Table 2.
Quality of evidence-based on GRADE assessment.
Outcome | Risk of bias | Inconsistencya | Indirectness | Imprecision | Publication biasb | Decision |
---|---|---|---|---|---|---|
Insulin sensitivity | Some concerns | Substantial | Some indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
HOMA-IR | Some concerns | Substantial | Some indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
Fasting glucose | Some concerns | Substantial | Some indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
Fasting insulin | Some concerns | Substantial | Some indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
Postprandial glucose | Low | Consistent | No serious indirectness | Moderate imprecision | Undetected | ⊕⊕⊕∅ Moderate |
Postprandial insulin | Low | Substantial | No serious indirectness | Moderate imprecision | Undetected | ⊕⊕⊕∅ Moderate |
HbA1c | Some concerns | Moderate | Some indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
Pancreatic beta-cell function | Low | Consistent | No serious indirectness | Moderate imprecision | Undetected | ⊕⊕∅∅ Low |
GLP-1 | Low | Substantial | Some indirectness | Moderate imprecision | Possible | ⊕∅∅∅ Very Low |
aBased on I2 using thresholds in Cochrane Handbook for Systematic Reviews of Interventions, Version 6. The Cochrane Collaboration, 2019. Available at: https://training.cochrane.org/handbook/current.
bBased on Egger’s regression.