|Home | About | Journals | Submit | Contact Us | Français|
This article reviews rational approaches to treating attention-deficit/hyperactivity disorder (ADHD) in preschool children, including pharmacological and nonpharmacological treatments. Implications for clinical practice are discussed.
We searched MEDLINE, PsychINFO, Cumulative Index to Nursing & Allied Health, Educational Resources Information Center, Cochrane Database of Systematic Reviews and Database of Abstracts of Reviews of Effects for relevant literature published in English from 1967 to 2007 on preschool ADHD. We also reviewed the references cited in identified reports.
Studies were reviewed if the sample included at least some children younger than 6 years of age or attending kindergarten, the study participants had a diagnosis of ADHD or equivalent symptoms, received intervention aimed at ADHD symptoms, and included a relevant outcome measure.
Studies were reviewed for type of intervention and outcome relevant to ADHD and were rated for the level of evidence for adequacy of the data to inform clinical practice.
The current level of evidence for adequacy of empirical data to inform clinical practice for short-term treatment of ADHD in preschool children is Level A for methylphenidate and Level B for parent behavior training, child training, and additive-free elimination diet.
Attention deficit/hyperactivity disorder (ADHD) frequently begins between 2 and 4 years of age (Connor 2002; Egger and Angold 2006). It is associated with significant impairment in terms of emotional distress for the preschool child and the caregivers (DuPaul et al. 2001), expulsion from daycare or early education settings (Blackman 1999), demands on the caregiver's time, exclusion from family events, and accident proneness and other safety concerns (Lahey et al. 2004; Rappley et al. 1999). Children with ADHD have comorbid mental health and chronic health problems and are frequent users of the healthcare system (Rappley et al. 2002). As shown by several prospective longitudinal follow up studies, behavior problems in preschool children persist to school-age years and continue to be associated with significant impairment (Campbell and Ewing 1990; Campbell et al. 2000; Egeland et al. 1990; Fischer et al. 1984; Lahey et al. 2004; Lavigne et al. 1998; McGee et al. 1991; Richman et al. 1982). In a recent study, 79.2% of the preschool children who met full diagnostic criteria for ADHD and 34.5% of the preschool children who met criteria in one situation only at initial assessment continued to meet full ADHD diagnostic criteria and exhibited global academic and social impairment three years later (Lahey et al. 2004).
Impairment from ADHD and persistence of problems at later ages underscores the need for early intervention in preschool children with ADHD (Beckwith 2000; Bierman et al. 2007; Campbell 2002; Conduct Problems Prevention Research Group 1999; Elliot et al. 2002; Petras et al. 2008; Shure et al. 2001). Recently, the Preschool Psychopharmacology Working Group (PPWG) reviewed pharmacological treatment studies in preschool children and proposed treatment algorithms for preschool psychiatric disorders (Gleason et al. 2007), however the nonpharmacological treatments for ADHD were not reviewed in detail. The primary objective of this paper is to review rational approaches to pharmacological and nonpharmacological treatment of ADHD in preschoolers and the implications for clinical practice. For the purposes of this paper, we define preschool age as prior to starting formal schooling, i.e., first grade. Hence, we reviewed studies that included children in kindergarten and/or younger than 6 years of age. The Food and Drug Administration (FDA) provides additional demarcation for children younger than 6 years. Most of the pharmacological agents for treatment of ADHD are approved by the FDA only for children older than 6 years (with the exception of amphetamines), and the FDA considers their use in children younger than 6 years as “off-label.”
We searched MEDLINE, PsychlNFO, Cumulative Index to Nursing & Allied Health, Educational Resources Information Center, Cochrane Database of Systematic Reviews and Database of Abstracts of Reviews of Effects for relevant literature on treatment of preschool ADHD. We also reviewed the references cited in identified reports to locate other relevant studies. Due to limited literature in this area, we reviewed both controlled and non-controlled studies. The reviewed studies met the following inclusion criteria: Published in English in the past 40 years (between 1967 to 2007); included at least some children younger than 6 years of age and/or attending kindergarten who had a diagnosis of ADHD; or exhibited behavior problems that are part of the ADHD diagnostic criteria, involved intervention aimed at ADHD symptoms, and included an outcome measure to monitor ADHD symptoms. To determine the level of evidence to inform clinical practice, we adapted the International Psychopharmacology Algorithm Project criteria (Jobson and Potter 1995) previously used by Judice and Mayes (2003) to categorize psychopharmacology treatments in preschool children. Since most of the reviewed child training studies were single case design experiments (state of current evidence for child training studies), based on the Task Force on Promotion and Dissemination of Psychological Procedures guidelines (Task Force on Psychological Intervention Guidelines 1995) previously used by Chorpita et al. (2002) to assess efficacy of psychosocial treatment studies in children and adolescents, we modified the criteria to include single case design experiments in addition to randomized controlled trials.
A treatment was considered to have evidence at Level A, if it demonstrated a significant difference on an ADHD outcome variable in a sample of preschoolers with a Diagnostic and Statistical Manual (DSM) diagnosis of ADHD in at least two randomized controlled trials (RCT) or two series of single case design experiments comparing randomly assigned active treatment to a comparison treatment or placebo. A treatment was considered to have evidence at Level B, if it demonstrated a significant difference on an ADHD outcome variable in a sample of preschoolers with a DSM diagnosis of ADHD in one RCT, two or more RCTs with mixed results, or one series of single case design experiments comparing randomly assigned active treatment to a comparison treatment or placebo. Level C was assigned to a treatment to indicate that data were based on uncontrolled trials, case reports, retrospective chart reviews, or informed clinical opinion.
Here we review the current available evidence for psychopharmacological and nonpsychopharmacological interventions (psychosocial and alternative treatments) and clinical implications for preschool children with ADHD. Prior to starting any treatment, it is important to conduct a comprehensive assessment that is contextually relevant and takes into account the rapid developmental changes occurring during preschool years. Because a comprehensive discussion of the assessment process for diagnosing ADHD in preschool children is beyond the scope of this paper, the reader is referred to several excellent reviews addressing preschool nosology, diagnosis, and assessment (Angold et al. 2004; Campbell 2002; Carter et al. 2004; Egger and Angold 2006; Emde et al. 1993; Task Force on Research Diagnostic Criteria: Infancy and Preschool 2003). In general, a multi-method and multi-informant evaluation extending over multiple appointments to assess symptomatology and impairment in multiple environments and caregiving contexts is recommended (Carter et al. 2004; Gleason et al. 2007). A combination of diagnostic interviews and parent and teacher rating scales, with psychometric data in preschoolers, are commonly employed to aid in the diagnostic and assessment process. Examples of the parent and teacher rating scales include Conners' Rating Scales-Revised (CRS-R) (Conners 2001), Swanson, Nolan and Pelham (SNAP) rating scale (Swanson 1992), Child Behavior Checklist-1½ (CBCL-1½) (Achenbach and Rescorla 2000), and ADHD Rating Scale (ADHD-RS) (DuPaul 1998; Gimpel and Kuhn 2000). The Preschool Age Psychiatric Assessment (PAPA) (Egger et al. 2006) is a reliable and valid semi-structured diagnostic parent-interview that is widely used in research studies of preschool pathology. However, clinical settings may find the cost and length of the PAPA training and administration to be a barrier for its use in routine clinical practice. Nonetheless, adequate history, mental status examination, and collateral information are important for developing an appropriate treatment plan that addresses the biopsychosocial issues specific to each family.
There are several challenges to psychopharmacological treatment of preschool children with ADHD. Preschool age is a period of continued rapid neuronal maturation including synaptic remodeling and construction. Cortical synaptic density reaches its maximum at age 3 and is substantially modified by the pruning process from ages 3 to 7 years (Huttenlocher 1990). Cerebral metabolic rate peaks between 3 and 4 years of age (Chugani 1987). Aminergic systems play an important role in neurogenesis, neuronal migration, axonal outgrowth, and synaptogenesis (Coyle 1997) and are also the targets of action for many psychopharmacological agents as indicated by studies in preclinical models. Thus, clinicians are faced with a dilemma. On one hand, whether it is prudent to recommend exposing the rapidly developing brain of a preschool child to psychopharmacological agents. On the other hand, clinicians also need to consider the consequences of an untreated disorder. Early exposure to adverse environmental circumstances and stress has been shown to result in long-lasting impact on the brain and emotional regulation of animals and humans (Graham et al. 1999; Matthews 2002; Nemeroff 2004).
Information about the use of psychopharmacological agents for treatment of ADHD is available mostly for school age children. In school age children, psychostimulants are the mainstay of treatment for ADHD; nonstimulant psychopharmacological agents are frequently recommended as a second-line treatment if a school age child's ADHD symptoms do not adequately respond to stimulants (Dulcan and Benson 1997). Recently, a nonstimulant psychopharmacological agent, atomoxetine, has been shown to be a safe and effective treatment of ADHD in school age children (Kratochvil et al. 2004; Michelson et al. 2001). Comparatively, there is limited information on the use of psychopharmacological agents for treatment of ADHD in preschool children. No pharmacokinetic and dose finding studies to identify dosage and frequency of drug administration in preschool children are available. Until recently, clinicians were left to extrapolate findings from older children to preschool children. However, medications used in older children may have specific toxicities in preschool children (Wigal et al. 2006), and there may be differences in efficacy in preschool children compared to school age children (Greenhill et al. 2006). Additionally, extrapolation to preschool children of data collected in older children is not always possible due to differences in development.
There are over 250 published studies of psychopharmacological agents in school age children with ADHD (Wilens et al. 2002). In contrast, there are a total of 24 published reports (blinded and open-label studies) on the use of psychopharmacological agents involving over 495 preschool children with ADHD. Of the 24 published reports (Tables 1 and and2),2), 20 published reports are on the use of stimulants, 2 published case reports on the use of α2 agonists, 1 published case report on the use of atomoxetine, and 1 published case report on the use of fluoxetine in preschool children with ADHD.
Of the 20 published studies on the use of stimulants in preschool children, 12 double-blind group treatment studies (one parallel groups, 10 crossover and 1 ABA design) included 417 preschool children treated with methylphenidate (MPH). Two double-blind MPH/placebo crossover studies included both preschool and older children, but did not specify the number of preschool participants (Barkley 1988; Fischer and Newby 1991); and one blinded time series treated one preschool child with dextroamphetamine (Speltz et al. 1988). The remaining five published reports are open-label studies or case reports involving a total of 61 preschool children treated with MPH or dextroamphetamine (Alessandri and Schramm 1991; Byrne et al. 1998; Cohen et al. 1981; Ghuman et al. 2001; Stiefel and Dossetor 1998).
Twelve of the fifteen blinded MPH studies treated typically developing preschool children with ADHD; seven studies included only preschool children (Barkley 1988; Conners 1975; Firestone et al. 1998; Greenhill et al. 2006; Musten et al. 1997; Schleifer et al. 1975; Short et al. 2004; Speltz et al. 1988), while the other six studies included both preschool and older children (Barkley et al. 1988; 1985; 1984; Chacko et al. 2005; Cunningham et al. 1985; Fischer and Newby 1991). Of the other two studies, one included a mixture of inpatient or outpatient preschool and school age children with ADHD who were either typically developing or had autism or other developmental disorders (Mayes et al. 1994), and the one remaining blinded study treated preschool children with developmental disorders (Handen et al. 1999). The diagnostic procedure used most frequently included a combination of clinical interview and dimensional rating scales. With the exception of the recent PATS study, sample size was small ranging from 11–59, duration of the psychostimulant trials ranged from 3–9 weeks, and stimulant dose ranged from 0.15–0.6mg/kg. Mixed outcomes were reported for efficacy. Based on direct observation of the preschool children's nursery school behavior, one study reported no improvement with MPH compared to placebo (Schleifer et al. 1975). Positive response to MPH was reported by other investigators in 80%–83% of typically developing preschool children (Conners 1975; Greenhill et al. 2006; Short et al. 2004) and 71%–73% of preschool children with developmental disorders (Handen et al. 1999). Data on side effect profile in preschool children was also divergent. Rates of side effects ranged from minimal or clinically negligible (Conners 1975) to 89% in typically developing preschool children (Schleifer et al. 1975) and 45%–50% in preschool children with developmental disorders (Handen et al. 1999; Mayes et al. 1994). Dysphoria, crying, whining, irritability, and solitary play were more frequently reported in preschool children than seen in older children.
The 6-site PATS study randomized 165 preschool children (3–5.5 years) diagnosed with ADHD in a placebo-controlled, double-blind crossover design, to one week each of 4 MPH doses (1.25mg, 2.5mg, 5mg and 7.5mg TID) and placebo. With the exception of the lowest MPH dose, improvements in parent- and teacher-rated ADHD symptoms were reported with MPH compared to placebo; the 7.5mg TID dose was found to be the most effective. The effect sizes (Cohen's d) in the intent-to-treat sample ranged from 0.4–0.8 and were smaller than those reported for school age children treated with MPH (Greenhill et al. 2006). Interestingly, secondary analyses of the PATS efficacy data showed that preschoolers with ADHD with no or one comorbid disorder (primarily oppositional defiant disorder [ODD]) had treatment responses (Cohen's d=0.89 and 1.00, respectively) at the same level as found in school age children (Ghuman et al. 2007). Preschoolers with 2 comorbid disorders had moderate treatment response (Cohen's d=0.56) and preschoolers with 3 or more comorbid disorders did not respond to MPH (Ghuman et al. 2007). However, caution is needed in the generalization of the findings as there were only 15 preschoolers (9% of the sample) with 3 or more comorbid disorders compared to 150 preschoolers with two comorbid disorders (n=34, 21% of the sample), one comorbid disorder (n=69, 42% of the sample) or no comorbid disorders (n=47, 28% of the sample). In addition to decreased appetite, stomach ache, and sleep difficulties usually seen in school age children, increased rates of social withdrawal and lethargy were reported especially at higher doses. A higher discontinuation rate (8.3%) due to MPH side effects was reported in the PATS study than the 0.5% discontinuation rate in the National Institute of Mental Health (NIMH) Multimodal Treatment of ADHD (MTA) study with school-age ADHD children (Wigal et al. 2006). Compared with the Center for Disease Control (CDC) norms, the preschool children with ADHD in the PATS were 2.0cm taller and 1.8 kg heavier at baseline. A 20% less than expected annual height gain (−1.38cm/year) and 55% less than expected annual weight gain (−1.32 kg/year) was reported for the children who continued MPH for a year in the open-label follow up phase (Swanson et al. 2006). Decrease in weight velocity was evident at the end of the 5-week double-blind crossover phase. Most preschoolers with ADHD were able to maintain improvement over 10 months of open-label follow up treatment (Vitiello et al. 2007).
Two of the open-label stimulant studies were prospective open-label treatment trials, two were case reports, and one was a retrospective chart review. Positive response to stimulants (MPH or dextroamphetamine) was reported in four of the open-label studies; one prospective treatment trial reported no treatment effect and reported a 30% discontinuation rate due to MPH adverse effects (Cohen et al. 1981).
The two published case reports of open-label treatment with α2 agonists in 5 preschool children with ADHD reported improvement in hyperactive and impulsive behavior (Cesena et al. 1995; Lee 1997), one published case report of open-label treatment with atomoxetine in 10 preschool children reported improvement in hyperactive, impulsive and inattentive symptoms (Kratochvil et al. 2007), and one published case report of open-label treatment with fluoxetine in one preschool child with ADHD reported improvement in attention span (Campbell et al. 1995) (Table 3).
Despite controversy and scarcity of empirical information regarding dose guidelines, safety, and efficacy, psychopharmacological agents are being prescribed to preschool children.
This is a serious public health concern and was identified as a research priority by the Surgeon General (National Institutes of Health 2000; US Public Health Service 2000) and the White House (Pear 2000).
In 1994, 226,000 MPH prescriptions were written for children under 6 years of age (US Food and Drug Administration 1997). A three-fold increase in MPH use in 2–4 year old children was reported from 1991–1995 (Zito et al. 2000); Marshall (2000) reported that 150,000 to 200,000 children between the ages of 2 and 4 years were estimated to be taking MPH. Outpatient prescription data from 7 state Medicaid programs revealed 67.3% stimulant and 26% α-agonist use in 2001 in 2–4 year old children treated with psychotropic agents (Zito et al. 2007).
Pre-school children with ADHD symptoms are more often treated with stimulants in the community than with any other drug. Those not treated with stimulants, are often given other psychotropic medications, including those that have not been shown to have any efficacy in ADHD, such as selective serotonin reuptake inhibitors (SSRIs) (Rappley et al. 2002) and those with some efficacy for ADHD but leading to severe long-term adverse events (e.g., tardive dyskinesia), such as neuroleptics (Minde 1998; Rappley et al. 2002). Health Maintenance Organization (HMO) and Medicaid database surveys in 1995 showed frequent use of psychotropic medications in preschool children diagnosed with ADHD. Frequency of psychotropic drug prescription in 1–3 year old children diagnosed with ADHD (N=223) in the Michigan Medicaid system was 57% (n=127) compared to 26% (n=47) for psychosocial intervention (Rappley et al. 1999). Psychotropic medications alone as a sole intervention strategy were utilized in 40% children (n=89); comparatively psychosocial intervention as a sole treatment was utilized in only 9% (n=21). Stimulants were prescribed for 93.7%, α2 agonists for 44.9%, tricyclic antidepressants for 33.1%, neuroleptics for 15.7%, and SSRIs for 11% of the preschool children receiving psychotropic medications. More than half of the children received treatment for longer than 6 months. Medication monitoring was inadequate—for 75 children (59%) follow-up visits occurred every 3 months and for 25 children (19%) at intervals greater than 6 months (Rappley et al. 2002).
There is evidence for short-term efficacy and long-term effectiveness and tolerability of psychostimulants, especially MPH, in preschool children with ADHD. Response is reported to be less robust and response rate is reported to be lower in preschool children with ADHD compared to older children. Preschool children with ADHD are sensitive to developing more side effects especially at higher doses and have unique adverse effect profile including more irritability and mood changes. This sensitivity to stimulant adverse effects may be a limiting factor in achieving an adequate and/or robust response in preschool children with ADHD. Only open-label information is available regarding effectiveness and tolerability of one non-stimulant, atomoxetine, in preschool children with ADHD. Additionally, pharmacological interventions may be effective in reducing core ADHD symptoms such as impulsivity, overactivity, and inattention among preschool children with ADHD (Greenhill et al. 2006); however, there is little evidence to suggest that psychostimulants improve long-term interpersonal relationships known to be important in predicting outcomes for children displaying disruptive behaviors (Coie and Dodge 1998; Pelham et al. 1998; Rubin et al. 2006). Moreover, no information is available about long-term safety and effects of psychopharmacological agents on brain development in preschool children.
Concerns about the short- and long-term safety of psychopharmacological agents especially on the developing brain of preschool children, coupled with ethical, societal, and political beliefs about manipulating behavior through medication and perceived overprescription (Jensen et al. 1999) often lead families and providers to favor other interventions for preschoolers (Dulcan and Benson 1997). In this section, we will review current evidence for success of psychosocial treatments in preschool children with ADHD.
There is evidence from studies in school-age children that long-term behavioral improvements may require psychosocial interventions (Ialongo et al. 1993). Inhibitory processes play a critical role in impaired functioning in children with ADHD (Barkley 1997; Barkley 1998; Barkley 2003; Nigg 2001; Nigg 2003; Quay 1997) and these, although rudimentary, are developing rapidly during the preschool period (Davidson et al. 2006; Diamond and Taylor 1996; Espy et al. 1999; Garon et al. 2008; Jones et al. 2003; Rueda et al. 2005). Psychosocial interventions targeting key executive functions, especially inhibitory processes, may be particularly helpful (Diamond et al. 2007; Dowsett and Livesey 2000).
Limited psychosocial intervention research has been conducted with preschool samples of children formally diagnosed with ADHD (Bryant et al. 1999; McGoey et al. 2002); however, there is considerable evidence that parent, child, and parent-child interventions can reduce problem behaviors in young children displaying a range of disruptive behaviors, including excessive hyperactivity and inattention as reviewed in the following section. We have grouped the psychosocial intervention studies into those that train parents in behavioral techniques and use the parents as the primary agent of change, and those that train children in a classroom setting to reduce problematic behaviors.
Among psychosocial interventions, parent training to help parents learn and implement behavioral treatment has the strongest evidence base showing positive effects for school age children with ADHD (Chorpita and Daleiden 2002). For preschool children with ADHD, parent training in behavior management is an especially helpful and the most appropriate psychosocial intervention (Stanley and Stanley 2005; Webster-Stratton et al. 2001). When children are young, parents have an enormous impact on their child's behavior (Capage et al. 1998; Eyberg et al. 1995; Funderburk et al. 1998; Hembree-Kigin and McNeil 1995) creating a window of opportunity to teach parents how to be positive and consistent in their parenting responses, help reduce noncompliant and aggressive behaviors, and help their child persist at a difficult task and provide successful experiences for their child, thus reducing risk for continued problems in later years. Parent behavior training programs for preschool children have been modeled after efficacious programs developed with older children (Anastopoulos et al. 1993; Dishion and McMahon 1998), and draw upon both social-learning and attachment theories to varying extent. Parent training programs may include sessions with parents, parent-child dyad, or a combination of parent sessions and work with the parent-child dyad to improve parent-child relationship, and increase the child's prosocial behaviors and decrease negative behaviors.
There are 15 published reports of parent behavior training treatment trials (either controlled, case series or case reports) that monitored outcomes in ADHD symptoms in preschool children with a DSM diagnosis of ADHD or preschool children displaying ADHD symptoms (Tables 4 and and5).5). However, there was a wide variation in the study design, type of control groups, inclusion criteria, diagnostic measures, type of psychosocial intervention, method of intervention delivery, and outcome measures employed in the studies. Nine studies used a randomized parallel groups design with control groups ranging from wait-list, community treatment, combination treatment to minimal treatment (Barkley et al. 2000; Bor et al. 2002; Corrin 2004; Jones et al. 2007; McGoey et al. 2005; Pisterman et al. 1992; 1989; Sonuga-Barke et al. 2001; Strayhorn and Weidman 1989; Strayhorn and Weidman 1991) and six studies did not employ any control group (Chang et al. 2004; Danforth 1999; Drash et al. 1976; Erhardt and Baker 1990; Henry 1987; Huang et al. 2003). Eight of the nine controlled studies included only preschool children and four of these eight studies selected the preschoolers based on a DSM diagnosis of ADHD through clinical or structured parent interview (Bor et al. 2002; Pisterman et al. 1992; 1989; Sonuga-Barke et al. 2001), and the other four selected preschoolers based on a rating scale cutoff. The ninth controlled study included both preschool-age and school-age children with a DSM diagnosis of ADHD. Sample sizes ranged from 20-50 per group. Most studies employed group-training sessions except for three studies that employed individual training sessions with the parents (Bor et al. 2002; Henry 1987; Sonuga-Barke et al. 2001). Training was conducted over 8–12 sessions, each parent training session lasting 1–3 hours. Most studies included both teaching/modeling sessions with the parents and work with the parent-child dyad with the exception of one study that included parent training sessions only (Jones et al. 2007), one study that included parent and/or child training sessions separately (Corrin 2004) and two studies that included only didactic teaching sessions with the parents (Barkley et al. 2000; Huang et al. 2003). Outcome assessments varied among the studies and included parent ratings of ADHD and disruptive behaviors, and direct behavior observation by independent raters for on-task behavior, child compliance with maternal commands, and parent-child interaction quality during structured play.
Improvements in ADHD symptoms were reported in three of the eight controlled studies that included only preschool children (Jones et al. 2007; Sonuga-Barke et al. 2001; Strayhorn and Weidman 1989) and only one of these studies (Sonuga-Barke et al. 2001) was in preschoolers formally diagnosed with ADHD. Preschoolers diagnosed with ADHD (N=78) were randomized to 8 weeks of parent training, parent counseling and support, or a wait-list group (Sonuga-Barke et al. 2001). Improvements in ADHD symptoms were reported with parent training compared to the other two conditions; positive effects were maintained at 6 month follow-up. No improvement in hyperactive, impulsive and/or inattentive symptoms on parent ratings or direct observation was reported in the remaining 5 controlled studies that included only preschool children. Improvements in ADHD symptoms were also reported in most of the non-randomized non-controlled case series (Drash et al. 1976; Erhardt and Baker 1990; Huang et al. 2003).
With the exception of Barkley et al (2000), most investigators reported improvements in parenting skills, parenting style of interaction, and child compliance. Barkley et al. (2000) reported poor treatment response with parent behavior training in kindergarteners (N=158) who met dimensional rating scale cutoff criteria for hyperactive, impulsive, inattentive, and aggressive behavior. The kindergarteners were assigned to one of four treatment groups: parent training only, classroom day treatment only, a combined condition, or a no treatment control group. The parent training intervention produced no effects.
There may be several reasons for the poor treatment response in this study. Inclusion criteria for the study were based on a rating scale cutoff and no clinical or structured diagnostic interviews were conducted for a diagnosis of ADHD. Most important, neither the children's caregivers nor their teachers had indicated impaired functioning in the kindergartners included in the study. There is evidence that psychosocial treatment approaches have greater impact on those children rated with higher levels of problems (Kellam et al. 1998; Wilson and Lipsey 2007). Furthermore, only 25% of the parents attended more than four parent behavior training sessions, and, finally, the parent behavior training was delivered in a didactic format.
Parent-Child Interaction Therapy (PCIT) (Eyberg 1988), a related yet distinct parent behavior training program, is an evidence-based intensive intervention for preschool children with disruptive behavior disorders. Parents are trained in behavioral management techniques within a play-therapy context; the therapist works with the parent-child dyad and provides “live” interactive coaching and immediate feedback to change interaction patterns within the dyad. A number of studies have reported positive effects of PCIT on outcomes for young children with externalizing problems, reducing hyperactive as well as disruptive behavior and improving compliance with maintenance of gains 6 years later (Hood and Eyberg 2003).
Although PCIT has not been used specifically to treat ADHD in preschoolers, many studies of PCIT included preschool children who met DSM diagnostic criteria for ADHD and reported benefit in parent- and/or teacher-rated ADHD symptoms (Eisenstadt 1993; Eyberg et al. 2001; Nixon 2001).
There have been a few studies in which behavioral management techniques have been applied within the preschool setting to treat individual preschool children presenting with hyperactive, inattentive and disruptive behaviors. We identified four published case series (Billings and Wasik 1985; Bornstein and Quevillon 1976; Bryant and Budd 1982; McGoey and DuPaul 2000) and two single-case reports (Allen et al. 1967; McCain and Kelley 1993) in preschool children with ADHD symptoms in which child training was conducted within the context of the classroom setting and ADHD outcomes were assessed with direct observation using a within-subject comparison design (Table 6). There was a wide variation in the inclusion criteria and the procedure used for ADHD diagnosis, study design, behavior techniques and outcome measures employed in the studies.
Only one study utilized a combination of a structured diagnostic interview with the parent and dimensional threshold on a rating scale for ADHD (McGoey and DuPaul 2000), one study included children who were given a diagnosis of ADHD prior to their study participation with no confirmation of the diagnosis by the study investigators (McCain and Kelley 1993), and four studies included children based on teacher complaints of hyperactive, inattentive, or disruptive behaviors. The two single-case reports and one case series (McCain and Kelley 1993) employed a reversal design, and three of the case series employed multiple baselines across subjects.
The sample sizes were small ranging from 1–4 children. Behavioral techniques used in the studies included self-instructional training to encourage on-task behavior in three studies, and shaping procedure with contingent reinforcement and/or response cost was employed to decrease hyperactive, disruptive, and impulsive behavior in three studies. Information regarding treatment integrity was reported only in one study (McGoey and DuPaul 2000) Treatment effect on number of activity changes, on-task behavior, and disruptive behavior was assessed through classroom observation (3 studies employed a blinded observer). All studies reported improved task-related attention and ability to delay activities and decreased disruptive behaviors.
Only three studies conducted follow-up assessment beyond the immediate intervention period, two studies reported maintenance of gains 2 weeks (McGoey and DuPaul 2000) and 22.5 weeks later (Bornstein and Quevillon 1976) and one study reported increase in attending behavior immediately following treatment, but gains were not maintained at follow-up 2 weeks later (Billings and Wasik 1985).
Behavioral approaches applied to the whole classroom affect all children in the classroom and reduce the sense of “unfairness” of ADHD children receiving special treatment. Whole classroom approaches reduce the burden of designing individual programs for multiple children with problem behaviors in the classroom. Finally, this approach may benefit whole class environment, producing a more positive climate (Conduct Problems Prevention Research Group 1999). In the classroom treatment group in the Barkley et al. (2000) study, the kindergarteners meeting dimensional threshold for hyperactive, impulsive, inattentive, and aggressive scores randomized to the special behavior treatment classrooms showed improvements over children randomized to the regular, no treatment classrooms. However, the positive effects did not generalize beyond the classroom setting and were not evident at two-year followup (Shelton et al. 2000).
There is evidence for short-term efficacy of psychosocial interventions, especially parent behavior training, in reducing disruptive behaviors in preschool children; evidence is much more limited for ADHD outcomes. Four randomized studies employed parent behavior training in preschool children formally diagnosed with ADHD and one of these studies reported improvement in ADHD symptoms. Evidence for efficacy of child training is limited to case series/case reports that included a total of 16 preschool children, only one of these studies employed a structured diagnostic procedure and blinded classroom observation, and reported improvement in ADHD symptoms.
Since it has been postulated that long-term behavioral improvements require psychosocial interventions (Ialongo et al. 1993), preschool years are an especially opportune time to promote appropriate inhibitory control by teaching positive and consistent parenting skills as well as training children directly (Diamond et al. 2007; Dowsett and Livesey 2000).
Here we review four alternative treatment options that may be more practical and/or effective during preschool years than later, may be better justified to try at this age, and may be undertaken in conjunction with standard treatments such as behavioral and pharmacological treatment. As with established treatments, information about alternative treatments is mostly available for school age children.
The quality of evidence ranges from randomized controlled trials to anecdote. In evaluating the evidence, or lack thereof, it is important to consider relative risk, difficulty, and expense. An intervention that is safe, easy, cheap, and sensible (SECS) can be accepted pragmatically on less evidence than one that is risky, difficult, or expensive (in terms of either cost or parental effort and time, which is a resource that should not be squandered on unproven intense treatments).
Of the myriad of alternative treatments advocated for ADHD, four seem especially applicable to preschoolers: elimination diets, vitamin/mineral and other dietary supplementation, vestibular stimulation, and massage. For a review of other popular alternatives, such as EEG and biofeedback that may be more appropriate for older children, see (Arnold 1999; Arnold et al. 2002).
At the time of the 1982 NIH Consensus Development Conference on Defined Diets and Childhood Hyperactivity (1982) most elimination diets (defined diets) were popularly known as Feingold diets. The Feingold (1975) hypothesis had stated that many children are sensitive to dietary salicylates and artificially added colors, flavors, and preservatives, and that eliminating the offending substances from the diet could ameliorate learning and behavior problems including ADHD. Despite a few positive studies (Swanson and Kinsbourne 1980; Williams et al. 1978), most controlled studies were interpreted by the investigators and reviewers as nonsupportive of the hypothesis (Conners 1980; Kavale and Forness 1983; Mattes 1983). The consensus panel called for more controlled research.
Since the 1982 NIH Consensus Development Conference, a literature search revealed 9 peer-reviewed reports on the use of elimination diets (additive free diet [food colors and/or preservatives] and/or few-foods diet) involving preschool children who either had a diagnosis of ADHD or displayed ADHD Symptoms (Table 7). Two studies included preschool children only (Bateman et al. 2004; Kaplan et al. 1989); the other 7 studies included both school age and preschool children with children ranging in age from 1.6–15 years (mean age ranging from 7.3–9.7 years). Only three of these seven studies specified the number of preschoolers participating in the studies (Boris and Mandel 1994; Egger et al. 1985; Rowe and Rowe 1994). A DSM diagnosis of ADHD was the required inclusion criteria in only three of the studies (Carter et al. 1993; Egger et al. 1992; Kaplan et al. 1989); one study was a population-based study (Bateman et al. 2004), and the remaining four studies included children with a DSM diagnosis of ADHD or ADHD symptoms.
Many of the studies did not use rigorous diagnostic procedures. Most studies employed a multi-phase design with open elimination diet (additive free diet and/or few-foods diet), open challenge to identify the incriminated food(s) followed by a randomized, placebo-controlled double-blind challenge with food additives (food colors and/or preservatives) in 5 studies (Bateman et al. 2004; Kaplan et al. 1989; Pollock and Warner 1990; Rowe 1988; Rowe and Rowe 1994) and/or the incriminated food in 4 studies (Boris and Mandel 1994; Carter et al. 1993; 1985; 1992).
In order to maintain the blind, food colors and/or food preservatives were administered in a capsule and/or meals were provided to the study participants. The few-foods or oligoantigenic diet most commonly included two meats (e.g., lamb and chicken), two carbohydrates sources (e.g., potatoes and rice), two fruits (e.g., banana and apple), vegetables (any brassica, e.g., cauliflower, cabbage, broccoli, or Brussels sprouts), cucumber, celery, carrots, parsnip, salt, pepper, water, calcium, and vitamins.
All studies demonstrated either significant improvement compared to a placebo condition or deterioration on placebo-controlled challenge of offending substances. The two studies that included preschoolers only are of particular relevance (Bateman et al. 2004; Kaplan et al. 1989). Both studies investigated additive free diet. Kaplan et al. (1989) conducted a within-subject, placebo-controlled double-blind crossover trial of elimination (additive free) diet and placebo control diet in 24 preschool children (3.56 years of age) with a DSM-III diagnosis of ADHD along with sleep problems and/or other symptoms (e.g., stuffy nose, stomach ache) indicative of food sensitivity. Food was provided for every member of the household. Parent Conners' Abbreviated Symptom Questionnaire ratings were significantly lower during the elimination diet phase compared to the placebo control diet phase (p<.01). Bateman et al. (2004) conducted a population-based, randomized, double-blind, placebo-controlled challenge study of elimination diet (additive free) in 277 3-year-old preschoolers (4 groups of preschoolers with hyperactivity crossed with atopy). Preschoolers in all four groups showed a general increase in parent rated hyperactivity symptoms with artificial food colors and benzoate preservatives, with effect size of d=0.5 compared to placebo challenge. There was no effect of prior levels of hyperactivity or by atopy. Table 7 presents a summary of the available studies on elimination diet that included preschool children with ADHD or ADHD symptoms and monitored ADHD outcome.
A related dietary strategy, simple elimination of sugar or candy, has not garnered convincing scientific support from repeated placebo-controlled acute challenge studies (Ferguson et al. 1986; Krummel et al. 1996; Wender and Solanto 1991; Wolraich et al. 1995) despite a few encouraging reports (Goldman et al. 1986). Even a well-controlled 3-week trial of a sugar-restricted diet found no effect (Wolraich et al.1994). However, Wesnes et al. (2003) did demonstrate in a sample of school children not diagnosed for ADHD that a wholegrain cereal and milk breakfast resulted in fewer inattentive symptoms over the course of the morning than the same number of calories in a glucose drink. It does not appear that sugar or candy restriction alone is a widely applicable treatment for ADHD. On the other hand, sugar is not an essential food group, and there does not appear to be any risk from restriction or elimination of candy and other densely sugared foods.
In summary, there is some evidence for efficacy of elimination diet, especially additive free diet, in preschool children with ADHD. The 4 studies showing efficacy of few-foods diet included a mixed age sample and either did not specify the number of preschoolers or did not report outcome separately for preschoolers, thus making it difficult to assess specific response of the preschoolers to the few-foods diet.
Dietary elimination (additive free diet and/or few-foods diet) may be more practical as well as more effective for preschoolers than for older patients because of better caregiver control of diet, and can be considered when there is a history of formula intolerance, food sensitivity, or general allergic diathesis. It is important to emphasize that an elimination diet trial should be implemented only under the supervision of the child's primary healthcare provider and a nutritionist to ensure that growing preschoolers do not suffer from nutritional deficiencies with the restricted diet. The restricted diet (additive free diet and/or few-foods diet) can be tried for 2 weeks (Egger et al. 1985). If there is no benefit from the restricted diet, it should be discontinued. A stringent elimination diet should not continue for more than 2 weeks without obvious benefit because of the danger of imbalance, especially of calcium and some vitamins. If there is benefit, start adding back the restricted foods weekly (Egger et al. 1985), one food component at a time to identify the problem foods to be excluded from a less restrictive permanent diet.
Unfortunately, there is no research on effects of Recommended Daily Allowance/Recommended Dietary Intake (RDA/RDI) of multivitamins/minerals in diagnosed ADHD children even though some reports suggest mild deficiencies in diet and blood levels that might be addressed. However, in a randomly assigned double-blind placebo-controlled trial of RDA vitamin and mineral supplementation in 47 6-year-old children not selected for ADHD, Benton and Cook (1991) found an 7.6 point IQ advantage (p<.001), mainly based on nonverbal ability increases. They also found increased concentration and decreased fidgeting on a frustrating task (p<.05), and advantage on a reaction time task reflecting sustained attention (Cohen's d=1.3, p<.05). These data warrant a controlled trial in ADHD, although the benefit may be confined to a subgroup with poor diets (Benton 2001).
Regarding a more specific nutrient, Metallinos-Kasaras et al. (2004) found in 3- and 4-year-old children with anemia (and serum lead levels of<50 ppb) not diagnosed with ADHD that iron supplementation (15mg/day) yielded significant improvements in selective attention compared to placebo. There were no effects of iron supplementation in preschoolers who were not anemic. As seen in Table 8, there is one case report in a 3-year-old child (Konofal et al. 2005) and one double-blind, randomized, placebo-controlled trial in 5- to 8-year-old non-anemic children with low serum ferritin levels, diagnosed with ADHD, reporting improvement in ADHD symptoms with iron supplementation (Konofal et al. 2008).
Another nutritional consideration is essential fatty acids, especially omega-3 long-chain polyunsaturated fatty acids. Omega-three deficiency in infants impairs visual attention e.g., (Neuringer 1998). There have been 7 placebo-controlled trials of essential fatty acids relevant to ADHD, all in school-age children.
Five of the seven studies were in children diagnosed with ADHD showing equivocal or no effect in three (Aman et al. 1987; Arnold et al. 1989; Voight et al. 1998), and promising results in two (Sinn and Bryan 2007; Stevens et al. 2003). The other two were in children with dyslexia and developmental coordination disorder, both of which have large overlap with ADHD.
Preschoolers, with their rapid growth/metabolism and smaller bulk for storage of nutrients, may have a special need for nutritional attention. Adequate nutrition becomes even more of a concern when the child is given an appetite-suppressing stimulant for ADHD. It is important in evaluating a preschool child with ADHD to take a careful diet history. If history reveals a diet poor in iron sources, a blood test for iron may be advisable (Konofal et al. 2008). Similarly, if the diet appears unbalanced in other ways, one might suspect other nutritional deficiency.
Recommendation of RDA/RDI multivitamin/minerals is well within the purview of conservative medical practice, at least until the child's diet can be balanced. Consumption of wild ocean fish a couple of times a week is recommended by the American Heart Association to protect against omega-3 deficiency (Kris-Etherton et al. 2002).
Mulligan (1996) reported significant impairment of vestibular processing in 309 children with ADHD compared to 309 matched 4- to 8-year-old children without ADHD (p<0.01). As seen in Table 9, improvement in Conners' teacher ratings from vestibular stimulation compared to a sham condition was reported in two randomized studies in a mixed-age sample (school-age and preschool age children) with symptoms of ADHD (Arnold et al. 1985; Bhatara et al. 1981). Bhatara et al (1981) mentioned that the largest effect was found in the younger children.
Vestibular stimulation is not a proven treatment, but the SECS rule may apply here. The vestibular stimulation of rocking, spinning, piggyback and horsie rides, and swings is a natural environment for preschoolers and can be augmented by sit-and-spin toys, swivel chairs, and rotational games.
The tactile and deep pressure stimulation of massage has been reported to elicit several benefits. Of relevance to ADHD, massage increased on-task behaviors of 3- to 6-year-old autistic children (Escalona et al. 2001), and attentiveness/responsivity and increased vagal activity were associated with increased attention span in 22 preschool children with autism (Field et al. 1997). In adults, it improved math performance (Field et al. 1996), which is sometimes used as an objective outcome measure in pharmacological treatment of ADHD.
The only study in diagnosed ADHD was a randomized, controlled trial in 28 adolescent boys with DSM-III-R ADHD. Massage was reported to reduce teacher-rated Conners' 10-item scale scores from 28 at baseline to 11.3 while the relaxation therapy controls deteriorated from 19.6 to 28.5 (Field et al. 1998). There are no studies on effects of massage in preschool children with ADHD.
Massage of the child by the parent appears to be a safe and cheap intervention that at least should improve parent-child relationship and fits naturally into the cuddling, roughhousing, and other tactile stimulation appropriate for preschoolers. Massage may be especially helpful at bedtime.
Of the four alternative interventions described, only the elimination diet has some convincing evidence at this point, and it is probably applicable to only a minority of children with ADHD (although probably a larger percent of preschoolers than of older children (Dulcan and Benson 1997). However, the other three interventions have some controlled evidence (albeit not conclusive), are safe, easy, cheap, and sensible, seem more widely applicable in preschool children, and can be implemented in conjunction with standard behavioral and/or pharmacological treatment.
Pharmacological intervention studies outnumber nonpsychopharmacological intervention studies. Pharmacological interventions have been studied in comparatively larger samples of preschool children and have tended to use more rigorous methodology than nonpsychopharmacological interventions.
Specifically, MPH demonstrates efficacy compared to placebo in treating ADHD symptoms in preschool children with DSM diagnosis of ADHD in at least five double blind, randomized, controlled, group treatment trials (Barkley 1988; Conners 1975; Greenhill et al. 2006; Musten et al. 1997; Short et al. 2004). Thus for informing clinical practice, adequacy of MPH efficacy data in preschool children is at Level A. In terms of safety data, preschoolers are reported to be sensitive to the adverse effects of MPH with increased rate of irritability, mood changes, withdrawal, and lethargy (Wigal et al. 2006) and a decreased rate of height and weight velocity (Swanson et al. 2006). Additionally, no information about long-term safety and effects of MPH on brain development of preschool children is available. Hence, caution is needed when considering pharmacological treatments in preschool children with ADHD. Information about the efficacy and safety of other stimulants and non-stimulants in preschool children is at Level C.
Among nonpsychopharmacological interventions, there are comparatively more studies for parent behavior training than child training, and evidence is sparse for alternative interventions for treating ADHD in preschool children. Adequacy of parent behavior training (PBT) data is at Level B as shown by improvement in ADHD symptoms with PBT (n=30) compared to a Parent Counseling and Support Group (n=28) and a Waiting List Control Group (n=20) in preschool children meeting DSM-IV diagnostic criteria for ADHD (Sonuga-Barke et al. 2001). Efficacy of child training in preschool children with ADHD was supported in one single case design series that included preschool children with a formal diagnosis of ADHD (N=4) and showed reductions in hyperactivity and disruptive behaviors with both token reinforcement and response cost in a reversal design (McGoey and DuPaul 2000). Hence, adequacy of the child training data is at Level B, the small number of subjects limits the generalizability of the findings.
Adequacy of additive free elimination diet data is at Level B as shown by one double-blind crossover study that included 24 preschool children with a formal diagnosis of ADHD and showed improvement in ADHD symptoms with the elimination diet compared to the placebo diet (Kaplan et al. 1989). The evidence for other alternative treatments is generally at Level C.
In summary, the level of evidence to support short-term treatment of preschool ADHD with MPH is Level A and with parent behavior training, child training and additive-free elimination diet is Level B. It is important to emphasize that the difference between the level of evidence for adequacy of pharmacological and nonpharmacological treatments for preschool ADHD does not necessarily indicate difference in efficacy between these treatments; rather it indicates relative paucity of adequate research with nonpharmacological treatments compared to pharmacological treatments. There is only one study that followed preschoolers, formally diagnosed with ADHD, prospectively for 10 months and reported long-term effectiveness of MPH. There are no studies of comparative and/or combined efficacy or long-term safety of any of the treatment interventions for ADHD in preschoolers.
Given the short- and long-term safety concerns and lack of information about effect of psychopharmacological interventions on brain development of preschoolers, there is a strong clinical consensus that psychosocial interventions should be tried first in preschoolers with ADHD (Dulcan and Benson 1997; Gleason et al. 2007; Kollins et al. 2006). A psychosocial intervention plan should address child's behavior problems both at home and at school. Parent behavior training should be offered to the caregivers, and parents should be encouraged to work with their child's preschool or daycare teacher to integrate coordinated behavior management strategies at home and at preschool or daycare. Direct child training in the classroom can be implemented as indicated. Comorbid disorders should be identified and appropriate work-up and interventions (e.g., speech, language, and communication assessment and treatment for preschoolers presenting with language delays) should be included in the treatment plan. It is important to assess and support treatment and social support needs of the caregivers. If the caregivers believe that their child's behavioral symptoms become worse with food additives and/or certain foods, a careful trial of additive free and/or the restricted diet can be implemented under the supervision of a nutritionist and the child's pediatrician, as described previously.
Pharmacological intervention can be considered when psychosocial intervention has been unsuccessful (Dulcan and Benson 1997) or only partially successful. Care and caution should be exercised in selecting medication dosage for preschool children. Practitioners need to consider the unique sensitivity of preschool children to adverse events and should follow the rule of “start low, go slow” allowing sufficient time on a particular dose to estimate adverse effects and efficacy. At the same time care should be taken to avoid undertreatment with lower doses. Preschoolers should be followed closely for monitoring of possible emergence of adverse effects and dosage adjustment with weekly or biweekly (every other week) visits for the first 1–2 months and then monthly for maintenance visits once the preschooler is on an optimal dose.
Parent and teacher rating scales (e.g., CRS, SNAP, CBCL-1½, ADHD-RS) should be collected for baseline behaviors and repeated regularly for ongoing monitoring of treatment response during follow-up visits.
As mentioned previously, MPH has the best evidence (Greenhill et al. 2006) and is most frequently started at 2.5mg BID and increased to 7.5mg BID or TID over the course of 2–4 weeks depending on the child's response and any side effects. It is important to note that there are a minority of preschoolers who may benefit from 1.25mg TID of MPH; 15% of the preschoolers in the PATS were reported to be best responding to 1.25mg TID, and teachers reported improved ADHD symptoms with 1.25mg TID compared to placebo (Greenhill et al. 2006).
Decision for a BID or TID dose may be based on the child's and family's needs. For example, some parents want their preschooler to take medication only when the child attends school and hence may prefer BID dose instead of the TID dose used in the PATS (Greenhill et al. 2006).
There are no controlled efficacy data for long acting MPH or other psychostimulants or non-stimulants in preschoolers with ADHD. If a preschooler does not respond to MPH, clinicians are left to extrapolate data from older school-age children. Based on school-age ADHD treatment data, if a child does not respond to a trial of one class of stimulants (e.g., MPH), switching to the other class (e.g., amphetamines) is recommended before using another drug class (Arnold et al. 1978; Dulcan and Benson 1997; Elia et al. 1991). There are no empirical data to guide dosing schedules for amphetamines in preschoolers with ADHD; it has been suggested that amphetamines are twice as potent as MPH (Pelham et al. 1999). If a preschooler does not respond to stimulants and/or has unacceptable side effects, a trial of atomoxetine or alpha-agonists is recommended (Gleason et al. 2007). As mentioned previously, there are no controlled efficacy or dose response data for atomoxetine or alpha-agonists in preschoolers with ADHD. Improvement in ADHD symptoms in 22, 5- and 6-year-old children (mean age 6.1±0.58 years) was reported in a prospective open-label trial with 10–40mg/day or 0.47–1.88mg/kg/day of atomoxetine (mean dose=1.25mg/kg/day±0.35mg/kg/day) administered as a single morning dose or BID (morning and afternoon) (Kratochvil et al. 2007). Adverse effects included mood lability in 54.5% and decreased appetite in 50% of the children. Regarding alpha-agonists, there are only 2 case reports of open-label treatment with clonidine (0.025mg TID) and guanfacine (0.25mg BID to 0.5mg BID) both reporting improvement in ADHD symptoms and side effect of sedation early in treatment.
The limited evidence for efficacious treatment options relative to the frequency with which preschool children are referred for treatment of ADHD is striking. Additionally, it is noteworthy that as noted earlier, nonpharmacological treatment investigations lag behind pharmacological treatment studies in preschoolers formally diagnosed with ADHD. This is especially salient since there are short- and long-term safety concerns and there is little information regarding the effect of pharmacological agents on the brain development of preschool children with ADHD. Because of the fewer safety concerns compared to pharmacological treatments, clinicians, caregivers of preschoolers with ADHD, professional organizations making treatment recommendations, and the community at large prefer psychosocial interventions as a first line of treatment for preschool ADHD. This calls for more research to find the best possible treatments matched with parent preferences in order to expand the limited intervention options that are currently available for treating ADHD in preschoolers. There is an urgent need for well-designed, blind, randomized, controlled, between-group treatment trials to study comparative and combined efficacy and safety of psychopharmacological, psychosocial and alternative treatments in well characterized samples of preschool children with ADHD. It is important to study the long-term outcome and safety of treatment interventions and their impact on the developing brain of preschool children with ADHD.
This paper was supported by a grant from the National Institute of Mental Health K23 MH01883 and an award from the Arizona Institute of Mental Health Research to JKG and a grant from National Institute of Mental Health K23 MH01899 to BJA.
Dr. Ghuman has received research funding from Bristol-Myers-Squibb. Dr. Arnold has research funding from Shire, Neuropharm, and Autism Speaks, consults with Shire, Neuropharm, Novartis, Targacept, and Organon, and is on the speaker's bureau for Shire, Novartis, and McNeil. Dr. Anthony has no conflicts of interest or financial ties to report.
This paper is based in part on a symposium presented at the October 2004 Annual Meeting of the American Academy of Child and Adolescent Psychiatry held in Washington, DC. The authors thank Drs. Harinder Ghuman and Alan Gelenberg for their helpful suggestions, support, and encouragement in writing this paper.