. Author manuscript; available in PMC: 2006 Jan 25.

Published in final edited form as: Blood. 2004 Dec 21;105(7):2973–2978. doi: 10.1182/blood-2004-09-3660

Table 4.

Logistic regression model to examine H–Y antibodies and transplantation outcomes

	≥1 H–Y antibody detected
Clinical outcomes	Yes n = 39 (%)	No n = 36 (%)	Unadjusted odds ratio (95% CI)	Adjusted odds ratio^* (95% CI)
Acute GVHD II-IV	10 (26)	15 (42)	0.48 (0.18-1.28)	0.24 (0.05-1.23)
Acute GVHD 0-I	29 (74)	21 (58)	P = .15	P = .11
Chronic GVHD	34 (87)	11 (31)	15.5 (4.8-50.1)	56.5 (7.5-425)
No cGVHD	5 (13)	25 (69)	P < .0001	P < .0001
Relapse	0	13 (36)	0.02 (0.001-0.39)	—
No relapse	39 (100)	23 (64)	P < .0001
cGVHD and no relapse	34 (87)	11 (31)	7.4 (2.1-25.8)^†	22.2 (2.4-206)^†
No cGVHD and no relapse	5 (13)	12 (33)	P = .001^‡	P = .006^‡
Relapse and no cGVHD	0 (0)	13 (36)	0.08 (0.004-1.7)^†	—

—indicates odds ratio is not determinable.

Logistic regression model variables included patient age, donor age, related versus unrelated donors, bone marrow versus peripheral blood stem cell source, myeloablative versus nonablative conditioning regimens, T-cell depletion, disease at conditioning, good prognosis at conditioning, and acute GVHD.

^†

“No cGVHD and no relapse” is the referent group.

^‡

cGVHD and no relapse versus no cGVHD and no relapse.