SGLT2 and GLP-1 on arrhythmias: 32 trials

A network meta-analysis published in Diabetes, Obesity and Metabolism on 9 July 2026 pooled 32 randomised controlled trials and 140,156 participants to ask a specific question: do SGLT2 inhibitors and GLP-1 receptor agonists affect the risk of ventricular arrhythmias (VA) and other cardiovascular events, and does the answer change depending on whether the underlying condition is type 2 diabetes or heart failure^[1]?

The study, led by Wen-Ting Sun and colleagues, is notable for its disease-stratified design. Instead of collapsing all populations into one network, the authors ran separate analyses for the type 2 diabetes (T2DM) and heart failure (HF) settings. That matters because the two populations carry different baseline arrhythmia risk and respond differently to the same molecule^[1].

What the analysis found on ventricular arrhythmias

Across most drugs in the network, there was no statistically significant change in VA risk. The exception was empagliflozin in the T2DM network, which showed a statistically significant but explicitly exploratory signal: OR 0.31 (95% CI 0.11 to 0.86)^[1]. In plain terms, empagliflozin was associated with roughly a 69% lower odds of a ventricular arrhythmia event compared with control in the T2DM trials included in the network. The authors categorised this as exploratory because it was not based on adjudicated arrhythmia endpoints across all included trials, and the confidence interval is wide.

The signal is biologically plausible. Empagliflozin reduces cardiac preload and afterload through osmotic diuresis and natriuresis, lowers sympathetic nervous system activation, and has been shown in experimental models to stabilise cardiomyocyte membrane potential. Whether any of those mechanisms translates into a clinically meaningful arrhythmia benefit in humans requires prospective trials with adjudicated arrhythmia endpoints, which this meta-analysis cannot substitute for.

Heart failure and mortality findings

The cardiovascular mortality findings are more consistent with prior large outcome trials. Dapagliflozin in the HF network and empagliflozin and liraglutide in the T2DM network were each associated with reduced cardiovascular and all-cause mortality^[1]. SGLT2 inhibitors consistently reduced hospitalisation for heart failure (HHF) risk across both the HF and T2DM networks, which aligns with what EMPA-REG OUTCOME^[2] and DAPA-HF^[3] established in their respective primary readouts.

For the GLP-1 class, liraglutide's mortality signal in this network is consistent with the LEADER trial track record in T2DM. The network meta-analysis does not directly address semaglutide, whose SELECT trial in 17,604 adults with obesity and pre-existing cardiovascular disease reported a 20% relative reduction in the primary MACE composite (HR 0.80, 95% CI 0.72 to 0.90)^[4]. SELECT enrolled people without diabetes, which falls outside the T2DM network in Sun et al., so the two analyses are complementary rather than redundant.

Safety signals

On safety, dapagliflozin was linked to a lower risk of acute kidney injury compared with control^[1]. That finding fits the broader nephroprotective evidence base for SGLT2 inhibitors and is consistent with DAPA-CKD and CREDENCE trial data. For GLP-1 agents, albiglutide and liraglutide were associated with lower hypoglycemia risk relative to comparators in the network, likely reflecting their glucose-dependent insulin-stimulation mechanism.

What the study cannot tell us

The authors are explicit about the exploratory nature of the ventricular arrhythmia findings. Three limitations stand out. First, not all trials used adjudicated VA endpoints, so the pooled OR for empagliflozin reflects heterogeneous ascertainment methods across contributing trials. Second, the network is sparse for some pairwise comparisons, meaning indirect evidence dominates certain estimates. Third, the two networks (T2DM and HF) carry different baseline risk profiles, and the disease-stratified design, while appropriate, still cannot fully account for variation in background therapy and trial era.

The finding that most GLP-1 agents showed no significant VA signal is also not the same as a finding of safety. VA events are rare and trials are powered for composite endpoints, not arrhythmia sub-endpoints. Absence of a signal in this network does not rule out a meaningful effect in either direction.

What this means for the broader GLP-1 cardiovascular picture

For clinicians and readers following the incretin class, this meta-analysis adds one more data point to a growing cardiovascular evidence base. The SGLT2 and GLP-1 classes now have overlapping cardiovascular outcome trial data across T2DM, heart failure, and (for semaglutide) obesity without diabetes. The question of arrhythmia specifically remains open. The empagliflozin VA signal in this analysis is hypothesis-generating, not practice-changing. The authors call for prospective trials with pre-specified arrhythmia endpoints before drawing clinical conclusions. That is the appropriate caution. For current prescribing decisions, the established MACE and HHF benefits of SGLT2 inhibitors in high-risk patients remain the load-bearing evidence, not a single exploratory OR from a network meta-analysis^[1].

This article is educational. It does not constitute medical advice. If you have questions about cardiovascular medications, speak with a qualified clinician.