Incorporating persistent pain in phenotypic frailty measurement and prediction of adverse health outcomes

Abstract

Background: Frailty, a syndrome of physiological deficits, is prevalent among older adults and predicts elevated risk of adverse health outcomes. Although persistent pain predicts similar risk, it is seldom considered in frailty measurement. This article evaluated the construct and predictive validity of including persistent pain in phenotypic frailty measurement. Methods: Frailty and persistent pain were operationalized using data from the Health and Retirement Study (2006-2012 waves). Among a subset of adults aged 65 and older (n = 3,652), we used latent class analysis to categorize frailty status and to evaluate construct validity. Using Cox proportional hazards models, we compared time to incident adverse outcomes (death, fall, hospitalization, institutionalization, and functional disability) between frailty classes determined by either including or excluding persistent pain as a frailty component. Results: In latent class models, persistent pain occurred with other frailty components in patterns consistent with a medical syndrome. Frail and intermediately frail classes determined by including persistent pain were more strongly associated with all adverse outcomes compared with frail and intermediately frail classes determined excluding persistent pain. Frail respondents had significantly greater risk of death compared with nonfrail respondents when frailty models included rather than excluded persistent pain (respectively, hazard ratio [HR] = 3.87, 95% confidence interval [CI] = 2.99-5.00 (including pain); HR = 2.10, 95% CI = 1.71-2.59 (excluding pain). Conclusions: Findings support consideration of persistent pain as a component of the frailty phenotype. Persistent pain assessment may provide an expedient method to enhance frailty measurement and improve prediction of adverse outcomes.