TABLE 2.
Multivariable logistic regression coefficients of the stepwise model.
| Variable | β | SE | P-value |
| Age | 1.057 | 0.186 | 0.000 |
| Weight | –0.360 | 0.171 | 0.036 |
| Education | –0.308 | 0.132 | 0.020 |
| Depression | 0.248 | 0.133 | 0.061 |
| Stroke | 0.209 | 0.121 | 0.086 |
| APOEe4 | 0.311 | 0.137 | 0.023 |
| Leather | –0.255 | 0.157 | 0.104 |
| Peppermint | –0.331 | 0.111 | 0.003 |
| Banana | –0.310 | 0.148 | 0.036 |
| Lemon | 0.261 | 0.158 | 0.098 |
| Pineapple | 0.406 | 0.166 | 0.014 |
| Rose | –0.241 | 0.149 | 0.107 |
| MMSE | –0.693 | 0.133 | 0.000 |
SE, standard error; APOE, apolipoprotein; MMSE, Mini-Mental State Examination. β corresponds to the change per year for age and education, per unit for weight and MMSE, with versus without depression or stroke, and positive versus negative for APOE-ε4 and individual odor tests.