Table 3.

Odds ratios (95% confidence intervals) of breast cancer by quintile of prediagnostic plasma anti-Mullerian hormone (pg/mL), Nurses' Health Study and Nurses' Health Study II

	Q1	Q2	Q3	Q4	Q5	p- trend
N cases/controls	102/94	99/94	81/94	124/94	133/95
Cutpoints (pg/mL)	<225	225–<575	575–<1053	1053–<2213	≥2213
Simple* OR (95% CI)	1.00 (ref)	1.41 (0.91–2.20)	1.24 (0.78–1.97)	2.06 (1.29–3.29)	2.38 (1.46–3.88)	0.0004
Multivariate^ OR (95% CI)	1.00 (ref)	1.29 (0.82–2.03)	1.17 (0.73–1.88)	1.92 (1.19–3.10)	2.20 (1.34–3.63)	0.001
							p-het
Invasive (N=369 cases)	1.00 (ref)	1.29 (0.80–2.10)	0.93 (0.55–1.56)	1.71 (1.01–2.89)	2.30 (1.34–3.95)	0.001
In situ (N=150 cases)	1.00 (ref)	1.49 (0.69–3.20)	1.54 (0.71–3.32)	2.59 (1.19–5.62)	2.08 (0.90–4.82)	0.21	0.27
ER+ (N=292 cases)	1.00 (ref)	1.55 (0.91–2.64)	1.18 (0.67–2.08)	2.13 (1.21–3.77)	2.78 (1.54–5.02)	0.001
ER− (N=64 cases)	1.00 (ref)	0.89 (0.37–2.17)	0.41 (0.14–1.18)	0.99 (0.37–2.68)	1.62 (0.57–4.58)	0.14	0.33
Triple negative** (N=30 cases)	1.00 (ref)	1.02 (0.26–3.99)	0.43 (0.08–2.44)	2.16 (0.54–8.59)	2.05 (0.45–9.29)	0.20

^*Simple model adjusted for matching factors: age at blood (continuous), year of blood (1989, 1990, 1996, 1997, 1998, 1999), time of blood (1–8am, 9am-noon, 1pm-midnight), fasting (yes/no), luteal day (3–5, 6–7, 8–9, 10–28, untimed), race (Caucasian, other); and age at blood squared (continuous)

^{^}Multivariate model additionally adjusted for the following variables (categories): age at menarche (<12, 12, 13, ≥14y), parity/age at first birth (nulliparous, 1–2 children/<25y, 1–2 children/≥25y, ≥3 children/<25y, ≥3 children/≥25y), BMI at age 18 (<21, 21–<23, ≥23), oral contraceptive use (ever/never), family history of breast cancer (yes/no), history of benign breast disease (yes/no)

^**Triple negative model adjusts for reduced set of matching factors only: age at blood (continuous), age at blood squared (continuous), luteal day (3–5, 6–7, 8–9, 10–28, untimed)