Abstract Background Excess fat, particularly visceral fat, plays an important role in the development of heart failure with preserved ejection fraction (HFpEF). Sodium glucose cotransporter-2 inhibitors improve hemodynamics, reduce symptom severity, and decrease risk of hospitalization, but their impact on body composition, as well as potential relationships between changes in body composition and hemodynamic effects, are unknown in HFpEF. Purpose Evaluate the effect of dapagliflozin on body and blood composition and examined the potential correlation of such changes on cardiac hemodynamics. Methods The CAMEO-DAPA trial, a single-center, double-blind, randomized, placebo-controlled study, showed that 24 weeks of dapagliflozin reduced resting and exercise pulmonary capillary wedge pressures (PCWP) in patients with HFpEF. This prespecified secondary analysis reports the effects of dapagliflozin on body composition using dual energy x-ray absorptiometry (DEXA), blood composition (radiolabeled iodinated albumin), and cardiac hemodynamics using high-fidelity micromanometers. Analysis was performed based on the intention-to-treat. Relative change was calculated by dividing the absolute change on the baseline value. Results Overall, 37 patients (67 years, 65% women, 70% with obesity, 27% overweight) completed the trial. As compared with baseline, dapagliflozin decreased total body fat compared with placebo [mean (SD) absolute change: -2.29 (2.94) kg vs -0.32 (1.84) in placebo, p=0.03; relative change: -4.6% (5.4) vs -0.8% (4.3), p=0.02], Figure 1. This was primarily due to greater loss of upper body trunk fat [absolute change: -1.53 (1.70) kg vs +0.31 (0.99) in placebo, p<.001; relative change: -5.2% (5.6) vs +1.1% (4.1), p<.001], Figure 1. Reductions in trunk fat were correlated with decreases in PCWP (r=0.41, p=0.03), Figure 2. Total body fat-free mass (FFM) tended to decrease in the dapagliflozin group [absolute change: -0.63 (1.85) kg vs +0.34 (1.76), p=0.12]; reductions in leg, android, and gynoid FFM were all greater than placebo (p<0.05). Dapagliflozin reduced plasma volume [absolute change: -170 (343) ml vs +112 (346), p=0.02] compared with placebo, and the decrease in plasma volume was correlated with the reduction in FFM (r=0.47, p=0.006), Figure 2. Treatment with dapagliflozin had no effect on resting metabolic rate but was associated with modest reduction in bone mineral content [-16 (47) g vs +26 (61), p=0.03]. Conclusions In patients with HFpEF, dapagliflozin decreases body fat, particularly trunk fat, suggesting greater effect on visceral adipose, which is correlated with favorable hemodynamic effects. Dapagliflozin also tends to reduce FFM, which may relate to reduction in total body water with or without reduction is muscle mass, in tandem with modest decreases in bone mineral content.Figure 1:Relative change in fat massFigure 2