Development and validation of a frailty index in the Longitudinal Aging Study Amsterdam

Background Frailty is a state of increased vulnerability to adverse outcomes. The frailty index (FI), defined by the deficit accumulation approach, is a sensitive instrument to measure levels of frailty, and therefore important for longitudinal studies of aging. Aims To develop an FI in the Longitudinal Aging Study Amsterdam (LASA), and to examine the predictive validity of this FI for 19-year mortality. Methods LASA is an ongoing study among Dutch older adults, based on a nationally representative sample. A 32-item FI (LASA-FI) was developed at the second LASA measurement wave (1995-1996) among 2218 people aged 57-88 years. An FI score between 0 and 1 was calculated for each individual. The LASA-FI included health deficits from the physical, mental and cognitive domain and can be constructed for most LASA measurement waves. Associations with 19-year mortality were assessed using Kaplan-Meier curves and Cox proportional hazards models. Results The mean LASA-FI score was 0.19 (SD = 0.12), with a 99% upper limit of 0.53. Scores were higher in women than men (women = 0.20, SD = 0.13 vs. men = 0.17, SD = 0.11, p < 0.001). The average age-related increase in the log-transformed LASA-FI score was 3.5% per year. In a model adjusted for age and sex, the FI score was significantly associated with 19-year all-cause mortality (HR per 0.01 = 1.03, 95% CI 1.03-1.04, p < 0.001). Discussion/conclusions The key characteristics of the LASA-FI were in line with findings from previous FI studies in population-based samples of older people. The LASA-FI score was associated with mortality and may serve as an internal and external reference value.