Mortality and morbidity hazards associated with cognitive status in seniors: a Canadian population prospective cohort study.

INTRODUCTION

Although cognitive impairment is widely accepted as a leading indicator of dementia, influences of cognitive status on incident dementia and mortality remain unclear. The present study investigated the morbidity hazard associated with cognitive impairment and the mortality hazard associated with dementia in comparison to cognitively intact seniors.

METHODS

A population-based sample of 2914 seniors with clinically diagnosed cognitive status at Wave I (1991-1992) of the Canadian Study of Health and Aging (CSHA) were followed-up 5 years later (1996-1997). At Wave I, there were 921 cognitively intact, 861 cognitively impaired but not demented (CIND), and 1132 seniors with dementia, respectively. The primary outcome measures 5 years later were being cognitively intact, CIND, dementia and death. Kaplan-Meier estimates, log-rank tests, and Cox's proportional models were used in the analyses.

RESULTS

Respondents with CIND at Wave I were 2.191 times (95%CI 1.706-2.814) more likely to have dementia 5 years later than cognitively intact seniors. After adjusting for confounding socio-demographic and health status factors, the odds ratio was reduced to 2.147 times (95%CI 1.662-2.774), but remained significant. Respondents with CIND had a mortality rate 1.869 times (95%CI 1.602-2.179) and seniors with dementia 3.362 times greater (95%CI 2.929-3.860) than that of seniors who were cognitively intact. After controlling the confounders, the odds remained significant at 1.576 (95%CI 1.348-1.843) for CIND respondents and 2.415 (95%CI 2.083-2.800) for seniors with dementia.

DISCUSSION

CIND increases both the risk of dementia and mortality. Early intervention with CIND is warranted to reduce both dementia incidence and mortality.