Aging Clinical and Experimental Research · Oct 30, 2025

Frailty modulates the predictive value of performance status in older adults living with cancer

By Patino-Hernandez D, Pérez-Zepeda MU, Sánchez-Garrido N et al.

BACKGROUND: Cancer prognosis in older adults is complicated by age-related vulnerabilities such as frailty, which may impact the performance of traditional prognostic tools, which may not accurately reflect risk in this population. AIMS: This study aimed to assess whether frailty modifies the association between a modified performance status (PS) -constructed from days spent in bed, physical activity, and disability-and mortality risk in older adults with cancer. METHODS: We conducted a secondary analysis of the Mexican Health and Aging Study (MHAS), including individuals ≥50 years with incident cancer and active treatment, with follow-up through 2021. A frailty index (FI) was calculated using a 33-item deficit accumulation model. Modified PS was constructed using data on days in bed, functional limitations, and physical activity. Participants were stratified into low (<0.25) and high (≥0.25) frailty levels. Cox regression and Kaplan-Meier curves were used to assess associations with all-cause mortality. RESULTS: Among 318 participants (mean age 68.0 years; 62.6% women), frailty and performance status were both associated with mortality. In individuals with low frailty, modified PS categories 2 and 4 were significantly associated with higher mortality (HR 5.46 [95% CI: 1.24-24.04], and HR 8.01 [1.50-42.66], respectively). Among those with high frailty, only age remained a significant predictor of mortality (HR 1.04 [1.00-1.07]). DISCUSSION: Frailty significantly modifies the prognostic value of performance status. In frail older adults with cancer, the predictive utility of performance status tools is reduced. CONCLUSIONS: Incorporating frailty assessment may enhance risk stratification and clinical decision-making in older adults with cancer.

doi.org/10.1007/s40520-025-03203-4 Open access PubMed 41165923 Biomarkers