Predictive value of frailty indices for adverse outcomes in older adults

Background: There are two widely used tools to classify frailty in older adults: the frailty phenotype and the frailty index. Both have been validated for prediction of adverse outcomes. Objective: To assess the ability of different frailty indices to predict a number of adverse outcomes (falls, disability, and mortality) by adding deficits in a fixed sequence (with the first five deficits as in the frailty phenotype: weakness, weight loss, slowness, exhaustion and low physical activity) or randomly. Methods: This is an analysis of the Costa-Rican Longevity and Healthy Aging Study in which > 60-year-old adults were included and followed up for four years. Frailty indices were constructed, including the frailty phenotype components in the first five indices followed by the random addition of other deficits and estimating for each one the odds ratios for falls and disability and hazard ratios for mortality, adjusted for age and sex. Results: We included 2,708 adults; mean age was 76.31 years, 54.28% were women. Indices with the highest number of deficits had the highest estimates for each adverse outcome, independent of the deficit. Conclusion: The higher the number of deficits in an index, the higher the estimates for adverse outcomes, independent of the type of deficit added.