The combined effect of mammographic texture and density on breast cancer risk: a cohort study

Breast Cancer Research

Table 3 The association between breast measures and breast cancer risk

Variables in the model		HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)	p value for trend	c-index (95% CI)
		per one SD	Q2	Q3	Q4
Model 1	DV	1.32 (1.18–1.48)	1.24 (0.87–1.78)	1.53 (1.08–2016)	1.85 (1.32–2.59)	<0.001	0.56 (0.53–0.59)
Model 1a	DV	1.32 (1.18–1.47)	1.40 (0.97–2.01)	1.75 (1.23–2.48)	1.98 (1.41–2.79)	<0.001	0.62 (0.58–0.65)
Model 1a	Texture residuals (DV) ^a	1.38 (1.23–1.56)	1.68 (1.15–2.44)	2.40 (1.68–3.43)	2.69 (1.87–3.88)	<0.001	0.62 (0.58–0.65)
Model 2	PDV	1.34 (1.20–1.50)	1.49 (1.03–2.15)	2.07 (1.46–2.96)	2.17 (1.51–3.12)	<0.001	0.58 (0.54–0.61)
Model 2a	PDV	1.36 (1.21–1.53)	1.50 (1.04–2.17)	2.00 (1.41–2.86)	2.15 (1.49–3.10)	<0.001	0.60 (0.57–0.63)
Model 2a	Texture residuals (PDV) ^b	1.27 (1.13–1.42)	1.28 (0.90–1.82)	1.69 (1.21–2.37)	1.92 (1.37–2.70)	<0.001	0.60 (0.57–0.63)
Model 3	Texture	1.46 (1.30–1.64)	1.69 (1.15–2.50)	2.65 (1.83–3.84)	3.16 (2.16–4.62)	<0.001	0.61 (0.57–0.64)

Difference c-index model 1 and 1a, p < 0.001; difference c-index model 2 and 2a, p = 0.054
SD standard deviation, Q quartile, DV dense volume, PDV percentage dense volume
^aTexture residuals (DV): residuals of texture pattern scores regressed on natural logarithm (Ln) transformed DV using a linear regression model
^bTexture residuals (PDV): residuals of texture pattern scores regressed on Ln transformed PDV using a linear regression model

ISSN: 1465-542X