The influence of mammogram acquisition on the mammographic density and breast cancer association in the mayo mammography health study cohort

Table 2 Association between mammographic density estimates on the enrollment (screen-film) mammograms and risk of breast cancer in the Mayo Mammography Health Study cohort (2003 to 2009)

	Number of cases (total n= 272)^a	Number of person-years (total n= 9,958.8)^{b, c}	Multivariate adjusted hazard ratio (95% CI)^c
Percent density
0.0-5.0	43	2,193.8	1.00 (Ref.)
5.1-15.0	79	2,746.1	2.02 (1.37, 2.97)
15.1-25.0	59	2,386.4	2.03 (1.32, 3.11)
25.1+	91	2,632.5	3.80 (2.46, 5.86)
P trend			< 0.001
BI-RADS lexicon
1: Almost entirely fat	44	2,181.6	1.00 (Ref.)
2: Scattered fibroglandular densities	111	3,904.5	1.61 (1.13, 2.29)
3: Heterogeneously dense	94	3,210.1	2.02 (1.38, 2.95)
4: Extremely dense	23	662.7	2.96 (1.73, 5.07)
P trend			< 0.001

Analyses adjusted for age, menopausal status, post-menopausal hormone use, and body mass index. ^aForty-six cases were excluded from the original 318 cases for the following reasons: digital mammogram (13), implants (12), and inability to read density (21). Percent density and BI-RADS (Breast Imaging Reporting and Data System) analyses were restricted to the same women. ^bThree hundred two in the subcohort were excluded for the following reasons: digital mammogram (123), implants (62), and inability to read density (117). Percent density and BI-RADS analyses were restricted to the same women (272 cases and 1,957 in the subcohort).

ISSN: 1465-542X