Table 1 Odds ratios corresponding to quartiles defined using the internal validation ovarian cancer dataset.

From: The DNA methylome of cervical cells can predict the presence of ovarian cancer

Quartile

Control (N = 297)

Internal validation ovarian cancer (N = 83)

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

(−2.38, −0.57)

75

3

1.00 (reference)

1.00 (reference)

(−0.57, −0.21)

74

10

3.25 (0.93,15.68)

2.38 (0.56, 12.99)

(−0.21, 0.17)

74

13

4.20 (1.27,19.79)

3.65 (0.87, 19.42)

(0.17, 2.21)

74

57

18.20 (6.33,79.95)

10.26 (2.89, 49.1)

Quartile

Control (N=225)

External validation ovarian cancer (N=47)

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

(−2.38, −0.57)

59

2

1.00 (reference)

1.00 (reference)

(−0.57, −0.21)

57

6

2.94 (0.62,22.94)

4.88 (0.85, 41.76)

(−0.21, 0.17)

49

6

3.42 (0.72,26.71)

4.57 (0.77, 40.12)

(0.17, 2.21)

60

33

14.99 (4.26,103.18)

26.25 (5.89, 194.92)

Quartile

Control (N=297)

Endometrial cancer (N=217)

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

(−2.38, −0.57)

75

4

1.00 (reference)

1.00 (reference)

(−0.57, −0.21)

74

6

1.50 (0.4,6.32)

0.72 (0.17, 3.26)

(−0.21, 0.17)

74

15

3.68 (1.25,13.77)

0.92 (0.24, 3.92)

(0.17, 2.21)

74

192

46.44 (18.41,159.14)

11.20 (3.91, 40.51)

Quartile

Control (N= 297)

Breast cancer (N = 329)

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

(−2.38, −0.57)

75

28

1.00 (reference)

1.00 (reference)

(−0.57, −0.21)

74

49

1.77 (1.01,3.14)

1.50 (0.83, 2.76)

(−0.21, 0.17)

74

91

3.27 (1.94,5.64)

2.30 (1.29, 4.14)

(0.17, 2.21)

74

161

5.78 (3.49,9.8)

5.27 (2.91, 9.78)

Quartile

Control (N= 114)

BRCA1 cases (N= 87)

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

(−2.38, −0.57)

41

12

1.00 (reference)

1.00 (reference)

(−0.57, −0.21)

26

15

1.95 (0.79,4.95)

1.66 (0.64, 4.31)

(−0.21, 0.17)

31

12

1.32 (0.51,3.39)

1.20 (0.42, 3.4)

(0.17, 2.21)

16

18

3.76 (1.49,9.88)

2.59 (0.84, 8.08)

  1. Adjustment was based on a logistic regression model with age, menopausal status, age at menarche, number of first degree relatives with ovarian cancer, and BMI included as covariates for the ovarian cancer datasets. For endometrial cancers and the BRCA1 dataset age and menopause were included as covariates. For the endometrial cancers adjusted estimates are unavailable as the logistic regression model failed to converge. In addition, it was assumed that there was 1 cancer case in the first quartile in order to estimate ORs for the remaining quartiles.