Vagal activity related events and GLP-1 receptor agonists

Background
Glucagon-like peptide-1 receptor agonists (GLP-1 RA) have been associated with increased vagal activity leading to higher incidence of dizziness and nausea. However, data regarding risk of potential subsequent outcomes and clinical implications are lacking.

Purpose
To investigate risk of vagal activity related (VAR) events in patients with type 2 diabetes mellitus (T2DM) treated with GLP-1 RA compared with sodium-glucose cotransporter-2 inhibitors (SGLT-2i).

Methods
By cross referencing Danish health registers, patients with T2DM first-time initiated on GLP-1 RA or SGLT-2i were identified between January 2010 and October 2022. Patients treated for other indications than T2DM (obesity, heart failure, and chronic kidney disease) as well as patients with history of bradyarrhythmia or cardiac device implantation were excluded. Standardized 1-year absolute risks of VAR events (syncope, hip fracture, bradyarrhythmia, or cardiac device implantation) were computed to compare risk associated with GLP-1 RA or SGLT-2i initiation.

Results
During the study period, 50 076 and 73 138 patients with T2DM were initiated on GLP-1 RA (47% women and median age: 59 years [IQR: 50–68]) and SGLT-2i (39% women and median age: 64 years [IQR: 55–72]). Metformin (90% vs 89%), insulin (19% vs 14%), dipeptidyl peptidase-4 inhibitors (13% vs 17%), and sulfonylureas (9% vs 11%) were the most common concomitantly used drugs among patients first-time initiated on GLP-1 RA or SGLT-2i, respectively, and 33% vs 37% were on more than two anti-diabetics. Correspondingly, comorbidity was equally prevalent: 54% vs 59% with hypertension, 13% vs 17% with ischemic heart disease, and 6% vs 8% with stroke.

The standardized 1-year absolute risk of syncope was 0.58% (95% CI: 0.51%–0.66%) among patients initiated on GLP-1 RA and comparable to patients initiated on SGLT-2i (0.56% [95% CI: 0.50%–0.62%]) (Figure 1). The standardized 1-year risks of hip fracture, bradyarrhythmia, and cardiac device implantation were very low (<0.5%) among patients initiated on GLP-1 RA and not elevated when compared with SGLT-2i with corresponding risk ratios of 0.88 (95% CI: 0.63–1.13), 0.97 (95% CI: 0.67–1.28), and 0.86 (95% CI: 0.62–1.10) (Figure 1). Exploring the same associations in the following subgroups yielded similar results: >59 years of age, with hypertension, with ischemic heart disease, or on insulin (Table 1).

Conclusions
In a nationwide cohort of patients with T2DM, initiation of GLP-1 RA was associated with a very low and similar risk of syncope, hip fracture, bradyarrhythmia, and cardiac device implantation compared with SGLT-2i. Despite a proposed augmentation of vagal activity, GLP-1 RA use seemed safe to use—even in subgroups of suspected higher inherent risk.