The challenge of estimating the prevalence of dementia in the elderly.

AIM: To examine the validity of the estimation of the prevalence of dementia in the elderly based on long-term care insurance (LTCI) data. METHODS: Japanese LTCI data among frail elders over 65 years of age (193 men, 360 women) from a central area of Osaka prefecture in 2005 was analyzed using the dementia discrimination criteria of a dementia autonomy level higher than level 2. Of those, no cases with a prior diagnosis of dementia were reexamined to ascertain dementia-equivalent status. RESULTS: Of the 70 reexamined cases (25 men, 45 women), 6 men and 16 women scored as dementia-equivalent. In addition, 7 men and 7 women experienced severe sequelae of cerebrovascular disease (CVD). Two men and 8 women showed symptoms equivalent to mild cognitive impairment (MCI). Ten men and 14 women showed no dementia/MCI. Based on the sum of subjects previously diagnosed with dementia and those with symptoms equivalent to dementia upon reexamination, the estimated prevalence of dementia in subjects over 65 years in the study region was 5%. CONCLUSIONS: Assuming a 50% capture rate of dementia based on the LTCI data (N = 553) from a review of previous research, the estimated prevalence of dementia was increased to 10% after incorporating the results of this study. The prevalence of elderly dementia according to LTCI data should be estimated based on a determination of dementia-equivalence in terms of the level of care required: dementia, equivalent to dementia (i.e., requiring daily care similar to that for dementia), and severe sequelae of CVD. Furthermore, the capture rate should be carefully considered.