Table 3.
Grading of Recommendations, Assessment, Development and Evaluation (GRADE) assessment of polyphenol supplementation compared to placebo for improving cardiovascular risk factors in haemodialysis patients.
| Quality Assessment | Number of Patients | Effect | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Number of Studies | Study Design | Risk of Bias | Inconsistency | Indirectness | Imprecision | Other Considerations | Polyphenol | Placebo | Absolute (95% CI) | |
| Oxidative stress (myeloperoxidase) | ||||||||||
| 2 | Randomised trials | Not serious | Not serious | Not serious | Serious a | Strong effect | 82 | 44 | MD 90.1 SD lower (135.8 lower to 44.4 lower) | MODERATE |
| Oxidative stress (other markers—not pooled) | ||||||||||
| 5 | Randomised trials | Not serious | serious b | Not serious | Serious c | None | - | - | See comment | MODERATE |
| Inflammation (CRP) | ||||||||||
| 5 | Randomised trials | Not serious | very serious d | Not serious | Serious e | None | 96 | 99 | MD 1.9 mg/dL higher (2.2 lower to 6.1 higher) | VERY LOW |
| Inflammation (IL-6) | ||||||||||
| 2 | Randomised trials | Not serious | very serious d | Not serious | Serious e | Both pomegranate studies | 79 | 49 | MD 1.6 mg/dL lower (5.1 lower to 1.96 higher) | VERY LOW |
| Inflammation (AOPP) | ||||||||||
| 3 | Randomised trials | Not serious | very serious d | Not serious | Serious e | All pomegranate studies | 96 | 57 | MD 17.7 mg/dL lower (46.5 lower to 11.1 higher) | VERY LOW |
| Diastolic blood pressure | ||||||||||
| 4 | Randomised trials | Not serious | Not serious | Not serious | Not serious | None | 133 | 112 | MD 5.6 mmHg lower (8.47 lower to 2.78 lower) | HIGH |
| Systolic blood pressure | ||||||||||
| 4 | Randomised trials | Not serious | Serious f | Not serious | Serious e | None | 107 | 86 | MD 10 mmHg lower (21.4 lower to 1.4 higher) | LOW |
| Hemodynamic measures (other markers—not pooled) | ||||||||||
| 4 | Randomised trials | Not serious | Serious g | Not serious | Serious h | None | - | - | Not pooled. See Table 2 | LOW |
| Lipid profile (TG) | ||||||||||
| 4 | Randomised trials | Not serious | Serious f | Not serious | Very serious e | None | 110 | 81 | MD 26.5 mg/dL lower (47.2 lower to 5.8 lower) | VERY LOW |
| Lipid profile (Total-C) | ||||||||||
| 4 | Randomised trials | Not serious | Very serious d | Not serious | Very serious e | None | 110 | 81 | MD 11.2 mg/dL lower (24.8 lower to 2.3 higher) | VERY LOW |
| Lipid profile (HDL-C) | ||||||||||
| 4 | Randomised trials | Not serious | Not serious | Not serious | Serious e | None | 110 | 81 | MD 2.4 mg/dL higher (0.1 lower to 4.8 higher) | MODERATE |
| Lipid profile (LDL-C) | ||||||||||
| 4 | Randomised trials | Not serious | Not serious | Not serious | Serious e | None | 110 | 81 | MD 3.3 mg/dL lower (14.5 lower to 7.8 higher) | MODERATE |
CI: Confidence interval; dL: decilitre; HDL-C, High Density Lipoprotein Cholesterol; LDL-C, Low Density Lipoprotein Cholesterol; MD: Mean difference; Total-C, Total Cholesterol; TG, Triglycerides. a Two studies were pooled in a meta-analysis for the outcome of myeloperoxidase. The 95%CI for the outcome was substantial (−135 to −44). b Although most measures of oxidative stress showed improvement from baseline in the intervention group; a number of control groups also showed improvement in this outcome. Therefore, decreasing confidence in the consistency of results for this outcome. c Although some had narrow measures of variance, other did show substantial imprecision in their measures of variance. d Heterogeneity was high (>70%). e The 95%CI was substantial. f Heterogeneity was substantial (50–70%). g Studies showed some improvement in mean arterial pressure and flow-mediated dilation; but other studies found no improvement. No improvement was seen in other measures. h Although individual studies and outcomes did not show significant variance; there is overall a small number of participants for each outcome as well as combined. This decreases confidence in the precision of the effects.