# Table 2 Odds Ratios of breast cancer risk in terms of A) pregnancy and breastfeeding (either status or duration) and B) breastfeeding and number of children before and after adjusting for the effect of each other and all other risk factors in multivariable logistic models

OR (95%CI) p-value OR (95%CI) p-value OR (95%CI) p-value
A. Pregnancy
No 1.00
Yes 0.69 (0.49,0.97) 0.03 0.89 (0.62,1.28) 0.54 1.07 (0.66,1.72) 0.79
Breastfeeding 1
No 1.00
Yes 0.69 (0.57,0.82) <0.001 0.63 (0.45,0.89) <0.01 0.72 (0.59,0.90) <0.01
Breastfeeding by duration 1, 2
Never 1.00
≤6 months 0.71 (0.58,0.88)   0.65 (0.45,0.94)   0.73 (0.57,0.93)
7-12 months 0.76 (0.58,0.98)   0.69 (0.47, 1.03)   0.86 (0.64,1.16)
>1 year 0.61 (0.48,0.77) <0.0013 0.56 (0.38,0.81) <0.0013 0.68 (0.51,0.91) 0.043
B. Breastfeeding
No 1.00
Yes 0.69 (0.57,0.83) 0.03 0.72 (0.58,0.88) <0.01 0.72 (0.58,0.90) <0.01
Number of children
None 1.00
One 0.90 (0.59,1.37)   1.06 (0.69,1.64)   1.29 (0.76,2.18)
Two 0.84 (0.61,1.15)   1.04 (0.74,1.48)   1.27 (0.76,2.18)
Three 0.68 (0.49,0.94) <0.013 0.90 (0.62,1.29) 0.253 1.14 (0.82,1.99) 0.87
1. Notes: 1In the adjusted models, odds ratio refers to the combined effect of pregnancy and breastfeeding, status or duration respectively, as estimated in a model with an interaction term between the two variables but no main effect for breastfeeding. 2p-values of Likelihood Ratio Test comparing the goodness of fit of model including breastfeeding duration Vs simpler model with breastfeeding expressed as binary variable = 0.27, 0.19 and 0.33 respectively, indicating no better fit of full model. 3P-value for trend across ordered categories. 4Due to collinearity, the variables pregnancy and number of children were mutually excluded from the fully adjusted models.