Visuospatial functions

The term ‘visuospatial functions’ is used to describe a wide range of cognitive functions measured by several different tests. The unifying theme of the term is that these tests rely on visual perception, visual representation, or a visually guided response. The term can be misleading, as it lumps together very different cognitive processes, perhaps too broadly. Ungerleider and Mishkin [49] proposed two distinct cortical processing pathways of visual information, the occipito-temporal pathway (the ‘what is it’ pathway) and the occipito-parietal pathway (the ‘where is it’ pathway). It has since become clear that visual processing is performed in several widely distributed cortical areas [50, 51] and that spatial cognition is very different from the perception of visual form and object categorization [50]. In the clinical neuropsychology literature, many ‘visuospatial tests’ are tests of the ability to draw complex figures (visual construction). Unfortunately, the performance of such tasks demands attentional/executive control as well as fine-motor control. In addition, such tests do not usually assess spatial cognition well. These issues are problematic in the research on PD-D, as it has been difficult to interpret findings. An overview of studies on PD-D in which visuospatial functions were reported is given in Table 4.2.






Pentagon drawing [52]




DRS construction [23]


Rey-Osterrieth complex figure test:

Copy [17]


Copy [20]



Poppelreuter test [17]


CERAD neuropsychological battery copy test [21]




Construction [15]





Perception [15]





BVRT: matching [16]



Rosen drawing test [16]



ADAScog: construction [26]


RBANS: visuospatial [24]



Benton visual retention test (copy) [22]




Average effect-size


-0.35, -0.31a



Numbers represent Cohen's d, relative to PD-D.

aWeighted meta-analytic average effect size, according to sample size; p = 0.009.

CAMCOG, Cambridge Cognitive Examination [27]; DRS, Dementia Rating Scale; BvRT: Benton visual Retention Test [53]; ADAScog, cognitive part of Alzheimer's Disease Assessment Scale [54]; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status [28].

PD-D patients have more pronounced visuospatial dysfunctions than patients with AD, with a medium effect size (Cohen’s d = 0.44). It is also evident that according to the fixed-effect metaanalysis, visuospatial functions are statistically significantly less severely affected in PD-D than in DLB, approaching a medium effect size (d = -0.35, average weighted fixed effect = -0.21, p =

0.009) (see Fig. 4.2 for a graphical depiction).

Some of the studies used perception tasks that required no eye-hand coordination or motor control, and less executive control; for instance there was a larger deficit in DLB compared with PD-D on the Cambridge Cognitive Examination (CAMCOG) perception task relative to the CAMCOG construction task [15]. Patients with PD-D also performed better than patients with DLB on other perception tasks [16, 17].

The study by Mosimann et al. [15] is probably the most comprehensive study published to date regarding visuospatial functions in PD-D compared with other patient groups. In addition to the CAMCOG results listed in Table 4.2, the study included tests of visual object-form perception, space-motion perception, and visual discrimination of various lengths and sizes. DLB, AD, PD, and HC groups were compared with PD-D, and separate analyses were conducted for patients with and without hallucinations. No statistically significant differences between DLB and PD-D patients were found on any tasks, and both these groups performed worse than patients with AD. Patients with visual hallucinations performed worse than non-hallucinators in both the DLB and

Fixed-effects meta-analysis

Fig. 4.2 Fixed-effects meta-analysis: DLB versus PD-D, visuospatial functions. Studies as in Fig. 4.1 plus Cormack et al. [52].

PD-D group. While the authors concluded that the performance of the PD-D and DLB groups was worse in the object-form perception tasks than in the space-motion perception task, the very low error scores in the AD, PD, and HC groups indicate a probable floor-effect that is likely lead to underestimation of the real differences. Thus, it is difficult to draw conclusions regarding relative impairment in subdomains of visuospatial functioning in PD-D based on this study.

The previously mentioned study on PD-D, AD, DLB, and PSP by Aarsland et al. [29] also showed that in the groups with mild to moderate dementia the DLB patients performed worse than the PD-D group on the Mattis DRS construction (drawing) tasks. No difference was found in the severe dementia groups.

In conclusion, PD-D is associated with visuospatial dysfunctions that are more severe than in AD but very likely less severe than in DLB.

< Prev   CONTENTS   Source   Next >