Table 2.
Univariable meta-regression models to assess the influence of various study-level factors among articles investigating the association between MetS and risk of incident all-cause dementia (eligible results N = 11)
| Study characteristic | Estimate | 95% CI | Q (m) | p value | Adjusted p* |
|---|---|---|---|---|---|
| Publication year | 0.004 | -0.007–0.014 | 0.425 | 0.514 | 0.249 |
| Mean/median follow-up (in years) | -0.001 | -0.006–0.004 | 0.117 | 0.732 | 0.646 |
| Mean/median age at baseline (in years) | -0.003 | -0.008–0.002 | 1.036 | 0.309 | 0.043 |
| Used NCEP-ATP III MetS criteria (yes vs. no) | -0.145 | -0.336–0.046 | 2.212 | 0.137 | 0.019 |
| % female sex | -0.002 | -0.007–0.003 | 0.644 | 0.422 | 0.116 |
| Adjusted for APOE-ε4 carrier status (yes vs. no) | -0.017 | -0.064–0.030 | 0.500 | 0.480 | 0.225 |
| Ethnicity (predominantly White vs. not) | -0.011 | -0.057–0.036 | 0.197 | 0.657 | 0.570 |
| Geographic region (Western vs. non-Western) | -0.015 | -0.061–0.031 | 0.432 | 0.511 | 0.345 |
Abbreviations: NCEP-ATP III = National Cholesterol Education Program – Adult Treatment Panel III; Q (m) = Q for moderators; MetS = Metabolic Syndrome; APOE = Apolipoprotein-E
Note: Pooled estimates were calculated using random-effects meta-analysis models
*Using permutation test