Epidemiology of Alzheimer's disease. Issues of etiology and validity.

This thesis concerns the epidemiology of Alzheimer's disease (AD) and some aspects of the validity of such studies. AD is a common and chronic dementing disorder among elderly people. Due to the lack of treatment and to the invalidating nature, the social impact of this disease is high in all the societies in which the proportion of elderly is increasing. Three studies on AD etiology have been performed. The first is a case-control study on early-onset AD and a wide range of putative risk factors. The cases were gathered from a clinical study on AD carried out in Italy. The information on the exposure obtained from a next-of-kin of 116 cases was compared with the information similarly collected from the next-of-kin of 116 hospital and 97 population controls. The other two etiological studies deal with late-onset AD and are a prevalence study on sociodemographic variables and a case-control study on selected putative risk factors. These two studies were performed within a population-based study on ageing and dementia that is ongoing in Stockholm, Sweden. The study on sociodemographic variables included 116 AD cases among 1810 people. The case-control study compared the information obtained by the informants of 98 AD cases and 266 controls. The main results of these three investigations are: (1) The prevalence of AD increases with age, even in advanced ages. (2) The prevalence of AD does not vary by gender and education. (3) The main risk factor for both early- and late-onset AD is the familial aggregation of dementia (relative risk of 2.6 and 3.2, respectively). (4) A second risk factor for early-onset AD may be the advanced age of the mother at index delivery, but this result needs confirmation. No other risk factors reported by others emerged in our study. (5) High relative risks were found for alcohol consumption and manual work in late-onset AD. Manual work could be an indicator of occupational exposures as well as life conditions or life habits. Although both these results may be affected by bias, the results are provocative for future research. Three validation studies were carried out on three different aspects: diagnosis, case ascertainment, and exposure assessment. The first study investigated the reproducibility of AD diagnosis according to the DSM-III-R diagnostic criteria. The diagnoses made by the examining physicians were compared with the diagnosis made independently by another clinician on the subjects' clinical records.(ABSTRACT TRUNCATED AT 400 WORDS)