Objective: to identify the predictors of mortality in a cohort of nonagenarians inside the "Mugello study" after 10 years follow-up.
Methods: Information on sociodemographic data, cognitive and functional status, lifestyle, medical history, and drug use was collected from 433 non-selected participants aged 90-99 years, living in the Mugello area (Italy). Participants were followed over 10 years and their dates of death were retrieved from the municipal registers. Cox regression analysis was used to determine significant potential prognostic factors.
Results: The mortality rate was 96.5%. Cox proportional hazards analysis showed that a lower cognitive status was significantly associated with higher mortality as well as a poorer functional status, a higher comorbidity, and a higher number of drugs consumption.
Discussion: Impaired cognitive function, loss of functional independence, higher comorbidity, and higher drugs intake were the stronger predictors of mortality.
Keywords: cognitive impairments; comorbidity; functional status; longevity; nonagenarians.