Weight changes in heart failure with preserved ejection fraction: findings from TOPCAT.

BACKGROUND

Weight loss has been associated with poor outcomes in patients with heart failure (HF). However, few data are available for patients with heart failure with preserved ejection fraction (HFpEF). The impact of weight gain on outcomes has not been frequently reported either.

AIMS

To study post-randomization weight changes and how these could impact outcomes and the effect of spironolactone in patients with HFpEF enrolled in the TOPCAT-Americas trial (N = 1767).

METHODS

Mixed-effects regressions and time-updated Cox models to assess the factors associated with weight changes and their impact on subsequent outcomes.

RESULTS

Over a median follow-up of 3 years, 824 (47%) patients experienced weight loss ≥ 5% and 390 (22%) experienced weight loss ≥ 10%. Patients experiencing weight loss were older and more frequently women with severe HF symptoms. Spironolactone slightly reduced body weight before 12 months of follow-up: β =  - 0.55 (- 0.82 to - 0.29) kg, without effect on weight afterwards: β = 0.01 (- 0.66 to 0.68) kg; treatment-by-time interaction P = 0.0015. Spironolactone did not increase the odds of weight loss but reduced the odds of weight gain. Weight loss ≥ 5% was associated with a higher risk of cardiovascular and all-cause death irrespective of baseline body mass index: HR = 1.47, 95%CI = 1.07-2.01 and HR = 1.84, 95%CI = 1.46-2.31, respectively. Weight gain was not associated with an increased risk of any outcome.

CONCLUSION

Weight loss ≥ 5% was frequent and independently associated with an increased risk of subsequent mortality. Spironolactone induced only slight body weight reductions early after its introduction and up to a maximum of 8-12 months of follow-up. Association between body weight changes and subsequent death. Legend: HR, hazard ratio from time-updated Cox models. Model adjusted on age, sex, race, NYHA class, systolic blood pressure, diabetes, atrial fibrillation, previous myocardial infarction, previous heart failure hospitalization, estimated glomerular filtration rate, diuretic use, and baseline weight.