| Sumario: | Background & Significance: Cardio-oncology patients with both cardiovascular disease (CVD) and cancer face complex health outcomes due to overlapping symptoms. Cancer therapy-related cardiac dysfunction (CTRCD), including heart failure and cardiomyopathy, is a key outcome influencing prognosis. These patients often experience high symptom burden, while adverse social determinants of health (SDOH), such as neighborhood disorder and discrimination, may further increase vulnerability. Yet few studies have examined how symptoms, perceived stress, social support, and SDOH jointly related to CTRCD in this population. Purpose: This study aimed to examine the associations and identify predictors of CTRCD among cardio-oncology patients while accounting for sociodemographic variables. Methods: This cross-sectional secondary analysis used data from the NIH All of Us Research Program, including 12,888 adults with documented diagnoses of both CVD and cancer and complete electronic health record and survey data. Five predictors and three co-variates were examined. Missing data were addressed using random forest imputation. Logistic regression with inverse probability weighting was conducted in R to account for class imbalance, and model performance was evaluated using 10-fold cross-validation (CV). Findings and Interpretations: Among final sample, 17% had CTRCD, while the remaining 83% did not. Across 10-fold CV, the model demonstrated moderate discrimination (AUC = 0.65) with balanced accuracy (0.61). Weighted regression model showed symptom burden (OR = 1.72, 95% CI [1.62-1.82]), perceived stress (OR = 1.02, 95% CI [1.01-1.02]), and older age (OR = 1.03, 95% CI [1.03-1.04]) increased odds of CTRCD, whereas being a woman (OR = 0.55, 95% CI [0.51-0.60]) and being partnered (OR = 0.74, 95% CI [0.68-0.80]) were protective (all p < .001). Social support, neighborhood disorder, and discrimination were not significant. Discussion: CTRCD was most strongly associated with symptom burden, perceived stress, and older age, whereas being a woman and being partnered appeared protective. Symptom count may serve as an early clinical signal, and stress as a modifiable risk factor. Oncology nurses should incorporate structured symptom and stress surveillance, targeted monitoring of high risk-groups, and partner-inclusive approaches. However, given the imbalance in CTRCD within the sample, it is possible that other predictors did not reach statistical significance, underscoring the need for cautious interpretation. Future research should employ more balanced cohorts to refine prediction and prevention strategies.
|
|---|