Table 3. Association between cigarette smoking variables and breast cancer risk in post-menopausal nonobese and obese women, multi-case-control study (MCC-Spain), 2008–2013, Spain (N=2191)

Smoking variables	Under/normal weight (N=908)				Overweight/obese (N=1283)				p-int c
	Ca	Co	AOR b	95% CI	Ca	Co	AOR b	95% CI
Smoking status (1 year before the interview)
Never smoker	214	305	1.00		447	430	1.00
Former ≥10	46	55	1.00	0.57–1.75	58	65	0.55	0.34–0.90
Former <10	37	52	0.99	0.53–1.86	47	49	0.64	0.35–1.16
Active smoker	85	114	0.89	0.55–1.45	101	81	0.77	0.49–1.19	0.534
p-trend				0.663				0.107
Age at smoking initiation (years)
Never smoker	214	305	1.00		447	430	1.00
≥18	86	128	0.89	0.57–1.38	131	124	0.67	0.46–0.97
<18	80	92	1.01	0.60–1.70	78	70	0.66	0.40–1.09	0.554
p-trend				0.932				0.031
Duration of tobacco use (years)
Never smoker	214	305	1.00		447	430	1.00
<20	40	40	1.02	0.54–1.93	51	52	0.67	0.40–1.12
20–30	43	42	1.57	0.83–2.96	54	44	0.64	0.36–1.15
>30	83	138	0.75	0.47–1.18	100	98	0.68	0.44–1.03	0.281
p-trend				0.360				0.036
Intensity (cigarettes/day)
Never smoker	214	305	1.00		447	430	1.00
<15	82	105	0.88	0.55–1.42	85	87	0.59	0.39–0.91
≥15	81	105	1.05	0.65–1.70	118	101	0.74	0.49–1.13	0.427
p-trend				0.917				0.065
Cumulative exposure (pack-years)
Never smoker	214	305	1.00		447	430	1.00
<10	55	63	0.90	0.51–1.58	52	58	0.56	0.34–0.92
10–25	40	65	0.77	0.42–1.40	54	58	0.53	0.31–0.92
>25	67	82	1.14	0.69–1.90	92	71	0.92	0.58–1.48	0.611
p-trend				0.836				0.202
Years smoking before first birtha
Never smoker	168	250	1.00		397	379	1.00
<10	63	82	0.81	0.46–1.41	73	70	0.56	0.33–0.95
≥10	49	70	0.98	0.51–1.87	52	50	0.76	0.43–1.34	0.533
p-trend				0.773				0.131
Cigarettes per day before first birtha
Never smoker	168	250	1.00		397	379	1.00
<15	42	69	0.74	0.40–1.37	50	52	0.54	0.31–0.96
≥15	44	65	1.01	0.55–1.87	54	50	0.73	0.40–1.35	0.695
p-trend				0.849				0.118

AOR: adjusted odds ratio. Multivariable logistic regression analyses were used to test whether each tobacco variable was associated with breast cancer risk. The p-trend values were calculated by incorporating the categorized variable as a continuous variable in the multivariable models. The p-value of the interaction term between weight status and the corresponding variable was calculated using the likelihood ratio test.

^aOnly parous women.

^bAOR and 95% CI adjusted for age, education level, region, previous biopsies, family history of breast cancer, age at first birth, number of children, alcohol use, history of hormone replacement therapy and age at menopause.

^cp-int.: p-value of the interaction term.