Table 2.
Subgroup analyses of vegetable consumption and metabolic syndrome (MetS)
| Stratification | No. of studies | Pooled RR | 95 % CI | Heterogeneity |
|---|---|---|---|---|
| All studies | 16 | 0·89 | 0·85, 0·93 | P=0·12; I 2=30·1 % |
| Type of vegetable | ||||
| Green vegetables | 3 | 1·10 | 0·98, 1·24 | P=0·47; I 2=0 % |
| Gender | ||||
| Male | 6 | 0·91 | 0·81, 1·03 | P=0·005; I 2=70 % |
| Female | 8 | 0·89 | 0·83, 0·95 | P=0·12; I 2=39 % |
| Design | ||||
| Cross-sectional or case–control | 13 | 0·89 | 0·85, 0·94 | P=0·07; I 2=40 % |
| Cohort | 3 | 0·89 | 0·81, 0·98 | P=0·50; I 2=0 % |
| Diagnostic criteria of MetS | ||||
| NCEP ATP III | 14 | 0·89 | 0·85, 0·94 | P=0·09; I 2=35 % |
| IDF | 1 | 0·99 | 0·80, 1·23 | – |
| Other | 1 | 0·86 | 0·77, 0·96 | – |
| Geographical region | ||||
| Asia | 11 | 0·92 | 0·87, 0·96 | P=0·13; I 2=34 % |
| Europe | 3 | 0·83 | 0·76, 0·90 | P=0·38; I 2=0 % |
| South America | 2 | 1·09 | 0·59, 1·99 | P=0·87; I 2=0 % |
| Sample size | ||||
| <1000 | 5 | 0·88 | 0·76, 1·01 | P=0·98; I 2=0 % |
| >1000 | 11 | 0·89 | 0·83, 0·96 | P=0·02; I 2=52 % |
| Age of population | ||||
| Adult | 14 | 0·92 | 0·87, 0·97 | P=0·21; I 2=22 % |
| Adolescents | 2 | 0·83 | 0·77, 0·90 | P=0·38; I 2=0 % |
RR, relative risk; NCEP ATP III, National Cholesterol Education Program–Adult Treatment Panel III; IDF, International Diabetes Federation.