Background: American Indian and Alaska Native (AIAN) people experience lower cancer survival than other racial/ethnic groups for several cancer sites. We explored predictors of cancer survival among AIAN people, with a specific focus on the potential impact of comorbidities. Methods: We analyzed the Surveillance, Epidemiology, and End Results–Medicare linked database. AIAN people diagnosed with breast (female only), lung, and colorectal cancers from 2000 to 2019, ages 66+ years at diagnosis, and with at least 1 year of Medicare coverage prior to diagnosis were included in the study. We used multivariable-adjusted Cox proportional hazards models to assess associations of comorbidity class, previously defined using latent class analysis, with overall and cancer-specific survival, adjusted for sociodemographic characteristics and area-based social measures. Results: The sample included 831 individuals with female breast cancer, 1,009 individuals with colorectal cancer, and 1,195 individuals with lung cancer. Comorbidity class was associated with 5-year hazard of death from any cause for patients with breast and colorectal cancers but not patients with lung cancer. Furthermore, we only observed associations of comorbidity class with 5-year cancer-specific death for lung cancer, in which the class associated with heart and vascular-related conditions experienced increased hazard of death. Conclusions: Higher comorbidity prevalence among AIAN people compared with other racial/ethnic groups did not contribute to observed disparities in cancer-specific survival; however, AIAN people with comorbid conditions may be more likely to experience death from any cause than those without these conditions. Impact: Differences in prevalence of comorbid conditions are unlikely to drive disparities in cancer-specific survival observed among AIAN people.
9 September 2025 in Cancer Epidemiology, Biomarkers & Prevention
Background: Comorbidities may affect the incidence and management of cancers. The burden of comorbidities among American Indian and Alaska Native (AIAN) patients with cancer and survivors is unknown. Methods: Using Surveillance, Epidemiology, and End Results–Medicare, we identified AIAN people aged ≥66 years diagnosed with female breast, lung, and colorectal cancers (2000–2019), with at least 1 year of Medicare coverage prior to diagnosis. We used latent class analysis (LCA) to identify comorbidity profiles and multinomial logistic regression to examine associations of class membership with sociodemographic characteristics. Results: The mean NCI weighted comorbidity index was 0.53 (±0.59) for those with breast cancer, 0.59 (±0.65) for those with colorectal cancer, and 0.76 (±0.68) for those with lung cancer. For breast and colorectal cancers, LCA identified three classes (low, medium, and high comorbidities). For lung cancer, LCA identified four classes (two “medium” comorbidity classes differed in the proportion of chronic obstructive pulmonary disease and cardiovascular-related diseases). Sex, age, residence in a purchased/referred care delivery area, and year of diagnosis were associated with class membership for each cancer. Colorectal cancer class membership was associated with the proportion of the population with limited access to a grocery store; lung cancer class membership was associated with the proportion of the AIAN population with limited fluency in the English language. Conclusions: The NCI comorbidity score among AIAN people with cancer was slightly higher than published reports for those of non-Hispanic Whites. Impact: Managing comorbidities and cancer treatment may be complicated for AIAN people due to the provision of noncancer care within the Indian Health Service and the need to coordinate across multiple care systems.
Supplementary Table 3 gives univariable associations of sociodemographic and area-based social measures with class membership derived from latent class analysis for female breast cancers
Supplementary Table 4 gives univariable associations of sociodemographic and area-based social measures with class membership derived from latent class analysis for colorectal cancers