Attention deficit hyperactivity disorder (ADHD) remains a topic of much interest across clinics, social media, academic research and the Mental Elf woodland with five blogs on this topic just this year (Calvosa, 2025; Es, 2025; Karmakar, 2025; Suetani, 2025; Wainwright, 2025).
Research surrounding ADHD has been shifting to explore the broader impacts beyond the core clinical symptoms described in diagnostic manuals. A recent Mental Elf blog by Jack Wainwright in January 2025 discussed the findings from an umbrella review highlighting the mental health, physical health, social and lifestyle risks that are associated with ADHD (French et al., 2024). The results suggest that ADHD may have far reaching adverse impacts across a variety of domains in life.
This then raises the crucial question; could medication mitigate the impacts beyond core ADHD symptoms? Or more specifically, does pharmacological treatment of ADHD reduce the risks of suicidal behaviours, substance misuse, accidental injuries, transport accidents and criminality? In their comprehensive study, Zhang et al. (2025) utilised data from multiple national Swedish registers and a novel study design – target trial emulation study – to answer this very question.
This target trial emulation study may sound like something from science-fiction, but is a novel approach to address the limitations of randomised controlled trials.
Methods
Data from multiple national Swedish registers were used to identify all residents aged six to 64 who were newly diagnosed with ADHD between 2007 and 2018. These individuals were then split into two cohorts:
- Initiators: those who started pharmacological treatment (amphetamine, atomoxetine, dexamphetamine, guanfacine, lisdexamfetamine and methylphenidate) within three months of diagnosis and remained on this medication for the follow-up period, or,
- Non-initiators: those who did not start pharmacological treatment during the follow-up period.
Five outcomes were explored: suicidal behaviours, substance misuse, accidental injuries, transport accidents and criminality. These outcomes were obtained from national registers.
The ‘special sauce’ in this study was the use of a target trial emulation design to emulate a randomised controlled trial (RCT), which was achieved through cloning, censoring and weighting. All eligible individuals were ‘cloned’ at the time of diagnosis and one of these clones was placed into each study cohort. At monthly intervals, each clone was assessed to see if it still met the cohort definition – continuing the medication or remaining off treatment respectively – and were censored if not. Lastly, weighting was applied to account for selection bias from the artificial censoring (see the figure from the Supplementary material – PDF).
Results
Overall, 148,581 individuals were included in the study with 84,282 in the initiation cohort and 64,377 in the non-initiation cohort. The main findings are summarised as below:
For first events, pharmacological treatment of ADHD was associated with statistically significant reduced rates in four of the five outcomes examined:
- Suicidal behaviours (incident rate ratio [IRR] 0.83; 95% confidence interval [CI] 0.78 to 0.88)
- Substance misuse (IRR 0.85; 95% CI 0.83 to 0.87)
- Transport accidents (IRR 0.88; 95% CI 0.82 to 0.94)
- Criminality (IRR 0.87; 95% CI 0.83 to 0.9)
There was no statistically significant association with pharmacological treatment and accidental injuries (IRR 0.98; 95% CI 0.96 to 1.01).
For recurrent events, pharmacological treatment was associated with statistically significant reduced rates for all outcomes examined:
- Suicidal behaviours (IRR 0.85; CI 0.77 to 0.93)
- Substance misuse (IRR of 0.75; CI 0.72 to 0.78)
- Accidental injuries (IRR 0.96; CI 0.92 to 0.99)
- Transport accidents (IRR 0.84; CI 0.76 to 0.91)
- Criminality (IRR 0.75; CI 0.71 to 0.79)
Stimulant medications were found to have lower rates of first and recurrent events of all outcomes when compared to non-stimulants with IRRs ranging from 0.71 to 0.97.
In the study population, 8.7% to 53.1% of individuals had prior events of the outcomes of interest. In the subgroup analysis, reduced rates of suicidal behaviours and transport accidents were seen with pharmacological treatment in individuals without a history of those events. There was no difference seen in substance misuse, accidental injuries or criminality when there was no history of those events. Pharmacological treatment was associated with decreased rates across all outcomes for individuals where there was history of the events with IRRs ranging from 0.79 to 0.97. A significant difference in the rate ratio was found for those with a history of substance misuse (P<0.01) compared to those without and for those with a history of criminality (P=0.02) compared to those without.
Pharmacological treatment of ADHD is associated with reductions in the rates of suicidal behaviours, substance misuse, transport accidents and criminality in a large target trial emulation study.
Conclusions
The authors concluded:
…these results provide evidence on the effects of ADHD drug treatment on important health related and social outcomes that should inform clinical practice and the debate on drug treatment of ADHD.
Clinicians should consider the broader social effects of pharmacological treatment of ADHD.
Strengths and limitations
This study has several strengths owing to the large study population and results that may be more applicable to the general population of people with ADHD seen in clinical practice. As the authors explain, this study found more modest reductions in the outcomes of interest compared to previous observational studies. This is to be expected with the large dataset in this study, but also may be due to the cohort definitions with individuals required to remain on the treatment consistently for the follow-up period or be censored out. This may have resulted in the exclusion of more severe cases with individuals who may, due to their symptoms, not be able to attend appointments, obtain prescriptions and regularly obtain the medications.
This also raises the question of other socio-economic factors that may impact on a person’s ability to obtain a diagnosis of ADHD and adhere to a treatment regime. While the situation may be different in Sweden, certainly in where we practice, in Australia, the diagnosis and treatment of ADHD seems to be heavily dependent on people’s ability to afford such services.
Similarly, the non-initiation cohort may encompass a cohort of people who have a less severe or very severe illness if they were never commenced on pharmacological treatment. Were these individuals not treated due to a lack of functional impairment? Were there other factors that would preclude treatment such as concerns for medication misuse or comorbid psychiatric conditions which may represent a more severe cohort which would have inflated the rate reductions seen in this study?
A key limitation, as mentioned by the authors, is that the outcome data was obtained from national registers. This means less severe outcomes where individuals did not come to the attention of services would not have been recorded.
Another limitation is how generalisable this study may be to other settings and countries.
Further, as mentioned above, how the cohorts were defined may make it difficult to translate the findings into our everyday clinical practice. A recent study by Ferrin et al (2025) found that adherence rates to ADHD medication in children and adolescents may be as low as 22% at 12-month follow-up. This raises a significant question on how many individuals were censored out of this study due to poor adherence? It is difficult to appraise how generalisable the results are if the population reflects a fraction of those seen in routine clinical practice.
Cohort definitions in this study may limit the generalisability and application to routine clinical practice.
Implications for practice
This is a very interesting study which utilises a novel methodology to leverage large national registers to answer a clinically relevant question in respect to ADHD treatment at a time when there are increasing concerns about the suitability of RCTs particularly in the mental health field. As Suetani (2025) mentioned in a Mental Elf blog earlier this year, “RCTs may no longer be the gold standard of clinical research in psychiatry”, could target trial emulation studies then be a potential contender to take the mantle?
Ultimately, despite the limitations, this study does support the clinical observation that if we provide appropriate pharmacological treatment for a person with ADHD, they will experience far-reaching benefits beyond their core symptoms into other aspects of their life.
Of note, one interesting difference between the study and our clinical practice in Australia is the type of psychostimulants we use. In the most recent data available on ADHD medications from the Australian Institute of Health and Welfare (AIHW, 2025), 85% of ADHD prescriptions in 2023-2024 were for psychostimulants (dexamfetamine, methylphenidate and lisdexamfetamine). This is similar to the rates of 91% in the current study. However, unlike in Sweden where the vast majority of individuals in the study were on methylphenidate (88%), there are similar rates of prescription of methylphenidate and lisdexamfetamine in Australia (AIHW, 2025). Overall, it could be expected that all psychostimulants would convey some level of benefit as suggested by this study, but the preponderance of methylphenidate should be kept in mind.
Furthermore, if the results of this study hold true and pharmacological treatments result in a greater reduction of criminality and substance misuse than other treatments and even more so when there is a history of these events, then should we be offering this as a treatment more readily to individuals with a criminal or substance misuse history? With the current national guidelines (AAPA, 2022) and level of restrictions around stimulants in Australia, this vulnerable cohort of people with ADHD would continue to face many barriers obtaining stimulant medications. But could we provide a monitored and regulated avenue for these individuals to obtain stimulants, similar to that seen with methadone? Is the system sufficiently structured enough currently to trial such interventions? Probably not yet.
Nevertheless, the current study by Zhang et al. empowers us to see beyond the core symptoms of ADHD using a novel study method. We now need to use the findings and knowledge we obtain to make systematic changes that will improve outcomes for people with ADHD at the population level.
Barriers to accessing pharmacological treatment of ADHD should be considered in light of the improved outcomes at a population level.
Statement of interests
Jordan Budgen has no conflicting interests with respect to the content of this blog post.
Shuichi Suetani is a member of the Royal Australian and New Zealand College of Psychiatrists ADHD Network, and Australasian ADHD Professionals Association.
Edited by
Dr Dafni Katsampa
Links
Primary paper
Zhang, L., Zhu, N., Sjölander, A., Nourredine, M., Li, L., Garcia-Argibay, M., Kuja-Halkola, R., Brikell, I., Lichtenstein, P., D’Onofrio, B. M., Larsson, H., Cortese, S., & Chang, Z. (2025). ADHD drug treatment and risk of suicidal behaviours, substance misuse, accidental injuries, transport accidents, and criminality: emulation of target trials. BMJ, 390, e083658.
Other references
Australian Institute of Health and Welfare (AIHW). ADHD Medications dispensed 2004-05 to 2023-24, last accessed 21 Sep 2025.
Australasian ADHD Professionals Association (2022). Australian Evidence-Based Clinical Practice Guideline for Attention Deficit Hyperactivity Disorder (ADHD). Last accessed 21 Sep 2025.
Calvosa S. Alarm bells ringing: how do people with ADHD process sensory information? The Mental Elf, 11 Aug 2025.
Es T. The genetic relationship between ADHD and depression. The Mental Elf, 26 Feb 2025.
Ferrin, M., Häge, A., Swanson, J. et al. (2025) Medication adherence and persistence in children and adolescents with attention deficit hyperactivity disorder (ADHD): a systematic review and qualitative update. Eur Child Adolesc Psychiatry 34, 867–882. https://doi.org/10.1007/s00787-024-02538-z
Karmakar S. “It’s not just for attention”: New research highlights the increased risk of PMDD in women with ADHD. The Mental Elf, 10 Sep 2025.
Suetani S. Will this ADHD medication help my patient? Bridging the gap between efficacy and effectiveness. The Mental Elf, 7 Jan 2025.
Wainwright J. The risks arising from having ADHD: physical health, mental health, social and lifestyle. The Mental Elf, 14 Jan 2025.
