Objective
To evaluate the 5 components of the Fried frailty phenotype (self-reported unintentional weight loss, physical activity questionnaire, gait speed, grip strength, and self-reported exhaustion) for long-term outcomes in elderly survivors of acute coronary syndrome.
Methods
A total of 342 consecutive patients (from October 1, 2010, to February 1, 2012) were included. The 5 components of the Fried score and albumin concentration, as malnutrition index, were assessed before hospital discharge. Patients were followed up until April 2020 (median follow-up, 8.7 years). The end point was postdischarge all-cause mortality.
Results
Mean ± SD age was 77±7 years and mean ± SD Fried score was 2.0±1.1 points. A total of 216 (63%) patients died. After adjusting for clinical covariates, the Fried phenotype was associated with mortality (per points, hazard ratio [HR], 1.35; 95% CI, 1.17 to 1.57;
P
<.001). Among Fried components, physical activity (HR, 2.21; 95% CI, 1.34 to 3.65;
P
=.002) and gait speed (HR, 1.77; 95% CI, 1.29 to 2.43;
P
<.001) were the deficits independendtly associated with mortality. Albumin level provided further prognostic information (per increase in g/dL; HR, 0.63, 95% CI, 0.45 to 0.88;
P
=.007). The model adding the components of the Fried score and albumin level to the clinical model showed the highest risk reclassification (integrated discrimination improvement, 0.040; 95% CI, 0.018 to 0.075;
P
=.001; continuous net reclassification improvement, 0.291; 95% CI, 0.132 to 0.397;
P
=.001) in comparison with the model using clinical covariates alone.
Conclusion
Frailty assessment using the Fried phenotype has prognostic value for long-term mortality in elderly survivors of acute coronary syndrome. Physical activity and gait speed are the predictive components of the Fried score. Albumin level provides incremental prognostic information.