Table 5. Statistics for plasma features alone & cognitive alone and paired with best plasma feature.
| Dxb_12x27 | cognitive_alone | cognitive_and _Dxb_12x27 | kit_cardiac_ alone | kit_inflamma-tory_alone | kit_metabolic_alone | kit_neuro _alone | |
|---|---|---|---|---|---|---|---|
| Mann-Whitney U | 437.500 | 1001.500 | 250.500 | 784.500 | 1396.000 | 1990.000 | 893.000 |
| Wil Wilcoxon W | 1427.500 | 1991.500 | 1240.500 | 1774.500 | 2386.000 | 2980.000 | 1883.000 |
| ZZ Z | -8.327 | -6.362 | -8.988 | -7.101 | -4.940 | -2.841 | -6.718 |
| Asymp. Sig. (2-tailed) | .000 | .000 | .000 | .000 | .000 | .004 | .000 |
Goodness of discrimination of NLs from stable MCI using RF techniques. The putative best plasma cohort (Dxb_12x27) is better than Cog alone, or MRI feature groups, or other plasma groups based on p and Z values.