Table 1:
Demographics grouped by dataset split.
| Grouped by Dataset Split | |||||||
|---|---|---|---|---|---|---|---|
| Missing | Overall | Test Set | Training Set | Validation Batch | P-Value | ||
| n | 45 | 7 | 15 | 23 | |||
| Sex, n (%) | Female | 0 | 26 (57.8) | 3 (42.9) | 9 (60.0) | 14 (60.9) | 0.684 |
| Male | 19 (42.2) | 4 (57.1) | 6 (40.0) | 9 (39.1) | |||
| Age at Death, Yrs median [Q1,Q3] | 0 | 83.0 [77.0,88.0] | 83.0 [78.0,86.0] | 84.0 [78.0,90.0] | 82.0 [78.0,86.5] | 0.574 | |
| Hispanic ethnicity, n (%) | Yes | 0 | 13 (28.9) | 1 (14.3) | 4 (26.7) | 8 (34.8) | 0.562 |
| No | 32 (71.1) | 6 (85.7) | 11 (73.3) | 15 (65.2) | |||
| Hispanic origins, n (%) | Cuban | 32 | 1 (7.7) | 1 (12.5) | 0.785 | ||
| Dominican | 2 (15.4) | 1 (25.0) | 1 (12.5) | ||||
| Mexican | 4 (30.8) | 1 (25.0) | 3 (37.5) | ||||
| Puerto Rican | 3 (23.1) | 1 (25.0) | 2 (25.0) | ||||
| Unknown | 3 (23.1) | 1 (100.0) | 1 (25.0) | 1 (12.5) | |||
| Race, n (%) | African American | 0 | 1 (2.2) | 1 (6.7) | 0.496 | ||
| Other | 4 (8.9) | 2 (13.3) | 2 (8.7) | ||||
| Unknown | 3 (6.7) | 2 (13.3) | 1 (4.3) | ||||
| White | 37 (82.2) | 7 (100.0) | 10 (66.7) | 20 (87.0) | |||
| Years of Education, median [Q1,Q3] | 0 | 15.0 [12.0,16.0] | 14.0 [12.5,15.5] | 16.0 [10.5,16.0] | 14.0 [12.0,16.0] | 0.902 | |
| CERAD-like* NFT Score, n (%) | 0 | 0 | 1 (2.2) | 1 (14.3) | 0.044 | ||
| 1 | 8 (17.8) | 1 (14.3) | 5 (33.3) | 2 (8.7) | |||
| 2 | 16 (35.6) | 4 (57.1) | 5 (33.3) | 7 (30.4) | |||
| 3 | 20 (44.4) | 1 (14.3) | 5 (33.3) | 14 (60.9) | |||
| PMI, median [Q1,Q3] | 34 | 6.0 [2.8,21.9] | 24.0 [22.4,36.0] | 6.0 [3.1,14.51] | 5.2 [0.0,6.0] | 0.057 | |
| Primary Clinical Diagnosis, n (%) | AD | 2 | 20 (46.5) | 4 (57.1) | 4 (30.8) | 12 (52.2) | 0.169 |
| CERAD Definite AD | 2 (4.7) | 1 (14.3) | 1 (7.7) | ||||
| CERAD Probable AD | 1 (2.3) | 1 (7.7) | |||||
| Definite AD | 11 (25.6) | 5 (38.5) | 6 (26.1) | ||||
| LB variant AD | 8 (18.6) | 1 (14.3) | 2 (15.4) | 5 (21.7) | |||
| Probable AD | 1 (2.3) | 1 (14.3) | |||||
| ADRC Origin, n (%) | Columbia | 2 | 20 (46.5) | 2 (28.6) | 7 (53.8) | 11 (47.8) | 0.642 |
| UC Davis | 9 (20.9) | 1 (14.3) | 3 (23.1) | 5 (21.7) | |||
| UC San Diego | 14 (32.6) | 4 (57.1) | 3 (23.1) | 7 (30.4) | |||
| Braak NFT Stage, n (%) | III | 15 | 1 (3.3) | 1 (14.3) | 0.323 | ||
| IV | 4 (13.3) | 1 (14.3) | 2 (25.0) | 1 (6.7) | |||
| V | 6 (20.0) | 2 (28.6) | 4 (26.7) | ||||
| VI | 19 (63.3) | 3 (42.9) | 6 (75.0) | 10 (66.7) | |||
| CERAD, n (%) | Sparse | 12 | 2 (6.1) | 1 (10.0) | 1 (5.9) | 0.496 | |
| Moderate | 4 (12.1) | 2 (33.3) | 1 (10.0) | 1 (5.9) | |||
| Frequent | 20 (60.6) | 4 (66.7) | 5 (50.0) | 11 (64.7) | |||
| Not assessed | 7 (21.2) | 3 (30.0) | 4 (23.5) | ||||
| Thal, n (%) | 4 | 31 | 1 (7.1) | 1 (25.0) | 0.559 | ||
| 5 | 10 (71.4) | 2 (50.0) | 3 (75.0) | 5 (83.3) | |||
| Not assessed | 3 (21.4) | 1 (25.0) | 1 (25.0) | 1 (16.7) | |||
Demographics, neuropathologic variables, and clinical diagnoses. P-values were calculated between datasets using appropriate statistical tests; continuous and normally distributed variables were tested using one-way ANOVA, continuous and non-normal variables were tested using Kruskal-Wallis, and categorical variables were tested via Chi-squared test. PMI: Post-mortem interval; LB: Lewy body.