Loading ...

An expert panel has released the first formal guidance on deprescribing stimulant medications in adults with attention-deficit/ hyperactivity disorder, recommendations that come as US stimulant prescribing increased by more than 50% over the past decade with little to no direction on when treatment should be tapered or stopped. 

The consensus statement, developed by an American Society of Clinical Psychopharmacology (ASCP) international task force of 45 psychopharmacology experts, identifies clinical scenarios in which deprescribing warrants consideration, organized into three categories: inadequate efficacy, associated harm, and misuse. 

It was published online May 20 in European Neuropsychopharmacology.

The recommendations are reinforced by a separate study, published online May 15 in Lancet Psychiatry, that mapped dose-response curves across all major ADHD medications for the first time. It identified dose thresholds beyond which higher doses offered little additional relief of symptoms but increased the risk of side effects. 

Together, the two publications address both sides of the dosing equation: where to stop going up, and when to come back down, said David W. Goodman, MD, assistant professor of psychiatry at the Johns Hopkins University School of Medicine, Baltimore, and co-lead investigator on ASCP guidance.

"As mental health and psychiatric treatment move towards non-psychiatrists in order to provide access to a greater number of people, we needed to provide some overall guidance," he told Medscape Medical News.

"Taken together, the dose-effect study demonstrates the range of dose efficacy and tolerability in ADHD, while the ASCP deprescribing stimulant expert consensus for adult ADHD prompts prescribers to regularly evaluate whether the medication remains effective and tolerable for the patient, and if not, consider deprescribing and re-evaluating alternatives," added Goodman, who was not involved in the dose-effect review. 

When to Reassess Stimulant Use

Until now, no practice guideline or consensus statement has offered guidance on when to deprescribe stimulant medications in adult ADHD. 

"The paper should not be interpreted as an advocacy for taking patients off medications," Goodman said. "The point of the paper is to consider whether or not the psychiatric medications that you're prescribing are still benefitting the patient, or whether it's time to reassess."

The task force used a two-round Delphi survey, reaching consensus — defined as 75% or greater agreement — on 10 of 11 statements.

For inadequate efficacy, the task force recommended deprescribing when a previous ADHD diagnosis is deemed inaccurate, when a stimulant fails to produce optimal response despite dose optimization, or when tolerance persists despite dose increases. 

For associated harm, the task force recommended deprescribing when stimulants exacerbate a concurrent psychiatric or medical condition, when adverse effects cannot be managed with dose reductions, or when a change in medical status shifts the risk-benefit ratio.

For misuse, the task force recommended deprescribing in cases where patients persistently exceed their prescribed dose, divert their medication to others, or use the medication for enhancement beyond ADHD treatment. 

The one statement that fell short of consensus involved cannabis. Most panelists (71%) felt that regular cannabis use alone should not be sufficient reason to deprescribe, but this fell just below the 75% threshold. "The literature is just not there" to justify reaching the preset threshold for consensus, Goodman said.

The task force found no randomized discontinuation trials of stimulants beyond 6 months in adults, and no empirical data on tapering protocols.

Goodman said stimulant withdrawal is generally less severe than with antidepressants. "Patients lose their energy, their motivation, or might be a little more reactive. That persists for a few days, and then within a week it dissipates," he said.

The task force found that clinicians sometimes prescribe stimulants based on subjective complaints of inattention without a comprehensive DSM-5-TR evaluation. The statement also found no evidence to support combining stimulants from two different chemical classes, such as amphetamine and methylphenidate, and recommended deprescribing one of the medications if this occurs.

"There's no reason to combine the two," Goodman said, adding that the recommendation was directed particularly at prescribers "who are not trained in this field and are not knowledgeable about how to use these medications."

Where Clinical Benefit Plateaus

The new guidance comes soon after publication of a dose-effect review of ADHD medications that reinforced the need for stronger evidence on dosing decisions. 

The review sought to address the lack of guidance on medication titration. This is especially relevant for children, as a substantial proportion of kids receive low medication doses without appropriate upward titration, said Samuele Cortese, MD, PhD, senior author of the new review and National Institute for Health and Care Research (NIHR) research professor and professor of child and adolescent psychiatry at the University of Southampton, United Kingdom.

"In many countries, for children, there's always been a mistrust or reluctance to use these medications," he told Medscape Medical News, adding that after 1 year of treatment, only about 20% of patients remain adherent. That figure is linked partly to poor dose optimization, he said.

The review provides dosage clarification that is missing from earlier analyses, added first author Mikail Nourredine, PhD, Department of Biostatistics at Hospices Civils de Lyon in France. 

"Clinicians don't just choose a medication, they choose a dose. We wanted to give them evidence for that decision," he told Medscape Medical News.

Unlike previous reviews that compared drugs to placebo in isolated pairs, Nourredine and colleagues used a network meta-analysis framework, synthesizing data from 113 double-blind randomized controlled trials with more than 25,000 participants.

It drew on both direct and indirect evidence across the full trial network to estimate dose-response curves for all medications simultaneously. Because the same drug comes in different formulations that release differently, the team converted all doses to a common scale to enable comparison across trials. 

Amphetamine doses were expressed as dextroamphetamine equivalents and methylphenidate as immediate-release methylphenidate hydrochloride equivalents.

In children and adolescents, methylphenidate reached peak efficacy at approximately 45 mg/day (standardized mean difference [SMD], -0.89; 95% credible interval [CrI], -1.18 to -0.60), with no evidence of additional benefit at higher doses. 

Amphetamines peaked at approximately 25 mg/day (SMD, -1.06; 95% CrI, -1.35 to -0.78), corresponding to roughly 60-70 mg/day of lisdexamfetamine. Guanfacine peaked at 4 mg/day.

In adults, amphetamines plateaued above approximately 50 mg/day, while methylphenidate showed continued gains without a clear ceiling, though Nourredine cautioned that high-dose adult data were sparse. 

"The absence of a plateau for methylphenidate could be a real signal, but it could also reflect the limited data we had at higher doses," he said.

Online Interactive Dosing Tool

No evidence was found across any medication or age group that exceeding FDA-licensed maximum doses improved mean efficacy. Tolerability patterns mirrored the efficacy findings. In children, amphetamine discontinuation risk increased above 25 mg/day, coinciding with the efficacy plateau. 

In adults, both stimulant classes exceeded placebo discontinuation risk above 50 mg/day. The methylphenidate curve was particularly steep: At the FDA maximum of 60 mg/day, estimated discontinuation risk was 7.3%, rising to 10% at 90 mg/day, while efficacy improved only modestly.

"The medication choice is important, of course, but once you choose the medication, the dose is also very important," said Cortese.

Researchers emphasized that the curves are population-level benchmarks, not prescriptive thresholds. "At the individual level, we know that there is an important variability," he said. 

But he stressed that the findings should not discourage clinicians from titrating assertively within the effective range, noting that choosing the right medication is not enough if the dose is wrong. "Optimize the highest dose that's well tolerated, so that you maximize the benefit," Cortese said.

To help clinicians, the team created a free interactive online tool showing estimated efficacy and tolerability across doses for shared decision-making. 

Cortese said his group is now working with individual participant data to develop personalized dose-response models. "That will hopefully be a game changer," he said.

Limited Clinical Utility

While the dose-effect analysis represented a legitimate scientific effort, Goodman said it has limited bedside utility. 

"Pulling this research together under one roof in one publication is important," he said. "However, its clinical implications are moderated even by their own discourse in the limitation section."

Andrew Adesman, MD, chief of developmental and behavioral pediatrics at Cohen Children's Medical Center, Northwell Health, New Hyde Park, New York, offered a similar assessment. 

"I'm not a fan of the study as being especially instructive about guiding clinicians from a treatment standpoint, other than to be wary of not dosing kids adequately," Adesman, who was not involved in either publication, told Medscape Medical News.

The dose-effect review did not account for ADHD comorbidity, symptom severity, adjunct therapy, or outcomes beyond an 8-week timeframe, he added. Because the analysis grouped children as young as 6 with adolescents as old as 17, it could not capture the role of different medication delivery systems or the practice of adjusting the timing and size of individual doses throughout the day, he said. 

"If someone doesn't do well with 20 milligrams three times a day, you could do 20, 20, and 15, or 20, 20, and 10," Adesman added. "There are things that clinicians can do that aren't going to be captured in a randomized controlled trial."

He agreed, however, with the review's goal to help clinicians be "mindful about potentially underdosing," adding that they will be most successful in treating ADHD "if they have an appreciation about the many different medication options there are, and what some of the differences are among them."

Cortese, who chairs the European ADHD Guidelines Group and serves on the panel developing the forthcoming American Professional Society of ADHD and Related Disorders (APSARD) adult ADHD clinical practice guidelines, said the dose-effect findings could inform those efforts. 

"There is quite a lot of discussion on which medication to prescribe, and which side effects to monitor, but there is not much detail on the dose," he said. "We think that this paper may contribute to the forthcoming guidelines."

The APSARD guidelines were announced in 2022 but have been repeatedly delayed. As previously reported by Medscape Medical News, they are now expected later this year. APSARD did not respond to a request for comment on how the new consensus recommendations and dosage review might impact their much-anticipated adult ADHD guidelines.

The ASCP consensus statement received no external funding. The dose-effect review was funded by the National Institute for Health and Care Research. Disclosure information for consensus statement and study authors is available in the original publications. Goodman reported consulting relationships with the National Football League, the World Anti-Doping Agency, APSARD, Children and Adults with Attention-Deficit/Hyperactivity Disorder (CHADD), Medscape, and others. Adesman reported no relevant financial relationships.