Table 4.

Associations between dairy variables and breast cancers classified by hormone receptor status.^a Models each for total dairy and dairy milk, supporting results for: (a) estrogen receptor (ER) status; (b) progesterone receptor (PR) status; (c) combinations of ER/PR

Dairy variable		Total dairy (kcal/d)				Dairy-milk (kcal/d)
Cancer receptor(s)	n(cases)	HRb	95% CI	p1d	p2d	HRb	95% CI	p1d	p2d
ER+	737	1.25	1.06–1.48	0.008	0.79	1.36	1.08–1.71	0.008	0.55
ER−	172	1.31	0.98–1.76	0.07	–	1.52	1.03–2.24	0.04	–
PR+	615	1.30	1.08–1.56	0.006	0.43	1.45	1.13–1.86	0.003	0.36
PR−	282	1.17	0.94–1.56	0.16	–	1.22	0.89–1.66	0.22	–
ER+/PR−	116	1.06	0.77–1.47	0.71	0.56	0.98	0.62–1.53	0.92	0.42
ER+/PR+	608	1.29	1.07–1.56	0.007	0.88	1.43	1.11–1.83	0.005	0.94
ER−/PR−	165	1.27	0.94–1.71	0.11	0.42	1.43	0.97–2.11	0.074	0.18
ER−/PR+	7	c	c	c	-	c	c	c	-

^aSame covariates as those listed as footnotes to Table 1 for model 2.

^bComparing medians of extreme quintiles of dairy intake.

^cInsufficient cases.

^dp1 relates to the test of whether the dairy variable has a CI that includes the null for that named cancer subtype; p2 relates to a test of whether the slope of association between the dairy variable differs between the named cancer subtype and that in the row below it. An exception is that the second to last line tests ER− /PR− vs ER+/PR−. Chi square joint Wald tests of no difference between the ER/PR combinations gave p = 0.80 (dairy kcals) and p = 0.58 (dairy milks).