Table 2.

Multivariate logistic regression analysis of intrracerebral hemorrhage in CSVD patients.

	Model 1			Model 2			Model 3			Model 4
	Variables	Adjusted OR (95%CI)	P-value	Variables	Adjusted OR (95%CI)	P-value	Variables	Adjusted OR (95%CI)	P-value	Variables	Adjusted OR (95%CI)	P-value
	IS/TIA	4.734 (2.384–9.403)	< 0.001	IS/TIA	4.778 (2.410–9.471)	< 0.001	IS/TIA	4.983 (2.495–9.952)	< 0.001	IS/TIA	4.991 (2.500–9.962)	< 0.001
Step1*	Total CSVD burden	1.363 (1.045–1.779)	0.022	Total CSVD burden	1.335 (1.025–1.738)	0.032	Total CSVD burden	1.354 (1.037–1.768)	0.026	Total CSVD burden	1.350 (1.034–1.762)	0.027
	p		NE	p		NE	Ca * P	0.485 (0.248–0.951)	0.035	Adjusted Ca * P	0.730 (0.543–0.998)	0.049
	Ca	0.044 (0.002–0.856)	0.039	Adjusted Ca	0.507 (0.259–0.995)	0.048
Step2#	IS/TIA	5.172 (2.614–10.234)	< 0.001	IS/TIA	5.470 (2.749–10.887)	< 0.001	IS/TIA	5.274 (2.667–10.430)	< 0.001	IS/TIA	5.189 (2.572–10.467)	< 0.001
	HPVS	2.430 (1.282–4.609)	0.007	HPVS	2.346 (1.230–4.473)	0.010	HPVS	2.257 (1.192–4.273)	0.012	HPVS	2.051 (1.044–4.027)	0.037
	p		NE	p		NE	Ca * P	0.497 (0.257–0.964)	0.039	Adjusted Ca * P	0.728 (0.531–0.998)	0.049
	Ca	0.034 (0.002–0.070)	0.029	Adjusted Ca	0.503 (0.259–0.978)	0.043

Model 1–4: There was covariance between Ca, Adjusted Ca, Ca * P, Adjusted Ca * P, and thus it needed to be put into the model separately.

Step 1–2: There was covariance between CSVD total burden and HPVS, high-grade white matter hyperintensities, Microbleeds, Subtentorial, and therefore it needed to be corrected separately.

IS ischemic stroke, TIA transient ischemic attack, Ca calcium, P phosphorus, Ca * P calcium-phosphorus product, HPVS high-grade perivascular space, CSVD cerebral small vessel disease. NE not entry, OR odds ratio, CI confidence interval; p < 0.05 was considered meaningful.

*Adjusted for diastolic blood pressure, IS/TIA, total CSVD burden.

^#Adjusted for diastolic blood pressure, IS/TIA, HPVS, high-grade white matter hyperintensities, Microbleeds, Subtentorial.