Purpose of Study: We used multiple mediation analysis to examine why African-American women have poorer ovarian cancer survival compared to white women.

Methods: We examined data from the OCWAA consortium which harmonized questionnaire-based data on 1,075 African-American women and 3,272 white women with ovarian cancer from seven U.S. studies. We first selected potential mediators and confounders by examining their association with race and survival and then incorporated these variables into a sequential multiple mediation analysis. We used multiple imputation and bootstrapping to fit and pool natural effects models with counterfactual exposures and survival times. In order to distinguish the effect of race through mediators and otherwise, we implemented log-normal accelerated failure time and Cox proportional hazards (PH) models.

Results: In OCWAA, 67.8% (n=729) of African-American women and 69.7% (n=2,282) of white women were deceased. The hazard ratio for African-American compared to white women was 1.29 (95%CI=1.15-1.46). In our final model (Table 1), mediators of this disparity in survival included college education, nulliparity, smoking status, body mass index, diabetes, diabetes/race interaction, postmenopausal hormone (PMH) therapy duration, PMH duration/race interaction, PMH duration/age interaction, histotype, and stage. These mediators explained 46.4% (95%CI=30.9-60.0%) of the overall racial disparity, and histotype/stage and PMH duration accounted for the largest fraction.

Conclusions: In the OCWAA consortium, almost half of the disparity in ovarian cancer survival between African-American and white women is explained by education, lifestyle factors, diabetes, PMH use, and tumor characteristics. A more complete understanding of the remaining difference in racial disparity is needed in order to ultimately improve outcomes and reduce disparities.

Table 1.

Sequential multiple mediation of racial differences in ovarian cancer survival in OCWAA

StepaMediator(s) AddedIndirect Effect HRbDirect Effect HRcTotal Effect HRd% Mediatede
College graduate 1.015 (0.960, 1.073) 1.308 (1.265, 1.351) 1.327 (1.268, 1.389) 4.5 (-16.9, 22.0) 
Nulliparity 1.018 (0.963, 1.076) 1.304 (1.261, 1.347) 1.327 (1.268, 1.388) 5.8 (-15.6, 22.9) 
Smoking 1.023 (0.968, 1.082) 1.296 (1.254, 1.340) 1.326 (1.266, 1.387) 7.5 (-13.3, 24.8) 
BMI,Diabetes,Diabetes/Race Interaction 1.051 (0.995, 1.111) 1.263 (1.222, 1.305) 1.327 (1.268, 1.389) 17.0 (-2.2, 33.2) 
PMH Duration,PMH Duration/Race Interaction,PMH Duration/Age Interaction 1.086 (1.025, 1.149) 1.195 (1.157, 1.235) 1.298 (1.238, 1.359) 31.0 (11.2, 47.3) 
Histotype,Stage 1.129 (1.073, 1.189) 1.149 (1.114, 1.184) 1.297 (1.243, 1.353) 46.4 (30.9, 60.0) 
StepaMediator(s) AddedIndirect Effect HRbDirect Effect HRcTotal Effect HRd% Mediatede
College graduate 1.015 (0.960, 1.073) 1.308 (1.265, 1.351) 1.327 (1.268, 1.389) 4.5 (-16.9, 22.0) 
Nulliparity 1.018 (0.963, 1.076) 1.304 (1.261, 1.347) 1.327 (1.268, 1.388) 5.8 (-15.6, 22.9) 
Smoking 1.023 (0.968, 1.082) 1.296 (1.254, 1.340) 1.326 (1.266, 1.387) 7.5 (-13.3, 24.8) 
BMI,Diabetes,Diabetes/Race Interaction 1.051 (0.995, 1.111) 1.263 (1.222, 1.305) 1.327 (1.268, 1.389) 17.0 (-2.2, 33.2) 
PMH Duration,PMH Duration/Race Interaction,PMH Duration/Age Interaction 1.086 (1.025, 1.149) 1.195 (1.157, 1.235) 1.298 (1.238, 1.359) 31.0 (11.2, 47.3) 
Histotype,Stage 1.129 (1.073, 1.189) 1.149 (1.114, 1.184) 1.297 (1.243, 1.353) 46.4 (30.9, 60.0) 

Abbreviations: OCWAA, Ovarian Cancer in Women of African Ancestry; PMH, Postmenopausal hormone; BMI, body mass index.

Bold numbers reflect statistically significant values.

aSequential order in which the selected mediators were added to the model. Model also includes variables in all preceding steps.

bThe indirect effect is the cumulative indirect effect of the variables selected as mediators in a given step and all preceding steps.

cThe direct effect is the remaining direct effect through all other non-mediated pathways.

dTotal effect is calculated as log(Indirect Effect HR) + log(Direct Effect HR).

ePercent (%) mediated is calculated as log(indirect effect HR)/log(total effect HR).

Citation Format: Kristin A. Guertin, Holly R. Harris, Tareq F. Camacho, Courtney E. Johnson, Anna H. Wu, Patricia G. Moorman, Evan Myers, Traci N. Bethea, Elisa V. Bandera, Charlotte E. Joslin, Heather M. Ochs-Balcom, Lauren C. Peres, Will T. Rosenow, Veronica W. Setiawan, Alicia Beeghly-Fadiel, Lauren F. Dempsey, Lynn Rosenberg, Joellen M. Schildkraut. Racial disparities in epithelial ovarian cancer survival: An examination of contributing factors in the Ovarian Cancer in Women of African Ancestry (OCWAA) consortium [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 33.