Fleur Delva1, Celia Touraine2, Pierre Joly2, Arlette Edjolo2, Hélène Amieva2, Claudine Berr3, Olivier Rouaud4, Catherine Helmer2, Karine Pérès2, Jean-François Dartigues5. 1. Centre INSERM U897, ISPED, Bordeaux, France; Université Bordeaux, Bordeaux, France. Electronic address: fleur.delva@isped.u-bordeaux2.fr. 2. Centre INSERM U897, ISPED, Bordeaux, France; Université Bordeaux, Bordeaux, France. 3. Centre INSERM U1061, Université Montpellier 1, Hôpital La Colombière, Montpellier, France; CMRR Languedoc-Roussillon, CHU Montpellier, Hôpital Gui de Chauliac, Département de Neurologie, Montpellier, France. 4. CMMR, CHU BOCAGE, DIJON, France. 5. Centre INSERM U897, ISPED, Bordeaux, France; Université Bordeaux, Bordeaux, France; CMRR d'Aquitaine, CHU de Bordeaux, Bordeaux, France.
Abstract
INTRODUCTION: Transition to bathing or dressing disability is a milestone in the evolution of dementia. We examined the transition to disability in these specific activities and considered death to be a competitive event and age and sex to be prognostic factors. METHODS: From a large cohort of 570 incident dementia cases screened in two prospective population-based cohorts, the Paquid study, and the Three-City study, we estimated the probabilities of remaining nondisabled, becoming disabled in bathing or dressing, or dying after the diagnosis using an illness-death model. RESULTS: On average, approximately half of the period (3 years) of living with dementia was free of disability. In women, a higher survival rate was associated with an average of 1 additional year with disability. DISCUSSION: The joint prediction of death and disability in dementia by an illness-death model gives original and useful parameters for the prognosis and management of dementia.
INTRODUCTION: Transition to bathing or dressing disability is a milestone in the evolution of dementia. We examined the transition to disability in these specific activities and considered death to be a competitive event and age and sex to be prognostic factors. METHODS: From a large cohort of 570 incident dementia cases screened in two prospective population-based cohorts, the Paquid study, and the Three-City study, we estimated the probabilities of remaining nondisabled, becoming disabled in bathing or dressing, or dying after the diagnosis using an illness-death model. RESULTS: On average, approximately half of the period (3 years) of living with dementia was free of disability. In women, a higher survival rate was associated with an average of 1 additional year with disability. DISCUSSION: The joint prediction of death and disability in dementia by an illness-death model gives original and useful parameters for the prognosis and management of dementia.