Frailty as a predictor of mortality in the oldest old: A systematic review and meta-analysis

Abstract

Aim

Frailty is highly prevalent in old age and is associated with a high risk of mortality. Few studies have evaluated frailty as a predictor of mortality in the oldest old. This systematic review and meta-analysis aims to determine the mortality risk associated with frailty in this age group.

Methods

An electronic systemic literature search was performed in May 2023 on three databases (Medline/PubMed, Embase, and Cochrane Library) for studies investigating the risk of mortality with frailty. A meta-analysis was done to calculate pooled mortality estimates.

Results

Frail participants had significantly lower overall survival (OS) compared with non-frail participants (hazard ratio [HR]: 1.81; 95% confidence interval [CI]: 1.32 to 2.50; P < 0.01, I² = 100%; risk ratio [RR]: 4.15; 95% CI: 2.50 to 6.88; P < 0.01, I² = 97%). Among participants aged 90 and above, a higher percentage of male participants was associated with poorer OS in frail participants. While the pooled association of frailty with OS remained significant across studies in participants aged less than 90 years old (HR: 2.09; 95% CI: 1.55 to 2.82; I² = 75%), frailty was not significantly associated with OS in studies for participants aged 90 and above. The pooled association of frailty with OS was only significant for the Fried Frailty Phenotype (HR: 2.73; 95% CI: 1.05 to 7.12; I² = 93%) and not for the Frailty Index. The pooled association also remained significant among studies that adjusted for age (HR: 1.74; 95% CI: 1.50 to 2.02; I² = 0%) and sex (HR: 1.77; 95% CI: 1.48 to 2.11; I² = 94%) as a covariate.

Conclusions

Frailty was significantly associated with a poorer OS in participants below the age of 90. This association was not statistically significant in those older than 90 years, with sex-differentiated effects observed. Geriatr Gerontol Int 2025; 25: 102–107.