Table 2.

Univariate models of the probability of screen-detected or interval invasive, or DCIS, second breast cancer (no second cancer as referent) within 1 year of screening (based on 13,958 screening mammograms) in PHBC women with history of DCIS (BCSC 1996–2008)

Variable (proportion missing data for variable where applicable)Screen-detected invasive second breast cancer OR (95% CI)Interval invasive breast cancer OR (95% CI)DCIS second breast cancer OR (95% CI)Pa
Age at mammography, y
 <501.98 (0.79–4.96)4.68 (1.7112.77)2.10 (0.99–4.45)0.0101
 50—591.79 (0.86–3.75)1.44 (0.51–4.04)1.38 (0.71–2.67)
 60—69ReferentReferentReferent
 70—792.52 (1.245.14)2.66 (1.026.94)1.00 (0.47–2.10)
 80+2.56 (1.125.84)0.78 (0.16–3.90)1.32 (0.55–3.13)
Race/ethnicity (3.5%)
 White, non-HispanicReferentReferentReferent0.74
 Black, non-Hispanic1.25 (0.28–5.51)1.54 (0.33–7.23)0.60 (0.08–4.64)
 Hispanic1.03 (0.43–2.42)0.75 (0.16–3.48)0.96 (0.27–3.42)
 Asian, Pacific Islander0.52 (0.07–3.82)2.55 (0.57–11.37)2.70 (0.92–7.87)
First-degree family history of breast cancer (17.6%)
 NoReferentReferentReferent0.021
 Yes2.15 (1.273.64)1.29 (0.62–2.70)1.37 (0.78–2.38)
Menopausal status (9.5%)
 PostReferentReferentReferent0.0004
 Pre, Peri1.56 (0.76–3.21)2.36 (1.065.28)2.89 (1.615.18)
BI-RADS breast density (19.6%)
 1—almost entirely fatty0.73 (0.17–3.17)1.35 (0.39–4.76)0.33 (0.04–2.49)0.60
 2—scattered fibroglandular tissuereferentReferentReferent
 3–heterogeneously dense1.42 (0.77–2.59)1.09 (0.52–2.29)1.60 (0.90–2.82)
 4—extremely dense1.31 (0.44–3.91)0.90 (0.20–4.00)1.99 (0.79–5.01)
BMI (36.6%)b
 Normal (18.5–24.9)ReferentReferentReferent0.66
 Overweight(25–29.9)1.58 (0.87–2.88)0.89 (0.37–2.14)1.35 (0.68–2.69)
 Obese I (30–34.9)1.87 (0.87–4.00)1.43 (0.51–4.05)1.64 (0.70–3.82)
 Obese II–III (35+)2.20 (0.88–5.52)1.11 (0.25–4.98)1.14 (0.33–3.93)
Time since last mammogram (1.7%)
 9–14 moReferentReferentReferent0.58
 15–23 mo0.96 (0.49–1.89)0.75 (0.26–2.13)0.95 (0.45–2.01)
 24+ mo0.80 (0.25–2.57)2.23 (0.77–6.45)1.90 (0.75–4.82)
Type of mammogram (0.02%)
 Film screenReferentReferentReferent0.73
 Digital0.79 (0.31–2.05)0.54 (0.15–1.93)1.16 (0.55–2.42)
Time since first breast cancer diagnosis
 <1 y (6–11 mo)1.22 (0.34–4.47)0.64 (0.14–2.89)1.43 (0.51–4.04)0.66
 1–2 yReferentReferentReferent
 3–4 y1.53 (0.70–3.35)0.76 (0.33–1.76)1.35 (0.66–2.74)
 5–6 y2.62 (1.245.53)0.80 (0.32–1.95)1.32 (0.62–2.82)
 7–9 y1.66 (0.73–3.76)0.67 (0.25–1.81)1.05 (0.46–2.42)
 ≥10 y1.04 (0.41–2.61)0.51 (0.16–1.66)1.28 (0.54–3.02)
Mode of detection of first cancer (41.1%)
 Screen-detectedReferentReferentReferent0.67
 Interval cancer in screening1.64 (0.68–3.94)1.16 (0.27–5.07)0.89 (0.27–2.91)
 Clinical/diagnostic detected1.38 (0.48–3.92)2.46 (0.81–7.45)1.61 (0.62–4.17)
 Other0.81 (0.25–2.68)2.26 (0.75–6.85)1.26 (0.44–3.62)
Age at first breast cancer, y
 <401.16 (0.26–5.14)2.35 (0.65–8.49)2.28 (0.64–8.11)0.17
 40—491.68 (0.86–3.30)1.14 (0.52–2.52)1.75 (0.88–3.48)
 50—590.93 (0.44–1.96)0.60 (0.24–1.47)1.48 (0.74–2.94)
 60—69ReferentReferentReferent
 70—792.36 (1.194.66)0.64 (0.22–1.82)1.21 (0.53–2.78)
 80+1.51 (0.43–5.32)0.55 (0.07–4.25)0.54 (0.07–4.18)
Primary surgery (7.7%)
 MastectomyReferentReferentReferent<.0001
 Breast conserving with radiation2.37 (1.154.88)4.44 (1.4913.25)2.74 (1.385.45)
 Breast conserving without radiation3.98 (1.928.25)6.30 (2.0919.01)2.68 (1.295.57)
Radiation therapy (12.5%)
 NoneReferentReferentReferent0.15
 Any0.59 (0.360.98)0.70 (0.36–1.36)1.09 (0.64–1.85)
Adjuvant systemic therapy (13.7%)
 NoneReferentReferentReferent0.035
 Endocrine therapyc0.33 (0.10–1.04)1.74 (0.79–3.87)0.40 (0.14–1.10)
Self-reported history of breast implant, reduction, or reconstruction (27.0%)b
 NoRefRefRef0.89
 Yes0.58 (0.13–2.53)1.17 (0.15–9.09)0.82 (0.18–3.67)

NOTE: Each univariate model included the variable of interest and was adjusted for mammography registry and primary surgery received for first breast cancer. OR and P values shown in bold indicate statistically significant association.

  • aP is the global P value for each univariate analysis in which joint modeling was conducted for the 3 outcomes (screen-detected or interval invasive, or DCIS, second breast cancer) with no cancer as the referent category.

  • bOne site was removed due to high proportion of missing values for variable at that site (for BMI, 141 women who were reportedly “underweight” were excluded).

  • cIncludes 19 women with a prior history of DCIS who reportedly had chemotherapy (12 with chemotherapy only; 7 with both endocrine and chemotherapy).