Longitudinal Dynamics in Indicators of Frailty: Predictors and Long-Term Outcomes

Abstract

Frailty is a common geriatric condition with a wide array of sequelae, including increased risks of mortality, morbidity and disability. Despite its long conceptual and operational history in research and publications, both frailty and mechanisms of frailty development are still poorly understood. A detailed description of trajectories of frailty indicators could offer vital insights on unfolding longitudinal dynamics involved in the development of frailty. Such a longitudinal modeling could also provide researchers and clinicians with a better foundational understanding of the phenomenon and facilitate targeted care approach. The purpose of this study was to: 1. Describe longitudinal (~10 years) trajectories of change in musculoskeletal and neuro-cognitive indicators of frailty in older (>/=65 years) women enrolled in the Women's Health Initiative Clinical Trial. 2. Estimate the extent to which baseline factors (e.g., demographic characteristics, health status and behaviors) conjointly were associated with a likelihood of membership in the derived longitudinal clusters. 3. To determine the extent to which membership in longitudinal trajectories predicts the incidence of clinically relevant geriatric health outcomes (i.e., mortality and hospitalization) over 5-years of follow up (2005-2010 WHI Extension Study) in a model adjusted for all other baseline predictors. The study findings demonstrated a high degree of heterogeneity in longitudinal dynamics of individual frailty criteria. We also showed that age, socio-demographic variables, health status, health behavior, environmental factors and personality traits are important determinants of individual frailty criteria. However the effect of these determinants on frailty phenotype is complex, presumably due to the multidimensional nature of frailty phenomenon. Thirdly, we found that the magnitude of risk carried by a membership in a certain longitudinal group for each of the defining elements of frailty is closely linked to the distance of that trajectory estimates from the one that represents the most optimal criterion-specific functioning over time. The further the distance between trajectory estimates of an individual who maintained the highest level of performance (specific to that indicator) and those who demonstrated less optimal functioning, the higher the risk of incidence of adverse health events. Finally, we empirically determined that distribution based cross sectional partitioning of frailty criteria seems to be a valid method for defining frailty given that elderly women maintained approximately similar levels of functioning over time without demonstrating clear increasing or decreasing longitudinal patterns.