Table 2.

Hazard Ratios and 95% Confidence Intervals for the Associations Between Baseline Hormonal Factors and Breast Cancer Risk Among Postmenopausal Caucasian Women in the Aggregated Sample From 4 US Prospective Studies, 1979–2006

Hormonal Factor	Nulliparous Women (32,641 Women)			Parous Women Aged <25 Years at First Birth (139,255 Women)				Parous Women Aged ≥25 Years at First Birth (65,015 Women)
	No. of Casesa	HRb	95% CI	No. of Casesa	HRbc	95% CI	Pheterogeneityd	No. of Casesa	HRbc	95% CI	Pheterogeneitye
Age at menarchef
Early	642	1.00	Referent	1,993	1.00	Referent		1,099	1.00	Referent
Middle	739	0.93	0.83, 1.04	2,345	0.92	0.86, 0.98		1,477	0.96	0.88, 1.04
Late	215	0.95	0.80, 1.11	726	0.89	0.82, 0.98	0.83	483	0.94	0.84, 1.05	0.91
Age at a natural menopause, years
<45	128	0.81	0.67, 0.99	223	0.78	0.68, 0.90		161	0.81	0.68, 0.95
45–49	325	0.91	0.79, 1.05	751	0.96	0.88, 1.05		519	0.91	0.82, 1.01
50–54	499	1.00	Referent	1,395	1.00	Referent		1,045	1.00	Referent
≥55	106	1.17	0.95, 1.45	392	1.17	1.04, 1.31	0.88	260	1.09	0.95, 1.25	0.95
Surgery/other/unknowng	548	0.77	0.67, 0.87	2,313	0.80	0.74, 0.86		1,082	0.88	0.80, 0.96
Body mass index, kg/m2
<25	858	1.00	Referent	2,316	1.00	Referent		1,653	1.00	Referent
25–<30	450	1.09	0.97, 1.22	1,635	1.11	1.04, 1.18		928	1.12	1.03, 1.21
≥30	254	1.10	0.95, 1.27	1,031	1.22	1.13, 1.32	0.44	437	1.11	1.00, 1.24	0.94
Menopausal hormone therapy useh
Never	554	1.00	Referent	1,423	1.00	Referent		1,010	1.00	Referent
Former	187	0.99	0.84, 1.18	562	1.10	0.99, 1.21		395	1.08	0.96, 1.22
Current, <5 yearsi	163	1.38	1.15, 1.66	527	1.33	1.19, 1.47		362	1.57	1.38, 1.78
Current, ≥5 yearsj	408	1.59	1.37, 1.83	1,391	1.60	1.47, 1.74	0.71	711	1.53	1.37, 1.70	0.46
First-degree family history of breast cancer
No	1,286	1.00	Referent	4,010	1.00	Referent		2,459	1.00	Referent
Yes	278	1.37	1.20, 1.56	965	1.56	1.45, 1.67	0.09	554	1.42	1.29, 1.56	0.65
Personal history of benign breast disease
No	870	1.00	Referent	2,923	1.00	Referent		1,783	1.00	Referent
Yes	681	1.31	1.18, 1.45	1,638k	1.41k	1.33, 1.51	0.25	869k	1.42k	1.30, 1.55	0.25
Weekly consumption of alcohol
None	432	1.00	Referent	1,547	1.00	Referent		853	1.00	Referent
<7 drinks	801	1.05	0.93, 1.18	2,563	1.04	0.98, 1.11		1,553	1.01	0.93, 1.10
≥7 drinks	268	1.30	1.11, 1.52	609	1.22	1.11, 1.35	0.79	450	1.33	1.19, 1.50	0.75

Abbreviations: BCDDP, Breast Cancer Detection Demonstration Project Follow-up Study; CI, confidence interval; HR, hazard ratio; NIH-AARP, National Institutes of Health-AARP Diet and Health Study; PLCO, Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial; USRT, US Radiologic Technologists Study.

^aFirst primary invasive and in situ breast cancers.

^bEstimated from Cox proportional hazards models (SAS PHREG; SAS Institute, Inc., Cary, North Carolina) that used attained age as the time scale. Separate models were fit to nulliparous women, parous women aged <25 years at first birth, and parous women aged ≥25 years at first birth. Models include all variables presented in the table as well as birth year (before 1927, 1927–1932, 1933–1938, 1939 or later), calendar year of entry (before 1987, 1987–1995, 1996 or later), oral contraceptive use (ever, never, missing), and study. Models include missing category for each variable (data not shown).

^cFurther adjusted for number of livebirths (<3, ≥3, missing).

^dP_{heterogeneity} between nulliparous women and parous women aged <25 years at first birth based on a Wald test.

^eP_{heterogeneity} between nulliparous women and parous women aged ≥25 years at first birth based on a Wald test.

^fAge at menarche groups: NIH-AARP and BCDDP—early, ≤12 years; middle, 13–14 years; late, ≥15 years; PLCO and USRT—early, ≤11 years; middle, 12–13 years; late, ≥14 years.

^gIncludes hysterectomy without bilateral oophorectomy. This category not included in heterogeneity tests.

^hFollow-up censored at June 30, 2002.

ⁱIncludes <6 years for PLCO.

^jIncludes ≥6 years for PLCO.

^kEstimates and case numbers presented for NIH-AARP, PLCO, and USRT.