TABLE 2.

The association between the incidence of postoperative delirium and the preoperative homocysteine, postoperative C-reactive protein, or their interaction.

	OR (95% CI)	P-value	OR (95% CI)	P-value	OR (95% CI)	P-value
	Univariate		Main effect		Interaction effect
Postoperative delirium incidence before adjustment
Postoperative CRP	1.47 (1.04, 2.09)	0.030	1.43 (1.01, 2.04)	0.047	1.53 (1.06, 2.22)	0.023
Preoperative homocysteine	1.34 (0.94, 1.89)	0.102	1.28 (0.90, 1.83)	0.166	1.42 (0.98, 2.05)	0.064
Interaction of CRP and Homocysteine	N.A.		N.A		0.73 (0.53, 1.00)	0.051
Postoperative delirium incidence after adjustment for age, preoperative MMSE, sex, and day of postoperative CRP measurement
Postoperative CRP	1.51 (1.05–2.16)	0.026	1.47 (1.02–2.12)	0.038	1.57 (1.07, 2.30)	0.020
Preoperative homocysteine	1.28 (0.88–1.87)	0.189	1.22 (0.83–1.79)	0.308	1.39 (0.93, 2.08)	0.126
Interaction of CRP and Homocysteine	N.A.		N.A.		0.71 (0.51, 0.99)	0.044

CI, confidence interval; CRP, C-reactive protein; N.A., Not applicable; OR, odds ratio; MMSE, Mini-Mental State Exam. The variables were scaled with Z scores in the models. Binary logistic regression was used to analyze the data, assuming an odds ratio of 1 under the null hypothesis.