Components of cognitive–behavioural therapy for mitigating core symptoms in attention-deficit hyperactivity disorder: a systematic review and network meta-analysis

Abstract

Question

Cognitive–behavioural therapy (CBT) is frequently implemented for individuals with attention-deficit hyperactivity disorder (ADHD). It is still unknown which specific components are effective, because CBT is a complex intervention with several components. The objective of this review was to assess the efficacy of CBT components for ADHD.

Study selection and analysis

Building on definitions of the components used in CBT strategies, we conducted network meta-analyses to evaluate the efficacy at both the treatment level (ie, combinations of these components) and the component level (ie, individual component). Primary outcome was treatment response of ADHD core symptoms. Secondary outcomes were total ADHD symptoms, inattention symptoms, hyperactivity-impulsivity symptoms and dropout from the treatment. We searched MEDLINE (via PubMed), Embase, PsycINFO, ClinicalTrials.gov and Cochrane Library for literature published up to 31 March 2022. This review included only randomised controlled trials.

Findings

43 trials with 3817 participants were included in the network meta-analyses. Third-wave therapy (OR=4.80, 95% credible interval (CrI) 2.50 to 9.10), behaviour therapy (OR=3.50, 95% CrI 1.70 to 7.30) and CBT (OR=3.10, 95% CrI 1.70 to 5.70) were treatments that demonstrated superior efficacy compared with placebo. The component-level analyses showed organisational strategies (incremental OR (iOR)=2.03, 95% CI 1.27 to 3.24) and third-wave components (iOR=1.95, 95% CI 1.30 to 2.93) were associated with increase in treatment response. Problem-solving techniques (incremental standardised mean difference=0.42, 95% CI 0.01 to 0.83) were associated with a reduction in inattention symptoms.

Conclusions

Our findings suggest that the optimal treatment package for ADHD may include organisational strategies, third-wave components and problem-solving techniques.

PROSPERO registration number

CRD42022323898.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Previous pairwise meta-analyses have demonstrated the efficacy of cognitive–behavioural therapy (CBT) in alleviating core symptoms of attention-deficit hyperactivity disorder (ADHD).

WHAT THIS STUDY ADDS

• This study provides the first evidence of a treatment ranking and promising components for mitigating core symptoms of ADHD by conducting network meta-analyses at both the treatment level and the component level within CBT.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Our findings may be useful guide for decision-making.

Background

Attention-deficit hyperactivity disorder (ADHD) is a neurodevelopmental disorder characterised by symptoms of inattention and hyperactivity-impulsivity due to impairment of brain functions.¹ People with ADHD experience various challenges in both childhood and adulthood, which can adversely affect various aspects of daily life stability, maintenance of interpersonal relationships and career development.²

The current evidence-based guidelines for ADHD treatment recommend physiological approaches such as pharmacological interventions³; although these can improve inhibitory function to some extent, complete remission is rare, and the core symptoms of ADHD persist even with physiological treatment in most cases.4,6 The National Institute for Health and Care Excellence guideline also recommends psychotherapy, especially cognitive–behavioural therapy (CBT) which has been reported to be effective in improving the core symptoms of ADHD.7,9

CBT is a complex intervention that includes various cognitive and behavioural components, and therapists assess the symptoms and characteristics of each patient individually and prescribe CBT techniques to address their problems accordingly. Thus, determining which CBT components are beneficial or which ones are detrimental for ADHD is crucial for optimising treatment. However, discussions regarding which components of CBT are more effective or potentially harmful in the context of the diverse challenges associated with ADHD have not progressed appreciably.

Objective

To discuss the efficacy of CBT for ADHD, it is important to integrate evidence on the specific effects of CBT components on core symptoms of ADHD and to examine, both theoretically and clinically, whether specific components of CBT are superior to others in the treatment of ADHD. The objective of this study was to assess the efficacy of CBT components for ADHD.

Study selection and analysis

Protocols and registration

This review protocol was formulated through discussions among all authors (KM, SH, YK and YM). The protocol for this review was registered in PROSPERO (registration number: CRD42022323898), was published as a protocol paper¹⁰ and revisions were presented in online supplemental appendix 1. This review was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Network Meta-Analyses.¹¹ Table 1 presents a list of interventions related to various levels of CBT, along with a description of the components frequently provided in the treatment. The requirements and definitions for the interventions and components included in these lists were jointly determined by two authors (KM and SH), CBT experts.

Table 1. Definitions of components and treatment-level interventions.

Treatment and component	Definition and description
Treatments
Psychoeducation	Offering fundamental information about ADHD to individuals and/or their parents. Component combinations: pe±pl±sup.
Behaviour therapy	Treatments focused on inappropriate behaviour change, based on learning psychology. The intervention includes organisational strategies for stimulus manipulation and reinforcement, and applied behaviour analysis. Progressive muscle relaxation and breathing techniques are now part of behaviour therapy, given their history in psychotherapy. Component combinations: or±pl/pe/prob/relax/sup.
Cognitive therapy	Cognitive therapy is a therapeutic approach that aims to modify negative thoughts or irrational beliefs in order to produce corresponding changes in adaptive behaviours and positive emotional states. Component combinations: cr±pl/pe/prob/relax/sup.
Third-wave therapy	The third-wave component was designed to enhance mindful engagement, acceptance, value recognition, alignment and cognitive flexibility. Intervention approaches encompass mindfulness meditation, acceptance and commitment therapy, dialectical behaviour therapy, mindfulness-based cognitive therapy, functional analytical psychotherapy, meta-cognitive therapy and several others. Attention training is a specific treatment technique aimed at improving meta-cognition. Component combinations: 3w±pl/pe/or/cr/sup.
CBT	Cognitive–behavioural therapy is a psychotherapeutic approach founded on principles of behavioural and cognitive theories. It encompasses both behavioural and cognitive techniques and additionally incorporates brief third-wave techniques such as mindfulness. Component combinations: or+cr±pl/pe/3w/prob/relax/sup.
Problem-solving therapy	Problem-solving therapy involves a process of identifying the factor of specific problems and experimentally testing potential solutions. Component combinations: prob±pl/pe/relax/sup.
Placebo control	Placebo control was defined as a condition where participants might experience a positive psychological response due to the intervention itself. The waitlist condition, where participants have the expectation of receiving a promising new treatment in the future, was also considered a potential source of positive response.31 We included the active control condition, which provided minimal levels of psychoeducation and relaxation, within the placebo control category. We included a pill placebo. Component combinations: pl/relax.
Treatment as usual	Standard care and treatment as usual served as a nocebo control. Participants who expect a new intervention and instead receive standard care may become disillusioned, believing that their condition will not improve. This can lead to a nocebo effect on the outcomes in participants allocated treatment as usual.31 This included facilitated group sessions or supportive care components. Component combinations: tau/sup±pl.
Components
Psychoeducation	Psychoeducation (pe) involves providing information about ADHD. This component was delivered though leaflets or self-help books.
Organisational strategies	Organisational strategies (or) were a set of techniques based on applied behaviour analysis, involving manipulation of stimuli and reinforcement schedules. The component included techniques like organisation, planning, time management and applied behaviour analysis.
Cognitive restructuring	Cognitive restructuring (cr) was a technique to identify and modify irrational or maladaptive thoughts using strategies such as Socratic questioning, thought recording and guided imagery.
Third-wave components	The third-wave component (3w) was designed to enhance mindful engagement, acceptance, value recognition, alignment and cognitive flexibility. Intervention approaches encompass mindfulness meditation, acceptance and commitment therapy, dialectical behaviour therapy, mindfulness-based cognitive therapy, functional analytical psychotherapy, meta-cognitive therapy and several others.
Problem-solving techniques	Problem-solving technique (prob) assists people in breaking down a problem and considering specific solutions.
Relaxation techniques	Relaxation techniques (relax), such as progressive muscle relaxation and breathing exercise, are components that regulate automimic nervous system function through a process of desensitisation.
Supportive components	Supportive care (sup) involves a supportive stance, facilitating client self-resolution through empathy, reassurance and guidance.
Placebo components	Interventions that have the potential to induce positive responses have been defined as the placebo component (pl).
Nocebo components	Interventions that have the potential to induce negative responses have been defined as the nocebo components.31 In this study, treatment as usual (tau) was a typical nocebo.

Symbols: ‘+’ means ‘and’; ‘±’ means ‘with or without’; ‘/’ means ‘and/or’.

ADHD, attention-deficit hyperactivity disorder; CBT, cognitive–behavioural therapy.

Eligibility criteria

The inclusion criteria were as follows: studies must include participants with ADHD, be published in English before 29 February 2024, employ a CBT-based intervention, have participants aged between 10 and 60 years and use a randomised controlled design. The use of pharmacotherapy was permitted if it was balanced across groups within the study. Studies were only included if a fixed combination of CBT components was mandated for each treatment group. The exclusion criteria were as follows: control condition differed solely in the duration of the intervention while delivering identical CBT components to the intervention group, the intervention or control condition incorporated specialised interventions such as neurofeedback or specific educational programmes, or if all participants presented with any physical disorders.

Literature search strategy

We searched MEDLINE (via PubMed), Embase, PsycINFO, ClinicalTrials.gov and Cochrane Library databases from database inception to 31 March 2022, with an updated search on 29 February 2024. The particulars of the search terms are furnished in online supplemental appendix 2. To include more randomised controlled trials (RCTs) in this review, the reference lists of several systematic reviews with/without a meta-analysis were checked.12,14 Grey literature was not searched.

Outcomes

The primary outcome was a dichotomous variable indicating clinically meaningful change, defined as treatment response. If the number or percentage of treatment responders was reported in the selected trials, we extracted these data. When trials lacked the data on treatment responders, we estimated response rates using the normal approximation method described by Furukawa et al,¹⁵ considering preintervention and postintervention outcome changes, SD, sample size and a 95% CI. Response rates were imputed using Jacobson and Truax’s¹⁶ reliable change index. We extracted data at baseline and at the assessment time point closest to 6 months after intervention. Secondary outcomes were assessed using measures of total ADHD symptoms, inattention symptoms and hyperactivity-impulsivity symptoms, in addition to collecting data on dropout from the treatment. The dropout rate was a binary variable, while the other secondary outcomes were continuous variables. For detailed information on the outcome measures used to assess ADHD’s core symptom and those priority ranking, please refer to online supplemental appendix 3.

Data extraction procedure

The first and second authors (KM and SH) checked the titles and abstracts (in the first round) and full texts (in the second round) of all the articles for extraction of the relevant data on the characteristics of the included studies. If there were multiple assessment points, data from the time point closest to 6 months from the start of the intervention were used. When multiple raters were involved assessing the symptom, clinician rating was prioritised, followed by maternal ratings. The corresponding authors were contacted for studies for which relevant data were unavailable. If we did not hear back, we contacted them again after 1 month.

Grading of evidence

The first and second authors (KM and SH) independently assessed the risk of bias in each included study using the revised Cochrane risk-of-bias tool (RoB 2) for randomised trials.¹⁷ The inconsistencies for risk of bias in each trial were resolved through discussions among the reviewers. The RoB 2 was rated as low, high or some concerns in each of the following five domains: randomisation process, deviations from the intended interventions, missing outcome data, measurement of the outcome and selection of the reported results.

Statistical analysis

Treatment-level network meta-analysis

To assess transitivity, boxplots of potential effect modifiers (publication year, pharmacotherapy, age, comorbidity, sex and treatment duration) were generated to examine their distributions across treatment comparisons visually. The Confidence In Network Meta-Analysis (CINeMA) was used to evaluate confidence of the results from network meta-analysis.¹⁸ Publication bias was assessed for all interventions and for 10 or more specific comparisons using a visual approach with contour-enhanced funnel plots and Egger and colleagues’ method.¹⁹ A network diagram of treatment level was constructed to assess the connections within the network for the primary outcome. Incoherence was evaluated using global network methods by using a design-by-treatment test and local network methods by using the node-splitting method.²⁰ Divergent estimates obtained from the direct and indirect evidence provided by the node-splitting method indicate the presence of incoherence. We conducted a Bayesian network meta-analysis of random effects at the treatment level. To estimate the results of the network meta-analysis, Markov chain Monte Carlo (MCMC) sampling was implemented to infer the posterior distributions of the parameters. The number of burn-in iterations (n.adapt), actual simulation iterations (n.iter) and the thin arguments was set to 500; 10 000; and 10; respectively. To assess model convergence, potential scale reduction factor (PSRF) was referenced using Gelman-Rubin plots. The PSRF for a modelled parameter value of less than 1.05 is generally considered indicative of convergence. To investigate which treatment was the most effective, surface under the cumulative ranking (SUCRA) scores were computed as a metric to evaluate the effectiveness of each treatment within the generated network. A forest plot was generated to present the estimated effect size (OR) and 95% credible intervals (CrIs) for treatment responses at the treatment level, with placebo serving as the reference. The statistical analysis was conducted using the ‘gemtc’ package in R.²¹

Component network meta-analysis

We conducted frequentist component network meta-analysis based on an additive model.²² In other words, we assumed no interaction among the CBT components, and the intervention effects were the sum of the effects of the included components. For each component, we estimated the incremental OR (iOR) for binary outcomes and the incremental standardised mean difference (iSMD) for continuous outcomes measuring each dimension of ADHD core symptoms. To demonstrate additional benefits and harms by incorporating each component into the intervention, we colour coded corresponding 95% CIs for each estimate using a shading scheme—where colour represents the effect and shading indicates the strength of evidence.²³ To quantify the effects of different component combinations, we computed ORs and their 95% CIs relative to the placebo control and estimated the incremental effect of each combination. The statistical analysis was conducted using the ‘netmeta’ package in R.²⁴

Sensitivity analysis

Two sensitivity analyses were implemented. First, for treatment level, a frequentist network meta-analysis was implemented by using the ‘netmeta’ package in R.²⁴ Second, for component level, we also performed sensitivity analyses by excluding small studies (n<30), the trials with dropout rates >30% and the trials with participants without a formal ADHD diagnosis. We performed subgroup analyses by age group for adolescents and adults. We also performed a sensitivity analysis that used estimated values for treatment responders for all trials.

Findings

Study selection

We identified 18 658 studies, assessed 139 full texts and included 43 trials with a total of 3817 participants. The 43 studies included in this review comprised a total of 92 arms. Flow diagrams detailing the study selection process for this review are provided in online supplemental appendix 4. The list of included and excluded studies is presented in online supplemental appendices 5 and 6. The participants were generally young adults (mean age 29.9, SD=10.1) who were undergoing pharmacotherapy and had severe ADHD symptoms (table 2). The most common CBT components were psychoeducation (50 of 92 arms (54.3%)), organisational strategies (43 (46.7%)), cognitive restructuring (23 (25.0%)), third-wave components (22 (23.9%)), problem-solving techniques (20 (21.7%)) and relaxation (11 (12.0%)). The detailed characteristics of the included studies and each component in each group are summarised in online supplemental appendices 7 and 8.

Table 2. Characteristics of included participants and trials.

Characteristics	Value	n
Participant		3817
Age, mean (SD), year	29.5 (10.1)
Adolescents, n (%)	43.2%	1648
Adults, n (%)	56.8%	2169
Sex, n (%)
Male	51.3%	1920 of 3745*
Female	48.7%	1825
Medicine usage	60.3%	2184 of 3619†
Comorbidity	45.8%	1346 of 2938‡
Trial characteristics
Region
	North America (USA)	12
	Europe	23
	East Asia	5
	Middle East Asia	2
	Oceania	1
	South America	0
	Africa	0
Number of arms		92
	Two-arm trial	38
	Three-arm trial	4
	Four-arm trial	1
Publication year, median (range)	2018 (1992–2024)	–
Treatments		92
	CBT	23
	Third-wave therapy	15
	Behaviour therapy	12
	Psychoeducation	4
	Problem-solving therapy	3
	Cognitive therapy	1
	Placebo control	16
	Treatment as usual	18
Components
	Psychoeducation	50
	Organisational strategies	43
	Cognitive restructuring	23
	Third-wave components	22
	Problem-solving techniques	20
	Relaxation techniques	11
	Placebo components	14
	Nocebo components	18
	Supportive components	56

^*Number of participants in trials without information on gender ratio was excluded from the calculation.

^†Number of participants in trials without information for ADHD medicine was excluded from the calculation.

^‡Number of participants in trials without information for comorbidity was excluded from the calculation.

ADHD, attention-dicit hyperactivity disorder; CBT, cognitive–behavioural therapy.

Transitivity assessment

The publication period ranged from 1992 to 2024 (median, 2018). 31 studies reported comorbidities, with 1346 out of 2938 participants (45.8%) having at least one coexisting mental disorder, such as major depressive disorder or anxiety disorder. The treatment duration ranged from 4 to 20 weeks (median, 11 weeks). Since distributions of the effect modifiers were evenly distributed across all comparisons (online supplemental appendix 9), it was concluded that the transitivity assumption holds.

Risk of bias assessment

The RoB 2 assessment revealed that the included studies were subject to some concerns for risk of bias. The overall risk of bias for each study and bias risk across the five dimensions was reported in online supplemental appendix 10. The assessment of risk of bias in individual studies demonstrated substantial inter-rater agreement (online supplemental appendix 11). The overall risk of bias according to the RoB 2 was low in 14 studies (32.6%), with some concerns in 24 (55.8%) and high in 5 (11.6%).

Efficacy of the treatment-level interventions

The CINeMA assessment suggested that the evidence from this network meta-analysis was relatively robust. The global test showed no evidence of incoherence (online supplemental appendix 12). For details on the reliability assessment of the network conducted in CINeMA, please refer to online supplemental appendix 12. The Egger’s test and funnel plot did not reveal any evidence of asymmetry; we found no evidence of publication bias (online supplemental appendix 13). The convergence of the model generated through MCMC sampling was confirmed by examining the continuity of trends across the time series, assessing the model convergence, and the PSRF values depicted in the Gelman-Rubin plots (online supplemental appendix 14). For local test, the node-splitting method revealed no evidence of incoherence between the direct and indirect evidence-based estimates for each treatment (online supplemental appendix 15).

The treatment-level network for the primary outcome was successfully constructed (figure 1). In the Bayesian network meta-analysis, third-wave therapy, behaviour therapy and CBT demonstrated superior efficacy, indicating a high probability of being one of the most effective treatments (online supplemental appendix 16). The OR for third-wave therapy has a SUCRA of 0.89 and its OR to placebo was 4.80 (95% CrI 2.50 to 9.10). Behaviour therapy has a SUCRA of 0.70, and its OR relative to placebo was 3.50 (95% CrI 1.70 to 7.30). CBT had a SUCRA of 0.63, and its OR relative to placebo was 3.10 (95% CrI 1.70 to 5.70). Figure 2 presents the estimated effects for the primary outcome. The relative effects table is presented in online supplemental appendix 17.

Figure 1. Network diagram for the primary outcome at treatment level. The width of the lines connecting each treatment reflects the number of trials in which the corresponding comparison was conducted. The number of trials is displayed above the line connecting each treatment. If two specific treatments are not connected, this indicates that there were no trials conducted for this comparison.

Figure 2. Estimated effect size from treatment-level network meta-analysis. 3W, third-wave therapy; BT, behaviour therapy; CBT, cognitive–behavioural therapy; CrI, credible interval; CT, cognitive therapy; PE, psychoeducation; ProbSolv, problem-solving therapy; TAU, treatment as usual.

The sensitivity analysis, which excluded certain trials based on predefined criteria, supported the findings of the primary analysis (online supplemental appendix 18). The results of the frequentist network meta-analysis were consistent with the primary analysis (online supplemental appendix 19). Subgroup analyses were largely consistent with the primary analysis (online supplemental appendix 20). Behaviour therapy, third-wave therapy and CBT were superior to placebo among adults. In adolescents, third-wave therapy, CBT and psychoeducation outperformed placebo. The sensitivity analysis, which used estimated values for treatment responders for all trials, was consistent with the primary analysis for all trials (online supplemental appendix 21).

Efficacy of the components

The component-level network consisted of one large network and two smaller networks. The large network included a variety of components (online supplemental appendix 22). Figure 3 illustrates the components associated with the increment and decrement of all outcomes. The heterogeneity variance for the primary outcome was estimated at τ²=0.21, with an I² value of 38.5% (95% CI 9.9% to 58.1%), indicating the presence of moderate heterogeneity overall (Q=63.47, p=0.008). Implementation of organisational strategies (iOR=2.03, 95% CI 1.27 to 3.24) and inclusion of third-wave components (iOR=1.95, 95% CI 1.30 to 2.93) were associated with improved treatment response. In contrast, the placebo component (iOR=0.26, 95% CI 0.09 to 0.80) and the nocebo component (iOR=0.24, 95% CI 0.10 to 0.59) were associated with decreased treatment response. In secondary outcomes, the component network meta-analysis suggested that third-wave components (iSMD=0.24, 95% CI −0.001 to 0.47) were associated with the alleviation of total ADHD symptoms. The component of problem-solving techniques (iSMD=0.98, 95% CI 0.001 to 0.83) was associated with the alleviation of inattention symptoms. There were no components related to dropout rates. Sensitivity analyses confirmed the findings of the primary component network meta-analysis for all outcomes (online supplemental appendix 23).

Figure 3. Estimated associations for all outcomes from component network meta-analyses. For iOR, a value below 1.00 indicates a decrease in treatment response rate, while a value above 1.00 indicates an increase in treatment response rate. For iSMD, the sign of the value indicates the effect—negative values (‘−’) signify a detrimental effect, while positive values indicate a beneficial effect. ADHD, attention-deficit hyperactivity disorder; iOR, incremental OR; iSMD, incremental standardised mean difference.

Our findings suggest that combining multiple components from among organisational strategies, third-wave components, or problem-solving techniques resulted in more favourable therapeutic outcomes. Treatment estimates compared with the placebo conponents and incremental effects for existing combinations were provided in online supplemental appendix 24. Sensitivity analyses provided support for the primary analysis results regarding these component combinations (online supplemental appendix 25).

Conclusions and clinical implications

To our knowledge, this is the first systematic review and component network meta-analysis evaluating CBT for core symptoms of ADHD from adolescence to adulthood. The main contribution of this study lies in demonstrating, for the first time, the associations of specific CBT components with improvements in the core symptoms of ADHD and treatment responsiveness.

Our findings support the idea that behaviour therapy, CBT and third-wave therapy operate through different mechanisms.^{13 25} Our results suggest that third-wave therapy, CBT and behaviour therapy may be more effective than placebo in improving core symptoms of ADHD. CBT is a complex intervention. Identifying specific, particularly effective treatment components could significantly contribute to developing optimal treatment for individuals with ADHD.

Previous research had not definitively identified the specific components of psychotherapy most effective in ameliorating core symptoms of ADHD. Our findings significantly contribute to filling this gap in the literature, suggesting that organisational strategies, third-wave components and problem-solving techniques may be particularly beneficial in mitigating ADHD core symptoms. Organisational strategies can be helpful for individuals with ADHD by maintaining motivation through stimulus adjustment through environmental adjustment, time management and optimisation of rewards, which may enhance executive functioning.²⁶ Third-wave components encompass a range of techniques that emphasise regulating responses to stimuli, enhancing awareness of one’s experiences, transforming dysfunctional behaviour and strengthening meta-cognitive skills. These third-wave components may ameliorate ADHD symptoms through the cultivation of attention control and adaptive behaviours. However, as third-wave techniques are relatively new, and effect sizes reported in recent trials have been prone to overestimation, cautious interpretation of their efficacy is warranted.²⁷ Problem-solving techniques have been shown to be associated with a reduction in inattention symptoms. Individuals with ADHD often experience frequent job changes and relational instability,²⁸ as well as financial distress such as late bill payments and low savings.²⁹ These real-world problems lead to chronic stress, which may impair performance levels in individuals with ADHD.³⁰ Therefore, it is logical to assume that successful problem-solving techniques can help reduce inattention symptoms.

This network meta-analysis, encompassing various CBT components, systematically reviewed the previous RCTs, evaluating treatment response, dropout from the treatment and severity of ADHD core symptoms. It also incorporated additional outcome data by contacting the authors and employed advanced network meta-analysis methods to seek optimal components of CBT for the treatment. There was no evidence for publication bias which potentially inflates effects of smaller trials. The robustness of the evidence presented in this network meta-analysis would be supported by the generally consistent findings from sensitivity analyses. Several limitations warrant cautious interpretation of our findings. This study exclusively included research reported in English and did not encompass trials documented in other languages. Future studies could consider including grey literature, with several languages. The third-wave components encompass a diverse range of cognitive–behavioural techniques. This component network meta-analysis could not elucidate the effectiveness of specific techniques within individual third-wave components. Therefore, our findings should be considered as hypotheses, and properly designed trials are warranted to evaluate whether specific combinations of components are superior.

In conclusion, our findings suggest that the third-wave therapy, behaviour therapy and CBT may be effective in addressing the core symptoms of ADHD. Our analysis further indicates that an optimal treatment package for ADHD may include organisational strategies and third-wave components. Problem-solving techniques may also contribute to reducing symptoms of inattention.

The funders had no role in the study design, execution and analysis, and manuscript conception, planning, writing and the decision to publish.

Data availability statement

Data are available upon reasonable request.