Cause of death in mild cognitive impairment: a prospective study (NEDICES).

BACKGROUND AND PURPOSE: Previous studies have reported the occurrence of increased mortality rates among individuals with mild cognitive impairment (MCI), but possible links between MCI subtypes and cause-specific mortality need to be explored. To examine short-term mortality (5 years), long-term mortality (13 years) and cause-specific mortality of individuals over 65 years of age suffering from MCI compared with cognitively unimpaired individuals in the Neurological Disorders in Central Spain (NEDICES) cohort.

METHODS: Mild cognitive impairment was classified using standardized psychometric and functional assessment in accordance with diagnostic convention. Cox's proportional hazards models, adjusted by sociodemographics and comorbidity factors, were used to assess the risk of death at 5 and 13 years of MCI subtypes compared with a reference group of older people without cognitive impairment (N = 2329). Causes of death were obtained from the National Population Register of Spain.

RESULTS: There were 1484 deceased individuals at 13 years. MCI subtypes were defined as amnestic single domain (N = 259), amnestic multiple domain (N = 197) and non-amnestic (N = 641). After adjusting for covariates, only the amnestic multiple domain MCI subtype showed an increased hazard ratio (HR) for mortality at 5 years versus the reference group. However, the HR for mortality at 13 years was increased for all MCI subtypes. The HR by MCI subtype was 1.19 in the non-amnestic subtype (95% CI 1.05-1.36), 1.31 in the amnestic single domain subtype (95% CI 1.10-1.56) and 1.67 in the amnestic multiple domain subtype (95% CI 1.38-2.02). In terms of cause-specific mortality, the chance of death from dementia was statistically higher in all MCI subtypes.

CONCLUSION: Amnestic multiple domain MCI showed the greatest risk of mortality in comparison with other MCI subtypes at different intervals. Dementia was the only cause-specific mortality that was increased in MCI individuals.