Table 5.
Differences in the AD likelihood as assigned by the statistical regression model (%) for each variable among persons diagnosed with SCI and MCI
| SCI | MCI | |||
|---|---|---|---|---|
| β | p | β | p | |
| Number of cardiovascular risk factors | 2.47 | 0.044 | −0.63 | 0.770 |
| Arterial hypertension | 6.67 | 0.002 | −3.80 | 0.361 |
| Stroke/TIA | 3.60 | 0.544 | −26.24 | 0.011 |
| ASA treatment | −0.90 | 0.726 | −8.49 | 0.086 |
| Years of education | 0.53 | 0.056 | 0.31 | 0.586 |
| One or more ApoE4 allele | 5.86 | 0.060 | 27.97 | <0.001 |
| Central atrophy | 7.32 | 0.006 | −12.43 | 0.005 |
| MTA R <3 | 53.47 | <0.001 | −6.34 | 0.514 |
| MTA L >3 | 21.10 | 0.007 | −16.56 | 0.023 |
| WML | 2.57 | 0.101 | −5.78 | 0.042 |
| Confluent WML | 9.00 | 0.033 | −11.47 | 0.045 |
Results of linear regression with AD likelihood as outcome. The outcome is the probability of being diagnosed with AD (ranging from 0.0 to 99.2%). The results are presented as β coefficients and p values. The β coefficients show the percent differences in the likelihood of being diagnosed with AD. Results are controlled for age and sex. TIA = Transient ischemic attack; ASA = acetylsalicylic acid; MTA R >3 = MTA rating on the right side >3 (dichotomous variable); MTA L >3 = MTA rating on the left side >3 (dichotomous variable); WML = WML as a trichotomous variable: none, few and confluent; confluent WML = presence vs. absence of confluent WML.