. Author manuscript; available in PMC: 2012 May 1.

Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2011 Apr 5;20(5):934–938. doi: 10.1158/1055-9965.EPI-11-0138

Table 2.

Multivariate hazard ratios (HR) and 95% confidence intervals (CI) for risk of epithelial ovarian cancer by years of rotating night shift work in NHS and NHSII

	NHS				NHSII				NHS and NHSII
	Cases	Person- years	Age-adjusted HR	Multivariate^a HR (95% CI)	Cases	Person- years	Age-adjusted HR	Multivariate^a HR (95% CI)	Cases	Person- years	Age-adjusted HR	Multivariate^a HR (95% CI)	p-het^b

Years of rotating shift work
none	202	454,004	1.00	1.00 (ref.)	68	602,593	1.00	1.00 (ref.)	270	1,056,596	1.00	1.00 (ref.)
1-2	143	270,522	1.20	1.20 (0.97-1.49)	54	591,716	0.82	0.80 (0.56-1.14)	197	862,237	1.08	1.07 (0.89-1.29)	0.06
3-5	80	183,352	0.96	0.95 (0.73-1.23)	35	386,364	0.82	0.79 (0.52-1.18)	115	569,716	0.93	0.90 (0.72-1.13)	0.46
6-9	33	74,671	0.98	0.96 (0.67-1.40)	18	178,426	0.89	0.80 (0.47-1.35)	51	253,098	0.96	0.92 (0.68-1.25)	0.55
10-14	25	48,771	1.12	1.06 (0.70-1.62)	14	76,496	1.42	1.25 (0.70-2.24)	39	125,268	1.22	1.14 (0.81-1.60)	0.65
15-19	19	29,849	1.33	1.30 (0.81-2.10)	5	24,001	1.41	1.21 (0.48-3.02)	24	53,850	1.34	1.28 (0.84-1.94)	0.90
20+	22	47,034	0.90	0.88 (0.56-1.37)	0	6,873	NA	NA	22	53,907	0.84	0.80 (0.51-1.23)	NA
p-trend^c			0.97	0.84			0.77	0.78			0.88	0.74	0.87

Multivariate analyses adjusted for age (continuous), duration of oral contraceptive use (continuous), parity (continuous), BMI (continuous), smoking status (current/past/never/missing), tubal ligation history (yes/no), menopausal status (postmenopausal/premenopausal or unknown), family history of ovarian cancer (yes/no), duration of breastfeeding (none/≤6 months/7-11 months/12-17 months/18+ months/missing), and cohort (combined analyses only).

P-heterogeneity by cohort was assessed using the DerSimonian and Laird random effects model.

P-value for multivariate model including duration of shift work as a continuous variable.