Table 3.

Relative risks and 95 % CIs for smoking status and endometrial carcinoma risk by tumor characteristics

Tumor characteristicsa	Smoking status						pheterogeneity
	Never (n = 49,768)		Former (n = 41,824)		Current (n = 18,712)
	Cases	RR (95 % CI)b	Cases	RR (95 % CI)b	Cases	RR (95 % CI)b
Histology
Type I	653	1 (ref)	490	0.92 (0.82, 1.03)	127	0.68 (0.56, 0.82)	0.15c
Low-grade endometrioid	333	1 (ref)	244	0.90 (0.76, 1.06)	56	0.61 (0.46, 0.82)	0.34d
High-grade endometrioid	42	1 (ref)	36	1.03 (0.66, 1.61)	12	1.04 (0.54, 2.00)
Type II	63	1 (ref)	36	0.74 (0.49, 1.12)	8	0.41 (0.20, 0.87)
Stage							0.92e
Localized	417	1 (ref)	288	0.85 (0.73, 0.98)	72	0.59 (0.46, 0.76)
Regional/distant	111	1 (ref)	78	0.91 (0.68, 1.21)	17	0.53 (0.32, 0.89)

^aAssociations were evaluated for subgroups defined by histology/grade and SEER summary stage using models to predict risk for one subgroup. Cases of other subgroups were censored at the time of diagnosis

^bAdjusted for race (white, non-white), age at menarche (≤12, 13–14, ≥15), BMI (normal, overweight, obese), parity (nulliparous, 1–2, ≥3), oral contraceptive use (never, ever), age at menopause (premenopausal,<45, 45–49, 50–54, ≥55), menopausal hormone use (never, current, former), diabetes (never, ever), physical activity (never/rarely, 1–3 times/month, 1–2 times/week, 3–4 times/week, ≥5 times/week). Unknown was set as a separate category within each factor

^cp_{heterogeneity} is from logistic regression of case-only analysis comparing smoking status between Type I and Type II endometrial cancer cases

^dp_{heterogeneity} is from logistic regression of case-only analysis comparing smoking status between low-grade endometrioid and high-grade endometrial cancer cases

^ep_{heterogeneity} is from logistic regression of case-only analysis comparing smoking status between localized and regional/distant endometrial cancer cases