School of Pharmacy, Sungkyunkwan University, Suwon, South Korea, Republic of Korea
Background: Obstructive sleep apnea (OSA) is a common yet underdiagnosed complication of type 2 diabetes (T2D), contributing to increased cardiometabolic risks. Sodium-glucose cotransporter-2 inhibitors (SGLT2i) induce diuresis and natriuresis, potentially mitigating OSA risk by reducing fluid retention and improving airway patency. However, real-world evidence on this potential protective effect remains limited.
Objectives: To evaluate the risk of OSA among patients with T2D initiating SGLT2i versus glucagon-like peptide-1 receptor agonists (GLP-1RA).
Methods: Using a National Health Insurance Service (NHIS) claims database (2014-2022) of Korea, we constructed an active-comparator, new-user cohort of patients with T2D. The cohort compared initiators of SGLT2i to GLP-1RA, with the index date defined as the date of drug initiation. We defined a composite primary outcome comprising incident diagnosis of OSA following polysomnography, and OSA-related surgical procedures, identified through NHIS database. Patients were followed from the day after index date until the earliest of outcome occurrence, drug discontinuation (90-day grace period) or switch, death, or study end date (December 31, 2022). We applied propensity score fine stratification to achieve comparability of treatment groups. Incidence rates (IRs) per 1,000 person-years were estimated based on Poisson distribution. Hazard ratios (HR) with 95% confidence intervals (CI) were estimated by weighted Cox models. We repeated the main analysis among prespecified subgroup patients with a diagnosis of hypertension, heart failure, asthma/chronic obstructive pulmonary diseases (COPD), or with obesity defined as body mass index > 25 kg/m2, to assess the role of fluid redistribution in mediating the observed associations.
Results: We identified 742,074 and 45,616 new users of SGLT2i (58.7% dapagliflozin; 36.5% empagliflozin) and GLP-1RA (82.3% dulaglutide), respectively, with a mean age of 57.3 years and 60.2% males. Over a mean follow-up of 1.8 years, SGLT2i presented a 56% lowered risk of OSA compared to GLP-1RA (weighted IR 2.3 vs. 5.6 per 1,000 person-years; weighted HR 0.44, 95% CI 0.28-0.70). Results remained consistent in patients with obesity (weighted HR 0.44, 95% CI 0.28-0.70) and heart failure (0.18, 0.04-0.80). A modest attenuation of effect was observed in patients with hypertension (HR: 0.43, 95% CI 0.17–1.10), while no significant association was found in those with asthma/COPD (HR: 1.06, 95% CI 0.32–3.47).
Conclusions: Findings indicated that the use of SGLT2i was associated with a lowered risk of OSA compared to GLP-1RA among patients with T2D. Further studies are warranted to identify possible mechanisms for the non-significant associations in some patient subgroups.
