Trends in Racial and Ethnicity Disparities in the Health-Related Quality of Life in Older Adults with Breast Cancer: A SEER-MHOS National Database Study

Abstract

To examine racial and ethnic disparities in Health-Related Quality of Life (HRQOL) in older adults with breast cancer, both pre- and post-diagnosis. Using the Surveillance, Epidemiology and End Results-Medicare Health Outcomes Survey (SEER-MHOS) database, we included patients > 65 years old with breast cancer diagnosed from 1998-2019, who had completed the Health Outcomes Survey within 24 months pre- and post-diagnosis, and who had a reported ethnicity of non-Hispanic White (W), non-Hispanic Asian or Pacific Islander (API), non-Hispanic Black or African American (B), or Hispanic (H). HRQOL data was measured by the Physical Component Summary (PCS) and Mental Component Summary (MCS) scores from the Medical Outcomes Study Short Form-36 and the Veterans RAND 12-Item Health Survey instruments. Univariable (UVA) and multivariable (MVA) linear regression models were fitted for MCS, PCS, and TCS (composite MCS and PCS) as response variables to assess for potential disparities between races and ethnicities. MVA models were adjusted for sex, age, marital status, education, income, number of comorbidities, limitations in activities of daily living, national region, histology, and treatment type (in post-diagnosis cohort only). We identified 1,008 in the pre-diagnosis cohort (56.6% stage I, 29.8% stage II, 9.3% stage III, and 4.3% stage IV) and 991 patients in the post-diagnosis cohort (58.3% stage I, 28.6% stage II, 9.3% stage III, 3.8% stage IV). On UVA, differences in mean scores of TCS, PCS, and MCS, respectively, were found among W, API, B, and H patients in both pre-diagnosis and post-diagnosis cohorts (pre-diagnosis - TCS: 93.4, 93.0, 87.0, 86.3; PCS: 40.3, 40.0, 37.0, 37.5; MCS: 53.1, 52.9, 50.0, 48.7; post-diagnosis - TCS: 90.8, 90.3, 84.3, 83.5; PCS: 38.5, 39.2, 36.8, 34.5; MCS: 52.3, 51.1, 49.8, 46.8; all p < .001). A drop in PCS, MCS, and TCS scores numerically was observed for all groups. Among patients in the pre-diagnosis cohort who would be diagnosed with stage IV breast cancer, race was found to be a predictor of PCS with overall significance (p = 0.04) within MVA linear regression. Specifically, on the local test, compared to B, W had higher pre-diagnosis PCS scores (+13.32, p = 0.03). Race was not found to be a predictor in PCS, MCS, or TCS scores otherwise. Among older patients who would be diagnosed with stage IV breast cancer, W had better physical HRQOL as compared to B patients' pre-diagnosis. The decrease in numerical HRQOL scores of the physical domain in all groups post-diagnosis highlights the potential negative physical impact breast cancer has on patients, demonstrating the need for determining the proper resources and support to improve physical HRQOL following diagnosis.