Table 2. Relationship between tumor p53 status and annual household income in the adjusted analysis.

	Adjusted logistic regression** OR (95% CI) (n  =  143)	OR (95% CI) additionally adjusted for smoking (n  =  136)	OR (95% CI) further adjusted for age, education, and BMI (n  =  131)
Odds of acquiring a p53 mutation with increasing household income*	1	1	1
	0.42 (0.18 to 0.97)	0.4 (0.17 to 0.94)	0.32 (0.10 to 0.99)
Odds of acquiring a p53 IHC-positive tumor with increasing household income*
	1	1	1
	0.57 (0.30 to 1.08)	0.63 (0.33 to 1.21)	0.88 (0.40 to 1.97)

OR  =  odds ratio; CI  =  confidence interval; IHC  =  immunohistochemistry

^*Trend test. Shown is the OR for the stepwise increase in household income (reference: low income). Income coded as 0 (< $15,000), 1 ($15,000 to $60,000), and 2 (> $60,000); adjustments: smoking (pack years), age, and body mass index (BMI) were used as continuous data; other covariates were dichotomized for the analysis, as shown in Table 1

^**adjusted for race/ethnicity, node status, tumor estrogen receptor status, and tumor grade