Cytochrome P450 1A2 (CYP1A2) activity, mammographic density, and oxidative stress: a cross-sectional study

Breast Cancer Research

Table 2 CYP1A2 activity and mammographic density

				Quartiles of CMR: mean percentage density (95% confidence interval)
Model adjustments	CMR: β (SE)¹	F	P	Q1	Q2	Q3	Q4	P for trend
Premenopausal (n = 146)³
Age, ethnicity, smoking²	0.29 (0.52)	0.31	0.58	26.7 (14.5–42.6)	28.4 (16.9–42.8)	30.7 (19.0–45.2)	29.9 (17.9–45.0)	0.57
+ total estrogen	0.36 (0.52)	0.48	0.49	26.1 (13.8–42.3)	28.0 (16.5–42.5)	30.2 (18.4–44.8)	30.3 (18.0–45.8)	0.49
+ other covariates	-0.16 (0.38)	0.17	0.68	20.1 (10.3–33.1)	20.9 (11.7–32.8)	21.2 (12.4–32.2)	18.5 (10.0–29.6)	0.72
Postmenopausal (n = 149)⁴
Age, ethnicity, smoking	1.57 (0.60)	6.89	0.01	14.1 (6.6–24.3)	30.2 (18.4–45.1)	24.8 (14.7–37.5)	31.0 (20.9–43.2)	0.01
+ total estrogen	1.30 (0.64)	4.05	0.05	12.8 (5.6–22.9)	27.2 (15.7–41.7)	22.5 (12.9–34.8)	27.4 (17.6–39.4)	0.03
+ other covariates	0.94 (0.61)	2.36	0.13	14.6 (5.7–27.6)	25.0 (13.0–41.0)	19.0 (8.6–33.5)	28.2 (15.8–44.1)	0.05

^1β associated with log transformed caffeine metabolic ratio (CMR) as the independent variable and square-root transformed percentage breast density as the dependent variable. ²Statistical models are initially adjusted for age, ethnicity, and smoking status. Models are subsequently adjusted for total estrogen, and then for other covariates, which include adjustments for age at menarche, age at menopause (postmenopausal only), number of live births, body mass index (BMI), waist–hip ratio (WHR), and family history of breast cancer. ³ n = 144 for estrogen adjusted analysis and n = 143 for covariate adjusted analysis. ⁴ n = 127 for estrogen adjusted analysis and n = 118 for covariate adjusted analysis.

ISSN: 1465-542X