Table 3.

Logistic regression models for ambient air pollutants impact on mammographic breast density considering confounder variables.

		P-value	OR	95 C.I. for OR
				Lower	Upper
Model 1	CO	0.001	0.331	0.172	0.637
	NO2	0.003	1.039	1.013	1.066
	Constant	<0.001	1.758
Model 2	CO	0.020	0.411	0.195	0.868
	NO2	0.026	1.034	1.004	1.064
	Age	0.000	0.942	0.915	0.971
	BMI	0.000	0.884	0.851	0.919
	Smoking	0.063	0.602	0.353	1.029
	History of OCP	0.000	0.526	0.375	0.738
	Menopause	0.002	0.522	0.348	0.782
	Parity	0.708	0.975	0.856	1.111
	History of Breast disease	0.106	0.728	0.495	1.070
	Constant	<0.001	2454.3
Model 3	CO	0.018	0.404	0.190	0.856
	NO2	0.022	1.035	1.005	1.066
	Age	0.000	0.947	0.919	0.976
	BMI	0.000	0.884	0.851	0.920
	Smoking	0.088	0.624	0.363	1.072
	History of OCP	0.000	0.536	0.381	0.753
	Menopause	0.004	0.544	0.360	0.823
	Parity	0.938	0.995	0.872	1.135
	History of Breast disease	0.124	0.735	0.496	1.088
	Metformin	0.726	0.913	0.547	1.524
	Aspirin	0.070	0.641	0.397	1.037
	Vitamin D	0.040	1.426	1.016	2.001
	Calcium	0.203	0.795	0.558	1.132
	Constant	<0.001	1710.8

Model 1, only significant pollutants, Model 2, significant pollutants with medical (history of breast disease, menopause statues, history of OCP use, and parity) and demographic (age, BMI, and smoking) risk factors, Model 3, significant pollutants with medical and demographic risk factors also with medicine and supplement use (metformin, aspirin, vitamin D, calcium). All models were non stepwise multivariable logistic regression. OR = Odds ratio, C.I = Confidence interval, NO₂ = Nitrogen dioxide (NO₂), SO₂ = Sulfur dioxide, CO = Carbon monoxide, O₃ = Ozone, PM = Particulate matter.