Inter-relationships between Gender, Frailty and 10-Year Survival in Older Italian Adults: an observational longitudinal study

Abstract

Aim of the present study was to assess the impact of gender on the relationship between long-term mortality and clinical frailty. In an observational, longitudinal study on 10-year mortality, we examined 1284 subjects. The Frailty Staging System was used to assess frailty. The Cox model was employed to assess variables independently associated with survival using a backward stepwise algorithm. To investigate the possible interactions between gender and the selected variables, an extension of the multivariable fractional polynomial algorithm was adopted. Women were more likely to be older, have a higher disability, present with more comorbidities, consume more drugs, be frail and have a higher rate of survival at the follow-up than were men. At the Cox multivariate analysis only age (HR 2.26), female gender (HR 0.43), and number of drugs (HR 1.57) were significant and independent factors associated with all-cause mortality. In the survival analyses, only frailty (vs no frailty) showed significant interaction with gender (p < 0.001, HR = 1.92). While the presence of frailty reduced the survival rate in women, no effect was observed in men. Importantly, frail women showed higher survival rates than did both frail and no frail men. The main finding of the present study is that gender shapes up the association between frailty and long-term survival rates.

Figure 1

Gender and frailty interaction directly adjusted survival curves. Each directly adjusted curve, estimated at a specific factor and interaction value, is compared to the overall observed survival Kaplan Meier curve (red continuous line) and exemplifies the survival that would be observed if all patients in the study population had had the given specific factor and interaction value. Continuous black line = no frail female. Dashed black line = frail female. Continuous blue line = no frail male. Dashed blue line = frail male. The male lines fully overlap.