The World Health Organization (WHO) defines healthy ageing as the development and maintenance of functional ability that fosters well-being. Optimizing intrinsic capacity (IC) is an important pathway to achieving healthy ageing. IC, encompassing all physical and mental capacities an individual possesses, marks a shift from focusing solely on diseases to an integrated, older individual-centred approach that emphasizes functional capabilities. It includes five key dimensions: locomotion, vitality, sensory, cognitive, and psychological.1,2 Evaluating IC is crucial for designing personalized care plans that integrate strategies to reverse, slow, or prevent further declines in capacity, address social care needs, and thereby foster healthy ageing.

Existing evidence strongly indicates that IC is a reliable and robust predictor of adverse health outcomes, including functional disability, nursing home admissions, hospitalization, and mortality, even after accounting for frailty, comorbidity, and chronological age.3, 4, 5 Nevertheless, previous research has typically utilized composite IC scores, potentially masking more detailed associations between IC dimensions and negative outcomes.5,6 IC is a multidimensional construct, and individuals do not always have balanced scores across its dimensions. Individuals with the same total score of IC may exhibit significant differences in the specific dimensions affected. For example, some individuals may have high deterioration with cognitive impairment, while others may have high deterioration with psychological and visual impairment.3 Therefore, from the perspective of heterogeneity, further characterization of IC is crucial to our understanding of patterns of impairment presentation and the association with adverse outcomes, for developing targeted, cost-effective intervention strategies to prevent IC decline.

Although there have been studies exploring the relationship between IC heterogeneity and adverse outcomes, important gaps remain. For instance, a study on Chinese community-dwelling older adults suggested that individuals in subgroups with sharp declines in specific IC-domain combinations had a higher risk of adverse outcomes compared to those in relatively healthy subgroups.7 However, most previous research has focused on static assessments rather than its longitudinal evolution, thereby overlooking the dynamic nature of IC transitions and their implications. IC is not a fixed state but a continuous process that reflects an individual’s aging trajectory over time rather than a single measurement.2,3 More importantly, individuals can move between latent IC profiles, experiencing improvements or declines, rather than following a predetermined profile. A cohort study involving participants from low- and middle-income countries that examined the natural history of IC impairment found that 61% of individuals worsened over time, while 3% improved to a healthier status.3 This underscores the need to investigate not only the overall progression of IC but also specific transitions between distinct profiles, as these shifts may serve as early indicators of functional decline or resilience. Understanding these transitions is essential for leveraging IC as a predictive tool in geriatric care, allowing for timely, personalized interventions before the onset of frailty or disability.8

In the current study, we employed an individual-centred approach that classifies older adults into discrete profiles based on a given set of observed indicators, ideal for capturing IC heterogeneity. Specifically, this study aimed to explore the continuity and discontinuity of these distinct profiles of IC over time and to assess whether specific transitions predict the risk of adverse outcome using longitudinal data from a nationally representative sample of older Chinese adults.