Association of Sodium-Glucose Cotransporter 2 Inhibitor Use With Risk of Osteoporotic Fracture Among Older Women: A Nationwide, Population-Based Cohort Study

초록

Sodium-glucose cotransporter-2 inhibitor (SGLT2i) is a widely recommended anti-diabetic medication (ADM) due to its cardioprotective and renoprotective benefits. However, the effects of SGLT2i on fracture risk have been debated. The Canagliflozin Cardiovascular Assessment Study showed that SGLT2i increased hip fracture risk during the mean 6-year follow-up period. However, in most other large randomized controlled trials of SGLT2i, SGLT2i use was not associated with increased fracture risk. Although the fracture risk associated with SGTL2i use is controversial, there are concerns regarding fracture risk due to the increased risk of falls caused by volume depletion or disturbance of calcium phosphate balance. Therefore, in the present study, the association between SGLT2i use and fracture risk in elderly women (>= 65 years of age) who were newly prescribed ADMs was investigated using a national cohort database.?We used the data for this population-based cohort study obtained from the National Health Insurance Service of Korea between January 1, 2013, and December 31, 2020. Women older than 65 years old diagnosed with type 2 diabetes mellitus who were newly prescribed ADMs other than glucagon-like peptide-1 receptor agonists and thiazolidinedione and who had available comprehensive health check-up data were included.A total of 1,333 SGLT2i users was analyzed and compared with covariate (e.g., body mass index (BMI), age-matched non-SGLT2i users at a 1:2 ratio (n = 2,626). After propensity score matching, mean age (70.3 years vs. 70.3 years, p = 0.830), BMI (27.4 kg/m2 vs. 27.4 kg/m2, p = 0.924), number of oral ADMs (1.4 vs. 1.3, p = 0.092), and other covariates were well-balanced between SGLT2i users and non-SGLT2i users. During the follow-up period, SGLT2i use was associated with a significantly increased risk of vertebral fracture compared with non-SGLT2i use (incidence rate 19.2 vs. 13.8 per 1,000 person-years; hazard ratio 1.40, 95% confidence interval 1.00-1.96, p = 0.049). Regarding the risk of other types of fracture, any fractures, hip fractures, and non-hip or non-vertebral fractures, significant differences were not observed between SGLT2i users and non-SGLT2i users.In conclusion, SGLT2i use showed a 40% higher risk of vertebral fracture than non-SGLT2i use in elderly women. Although further validation is required, SGLT2i should be cautiously prescribed in older women due to the potential association with fracture risk in this vulnerable population.