A new study published today [14 May] in The Lancet Psychiatry provides the most comprehensive view of dosage effects for five commonly used medications for ADHD.

To help patients and clinicians choose the right dosage, the international research team led by Professor Samuele Cortese from the University of Southampton has also developed a free online tool based on the findings.

The research was funded by the National Institute for Health and Care Research (NIHR).

Attention Deficit Hyperactivity Disorder (ADHD) is one of the most common neurodevelopmental conditions, affecting about five per cent of school-age children and two to three per cent of adults.

Medication is a key part of treatment, and prescriptions have increased substantially in recent years. However, most clinical guidelines provide limited guidance on the most effective dosages.

Finding the right dose is important to avoid dosages that are too low to be effective or too high, causing unwanted side effects. To identify this dosage ‘sweet spot’, the research team analysed data from 113 clinical trials, including more than 25,000 participants.

They used an advanced method called dose–effects network meta-analysis, which allowed them to estimate how different doses of each medication affect both effectiveness and side effects.

The results show that patterns differ between medications and age groups.

Dr Mikail Nourredine from the University of Lyon, first author of the study, said: “Overall, our findings suggest that clinicians should avoid using doses that are too low to be effective. If symptoms are not well controlled, the dosage may need to be increased.

“We also found no evidence that going beyond the licensed maximum doses improves average effectiveness, and higher doses are usually linked to more side effects. However, our results derive from group averages. Specific individuals with ADHD may benefit from and tolerate well unlicensed doses.”

Evidence from other studies shows that a substantial proportion of children and adolescents are prescribed low dosages without appropriate increases. That’s despite timely and adequate dose adjustments being associated with better adherence to treatment.

Professor Cortese, an NIHR Research Professor at the University of Southampton, commented: “Our study and the tool have the potential to support shared decision-making between clinicians, patients, and families when choosing the best dose. It is not only a clinician’s decision – patients and caregivers should be involved.

“The tool helps show what can be expected from each dose so that the patient knows why that particular dose has been chosen. We are continuing research to further personalise these recommendations based on individual patient characteristics.”

The study Pharmacological interventions for ADHD: a systematic review and dose-effect network meta-analysis is published in The Lancet Psychiatry and is available online.

Steve Williams, Media Manager, University of Southampton, press@soton.ac.uk or 023 8059 3212.

1. The study Pharmacological interventions for ADHD: a systematic review and dose-effect network meta-analysis will be published in The Lancet Psychiatry. An advanced copy is available upon request.
2. The online tool is available here: https://mikailnourredine.github.io/ADORMA/
3. For interviews, please contact Steve Williams, Media Manager, University of Southampton, press@soton.ac.uk or 023 8059 3212.

Previous meta-analyses looking at the use of antidepressants during pregnancy and risk of neurodevelopmental disorders in children were conducted nearly a decade ago and limited by small study numbers and a lack of controlling for additional factors. This new meta-analysis provides the best evidence to date that the small increase in risk of autism or ADHD in the children of women who used antidepressants when pregnant identified in many studies is not caused by the medication.

“We know many parents-to-be worry about the potential impact of taking medication during pregnancy; our study provides reassuring evidence that commonly used antidepressants do not increase the risk of neurodevelopmental disorders such as autism and ADHD in children. While all medications carry risks, so too does stopping antidepressants during pregnancy due to an increased risk of relapse. Therefore, for women with moderate-severe depression, doctors and patients must carefully weigh the potential risks and benefits of continuing antidepressant treatment during pregnancy against the potential harms of untreated depression,” says author Dr Wing-Chung Chang, University of Hong Kong.

He continues, “Although our study found a small increase in the risk of autism and ADHD in the children of women who had used antidepressants during pregnancy, it also found that this risk disappeared when we accounted for other factors. The increased risk was also seen in the children of fathers who took antidepressants and of mothers with antidepressant use before, but not during, pregnancy. Together, this suggests that it is not the antidepressants themselves causing an increased risk in autism and ADHD but it is more likely to be due to other factors, including genetic predisposition to conditions such as ADHD, autism, and mental health conditions.”

Authors pooled data from 37 studies which included more than 600,000 pregnant women taking antidepressants and almost 25 million pregnancies with no antidepressant use.

Before controlling for key factors such as mental health conditions, the analysis found that antidepressant use by the mother during pregnancy was associated with a 35% increased risk of ADHD and a 69% increased risk of autism. However, this became greatly reduced or non-significant in analyses that better controlled for confounding factors. Use of antidepressants during pregnancy by the father was associated with a 46% increase in the risk of ADHD and a 28% increase in the risk of autism.

Among studies with analyses restricted to mothers with mental health disorders, all selective serotonin reuptake inhibitors (SSRIs) were found to not be associated, only amitriptyline/nortriptyline remained associated with increased ADHD and autism risk. Amitriptyline/nortriptyline are currently considered second or third options as treatments for depression and are often prescribed for treatment-resistant depression. Therefore, women treated with these may have more severe, chronic, or complex underlying mental health conditions than those receiving more common antidepressants, which could be influencing the association between amitriptyline/nortriptyline and increased ADHD and autism risk.

The study found no difference in risk between high and low doses of antidepressants.

“The evidence suggests a link between either parent having a mental health condition and a slightly higher risk of ADHD or autism. In addition to genetic factors, this link could be explained by the home and social environment as ongoing family stress, changes in how the family functions, and differences in how parents behave and care for their children may influence neurodevelopment. There is a need to ensure both parents have access to support and treatment for mental health conditions; for their own sake and to support neurodevelopment of their child,” says Dr. Joe Kwun-Nam Chan, University of Hong Kong.

The researchers note some limitations of their study, including that data on important factors such as socioeconomic status, lifestyle risk factors and low birth weight was lacking in the studies. Additionally, there was only a small number of studies looking at antidepressant use in specific trimesters or exact doses and dose changes, which makes it harder to draw conclusions about these. Finally, women who are prescribed antidepressants tend to have more severe depression than those who are not, so some bias may remain even after controlling for mental health status.

Writing in a linked Comment, Lisa Vitte, Emmanuel Devouche and Gisele Apter from University Rouen Normandy (France), who were not involved in the study, say, “Chang and colleagues’ study adds knowledge and confirms some of the pre-existing knowledge on the use of antidepressants during pregnancy: that they should continue to be taken as they protect maternal mental health and do not harm fetal development. This result is of considerable impact after many contradictory and controversial studies.”

There was no funding source for this study. The study was conducted by researchers from the University of Hong Kong.

The labels have been added to this press release as part of a project run by the Academy of Medical Sciences seeking to improve the communication of evidence. For more information, please see: http://www.sciencemediacentre.org/wp-content/uploads/2018/01/AMS-press-release-labelling-system-GUIDANCE.pdf. If you have any questions or feedback, please contact The Lancet press office at pressoffice@lancet.com

Quotes from Authors cannot be found in the text of the Article but have been supplied for the press release. The Comment quote is taken directly from the linked Comment.

https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(26)00089-1/fulltext

The study comes amid more than one million prescription approvals and a tripling of sales of cannabinoid medications (including both cannabidiol (CBD) and tetrahydrocannabinol (THC) products) in Australia over the past four years, often for the treatment of mental health and substance-use disorders.

The study’s lead author, Dr Jack Wilson at the University of Sydney’s The Matilda Centre, said the results call into question the approval of medicinal cannabis for the treatment of depression, anxiety and PTSD.

“Though our paper didn’t specifically look at this, the routine use of medicinal cannabis could be doing more harm than good by worsening mental health outcomes, for example a greater risk of psychotic symptoms and developing cannabis use disorder, and delaying the use of more effective treatments,” he said.

More than 700,000 Australians have reported using medicinal cannabis to treat over 250 different health conditions. The research found evidence to suggest that medicinal cannabis could potentially be beneficial for some conditions – such as the treatment of cannabis use disorder (otherwise known as cannabis dependency), autism, insomnia, and tics or Tourette’s syndrome.

Dr Wilson said: “But the overall quality of evidence for these other conditions, such as autism and insomnia, was low. In the absence of robust medical or counselling support, the use of medicinal cannabis in these cases are rarely justified.

“There is, however, evidence that medicinal cannabis may be beneficial in certain health conditions, such as reducing seizures associated with some forms of epilepsy, spasticity among those with multiple sclerosis, and managing certain types of pain, but our study shows the evidence for mental health disorders falls short.

“In the case of autism specifically, while the study showed some evidence medicinal cannabis could assist with a reduction in symptoms, it is worth noting that there is no one – or universal – experience of autism, so this finding should be treated with caution.”

The study found that medicinal cannabis was not effective for every type of substance-abuse disorder. While medicinal cannabis may help with cannabis dependence, it was found to increase cocaine cravings among people with cocaine-use disorder

“Similar to how methadone is used to treat opioid-use disorder, cannabis medicines may form part of an effective treatment for those with a cannabis-use disorder. When administered alongside psychological therapy, an oral formulation of cannabis was shown to reduce cannabis smoking,” Dr Wilson said.

“However, when medicinal cannabis was used to treat people with cocaine-use disorder, it increased their cravings. This means it should not be considered for this purpose and may, in fact, worsen cocaine dependence,” he said.

Researchers urge greater regulation for prescribing of medicinal cannabis 

The rapid expansion in medicinal cannabis use and prescribing rates has raised concerns among major medical bodies, including the Australian Medical Association (AMA) and the Pharmacy Guild of Australia, about the largely unregulated growth in prescribing and the uncertainty surrounding the efficacy and safety of these products.

In response, the Therapeutic Goods Administration (TGA) initiated a review of the regulatory oversight of medicinal cannabis, with more than 500 responses published in February.

“Our study provides a comprehensive and independent assessment of the benefits and risks of cannabis medicines, which may support the TGA and clinicians to make evidence-based decisions, helping to ensure patients receive effective treatments while minimising harm from ineffective or unsafe cannabis products,” Dr Wilson said.

The systematic review and meta-analysis included results from 54 randomised controlled trials (RCTs) published over a 45-year period (1980-2025) worldwide.

Research: Wilson, J et al, ‘The efficacy and safety of cannabinoids for the treatment of mental disorders and substance use disorders: a systematic review and meta-analysis’ (The Lancet Psychiatry, 2026)

DOI: https://doi.org/10.1016/S2215-0366(26)00015-5

In a paper, published today in the British Journal of Psychiatry, a group of experts led by Professor Samuele Cortese from the University of Southampton say there is no robust evidence that ADHD is over-diagnosed in the UK.

The new paper refutes the view that ‘nowadays everyone has ADHD’ which is gaining traction in public discourse and has been amplified by some leading politicians, as demand rises for NHS assessments and services.

Bringing together academics, clinicians, people with lived experience and carers, the group say this narrative risks misleading the public and policymakers and overshadows a more pressing concern – unmet need.

Professor Cortese said: “Rather than focusing on increases or decreases in diagnostic rates, attention should be directed toward the extent to which those with ADHD are being adequately diagnosed and treated.

“While misdiagnosis and inappropriate diagnosis do occur, the available evidence indicates that under diagnosis and under treatment remain the predominant challenges.”

When standardised diagnostic criteria are applied, the prevalence of ADHD internationally is around 5 per cent in children and 3 per cent in adults.

While prevalence has increased over time, NHS administrative data in England remains substantially below these expected levels, suggesting that many people with ADHD are living without a diagnosis and adequate support.

The group acknowledge that misdiagnosis can occur in some cases, particularly where assessments rely heavily on self-reporting or where alternative conditions are not fully considered.

Professor Tamsin Ford, Head of the Department of Psychiatry at the University of Cambridge, a coauthor on the paper, commented: “While many more people with ADHD are being recognised and treated, we are failing to support many more. Overdiagnosis is not a problem, but misdiagnosis may be as people are driven into the private sector by long waits; and sadly, missed diagnoses remain common.”

The researchers stress that the absence of biological diagnostic markers means that thorough, multidisciplinary clinical assessment is essential. Field trials show that when clinicians are properly trained, an ADHD diagnosis is among the most reliable for a mental health condition.

“Similar to physiological traits, such as blood pressure or weight, ADHD symptoms are distributed along a continuum,” says coauthor Professor Chris Hollis from the University of Nottingham. “But as with hypertension or obesity, there are diagnostic severity thresholds that determine health risks and what interventions should be used. Similarly, in ADHD a risk-stratified stepped-care approach may be useful.”

Professor Cortese and colleagues highlight significant pressure on UK services, with long waiting times and growing demand, especially among adults who were not diagnosed in childhood.

They point to figures showing that around 27 per cent of children and young people diagnosed with ADHD reported waiting one to two years, while 14 per cent waited two to three years.

Evidence shows that untreated ADHD is associated with serious long-term risks, while effective treatments are available, backed by strong evidence, and generally well tolerated.

“The costs of untreated ADHD are often overlooked,” says Professor Cortese. “They include increased risk of academic failure, suicidal behaviour, substance abuse, criminality, injury and death. The failure to provide treatments which have been shown to reduce these risks represents a major ethical issue that needs to be urgently addressed.”

The authors call for improved funding, workforce training and a more balanced, evidence-based conversation to ensure accurate diagnosis while expanding access to care for those who need it.

The researchers are supported by the National Institute for Health and Care Research (NIHR), UK Research and Innovation (UKRI) and the Office for Life Sciences.

ADHD (over) diagnosis: fiction, fashion, and failure is published in the British Journal of Psychiatry and is available online.

Contact

Steve Williams, Media Manager, University of Southampton, press@soton.ac.uk or 023 8059 3212.

1. In 2018, the administrative prevalence was 2.5% in boys and 0.7% in girls, and 0.7% in men and 0.2% in women, respectively. Current post-pandemic administrative prevalence data are unavailable.
2. Figures on wait list times come from an online survey of 7,340 people, conducted by the House of Commons Petitions Committee between 2021 and 2022.
3. ADHD (over) diagnosis: fiction, fashion, and failure is published in the British Journal of Psychiatry here:ADHD (over) diagnosis: fiction, fashion and failure | The British Journal of Psychiatry | Cambridge Core
4. For interviews, please contact press@soton.ac.uk or 023 8059 3212.

The researchers analysed Finnish national registry data on 209,346 older adults living at home who did not have Alzheimer’s disease. They compared three groups: off-label users, people taking antipsychotics for an official indication, and those not using the medications at all.

The results were striking: 70 percent of all antipsychotic use fell into the off-label category. Use of these medications off-label also increased over the course of the study period. Risperidone and quetiapine were the most commonly used antipsychotics off-label.

Physicians often prescribe antipsychotics to older patients to manage prolonged behavioural symptoms or insomnia, despite the absence of official approval for these conditions. Previous research has even suggested that using these drugs without alternative non-drug treatments can constitute a form of elder abuse. Limited resources in elderly care and a lack of non-pharmacological options appear to have contributed to the growing use of these medications.

The study found that off-label users were more likely than other groups to have cardiovascular diseases and cerebrovascular disorders. They were also more likely to be using opioids.

“Surprisingly, off-label users had cardiovascular and cerebrovascular conditions even more frequently than patients prescribed antipsychotics for an approved indication, such as psychotic disorders,” said postdoctoral researcher Tuomas Majuri of the University of Oulu.

Majuri noted that this finding could reflect insufficient monitoring of the metabolic side effects of these medications when used off-label.

The researchers say their findings underscore the need for alternative treatments for symptoms currently managed with off-label antipsychotic use.

“Clearer guidelines are needed for monitoring metabolic side effects, and measures must be taken to promote safer prescribing practices,” Majuri added.

The study was funded by the Päivikki and Sakari Sohlberg Foundation and the Research Council of Finland.

Article: Majuri T*, Silvan J*, Tolppanen A-M, Hartikainen S, Huotari T, Rautio N, Miettunen J, Nordström T, Jääskeläinen E, Haapea M. Characteristics and predictors of antipsychotic medication off-label use among community-dwelling older people. Nord J Psychiatry 2026. Published online 16 Feb 2026.

Indeed, a research letter by Rutgers and Columbia University researchers in JAMA Psychiatry shows those prescriptions are becoming more common in the United States, even though antipsychotic drugs do little for dementia and carry a black-box warning on their labels stating they increase the risk of death in senior patients.

Using a national prescription-claims database that captures more than 90% of retail pharmacy fills, researchers tracked antipsychotic use among adults 65 and older from 2015 through 2024 and found that the annual rate of any antipsychotic use increased nearly 52% to 4.05 per 100 from 2015 to 2024. Long-term use, defined as at least 120 days a year, rose 65% to 2.45 per 100 older adults. Rates were highest among people 75 and older, rising from 3.42 to 5.12 per 100.

The trend is striking because antipsychotics have limited proven effectiveness in people 65 and older and serious risks, including falls, fractures, cardiovascular and cerebrovascular events, pulmonary embolism and death.  Antipsychotics may be used as a last resort to manage severe behavioral and psychological symptoms of dementia, such as aggression, agitation, hallucinations, or delusions, especially when these symptoms pose a risk to the safety of the individual or others.  However, such use carries substantial risk and should be avoided in most cases and limited to short-term use whenever possible.

“The evidence is pretty solid on the risks,” said Stephen Crystal, the letter’s co-author and director of the Center for Health Services Research at the Rutgers Institute for Health, Health Care Policy and Aging Research.

The claims data don’t include diagnoses, so the researchers couldn’t determine why each prescription was written or whether it was appropriate. Antipsychotics remain essential for some people, including those with schizophrenia, bipolar disorder with psychosis or other severe psychiatric illnesses.

However, Crystal noted those conditions aren’t common enough in older populations to explain the surging number of antipsychotic prescriptions.

“We think that conditions like schizophrenia that have FDA-approved indications for antipsychotic treatment are unlikely to account for the majority of the rates of use that we observed,” he said.

The biggest concern for the researchers is using antipsychotics for the behavioral and psychological symptoms that can accompany dementia: agitation, wandering, acting out behavior and shouting. In many cases, the medications are used to “damp down” behaviors that are distressing to caregivers and disruptive to facilities, said Crystal, who also holds endowed professorships at the Institute for Health and Rutgers School of Social Work.

Because the drugs can be highly sedating, they reduce the tendency to roam and act out, but that sedation comes with a steep tradeoff for frail patients, increasing fall risk and reducing physical activity.

The study also reveals a shift in who manages cases. Among patients who took an antipsychotic in a given year, the share with at least one prescription from a psychiatrist fell from 30% in 2015 to 20% in 2024. Over the same period, the share who filled an antipsychotic from a pharmacy in a long-term care facility rose from 14% to 21%.

Crystal said the decline in psychiatrist involvement matters because optimal care for behavioral symptoms in dementia often starts with careful evaluation rather than a quick prescription. Clinicians may need to confirm the diagnosis and look for treatable causes that can mimic or worsen confusion, including medication interactions, infections, depression and unmanaged pain. Even when dementia is the main driver, nondrug approaches can work, but they require training, staffing and time.

“This can look like managing symptoms,” Crystal said. “Which is common because it’s so much easier to write a prescription than do the work of addressing the underlying condition, particularly at nursing home and assisted living facilities that are dangerously short-staffed.”

There was one potentially encouraging sign in the data: the use of first-generation antipsychotics, which are associated with higher mortality risk in older patients than second-generation medications, declined from 22% to 14%.

Still, the overall rise in use and the growth in long-term prescribing suggest a system leaning more heavily on medication to solve problems that are often social, environmental and staffing-related. The authors have called for renewed efforts to evaluate and spread nonpharmacological interventions that can reduce reliance on antipsychotics in older adults.

For families contending with a new prescription, the study’s lead author, Mark Olfson of the Columbia University Department of Psychiatry, said that it is reasonable to ask what problem the drug is meant to address and what other steps have been tried. Just as important is what happens next: whether the clinician has a plan to reassess, taper and stop the medication once a crisis has passed.

“These are high-stakes decisions,” he said.

The authors emphasise that these findings do not suggest that antidepressants are unnecessary or that psychotherapy alone is adequate. Instead, they say the results underscore the importance of tailoring deprescribing to each individual, with a gradual individualised tapering of antidepressants alongside structured psychological support.

“Depression is often a recurring condition and without ongoing treatment as many as three out of four people with recurrent depression relapse at some point,” said lead author Professor Giovanni Ostuzzi from the University of Verona in Italy. “Clinical guidelines recommend continuing antidepressants for a certain period after remission, then considering discontinuation once the person has remained well. Yet in everyday practice, treatment is often prolonged far beyond what guidelines suggest. Many people do not wish to stay on medication indefinitely, and some experience troublesome side-effects. Still, there has been little rigorous research into the safest and most effective approaches to discontinue treatments.”

He continued, “By incorporating a substantially larger evidence base, a broader range of deprescribing strategies, and direct head-to-head comparisons, our new review clarifies the scientific evidence about the most effective way to come off antidepressants for individuals successfully treated for depression and could change how coming off antidepressants is managed globally.”

He added, “We encourage anyone considering coming off antidepressants to discuss the process with their doctor first to jointly find the best strategy for them.”

Evidence-based guidelines for moderate-to-severe depression and anxiety disorders typically recommend that antidepressants be taken for six to nine months after a first episode, or for up to two years in people who have experienced multiple recurring episodes or have specific risk factors for relapse. But there are concerns about overprescribing, long-term use, and withdrawal symptoms after discontinuation which underscore the need for evidence-based deprescribing strategies.

To address this evidence gap, researchers conducted the largest and most rigorous analysis of randomised controlled trials to date, examining antidepressant deprescribing strategies up to April 2025.

They analysed data from 76 randomised controlled trials involving 17,379 adults (average age of 45 years, 68% female, 88% White) with fully or partially remitted depression (defined as mild-to-negligible symptoms but diagnostic criteria not fulfilled; 60 studies) or anxiety disorders (16 studies) receiving antidepressant treatment (mostly selective serotonin-reuptake inhibitors [SSRIs] or serotonin norepinephrine reuptake inhibitors [SNRIs]) who were followed for an average of 10 to 11 months.

The network meta-analysis directly compared the effectiveness of all major strategies for deprescribing on the rate of relapse (a new episode of depression or anxiety) each with or without psychological support:

• sudden stopping of an antidepressant (replaced by a placebo pill)
• fast tapering (reducing the medication dose over four weeks or less)
• slow tapering (reducing the medication dose over more than four weeks)
• dose reduction (50% or less of the minimal effective dose)
• or continuing antidepressant treatment

Overall, the analysis found moderate-quality evidence that after successful treatment of depression, slow tapering of antidepressants combined with psychological support prevented relapse over the following year to a similar extent as remaining on an antidepressant at standard doses (either with or without psychological support). These were the most effective strategies for reducing the rate of depression relapse. (See figure 3 in the paper for the full comparison of strategies).

The researchers estimated that slow tapering of antidepressants plus psychological support could prevent one relapse in every five individuals compared with abrupt stopping or fast tapering (the two least effective strategies) – offering a clinically meaningful benefit.

The analysis also found that continuing with reduced antidepressant doses was better than abruptly stopping and fast tapering for reducing relapse, but the evidence for this was less strong. Additionally, combining psychological support with either slow or fast tapering was significantly more effective at reducing relapse than fast or slow tapering alone, however the quality of evidence for this was low.

No significant differences in reported side effects or number of dropouts were seen between the deprescribing strategies. However, information on withdrawal symptoms and other clinically relevant factors was scarce or missing, limiting insights into the potential influence of withdrawal symptoms, as well as quality of life and social functioning.For anxiety, the authors note that the evidence was less robust and requires confirmation in dedicated trials to provide doctors and patients with accurate evidence-based information on which to base their decisions.

“Our findings suggest that while antidepressants are effective in preventing depressive relapses, they do not need to be a long-term treatment for everyone,” said co-author Dr Debora Zaccoletti from the University of Verona. “Safe alternative treatments like psychological support, including cognitive behavioural and mindfulness-based therapies, can be a promising tool—even in the short-term. However, considerable healthcare resources are needed to develop and implement dedicated psychotherapy approaches in clinical practice so more cost-effective, short-term, scalable and remotely delivered interventions should be tested and prioritised.”

She added, “These findings highlight the need for clinical guidelines to be updated to promote regular treatment reviews and individualised deprescribing with gradual tapering and structured psychological support for patients with depression who are feeling better and wish to come off their medication.”

Writing in a linked Comment, Dr Jonathan Henssler, Charité–Universitätsmedizin Berlin (Germany), who was not involved in the study, said: “One of the most consistent findings of the analysis is that adjunctive psychological support or psychotherapy resulted in improved outcomes for all the different pharmacological strategies. Although this might seem a trivial finding, it is of substantial clinical importance.

He continues, “Another crucial finding from Zaccoletti and colleagues’ analysis […] warrants emphasis: the best patient outcomes were achieved with strategies that maintained antidepressant therapy. This finding serves as a reminder of the severity and chronicity of depression in many cases and, notwithstanding the preference for minimising pharmacotherapy duration, highlights the limitations of the current curative capacity of antidepressant therapies. Importantly, this limited capacity applies equally to psychotherapy, whose effect sizes are no better than those of pharmacotherapy.”

To diagnose a mental illness, medical practitioners rely on a variety of factors, including the symptoms reported by patients and recorded on clinical questionnaires. The precise wording of individual questions on these questionnaires is often crucial for making the correct diagnosis. However, standard questionnaires often vary considerably. Researchers have found evidence of overlaps and deviations in the content of questions used to identify depression, bipolar disorder, and the risk of psychosis, which makes accurate diagnosis difficult.

In addition, doctors rely on their clinical experience. This means that they associate individual symptoms with a specific illness that corresponds to their experience. However, as different illnesses can produce the same or similar symptoms, this can also increase the risk of misdiagnosis. “We know surprisingly little about whether – and how – the wording of clinical questionnaires triggers certain associations in doctors,” says Professor Joseph Kambeitz. Inconsistent findings could also result from differences among patients in the same diagnostic group or, alternatively, from differences between questionnaires.

Using large language models (LLMs) is one approach to analysing language-mediated illness descriptions. The team used the LLMs GPT-3, Llama and BERT to analyse both the structure and content of four clinical questionnaires. The study was based on data from over 50,000 questionnaires on depression, anxiety, psychosis risk, and autism.

In clinical practice, symptoms often occur simultaneously, such as the empirical association between a lack of drive and a loss of pleasure. The analysis showed that the LLMs ‘recognize’ which symptoms commonly occur together. Even without access to specific empirical data, the same symptom associations are evident in LLMs based purely on the questionnaire formulations.

This suggests new ways in which artificial intelligence could improve psychological questionnaires in future, by avoiding redundant items and making diagnosis and understanding of mental illnesses more efficient. LLMs can be used to develop questionnaires that are both precise (i.e. that reliably recognize psychological symptoms) and efficient, asking only as many questions as necessary in order to simplify the process for patients and practitioners.

“AI can map both medical knowledge and the structures of mental illnesses. This is an important step in bringing digital methods and neuroscience closer together, and in advancing the development of diagnostics and research in psychiatry,” says Professor Kai Vogeley.

Professor Joseph Kambeitz concludes: “In psychiatry, the ‘spoken word’ plays an important role in diagnosis and therapy. There are currently many promising projects that are investigating how we can use LLMs in psychiatry, from diagnostics via the writing and amending of reports to the simulation of therapy sessions. We can expect many more exciting research results in this field.”

Acetaminophen (paracetamol) is the most commonly used over-the-counter pain reliever during pregnancy, taken in an estimated 70% of all pregnancies. In 2021, a commentary about a possible increased risk of NDDs in children exposed to acetaminophen in utero was published, leading to substantial concerns in the general population and among prescribers. However, this publication—as well as a subsequent narrative review by the same group of researchers—was met with criticism of its methodology, choice and quality of studies considered, and lack of mechanistic data.

Acetaminophen exposure during pregnancy is difficult to assess in epidemiological studies because it is available over the counter without a prescription and is used as needed. Hence, there is inconsistency between studies regarding the risk of NDDs associated with its use in pregnancy.

“Given the significant methodological challenges in this area, additional individual studies will inevitably suffer the same limitations as are already present in published studies,” explains lead author of the current systematic review and meta-analysis Anick Bérard, PhD, University of Montreal and CHU Sainte-Justine, Montréal, Quebec, Canada. “An integrative approach, summarizing the present state of knowledge and quantifying specific methodological areas of biases, as our study does, is needed in order to have a significant impact in future studies performed as well as on the causal effect of acetaminophen use during pregnancy on the risk of specific NDDs.”

Dr. Bérard and an international group of experts in the field conducted a comprehensive search of major bibliographic databases and grey literature to identify human studies evaluating the association between prenatal acetaminophen exposure and the risk of NDDs in offspring. Sixteen studies retrieved met eligibility criteria. In the resulting novel systematic review and meta-analysis, the investigators applied rigorous systematic review methodologies to determine the extent to which current data can support an association between prenatal exposure to acetaminophen and the risk of NDDs in children. They also used quantitative bias analysis to provide an estimate of the direction, magnitude, and uncertainty arising from systematic errors when assessing acetaminophen use during pregnancy and the risk of ADHD in children and performed sensitivity analyses.

Dr. Bérard notes that while a modest but significant association between acetaminophen use during pregnancy and ADHD in children was initially observed, concerns about its reliability remain. She explains, “This association is unlikely to be explained by possible confounding factors that were assessed but were not seen when sibling controls (considered to be the most reliable study design for assessing this risk) were used.”

The authors conclude that the observed increase in ADHD risk is likely not causal and can be explained by inherent biases and underlying genetic factors. They recommend further studies to examine this association more robustly.

Commenting on the study, noted expert in the field David Coghill, MD, Financial Markets Foundation Chair of Developmental Mental Health in the Departments of Paediatrics and Psychiatry at the University of Melbourne, observes, “This is a landmark study that highlights the importance of good science and strong methods. The findings of no association between acetaminophen use during pregnancy and the risk of autism and ADHD in the child are not unexpected. These new findings support the position of professional organizations and regulatory bodies around the world that women should continue to use acetaminophen during pregnancy and do so without fear. The fact that they contradict the recent announcements by the US government must be acknowledged and acted on.”

Participants enrolled in a total diet replacement (TDR) programme experienced significant improvements in eating disorder symptoms which persisted six months after the programme finished when participants had regained some weight. This directly addresses a research gap highlighted by recent National Institute for Health and Care Excellence (NICE) guidance on the use of ‘Low-energy and very-low-energy diets for adults’ in Overweight and obesity management, which noted a lack of evidence on the psychological impact of restrictive diets.

“As a dietitian, I’ve seen first-hand how both eating disorders and type 2 diabetes can impact the quality of people’s lives,” said lead author Dr Elena Tsompanaki, a registered dietitian who led the study as part of her doctoral research at the Nuffield Department of Primary Care Health Sciences, University of Oxford. “Our findings challenge the assumption that weight loss programmes worsen eating disorder symptoms in vulnerable people, potentially opening up important treatment options that many patients might have previously been denied.”

The researchers recruited 56 participants with type 2 diabetes (diagnosed within the last six years), overweight or obesity, and existing eating disorder symptoms identified via a validated questionnaire (called Eating Disorders Examination Questionnaire or EDE-Q). Half were assigned to receive a low-energy (~860 kcal/day) TDR programme with behavioural support for six months – similar to the NHS Path to Remission programme- while the control group received usual diabetes care from their GP. It assessed four types of symptoms: restraint eating, eating concerns, shape concerns, and weight concerns.

Key findings:

• Eating disorder symptoms significantly improved in the TDR group compared to usual care at 6 months (-0.8 points EDE-Q difference) and this benefit persisted at 12 months (-0.7 points).
• The TDR group also saw significant reductions in depression and diabetes distress symptoms at 6 and 12 months compared to usual care.
• The TDR group lost more weight at 6 months (-13.9kg vs -3.7kg), but the difference between groups was no longer statistically significant at 12 months.
• Crucially, no participants were suspected of developing a new eating disorder.

Professor Susan Jebb, OBE, study author and Professor of Diet and Population Health at the Nuffield Department of Primary Care Health Sciences, who was also involved in evaluating the NHS Path to Remission programme, commented: “The NHS Path to Remission programme is already showing promising results, with 27% of participants achieving diabetes remission at one year. Our new findings provide reassurance that these programmes can be safely offered to patients with symptoms of eating disorders, providing an opportunity for them to benefit from these effective treatments.”

The findings are important as there have been concerns that weight loss programmes, like the NHS Path to Remission, might not be safe or appropriate for people at high risk of eating disorders who might otherwise benefit from weight loss.

Dr Dimitrios Koutoukidis, senior author also based at the Nuffield Department of Primary Care Health Sciences, added: “This study provides crucial reassurance that this type of intervention does not appear to harm, and may even benefit, people with type 2 diabetes already experiencing eating disorder symptoms. It suggests that receiving structured support to lose weight may improve not only the physical health of this vulnerable group but also aspects of their mental health too.

“Low-energy total diet replacement programmes can be a very effective option for those who wish to give them a go to manage their diabetes,” Dr Tsompanaki concluded. “I hope these results help patients and healthcare professionals feel more confident in navigating the discussion around these treatment options.”

Limitations include the small sample size and participant demographics (see Notes). The team, who worked with patient advisors throughout the study, are continuing to follow-up participants to explore how their eating disorder symptoms change in the longer term.

• Publication: ‘An intensive weight loss programme with behavioural support for people with type 2 diabetes at risk of eating disorders: the ARIADNE non-inferiority randomised controlled trial’ The Lancet Psychiatry
  https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(25)00126-9/fulltext  (after embargo lifted at 00:01 BST Wednesday 11 June 2025)
• Media contact:  Communications Team, Nuffield Department of Primary Care Health Sciences, University of Oxford.  Email: communications@phc.ox.ac.uk
• Funding: This trial was funded by the Novo Nordisk UK Research Foundation (Doctoral Fellowship to ET). Further support from NIHR Advanced Fellowship (DK), NIHR Oxford Biomedical Research Centre (SJ, PA, DK), NIHR Applied Research Collaboration Oxford and Thames Valley (SJ). Medichecks donated HbA1c kits (used to diagnose diabetes, diabetes risk, or monitor blood sugar levels).  The funders had no role in study design, data collection, data analysis, data interpretation, or writing of the study. Views are authors’ own.
• Limitations: Limitations noted by the authors include online recruitment; a modest skew towards more affluent, White ethnic background participants vs UK diabetes population; self-reported weight; small sample size (n=56), relatively short-term follow-ups.
• Patient and Public Involvement:
  Eight public members advised on study design, materials, and delivery.
• The NHS Path to Remission programme is offered to people within the first 6 years of diagnosis of type 2 diabetes. It uses a similar total diet replacement programme and early evaluation has shown 27% of people with available data achieved remission at 1 year. https://pubmed.ncbi.nlm.nih.gov/39116897/

The findings, published in JAMA Psychiatry, showed that weekly injections of semaglutide – compared with placebo injections – reduced alcohol craving, drinking quantity and the frequency of heavy drinking days in adults with symptoms of alcohol use disorder. Results also showed that after treatment, those in the semaglutide group consumed lower amounts of alcohol in the laboratory, as measured by grams of alcohol consumed and breath alcohol concentration.

The discovery could help address an important treatment gap: An estimated 178,000 U.S. deaths per year can be attributed to alcohol, which is linked to liver disease, cardiovascular disease and is a known cause of cancer, as noted recently by the U.S. Surgeon General. A significant proportion of American adults have met criteria for alcohol use disorder at some point in their lives — yet relatively few seek or receive treatment. The three drugs currently approved by the FDA to treat alcohol use disorder are under-utilized. The popularity of Ozempic and other GLP-1 receptor agonists could increase the chances of broad adoption of these treatments for alcohol use disorder.

The Clinical Trial

For the trial, researchers recruited 48 adults with alcohol use disorder who weren’t actively seeking treatment. Alcohol use disorder is defined by a range of possible symptoms, including the inability to stop or control one’s drinking despite negative consequences.

One week prior to the first injection, researchers invited participants to drink their preferred alcohol beverages over a two-hour period in a comfortable lab setting, with instructions to delay drinking if they wished. Researchers documented delays and drinks consumed.

Participants were then randomly assigned to receive weekly injections of Ozempic or a placebo for nine weeks, during which time their weekly drinking patterns were also measured. Afterward, participants and researchers returned to the drinking lab to repeat the process.

Results showed that after treatment, those in the semaglutide group consumed lower amounts of alcohol in the laboratory, as measured by grams of alcohol consumed and breath alcohol concentration.

Clinical assessments also indicated that semaglutide (compared to placebo injections) reduced weekly alcohol craving, reduced average drinks on drinking days, and led to greater reductions in heavy drinking days, relative to placebo. A key finding was that the magnitude of semaglutide’s effects on several drinking outcomes appeared potentially greater than is often seen in similar studies with existing AUD medications, even though semaglutide was only administered at the lowest clinical doses.

Among a small subgroup of participants who smoked cigarettes at baseline, those treated with semaglutide had significantly greater reductions in average cigarettes per day compared to those in the placebo group. This finding is potentially important because there are no medications currently approved for both alcohol reduction and smoking cessation.

“The first clinical trial testing the impact of an older GLP-1 receptor agonist on alcohol use in humans was inconclusive,” said Klein. “However, as prescription of semaglutide and similar medications escalated, anecdotal reports of reduced alcohol use became very common, and suggested the potential of these more potent therapies for treatment of alcohol use disorder.”

Research is needed to understand the mechanisms by which GLP-1 receptor agonists reduce alcohol cravings. Dr. Klein says that preclinical studies suggest that these effects are likely mediated in the brain and involve changes in reward processing.

Although exciting, these data are preliminary and there is a need to learn more. As clinical use of this medication increases, the findings call for additional studies to evaluate long-term effect on alcohol consumption, and the ideal doses and treatment durations, which may be different from the current recommendations for people living with diabetes and obesity.

“These data suggest the potential of semaglutide and similar drugs to fill an unmet need for the treatment of alcohol use disorder,” said Klein. “Larger and longer studies in broader populations are needed to fully understand the safety and efficacy in people with alcohol use disorder, but these initial findings are promising.”

About this Study

In addition to Hendershot and Klein, other authors of the study are Michael Bremmer, Michael Paladino, Georgios Kostantinis, Thomas Gilmore, Neil Sullivan, Amanda Tow and Robyn Jordan, all of University of North Carolina at Chapel Hill; Sarah S. Dermody of Toronto Metropolitan University; Mark Prince of Keck School of Medicine; Sherry A. McKee of Yale School of Medicine; Paul J. Fletcher of University of Toronto; and Eric D. Claus of Pennsylvania State University.

This research was supported by National Institute on Alcohol Abuse and Alcoholism grant R21AA026931.

A new national survey of 1,000 American adults commissioned by The Ohio State University Wexner Medical Center and College of Medicine finds that 25% of adults now suspect they may have undiagnosed ADHD. But what worries mental health experts is that only 13% of survey respondents have shared their suspicions with their doctor.

That’s raising concerns about the consequences of self-diagnosis leading to incorrect treatment.

“Anxiety, depression and ADHD – all these things can look a lot alike, but the wrong treatment can make things worse instead of helping that person feel better and improving their functioning,” said psychologist Justin Barterian, PhD, clinical assistant professor in Ohio State’s Department of Psychiatry and Behavioral Health.

An estimated 4.4% of people ages 18 to 44 have ADHD, and some people aren’t diagnosed until they’re older, Barterian said.

“There’s definitely more awareness of how it can continue to affect folks into adulthood and a lot of people who are realizing, once their kids have been diagnosed, that they fit these symptoms as well, given that it’s a genetic disorder,” Barterian said.

The survey found that younger adults are more likely to believe they have undiagnosed ADHD than older generations, and they’re also more likely to do something about it.

Barterian said that should include seeing a medical professional, usually their primary care provider, to receive a referral to a mental health expert to be thoroughly evaluated, accurately diagnosed and effectively treated.

“If you’re watching videos on social media and it makes you think that you may meet criteria for the disorder, I would encourage you to seek an evaluation from a psychologist or a psychiatrist or a physician to get it checked out,” Barterian said.

What is Adult ADHD?
Adults struggling with ADHD will have problems with paying attention, hyperactivity and impulsivity that are severe enough to cause ongoing challenges at school, work and home. These symptoms are persistent and disruptive and can often be traced back to childhood.

Adult ADHD occurs in:

• Adults who were diagnosed as children, but symptoms continue into adulthood.
• Adults who are diagnosed for the first time, despite experiencing symptoms since they were younger that had been ignored or misdiagnosed.

Hyperactivity as a symptom is typically less present in adults than in children. Many adults with ADHD struggle with memory and concentration issues. Symptoms of ADHD often worsen with stress, conflict or increased demands in life.

What are common types of ADHD?
The three types of ADHD are:

• Inattentive ADHD – Inability to pay attention and distractibility. This also is known as attention-deficit disorder (ADD).
• Hyperactive and impulsive ADHD – Hyperactivity and impulsivity.
• Combined ADHD – This type causes inattention, hyperactivity and impulsivity.

ADHD can be difficult to diagnose in adults, because some of the symptoms are similar to those in other mental health conditions, such as depression or anxiety.

“Symptoms of ADHD can look different between different people,” Barterian said. “Some people might have more difficulty focusing on lectures or with organization, while others may have more social difficulties with impulsivity and trouble following along in conversations.”

Survey Methodology
This study was conducted by SSRS on its Opinion Panel Omnibus platform. The SSRS Opinion Panel Omnibus is a national, twice-per-month, probability-based survey. Data collection was conducted from August 16 – August 18, 2024, among a sample of 1,006 respondents. The survey was conducted via web (n=975) and telephone (n=31) and administered in English. The margin of error for total respondents is +/-3.8 percentage points at the 95% confidence level. All SSRS Opinion Panel Omnibus data are weighted to represent the target population of U.S. adults ages 18 or older.