. Author manuscript; available in PMC: 2010 Oct 1.

Published in final edited form as: Arterioscler Thromb Vasc Biol. 2009 Jul 16;29(10):1696–1701. doi: 10.1161/ATVBAHA.109.192179

Table 3.

Associations among LRNC size, presence of hemorrhage, and fibrous cap status. Ordinal logistic regression was used for the univariate and multivariate models, unless otherwise specified.

	Fibrous cap status			Univariate model		Multivariate model
	Thick	Thin	Ruptured	OR (95%CI)	p	OR (95%CI)	p
Slice based
N	565	194	39
Hemorrhage present (%)	27 (5)	54(28)	25(64)	12.78 (8.33-19.61)	<0.001	5.91 (2.66-13.12)	<0.001
%LRNC (OR: per 10%)	20.1 ±11.8^*	33.4 ±15.0^*	41.3 ±19.3^*	1.95 (1.53-2.47)	<0.001	1.49 (1.04-2.13)	0.02
Artery based
N	43	50	16
Hemorrhage present (%)	0 (0)	10(20)	14(88)	NA^†	<0.001^‡	NA^†	<0.001^‡
Maximum %LRNC (OR: per 10%)	22.5 ±11.0^*	39.4 ±12.2^*	53.1±13.8^*	3.33 (1.94-5.71)	<0.001	1.83 (1.18-2.84) ^§	0.003^§

Mean ± SD;

^†

OR cannot be computed for 3-level cap status variable from ordinal logistic regression; for ruptured vs. thin cap unadjusted OR = 28 (95% CI, 5-130, p < 0.001), adjusted OR = 17 (95% CI, 3-108, p = 0.002).

^‡

p-value from linear regression analysis;

^§

ordinal logistic regression stratified on hemorrhage presence/absence.