Objective
Survivors of multiple primary cancers (MPC) are at increased risk for poor health outcomes compared with survivors of single cancers. Using an adapted psychobehavioral stress‐response model, the study purpose was to identify pathways and individual risk factors associated with poor health outcomes in adults with MPC.
Methods
Adult MPC survivors (N = 211) with first cancers (stages I‐III) diagnosed within 1 to 10 years were recruited via tumor registry. Employing a cross‐sectional design, established questionnaires were used to operationalize patient characteristics and theoretical constructs including perceived stress, psychological and behavioral responses, financial hardship, social role function, and physical health. Disease and treatment data were obtained via registry and medical records. Structural equation modeling (SEM) was performed to fit, test, and modify the hypothesized psychobehavioral model.
Results
Following measurement model refinement, an SEM linking self‐management behaviors, distress, financial hardship, and functional health demonstrated a good fit: χ2 (200, N = 206) = 332.06, P < .01; Tucker‐Lewis index (TLI) = .95, comparative fit index (CFI) = .96, standardized root mean residual (SRMR) = .06, root‐mean‐square error of approximation (RMSEA) = .06. Less use of self‐management behaviors predicted higher distress which, in turn, predicted higher financial hardship; higher distress and financial hardship predicted poorer functional health. Several sociodemographic and personal factors predicted self‐management behaviors and distress.
Conclusions
The hypothesized stress‐response model was partially supported. Data supported pathways among self‐management behaviors, distress, financial hardship, and functional health. Self‐management and distress may represent modifiable intervention targets for MPC survivors. High body mass index (BMI), less education, greater neuroticism, and lower social support were associated with less use of self‐management behaviors and higher distress and should be further evaluated as potential markers of vulnerability.