The clinical dementia diagnosis has become more complex with increasing knowledge of the heterogeneity of the disorder and its different aetiological aspects. A clinical dementia population and a control group were investigated with the following aims: I. To study the CSF levels of tau phosphorylated at threonine 181 (P-tau), total tau (T-tau) and ß-amyloid1-42 (Aß42) in the different diagnoses. II. To study associations between dementia disorder, cobalamin and/or folate deficiency, and gastritis. III. To study the presence and severity of CT brain changes in different dementia diagnoses. IV. To investigate to what extent different biomarkers and disease history contribute to the diagnostics of clinical dementia.
I. CSF Levels of P-tau, T-tau and Aß42 were analysed with ELISA methods. Elevated CSF levels of P-tau were found in probable Alzheimer's disease (AD) patients compared with cognitively non-disturbed controls. Increased CSF T-tau, and decreased levels of Aß42 were found in both AD, mixed type of dementia, and vascular dementia (VaD) patients compared with the controls. Increased P-tau levels were more specific for AD pathology, but there was still an overlap with the controls, mixed dementia and VaD patients.
II. Serological markers for cobalamin and folate deficiencies, and for gastritis were assessed in patients with different dementia diagnoses. Hyperhomocysteinaemia were commonly found in dementia without predominance in any of the investigated categories. Low levels of serum cobalamin or blood folate rarely reflected the elevated Hey levels. A lack of association between serological markers for cobalamin and folate deficiencies and for gastritis was demonstrated.
III. A protocol for evaluation of the CT scans was used. Atrophy on the CT scans, although common in dementia, is an unspecific fmding in dementia of different backgrounds. White-matter changes and lacunes, indicating small-vessel disease, were common in dementia of different aetiologies. Dementia of mixed-type pathology was underestimated. More distinct criteria for this diagnostic category are warranted.
IV. Partial Least Squares Discriminant Analysis (PLS-DA) was used on a large number of variables covering cognitive and biological markers and disease history. There were good discriminations of subgroups of dementia from the controls. However, the included variables were not able to distinguish between the investigated groups, indicating that several clinical parameters used in diagnosing dementia are in fact observed across different subtypes of dementia.
It is concluded that there are no known biomarkers available that can provide a precise differential diagnosis of dementia. The clinical dementia diagnosis must still be based on a combination of a careful disease history, evaluation of risk factors, symptomatology, clinical findings, neurocognitive tests, blood analysis and other available methods such as CT and CSF markers.
Linköping: Linköpings universitet , 2004. , 60 p.
2004-03-12, Elsa Brändströms sal, Universitetssjukhuset, Linköping, 09:00 (Swedish)