Table 3.
Evidence quality rated using the GRADE approach.
| Outcomes | No. of Studies | Limitations | Inconsistency | Indirectness | Imprecision | Publication Bias | Evidence Quality | |
|---|---|---|---|---|---|---|---|---|
| Serum vitamin D levels | 16 | Serious 2,3,4 | Serious 5 | Not serious | Not serious | Not found | ⊕⊕⊝⊝ | Low |
| FBG 1 | 13 | Serious 2,3,4 | Serious 5 | Not serious | Serious 6 | Not found | ⊕⊝⊝⊝ | Very low |
| HbA1c 1 | 15 | Serious 3,4 | Serious 5 | Not serious | Serious 6 | Not found | ⊕⊝⊝⊝ | Very low |
| HOMA-IR 1 | 7 | Serious 3,4 | Serious 5 | Not serious | Not serious | Not assessed 7 | ⊕⊕⊝⊝ | Low |
| Fasting serum insulin | 4 | Serious 4 | Serious 5 | Not serious | Serious 6 | Not assessed 7 | ⊕⊝⊝⊝ | Very low |
1 Abbreviations: FBG: fasting blood glucose; HbA1c: hemoglobin A1c; HOMA-IR: homeostasis model assessment of insulin resistance. 2 Had single-blind randomized controlled trial design; 3 had reporting bias; 4 had attrition bias. 5 A significant heterogeneity was observed in this meta-analysis. 6 Wide confidence interval, including values in favour of the experimental group and values in favour of the control group. 7 Not assessed because a limited number of studies were included in the meta-analyses on HOMA-IR and fasting serum insulin.