Factors associated with late-life psychosis in primary care older adults without a diagnosis of dementia.

ABSTRACT

PURPOSE: The epidemiology of late-life psychosis (LLP) remains unclear comparatively to early-onset psychosis. The study aims to estimate the prevalence and incidence of LLP over a 3-year period and examine the correlates of LLP in community-living older adults aged ≥ 65 years recruited in primary care. METHODS: Study sample included N = 1481 primary care older adults participating in the Étude sur la Santé des Aînés (ESA)-Services study. Diagnoses were obtained from health administrative and self-reported data in the 3 years prior and following baseline interview. The prevalence and incidence of LLP (number of cases) were identified in the 3-year period following interview. Participants with dementia or psychosis related to dementia were excluded. Logistic regressions were used to ascertain the correlates of LLP as function of various individual and health system factors. RESULTS: The 3-year prevalence and incidence of LLP was 4.7% (95% CI = 3.64-5.81) and 2.8% (95% CI = 1.99-3.68), respectively. Factors associated with both prevalent and incident LLP included functional status, number of physical diseases, hospitalizations, continuity of care and physical activity. Older age and the presence of suicidal ideation were associated with incident LLP, while higher education, a depressive disorder and a history of sexual assault were associated with persistent cases. CONCLUSIONS: Results highlight the importance of LLP in primary care older adult patients without dementia. Health system factors were consistent determinants of prevalent and incident LLP, suggesting the need for better continuity of care in at-risk primary care older adults.