This abstract was recently published on PubMed (see pubmed.gov). The researchers used a test called the Boston Naming Test. The test was given to those with probable DLB and those with probable AD. Errors were classified into 5 types. "AD patients' overall performance on the BNT was significantly worse than DLB patients. In terms of error types, DLB patients made significantly more visuoperceptual errors while AD patients made significantly more semantic errors. ... Results suggest that error analysis of BNT responses may be useful in distinguishing between patients with DLB and AD."
Robin


Archives of Clinical Neuropsychology. 2007 Aug 1; [Epub ahead of print]

Boston naming performance distinguishes between Lewy body and Alzheimer's dementias.

Williams VG, Bruce JM, Westervelt HJ, Davis JD, Grace J, Malloy PF, Tremont G.
Brown Medical School, Providence, RI, United States; Rhode Island Hospital, Providence, RI.

Although naming impairment is common among persons with dementia, little is known about how specific error types on naming tasks may differ between dementias. Recent research has suggested that persons with dementia with Lewy bodies (DLB) have more visuospatial/visuoperceptual dysfunction than those with Alzheimer's disease (AD), which may impact their ability to correctly perceive and name objects. Our retrospective study evaluated the presence and frequency of error types among patients with DLB and AD on the Boston Naming Test (BNT). Errors on the BNT were classified into five types (i.e., visuoperceptual, semantic, phonemic, no response, and other), and performance was compared among 31 probable DLB patients and 31 probable AD patients matched for age, gender, education, and overall dementia severity. AD patients' overall performance on the BNT was significantly worse than DLB patients (p<.05). In terms of error types, DLB patients made significantly more visuoperceptual errors (p<.05) while AD patients made significantly more semantic errors (p<.001). Logistic regression revealed that the number of visuoperceptual and semantic errors significantly predicted group membership (p<.005), with an accuracy of up to 85%. Results suggest that error analysis of BNT responses may be useful in distinguishing between patients with DLB and AD.

PubMed ID#: 17681741



Here's a useful list of neuropsych tests:
http://www.brainsource.com/nptests.htm

The Boston Naming Test is described as: "Assessing the ability to name pictures of objects through spontaneous responses and need for various types of cueing. Inferences can be drawn regarding language facility and possible localization of cerebral damage."

Thanks to fellow Forum member Patty, I got a chance to read this article on the use of the Boston Naming Test (a common neuropsych test) to differentiate between DLB and AD. The article refers to research by Simard et al. (2003) suggesting that DLB might present in two forms.

Here are some excerpts:

"Postmortem, patients with DLB show alpha-synuclein-positive Lewy bodies in the brainstem and cortex as well as nigral Lewy body degeneration. Patients with DLB may also have concurrent Alzheimer’s disease pathology, making antemortem diagnostic accuracy challenging. The cognitive presentation of DLB and AD patients can be similar, perhaps in part due to the frequency of overlapping pathology."

"Utilization of McKeith et al.’s consensus criteria for the diagnosis of DLB has high specificity and can lead to diagnostic accuracy similar to that for AD (Geser et al., 2005). However, these criteria are thought to have variable and low sensitivity (Geser et al., 2005; McKeith et al., 2004), suggesting that further refinement of criteria and supportive data is needed."

"Pattern analysis of neuropsychological data may be one way of increasing diagnostic accuracy. It is now well accepted that patients with DLB demonstrate greater visuospatial and attentional deficits than patients with other types of dementia on standard neuropsychological tests."

"Simard et al. (2003) suggested that DLB might present in one of two forms; the anterior group, who present with extrapyramidal signs and working memory or attention deficits, and the posterior group, who present with visual hallucinations and have deficits in visual perception and construction. In their study, DLB subjects with visual hallucination had greater visual-perception errors on Benton’s Judgment of Line Orientation than both DLB subjects with predominant Parkinsonian features and AD subjects. This study was limited, however, by the small number of subjects."

"Numerous studies have described the nature and extent of naming deficits in AD. Research has suggested that the naming deficit in AD is related to impaired semantic and lexical processing. ... Qualitative aspects of naming performance have not been previously studied in DLB. It is possible that naming deficits observed in DLB may be related at least in part to problems with perceptual accuracy, leading to different types of errors than those seen in AD."

"Research has suggested that qualitative analysis of the BNT, including error analysis, can be useful in the clinical neuropsychological evaluation." (BNT = Boston Naming Test)

"As predicted, we found that AD patients performed worse on the BNT than patients with DLB. Given the breakdown of semantic networks described in AD and the importance of these networks in confrontation naming, this finding was not surprising. An evaluation of error type (semantic, visuoperceptual, phonemic, no response, perseverative) revealed that AD patients made significantly more semantic errors on the BNT than DLB patients. In contrast, we found that DLB patients made significantly more visuoperceptual errors on the BNT than patients with AD. Other error types did not significantly differentiate the groups. It is possible that the disproportionate visuospatial dysfunction described in DLB patients contributes to their ability to perceive objects accurately, leading secondarily to errors in naming."

"In conjunction with other clinical data, naming errors may provide clinicians with additional information regarding the differential diagnosis of AD and DLB. Additionally, the current study supports systematic investigation of error types, given the relative ease in which a clinician can analyze error types in their review of a BNT protocol."

"(If) a patient has three or more perceptual errors on BNT, they are unlikely to have Alzheimer’s disease (specificity = .91)."

Within the DLB group, "patients with visual hallucinations made significantly more visuoperceptual errors than those without visual hallucinations." This finding would support "Simard et al.’s (2003) assertions that DLB may present in one of two forms – the anterior group may have predominant working memory dysfunction while the posterior group (who present with visual hallucinations) will have predominant visuoperceptual dysfunction. ... However, this finding should be interpreted with caution given the small number of patients without visual hallucinations."