Table 4.

HR estimates and 95% CIs from Cox regression analysis of breast cancer–specific mortality, for the full baseline model versus the baseline model + specific domain variables, for the subsets of cases for which domain data were available

Contextual factors subseta (N = 9,635)Physical activity subset (N = 4,608)bBody size subset (N = 11,021)cComorbidity subset (N = 10,106)d
Racial/ethnic groupBaseline modeleBaseline + domain variablesfBaseline modeleBaseline + domain variablesfBaseline modeleBaseline + domain variablesfBaseline modeleBaseline + domain variablesf
NL White1 (Ref)1 (Ref)1 (Ref)1 (Ref)1 (Ref)1 (Ref)1 (Ref)1 (Ref)
African American1.17 (0.98–1.40)1.17 (0.98–1.40)1.27 (1.03–1.57)1.31 (1.06–1.61)1.10 (0.93–1.30)1.07 (0.90–1.27)1.21 (1.01–1.46)1.20 (1.00–1.44)
Latina0.90 (0.74–1.09)0.89 (0.73–1.09)0.91 (0.72–1.17)0.95 (0.74–1.22)0.77 (0.63–0.93)0.76 (0.62–0.92)0.81 (0.65–1.00)0.77 (0.62–0.96)
Asian0.73 (0.43–1.23)0.72 (0.43–1.23)(omitted)(omitted)0.54 (0.32–0.90)0.55 (0.32–0.93)0.61 (0.30–1.24)0.63 (0.31–1.27)
Disparity (D)0.140.140.120.120.230.220.200.19
% Change in D<±1%<±1%−3.70%−2.30%
Overall χ2(df)g31.8 (27)8.96 (7)3.22 (2)20.2 (6)
Pg0.240.250.200.003
Disparity χ2(df)h0.0 (3)0.120.07(3)0.16 (3)
Ph>0.95i>0.95>0.95
  • Abbreviation: df, degrees of freedom.

  • aLACE excluded.

  • bPhysical activity: CTS, LACE, MEC excluded.

  • cBody size: none excluded.

  • dComorbidity: MEC excluded.

  • eModel including racial/ethnic group and all other variables in Supplementary Table S1, stratified by study and adjusted for age at diagnosis. Analysis excludes cases from studies that did not collect variables relevant to the specific domain. In addition, some cases with missing data relevant to the domain are also excluded.

  • fIncludes all variables in the baseline model plus domain-specific variables.

  • gCox partial likelihood ratio χ2comparing left model to right model in each section.

  • hPortion of the likelihood ratio χ2 attributable to change in disparity measure (see Materials and Methods).

  • iCould not be computed because of parameter aliasing.