A normative study of the Italian version of the CERAD neuropsychological battery in young‐old and old‐old participating in population‐based studies

Abstract Background Prevalence of dementia increases with increasing age. Although cognitive assessment is central to the diagnosis of dementia, normative data on neuropsychological tests in non‐selected elderly populations are limited and those available in the oldest‐old rare. Objectives: To report the distribution of scores and normative data on CERAD neuropsychological test battery in a large elderly population including a sizeable number of very old individuals participating in population‐based studies. Method Subjects included in the analyses were non‐demented participants in three Italian population‐based studies (Monzino‐80 plus, Varese Province, Lombardy; Heath and Anemia, Biella municipality, Piedmont; and Moltrasio, Como province, Lombardy) and in the first normative study on CERAD battery (Lugli 2002). Participants fulfilling DSM‐IV criteria for dementia were excluded. Tests were administered and scored by specifically trained psychologists following standardized criteria, the same in all four studies. The Italian version of CERAD showed high test‐retest reliability (r = 0.89). Multivariable linear regression models were fitted to investigate the association of age (in years), sex, and education (years of schooling) with performance on CERAD tests. Result The total population consisted of 2,501 individuals (mean age: 83 years; women: 64%; average education: 6.8 years; institutionalized: 69 [3%]), 802 of whom were young‐old (65‐79 years of age) and 1,699 old‐old (80+ years: 1,155 octogenarians, 470 nonagenarians, and 74 centenarians). The Table shows the distribution of scores (means, SDs, and 5 th percentiles) on each test by the whole population as well as by each age class. Better performance on all tests was significantly associated with lower age and higher education; women scored significantly better than men on Word List Learning and Recall, and men better than women on short‐Boston Naming Test and Constructional Praxis Copy and Recall. Refusal to continue and fatigue were the main reasons put forward by elderly participants for not completing the test battery. Conclusion The present study provides normative data on a neuropsychological test battery commonly utilized in the diagnosis of dementia in a large, non‐selected population of young‐old and old‐old. Performance on CERAD test battery declines with increasing age and lower years of education; performance on some tests were also influenced by sex.