Criterion validity of brief, comprehensive, and latent measures of intrinsic capacity against frailty, disability, and mortality in older adults.
Intrinsic capacity (IC), the composite of physical and mental capacities, underpins the WHO Integrated Care for Older People (ICOPE) program. IC may be measured as a brief screen, a comprehensive assessment, or a latent factor score; whether the simpler measures retain the prognostic information of the complex ones remains untested within a single population.
In the nationwide Korean Frailty and Aging Cohort Study, 3011 community-dwelling adults aged 70-84 years (mean age, 76.0 [SD, 3.9] years; 52.5% women) were assessed for frailty and for activities of daily living (ADL) and instrumental ADL (IADL) disability and followed for mortality (median, 8.0 years). IC was measured by a Korean ICOPE screening score (K-ICOPE), a comprehensive clinical score (CIC), and a bifactor latent score (LIC); differences in discrimination were tested formally.
Discrimination for frailty rose with measurement intensity (area under the curve, 0.82, 0.86, 0.90; all pairwise P < .001) but not for ADL or IADL disability (all P ≥ .25). There were 500 deaths. Adjusted for sociodemographics and multimorbidity, each 1-SD higher IC was associated with lower mortality (hazard ratio 0.83 [95% CI, 0.75-0.91], 0.71 [0.64-0.78], 0.65 [0.58-0.73]), with closely similar discrimination (C statistic, 0.723, 0.731, 0.736; largest difference, 0.013 [0.005-0.021]). Discrimination from IC alone was lower and similar across measures (0.583, 0.604, 0.585).
Measurement intensity improved the characterization of frailty but added little to the ranking of long-term mortality risk. Because K-ICOPE was aligned by construction to the latent structure, these findings support the potential feasibility of brief ICOPE-style screening rather than validating the WHO instrument.
In the nationwide Korean Frailty and Aging Cohort Study, 3011 community-dwelling adults aged 70-84 years (mean age, 76.0 [SD, 3.9] years; 52.5% women) were assessed for frailty and for activities of daily living (ADL) and instrumental ADL (IADL) disability and followed for mortality (median, 8.0 years). IC was measured by a Korean ICOPE screening score (K-ICOPE), a comprehensive clinical score (CIC), and a bifactor latent score (LIC); differences in discrimination were tested formally.
Discrimination for frailty rose with measurement intensity (area under the curve, 0.82, 0.86, 0.90; all pairwise P < .001) but not for ADL or IADL disability (all P ≥ .25). There were 500 deaths. Adjusted for sociodemographics and multimorbidity, each 1-SD higher IC was associated with lower mortality (hazard ratio 0.83 [95% CI, 0.75-0.91], 0.71 [0.64-0.78], 0.65 [0.58-0.73]), with closely similar discrimination (C statistic, 0.723, 0.731, 0.736; largest difference, 0.013 [0.005-0.021]). Discrimination from IC alone was lower and similar across measures (0.583, 0.604, 0.585).
Measurement intensity improved the characterization of frailty but added little to the ranking of long-term mortality risk. Because K-ICOPE was aligned by construction to the latent structure, these findings support the potential feasibility of brief ICOPE-style screening rather than validating the WHO instrument.
Authors
Kim Kim, Luo Luo, Shi Shi, Park Park, Pritchard Pritchard, Kim Kim, Won Won, Kim Kim
View on Pubmed