Tsukuba, Japan—Older adults requiring long-term care frequently experience multiple disabilities, with considerable variation in the combinations of these impairments. Because of this diversity, interventions that focus on a single impairment may not be sufficient. Therefore, understanding the complex physical and cognitive conditions of older adults in need of care is essential for developing appropriate interventions.

In this study, researchers used unsupervised machine learning methods to classify individuals aged 65 and older who had begun using long-term care insurance services in two Japanese cities: Tsukuba City (Ibaraki Prefecture) and Kashiwa City (Chiba Prefecture). The classification was based on 74 items, primarily related to physical and cognitive functions, collected through a standardized care-needs certification survey. The researchers also examined the relationship between the classifications (functional subtypes) and prognosis, including outcomes such as death, hospitalization, admission to long-term care facilities, and care-need level deterioration.

The analysis of data from Tsukuba City identified five functional subtypes: i. mild physical, ii. mild cognitive, iii. moderate physical, iv. moderate multicomponent, and v. severe multicomponent. This classification was validated using data from Kashiwa City. In terms of prognosis, the severe multicomponent type showed a particularly high risk of death and long-term care facility admission; the moderate physical type was associated with a high risk of hospitalization; and the moderate multicomponent type was associated with a high risk of care-need level deterioration.

These findings may help individuals requiring long-term care and those involved in their care to consider appropriate treatment and care strategies. Future research into the most appropriate medical and long-term care services for each functional subtype is expected to contribute to the improvement of the quality and efficiency of care services.