TABLE 5.
Subgroup analysis for SMD of dietary beta-carotene in depression vs. control subjects.
| Stratification | Number of studies | Pooled SMD | 95% CI | P-value | Heterogeneity |
| All studies | 14 | −0.34 | −0.48, −0.20 | P < 0.001 | P < 0.001; I2 = 96% |
| Geographical region | |||||
| Asia | 9 | −0.43 | −0.67, −0.20 | P < 0.001 | P < 0.001; I2 = 97% |
| Non−Asia | 6 | −0.20 | −0.24, −0.16 | P < 0.001 | P = 0.96; I2 = 0% |
| Exposure assessment | |||||
| FFQ | 10 | −0.41 | −0.63, −0.18 | P < 0.001 | P < 0.001; I2 = 97% |
| Recall method | 4 | −0.20 | −0.24, −0.16 | P < 0.001 | P = 0.95; I2 = 0% |
| Sex | |||||
| Male | 5 | −0.63 | −1.32, 0.07 | P = 0.08 | P < 0.001; I2 = 99% |
| Female | 9 | −0.16 | −0.20, −0.12 | P < 0.001 | P = 0.33; I2 = 13% |
| Population | |||||
| Adolescent | 4 | −0.33 | −0.59, −0.06 | P = 0.02 | P = 0.01; I2 = 73% |
| Middle aged and elderly | 10 | −0.34 | −0.50, −0.17 | P < 0.001 | P < 0.001; I2 = 97% |
| Sample size | |||||
| <1,000 | 5 | −0.27 | −0.47, −0.07 | P = 0.008 | P = 0.02; I2 = 65% |
| >1,000 | 9 | −0.35 | −0.52, −0.18 | P < 0.001 | P < 0.001; I2 = 97% |