Neuropsychiatric Symptoms as Predictors of Falls in Long-Term Care Residents With Cognitive Impairment

Abstract

Listen

Translate article

ObjectivesFalls and neuropsychiatric symptoms (NPS) are common among long-term care residents with cognitive impairment. Despite the high prevalence of falls and NPS, little is known about their association. The aim of our study was to explore how NPS, particularly the severity of NPS and specific NPS subgroups, are associated with falls and how psychotropics modify this association. DesignLongitudinal cohort study. Setting and ParticipantsIn total, 532 long-term care residents aged 65 years or older in Helsinki, Finland. MethodsNPS were measured with Neuropsychiatric Inventory (NPI) at baseline. Participants were grouped into 3 groups: no significant NPS (NPI points 0‒3), low NPS burden (NPI 4‒12), and high NPS burden (NPI >12). The number of falls, injuries, fractures, and hospitalizations were collected from medical records over 12 months following baseline assessment. ResultsAltogether, 606 falls occurred during the follow-up year. The falls led to 121 injuries, 42 hospitalizations, and 20 fractures. Falls and injuries increased significantly with NPS burden (P < .001): 330 falls in the high NPS group (n = 184), 188 falls in the low NPS group (n = 181), and 88 falls in the no significant NPS group (n = 167). The risk of falling showed a curvilinear association with NPI total score. Of NPS subgroups, psychosis and hyperactivity were associated with a higher incidence rate ratio of falls, whereas apathy had a protective association even after adjustment for age, sex, and mobility. Affective symptoms were not associated with falls. Psychotropics did not modify the association between NPS burden and falls. Conclusions and ImplicationsThe results of this study show that NPS, especially NPS severity, may predict falls and fall-related negative consequences. Severity of NPS should be taken into account when assessing fall risk in long-term care residents with cognitive impairment.