Intrinsic Capacity (IC) is a multidimensional measure of physical and mental capacity in older individuals. Proposed by the World Health Organization, it is increasingly used internationally in research and clinical practice. Like frailty and multimorbidity, IC measures health in older age, but differs in its strengths-based orientation and its potential to identify ageing-related decline earlier, which may improve prediction of future care needs. However, its predictive performance relative to these measures has yet to be clearly established.
We address this gap by comparing the predictive performance of IC, frailty, and multimorbidity for future healthcare use. We use data from an 11-country sample of the Survey of Health, Ageing, and Retirement in Europe (SHARE), comprising 128,783 observations on 53,583 individuals across eight waves. IC is estimated using a factor analysis model combining multiple indicators across 5 domains, frailty using a 37-item frailty index, and multimorbidity using 12 chronic conditions. We predict hospitalisation and prescription medication use over 2-, 6-, and 10-year horizons, and doctor or nurse visits over a 2-year horizon. To assess out-of-sample predictive performance, we use three complementary validation strategies: individuals, later survey waves, and entire countries not used to develop the models. We also assess whether prediction accuracy differs across population subgroups.
By benchmarking IC against alternative measures of health in older age, this study evaluates whether its broader, capacity-based conceptualisation translates into improved prediction of future healthcare use. The findings inform the choice of health measures for understanding and anticipating healthcare needs in ageing populations.