A large-scale observational study found no evidence of a link between semaglutide use and increased neurological or psychiatric risks in patients with type 2 diabetes mellitus (T2DM). The study compared neuropsychiatric outcomes between semaglutide users and those prescribed other common antidiabetic medications.
In the retrospective cohort study, published in eClinicalMedicine, investigators examined the electronic health records of 65,176 patients with T2DM from the TriNetX US Collaborative Network. The study—spanning from December 1, 2017, to May 31, 2021—employed propensity-score matching to compare 12-month neuropsychiatric outcomes among patients prescribed semaglutide versus those prescribed sitagliptin (n = 23,386 per group), empagliflozin (n = 22,584 per group), or glipizide (n = 19,206 per group).
Key findings included:
- There was no increased risk for 22 neurological and psychiatric outcomes with semaglutide use.
- There were lower risks of cognitive deficits with semaglutide compared to sitagliptin (hazard ratio [HR], 0.72; 95% confidence interval [CI], 0.64-0.80) and glipizide (HR, 0.72; 95% CI, 0.63-0.81).
- There was a reduced risk of dementia with semaglutide compared to sitagliptin (HR, 0.52; 95% CI, 0.40-0.68).
- There was a reduced risk of nicotine misuse with semaglutide across most comparisons of other antidiabetic medications.
- There was no increased risk of depression or suicidality among patients receiving semaglutide.
- Lower all-cause mortality was observed with semaglutide compared to all three comparators (sitagliptin, empagliflozin, and glipizide), though the authors cautioned about incomplete linkage between the database and death registries.
Secondary analyses revealed consistent results when stratified by age groups and pre/post–COVID-19 pandemic periods, as well as similar trends at 2-year follow-up, with an additional potential association of lower psychosis risk for semaglutide compared to sitagliptin and glipizide.
The study's strengths included its large sample size and extensive propensity-score matching. The absence of differences in negative control outcomes between cohorts further strengthened the validity of the findings. However, limitations such as potential coding errors and unmeasured confounding factors were noted.
The authors discussed the possible neuroprotective and anti-inflammatory mechanisms of GLP-1 receptor agonists, as well as effects on dopaminergic pathways, as potential explanations for the observed outcomes.
The authors noted that these findings may be relevant to ongoing regulatory investigations. They also suggested that further clinical trials could explore semaglutide's effects on cognitive function and substance use disorders in patients with T2DM.
Declaration of interests can be found in the study.