Search

The first section of this paper describes methodology and major cognitive outcomes of the Canberra Longitudinal Study (CLS). This community study of 1045 Australians aged 70 years or over commenced in 1990. Participants were reassessed on three subsequent occasions. Its major themes were investigations of prevalence of dementia and depression, risk factors, inter-individual variability and instrument development. Over 60 papers arising from the study have been published. The second section of this paper describes the use of a Factor of Curves (FOC) latent growth model to examine the existence of a common factor responsible for age-related deterioration in cognitive and non-cognitive processes. This analysis is a logical progression in a series of investigations using the Canberra Longitudinal Study into risk factors and correlates of cognitive change using structural equation modeling techniques. The FOC model is described and is used to explore the nature of an hypothesized common factor and to determine its relationship with age, gender, education, pre-morbid intelligence and to the ApoE genotype. Latent growth models were developed for each of reaction time, Symbol Letter Modalities Test (SLMT), Grip strength, self-reported Sensory disability and memory from three waves of data. Second-order latent level slope factors were established based on the individual factor growth curve models. Although a common factor model could be fitted to the data, there is little support that it represents a single common cause. [ABSTRACT FROM AUTHOR]

Personality changes are included in the DSM-III-R and ICD-10 criteria for diagnosis of dementia. However, until quite recently, dementia research has emphasized cognitive decline, with little interest being shown in personality change. Recent studies looking at personality changes in dementia have usually used small clinical samples. In the study reported here, such changes have been assessed in a community sample of persons over 70 years. Information was obtained from a close relative or friend. A single dimension of personality change was found. There was a moderate but significant correlation between personality change and cognitive change. Reports of personality change were also correlated to neurotic symptoms in the informant. Personality change was unrelated to the gender of the informant, the relationship of the informant to their relative or whether they lived with their relative. Also, there was no association between personality change and the elderly subject's age, gender or level of education. Through logistic regression, the neurotic symptom score of the informant was found to be the best predictor of reported personality change. This suggests that caution may be needed in interpreting information about personality change obtained from informants. [ABSTRACT FROM AUTHOR]