SGLT2 inhibitors were associated with a lower or similar fracture Risk compared with incretin-based antidiabetic agents in postmenopausal women with type 2 diabetes, according to a claims-based study involving two independent Korean cohorts.

In the first, new users of an SGTL2 inhibitor had a lower risk for incident fractures compared with those started on a DPP-4 inhibitor (weighted HR 0.78, 95% CI 0.72-0.84), found Ju-Young Shin, PhD, of Sungkyunkwan University in South Korea, and colleagues.

And in the second cohort, the fracture risk with SGLT2 inhibitors versus GLP-1 receptor agonists was no different (weighted HR 0.92, 95% CI 0.68-1.24), the researchers reported in JAMA Network Open.

Given the higher fracture risk in this patient population, the findings should provide reassurance and help with clinical decision-making, Shin’s group wrote.

“Postmenopausal individuals are at high risk of fractures due to declining estrogen levels that disrupt the homeostasis of bone metabolism,” they explained. “If these individuals also have type 2 diabetes, risk of fractures could be further heightened given that type 2 diabetes itself is an independent risk factor for fracture.”

Fracture risk with SGLT2 inhibitors has been a lingering concern after a signal was seen in the CANVAS program testing canagliflozin (Invokana) — the first of these agents approved by the FDA in 2013. In this clinical program, there was a 26% higher risk for all fractures with canagliflozin versus placebo.

Since then, a few studies have looked into the potentially higher fracture rate with SGLT2 inhibiting agents, all consistently deeming them safe, including studies comparing the class to sulfonylureas, as well as DPP-4 inhibitors and GLP-1 receptor agonists. The current study was unique in testing this concern specifically in postmenopausal women.

“Biologically, [SGTL2 inhibition] may potentially disrupt calcium and phosphate homeostasis in serum by increasing proximal tubular reabsorption of phosphate, which might have detrimental effects on bone and lead to skeletal fragility,” the researchers explained.

In subgroup analysis comparing users of SGLT2 and DPP-4 inhibitors, a few factors appeared to modify the fracture risk. First, age played a role, as only women 61 and older saw a lower fracture risk with SGLT2 inhibitors. Postmenopausal women with osteoarthritis or osteoporosis in the SGLT2-inhibitor group had an even lower risk for fracture, as did women on other concomitant medications, including thiazolidinediones, proton-pump inhibitors, and selective serotonin reuptake inhibitors.

For the SGLT2 inhibitor versus GLP-1 receptor agonist comparison, a trend toward a higher risk of fractures was seen in the SGLT2-inhibitor group among women with osteoarthritis, a finding that warrants further attention, according to the researchers.

Data for the study came from the National Health Insurance Service, a single provider for health insurance in South Korea.

The first cohort included 37,532 new users of SGLT2 inhibitors and 332,004 new users of DPP-4 inhibitors. Mean follow-up was 1.45 years and 2.09 years, respectively. Fracture incidence reached 1.41 events per 100 person-years in the SGLT2-inhibitor group and 1.81 per 100 person-years for the DPP-4-inhibitor group.

Looking at specific agents in the SGLT2-inhibitor class, patients on dapagliflozin (Farxiga), empagliflozin (Jardiance), or ertugliflozin (Steglatro) all had a lower fracture risk compared with patients started on a DPP-4 inhibitor, which included alogliptin (Nesina), linagliptin (Tradjenta), saxagliptin (Onglyza), and sitagliptin (Januvia). The SGLT2 inhibitor canagliflozin was not included in this current analysis.

In the second cohort, Shin’s team compared 111,835 new users of SGLT2 inhibitors with 8,181 new users of a GLP-1 receptor agonist over mean follow-ups of 1.43 and 0.82 years, respectively. In this dataset, the incident fracture rate reached 1.67 per 100 person-years among patients prescribed SGLT2 inhibitors and 1.92 per 100 person-years for those on GLP-1 receptor agonists. GLP-1 receptor agonists included dulaglutide (Trulicity), exenatide (Byetta, Bydureon), and lixisenatide (Adlyxin).

Multivariable analysis adjusted for several risk factors for fracture, including diabetes severity, though researchers didn’t rule out the possibility of unmeasured covariates. Other limitations included the low number of patients using GLP-1 agonists and the use of age (45 and older) as a proxy for postmenopausal status.