Variations in case definition affect prevalence but not outcomes of mild cognitive impairment

Neurology. 2003 Nov 11;61(9):1179-84. doi: 10.1212/01.wnl.0000089238.07771.c7.

Abstract

Objective: To examine the prevalence estimates and 5-year outcomes of various case definitions of mild cognitive impairment (MCI).

Methods: The authors examined 1,790 adults 65 years of age or older who completed neuropsychological and clinical assessments in the Canadian Study of Health and Aging, a 5-year, representative, prospective cohort study.

Results: The most commonly used case definition of MCI yielded a population prevalence estimate of 1.03% (95% CI 0.66 to 1.40). Eliminating the requirements for subjective memory complaints and intact instrumental activities of daily living (IADL) increased the prevalence to 3.02% (CI 2.40 to 3.64). Five-year outcomes, including the risk of death, institutionalization, and dementia, were not distinctly different for various case definitions of MCI, but all were at increased risk of institutionalization (RR 2.3 to 5.2) and dementia (RR 9.3 to 19.7). Regardless of the case definition, most people with MCI developed dementia, chiefly Alzheimer disease (AD). Still, for each case definition, almost one third were considered to have no cognitive impairment after 5 years.

Conclusions: Memory complaints and intact IADL may be unnecessary requirements for a case definition of MCI in population-based samples. The MCI criteria identify people at increased risk of AD, but the potential for improvement of a substantial proportion of those with MCI needs to be acknowledged.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Canada / epidemiology
  • Cognition Disorders / diagnosis*
  • Cognition Disorders / epidemiology*
  • Cognition Disorders / mortality
  • Dementia / epidemiology
  • Disease Progression
  • Female
  • Humans
  • Institutionalization / statistics & numerical data
  • Male
  • Neuropsychological Tests / statistics & numerical data
  • Outcome Assessment, Health Care / statistics & numerical data*
  • Prevalence
  • Risk