GLP-1 Agonist Semaglutide Offers Greater Neuropsychiatric Benefits Than Other Antidiabetics, Study Suggests
The study highlighted semaglutide's lower risk of adverse neurological outcomes, making it a promising candidate for future neuroprotective therapies, including ongoing trials for Alzheimer disease.
Riccardo De Giorgi, MD, DPhil, MRCPsych
A recently published retrospective cohort study using electronic health records from the TriNetX US Collaborative Network showed that semaglutide (Novo Nordisk), an FDA-approved therapy for type 2 diabetes (T2D) and obesity, was not associated with higher 12-month risk of adverse neuropsychiatric outcomes relative to other antidiabetic medications. In fact, treatment with the therapy resulted in attenuated risk for several neurological and psychiatric outcomes, including cognitive deficit and nicotine misuse.
In the study, 3 cohorts of patients with T2D prescribed semaglutide from December 2017 to May 2021 were propensity-score matched with cohorts receiving antidiabetic medications sitagliptin, empagliflozin, and glipizide. Following application of inclusion/exclusion criteria, a total of 23,386 patients were included in each cohort (after matching) for the semaglutide vs sitagliptin comparison (mean age, 56.6 [±12.8] years), 22,584 patients in each cohort for the empaliflozin comparison (mean age, 57.6 [±12.4] years), and 19,206 patients in each cohort for the glipizide comparison (mean age, 56.3 [±13.0] years).
Led by Riccardo De Giorgi, MD, DPhil, MRCPsych, a clinical lecturer in the department of psychiatry at the University of Oxford, the study used a Cox regression analysis to compare the risks of 22 neurological and psychiatric outcomes within 1 year of starting treatment with any of the antidiabetic medications included. To assess for possible unmeasured confounding, the study authors also estimated the risk of 15 negative control outcomes.
Overall, treatment with semaglutide did not result in increased risk for neuropsychiatric outcomes. This is critical to the care of patients with T2D, who have a high burden of mental illness. In addition, the authors noted that the absence of negative associations between semaglutide use and depression or suicidality can help inform current investigations by the European Medicines Agency, MHRA, and FDA, although ongoing pharmacovigilance as well as studies in other at-risk subgroups were recommended.
The study showed that the risk of cognitive deficits was significantly lower after semaglutide compared with sitagliptin (HR, 0.72; 95% CI 0.64–0.80, Padj <0.0001) and glipizide (HR, 0.72; 95% CI 0.63–0.81, Padj <0.0001), but similar between semaglutide and empagliflozin (HR, 0.96; 95% CI 0.86–1.08, Padj = 0.51). Similarly, the risk of dementia was lower after semaglutide compared to sitagliptin (HR 0.52; 95% CI 0.40–0.68, Padj <0.0001) and glipizide (HR 0.63; 95% CI 0.46–0.86, although this did not reach significance after correction, Padj 0.075), but similar between semaglutide and empagliflozin (HR 0.91; 95% CI 0.69–1.21, Padj 0.53).
In the discussion, investigators concluded that the underlying mechanisms for why glucan-like peptide-1 (GLP-1) analogues like semaglutide positively impact cognition. “One possibility is that any protective effect of GLP1-RAs on cognition is mediated by their effect on cardiovascular morbidity," De Giorgi et al wrote.1 "However, the lack of consistent association with ischaemic strokes suggests that other mechanisms might also be at play. These might include other neuroprotective and anti-inflammatory mechanisms, supported by measurements of systemic inflammation in people exposed to semaglutide."
READ MORE: Exploratory PEGASUS Data Reveal AMX0035’s Impact on Neurodegeneration, Tau
Additional data from the study showed that the risk of nicotine dependence was significantly lower after semaglutide compared with glipizide (HR, 0.72; 95% CI 0.61–0.85, Padj = 0.0027) and empagliflozin (HR 0.77; 95% CI 0.65–0.90, Padj = 0.024), and lower compared to sitagliptin, although not after correction for multiple testing (HR; 0.82, 95% CI 0.70–0.95, Padj = 0.23). Semaglutide was also associated with a lower risk of first diagnosis of depression and ischemic stroke in comparison with sitagliptin. Furthermore, the risk of all-cause mortality was lower after semaglutide compared with sitagliptin (HR, 0.58; 95% CI 0.50–0.66, P < 0.0001), glipizide (HR, 0.55; 95% CI 0.47–0.64, P < 0.0001), and empagliflozin (HR, 0.86; 95% CI 0.75–0.99, P = 0.035).
Regarding the lowered risk of nicotine use disorder with semaglutide, the study authors concluded that, "This might reflect that people prescribed semaglutide are fundamentally different in their risk of nicotine misuse that was not captured in the propensity-score matching. Alternatively, it is possible that, unlike other antidiabetic agents, GLP1-RAs regulate dopaminergic pathways underlying reward sensitivity, which are at least partly responsible for their weight-loss action and their putative activity against addictive behavior."
De Giorgi et al continued, "The same reasoning could apply to other substance use disorders (e.g., stimulant misuse), for which however we only observed trend associations that did not hold against adjustment for multiple comparisons. Further research is therefore needed to investigate the potential role of semaglutide in the treatment of addictions."
There was one exception to the findings in that there was an increased risk of any diagnosis for migraine for semaglutide compared with glipizide (HR, 1.20; 95% CI, 1.08-1.33; Padj = 0.015).
Novo Nordisk is currently funding 2 international phase 3 studies to assess semaglutide in patients with early Alzheimer disease. Dubbed evoke (NCT04777396) and evoke+ (NCT04777409), each trial is expected to include 1840 amyloid-positive participants (aged 55-85 years) with mild cognitive impairment due to AD or mild AD dementia. Participants will be randomized 1:1 to either oral semaglutide 14 mg daily, escalated via 3- and 7-mg doses over 8 weeks, or placebo, for 156 weeks. Both trials, which use change in Clinical Dementia Rating-Sum of Boxes score as the primary end point, are expected to have data read out in 2025.2
