Heart failure & cardiomyopathiesMeta-analysis
Exercise Training for Patients with Heart Failure with Preserved Ejection Fraction: A Systematic Review and Bayesian Network Meta-Analysis
- Systematic Review and Bayesian Network Meta-Analysis
- Bayesian random-effects network meta-analysis
- eight nodes
Authors' conclusion
CART, HIIT and MCT carry the clearest evidence of improved exercise capacity in HFpEF. FES is a candidate where QoL is the priority. Several nodes rest on single small trials.
Study context
- Population
- Adults with heart failure with preserved ejection fraction (HFpEF) (ejection fraction ≥50%) in randomized controlled trials of exerciserandomized, 1,161 after deduplication, 20 reports of 15 randomized controlled trials (RCTs)
- Participants
- 1,554 adults
- Comparison
- Exercise training modalities in HFpEF compared and ranked for exercise capacity, cardiac function and quality of life (QoL)
Pooled result
As reported in the abstract
Peak VO2
SMD (95% CI)P value
CART
0.58(0.31–0.86)
P = 0.002HIIT
0.56(0.07–1.09)
P = 0.026MCT
0.46(0.23–0.70)
P = 0.001Further pooled results
QoL
| Comparison | SMD (95% CI) | |
|---|---|---|
| FES · 2 trials, n = 30; only node with a CrI excluding zero | 0.86 (0.25–1.48) | |
| −1012Difference (linear scale) |
E/A ratio
| Comparison | SMD (95% CI) | |
|---|---|---|
| FES · ranked first; CrI including zero | 0.68 (−0.05-1.41) | |
| −1012Difference (linear scale) |
LVEF, BNP, left atrial volume index
- No node affected left ventricular ejection fraction, B-type natriuretic peptide or left atrial volume index
Baseline characteristics
- mean age
- 69.0 ± 8.0 years
- women
- 63.0%
- trials enrolling NYHA class II-III
- 70%
Session completion
- ≤12-week trials
- 76-90%
- 12 months
- 38.1%
Heterogeneity, other outcomes
- I2 elsewhere
- ≤6.9%
Original publication
Exercise Training for Patients with Heart Failure with Preserved Ejection Fraction: A Systematic Review and Bayesian Network Meta-Analysis. ESC Heart Fail. Published 15 September 2026. doi:10.1093/eschf/xvag237
Summarised from the published abstract; every number is checked against it