12-month neurological and psychiatric outcomes of semaglutide use for type 2 diabetes: a propensity-score matched cohort study

Background: While semaglutide, approved for type-2 diabetes mellitus (T2DM), is being investigated as a treatment for brain disorders, concerns over adverse neuropsychiatric events have emerged. More data are therefore needed to assess the effects of semaglutide on brain health. This study provides robust estimates of the risk of neurological and psychiatric outcomes following semaglutide use compared to three other antidiabetic medications. Methods: This retrospective cohort study used electronic health records from TriNetX US Collaborative Network, covering >100 million patients in the USA. Due to the exploratory nature of this study, we did not use a pre-registered protocol or statistical analysis plan. Three cohorts with T2DM prescribed semaglutide between 1st December 2017 and 31st May 2021 were propensity-score matched (1:1 using a greedy nearest-neighbour algorithm with calliper distance of 0.1) with cohorts receiving sitagliptin, empagliflozin, and glipizide. Using Cox regression analysis, we compared the risks of 22 neurological and psychiatric outcomes within one year since the index prescription: encephalitis, parkinsonism, cognitive deficit, dementia, epilepsy/seizure, migraine, insomnia, nerve disorder, myoneural junction/muscle disease, intracranial haemorrhage, ischaemic stroke, alcohol misuse, opioid misuse, cannabis misuse, stimulants misuse, nicotine misuse, psychosis, bipolar disorder, depression, anxiety, obsessive-compulsive disorder, and suicidality. Negative control outcomes (NCOs) were used to assess unmeasured confounding. Findings: Each matched cohort included 23,386 (semaglutide vs sitagliptin), 22,584 (vs empagliflozin), and 19,206 (vs glipizide) patients. Semaglutide was not associated with an increased risk of neurological and psychiatric outcomes. Instead, after multiple-testing correction, semaglutide was associated with reduced risk for several such outcomes, notably cognitive deficit compared to sitagliptin (HR 0.72, 95% CI 0.64–0.80) and glipizide (HR 0.72, 95% CI 0.63–0.81), dementia compared to sitagliptin (HR 0.52, 95% CI 0.40–0.68), and nicotine misuse across most comparisons (HR 0.72, 95% CI 0.61–0.85 against glipizide; HR 0.77, 95% CI 0.65–0.90 against empagliflozin; HR 0.82, 95% CI 0.70–0.95 against sitagliptin, though the latter was no longer statistically significant after adjustment for multiple comparisons). Empagliflozin showed fewest differences from semaglutide. No differences in NCOs were observed between cohorts. Interpretation: Semaglutide is not associated with higher 12-month risk of adverse neuropsychiatric outcomes compared to other antidiabetic medications. Potential beneficial associations with some outcomes, especially cognitive deficit and nicotine misuse, should stimulate validation in clinical trials. Funding: National Institute for Health Research (NIHR) Oxford Health Biomedical Research Centre, Medical Research Council.