Neuropsychiatric and behavioural changes associated with GLP-1 receptor agonist therapy: a systematic review of case-based evidence

Background: GLP-1 receptor agonists (GLP-1RAs) have been widely used to treat diabetes and obesity. With more off-label use and evidence of their roles in reward processing and motivation, the risk and benefit associated with neuropsychiatric and behavioural outcomes remained uncertain. Most systematic reviews and meta-analyses which support their neuropsychiatric safety exclude case-based reports which may actually better capture rare and idiosyncratic side effects in vulnerable patients.

Objectives: To systematically synthesise case-based evidence from case reports and case series on neuropsychiatric and behavioural changes with concurrent GLP-1RAs use.

Methods: A systematic review of case reports and case series published in MEDLINE, Embase, PubMed, PsycINFO, and Google Scholar up to January 2026 was conducted according to the CABARET protocol with pre-registration on PROSPERO (CRD420261277870). Two independent reviewers screened titles, abstracts, full-texts, and extracted data using standardised forms on Covidence. Study quality was assessed using Joanna Briggs Institute (JBI) checklist. Causalities were evaluated using the Naranjo Adverse Drug Reaction Scale for adverse neuropsychiatric events and Modified Naranjo Criteria (MONARCH) for therapeutic outcomes.

Results: 834 studies were included. After screening, 20 case reports and case series with the total of 29 patients were included in the analysis. GLP-1RAs demonstrated a heterogeneous and potentially bidirectional neuropsychiatric profile. Therapeutic effects beyond metabolic benefits included reduction in binge eating, food preoccupation, compulsive eating, food-triggered aggression, and genitopelvic dysesthesia. They also improved impulse control and satiety. These therapeutic effects are supported by objective measurements such as Binge Eating Scale, intracranial electroencephalogram (iEEG) and rechallenge patterns. Additionally, the MONARCH reflects definite and probable causality regarding these outcomes. Conversely, adverse outcomes such as new-onset depression, apathy, anhedonia, and suicidal behaviours were found, especially within the first 4 weeks of the GLP-1RAs initiation. These symptoms resolved after GLP-1RAs discontinuation and were classified as probable causality in all cases. A temporal pattern of initial neuropsychiatric symptoms improvement followed by relapse or breakthrough symptoms were also observed. Results in substance use disorder cases were mixed.

Conclusions: This review reveals the heterogeneity and potentially bidirectional effect of GLP-1RAs offering both neuropsychiatric therapeutic effects and adverse outcomes. Beneficial effects were most consistent with dysregulated reward processing such as the stabilization of 9 cases from dysregulated eating pattern, 2 cases of substance use disorder, and 1 case with holoprosencephaly. For adverse effects: despite multiple studies claiming its safety, new onset-depression, apathy, anhedonia, were reported in 4 cases, and suicidal behaviours in 1 case. These symptoms resolved upon drug discontinuation. The temporal pattern of relapse or breakthrough of symptoms, particularly in food related behavior, suggests neuroadaptive changes or tolerance within reward circuits over time. These observations support the possibility that GLP-1RAs alter reward circuitry leading to either beneficial outcomes or adverse effects such as diminished motivation. The latter may be more pronounced in vulnerable population with prior psychiatric illnesses or other neurodegenerative diseases. These findings highlight the need for risk assessment, clinical vigilance, and individualised monitoring, as well as prospective controlled studies in heterogeneous population to identify neuropsychiatric outcomes in vulnerable population.

References