Initiation of sodium–glucose cotransporter‐2 (SGLT2) inhibitor therapy upon hospital discharge is associated with improved clinical outcomes among patients with acute heart failure (HF), especially those with frailty, according to the West Tokyo Heart Failure 2 (WET‐HF 2) registry in Japan.
Over a median follow‐up of 320 days after discharge, the primary outcome of a composite of cardiac death or HF rehospitalization within 1 year occurred less frequently among patients who received an SGLT2 inhibitor prescription at discharge than among those who did not (16 percent vs 19 percent; p=0.009). [J Am Heart Assoc 2026;doi:10.1161/JAHA.125.048748]
The protective effect of SGLT2 inhibitors was pronounced for patients with frailty (Clinical Frailty Scale [CFS] ≥4; 18 percent vs 23 percent; p=0.022) but disappeared for those without (CFS ≤3: 14 percent vs 13 percent; p=0.52).
In an analysis adjusted for baseline characteristics using propensity score–based inverse probability of treatment weighting, SGLT2 inhibitor use was associated with a 24-percent lower risk of the primary outcome (hazard ratio [HR], 0.76, 95 percent confidence interval [CI], 0.57–0.99; p=0.045) in the overall population and a 41-percent risk reduction in the subgroup of patients with frailty (HR, 0.59, 95 percent CI, 0.40–0.85; p=0.005). The association was null in the subgroup of those without frailty (HR, 1.18, 95 percent CI, 0.81–1.73; p=0.40).
Similar trends were observed for each component of the primary outcome.
No evidence of harm
“SGLT2 inhibitors have been shown to reduce body weight and fat mass, which is beneficial in diabetes management but may raise concerns regarding potential skeletal muscle mass and worsening sarcopenia or frailty,” the investigators noted.
They pointed out that in the WET‐HF 2 cohort, SGLT2 inhibitor use did not result in a decline in BMI and Geriatric Nutritional Risk Index (GNRI).
“Both BMI and GNRI increased significantly from discharge to 1 year in both the SGLT2 inhibitor user and nonuser groups, with no significant between-group differences. This trend was consistent among patients with CFS ≤3 or CFS ≥4,” they said.
Additionally, SGLT2 inhibitor use was associated with more favourable changes in estimated glomerular filtration rate, irrespective of frailty status.
“Our data suggest that treatment with SGLT2 inhibitors may be both safe and effective in patients with HF and frailty,” the investigators said.
They proposed several mechanisms that may explain the observed maintenance or modest increase in BMI and GNRI in HF patients with or without frailty.
“Although glucosuria induced by SGLT2 inhibitors leads to modest calorie loss, prior studies indicate that this energy deficit is partially compensated by triggered increases in food intake, resulting in less weight loss than predicted from urinary glucose loss alone,” the investigators explained.
“In addition, the symptomatic improvement associated with SGLT2i therapy—including reduced dyspnoea and improved exercise tolerance—may enhance appetite and physical activity, contributing to better nutritional status,” they said.
Careful interpretation needed
However, the investigators emphasized that the stronger association observed between SGLT2 inhibitor use and the primary outcome requires careful interpretation.
“Because patients with frailty have a substantially higher baseline risk of adverse outcomes, the relative risk reduction associated with SGLT2 inhibitor therapy may appear greater without necessarily indicating true biological effect modification,” they said.
“The lack of a clear association in patients without frailty may also be influenced by the higher use of guideline‐directed medical therapy in this group, which may attenuate the incremental benefit of SGLT2 inhibitors,” they added.
The present findings are hypothesis generating and should be validated in prospective studies, according to the investigators.
WET‐HF 2 included 5,579 patients (median age 79 years, 43 percent female, average BMI 23.1 kg/m2) hospitalized for acute HF. HF was driven by an ischaemic cause in 23 percent of patients, and left ventricular ejection fraction was 45 percent. Nearly a quarter (24 percent) received SGLT2 inhibitor prescription at discharge.