Correlates of circulating ovarian cancer early detection markers and their contribution to discrimination of early detection models: results from the EPIC cohort

Table 3 Multivariate adjusted associations between epidemiologic characteristics and CA125, CA15.3, and HE4 among controls postmenopausal at blood collection: EPIC^a

	CA125		CA15.3		HE4^b
	beta	p	beta	p	beta	p
Age at blood draw (continuous)					0.016	<0.0001
Duration of oral contraceptive use (continuous)					−0.007	0.009
Parity (continuous)^c	0.030	0.03			−0.036	0.03
Hysterectomy	−0.053	0.34
Uniliateral oophorectomy	−0.104	0.18
Age at menopause (continuous)	0.019	0.0009			−0.015	0.03
Type of HT (E alone vs. never)	−0.120	0.17
Ovulatory cycles^{c, d}(continuous)	−0.0001	0.76
BMI (kg/m²; continuous)			0.013	0.003
Smoking
Former vs never	0.006	0.87	0.081	0.03	−0.003	0.95
Current vs never	−0.164	<0.0001	−0.018	0.66	0.385	<0.0001

^a1 control missing CA125 and 8 missing CA153; ^b Restricted to 538 controls; ^cmodel was run once including parity and excluding ovulatory cycles, then run again excluding parity and including ovulatory cycles. With the exception of ovulatory cycles, the betas and p-values shown are for the model including parity; ^d time between menarche and menopause with time subtracted for oral contraceptive use, pregnancy and breastfeeding. Note: Italicized p values indicate statistically significant associations

ISSN: 1757-2215