Depression, physical illness and mortality in a Spanish community-dwelling elderly people

Abstract

ObjectivesThe aim of this study is to investigate the association between depression, physical factors and mortality in elderly people living in the community.MethodsProspective longitudinal multicenter study for 5 years. Cohort of 293 people aged 65 years and older living in the province of Huesca (Spain). Individual face-to-face interviews and with appropriate caregiver. The following information was collected:– demographic data;– psychosocial factors: sex, age, education, marital status, live-in family members, social relationships, life events;– physical factors: severity of physical illness, comorbidity (Cumlative Illness Rating Scale);– psychiatric factors: cognitive function (Spanish version of Mini-Mental State Examination), depression (Geriatric Depression Scale), diagnostic criteria according DSM-IV-TR.Statistical analyses:– a bivariate analysis;– a multivariate analysis. Cox regression model (explanatory variables).ResultsTwo hundred ninety-three participants, simple representative of people aged 65 years old or more in province of Huesca (Spain). Monitored 5 years follow-up study. Sixty-four people died (21.8%), annual mortality rate: 5.3%. Depression: 66 people (22.5%), (32.2% women, 13.3% men). Cognitive impairment: 51 people (17.4%). Bivariate: factors associated (P < 0.005) with mortality: functional impairment, living in nursing home, sensorial impairment, polypharmacy, severe physical illness and psychiatric comorbidity: depression (34.8 vs 18.1%), cognitive impairment (49.1 vs 15.8%). Association between some factors and mortality was nullified after multivariate statistical model; the case for depression (Hazard Ratio: 1.1), cognitive impairment (HR: 1.2) or functional impairment (HR: 1.3).ConclusionsDepression and cognitive impairment are associated with mortality in elderly community living people in bivariate analysis, therefore, this association disappears after multivariate analysis. Severity physical illness seems to nullify the effect of other variables, such as depressive symptomatology.Disclosure of interestThe authors have not supplied their declaration of competing interest.