Table 2.

Continuous and Cut Point Analyses of Prostate Cancer Risk with Decreased CAG-Repeat Length^[Note]

		Findings for CAG-Repeat Length
		Continuous			<22 versus ⩾22			<23 versus ⩾23
Subjects	No. of Controls/Cases	ORa	95% CI	Pb	ORa	95% CI	Pb	ORa	95% CI	Pb
All cases	2,160/2,036	1.016	.997–1.036	.11	1.084	.954–1.231	.22	1.115	.981–1.267	.09
Advanced cases	2,160/686	1.015	.988–1.044	.28	1.094	.913–1.310	.33	1.122	.936–1.345	.21
Local cases	2,160/1,239	1.013	.991–1.036	.25	1.068	.922–1.238	.38	1.087	.937–1.261	.27

Note.— There is no statistically significant association between the CAG-microsatellite polymorphism and prostate cancer risk for the entire population or when stratified by stage. Multiple analyses were performed, by using two cut points (22 and 23) as well as by modeling the CAG-repeat length as a continuous variable. These analyses demonstrate no significant association between prostate cancer risk and CAG-repeat length.

^aAdjusted for race/ethnicity and age.

^bTwo-sided P value.