[Mortality among the elderly in Tuscany in the period 1987-2003].

Aim: to assess cause-specific mortality and its temporal trend in Tuscan elderly residents.

Design: descriptive epidemiologic study based upon death certificates, collected and registered since 1987 by the Tuscan Regional Mortality Registry (RMR) according to standardized procedures.

Methods: Major groups of causes: temporal trends are described calculating annual truncated age-adjusted mortality rates (based on age-specfic rates subdivided in three classes: 75-79; 80-84 and > or =85; standard: European population) and estimating annual percent changes (EAPC, Estimated Annual Percent Change) using ]oinpoint regression models. Most frequent specific causes: number of deaths and truncated age-adjusted mortality rates (based on age-specific rates subdivided in three classes: 75-79; 80-84 and > or =85; standard: European population) are compared between the first and the last quinquennium (1987-1991 and 1999-2003) by percent change.

Results: during 1987-2003, an average of 26667 annual deaths (65% of total) occurred in persons aged 75 and over in Tuscany. In the same period the number of elderly residents increased (2003 vs 1987: men +43.6%; women +41.5%) with a consequent increase in number of deaths (2003 vs 1987: men +13.8%; women +15.9%). The truncated age-adjusted rates for all mortality causes decreased (EAPC = -1.35% in males; EAPC = -1.41% in females) while an increase was registered for few specific causes as Alzheimer's disease, senile dementia, arterial hypertension and lung cancer.

Conclusions: the decreasing mortality trend observed in persons aged 75 and over in Tuscany is consistent with similar trends in other developed countries. The opposing trends for few specific causes of death need to be further investigated.