Heart failure & cardiomyopathies
Fluid restriction vs liberal intake in patients with heart failure: a meta-analysis of randomized trials
Pooled result · All-cause mortalityResult — go to the resultRelative risk 0.54 95% CI 0.31–0.94Authors' conclusion
Fluid restriction reduces the risk of all-cause mortality but not HF rehospitalizations in HF patients. Further studies are warranted to definitively confirm the present findings and result on the suitable changes in recommendations and clinical practice.
Study context
- Population
- HF patients9 randomized controlled trials (RCTs), mean follow-up: 196.33 days · fluid restriction (FR) group 614 · non-FR group 657
- Participants
- 1,271 patients
- Comparison
- FR vs liberal fluid intake
- FRn = 614
- non-FRn = 657reference
Methodological detailssystematic literature search, random-effects model with the Hartung-Knapp-Sidik-Jonkman (HKSJ) adjustment
- Systematic literature search
- Random-effects model with the Hartung-Knapp-Sidik-Jonkman (HKSJ) adjustment
Pooled result
All-cause mortality
Relative risk
0.54
95% CI 0.31–0.94
P = 0.03
FR vs non-FR
RR 0.54 (95% CI 0.31–0.94); P = 0.03
As reported in the abstract
Further pooled results
| Outcome | RR (95% CI) | P value | |
|---|---|---|---|
| Heart failure hospital readmissions | 0.67 (0.28–1.65) | P = 0.31 | |
| 0.20.5125Relative risk (log scale) |
| Outcome | WMD (95% CI) | P value | |
|---|---|---|---|
| Perceived thirst | −6.89 (−22.86 to 9.08) | P = 0.26 | |
| −30−20−10010Difference (linear scale) |
| Outcome | WMD (pg/mL) (95% CI) | P value | |
|---|---|---|---|
| Serum BNP levels | 54.09 (−316.86 to 425.04) | P = 0.71 | |
| −400−2000200400600Difference, pg/mL (linear scale) |
| Outcome | WMD (mmol/L) (95% CI) | P value | |
|---|---|---|---|
| Sodium levels | 1.42 (−0.68 to 3.51) | P = 0.15 | |
| −2024Difference, mmol/L (linear scale) |
Mean age
Original publication
Fluid restriction vs liberal intake in patients with heart failure: a meta-analysis of randomized trials. ESC Heart Fail. Published 8 January 2026. doi:10.
Summarised from the published abstract; every number is checked against it