The AADPA Australian Evidence-Based Clinical Practice Guideline for Attention Deficit Hyperactivity Disorder (ADHD) was released in July 2022 (Bellgrove, 2022). The AADPA guideline has been endorsed by the Royal Australian College of General Practitioners (2023) and ADHD New Zealand (2023). The website of the Royal Australian and New Zealand College of Psychiatrists (2023) lists the guideline as a resource and the guideline’s recommendations have been formally accepted by the National Health and Medical Research Council (NHMRC) of Australia (Bellgrove, 2022). Clinicians and people affected by ADHD are increasingly aware of this new Australian guideline, with some proposing that it be adopted here (ADHD New Zealand, 2023; Swarbrick, 2022). We support the need for locally-based guidelines because Aotearoa has witnessed a vast increase in the number of people seeking ADHD diagnoses.

What Would a Good Aotearoa ADHD Guideline Look Like?

Guidelines have the potential to improve clinical work by summarising evidence in a way that supports clinicians to implement best practice and prevent misdiagnosis. A useful guideline for Aotearoa would be underpinned by Te Tiriti, be neutral and independent, and based on key evidence. We evaluated the AADPA guideline against both the best practice standards expected of any guidelines used in Aotearoa (Best Practice Advocacy Centre New Zealand, 2023; Nejstgaard et al., 2020) and the guideline authors’ own stated aims that the AADPA guideline should “promote accurate and timely diagnosis, and provide guidance on optimal and consistent assessment and treatment of ADHD” (p. 6).

Strengths of the AADPA Guideline

Although we have many concerns with the AADPA guideline, it does include some statements that are useful in a generic way. For example, recommendations that clinicians be adequately trained (2.1.1), that ADHD assessment is thorough (2.1.2) and that people are provided with appropriate information about their diagnosis after assessment (2.3.3). However, these recommendations are broad suggestions that are already expected best practice for all mental health issues. They are therefore unlikely to contribute to an improvement in the quality of ADHD diagnoses.

Guidelines Should be Underpinned by Te Tiriti o Waitangi

Psychologists in Aotearoa have committed to developing mental health responses with due regard for the spirit and intent of Te Tiriti (NZ Code of Ethics Review Group, 2002). Local developers of guidelines highlight the need for our unique variances, including our ethnic and cultural differences, to be taken into account with any adaption of guidelines (BPAC, 2023). The AADPA guideline was developed for an Australian population. Māori are not mentioned in the document and Te Tiriti informed principles and processes were absent in its development. This is reason enough to not adopt the AADPA guideline here, and is all the more important given that New Zealand research documents different experiences for Māori and non-Māori reporting ADHD symptoms (Cargo et al., 2022).

Guidelines Should Prevent Underdiagnosis and Overdiagnosis

A useful guideline should prevent both under- and overdiagnosis. Most information currently available publicly is focused on the risk for underdiagnosis. But overdiagnosis is a critical issue because of the current influx of people seeking ADHD diagnoses. Given the risks associated with diagnosis described below, it is concerning that the guideline’s evidence base excluded research such as that by Kazda et al. (2021) that found evidence for ADHD overdiagnosis and overtreatment in children and adolescents. The Kazda et al. research is Australian based (and therefore ‘local’ for the AADPA guideline), and the guideline authors were alerted to that research during consultation on the draft guideline (Australian Psychological Society, 2022). We are confused that key details that would help to prevent overdiagnosis are absent from the recommendations. For example, in a list of all other diagnostic criteria for ADHD, the requirement that inattentive or hyperactive-impulsive symptoms must have been present before the age of 12 years is missing. This is likely to increase the likelihood of overdiagnosis, as is the omission of information about relative age effects.

Include the Impact of Relative Age Effects

Relative age effects are absent from the guideline. Relative age effects or ‘late birthdate effects’ are terms used to describe the way the youngest children in a school year tend to fare worse than the oldest children in that same year. For example, in Aotearoa, a child born in June (technically the last birth month of a school cohort) is almost a year younger than the oldest children in their class (i.e. those born in July). At the age of 5 years, a younger new entrant has had 11 fewer months to learn and develop skills than the older new entrants, which is almost 20% of their life. Treating age-related behaviour as ‘ADHD’ could be actively harmful.

Relative age effects are consistently found in the diagnosis and medical treatment of ADHD. Large-scale studies that include data from between 300,000 and 7 million children and span cohorts in Germany, Denmark, Sweden, the Netherlands, Israel, Canada, Norway, Taiwan, the US and Australia have all found the same result (Whitely et al., 2019). For example, in British Columbia, the youngest boys in each class were 30% more likely to receive an ADHD diagnosis than the oldest boys in the class (Morrow et al., 2012). The youngest girls were 70% more likely to receive an ADHD diagnosis than the oldest girls. Moreover, the youngest boys and girls were 41% and 77% more likely to be prescribed a stimulant to treat ADHD, respectively.

Despite these findings, the AADPA guideline does not mention relative age effects. Moreover, the associated technical report (Australian ADHD Guideline Development Group, 2022) lists ‘relative age in school year’ (p. 11) as an excluded search term from a guideline literature review. It lacks any recommendation that teachers and clinicians be aware of the potential for age-related immaturity to be misdiagnosed as ADHD, and to adjust their teaching and diagnostic practices accordingly. These oversights are especially puzzling given local published data were available (Whitely et al., 2017), and a key goal of this guideline was to adapt the UK National Institute of Health and Care Excellence (NICE) (2018) guideline ‘to the Australian context’ (Bellgrove, 2022, p. 36).

Such significant omissions may have long-lasting consequences for the youngest children in each school year who are at risk for being misdiagnosed and treated simply for being younger. Relative age effects also raise critical questions about the conceptualisation of ADHD, which is not considered at all in the guideline. This is beyond the scope of this article but is discussed in detail elsewhere (Ophir, 2022a).

Guidelines Should be Neutral and Independent

Conflict of interest in psychopharmacology is well documented and known to influence the scope, outcomes and reporting of research (Hengartner, 2022; Nejstgaard et al., 2020; Ophir, 2022a). It is accepted good practice for guidelines to include a full conflict of interest declaration (Nejstgaard et al., 2020). In 2009, the NHMRC rejected the previous draft ADHD guidelines because it could not determine whether undisclosed sponsorship may have affected the findings of many of the publications relied on for the draft guideline (National Health and Medical Research Council, 2012). We reviewed the conflict statements in the guideline, on the AADPA website and in other relevant publications. We found that the guideline included a detailed description of a conflict of interest management process; however, the actual conflicts of interest are only available upon request from the authors. This is at odds with similar documents, including: the NICE (2018) ADHD guideline, where the extensive conflicts of interest are publicly available alongside the guideline itself; the World Federation of ADHD International Consensus Statement (Faraone et al., 2021), where over two pages of a 19-page article list all 81 authors’ relevant interests; and the Cochrane review of methylphenidate for ADHD (Storebø et al., 2023), where author declarations of interest are included in the body of the report.

The Australian Federal Government granted the AADPA $1.5 million to develop this guideline. The AADPA, board members and some guideline authors also received funding from the pharmaceutical industry. For example, this year’s AADPA conference is sponsored by Takeda and Novartis, manufacturers of lisdexamfetamine dimesylate (Vyvanse) and methylphenidate hydrochloride (Ritalin). The guideline’s lead author has received research funding from Eli Lilly (Australian ADHD Professionals Association, 2023) and has elsewhere declared receipt of travel support and speaker fees from Shire Pharmaceuticals (Faraone et al., 2021). No information is listed for the guideline collaborators. We suggest that the guideline’s authors should declare interests directly so that readers can easily draw their own conclusions about possible influence. The interests of guideline collaborators should also be listed. This information, which has not been made available in this case, provides vital context for understanding the guideline that has been produced.

Guidelines Should be Based on Evidence

The AADPA guideline emphasises the importance of basing findings on evidence. Long descriptions of the processes used to evaluate research have been included, and the term ‘evidence-based’ is in the guideline’s title. We therefore assessed the evidence cited for each of the 134 recommendations in the guideline. We found that only 9% of the guideline met the authors’ own standards. Instead of being ‘evidence-based’, 122 of the 134 recommendations are either ‘clinical consensus recommendations’ prepared by ‘expert subgroups of the GDG [Guideline Development Group]’ or ‘clinical practice points’, which are guidance based on expert opinion and clinical experience.

Of the 12 evidence-based recommendations, only three were considered to have ‘high’ or ‘moderate’ certainty. These terms mean that either the guideline’s development group was “very confident that the true effect lies close to that of the estimate of the effect” or it believed that “the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different” (pp. 10–11). While only three of the 12 recommendations were considered to be supported by ‘certain’ evidence, the guideline ‘strongly’ recommends eight of them. Our finding was consistent with the Cochrane review (Storebø et al., 2023) on the use of methylphenidate for childhood ADHD, which found that all 212 trials included in its meta-analysis were at high risk of bias. Calling a guideline ‘evidence-based’ when only 2% of its recommendations are based on certain evidence seems misleading. The phrase may encourage unwarranted confidence in the guideline by the community, practitioners and researchers.

Risks of Stimulant Medication

Before 1999, little was known about the effects of stimulants beyond a 3-month period (Jensen et al., 2007). This led to the establishment of the Multisite Multimodal Treatment Study of Children with ADHD (MTA), which was designed to compare the long-term efficacy of medication and behavioural treatments for ADHD (MTA Cooperative, 1999). After diagnosis with ADHD, 579 children aged 7–10 years were randomly assigned to behaviour therapy, medication management, a combination of these two (medication and behaviour therapy) or to referral to treatment provided as usual in the community. After the initial 14-month study, participants were followed for 17 years alongside a matched comparison group of 289 classmates without ADHD. At 14 months, greater reduction in ratings of ADHD symptom severity were found in the groups who used stimulants, offset by significantly reduced height gain. After 24 months, the groups that received stimulants had lower symptom scores compared with the groups that did not, although the effect sizes were reduced by about half. However, the benefit of medication was not maintained at 3 years after baseline (Jensen et al., 2003). Although substantial improvement had occurred in all groups by that time, the treatment groups did not differ significantly from the comparators on any measure. “The modest significant advantages we found at the 24-month assessment for [medication]…were completely lost by 36 months” (p. 998). Participants using stimulants in the 24–36-month period showed increased symptomatology during that interval relative to those not taking medication. The loss of the benefit from stimulants continued for 17 years from the 3-year point to the final assessments when participants were aged 25 years (Molina et al., 2009; Swanson et al., 2017). Follow-up assessments found no difference in symptom severity between the original four groups. However, they did continue to find lower heights for those in the medication groups. Although the MTA data provided support for the short-term reduction of symptoms with medication management, the long-term follow-up data did not support an advantage of stimulant treatment beyond 2 years for the majority of children.

The lack of positive outcomes from long-term stimulant use has been replicated in other research, such as the Western Australian Raine ADHD Study (Department of Health, 2010) and a study by the Bureau of Economic Research in Quebec (Currie et al., 2013). The Canadian study found an association between long-term stimulant use and academic failure, as well as a greater likelihood of mood problems. The Raine study found no significant differences between the medication and no medication groups for depression, self-perception, social functioning or core ADHD symptoms. However, children with ADHD who had used stimulants were much more likely to be identified as performing below age-level by a classroom teacher and had significantly higher diastolic blood pressure than children who had never received medication. It is particularly surprising that the Raine study was not reviewed for the guideline, both because its outcomes raise important questions about the risk/benefit profile of stimulant use for children, and because it is locally relevant. Translating research to local contexts is critical if a guideline is to be useful to the community it serves, which was identified as a specific aim of the guideline (Bellgrove, 2022, p. 36).

Excluding key research that documents the lack of long-term benefits of stimulants matters for two reasons. First, stimulants are not neutral. They produce unpleasant and disruptive side effects (Lakhan & Kirchgessner, 2012; Ophir, 2022b; Schein et al., 2023). Second, the focus on stimulants as treatment is likely to prevent research and clinical attention from addressing other important factors and ways of providing support.

Negative Outcomes of ADHD Diagnosis

Worldwide, health providers are experiencing increasing demands for ADHD diagnoses. People clearly hope that a diagnosis will help. However, although it may seem counterintuitive, research suggests that the diagnosis of a mental illnesses does not necessarily result in improvements. The guideline fails to consider whether a diagnosis of ADHD is in itself positive, when studies have found that diagnosis can lead to poorer outcomes. For example, Kazda et al. (2022) investigated whether an ADHD diagnosis in childhood was associated with improved quality of life in adolescence compared with individuals who had exhibited the same behaviours as children but who were not diagnosed. They found that adolescents who had been diagnosed with ADHD fared worse than those exhibiting the same ADHD behaviours but who had not been diagnosed with ADHD.

Kazda et al. (2022) used data from the Longitudinal Study of Australian Children. That representative, population-based sample included almost 400 adolescents who had been diagnosed with ADHD as children or young adolescents who were matched by age, sex and hyperactivity/inattention score with 393 adolescents who had not been diagnosed with ADHD. Compared with adolescents without diagnoses, there were no significant improvements in the diagnosed group on any measure including general health, happiness and peer trust. Furthermore, compared with adolescents without a diagnosis, diagnosed adolescents had a worse psychological sense of school membership, academic self-concept and self-efficacy, displayed more negative social behaviours and were more likely to harm themselves. The investigators had expected Australian children and adolescents to benefit from a diagnosis of ADHD and consequent treatment, resulting in better quality of life later. However, the results contradicted the study hypothesis. For these Australian representative adolescents, by the time they were adolescents, it was found that important aspects of well-being were worse for those who had been diagnosed with ADHD as children. Such outcomes need to be accounted for when working with people seeking an ADHD diagnosis. Guidelines therefore need to be careful about the assumptions underlying their recommendations, even if they are counterintuitive.

As more parents seek a diagnosis for their children, more children who exhibit mild and moderate symptoms will be diagnosed. For such children, it is more likely that any harms associated with an ADHD diagnosis will outweigh any benefits. The potential for improvement is smaller for children with fewer ADHD behaviours for whom large reductions are impossible. When symptoms are mild or moderate, any harms associated with the ADHD diagnosis (such as those discussed above) or from treatment (such as the adverse effects of stimulants) may outweigh the benefits of symptom reduction. To ensure safe and equitable practice and policy, practitioners should consider the balance of risks and benefits from diagnosis and treatment, especially when supporting less-affected individuals.

Conclusions

The AADPA Australian evidence-based clinical practice guideline is being recommended for adoption in Aotearoa New Zealand (ADHD NZ, 2023). A more useful guideline would be developed from Te Tiriti principles, reduce under- and overdiagnosis and be neutral, independent and based on evidence. The Australian guideline does not include perspectives from Te Ao Māori, ignores relative age effects and evidence of overdiagnosis, was not developed by an independent, neutral entity and omits critical evidence that amphetamine treatment does not sustain benefits in the long-term and that diagnosis itself can be harmful. In our opinion the flaws in the guideline undermine any confidence we can have in the guideline as a whole.

It is clear there is still a lot to learn about the difficult and contentious area of ADHD. More work is required, but ignoring existing knowledge and knowledge gaps will not help us consolidate a stable, shared understanding about ADHD’s aetiology, ADHD’s course and the best forms of support for those who are already struggling with ADHD symptoms.

Implications for Practice

Many people are distressed about their impaired concentration, attention and impulse control, and are seeking ADHD diagnoses in ever-increasing numbers. This review of the AADPA’s guideline has identified too many weaknesses and biases in its development and content for it to be accepted as guidance for clinicians supporting people seeking an ADHD diagnosis. In addition to the weak evidence base and lack of transparency about conflicts of interest, key research is missing. Examples include research about relative age effects and research that suggests diagnosis and stimulant use is at best neutral in its effects on symptoms, but can also make a person’s life worse. As written, the guideline will lead to continued overdiagnosis, misdirected treatment for people who have other unrecognised difficulties and unnecessary exposure to the harms of amphetamines. Until clinicians in Aotearoa have the time and resources to develop and agree upon guidelines that are relevant to our local population and well-grounded in global evidence, we suggest that practitioners be thoughtful about the evidence outlined in this evaluation and rely on their full range of assessment and formulation skills to support their clients.

Conflicts of Interest

The authors have no potential conflicts of interest to declare in relation to this article.