Premium
O3‐02‐03: Is dementia incidence declining? a birth cohort analysis of the einstein aging study (EAS) cohort, 1993‐2014
Author(s) -
Derby Carol A.,
Katz Mindy J.,
Lipton Richard B.,
Hall Charles B.
Publication year - 2015
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2015.07.244
Subject(s) - dementia , medicine , cohort , incidence (geometry) , demography , psychosocial , gerontology , depression (economics) , cohort study , pediatrics , disease , psychiatry , sociology , optics , physics , economics , macroeconomics
Background:Without any change in dementia incidence, the number of people living with dementia worldwide is expected to rise to 81 million by 2040. However, improvement of education and better prevention and care of vascular risk factors, which are important factors related to dementia, could call into question these forecasts. Indeed, few recent studies have suggested a decline in dementia incidence. Thus, we aimed to determine whether there was a decrease in dementia incidence between the 90’s and the 2000’s and to assess the potential impact of major dementia risk factors on any change in incidence patterns.Methods:Analyses were based on participants from two cohorts, PAQUID (the 1990’s population, n1⁄41469) and 3C (the 2000’s population, n1⁄42104), aged 65 years and older and living in the same urban area around Bordeaux (France). These two populations were sampled ten years apart and each of them was followed for 10 years (1988-1999 and 1999-2010 respectively) with similar repeated cognitive and functional assessments and systematic screening of dementia. In order to account for change in mortality between the two periods, dementia incidence was compared between the two populations, for men and women separately, using multi-states illness-death models. Two types of diagnosis, one clinical and one algorithmic basedon theMini-Mental State Examination and the InstrumentalActivities of Daily Living, were used. Results: The 2000’s participants had a higher education level and were more medicated against vascular risk factors and depression than the earlier population. Using the algorithmic diagnosis, dementia incidencewas significantly lower in the 2000’s population than in the 1990’s population, forwomenonly (age-adjusted HR1⁄40.62; IC1⁄40.48 – 0.80). Differences in educational level, vascular risk factors and depression accounted to some extent for the reduction, but not entirely (women fully adjusted HR1⁄40.73; IC1⁄40.57 – 0.95). Men’s incidence remained stable. No decreased trends were found when using the clinical diagnosis for either sex. Conclusions: Our study provides further support for a decrease in incidence of dementia between the 1990s and the 2000s, but only for women. Changes in diagnostic boundaries mask this reduction.Adjustment for potential risk factors onlypartially explained this decrease.