Date of Award


Degree Type


Degree Name




First Advisor

Rourke, Byron P.,


Psychology, Clinical.




Neuropsychological CERAD data from 960 patients diagnosed with probable Alzheimer's disease (AD) and 465 non-demented controls were examined, to investigate whether previously identified subgroups of patients could be replicated (i.e., externally validated) in this sample, and if so, whether these subgroups were specific to AD. Patient and comparison group data were subjected to separate Q-factor and cluster analytic procedures. Consistent with past research, three qualitatively distinct patient subgroups were reliably identified across classification methods: Subgroup 1 (LAD; n = 312) was characterized by severe naming impairment yet border normal figure copying skills; Subgroup 2 (LAD; n = 247) displayed average naming ability within the context of moderately impaired copying performance; Subgroup 3 (GAD; n = 161) evinced profound anemia and constructional dyspraxia. Members of Subgroup 1 were older and less educated than those of the other two subgroups. Subgroup 2 was the highest functioning group in terms of functional levels and overall dementia severity; Subgroup 3 was most impaired in these respects. Analyses of the control data revealed two clusters of individuals that did not resemble the patient subgroups, suggesting that the above 3-subgroup classification is unique to AD. Longitudinal patient data from 7 follow up examinations were also examined. Initial patterns of performance were generally stable across time for the LAD and GAD subgroups. Such data were less consistent for the RAD group, suggesting additional, and as yet unexplained, heterogeneity within this subgroup. Cross-sectional analyses revealed the presence of members from all three subgroups across mild, moderate and severe stages of the disease, regardless of the stratification method employed i.e., CDR or MMSE). For the LAD and GAD groups, longitudinal and cross-sectional analyses involving naming/copying discrepancy scores revealed findings in accord with those predicted by an earlier proposed soup-specific progression model. Although the results of such analyses for the RAD group did not support the model when the group data were averaged, it should be noted that a subset of RAD members did perform in a manner consistent with the model. The results of this investigation support an integration of the Subgroup and Stage models in conceptualizing this most complex disorder.Dept. of Psychology. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1995 .F57. Source: Dissertation Abstracts International, Volume: 61-09, Section: B, page: 4980. Adviser: Byron P. Rourke. Thesis (Ph.D.)--University of Windsor (Canada), 1998.