Recent large registry studies report associations between GLP-1 RA use and reduced substance-related harms. We synthesize this evidence, identify critical knowledge gaps, and define research priorities needed before clinical translation.

A targeted evidence synthesis was conducted across three domains:

1. Large registry studies examining outcomes in people with AUD or opioid use disorder (OUD) prescribed GLP-1 RAs.

2. Real-world cohort studies comparing overdose and hospitalisation rates between GLP-1 RA users and non-users.

3. Neuroscience literature describing GLP-1 receptor activity in reward-related brain regions.

Multiple observational studies show associations between GLP-1 RA use and reduced substance-related harms.

In a Swedish cohort of 227,866 individuals with AUD, semaglutide was associated with reduced alcohol-related hospitalisations (Lähteenvuo 2024). US data (>500,000 OUD, >800,000 AUD patients) showed lower opioid overdose and alcohol intoxication rates among GLP-1 RA users (Qeadan 2024; Wang 2024).

All current human findings are observational, but several human randomised controlled trials are now underway, including trials of semaglutide and exenatide for AUD, semaglutide for cocaine use disorder, and liraglutide for nicotine dependence.

The overlap between metabolic and addiction pathways may explain these associations, as GLP-1 signaling modulates reward processing relevant to both food intake and substance use. Preclinical studies demonstrate GLP-1R-mediated reduction in substance self-administration via reward pathway modulation in VTA, NAcc, and PFC. While this provides biological plausibility, observational human data cannot establish causality, and RCT evidence is needed.

Observational data suggest potential associations between GLP-1 RA use and reduced substance-related harms, but cannot establish causality due to confounding. Before clinical application, essential research includes: (1) completion of ongoing RCTs with addiction-specific outcomes, (2) safety and tolerability evaluation in actively substance-using populations, including assessment of nausea/vomiting risks and drug interactions with methadone, buprenorphine, and benzodiazepines, (3) adherence and implementation feasibility studies in populations with chaotic substance use, and (4) health economics modeling. Psychiatrists should monitor emerging trial results critically as this evidence base develops in the coming years.

Inattention and hyperactivity/impulsivity are frequently associated with intellectual disability (ID), contributing to significant behavioural dysregulation and functional impairment. Protein kinase C (PKC) has emerged as a key molecular target in the neurobiology of impulsivity, with PKC overactivity disrupting dopaminergic regulation and prefrontal cortical functioning, leading to impaired inhibition and executive dysfunction. Endoxifen, an active metabolite of Tamoxifen, is a selective PKC inhibitor that has demonstrated antimanic, anti-impulsive, and emotion-stabilising effects in adults. Emerging clinical evidence suggests benefit in behavioural dysregulation across several psychiatric conditions; however, paediatric use remains limited and largely off-label. This case report aims to evaluate the efficacy and tolerability of endoxifen for behavioural symptoms in an adolescent with ID who showed inadequate response to standard treatments.

A 14-year-old male with global developmental delay and significant speech–language impairment presented with repetitive verbal behaviours, hyperactivity, poor sitting tolerance, and frequent anger outbursts for two years. He was provisionally diagnosed with ID with behavioural impairment. Baseline BMI was 21 kg/m², and comorbid hypothyroidism was managed with replacement therapy.

Previous trials of risperidone (4 mg/day) with divalproex sodium (1500 mg/day) resulted in minimal improvement and adverse effects, including nausea, increased appetite, and weight gain. Subsequent treatment with oxcarbazepine (titrated to 1350 mg/day) and aripiprazole (10 mg/day) was complicated by hyponatremia, necessitating discontinuation of oxcarbazepine.

Following informed consent and assent, endoxifen was initiated at 8 mg/day and gradually titrated to 24 mg/day, while aripiprazole was continued at 5 mg/day. Clinical response and tolerability were monitored over 12 months.

Over one year of treatment, the patient demonstrated sustained reductions in impulsivity, hyperactivity, repetitive behaviours, and anger outbursts, with notable improvements in behavioural regulation and overall family functioning. The medication was well tolerated, with no significant metabolic, neurological, or systemic adverse effects, and adherence remained satisfactory throughout follow-up.

Endoxifen showed favourable efficacy and tolerability in managing behavioural dysregulation in an adolescent with ID who was refractory or intolerant to conventional psychotropics. Targeting PKC-mediated pathways may represent a promising therapeutic strategy for impulsivity and aggression in neurodevelopmental disorders. Larger controlled studies are needed to establish safety and long-term efficacy.

In a community mental health setting, we studied real-world relapse rates in patients with a history of psychosis, before and after switching from oral antipsychotics to a long-acting injectable antipsychotic (LAIA). We also aimed to evaluate the relevance of metabolic/demographic factors in relapse rate.

Twenty-five individuals with a past history of psychosis and variable medication concordance who were switched from oral antipsychotics to LAIAs were included in the study. Oral agents were converted to olanzapine equivalents whilst LAIA generics were converted to flupenthixol depot equivalents. The depot equivalent oral dose of olanzapine was calculated, assuming an oral bioavailability of 40%. Relapse-rates occurring over the time periods on oral therapy were compared with rates occurring following initiation of LAIAs. Relapse was defined as any psychiatric hospital admission.

A multivariable Poisson regression model including drug formulation/age/sex/smoking status/BMI were fitted with follow-up time as an offset to estimate adjusted incidence rate ratios and associations with relapse outcomes.

The mean±SD age of the 25 individuals was 50.6±13.9 years whilst 12 (60%) were male. Mean BMI of males/females was 28.1±9.3 kg/m2/ 32±6.4 kg/m2(higher in women) respectively. 65% of patients received either flupenthixol or zuclopenthixol depot with next most common depot being Risperidal Consta. The median (IQR) of daily oral olanzapine equivalent was 10mg (6.9 – 20.0mg)/day and after switching to depot was 16.0mg (5.4 – 40.7mg)/day. The 4-week depot equivalent dose for services users switched to depot was flupenthixol 480mg (200 – 1220mg) / 4-weeks.

Following adjustment for age/sex/smoking status/BMI, relapse rates were significantly lower during LAIA treatment compared with oral treatment (incidence rate ratio [IRR] 0.13, 95% CI 0.08–0.22, p<0.001). Thus depot antipsychotics were associated with an 87% reduction in relapse rates compared to oral therapy. Male sex/younger age/lower BMI were independently associated with higher relapse rates (Exp (beta) 1.9/0.98/0.94 respectively (p<0.05)).

In this real-world community mental health setting, this study highlights that depot antipsychotics provide a dramatic and sustained reduction of 87% in relapse rate for individuals with a history of psychosis. These findings support early prescription of LAIAs in prevention of psychosis relapse. Further evaluation is warranted to elucidate the potential of LAIAs as a treatment option after first presentation of psychosis.

Et al:

Ruth Parkman-Eason, Salford Royal Hospital, UK

Sophie Manttan, Salford Royal Hospital, UK

Yasitha llangasekera, University of Peradeniya, Sri Lanka

Various antipsychotics and antidepressants have pharmacogenomic guidelines available from international sources, recommending dose adjustments and/or alternative drug selection based on a patient’s genetic profile. However, pharmacogenomic implementation remains limited in UK healthcare.

Genetics and Environment in Mental Health Study (GEMS) is a prospective study of pharmacogenomic testing in psychosis; the first in the UK. Adults taking an antipsychotic medication have been recruited across England. Participants provided a DNA sample for genotyping on a multi-gene panel, selected from evidence-based pharmacogenomic guidelines. Pharmacogenomic reports, displaying the participants’ genetic profile and any alterations to usual prescribing guidelines were delivered to the prescribing clinician. Any changes to the participants’ prescription were at the discretion of the clinician.

A diverse sample of 584 people taking an antipsychotic medication has been recruited so far, consisting of participants aged 18-82 years, with an even split of male and female participants, and 32.5% of the sample from Black, Asian, and Minority Ethnic (BAME) backgrounds.

Across the four genes with pharmacogenomic guidelines in psychiatry (CYP2D6, CYP2C19, CYP2B6, and CYP3A4), 90.4% of the sample carried at least one actionable variant. At the time of recruitment, 20.9% of participants carried a pharmacogenomic variant that (based on current guidelines) was actionable for the antipsychotic and/or antidepressant they were currently prescribed. By medication class, 6.8% were taking an antipsychotic and 15.1% were taking an antidepressant for which they had a pharmacogenomic recommendation. A further 69.5% of the sample would have a recommendation for at least one antipsychotic or antidepressant medication.

Based on the estimated population prevalence of pharmacogenomic actionable variants and prescription data in England, approximately 45% of people would be expected to carry an actionable variant for antipsychotic medications and at least 61% would be expected to carry an actionable variant for antidepressants. This equates to approximately 285,000 people currently taking an antipsychotic medication that could have a pharmacogenomic recommendation and at least 4.2 million people taking an antidepressant with a pharmacogenomic recommendation in England.

Pharmacogenomics has the potential to personalise medication prescriptions and improve patient outcomes. Workforce training and implementation barriers present a challenge. However, in line with the NHS 10 Year Plan goal of 50% of healthcare interventions being genomics-informed by 2035, psychiatry is one of the most promising areas for the implementation of pharmacogenomic-guided prescribing.