Predictors of Mortality in Dementia: The PRIME Study

Article type: Research Article

Authors: Connors, Michael H.^{a; b} | Ames, David^{c; d} | Boundy, Karyn^e | Clarnette, Roger^f | Kurrle, Sue^g | Mander, Alastair^h | Ward, Johnⁱ | Woodward, Michael^j | Brodaty, Henry^{a; b; *}

Affiliations: [a] Dementia Collaborative Research Centre, School of Psychiatry, UNSW Australia, Sydney, Australia | [b] Centre for Healthy Brain Ageing, School of Psychiatry, University of New South Wales, Sydney, Australia | [c] University of Melbourne Academic Unit for Psychiatry of Old Age, Melbourne, Australia | [d] National Ageing Research Institute, Melbourne, Australia | [e] Queen Elizabeth Hospital, Adelaide, Australia | [f] School of Medicine and Pharmacology, University of Western Australia, Perth, Australia | [g] Sydney Medical School, University of Sydney, Sydney, Australia | [h] Faculty of Health, Deakin University, Geelong, Australia | [i] School of Medicine and Population Health, University of Newcastle, Newcastle, Australia | [j] Medical and Cognitive Research Unit, Austin Health, Heidelberg, Australia

Correspondence: [*] Correspondence to: Henry Brodaty, Dementia Collaborative Research Centre, Level 3, AGSM Bldg (G27), University of New South Wales, NSW 2052, Australia. Tel.: +61 2 9385 2585; Fax: +61 2 9385 2200; E-mail: h.brodaty@unsw.edu.au.

Abstract: Background: Dementia is a terminal illness. While various baseline characteristics of patients, such as age, sex, and dementia severity, are known to predict mortality, little research has examined how changes in patients’ symptoms over time predict survival. There are also limited data on patients seen in memory clinics, as opposed to other health care settings, and whether antipsychotic medications are associated with mortality in dementia once patients’ demographic and clinical features are controlled for. Objective: To identify predictors of mortality in patients with dementia. Method: Of 970 patients recruited from nine memory clinics around Australia, 779 patients had dementia at baseline. Patients completed measures of dementia severity, cognition, functional ability, neuropsychiatric symptoms, caregiver burden, and medication use at baseline and at regular intervals over a three-year period. Mortality data were obtained from state registries eight years after baseline. Results: Overall, 447 (57.4%) of the patients with dementia died within the eight years. Older age, male sex, more severe dementia and functional impairment at baseline, greater decline in dementia severity and functional impairment over six months, taking a larger number of medications, and use of atypical antipsychotic medication predicted earlier mortality. Conclusions: The findings confirm that demographic and diagnostic features predict the survival of patients with dementia. Importantly, the findings indicate that changes in dementia severity and functional impairment over time predict mortality independently of baseline levels, and provide further evidence for the higher mortality risk of patients taking antipsychotic medications.

Keywords: Death, dementia, longitudinal study, mortality, predictors

DOI: 10.3233/JAD-150946

Journal: Journal of Alzheimer's Disease, vol. 52, no. 3, pp. 967-974, 2016