TABLE 4.

Binary logistic regression models to investigate the variables that discriminate between “Good” and “Poor” cognitive performers.

	IL-1β				IL-6
	B	SE	Wald	Exp(B)	B	SE	Wald	Exp(B)
Sexa	–0.551	0.484	1.296	0.577	–0.452	0.481	0.884	0.636
Age	0.043	0.029	2.141	1.043	0.012	0.029	0.159	1.012
School years	–0.313	0.120	6.778**	0.732	–0.342	0.123	7.700**	0.710
GDS	0.506	0.240	4.452*	1.659	0.594	0.236	6.344*	1.811
Anti-inflammatory drugs	–0.345	0.625	0.305	0.708	–0.507	0.626	0.657	0.602
IL-1β	0.326	0.125	6.777**	1.385
IL-6 (log10)					3.906	1.851	4.453*	49.708
χ2 (df)	36.653 (6)***				36.505 (6)***
R2Nagelkerke (change)	0.370 (0.074)				0.363 (0.044)
Total hit rates (%)	69.000				68.700
	IL-8				IL-13
	B	SE	Wald	Exp(B)	B	SE	Wald	Exp(B)
Sexa	–0.558	0.475	1.384	0.572	–0.694	0.499	1.937	0.499
Age	0.030	0.028	1.139	1.031	0.039	0.029	1.847	1.040
School years	–0.333	0.125	7.095**	0.717	–0.371	0.122	9.178**	0.690
GDS	0.558	0.233	5.706*	1.747	0.498	0.243	4.213*	1.646
Anti-inflammatory drugs	–0.254	0.614	0.172	0.775	–0.173	0.634	0.074	0.841
IL-8	0.231	0.119	3.767*	1.260
IL-13					0.225	0.071	9.926**	1.252
χ2 (df)	34.754 (6)***				40.648 (6)***
R2Nagelkerke (change)	0.348 (0.035)				0.407 (0.107)
Total hit rates (%)	68.7				70.500

SE, standard error; GDS, Geriatric Depression Scale. ^aReference category is male. *p ≤ 0.05, **p < 0.01, ***p < 0.001.