Early Preventive Interventions for Attention-Deficit/Hyperactivity Disorder : A Systematic Literature Review

(1)

Early Preventive Interventions for

Attention-Deficit/Hyperactivity

Disorder

A Systematic Literature Review

Ann Thekla Pinkert

One year master thesis 15 credits Supervisor

Interventions in Childhood Andrea Ritosa

Examinator

(2)

SCHOOL OF EDUCATION AND COMMUNICATION (HLK) Jönköping University

Master Thesis 15 credits Interventions in Childhood Spring Semester 2020

ABSTRACT

Author: Ann Thekla Pinkert

Early Preventive Interventions for Attention-Deficit/Hyperactivity Disorder A Systematic Literature Review

Pages: 31

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common psychological disorders in childhood and can lead to many challenges for the children. Several forms of treatment exist, but treat-ment effects are not as impactful as would be desirable. A possibly promising approach for effective intervention is the prevention of the disorder before its onset. The aim of this systematic review was the identification and evaluation of empirically evaluated early preventive interventions for ADHD, imple-mented before school entrance. A systematic database search in different databases resulted in 12 articles. After applying a quality assessment, nine articles, evaluating seven interventions were included in the data analysis. All interventions targeted the behavioral and emotional development of preschool children, with ADHD symptomatology as part of the child problem behavior. All but one intervention resulted in a significant improvement in ADHD symptomatology of the intervention group compared to the control group, supporting the effectiveness of prevention for the development of ADHD. Interventions were most effective when addressing both, psychological factors (by teaching of behavioral and emotional skills), as well as social factors (by teacher and/or parent trainings). However, for most interventions, no long-term outcomes were assessed. Methodological challenges and limitations of this systematic review, as well as possibilities for future research, are addressed.

Keywords: Attention-deficit/hyperactivity disorder, Prevention, Early intervention, Preschool, Childhood, Systematic literature review

Postal address Högskolan för lärande och kommunikation (HLK) Box 1026 551 11 JÖNKÖPING Street address Gjuterigatan 5 Telephone 036–101000 Fax 036162585

(3)

Zusammenfassung

Autor: Ann Thekla Pinkert

Frühe präventive Interventionen für Aufmerksamkeitsdefizit-/Hyperaktivitätsstörung Eine systematische Übersicht der Literatur

Seiten: 31

Aufmerksamkeitsdefizit-/Hyperaktivitätsstörung (ADHS) ist eine der verbreitetsten psychologischen Störungen in der Kindheit und kann zu vielen Beeinträchtigungen im Leben der Kinder führen. Es gibt verschiedene Arten der Behandlung, doch Behandlungseffekte sind nicht so wirkungsvoll wie gewünscht. Ein möglicherweise vielversprechender Ansatz für effektive Intervention ist die Prävention der Störung vor Krankheitsbeginn. Das Ziel dieser Übersichtsarbeit war die Identifikation und Evaluation von empirisch evaluierten frühen präventiven Interventionen für ADHS, durchgeführt vor Schuleintritt. Eine systematische Datenbanksuche resultierte in 12 Artikeln. Nach Einschätzung der Qualität, wurden neun Artikel, die sieben Intervention evaluieren, in die Datenanalyse einbezogen. Alle Interventionen bezogen sich auf die behaviorale und emotionale Entwicklung von Vorschulkindern, mit ADHS als Teil des Problemverhaltens. Alle bis auf eine Intervention führten zu einer signifikanten Verbesserung der ADHS Symptomatik in der Interventionsgruppe im Vergleich zur Kontrollgruppe. Dies spricht für die Wirksamkeit von Prävention für die Entwicklung von ADHS. Interventionen zeigten die größte Wirksamkeit, wenn sie sowohl psychologische Faktoren (durch das Vermitteln von behavioralen und emotionalen Fähigkeiten), als auch soziale Faktoren (durch Lehrer- und/oder Eltern-Trainings) addressierten. Für die meisten Interventionen wurden jedoch keine Langzeitauswirkungen erfasst. Methodische Probleme und Limitationen dieser Übersichtsarbeit, sowohl wie Möglichkeiten für zukünftige Forschung werden diskutiert.

Schlüsselwörter: Aufmerksamkeitsdefizit-/Hyperaktivitätsstörung, Prävention, Frühe Intervention, Vorschule, Kindheit, Systematische Übersicht der Literature

(4)

1 Introduction

Attention‐deficit/hyperactivity disorder (ADHD) is one of the most common psychological disor-ders in childhood and the number of diagnoses has been rising steadily over the past years (Danielson et al., 2017). This neurodevelopmental disorder emerges below the age of 12 and is characterized by persistent patterns of inattention and/or hyperactivity-impulsivity (American Psychiatric Associa-tion, 2013). Children with ADHD can face many challenges in their daily life, like the impairment of global functioning, social behavior, and learning (Posner et al., 2007). Additionally, their parents and family as well as their teachers are often overwhelmed with their behavior and unsure how to best support them. ADHD can also lead to many long-term challenges in the development of children. For example, they are at greater risk to develop comorbid psychopathology, like antisocial behavior or substance use disorders. Furthermore, they are more likely to drop out of high school, have fi-nancial problems, or be unemployed (Barkley et al., 2006).

Most children with ADHD are diagnosed and start receiving intervention at school age (Danielson et al., 2018). However, more and more preschool children also receive the diagnosis (Danielson et al., 2017) and there is emerging literature, that ADHD like symptoms can even already be identified in infants and toddlers (Brown & Harvey, 2019). Therefore, it becomes evident that there is a need for intervention as early as possible. A possibly promising approach is intervening even before the onset of the disorder or before the symptoms are fully developed (Halperin et al., 2012; Sonuga-Barke & Halperin, 2010). Preventive approaches like this are already established and evaluated for some mental disorders, e.g. conduct disorder (Rapee, 2008), but less used and researched for ADHD (Sonuga-Barke et al., 2011).

The present study focuses on the systematic assessment of empirically evaluated early preventive interventions for ADHD.

(7)

2

2 Theoretical Background

2.1 Attention-Deficit/Hyperactivity Disorder (ADHD) 2.1.1 Characteristics and Diagnosis

Attention‐deficit/hyperactivity disorder (ADHD) is a chronic, neurodevelopmental psychological disorder, characterized by persistent patterns of inattention and/or hyperactivity-impulsivity. The manifestation and severity of the symptoms are heterogeneous. It is mostly diagnosed in children but can also pertain to adulthood (American Psychiatric Association (APA), 2013). Prevalence rates vary depending on the country, research methodology, age groups, and changes in diagnostic criteria (Holbrook et al., 2017). A large, recent meta-analysis reports a pooled worldwide prevalence of 7.2% for children under 18 years (Thomas et al., 2015).

Most health care providers decide on the diagnosis of ADHD based on the Diagnostic and Statistical Manual, currently in its fifth edition (DSM-V; APA, 2013). There, the main symptom domains of inattention on the one hand and hyperactivity-impulsivity on the other are differentiated. Examples for inattentive symptoms are when individuals are forgetful or easily distracted, quickly lose focus on tasks, or do not seem to listen when spoken to. Hyperactive-impulsive symptoms, for example, include restlessness, talking excessively, trouble waiting their turn, or interrupting others.

For children up to 16 years to be diagnosed with ADHD, at least six symptoms, which are inappro-priate for the developmental level, must have been present for the last six months. In addition, the symptoms must be present in multiple settings (e.g. home and school) and interfere with the func-tioning of the individual. Furthermore, several symptoms must have been present before the age of 12. Depending on the type of symptoms displayed, three presentations of ADHD are distinguished. Children with the predominantly inattentive presentation show more symptoms of inattentiveness, chil-dren with the predominantly hyperactive-impulsive presentation show more symptoms of hyperactivity-impulsivity, and children with the combined presentation show symptoms of both domains equally. 2.1.2 Developmental Course and Treatment

Historically, ADHD was presumed as a disorder of middle childhood, but research from the last 20 years shows, that the disorder can emerge much earlier (Brown & Harvey, 2019). However, there is still a lot unclear, about the early development of ADHD. Research about the early emergence is complicated due to the slow development of the disorder. In addition, not being able to focus on something for long or to be active and impulsive is developmentally typical for young children and

(8)

3 there is a lack of valid tools to measure uncommon levels of inattentive and hyperactive-impulsive behavior in children younger than four (Miller et al., 2020).

So far, different factors in children as young as infancy (younger than 2 years) were found to predict later ADHD. It is unclear if these are precursors or early symptoms of the disorder (Brown & Har-vey, 2019). One factor linked to later ADHD is difficult infant temperament, with higher levels of reactivity and negative affectivity, and lower levels of regulation and effortful control (Miller et al., 2020). This can, for example, manifest in feeding and sleeping problems. In addition, infants that are later diagnosed with ADHD can show a delay in motor and language development (Gurevitz et al., 2014).

In toddlerhood and preschool-aged children (2 to 6 years) the DSM-V symptoms of inattention and hyperactive-impulsive behavior can already partly be observed (Campbell et al., 2014). Children are likely to be overactive, difficult to calm down, and noncompliant. Furthermore, cognitive and lan-guage delay and impairment of executive functioning and school readiness may be present (DuPaul & Kern, 2011). Problems with peers can emerge, as these children have difficulties following rules or waiting for their turn (Campbell et al., 2014). The American Academy of Pediatrics recommends the assessment and treatment of ADHD starting from age 4, as there are no valid tools for earlier diagnosis (Wolraich et al., 2019). The first line of treatment for children from 4 to 6 years are behav-ioral interventions, i.e. parent training in behavior management and/or behavbehav-ioral classroom inter-ventions. If these do not lead to sufficient improvement, psychopharmacological medication in form of methylphenidate may be considered.

Most children are diagnosed with ADHD in school-age (6 to 12 years; Danielson et al., 2018). In fact, an increase of almost 30% in the number meeting ADHD criteria has been found for the tran-sitioning from preschool to elementary school (Curchack-Lichtin et al., 2014). Therefore, school entry seems to be a turning point for a lot of children with ADHD. There are several possible ex-planations for this (Halperin & Marks, 2019): First, the symptoms might increase as children get older. Secondly, the school context offers many new challenges for children with ADHD. They need to follow more and stricter rules and routines and focus on lessons and tasks for a longer period of time. Thus, the expectations from parents and teachers on the children grow. Inattentive or hyper-active behavior, that might not have been considered as problematic in the more unstructured pre-school context, is considered disruptive and pathological in pre-school. School-age children with ADHD often show disruptive behavior in class, have lower academic achievement and peer problems, and are likely to have impairments in executive functions (Campbell et al., 2014). Treatment includes

(9)

4 behavioral interventions with the family and school along with psychopharmacological medication. In addition, the children need educational interventions for the school context, like instructional and behavioral support (Wolraich et al., 2019).

Reaching adolescence (12 to 18 years), children with ADHD are at risk for several problems (Camp-bell et al., 2014). They are at greater risk than children without ADHD to develop further adolescent psychopathology, like antisocial behavior, substance use disorders, or personality disorders. Moreo-ver, they are more likely to show poorer academic performance, drop out of high school, and have more difficulties with intimate relationships and friendships. For adolescents with ADHD a combi-nation of medication and training or behavioral therapy, along with educational interventions are recommended as treatment (Wolraich et al., 2019).

In general, treatment effects are not as impactful as would be desirable. Only for a minority of chil-dren treatment shows lasting effects throughout the disorder (Molina et al., 2009). Parent training was found to improve the level of non-compliant behavior of the children but has no effect on the core symptoms of ADHD (Daley et al., 2018). Therefore, there is still a need for the development of more efficacious interventions. ADHD can also pertain to adulthood. However, the presentation of symptoms may change, with less obvious hyperactive-impulsive behavior, but more internal feel-ings of restlessness displayed (Campbell et al., 2014). Adults with ADHD are more likely to have lower academic achievement and more financial and occupational difficulties (Barkley et al., 2006). 2.1.3 Developmental Model

Despite extensive research, there is still a lot unknown about the etiology and development of ADHD. Well-established is the theory of a bio-psycho-social model of multiple causal pathways (Pauli-Pott et al., 2019). In their developmental psychopathology framework Campbell et al. (2014) attempt to integrate the findings regarding the etiology of ADHD and propose a dynamic interplay between different causal factors and pathogenic processes. According to the model, interactions between genetic and environmental risk factors cause structural and functional alterations to neural networks, which lead to impairment of neuropsychological functioning. These in turn lead to the development of ADHD symptoms.

Research regarding genetic risk factors is still inconclusive and suggests only small effects of specific genes on the variance in ADHD expression. Furthermore, no interventions targeting genes have been successfully developed yet (Faraone & Mick, 2010). Thus, the focus of intervention lies more on environmental risk factors. Associations to different pre-, peri-, and postnatal risk factors have been found, but the effects are small, and it is difficult to conclude causality (Taylor & Rogers, 2005).

(10)

5 Prenatal factors linked to ADHD are for example maternal smoking and alcohol consumption dur-ing pregnancy, while perinatal factors are prematurity and pregnancy complications (Campbell et al., 2014). A variety of postnatal factors has been found, many related to the family. For example, chil-dren living in socioeconomically disadvantaged families are at higher risk to develop ADHD (Russell et al., 2015). Furthermore, there seems to exist a transactional relationship between ADHD and parenting, with child ADHD symptomatology eliciting negative and harsh responses from parents, which in turn exacerbate the ADHD symptoms (Campbell et al., 2014).

Symptom patterns are highly heterogeneous between individuals and can also change over time within individuals. For example, the clinical presentation of combined, inattentive, or hyperactive-impulsive often changes with age (Campbell et al., 2014). Additionally, early- and late-emerging forms of ADHD can be distinguished. Thus, different developmental phenotypes can be assumed (Sonuga-Barke & Halperin, 2010).

2.2 Prevention

2.2.1 Prevention and Treatment

In the history of preventive sciences, there have been many conceptual and definitional issues, espe-cially regarding the difference between prevention and treatment (National Research Council & In-stitute of Medicine, 2009). The framework developed by the InIn-stitute of Medicine (IOM) in 1994 offers a definition of prevention and treatment for mental health. In this framework, treatment is defined as intervention after the onset of the disorder, i.e. when a disorder could be diagnosed. Prevention, on the other hand, is limited to “interventions that occur before the initial onset of a disorder” (IOM, 1994, p. 23) and whose aim is “the reduction of the occurrence of new cases” (IOM, 1994, p. 26).

However, applied in practice, it is still sometimes difficult to clearly distinguish between the concepts of prevention and treatment. For example, the treatment of a disorder might act as prevention for another disorder in a later developmental stage or the treatment of a mentally ill parent can also be conceptualized as prevention for developmental difficulties of the child (IOM, 1994). Therefore, prevention can be seen as a rather fluid construct (Rapee, 2008).

(11)

6 2.2.2 Classification of Prevention

The classification system most used for prevention of mental health disorders was presented by the IOM in 1994. It differentiates between universal, selective, and indicated prevention, depending on the targeted population Universal prevention targets an entire population without identifying indi-vidual risks, like school-based programs directed at all schools in a city. In order to be profitable, universal programs need to be low in cost per individual, as well as acceptable and with low risk for the population. Selective prevention targets individuals or groups who are at higher risk to develop a disorder than the average population, due to biological, psychological, or social risk factors. Pro-grams can be of moderate cost to be appropriate. Examples are preschool proPro-grams for children from poor neighborhoods or whose parents are divorced. Finally, indicated prevention targets indi-viduals, who possess biological markers for a disorder or are already showing detectable signs or symptoms of the disorder but who do not yet meet the diagnostic criteria. Programs, which only include children who score high on behavioral screening measures are an example here. Indicated preventive interventions are also often called early intervention. They can be profitable even if they are high in costs.

2.2.3 Prevention Research

To adequately evaluate the effects of preventive interventions, Gottfredson et al. (2015) established recommendations for prevention research. Studies should include a sufficient number of participants and at least one control group, that does not receive the intervention. The best standard is a random assignment to the groups. Furthermore, studies should include the assessment of pretest and posttest differences. To evaluate long-term outcomes, preferably follow-up assessments should be included. An interval of at least six months is recommended, but the timing depends on the intervention target. 2.2.4 Prevention of Mental Health Disorders

The idea of intervening in the development of a disorder before it has ever emerged and thus allevi-ating its symptoms or possibly even stop its onset altogether seems very appealing. However, there is still a lot unclear about the possibilities of prevention for mental health disorders. One reason is that an essential requirement for developing an efficacious preventive intervention is a clear devel-opmental model of risk factors (Rapee, 2008). Nevertheless, the current knowledge about psycho-pathological development varies greatly between disorders. For the development of conduct disor-der, for example, many clear risk factors could already be identified and in turn, efficacious

(12)

preven-7 tion programs targeting these established (Slough & McMahon, 2008). For other disorders, like anx-iety and depression, however, risk factors are much less known and the development of preventive interventions more difficult (Rapee, 2008).

2.3 Prevention of ADHD

Regarding the fact, that current treatment for ADHD only shows lasting effects throughout the disorder for a minority of children (Molina et al., 2009), there is a need for new, more efficacious types of intervention. A potentially profitable approach is to use prevention as a mean to intervene as early as possible. Sonuga-Barke and Halperin (2010) emphasize the potential benefits of early intervention or prevention of ADHD for the following reasons: Firstly, brain plasticity is highest in young children and can, therefore, be more easily altered when intervening early. Furthermore, the longer the disorder is untreated, the stronger the child’s impairing behavioral habits can grow. Early intervention can take effect before these habits are formed. Additionally, early intervention can pre-vent the development of negative attitudes of the family regarding the problem behavior of the child. These attitudes can otherwise impede family-based interventions. Lastly, the intervention can take place before the child experiences failure in the school context.

3 Rationale and Aim

Despite the need for alternative types of intervention for ADHD and the promising arguments for the implementation of a preventive approach before the school entry, to the author's knowledge, there are no reviews evaluating existing early preventive programs to date. Therefore, the aim of this systematic literature review is to identify and evaluate existing early preventive interventions for ADHD. The research questions are:

1. What is the form and content of early preventive interventions for ADHD related symptoms? 2. What are the outcomes of early preventive interventions for ADHD related symptoms?

(13)

8

4 Method

4.1 Systematic Literature Review

To identify and evaluate existing early preventive interventions for ADHD a systematic review of the literature was conducted. Systematic reviews are a type of research method, that provides a sys-tematic and transparent way to identify relevant literature, synthesize the findings, and appraise the quality of studies regarding a research question. Thereby, bias related to single studies and non-systematic reviews is aimed to be reduced (Sweet & Moynihan, 2007). The research process is char-acterized by a comprehensive database search of literature following strictly defined inclusion and exclusion criteria (Jesson et al., 2011).

4.2 Search Procedure

The database search for this systematic review was performed in July 2020. To ensure a broad search within different disciplines related to ADHD, the databases CINAHL, ERIC, and PsycINFO were used. CINAHL covers health sciences, ERIC provides educational resources and PsycINFO con-tains literature regarding the field of psychology. Thesaurus search was performed but suggested terms were the same for all databases. Therefore, the same combination of Thesaurus and free search terms were used across databases. The search string was (infan* OR toddler* OR child*) AND ("Atten-tion Deficit Hyperactivity Disorder" OR ADHD OR atten("Atten-tion OR inattenti* OR hyperactiv* OR impulsiv*) AND (“primary prevent*” OR “secondary prevent*” OR “tertiary prevent*” OR “universal prevent*” OR “selective prevent*” OR “indicated prevent*” OR “prevent* intervention*” OR “prevent* program*” OR “early interven-tion*”).

Each database search was limited to peer-reviewed articles and the English language. The search resulted in 897 publications in CINAHL, 402 in ERIC, and 1915 in PsycINFO.

4.3 Inclusion and Exclusion Criteria

Inclusion and exclusion criteria for the selection of articles were defined based on the research aim and questions. This review focuses on preventive interventions for ADHD in young children before entering school. Thus, to be included studies had to evaluate an intervention, which (a) was defined as preventive by the authors, (b) targeted children before elementary school entrance, and (c) targeted ADHD symptomatology (inattention and/or hyperactivity-impulsivity). No limitation was set on the type of prevention, to be able to assess the full picture of preventive interventions.

(14)

9 Additionally, studies had to include the measurement of pre- and post-intervention outcomes of ADHD symptomatology and utilize at least one control group, which received no or an alternative intervention. The precise inclusion and exclusion criteria are listed in Table 1.

Table 1

Inclusion and Exclusion Criteria

Inclusion criteria Exclusion criteria

Availability Full text available in English Only abstract available, any other language

Publication Research articles published in peer- reviewed journals

Reviews, book chapters, reports, theses, missing peer-review

Participants Children before school entrance and without ADHD diagnosis at the start of preventive intervention

School-aged children and/or children with an ADHD diagnosis at the start of the preventive intervention

Intervention Intervention defined as preventive, targeting the development of ADHD symptoms (inattention and/or hyperac-tivity-impulsivity) in children

Intervention specified as treatment

Measure Pre- and post-intervention measurement of ADHD symptomatology

(inattention and/or hyperactivity- impulsivity)

Only broad measurement of externalizing behavior problems

Only post-intervention measurement of ADHD symptomatology

Design Quantitative studies with at least one control group receiving no intervention or alternative preventive intervention At least pre- and post-intervention assessment

Qualitative studies, systematic reviews No control group

Within-subject control group

4.4 Selection Process – Title and Abstract

The database search yielded a total of 3,214 articles. These were imported into the online manage-ment tool Rayyan, which facilitates the screening process for systematic reviews (Ouzzani et al., 2016). First, 742 duplicates were identified and removed. Next, the titles and abstracts of the remain-ing 2,472 articles were screened regardremain-ing the inclusion and exclusion criteria. 2,352 articles were excluded in this step, mostly due to reasons of not addressing an intervention, the wrong population, or not including ADHD symptomatology. The remaining 120 articles were included in the full-text review.

(15)

10

4.5 Selection Process – Full Text

The 120 articles were read as full text and the inclusion and exclusion were applied again. Special focus was put on the method section, where the participants, the study design, the intervention, and the outcome assessments are described. One article was not available as full text, one had been retracted, one was a conference paper, one was a study protocol and one was only a summary. Other reasons for exclusion were the wrong population (N=11), wrong study design with no control group (N=25), or the intervention was not described as prevention (N=14). Most articles were excluded because they did not report outcomes of ADHD symptomatology (N =53). In the end, 12 articles remained for further analysis. The whole selection process is illustrated as a flowchart in Appendix A.

4.6 Data Extraction

In order to prepare the analysis of the data, relevant information from the articles was extracted with the help of an extraction protocol in Microsoft Excel. The protocol is shown in Appendix B. Data was extracted regarding the identification of the paper (i.e. authors, publication year, title, journal and study location), purpose, research questions and hypotheses of the study, participants, and study design. Furthermore, information regarding the aim, content, and delivery of the preventive inter-vention, measurement of outcomes, results, limitations, and implications were extracted. For three studies (Feil et al., 2016; Heinrichs et al., 2017; Raver et al., 2011), not all relevant information was reported in the included articles, as they were already described in other publications. Therefore, information about recruitment, study design, and intervention in these trials were partly extracted from older publications (Feil et al., 2014; Heinrichs et al., 2014; Raver et al., 2009).

4.7 Quality Assessment

To ensure validity and minimize bias, the quality of the studies was assessed with the Quality Assess-ment Tool for Quantitative Studies from the Effective Public Health Practice Project (EPHPP; Jack-son & Waters, 2005). This standardized assessment tool allows the evaluation of quantitative studies on the six methodological dimensions of (a) selection bias, (b) study design, (c) confounders, (d) blinding, (e) data collection methods, and (f) withdrawals and dropouts. Due to the nature of early childhood programs analyzed in this review, where often the parents or teachers deliver the inter-vention and it is often not possible to conceal the interinter-vention condition from the participants, the dimension of blinding was excluded from the quality assessment. All the questions administered in the assessment are shown in Appendix C.

(16)

11 The EPHPP tool includes a dictionary, which provides instructions on how to administer the rating. Each of the five coded dimensions received a rating of strong, moderate, or weak. The global rating for each study was determined by the combination of the component ratings. A study received a strong global score if none of the dimensions was rated as weak and a moderate global score if one weak rating was administered. Studies with two or more weak ratings received a weak global score and were excluded from further analysis due to a significant risk of bias.

From the 12 articles appraised for quality, six studies received a strong global rating, and three studies a moderate global rating. Three studies were rated as weak due to the risk of selection bias, not controlling for confounders or not reporting dropouts, and were therefore excluded from further analysis.

4.8 Ethical Considerations

As the present study did not include primary data of participants, no ethical review or approval was necessary. However, due to the sensitive nature of research about mental health, especially when including young children, the ethical principles from the Declaration of Helsinki were adhered to (World Medical Association, 2013). The declaration applies to all medical research involving human subjects and ensures the health, well-being, safety, and rights of all participants. All included studies had to take ethical considerations into account. This was ensured on the one hand, by only including peer-reviewed articles. On the other hand, all studies included the collection of informed consent from the parents, four studies mentioned the approval of a review board, and one study referred to the Declaration of Helsinki.

4.9 Data Analysis

Nine articles were included in the final data analysis. Each article was assigned an identification num-ber, that was used for reference in the following. First, a general overview of the studies was prepared (shown in Table 2). Then, to answer the first research question, more detailed information about the preventive interventions was collected and synthesized (shown in Table 3,4 and 5). The interventions were categorized regarding their characteristics and similarities and differences were emphasized. To answer the second research question, the intervention outcomes regarding ADHD symptomatology were analyzed and compared (shown in Table 6). The focus was put on whether the intervention yielded significant results.

(17)

12

5 Results

5.1 Overview of Included Studies

Through the selection process and quality assessment, nine articles were identified, that met all the inclusion criteria and were judged to be of adequate quality. An overview of the studies is given in Table 2. All were quantitative studies utilizing at least a pre-post-intervention design with a control group, to evaluate preventive interventions targeting the emotional and/or behavioral development, including ADHD symptomatology, in preschool children. The articles were published between 2000 and 2017 in journals related to psychology, child development, and prevention sciences. Six of the studies were implemented in the USA (1,2,3,4,8,9), while two studies were set in Germany (6,7) and one in Norway (5).

Regarding the study design, all the studies utilized a control group, who received care as usual. Seven of the studies implemented a randomized control trial (RCT) design (1,2,3,4,6,7,8), one had a quasi-experimental design (5) and one did not describe a method of randomization (9). For five studies (4,5,6,8,9) only pre-and post-intervention results are reported. One study also includes a 4-year fol-low-up (7). The remaining three articles report post-intervention (1), 1-year (2), and 4-year (3) follow-up results from the same data set. Thus, the nine included articles describe findings from seven original data sets.

In all studies, the participants were preschool children with a total age range of 2 to 6 years recruited via their preschools. In five articles (1,2,3,8,9) the recruitment was limited to classrooms which were part of Head Start. This is a large US American program established in 1965, offering educational, nutritional, health, social, and other services to promote the school readiness of children from low-income families (Office of Head Start, 2020). The other four studies did not limit their recruitment to specific family characteristics. However, in two studies (4,6) interventions were targeted at chil-dren with externalizing problem behavior. The gender of the participants was distributed approxi-mately equally in most studies. Just in the two studies targeting children with externalizing behavior (4,6), there was a much higher percentage of boys.

In the nine articles, seven different preventive interventions are evaluated (a different intervention for each original data set). To refer to the intervention implemented in the articles related to the same original data set (1,2,3), the identification number 123 will be used in the following. None of the interventions targeted ADHD specifically or was developed for children younger than three.

(18)

13 Table 2

Overview of Included Articles.

ID* Authors & Year

Country Participants in Total (Age & Gender)

Family Characteristics Intervention Quality Assessment 1 Bierman et al.,

2008 USA 356 preschool children, age 3–5, 54% female

Low-income families, 40% single parents, 31% with no high school degree, 79% in unskilled or semi-skilled labor categories, 25% African American, 17% Hispanic, 58% European American

PATHS (Promoting Alternative Thinking Strategies) as part of Head Start REDI (Research based, Developmentally Informed)

Strong

2 Bierman et al., 2014

USA Follow-up of study 1 in kindergarten

Strong

3 Nix et al.,

2016 USA Follow-up of study 1 from kindergarten to third grade Strong 4 Feil et al., 2016 USA 45 preschool children, age 3–5, 36% female Low-income families, 36% African American, 49% European American

Preschool First Step (PFS) Strong

5 Fossum et al., 2017

Norway 1049 preschool children, age 3–6, 49% female

Not reported Incredible Years Teacher Class-room Management Training (IY TCM)

Strong

6 Hanisch et al.,

2010 Germany 155 preschool children, age 3–6, 25% female

Not reported Prevention Programme for

Exter-nalizing Problem Behaviour (PEP) Moderate

(19)

14 Table 2 (continued).

ID Authors &

Year Country Participants in Total (Age & Gender) Family Characteristics Intervention Quality Assessment 7 Heinrichs et al., ₂₀₁₇ Germany 280 preschool _{children, age 2–6,}

49% female

Low-income families, 22% single parents, 11% migration background

Triple P Parent Training Moderate

8 Raver et al., 2011 USA 542 preschool children, age 2–6, 52% female Low-income families, 68% single parents, 66% African American, 26% Hispanic

Chicago School Readiness Project

(CSRP) Strong

9 Serna et al.,

2000 USA 84 preschool children, age 3–5, 44% female

71% Hispanic, 12% African American, 10% European American, 7% Native American

Self-determination intervention Moderate

(20)

15

5.2 Characteristics of Preventive Interventions 5.2.1 Target Population and Prevention Level

An overview of the intervention characteristics is given in Table 3. All seven preventive interven-tions targeted preschool children. According to the classification system of the IOM (1994), the majority of the interventions used a selective preventive strategy, as they were implemented in low-income Head Start preschool classes (123,8,9). Two programs did not limit their recruitment to certain characteristics and were, therefore, universal (5,7). Two studies did only include children, who were either identified as showing externalizing behavior by their teachers (4) or scored high on a screening instrument for externalizing behavior (6). Thus, these studies implemented indicated preventive strategies. Note, that some authors considered the prevention level as universal when the intervention was implemented in low-income Head Start classrooms. However, this is not in accordance with the IOM’s classification system (1994) used in the present study.

5.2.2 Delivery

Regarding the delivery of the preventive interventions, two interventions (123,5) were school-based, i.e. teachers received training to implement the program in school. One intervention (7) was parent-based, i.e. parents received training regarding parenting strategies. The remaining programs implemented either a combination of school and parent-based strategies (6), school-based strate-gies and an interventionist directly working with the children (8), or a combination of all three components (4,9). Interventionists were a trained behavioral coach (4), trained mental health con-sultants (8), or a trained masters-level preschool teacher (9).

The duration and frequency of the interventions differed as well. The longer interventions were implemented for a whole school year (123,8) or 6 months (5), while the shorter interventions lasted 12 weeks (9), 10 weeks (6), or 8 weeks (4,7). The frequency ranged from daily sessions (4) or two times a week (9) to weekly (123,4,6,8,9), bi-weekly (7), or monthly (5) sessions. The sessions lasted between 90min or a day long.

(21)

16 Table 3

Overview of Intervention Characteristics

Prevention Level

Aim Intervention ID Delivered to children by

Duration & Frequency

Universal Promote emotional and behavioral development

Incredible Years Teacher Classroom Management Training (IY TCM)

5 Teachers 6 months,

monthly 7-hour sessions with teachers Triple P Parent

Training

7 Parents 8 weeks,

bi-weekly 2-hour sessions with parents Selective Promote emotional and

behavioral development in children from low- income families

PATHS (as part of Head Start

REDI) 123 Teachers One school year, 3-day teacher training in beginning, 1-day booster training midterm, weekly lessons for children Chicago School Readiness Project

(CSRP) 8 Teachers and interventionist One school year, 30-hour teacher training across 3 months, weekly sessions by interventionist

Self-determination

intervention 9 Teachers, parents, and interventionist

12 weeks,

two 3-hour sessions per week in class, monthly sessions with parents

Indicated Identify and support children with externaliz-ing behavior

Preschool First Step (PFS) 4 Teachers, parents, and interventionist

8 weeks,

1-day teacher training in beginning, 30-day implementation by interventionist, weekly session with parents

Prevention Programme for

Exter-nalizing Problem Behaviour (PEP) 6 Teachers and parents 10 weeks, weekly 90min to 120min workshops with teachers and parents

(22)

17 5.2.3 Aim

All preventive interventions generally aimed to promote the emotional and behavioral development of preschool children. More specifically, the interventions implementing a selective preventive strategy (123,8,9) aimed to prevent the development of emotional and behavioral difficulties and increase school readiness in children from low-income families. One of them (123) also targeted the development of language/emergent literacy skills, but as this is not the focus of the present study, this will not be described further. The indicated preventive interventions (4,6) aimed to iden-tify high-risk children with externalizing behavior, including inattention and hyperactivity-impul-sivity, to improve their behavior and prevent further maladaptive development.

5.2.4 Activities

The preventive interventions implemented different activities to deliver the content, which are shown in Table 4. The majority of the interventions implemented a group-based teacher training in the form of a workshop (123,4,5,6,8). Another activity included in most programs was some form of mentoring by interventionists for the parents or teachers throughout the program (123,4,6,7,8). There, the teachers or parents received feedback and could ask questions, either in meetings in person or via telephone. Three interventions utilized a group-based parent training in the form of a workshop (6,7,9), while one implemented one-on-one meetings with the parents in their home (4). Other activities only included in one intervention were stress reduction workshops for teachers (8), curriculum-based lessons and activities implemented by the teachers (123), story-based teaching from an interventionist with the assistance of the teachers (9), and group or indi-vidual therapy by a mental health consultant for the children with the highest rate of emotional and behavioral problems (8).

(23)

18 Table 4

Intervention activities.

Activities ID: 123 4 5 6 7 8 9 Teacher training in a group-based workshop X X X X X Parent training in group-based workshop X X X One-on-one meetings with the parents X

Stress reduction workshops for the teachers X Curriculum-based weekly lessons and

extension activities X

Mentoring from an interventionist X X X X X Interventionists teaching children in a

story-based format with the assistance of teacher X Group therapy sessions by an interventionist for

the children with the highest rate of emotional and behavioral problems

X

Individual therapy sessions by an interventionist for the children with the highest rate of emotional and behavioral problems

X

5.2.5 Content

To achieve the aim to prevent the development of emotional and behavioral difficulties in pre-school children, different content was delivered to the parents, teachers, or children in the inter-ventions and listed in Table 5 below. The studies differed in regard to how detailed the content was described. While some only mentioned the general topics, others included detailed descrip-tions of all subtopics. Only the content that was explicitly named is reported in the present study. Most of the studies, except for two (6,7), described the delivery of strategies to promote certain skills in the children. More than half of these interventions included problem-solving skills (123,4,5,9), while prosocial friendship skills were promoted in three interventions (1,2,3 and 4,5). Skills, that were targeted in two interventions were communication and cooperation (4,9), emo-tional regulation (123,5), and self-control (123,8). Limit setting and self-confidence were each only promoted in one intervention (4). It can be seen that the PATHS intervention (123) and the Pre-school First Step program (4) promoted the most strategies, with each targeting more than half of the skills.

(24)

19 Table 5

Content of Interventions.

Content ID: 123 4 5 6 7 8 9 Strategies to promote the children’s:

Prosocial friendship skills X X X

Communication and cooperation X X Emotional regulation X X

Limit setting X

Self-control X X

Problem-solving X X X X Self-confidence X

Principles of positive classroom management X X X X X

Defining rules X X X

Strategies to effectively communicate commands X X Strategies to build positive relationship

with the child

X X Positive reinforcement of appropriate child

behavior (e.g. praise, rewards)

X X X X X X Adequate negative consequences for inappropriate

child behavior (e.g. ignoring, redirecting)

X X X Etiological and maintaining factors for child

problem behavior

X Strategies for stress reduction for teachers X Strategies for parental self-regulation X

All studies but one (8) mentioned the delivery of strategies for positive reinforcement of appropri-ate child behavior. This included, for example, verbal praise, rewards, or a formal motivational system with charts or graphs.

Strategies to adequately react to inappropriate child behaviorwere described for three interventions (5,6,7). These included negative consequences like ignoring mild inappropriate behavior or redi-recting the child. Only two studies explicitly mentioned strategies to build a positive relationship with the child (5,6).

All five of the interventions that included a teacher training, delivered principles of positive class-room management (123,4,5,6,8). Other strategies mentioned multiply were defining rules (4,6,7)

(25)

20 and how to communicate commands effectively and adequately (6,7). Psychoeducation regarding the etiological and maintaining factors for child problem behavior (7), strategies for stress reduction for teachers (8), and strategies for parental self-regulation (7) were only described in one interven-tion each.

5.3 Outcomes of Preventive Interventions

Across the studies, multiple instruments and methods were used to measure a variety of constructs including child problem and adaptive behavior, child learning behavior and pre-academic skills, parenting behavior, parental well-being, and classroom characteristics. None of the interventions was designed to specifically address ADHD symptomatology, but this was assessed in all studies as part of the child problem behavior. As the present study focuses on ADHD, only those out-comes are reported and presented in Table 6 below.

5.3.1 Measures of ADHD Symptomatology

To measure ADHD symptomatology, all studies used some form of a rating scale. These were either directly developed for ADHD based on the DSM (1,2,3,4,6,7) or included a subscale as-sessing both inattentive and hyperactive-impulsive behavior (4,8,9) or only inattention (5). In the three studies related to the same original data set (1,2,3), in the first two studies parents and teachers completed the rating scale, while in the third study only teachers’ ratings were included. One other study (4) reported parent- and teacher-rated outcomes for a measure specifically for ADHD symp-tomatology, but only teacher-rated outcomes for another measure assessing inattentive and hyper-active-impulsive behavior. Two studies only reported outcomes from teacher rating (5,9), while for the parent-based intervention (7) only parents rated the child behavior. Two studies reported out-comes rated by an independent observer, who implemented a standardized test regarding academic skills (6) or on executive functioning and effortful control (8) and afterward rated the inattentive and hyperactive behavior of the child throughout the test.

(26)

21 Table 6

Intervention Outcomes.

Intervention ID Measure Teacher-rated Parent-rated Observer-rated

PATHS (as part of

Head Start REDI) 1 ADHD Rating Scale (DuPaul, 1991) No significant effect No significant effect

2 ADHD Rating Scale

(DuPaul, 1991) Intervention effect pre to 1-year follow up, but only in schools with many low- achieving students (p < .01)

Intervention effect pre to 1-year follow up, but only in schools with many low-achieving students (p < .05)

3 ADHD Rating Scale

(DuPaul, 1991)

Children in IG were more likely than chil-dren in CG to follow the most optimal low–stable developmental trajectory of attention problems from pre to 4-year fol-low up (36% vs. 26%, p < .05, OR = 1.60) Preschool First

Step (PFS) 4 Conners’ ADHD Scales (CADS; Conners, 1999) Social Skills Improvement System Rating Scales (SSiS; (Gresham & Elliott, 2008) Subscale Hyperactivity/ Inattention

Children in IG were more likely than children in CG to improve on CADS (54% vs. 11%, p = .002, OR = 9.9) Large intervention effect on SSiS (p = <.001, Hedges' g = 1.10)

Children in IG were more likely than children in CG to improve on CADS (62% vs. 21%, p = .002, OR = 8.7) Incredible Years Teacher Classroom Management Train-ing (IY TCM)

5 Child Teacher Report Form

(C-TRF; Achenbach & Rescorla, 2000) Subscale Attention

Small intervention effect (p < .01, d = 0.14)

(27)

22 Table 6 (continued).

Intervention ID Measure Teacher-rated Parent-rated Observer-rated

Prevention Pro-gramme for Exter-nalizing Problem Behaviour (PEP)

6 Problem Checklist

Atten-tion-Deficit/Hyperactivity Disorder (PCL ADHD; Döpfner & Lehmkuhl, 2000)

No significant effect

Triple P Parent

Training 7 Problem Checklist Atten-tion-Deficit/Hyperactivity Disorder (PCL ADHD; Döpfner & Lehmkuhl, 2000)

Intervention effect pre to 4- year follow-up for mothers’ rating (p = .010) No significant effect on fathers’ ratings Chicago School Readiness Project (CSRP) 8 Preschool Self-Regulation

Assessor Report (PSRA; Smith-Donald et al., 2007) Subscale Attention/ Impulsive control Small to medium inter-vention effect (p < .05, d = 0.43) Self-determination intervention 9 IOWA Conners

(Loney & Milich, 1982) Subscale Inattention/ Overactivity

Small to medium intervention effect (p < .001)

IG improved pre to post (p < .000,

d = .56) and CG worsened (p < .023, d = −.40)

(28)

23 5.3.2 Outcomes of ADHD Symptomatology

Different methods of data analysis were implemented in the studies and the level of detail of the reported results varied. Therefore, the outcomes are not directly comparable between studies. Only the post-intervention results are displayed in Table 6 and effect sizes are only reported for studies that included them.

Regarding the pre-intervention measurement, all but one study (9) showed no significant difference in ADHD symptomatology between the groups. In study number 9, the intervention group showed higher levels of ADHD symptomatology before the implementation of the intervention than the control group.

Overall, only for one intervention, no significant intervention effect was found (6). In four studies (4,5,8,9), children in the intervention group improved significantly more on the measure of ADHD symptomatology than children in the control group. Reported effect sizes were small (5), small to medium (8,9), or large (4). In one study (7), children significantly improved on the mothers’ rating of symptomatology but not on the fathers’. Regarding the three studies related to the same original data set, the post-intervention assessment showed no significant effect (1). However, an interven-tion effect was found from pre-interveninterven-tion to 1-year follow up (2), but only for children in schools with many low-achieving students. In schools with few low-achieving students, no effect was found. For the 4-year follow up assessment (3), developmental trajectories for ADHD sympto-matology were calculated. Three different trajectories were found: (a) a low-stable trajectory, re-flecting low post-intervention scores, that did not change in follow-up assessments, (b) a moderate-increasing trajectory, reflecting moderate post-intervention scores, that increased over the follow-ing years, and (c) a high-increasfollow-ing trajectory, reflectfollow-ing high post-intervention scores, that also increased over follow-up assessments. Children in the intervention group were significantly more likely to follow the most optimal low–stable trajectory than children in the control group.

(29)

24

6 Discussion

There is a need for new, more efficacious interventions for ADHD. Preventive approaches, like already adopted for other mental health disorders, sound promising but are not well-researched for ADHD. Therefore, the aim of this systematic review was to identify and evaluate existing early preventive interventions for ADHD. In the selection process, nine studies evaluating seven pre-ventive interventions were identified. All the interventions target the behavioral and emotional development of preschool students and not ADHD specifically. However, ADHD symptomatol-ogy (inattention and/or hyperactivity-impulsivity) is included as part of the child problem behavior in all studies. All but one intervention resulted in a significant improvement in ADHD sympto-matology for children in the intervention group compared to children in the control group.

6.1 Reflections on Findings

Although the preventive interventions all have the same general aim of preventing difficulties in the behavioral and emotional development of preschool children, they nonetheless differ in their targeted population, prevention level, delivery, activities, and content. In the following, these inter-vention characteristics are set in relation to the outcomes of the interinter-ventions.

6.1.1 Target Population and Preventive Level

All three levels of prevention from the IOM (1994) are present, with two programs adopting a universal approach, three selective preventive programs targeting low-income Head Start class-rooms, and two indicated preventive interventions targeting children with high levels of external-izing behavior. There are possible advantages of all the three different approaches for the preven-tion of ADHD.

Universal interventions offer the possibility to include all children, irrespective of individual risk. Thereby, children who might not be identified as at risk (at least according to the current level of knowledge), but would still later develop ADHD, are targeted as well. Regarding the outcomes of the two universal programs in this review, Incredible Years TCM (Fossum et al., 2017) and Triple P (Heinrichs et al., 2017), both resulted in an intervention effect, with children in the intervention group significantly improving more in ADHD symptomatology than their control counterparts. For Triple P, however, this occurred only in mothers’ ratings of symptomatology and not in fa-thers’. Therefore, the effect has to be interpreted carefully. A possible reason for the difference in

(30)

25 parental ratings is that fathers are often less involved in childcare than mothers. Furthermore, fa-thers showed a lower participation rate in the Triple P parent training than mofa-thers. While for Triple P there is no effect size reported, the effect was only small for the Incredible Years TCM. This is, nevertheless, in accordance with expected effect sizes for universal programs, as they are usually implemented with a lot of participants (1049 for Incredibly Years TCM), of whom the majority is not affected by the targeted problem (Smedler et al., 2015).

Low socioeconomic status was the selection criteria in all three selective preventive interventions reviewed in the present study, the PATHS program (Bierman et al., 2008), CSRP (Raver et al., 2011), and the self-determination intervention (Serna et al., 2000). Socioeconomic status has been identified as one of the major environmental risk factors for ADHD (Russell et al., 2015). There-fore, regarding the bio-psycho-social model of multiple causal pathways for the development of ADHD (Pauli-Pott et al., 2019), the three selective interventions are based on environmental social risk factors. They all resulted in an improvement in ADHD symptomatology for the intervention group. For the PATHS intervention, however, a significant improvement could only be detected at the follow-up assessments one and four years later. In addition, at the 1-year follow, only children in schools with many low-achieving students showed a significant improvement. This might sug-gest, that skills gained in the intervention were more helpful for children in more adverse school contexts, and children in more protective school environments did not need these skills as much (Bierman et al., 2014).

Indicated interventions offer the possibility to target children, who already show signs of a disorder, but do not yet meet the diagnostic criteria. The indicated interventions in the present review, the PFS (Feil et al., 2016) and the PEP program (Hanisch et al., 2010), target children exhibiting high levels of externalizing behavior, including inattention and hyperactivity-impulsivity. Thus, they are the interventions targeting ADHD most closely. While the PFS program showed a large interven-tion effect, there were no significant findings for the PEP program.

None of the interventions target children younger than two years old. This is not surprising, as the review only includes studies, in which ADHD symptomatology of inattention and/or hyperactiv-ity-impulsivity was assessed pre- and postintervention. Although precursors or early symptoms of ADHD can already be identified in infants, they do not show a clear picture of symptomatology (Gurevitz et al., 2014; Miller et al., 2020). Additionally, none of the interventions target ADHD specifically. A possible reason is, that the development of efficacious preventive interventions re-quires a clear developmental model of risk factors (Rapee, 2008). ADHD, however, is one of the

(31)

26 disorders, for which there is still a lack of comprehensive knowledge about etiology and develop-ment (Pauli-Pott et al., 2019). Therefore, the developdevelop-ment of a prevention program only for ADHD might prove to be difficult.

6.1.2 Delivery, Activities, and Content

The preventive interventions vary greatly in their intensity. While some only implement a few hours of training workshops, others include weekly activities for a whole school year. Not surprisingly, the two universal programs (Fossum et al., 2017; Heinrichs et al., 2017) are amongst the ones with the least intensity, with 42 respectively 8 hours of workshops in total. This is in accordance with the requirement for universal interventions to be cost-effective (IOM, 1994). Contrary to expecta-tions, however, the indicated PEP program (Hanisch et al., 2010) is one with the least frequency as well, with only 15 to 20 hours of workshops. This might not be extensive enough for children already exhibiting behavioral problems and might be a reason for the nonsignificant results of this intervention.

Four out of seven interventions include the parents, either in group-based parent training or one-on-one meetings. Although the content is not the same in the interventions, they all include the teaching of some principles of positive parenting, like positive reinforcement of appropriate be-havior of the child or defining rules. This is in line with the importance of parenting as a risk factor for the development of ADHD. Targeting parenting behavior can interrupt the negative cycles of parent-child interaction that exacerbate ADHD symptoms (Campbell et al., 2014). Outcomes of the interventions including a parent-based component in this review support this theory, with three out of four resulting in an improvement for the intervention group. On the other hand, the great majority with six out of seven interventions, include teachers. School-based approaches like this are common in interventions for young children. Teachers are often important support figures for young children and the teacher-child relationship was found to influence the emotional and behav-ioral development of children (Webster-Stratton & Reid, 2004). Especially effective, nonpunitive classroom management has positive effects on children’s behavior. All reviewed interventions im-plementing teacher training, include the delivery of strategies for positive classroom management. The majority also delivered strategies for positive reinforcement in the form of rewards or motiva-tional systems. Again, all but one of the interventions including a teacher-based component led to an improvement in child symptomatology of the intervention group and therefore support the importance of teachers in the children’s development. Including parents and/or teachers is also in

(32)

27 line with the treatment of preschool children diagnosed with ADHD, where the first line of treat-ment are behavioral interventions like parent training or behavioral classroom interventions (Wolraich et al., 2019). In fact, the two programs implementing a universal preventive approach in this review, Incredible Years TCM (Fossum et al., 2017) and Triple P (Heinrichs et al., 2017), have been established as treatment programs for preschool ADHD (Halperin & Marks, 2019).

Apart from the parenting and teaching strategies, another major category of content are strategies to promote certain psychological, i.e. behavioral and emotional skills in the children, like prosocial friendship skills, emotion regulation, and problem-solving. Therefore, regarding the bio-psycho-social model (Pauli-Pott et al., 2019), most interventions address both bio-psycho-social as well as psychological factors of development. Biological factors are, however, not addressed. The two interventions not including strategies to promote children’s skills, Triple P (Heinrichs et al., 2017) and PEP (Hanisch et al., 2010), did not produce very conclusive outcomes. While the first led to an improvement of symptomatology only in mothers’ rating and not fathers’, the latter did not show any significant findings. This suggests that preventive interventions for ADHD symptomatology are more suc-cessful when including a combination of promoting certain skills in the children and delivering parenting and/or teaching strategies, thus addressing psychological as well as social risk factors. In summary, the findings suggest, that preventive interventions promoting the behavioral and emo-tional development of preschool children, can be effective in improving ADHD symptomatology. Regarding the form and content, interventions are most effective, when targeting psychological and social risk factors of the bio-psycho-social model of development, while also including teachers and/or parents. Psychological factors are addressed by teaching children behavioral and emotional skills, while social factors are addressed by teacher and/or parent trainings that include principles of positive classroom management and positive parenting. These findings are promising for the application of prevention. Individuals with ADHD can face a lot of long-term challenges and their whole life can be impacted. Intervening early with the effective preventive programs reviewed in this study can change the developmental course of the disorder and possibly prevent negative long-term outcomes. This would offer children at risk for ADHD a whole new, promising perspective and can affect future lives.

(33)

28

6.2 Methodological Challenges and Limitations

There are several methodological challenges, that limit the interpretation of the findings of this review. Firstly, the results of the studies are mostly not directly comparable, as different methods of measurement and data analysis were implemented and effect sizes are not always reported. Ad-ditionally, the preventive interventions were only implemented in three countries, the USA, Nor-way, and Germany. Therefore, the findings might not be generalizable to other countries.

Seven out of nine studies included teacher and parent ratings for the measurement of ADHD symptomology. However, as common for early childhood programs, the parents and teachers also delivered the intervention to the children and therefore knew if the child they rated was in the intervention group. Thus, their assessment might have been biased. The inclusion of different sources for ratings is a way to minimize bias. Only three out of nine studies included a combination of parent and teacher ratings. In all of these studies (Bierman et al., 2008, 2014; Feil et al., 2016) parents and teachers agreed in their ratings, which supports the validity of the outcomes. For two interventions, PEP (Hanisch et al., 2010) and CSRP (Raver et al., 2011), the assessment of blinded, independent observers was implemented. This allows a much more objective measurement and therefore represents a higher methodological standard (Daley et al., 2018). Thus, if the other inter-ventions would have included independent observers as well, the effects might have been smaller or not significant, like it was the case for the outcomes of the PEP. The CSRP led to a small to medium intervention effect even on observers’ ratings, which supports the effectiveness of the program.

Regarding the recommendations for prevention research from Gottfredson et al. (2015) for a suf-ficient number of participants, two of the studies (Feil et al., 2016; Serna et al., 2000) had a sample size of less than 100, which is relatively small for a preventive trial. While the PFS program (Feil et al., 2016) is an indicated intervention, where smaller sample sizes are expected, the inclusion of only 45 participants still limits the generalizability of the results. The results of the selective pre-ventive self-determination intervention (Serna et al., 2000), on the other hand, have to be inter-preted even more carefully. Apart from only including 84 participants, there is also no randomiza-tion described, which is the best standard for preventive research. Addirandomiza-tionally, the intervenrandomiza-tion and control group already showed a pretest difference in ADHD symptomatology, which was not controlled in the data analysis. Thus, a regression to the mean might be part of the small to medium intervention effect, that was found (Smedler et al., 2015).

(34)

29 Another recommendation from Gottfredson et al. (2015) is the implementation of follow-up as-sessments at least six months after the intervention. Only for the PATHS program (Bierman et al., 2008, 2014; Nix et al., 2016) and Triple P (Heinrichs et al., 2017), follow-up assessments were included. Both were implemented up to four years after the intervention, which allows a long-term evaluation after school entrance. Interestingly, for the PATHS program, no significant intervention effect was found until one year after the intervention. No estimation of long-term effects can be made for the other interventions and therefore it is unknown if some outcomes may have only become evident later in the development of the children as well.

Regarding the content and activities of the interventions, only explicitly mentioned concepts were reported in this review. However, the level of detail in the description varied, and especially for the longer interventions, it cannot be expected that the articles can give a comprehensive description of the content. Therefore, it is safe to assume, that the interventions included concepts, not men-tioned in this review.

This review was performed by a single reviewer. Thus, it is possible, that the results are biased. Peer-review, especially for the selection process and quality assessment of the studies, would have been a better methodological approach. Additionally, the inclusion and exclusion criteria were rel-atively strict and only included empirically evaluated programs. It is possible and likely, that there are other prevention approaches in practice, especially on individual child level.

6.3 Future Research

Although all but one of the preventive interventions led to an improvement of ADHD symptomol-ogy in the children, there is a clear need for more longitudinal research to assess long-term out-comes. Furthermore, no empirically evaluated intervention targeting children younger than two or targeting ADHD specifically were identified in this systematic review. It is clear, that there is still a lack of knowledge regarding the developmental model of ADHD, and future research about causal factors in the form of longitudinal birth cohort studies is necessary. However, several pre-, peri-, and early postnatal factors have already been identified (Campbell et al., 2014) and concepts of prevention for these proposed (Halperin et al., 2012). Instead of focusing on the symptom pattern of inattention and/or hyperactivity-impulsivity, that does not become evident until after the age of two, another approach for prevention programs could be to address early precursors of ADHD. This can include prenatal maternal smoking or alcohol consumption (Campbell et al., 2014) or difficult temperament (Miller et al., 2020) or motor and language delay in infants (Gurevitz et al.,

Early Preventive Interventions for Attention-Deficit/Hyperactivity Disorder : A Systematic Literature Review