Plasma S100β is a predictor for pathology and cognitive decline in Alzheimer’s disease

Table 7 Model comparisons of peripheral biomarkers vs. consensus diagnosis for CNI and AD patients

Predictor	AICc	Probability^a	χ2	Raw p value	Adjusted p value^a	R²
Plasma S100β levels (pg/mL)	5.200	0.915	17.323 (1)	< 0.001	0.001	1.000
Plasma NSE levels (ng/mL)	11.918	0.032	10.605 (1)	0.001	0.009	0.637
Plasma MMP-2 levels (ng/mL)	10.994	0.051	11.529 (1)	< 0.001	0.008	0.683
Plasma MMP-9 levels (ng/mL)	16.961	0.003	5.562 (1)	0.018	0.073	0.365

^aProbability of the model being “correct” vs. other models in set, derived from ΔAICc (Burnham & Anderson, 2002)
^bAdjusted by Benjamni-Hochberg FDR

ISSN: 2045-8118