Indirect relationship

Importantly, the authors clarify that they ‘believe that this is unlikely to be because of ADHD itself’ but rather ‘likely caused by modifiable factors such as smoking, unmet mental and physical health support, and unmet treatment needs’. While these potential mediators are plausible, none were examined in the study. Furthermore, the authors did not account for key confounders that are well documented in the existing literature on ADHD and mortality (see below), leaving critical factors unexplored. Instead, the study focused on a simple, unidimensional relationship between ADHD and mortality. While this approach might be suitable for exploratory research into novel phenomena, it is not justified when building on an established body of evidence – especially when the researchers themselves assume that the relationship between ADHD and mortality is indirect and mediated by other factors.

Socioeconomic status

Astonishingly, the failure to consider potential mediating variables extends even to basic sociodemographic factors – a fundamental step in medical research. Departing from standard scientific practice, the authors justify their decision not to adjust for socioeconomic status (SES) – a well-established determinant of health outcomes – by asserting that ’SES is best understood as part of the causal pathway between ADHD and premature mortality (i.e. SES is a mediator)’. While this reasoning may appear plausible, it does not absolve the authors from the responsibility of presenting a SES-adjusted model and empirically testing their mediation hypothesis. If SES is believed to act as a mediator, its role should have been explicitly examined using appropriate statistical methods. Without such analysis, the study overlooks an equally plausible – and arguably more parsimonious – alternative: that SES alone, independent of ADHD, may explain much of the observed premature mortality.

The treatment of choice

This omission of straightforward SES variables is striking, especially given that even literature reviews asserting that ‘ADHD is associated with elevated morbidity and mortality’ ^{Reference Nigg2} often acknowledge that ‘the life expectancy of individuals with ADHD… has not been characterized… Instead, studies addressing ADHD-related mortality issues have focused largely on the potential contributions of (a) treatment-related sudden death and (b) suicide’. ^{Reference Nigg2} Unfortunately, these well-documented mediating risks were neither considered nor acknowledged in the study at hand.

This represents a major limitation of the study, as the mortality risks associated with the first-line treatment for ADHD are, of course, not a direct consequence of the diagnosis itself. Moreover, these risks challenge the authors’ hypothesis that the observed mortality is ‘likely caused by modifiable factors such as… unmet treatment needs’. If the widely used stimulant medications – prescribed to individuals formally diagnosed with ADHD – are themselves associated with increased mortality risk, then the phrase ‘unmet treatment needs’ may need to be reconsidered. It could instead imply that stimulant medications are not the ‘needed’ treatment for many individuals with ADHD.

Final remarks

Of course, other notable mediators, such as comorbidities – which were not adequately examined in the study under discussion – as well as the previously mentioned SES, could also contribute to the observed risk of mortality. However, the current commentary has focused on the evidence regarding the treatment of choice for ADHD, as this represents the primary practical concern arising from the study by O’Nions et al. The authors’ warning that ‘adults diagnosed with ADHD are living shorter lives… likely caused by modifiable factors such as… unmet treatment needs’ might prompt efforts to urgently treat these individuals with first-line medications. Paradoxically, these very medications could exacerbate the exact risks the authors aim to mitigate.