Table 2.

Odds ratios for ductal and lobular breast cancer in relation to single locus genotype

Genotypea	00			01			11			PHWEc
Polymorphism	Cases/Controls	ORb	95% CI	Cases/Controls	ORb	95% CI	Cases/Controls	ORb	95% CI
Promoter TAn
Ductal	369/445	1	Ref	494/593	1.01	0.84–1.21	230/312	0.87	0.70–1.09	<0.001, <0.0001d
Lobular	52/445	1	Ref	81/593	1.30	0.88–1.91	33/212	0.92	0.57–1.49
Intron 1 c.454-397C→T
Ductal	347/384	1	Ref	542/651	0.93	0.77–1.12	229/313	0.80	0.64–1.01	0.32
Lobular	43/384	1	Ref	92/651	1.42	0.95–2.11	39/313	1.18	0.73–1.90
Intron 1 c.454-351A→G
Ductal	522/577	1	Ref	471/610	0.86	0.72–1.02	124/161	0.86	0.66–1.12	0.88
Lobular	66/577	1	Ref	89/610	1.36	0.96–1.92	19/161	1.05	0.60–1.84
Exon 3 c.729C→T
Ductal	1037/1253	1	Ref	84/93	1.07	0.78–1.45	1/3	0.41	0.04–4.00	0.20
Lobular	164/1253	1	Ref	10/93	0.92	0.46–1.83	0/3	-	-
Exon 4 c.975C→G
Ductal	722/805	1	Ref	346/477	0.81	0.68–0.96	54/67	0.93	0.64–1.35	0.87
Lobular	106/805	1	Ref	59/477	0.94	0.66–1.34	9/67	1.00	0.48–2.10

OR, odds ratio. ^aThe genotypes are as follows: TA_n, 00 ≤ 14/≤ 14, 01 ≤ 14/>14, 11 >14/>14; c.454-397C → T, 00 T/T, 01 T/C, 11 C/C; c.454-351A → G, 00 A/A, 01 A/G, 11 G/G; c.729C → T, 00 C/C, 01 C/T, 11 T/T; c.975C → G, 00 C/C, 01 C/G, 11 G/G. ^bLong-term users of menopausal hormones and diabetics were oversampled; thus, logistic regression models are adjusted for age and sampling category, which means that only those with complete information about hormone use and diabetes mellitus were included. ^cP value for χ² test of deviation from the assumption of Hardy–Weinberg equilibrium, among controls only. ^dFor this test for deviation from Hardy–Weinberg equilibrium either all microsatellite alleles were considered or lengths were categorised into ≤10, 11, 12 or 13, 14–16, 17, 18, 19, or ≥ 20.