Table 5.
The independent effects of admixture and ethnic identity upon dementia prevalence (weighted analysis)
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| APOE genotype | ||||
| One or more e4 allele | 2.21 (1.58-3.09) p < 0.0011 |
2.19 (1.54-3.10) p < 0.0011 |
2.13 (1.50-3.02) p < 0.0011 |
2.45 (1.77-3.40) p < 0.0011 |
| Admixture | ||||
| 100% African versus 100% European admixture | 1.53 (0.80-2.91) p = 0.201 |
- | 4.62 (1.48-14.5) p = 0.011 |
2.55 (0.75-8.61) p = 0.131 |
| Ethnic identity | ||||
| 'White' | - | 1 (ref) | 1 (ref) | 1 (ref) |
| 'Mixed' | - | 0.79 (0.47-1.33) p = 0.381 |
0.54 (0.30-0.96) p = 0.041 |
0.60 (0.34-1.09) p = 0.091 |
| 'Black' | - | 1.02 (0.68-1.51) p = 0.931 |
0.50 (0.25-1.00) p = 0.051 |
0.47 (0.22-1.02) p = 0.061 |
| Sociodemographic factors | ||||
| Age (years) | - | - | - | 1.10 (1.07-1.12) p < 0.0011 |
| Male sex | - | - | - | 0.84 (0.59-1.21) p = 0.36 |
| Education (per level) | - | - | - | 0.76 (0.64-0.90) p = 0.001 |
1. p-values are taken from the model z-scores. It was not possible to run likelihood ratio tests because robust variance estimates were required given the use of sampling weights
Model 1 African admixture and APOE
Model 2 Ethnic identity and APOE
Model 3 African admixture, ethnic identity and APOE
Model 3 African admixture, ethnic identity, APOE, age, sex and education