Results of fixed effects meta-analyses
| Number of studies | Odds ratio | 95% CI | p value | |
|---|---|---|---|---|
| *Female data: in Rasmusson (1995) neither cases nor controls had had a head injury so the odds ratio cannot be computed. | ||||
| †Male data: Fratiglioni (1993) had an estimated odds ratio of 0 and could therefore not contribute to the meta-analysis. | ||||
| All studies | 15 | 1.58 | 1.21 to 2.06 | 0.001 |
| P only | 10 | 1.56 | 1.12 to 2.18 | 0.01 |
| Q only | 6 | 1.42 | 0.75 to 2.67 | 0.28 |
| Post-Mortimer | 7 | 1.35 | 0.94 to 1.94 | 0.10 |
| Pre-Mortimer | 8 | 1.90 | 1.29 to 2.81 | 0.001 |
| Females* | ||||
| All studies | 7 | 0.91 | 0.56 to 1.47 | 0.69 |
| Post-Mortimer | 4 | 0.92 | 0.53 to 1.59 | 0.75 |
| Males† | ||||
| All studies | 8 | 2.29 | 1.47 to 3.58 | <0.001 |
| Post-Mortimer | 4 | 2.26 | 1.13 to 4.53 | 0.02 |