Executive functions (EFs) are the cognitive control systems that allow people to pursue goals, make decisions, and adapt to changing circumstances. Researchers generally break them into three overlapping capacities: working memory (holding and manipulating information in mind), inhibitory control (suppressing impulses and filtering out distractions), and cognitive flexibility (switching between tasks or mental frameworks). Strong EFs in childhood predict academic achievement, social competence, and long-term mental health; weaknesses in these areas that go unaddressed can persist into adulthood, undermining school performance, career prospects, and well-being. 

The Background:

Interest in training these skills has grown rapidly, but most research has been conducted in Western settings. China presents a distinctive context. Collectivist values make group-based programs culturally natural, and parental investment in academic outcomes is high. Both of these factors should, in theory, work in an intervention’s favor. At the same time, tightly scheduled school days (sessions typically capped at 30 minutes or less) constrain what is actually deliverable. A growing number of randomized controlled trials (RCTs) have tested EF interventions with Chinese children, but until now, no one has pulled that evidence together systematically. 

The Study:

A new network meta-analysis did exactly that. The researchers screened RCTs involving Chinese children aged 3–12, including both typically developing children and those showing subclinical signs of ADHD or autism spectrum disorder (ASD), for instance, siblings of children with an ASD diagnosis. Children who already carried a formal neurodevelopmental diagnosis were excluded. Fifty-two trials covering nearly 3,000 children met the inclusion criteria. Interventions fell into four categories: 

• Computerized adaptive n-back training with metacognitive coaching (strategy instruction and self-monitoring): 14 trials, 486 children 

• Stop-signal and rule-switching tasks targeting inhibitory control and cognitive flexibility, delivered face-to-face: 18 trials, 632 children 

• Hybrid physical-cognitive training moderate-to-vigorous aerobic exercise combined with concurrent cognitive demands (e.g., brisk walking while counting backward) in 20–30-minute sessions: 10 trials, 298 children 

• Computerized cognitive flexibility training (set-shifting, dual-task coordination), self-paced with progressive difficulty: 10 trials, 312 children 

The headline finding is that three of the four intervention types produced statistically significant improvements across all three EF domains. The exception was the hybrid physical-cognitive program, which did not reach significance for inhibitory control. Positive results across the board might sound encouraging until you look at the actual effect sizes. 

The Results:

The actual effects were negligible. Every significant result fell well below what methodologists define as a “small” effect (a standardized mean difference, or SMD, of 0.2). The largest effect size in the entire analysis was an SMD of 0.097  (less than half that threshold). The authors summarize the interventions’ effects as “modest,” but that is generous phrasing for numbers that, in practical terms, amount to very little. The analysis also showed signs of publication bias, meaning that studies with null or negative results may not have been published, potentially inflating even these modest figures. 

The Take-Away: 

It is important to note that these results don’t necessarily mean that this is the last word on EF training. The results apply specifically to Chinese children working within the time constraints of Chinese school schedules, and they exclude children with diagnosed ADHD, a population for whom cognitive interventions sometimes show larger effects. Generalizing beyond those boundaries is unwarranted. 

What the findings do suggest is that structured EF programs, as currently implemented in Chinese educational settings, are not delivering meaningful real-world benefits. Statistical significance, it is worth remembering, is not the same as practical significance, and the gap between the two is sharp here. 

Language is powerful. The words we choose not only reflect our understanding of the world but also actively shape it. Recently, this truth has been at the center of a growing debate in the mental health field regarding how we talk about ADHD.  

In a recent paper published in The Lancet Psychiatry titled “The Power of Words: Respectful Language in ADHD Research,” French and colleagues advocated for a shift toward "neurodiversity-affirmative language”. Rooted in the social model of disability, their proposal encourages researchers to abandon traditional medical terminology, e.g., words like disorder and deficit, in favor of more neutral terms such as condition and challenge.  

My colleague, Dr. Michael Miller, and I read this with great interest. We completely agree that revising language is essential to good science and that, both as researchers and as human beings, we are ethically bound to speak respectfully. However, we felt compelled to write a response. In our new paper, we argue that while language must evolve, it must do so scientifically. 

The Two Prerequisites for Language Change 

If we are going to fundamentally shift our scientific lexicon, two requirements must be met: 

1. A clear consensus among those with lived experience that the current language is harmful and that new language is needed. 

2. A commitment to scientific accuracy and precision in the new terms. 

Currently, the proposal by French and colleagues meets neither requirement. While they claim consensus is accumulating that certain terms are disrespectful, they provide zero empirical evidence that this view is shared by the community of individuals living with ADHD. Even proponents of patient-centered language admit there is surprisingly little data supporting specific language changes. 

More alarmingly, the recommended changes severely dilute the scientific accuracy of our field. Let’s look at two examples. 

Why a "Deficit" is Not Just a “Challenge" 

French and colleagues suggest replacing the term deficit with challenge. On the surface, challenge sounds softer and more affirming. But scientifically, these words are not interchangeable. 

For decades, the term deficit has been defined by a specific performance metric that falls substantially below an expected level. It is a measurable reality. A challenge, on the other hand, refers to a new or difficult task that tests someone's ability.  

Every single human being is "challenged" by complex neuropsychological tests, but only some individuals who face that challenge demonstrate scientifically significant deficits. If we relabel measurable deficits as universal challenges, we sacrifice the exactness required to communicate scientific findings and accurately measure the effects of life-changing treatments. 

ADHD is a Disorder, Not Just a "Condition" 

Another proposal is to replace the word disorder with condition. 

In mainstream psychiatry, a disorder is a clinically significant disturbance that causes distress or disability. The word purposefully separates natural human variation from the suffering (pathos) that gives pathology its meaning.  

Condition is a completely neutral term. Pregnancy is a condition. Being tall is a condition. Calling ADHD a condition distances the diagnosis from the profound suffering it can cause.   

French et al. argue against framing ADHD as a disorder because it exists on a spectrum without a clear cutoff, its manifestation is context-dependent, and its definition evolves. But if we apply that logic across all of medicine, the concept of disease unravels: 

• Are hypertension and osteoporosis no longer diseases because they rely on dimensional thresholds? 

• Is asthma no longer a disease because its manifestation depends heavily on environmental context? 

• Was multiple sclerosis not a disease before modern imaging allowed us to physically see brain lesions? 

‍

The Real-World Danger of Imprecise Language 

This is not merely an academic debate over semantics. The language we use has real-world implications. In the United States and across the globe, our healthcare, educational, and legal systems run on precise medical language. Terms like impairment, dysfunction, and disorder are legally and administratively required to justify support services, workplace accommodations, specialized educational therapies, and medications. The language of pathology in diagnostic manuals regulates the flow of these resources. 

If we reclassify ADHD as a neutral condition characterized only by challenges, we risk erecting massive bureaucratic barriers. Imprecise language could easily be used by institutions or insurance companies to deny vital care to the people who need it most. 

The Need for Lexical Discipline 

Attempting to characterize a clinical disorder entirely through its strengths happens in a scientific vacuum. We cannot ignore the vast body of rigorous evidence confirming that ADHD meets the long-standing criteria used by mental health science to identify clinical disorders. 

As professionals, our respect for the ADHD community demands a commitment to language that is clear, correct, and evidence-based. To build genuine consensus about how we talk about ADHD, we need meaningful, collaborative dialogue that integrates compelling empirical data and rigorous theory. 

This standard of "lexical discipline" is not just a technical preference.  It is a vital mechanism through which science and the mental health professions uphold their duty to society. 

‍

^{The First Comprehensive Dose-effect Network Meta-analysis of ADHD Medications:}

For many ADHD patients, getting properly diagnosed and starting meds is only half the battle. The next step is figuring out the exact right dose. Historically, clinical guidelines have provided scant guidance on this critical step. This lack of direction can inadvertently foster two extremes in clinical practice: therapeutic inertia (settling for a subtherapeutic dose that leaves symptoms undertreated) or uncritical escalation (driving doses higher and higher beyond licensed limits without meaningful benefit).

To clear up this pharmacological gray area, an international team of researchers published the first comprehensive dose-effect network meta-analysis of ADHD medications in The Lancet Psychiatry. By pulling together a massive vault of clinical trial data, they mapped out exactly how efficacy and tolerability shift as doses increase.

The Study:

Traditional meta-analyses evaluate head-to-head, pairwise data, comparing one drug at a specific dose directly against a placebo. However, this study utilized an advanced Bayesian hierarchical network model using restricted cubic splines.

This mathematical framework allowed the researchers to combine both direct trial data and indirect evidence simultaneously across 113 double-blind randomized controlled trials (RCTs). In total, the study evaluated data from 14,138 children/adolescents and 11,016 adults. By standardizing various formulations into basic equivalents (e.g., converting amphetamines to dextroamphetamine equivalents), they created a clear, unified map of dose ranges.

The Results: 

The study yielded distinct dose-response curves depending on the patient's age and the specific medication class. Rather than a linear trend in which "more medicine equals more benefit," most treatments reach a clear statistical plateau or ceiling.

For Children and Adolescents (under 18)

In the pediatric population, medications hit clear peak efficacy boundaries:

• Methylphenidate: Average efficacy peaked at roughly 45 mg/day. Beyond this, curves suggested a minor dip in efficacy, though with wide credible intervals (high uncertainty).
• Amphetamines: Reached their peak average benefit at approximately 25 mg/day
• Guanfacine: Maxed out its clinical benefit at around 4mg/day.

For both amphetamines and guanfacine, escalating the dosage past these points resulted in U-shaped curves, meaning further dose hikes yielded diminishing group-level symptom reduction.

For Adults (18 and older)

Adult profiles showed slightly different trajectories:

• Amphetamines: Reached a distinct clinical plateau at roughly 50 mg/day. Pushing the dose higher did not improve average symptom relief.
• Methylphenidate: Interestingly, adult data showed a continuous increase in efficacy across the observed dose range, though with diminishing incremental improvements as it approached 50 mg/day. The researchers noted this lack of a distinct plateau might be due to sparse trial data in higher-dose adult brackets.

The ultimate goal of this landmark analysis is to guide shared decision-making between clinicians, patients, and families. The results send a dual message to the medical community:

1. Avoid Therapeutic Inertia: Clinicians should not hesitate to optimize doses and titrate upward from low starting doses if a patient's ADHD symptoms remain insufficiently controlled. Subtherapeutic dosing remains a widespread issue that impairs long-term treatment adherence.
2. Rethink Routine Escalation: At the patient-group level, there is no compelling statistical evidence that routinely pushing past FDA-licensed maximum limits provides additional clinical benefit—but it reliably exposes patients to higher risks of side effects and reduced tolerability.

The Takeaway:

A medication's true efficacy hinges on its tolerability, typically measured by how often patients discontinue treatment due to severe side effects. For amphetamines, this dropout risk scales linearly with dosage, notably exceeding placebo in children above 25 mg/day and becoming prominent in adults past 50 mg/day. In contrast, methylphenidate shows no clear dose-dependent dropout risk in pediatric patients, whereas adults face a steep risk curve: increasing the dose from 60 mg/day to 90 mg/day raises the dropout risk from 7.3% to 10.0% for only modest symptom relief. Finally, youth taking guanfacine experience a sharp climb in discontinuation risks, reaching a 9.8% median risk at 4 mg/day before data limitations obscure further trends.  

‍

The authors strongly emphasize that these findings represent group averages. Because individual metabolism, genetics, and comorbidities vary widely, some specific patients may legitimately require and tolerate higher off-label doses. However, if an unusually high dose is needed, the study suggests it should prompt a careful clinical pause, either to reassess for co-occurring conditions (like anxiety, autism, or sleep disorders) or to manage realistic expectations regarding what the medication can achieve.

‍