Psychostimulants

Psychostimulants continue to provide control of ADHD symptoms and are generally well tolerated for months to years at a time. Concerns about exacerbation of tics with stimulants appear to be unfounded, although sample size in studies of tics remain small and this may result in a type II error. Some of the long-term research summarizes information based on short-acting formulations of psychostimulants, requiring multiple doses daily. The Barbaresi⁵⁹ study, for instance, reports that MPH is better tolerated than dextroamphetamine (DEX). However, direct comparison of once-daily agents, for example, OROS MPH and MAS XR is can be difficult. For example, the Hoare, et al.⁶⁰ study of OROS MPH included adolescents and those with ADHD inattentive type (ADHD-I), whereas the McGough, et al.⁶³ study of a MAS XR sample had more than 90 percent of participants with ADHD Combined type (ADHD-C). Comparison could be read as suggestive that OROS MPH is better tolerated than MAS XR, but both studies had 15 percent of participants withdraw because of adverse events. Also the methods for collecting adverse events may have been more sensitive in McGough, et al., as they were collected by both spontaneous reports and by investigator inquiry.⁶³ It is also possible that the Hoare, et al., study offered participants relatively less effective dose, thereby diminishing the likelihood of adverse events.⁶⁰ The agents have not been compared in the same long-term (over 12 months) trial and therefore, it is not possible to make direct comparisons of effectiveness and safety or tolerability.

Atomoxetine

Long-term extension trials show that ATX is both safe and effective for ADHD symptoms in children and teens over 12 to 18 months. The research examining its use considers global functional assessments as well as ADHD symptom change. In contrast to studies of other agents, the research offers direct comparison with placebo for examination of relapse prevention, offering evidence that benefits are maintained following discontinuation.^66,67,69 An important caveat to these statements appears in Newcorn, et. al.,⁷⁹ a study not meeting criteria for this review as the total length of treatment and followup was less than 12 months. This study compared effect sizes for ATX with OROS MPH and documented the psychostimulant as more efficacious than ATX for ADHD symptom control. Adler, et al.,⁶⁸ offer the only study of a pharmacologic intervention over an extended time period in adults with ADHD.

Guanfacine Extended Release

Open-label extension trials of GXR show it to be effective and generally safe.^70,71 Parents report benefit in reduced ADHD symptoms and global improvement for a substantial number of children and teens with ADHD. Somnolence, headache, and fatigue appear to interfere with its use, but these adverse events appear to diminish following several months of treatment, although this may be due to discontinuation by those who do not tolerate the agent.⁷⁰ Substantially fewer children completed the 12-month extension trial on GXR monotherapy than completed the psychostimulant trials and the ATX trials reviewed, suggesting less overall effectiveness and tolerability. Fewer adverse events are reported and adherence improved with concurrent administration of psychostimulants.⁷¹ These observations may also reflect improved symptom control.

Adverse Events

We examined studies regarding three areas of adverse events that required the use of articles that were not clinical trials comparing two or more interventions. The studies examined growth rates in comparison to standardized norms and rates of hospital and emergency department use for cardiac events and cerebrovascular events, such as cerebrovascular accidents (CVAs) and Transient Ischemic attacks (TIAs). In this review, the safety, tolerability, and adverse events of pharmacological agents is reported within the context of clinical trials, the information appears where the clinical trials of the specific agent are described.

Psychostimulant Medication Compared With Combination of Psychostimulant Medication and Psychosocial and/or Behavioral Treatment

The studies examining combined PBT and school or daycare interventions for children with ADHD suggest that adding classroom teacher consultation may be of greater importance for children in low SES communities, rather than for families with educated parents who live in communities with resources.^27,122,143 As a group, these studies offered some information about the benefits of PBT over a full school year, but also documented that many disadvantaged families do not attend PBT sessions even when transportation and babysitting are available.²⁷ When parents attend, children benefit.⁴⁰ One recent German study offered quality evidence about combining teacher behavior training and direct child training with and without PBT.⁴⁰ Synergies among some, but not all, aspects of the program were noted, and some benefits lasted a year beyond discontinuation of the intervention program. Additional studies of this type will confirm the best means of offering interventions, as well as which children to target.

Three cohorts were identified that examined stimulant medication and/or combined medication and psychosocial or behavioral treatment. One of these was a study in China,⁷⁷ and two were in North America,^{73,74,160,171} including the followup cohort extension study of the Multimodal treatment (MTA) study of ADHD, the largest RCT to date examining combinations of interventions.⁷³ The results from these three cohorts indicate that both psychostimulants and combined psychostimulants and behavioral treatment are effective in treating ADHD plus ODD symptoms in children, and also anxiety, primarily boys ages 7 to 9 years of normal intelligence with combined type of ADHD, especially during the first 2 years of treatment. Overall, the MTA study suggests that combined therapy may have a slight advantage over medication management during the first 14 months, and a clear advantage over behavior treatment,^72,165 especially for children with multiple comorbidities.⁸⁰ However, combined treatment is equivalent to medication alone in controlling ADHD and ODD symptoms for up to 2 years if the child shows an early favorable response to medication.⁷⁶ The MTA study also suggests that these two strategies may be superior to psychosocial/behavioral treatment alone or community care during the first 2 years,^73,74,169 although psychosocial/behavioral treatment is equally effective as treatments with psychostimulants for ADHD children with comorbid anxiety disorder during the first 14 months.⁸⁰ Combination therapy and medication management are effective in reducing ODD during the first 2 years of treatment,⁷⁵ and superior to psychosocial/behavioral treatment and Community Care.^73,74 It appears that psychosocial/behavioral treatment reduces the risk of substance use for 10 months following the intervention, but the effect appears to disappear by 22 months.⁸³ However a re-analysis of the data adjusting outcome for age, suggested that the reduced risk for substance use following behavioral intervention was maintained at 3 years. These results were formally presented, but not published (Molina, October 2010). No treatment strategy is clearly superior in reducing other comorbid psychiatric disorders at 14 months or 3 years.^81,168

Combining medication with psychosocial/behavioral treatment may reduce the dose of medication required, improve retention of patients in treatment, and improve positive parenting. So, et al., in a study involving Chinese children, set the mean daily dose of stimulant medication to less than half that used in the MTA study, and many fewer families who were offered medication alone continued in care.⁷⁷ However, there may be genetic and cultural differences between samples studied that make direct comparison with children in North America complex. Abikoff ‘s 2004 study suggests that it may be cost-effective to treat stimulant-responsive children free of learning and conduct problems with medication alone, although families in both groups had frequent contact with clinicians.⁷⁶ Treatment with psychostimulants, intensive behavioral treatment or combination of the two can reduce negative parenting, but combined treatment may be the most effective in improving positive parenting.^89,161–163 Too few long-term studies examining combinations of medication management and psychosocial/behavioral interventions are available to clarify what subgroups of children do best with which interventions. For some subgroups, multiple interventions are synergistic, but perhaps not for all. Synergies may result in improved effectiveness due to increased treatment adherence, continuity of care, and proactive approaches to new onset of mental health concerns over extended periods of time.

Using intention to treat analyses, the MTA study suggests a loss of superiority of any individual intervention 2 years after treatment has ended.¹⁶⁰ However, secondary analyses such as mixed effects models, propensity score analysis, and growth mixture model analysis have provided additional findings. These secondary studies document that most children with ADHD receiving any of the interventions generally maintained improvement for up to 8 years, while a small proportion began to worsen after the interventions discontinued. On the other hand, while most of the children experienced improved symptoms and functioning, they did not reach levels of functioning comparable to their nonclinical community peers.⁸²

We also examined longitudinal cohort studies that followed children for multiple years following initial treatment. The outcomes and time frames varied extensively across studies. Biederman, et al.,⁸⁶ and Wilens, et al.,¹⁸¹ studied an exclusively female cohort, and all others studied an exclusively or predominantly male sample. Although any conclusions can only be seen as preliminary, it appears that stimulant medication might protect against psychiatric disorders (e.g., ODD, CD, depression, anxiety disorder) at 10 years. Some studies suggest that stimulant medication reduces substance use disorders in late adolescence or adulthood,^87,88,181 while one paper reported no benefit.¹⁷⁹ Two studies suggested that stimulant medication may protect against nicotine use.^176,181 Treatment with stimulant medication, especially at an early age, may delay the onset of smoking and reduce substance use disorder.^88,177,180 Given the challenges inherent in pursuing long-term outcomes studies, with lack of ability to control for co-interventions and significant life events, such information can only be seen as hypothesis generating.

We found three reports on two cohorts that examined academic achievement as the primary outcome following classroom-based interventions. Other studies reported on academic outcome as one of multiple secondary outcome measures. The review of the academic outcomes with long-term followup of treatment interventions revealed benefits, albeit limited, with medication interventions in some aspects of reading and arithmetical skills.^86,174,288 Combining psychobehavioral and academic skills interventions with medication offers no additional gains than medication alone, at least for children with ADHD without comorbid learning disabilities. Interventions for academic skills in classroom-based programs result in academic enhancement, but the findings support the need for sustained intervention to improve academic functioning over time.^91,92,175

Preschool Interventions

While the overall evidence for preschool interventions is strongest for PBT for disruptive behavior including ADHD, very few RCTs offer information about PBT interventions designed specifically for preschoolers with ADHD. Despite this, seven of the eight PBT intervention studies documented improvement in ADHD symptoms. We chose to emphasize similarities among manualized PBT programs, although differences are also noted. Further research will be required to document whether the programs as currently running are successful in addressing aspects of functional impairment due to ADHD symptoms. Although short-term trials show the efficacy of PBT, evidence for lasting benefits are less robust. While it appears that PBT benefits may last several years, no extension study included untreated comparison groups, and attrition over the followup period ranged from 24 percent at 18 months²⁶ to 54 percent at 3 to 6 years,^21,29 limiting interpretation of the results.

Investigations of psychostimulant medication use in preschoolers are generally short-term trials with very small samples. The PATS study addresses a number of important methodological and clinical concerns, examining the potential additional benefit of medication following a series of 10 PBT sessions. Careful attention to details regarding adverse events and the impact of these on medication adherence offers clear information about long-term effectiveness and safety. Interestingly, clinicians documented improved global functioning concurrently with parents noting increased mood problems.⁵¹ While parent and teacher ADHD symptom scales measuring dysfunction noted improvement, those measuring strengths as well as weaknesses in behavior showed no overall behavioral benefit from the addition of stimulant medication. The PATS study offers information about both the potential benefits and limitations of stimulant medication use in young children. Limitations are: 1) younger children experience more dose related adverse events than older children, 2) stimulants interfere with rates of growth, and 3) not all parents agree with ongoing use following medication titration.^7,53,54 Also, the presence of three or more comorbid conditions interfered with the effectiveness of psychostimulant medication following PBT.⁵² Only 54 percent of those initially enrolled in the study opted to enter the medication titration component following PBT, suggesting that parent preferences play an important role in providing optimum care for young children with ADHD.

Future work should examine the appropriate place of PBT as a specific intervention for ADHD in preschoolers. A focus of such studies should include different SES and ethnocultural groups, as well as the presence of comorbid conditions in the children. Adverse events are not discussed in reports of PBT trials or teacher training/classroom intervention trials. Outcomes examined should include global functioning and school readiness as well as behavior symptom counts. Specific attention to the circumstances surrounding parental reluctance to engage in treatment or parent attrition from PBT is warranted as that appears to be a primary barrier to success. Additional awareness and understanding of parent preferences may be especially important in this age group.

Extended Studies

Studies conducted over long periods of time face challenges in controlling for many confounders which may affect the outcomes studied. Several of these longer-term studies either did not enroll representatives from lower SES at risk for psychosocial adversity or those who were less able to be contacted for followup. Some studies did not systematically collect or report important confounders, such as socioeconomic demographics, family psychiatric history, childhood abuse, adherence to treatment, or co-interventions. The retrospective studies face problems with recall and documentation bias, both of which prospective longitudinal studies face as well if the time intervals between data collection are lengthy. An important challenge is the documentation of treatment adherence and co-interventions, both formal and informal, which affect treatment outcomes.

A considerable limitation to evaluating academic outcomes following interventions is that classroom-based or teacher consultation-based interventions are by nature difficult to investigate, as it can be challenging to coordinate cross-sector research and to develop informative comparison interventions that are ethically acceptable. In addition, few of the studies reviewed controlled for learning disabilities and IQ, important confounding factors for academic outcomes in an ADHD population. Additional aspects to consider in future studies will be the challenges inherent in coordinating and tracking the co-interventions offered in school settings along with those offered in health care settings.

The most commonly studied population in the extended interventions studies were children, primarily boys, ages 7 to -9 years, with ADHD-C at the time of documented treatment. It is not clear whether the same intervention outcomes apply to community samples across different geographical regions, cultures, and to both genders, other ADHD subtypes, and different age groups. In addition, for the most rigorous studies, there was no comparison group of children with untreated ADHD, as this would be an ethical challenge. It is therefore difficult to be fully confident that the improvements seen over time were due to treatment effects rather than subsequent co-interventions, maturation or other unmeasured effects.

A major gap in the available literature is the lack of clinical trials and extensions of clinical trials examining non-pharmacological interventions targeting the functional impairment associated with ADHD symptoms in a variety of sample populations.

Prevalence and Health Services Studies

Determining prevalence of ADHD across all age categories in the population is necessary to understand the burden that the condition poses. From this, we can identify gaps in service and develop responses which will help patients and their families in the shorter-term and allow patients to meet their potential in all areas of their lives, such as maintaining fulfilling relationships and finding success in school and workplace environments. There are several methodological factors that influence the calculation of prevalence estimates – namely, the diagnostic criteria employed, along with the informant type, and the data source.²⁸⁹ As described by a recent systematic review/meta regression of the worldwide prevalence of ADHD,⁹³ key methodological differences between studies accounted for much of the variability in the pooled prevalence estimate, highlighting the need for a standardized, methodological approach in order to improve comparability of estimates and epidemiological trends reported over time and in different geographical areas.

To date, the prevalence of ADHD among both adolescents and adults is not well delineated in the literature. Adolescents tend to be subsumed under children, though the burden in this age group may well be different and/or incorrectly approximated by current diagnostic methods. It is also unclear whether the diagnostic criteria are appropriate for use with adults. University-aged individuals with ADHD may be worth examining further, as a special group. Other special populations that warrant further interest include diverse cultural groups and/or ethnic minorities, and other vulnerable groups such as immigrants and families of low SES.

To develop an understanding of who is identified and treated for ADHD in community practice, the types of data used most frequently were epidemiologic surveys and administrative claims and prescription databases. The first type of data is limited to relatively smaller numbers of volunteers, although specific research questions about risk and protective factors can be asked.

The administrative claims database is limited in the sense that it represents only services reimbursed whereas the prescription database takes into account only those who use prescription medication. Nevertheless, each provides a depiction of what happens in community practice, identifying enrollee characteristics as well as clinician diagnosis and treatment plans, or in the case of prescription databases, dispensed medication. Similar to epidemiology studies for prevalence, issues of case identification, informant, quality of interventions, and outcome measures limit interpretations of the results. For the purposes of understanding who is receiving what kind of treatment, a significant shortcoming of the current literature is the lack of information on other forms of treatment for ADHD besides the use of psychostimulants or other medications. This renders the task of capturing all aspects of treatment use difficult. In addition to addressing this gap, more attention should be paid to uncovering whether or not certain groups (e.g., those of lower SES, ethnic minorities, children in foster care, or those living in more isolated or rural areas) are being under-recognized and/or under treated for ADHD.

Some of the potential vulnerable groups appear to be identified and prescribed medication, if not actually treated, to a greater degree than the norm. Overall, the rates of identification and treatment with ADHD medications is high in the United States relative to other areas globally, and higher in some regions of the United States than others, raising issues about the possibility that some practitioners are identifying too many children and youth, while others may be identifying too few. Evidence suggests that it is not only characteristics of the patients but also characteristics of the providers that influences rates of diagnosis and medication treatment. Patterns suggest that cultural biases exist suggesting that increased information about patient preferences could improve the match between what interventions are offered and what treatments are accepted. As it currently stands, many sufferers may be identified, but a large proportion of those in need do not utilize the treatments offered, even if they can be accessed.