Table 3. HLA data used in the model ^# .

				2HB+		1HB+	HLA−
			Canadian Data
			Observed Frequency – Cases (HLA+ and HLA−)##		0.55		0.45
			Observed Frequency – Controls (HLA+ and HLA−)##		0.24		0.76
			OR – (2HB+ & 1HB+) vs. (HLA−)*		3.9
			UCSF #1
			Observed Frequency – Cases†	0.10		0.46	0.44
			Predicted HWE frequencies – Cases††	0.11		0.45	0.44
			Predicted Controls – HWE at: P(HLA+) = 0.24	0.016		0.224	0.76
			OR – (2HB+) vs. (HLA−) & (1HB+) vs. (HLA−) *	10.4		3.6
			OR – (2HB+ & 1HB+) vs. (HLA−)*		4.0
			UCSF #2
			Observed Frequency – Cases†	0.07		0.39	0.54
			Predicted HWE frequencies – Cases††	0.07		0.39	0.54
			Observed Frequency – Controls†	0.012		0.186	0.80
			Predicted HWE frequencies – Controls††	0.011		0.186	0.80
			OR – (2HB+) vs. (HLA−) & (1HB+) vs. (HLA−) *	9.3		3.1
			OR – (2HB+ & 1HB+) vs. (HLA−)*		3.5

^#Numbers listed are genotype frequencies.

2HB+ = carrier of 2 copies of the DRB1*1501 allele (homozygous carrier).

1HB+ = carrier of 1 copies of the DRB1*1501 allele (heterozygous carrier).

HLA− = carrier of 0 DRB1*1501 alleles.

(HLA+) = (2HB+)+(1HB+).

^##Canadian HLA data: D Sadovnick (personal communication).

Based on ∼3,000 cases and ∼400 Controls (% women not available). Control rates confirmed in a much larger transplant database.

^*Odds ratio (OR) versus controls. Calculated as odds of genotype in cases divided by odds of the same genotype in controls.

^†UCSF Databases: J Oksenberg (personal communication).

UCSF #1 (IMSGC) – 779 cases (76% women); No observed controls.

UCSF #2 (GeneMSA) – 485 cases (68% women) and 431 Controls (66% women).

^††Hardy Weinberg Equilibrium (HWE) values predicted based on the observed P(2HB+) in Cases or Controls.