Association of frailty with mortality in cancer survivors: results from NHANES 1999-2018

Cancer survivors are vulnerable to frailty. While few studies have focused on the association of frailty with mortality risk among cancer survivors, the current study aimed to reveal this association. In this cohort study, 4723 cancer survivors were enrolled from the National Health and Nutrition Examination Surveys (NHANES, 1999-2018). Frailty status was quantified using the 53-item frailty index. Death outcomes were linked to National Death Index mortality data (as of December 31, 2019). Cox proportional hazard models were used to estimate HRs (95% CIs). The median (IQR) frailty score was 0.190 (0.132, 0.277). During the median follow-up of 6.7 years, 1775 all-cause deaths (including 581 cancer deaths and 385 cardiac deaths) were documented. Compared to the lowest tertile of frailty scores, the adjusted HRs (95% CIs) for the highest tertile were 2.698 (2.224, 3.272) for all-cause mortality (P trend < 0.001), 2.145 (1.547, 2.973) for cancer mortality (P trend < 0.001), and 3.735 (2.231, 6.251) for cardiac mortality (P trend < 0.001). Moreover, a positive doseresponse association between the frailty score and mortality risk was determined. Each per-unit increase in the frailty score (natural logarithm transformed) was found to increase all-cause mortality by 159% (P < 0.001), cancer mortality by 103% (P < 0.001), and cardiac mortality by 256% (P < 0.001). A consistent result was shown when stratifying by age, sex, race, body mass index, and type of cancer. This study suggested that the frailty index was positively associated with all-cause mortality and cause-specific mortality (including cancer and cardiac deaths) among cancer survivors.