also demonstrated a slower clearance of a single dose of methylphenidate in 4- and 5-yr-old children compared to school-aged children (Wigal et al.2007). Tolerability seems to have age-related variability, with younger children demonstrating more emotional adverse events (e.g. crabbiness, irritability and proneness to crying) than school-aged children. Thus, slower titration, closer monitoring and smaller doses of stimulants are advised when treating preschool children (Pliszka, 2007).

Adverse effects

All formulations of the stimulant medications have similar adverse-event profiles (Greenhill et al.2002). Delayed sleep-onset, decreased appetite, weight loss, headache, stomach upset and increased heart rate and blood pressure are common. Emotional outbursts and irritability have also been frequently reported in younger children (Wigal et al.2006).

Concerns with cardiovascular safety of ADHD pharmacotherapies have led to specific recommendations for pre-treatment evaluation, treatment selection and monitoring. Much scrutiny is given to the risks present for children with structural cardiac abnormalities, but potentially medication-related changes in heart rate and blood pressure are also observed in healthy children with ADHD. In a study of 10 yr of Florida Medicaid claims, stimulant use in patients with ADHD was associated with 20% more emergency-room visits, and 21% more office visits for cardiac symptoms (Winterstein et al.2007).

Gould et al. reported that the rate of sudden death in paediatric patients taking a psychostimulant was the same as that seen in the general population, with 11 sudden deaths reported between 1992 and 2005. However, in a matched case-control study, a significant association of stimulant use with sudden death was seen when comparing data for 564 reports of sudden death in 7- to 19-yr-olds with the deaths of 564 same-aged patients who died in motor vehicle accidents (odds ratio 7.4, 95% confidence interval 1.4–74.9) (Gould et al.2009).

The AAP (Perrin et al.2008) recommends that a targeted cardiac history and physical examination be part of the assessment of a child prior to initiating ADHD treatment. Questions regarding a prior patient history of heart disease, palpitations, syncope or seizures, or a family history of sudden death in children or young adults, cardiomyopathy or long-QT syndrome should be asked. If these are present, an ECG and/or referral to a cardiologist may be warranted prior to initiating treatment. These cardiovascular risks may become more of an issue in the treatment of adults who may have concurrent hypertension and cardiovascular disease.

Atomoxetine

Atomoxetine, which selectively blocks re-uptake at the noradrenergic neuron, was the first non-stimulant medication approved by the FDA for the treatment of ADHD. Two large, DBPC efficacy studies demonstrated significant improvement in ADHD symptoms with atomoxetine compared to placebo, with 64.1% and 58.7% of atomoxetine subjects responding (Spencer et al.2002). More than a dozen DBPC trials have provided evidence supporting the safety and efficacy of atomoxetine dosed both once and twice-daily for the treatment of ADHD in children, adolescents, and adults (Kelsey et al.2004; Michelson et al. 2001, 2002, 2003; Spencer et al.2002; Weiss et al.2005).

The FDA-approved target therapeutic dose of 1.2 mg/kg.d was selected following a dose-finding study which observed a graded dose-response to atomoxetine 0.5 mg/kg.d and 1.2 mg/kg.d, but no significant difference between 1.2 mg/kg.d and 1.8 mg/kg.d for reduction of core ADHD symptoms. Improvements in psychosocial functioning, however, were seen when the dose was increased to 1.8 mg/kg.d without any significant difference in adverse events (Michelson et al.2001).

Atomoxetine is not approved for use in children aged <6 yr. However, there has been one DBPC trial (n=101), examining the use of atomoxetine in 5- and 6-yr-olds. Improvements were noted on parent and teacher ADHD-IV ratings for children assigned to atomoxetine compared to those on placebo (p<0.05). Three subjects withdrew from the study due to adverse events (atomoxetine=0, placebo=3). The mean final daily dose of atomoxetine was 1.38 mg/kg.d. Despite statistically significant improvements in ADHD symptoms, and the fact that the parents received concomitant education on ADHD and behavioural interventions as a part of the study, the children continued to have ADHD-IV (parent) scores above the 86th percentile for age and gender at study completion (Kratochvil et al.2008b).

Adverse effects

Common acute adverse effects of atomoxetine include sedation, loss of appetite, nausea, vomiting, irritability, and headaches. In an analysis of the efficacy and tolerability of atomoxetine in young vs. older children, no significant differences were noted in the adverse-event profile or response to atomoxetine (Kratochvil et al.2008a).

Atomoxetine carries additional warnings for hepatotoxicity and suicidality risk. An analysis of laboratory data from 7961 adult and paediatric subjects in atomoxetine clinical trials revealed 41 instances of elevations in AST and ALT. There were 351 spontaneous reports of hepatic events in the first 4 yr atomoxetine was on the market. Of these, three suggested atomoxetine as a probable cause, and 1/3 had a positive re-challenge. In all three cases, symptoms resolved following discontinuation of atomoxetine. These data resulted in recommendation that atomoxetine be discontinued if jaundice or elevations in hepatic enzymes are present (Bangs et al.2008a). A 2008 analysis of data from 14 studies of atomoxetine by Bangs and colleagues demonstrated that suicide ideation was more common in subjects receiving atomoxetine (0.37%, 5/1357 subjects) compared to those receiving placebo (0%, 0/851 subjects). To place the risk of suicidality in context, the number needed to harm (NNH) was 227, whereas the number needed to treat (NNT) to achieve remission of ADHD symptoms was five. No suicides occurred in any of the trials in the analysis (Bangs et al.2008b).

Stimulant and atomoxetine comparator trials

Atomoxetine and osmotic release oral system (OROS) methylphenidate

In a comparator trial in 516 children and adolescents aged 6–16 with ADHD, subjects were randomized to 6 wk of treatment with either atomoxetine up to 1.8 mg/kg.d (n=222), OROS methylphenidate up to 54 mg/d (n=220) or placebo (n=74). Atomoxetine and OROS methylphenidate were both superior to placebo, with 45% (p<0.003) and 56% (p<0.001) responding, respectively. Effect sizes were 0.6 for atomoxetine and 0.8 for OROS methylphenidate. Decreased appetite was the only adverse event separating from placebo for both active treatments (p<0.05). Subjects receiving OROS methylphenidate reported experiencing insomnia, while those assigned to atomoxetine had more frequent complaints of somnolence. Weight loss and increased diastolic blood pressure (p<0.05) were noted to be significant for both drugs compared to placebo, and an increased pulse rate was significant in the atomoxetine group compared to OROS methylphenidate and placebo (p<0.05) (Newcorn et al.2008).

For the stimulant-naive patients (n=191) participating in this trial, response rates to atomoxetine (57%, p=0.004) and methylphenidate (64%, p≤0.001) were comparable (p=0.43), but those subjects with prior exposure to stimulants (n=301), had better responses to methylphenidate (51%, p=0.002) than to atomoxetine (37%, p=0.09) (p=0.03) (Newcorn et al.2008). The effect size for atomoxetine was greater in stimulant-naive patients (0.9), compared to patients previously treated with stimulants (0.5), while the effect-sizes for OROS methylphenidate in patients not previously treated with a stimulant and with prior exposure were 1.0 and 0.8, respectively (Newcorn et al.2008).

Subjects initially assigned to OROS methylphenidate were then switched to atomoxetine at the end of the 6-wk acute treatment phase of the study. Forty-two percent (29/69 subjects) who did not respond to atomoxetine in the second phase of the study had previously responded to OROS methylphenidate during acute treatment, while 43% of subjects who did not respond acutely to OROS methylphenidate (30/70 subjects) went on to respond to atomoxetine. This may indicate a differential response to treatment for some patients (Newcorn et al.2008).

Atomoxetine and mixed-amphetamine salts

In a 3-wk laboratory school comparison of atomoxetine and extended-release mixed amphetamine salts in 6- to 12-yr-olds with either combined or hyperactive-impulsive type ADHD, improved attention and academic performance were noted with both treatments. Mixed amphetamine salts-treated subjects had greater improvements than those who received atomoxetine (p<0.001). The difference at endpoint was statistically and clinically significant; however, the relatively short 3-wk duration of the study may not have been sufficient to demonstrate the full effect of atomoxetine treatment. The mixed amphetamine salts group reported experiencing insomnia, decreased appetite, upper abdominal pain, anorexia and headache, while the most common adverse events reported in the atomoxetine group were somnolence, appetite decrease, upper abdominal pain, vomiting and headache. Vital sign changes were similar for both groups and were not statistically significant (Wigal et al.2005).

α₂ agonists

The α₂ adrenergic agents, clonidine (Catapres) and immediate-release guanfacine (Tenex), have been used relatively commonly over the past decade as second-line or adjunctive treatments in the USA. International comparisons (Winterstein et al.2008), however, show very different co-medication patterns between the USA and European countries where α₂ adrenergic agents are rarely used. Clonidine has been shown to reduce ADHD symptoms in patients with comorbid tics, aggression and conduct disorder. Immediate-release clonidine is short-acting and requires multiple divided doses throughout the day (Brown et al.2005). In the USA clonidine is also available as a transdermal patch, allowing for once-weekly application. An extended-release formulation (Kapvay™) was approved by the FDA in September 2010, for the treatment of ADHD in children and adolescents aged 6–17 yr. Kapvay received approval as both monotherapy and in combination with a stimulant.

Guanfacine is a more selective α₂-adrenergic agonist with less sedation and a longer duration of action (Biederman & Spencer, 2008). A small open-label study of immediate-release guanfacine showed improvements in hyperactivity and inattention, with transient sedation as the most common adverse event (Hunt et al.1995), and additional studies have demonstrated its utility and good tolerability in treating ADHD with co-occurring tic disorders and Tourette's (Chappell et al.1995; Scahill et al.2001). An extended-release form of guanfacine was given FDA approval in 2009 as monotherapy for paediatric ADHD following two controlled trials (study 1: n=345, ages 6–17 yr; study 2: n=324, ages 6–17 yr). Adverse events were largely dose-dependent. Both studies had similar tolerability data, with the most common treatment-emergent adverse events being headache, somnolence, fatigue, sedation, and upper abdominal pain. No clinically significant vital sign or ECG changes were seen (Biederman et al.2008b; Sallee et al.2009b). Dose-based effect sizes ranged from 0.43 to 0.86, and response rates were 43% for the 3-mg dose and 62% for the 4-mg dose.

Guanfacine's most common acute adverse effects include somnolence, headache, fatigue, upper abdominal pain and sedation. Bradycardia was reported in long-term studies (Biederman et al.2008a; Sallee et al.2009a)

What is the impact of ADHD pharmacotherapy?

The benefits of pharmacotherapy are most evident in reduction of the core symptoms of ADHD. By reducing inattention, hyperactivity and impulsivity, patients with ADHD are better able to perform academically and socially. Studies have demonstrated that children treated with stimulants have improved attention to school work, decreased disruptive behaviours and decreased non-compliance. Short-term data also shows improvements in academic performance and productivity (Barkley, 1998). Some data suggest that children with ADHD treated with psychostimulants demonstrate better academic outcomes as evidenced by WIAT-II subtests and high school grade point average (GPA) than children with ADHD who were not treated. However, the treated children did not do as well as non-ADHD controls. It is unclear if pharmacotherapy alone translates to long-term academic success (Powers et al.2008).

Social interactions between affected children and their parents, teachers and peers are significantly improved with stimulant treatment. Treated children are more compliant with commands and more appropriately responsive to interactions with others, with less negative and off-task behaviour. As a result, adult redirections and supervision needs decrease, and praise and positive attention to the child increase. ADHD children treated with stimulants also appear to be better accepted by peers, probably as a result of reduced negative and aggressive behaviour (Barkley, 1998). Health-related quality-of-life outcomes measured by the Child Health Questionnaire (CHQ) were improved along with ADHD symptoms in children treated with atomoxetine in a DBPC dosimetry study in children and adolescents aged 8–18 yr (Michelson et al.2001).

Early treatment with methylphenidate does not appear to increase risk for negative outcomes, and may have beneficial long-term effects (Mannuzza et al.2008). However, long-term data from the MTA study notes that benefits of pharmacotherapy are sustainable up to 2 yr for the majority of subjects followed, but by the third year of follow-up, only about one third of subjects demonstrated ongoing benefit with medication treatment (Swanson et al.2008). Despite decreases in ADHD symptoms, the MTA subjects as a group still had relatively poorer ratings of behaviour, academic and overall functioning compared to normal controls at 6- and 8-yr follow-ups.

How long should treatment last?

Epidemiological surveys of community samples indicate that 2–6% of preschool children meet diagnostic criteria for ADHD (Angold et al.2000; Lavigne et al.1996), with prevalence rates in school-aged children conservatively estimated to be between 3% and 7% (APA, 2000). As children grow into adolescence and adulthood the prevalence of ADHD decreases, yet still persists in significant numbers, estimated at approximately 3–4% in adults (Fayyad et al.2007). Even though the presentation may vary from early childhood to adulthood, the impairment there is no less significant (Kessler et al.2006). A multitude of studies have demonstrated a correlation between ADHD in adults and global impairment in functioning, including: smoking and substance abuse, diminished rates of college graduation, occupational/vocational difficulties, motor vehicle accidents, legal problems, unplanned pregnancies, and relationship problems (Barkley, 2006).

In a 10-yr case-controlled follow-up study of 112 male adults with ADHD, potential protective factors of stimulant treatment for ADHD were assessed. Biederman et al. (2008c, 2009) found no evidence that stimulant treatment in childhood or adolescence either increased or decreased the risk for development of substance use disorders in young adulthood, but that ADHD patients treated with stimulants were at significantly less risk of developing depressive and anxiety disorders, disruptive behaviour, and repeating a grade in school than the ADHD patients who were not treated. Daviss et al. also demonstrated a similar finding of ADHD pharmacotherapy reducing the risk of later major depression (Daviss et al.2008).

With the longitudinal course of ADHD documented, the AACAP Practice Parameter recommendations serve as a reminder to periodically evaluate the need for ongoing treatment of ADHD with pharmacotherapy. Follow-up clinic visits ensure that medication remains effective, dosing is optimal, and adverse events are minimized. The AACAP recommends that ADHD treatment be individualized, and that the duration of treatment should continue as long as impairing symptoms are present (Pliszka, 2007).

Considerable evidence demonstrating the efficacy of psychostimulants in treating adults with ADHD (Asherson, 2005), led to FDA approval of both methylphenidate (extended-release methylphenidate and d-methylphenidate) and amphetamine (extended-release mixed amphetamine salts). Atomoxetine has also received FDA approval for adults with ADHD based on two DBPC studies (Buitelaar et al.2007).

Adverse effects

A specific concern with long-term pharmacotherapy is impact on growth, so much so that the AACAP Practice Parameter for ADHD treatment includes a specific recommendation regarding regular height and weight monitoring, including serial plotting of growth parameters. The AACAP advises that a change in height or weight crossing two percentile lines is suggestive of abnormal growth and warrants a medication holiday, dose adjustment or change. Reductions in growth must be balanced with benefits of treatment (Pliszka, 2007).

Swanson et al. (2005) demonstrated that children treated with stimulants grew more slowly and appeared to gain less weight than expected; however, they also theorized that, in general, children with ADHD may have different growth trajectories than their ‘normal’ peers. Statistically significant delays in height and weight were also seen with stimulant treatment in a meta-analysis of 22 studies by Faraone et al. (2008) . The pooled data showed that the weight deficits were more significant than the deficits seen in height (p=0.002), although, both appeared to normalize over time (Faraone et al.2008).

Based upon a qualitative meta-analysis, Faraone et al. suggested that the effects on weight and height may be dose-dependent. There was no apparent difference, however, in the growth effects between methylphenidate and amphetamine, and cessation of treatment appeared to normalize growth (Faraone et al.2008). This normalization of growth with breaks over the summer or with drug discontinuation has been demonstrated in additional studies (Gittelman et al.1988; Kaffman et al.1979; Klein & Mannuzza, 1988; Safer et al.1975); although analysis of data from the MTA study (MTA Cooperative Group, 2004) showed that while discontinuation of methylphenidate treatment did not reverse losses in expected height, it did have a beneficial effect on weight gain.

Atomoxetine has also been clearly linked with changes in height and weight trajectories, which for the group appeared to dissipate over time, despite continued treatment (Spencer et al. 2005, 2007). These data appear to indicate that for most children growth suppression, if present, will be transient and not clinically significant over time. Nonetheless, there is clearly an effect of these medications on growth. Therefore, while group averages over time may not be overly concerning, close monitoring of individual children taking ADHD medication is clearly indicated.

What is the management of treatment-resistant cases?

The vast majority of patients with ADHD will respond to one of the FDA-approved pharmacotherapies for the treatment of ADHD. If a patient does not respond adequately to appropriate trials (adequate duration and optimal dose) of these agents, a re-assessment of the diagnosis is warranted both to confirm the diagnosis of ADHD and to re-examine for missed co-occurring disorders (AACAP recommendation). Co-occurrence of learning disorders, developmental disorders and other psychiatric conditions can affect response to treatment and/or complicate treatment planning, and the addition of behavioural therapy to a medication regimen may be required. Non-FDA-approved pharmacotherapies (e.g. bupropion or tricyclic antidepressants) may be tried if interventions with a greater evidence base are either ineffective or contraindicated. Finally, combination therapy with FDA-approved agents and/or non-approved agents might be clinically indicated. Use of medications not approved for the treatment of ADHD, and treatment with more than one medication simultaneously elevates potential risks, however, and these risks as well as other treatment options must be discussed with the patient and caregivers, and if employed monitored closely (Pliszka, 2007).

Conclusion

ADHD is one of the best studied disorders in psychiatry. Reliable diagnosis at a young age is possible, and recognition of ADHD as a potentially life-long impairing disorder is increasing. As data emerge which describe the physiological evidence behind the historically ‘behavioural’ diagnosis, acceptance of the role of pharmacotherapy has increased for preschool children through adults. Guidelines such as those from the AACAP provide clear recommendations to the practising clinician for diagnosing, treating and monitoring patients with ADHD, in a manner which maximizes effectiveness, tolerability, and ultimately, functionality of the patient. As improvements are made in the delivery systems and durations of effect of the various psychostimulant agents, clinicians and patients will still be faced with what to do for those who do not respond. Research is expanding into non-stimulant agents, and the specific role these may have. Further examination of a potential ‘differential response’, as suggested in the comparator trial of atomoxetine and OROS methylphenidate, may ultimately better inform clinicians as to the selection of a specific pharmacotherapy for a specific individual. In the interim, appropriate diagnosis, informed prescribing, clinical monitoring and collaborative treatment planning, can all help to optimize outcomes in ADHD management.

Acknowledgements

None.

Statement of Interest

Dr Kratochvil was supported by NIMH Grant 5K23MH06612701. Received grant support from Eli Lilly, McNeil, Shire, Abbott, Somerset and Cephalon; was a consultant for Eli Lilly, AstraZeneca, Abbott Quintiles and Pfizer. He is Editor of the Brown University Child & Adolescent Psychopharmacology Update, member of the REACH Institute Primary Pediatric Psychopharmacology Steering Committee, member of the American Professional Society for ADHD & Related Disorders Board of Directors, and received study drug for a NIMH-funded study from Eli Lilly and Abbott.

Ms. Vaughan has received salary support from NIMH Grant 5K23MH06612701, Eli Lilly, and Bristol–Myers Squibb.

Dr March has served as a consultant or scientific advisor to Pfizer, Eli Lilly, Wyeth, GlaxoSmithKline, BMS, Johnson and Johnson, Psymetrix, Scion, Transition Therapeutics and MedAvante; has received research support from Eli Lilly and Pfizer; has received study drug for a NIMH-funded study from Eli Lilly and from Pfizer; is an equity holder in MedAvante; receives royalties from Guilford Press, Oxford University Press and MultiHealth Systems; and receives research support from NARSAD, NIMH and NIDA.

References