Article Text


Orientation agnosia in pentagon copying
  1. S Della Sala1,
  2. O Turnbull2,
  3. N Beschin3,
  4. M Perini3
  1. 1Department of Psychology, King's College, University of Aberdeen, Aberdeen AB24 2UB, UK
  2. 2School of Psychology, University of Wales, Bangor, Wales, UK
  3. 3Department of Rehabilitation and Department of Neurology, Gallarate Hospital, Italy
  1. Correspondence to:
 Professor S Della Sala;
  1. T A Ala4
  1. 4Center for Alzheimer Disease and Related Disorders, Southern Ilinois University Medical School, P O Box 19643, Springfield, IL 62794-9643, USA

    Statistics from

    Ala et al1 have recently claimed that patients with Alzheimer's disease (AD) were less likely to make errors on the pentagon copying component of the mini mental state examination (MMSE) than patients affected by Lewy body (LB) dementia. Using (standard and modified) MMSE scoring criteria they found normal copies in 16 of 27 patients with AD but in only two of 13 patients with LB. Although this difference was non-significant,1 they concluded that this feature might be useful in differential diagnosis. It is of some interest that the MMSE scoring criteria employed by Ala et al1 did not regard errors of rotation as abnormal. In the light of recently identified visuospatial deficits which affect the domain of orientation, it may well be that the criteria of Ala et al for a “normal” copy are inappropriate. This has implications for the claimed diagnostic power of their measure.

    Patients with orientation agnosia are able to recognise objects, but have a selective visuospatial impairment, in that they have lost knowledge of object orientation.2 The most striking feature of this disorder is shown in the patients' drawing of objects, where they rotate their copy by 90 or 180 degrees relative to the original. This previously underinvestigated deficit is far from rare. In one study, seven of 63 (11%) patients with stroke grossly rotated their copy of the MMSE.3 All had a right hemisphere lesion and showed some evidence of left visuospatial neglect. In a second study4 16 of 240 (7%) of a series unselected by pathology showed the same sign, but without clear lateralising features.

    To better investigate this issue in degenerative disease, we reviewed the records of a consecutive series of 134 patients with AD. All had been given the MMSE, which includes the interlocking pentagon figure (aligned horizontally) as a copy task, and a battery of neuropsychological tests. Errors in copying the geometrical figure were systematically graded. In general agreement with the findings of Ala et al, 57 of 134 (43%) drew unrecognisable figures, adding or omitting details. Three showed signs of closing in, and seven (5% of the sample) made clear rotation errors, misplacing one or both pentagons by 90, 120, or 180 degrees relative to the original. These seven patients did not differ from the overall sample in terms of general severity of dementia, as measured by the MMSE total score, and by daily activity scales. Eight patients (6%) presented with signs of visuospatial neglect,5 but patients with neglect and orientation agnosia did not overlap in this sample, suggesting that orientation agnosia should be considered an independent sign. Our data thus suggest that orientation agnosia is a significant visuospatial deficit in degenerative disease, which is not identified as such in the study of Ala et al by virtue of the criteria by which they define an abnormal performance.

    In figure 2 of the paper of Ala et al, it is clear that some copies, scored as errorless, showed clear signs of orientation agnosia (AD cases 3, 9, 15, 18, 19, and 24). If these errors are considered, the claimed difference between AD and LB dementia disappears, confirming the results of previous studies which showed no significant difference in drawing or constructional abilities between AD and LB dementia.6

    Figure 1

    Interlocking pentagon models and copies from the mini mental state examinations (MMSEs) of six patients with AD. For each pair, the model shown to the patient is on the left and the patient's copy is on the right. Each patient was included among the patients with AD reported in our original article,1 identified by the same code. Orientation to the bottom of the page and relative sizes within each pair are preserved. Each model-copy pair has been reduced in size for this figure.


    Authors' reply

    We were very pleased to see the interest that Della Sala and colleagues have shown in our study.1 Orientation agnosia is certainly a fascinating subject, and we applaud their research into its neuropsychology. We think, however, that their analysis of our our study has several weaknesses.

    Firstly, because we had not included the original interlocking pentagon models with the copies in our figures, they incorrectly inferred that six of our patients with AD made rotational errors. As mentioned in our results,1 many different printed models of the interlocking pentagons were used in our study. To clarify for your readers exactly what the patients were shown to copy and how they performed, Figure 1 presents the models and the patients' copies for AD cases 3, 9, 15, 18, 19, and 24. To allow the study of rotational changes, the exact orientation to the bottom of the page is preserved for each model and for each copy. Although all have been reduced for publication, the relative sizes of the models and the copies are also preserved. By our inspection none of the questioned cases exhibits any significant rotational error. We also studied the copies of the remaining patients with AD and the patients with DLB in our study, and we could not form any conclusions about rotational errors with either group.

    It should also be noted that we specifically used the original published criteria2 for grading the copies of the interlocking pentagons in the MMSE. These criteria state that “tremor and rotation are ignored” (table 1 of our article1).

    The diagnoses of our patients were neuropathologically proved. It would be interesting to know if the diagnoses of the 134 patients with AD that Della Sala and colleagues mention were also neuropathologically proved. If their patients were diagnosed clinically, we would not be surprised if some patients with DLB were included in the 134. Perhaps many of those who had the orientation agnosia were actually patients with DLB who did not exhibit the classic DLB clinical symptoms of fluctuation, visual hallucinations, and parkinsonism.

    Also, the study3 that Della Sala and colleagues cite as evidence that there is “no significant difference in drawing or constructional abilities between AD and LB dementia” cannot be easily compared with ours because of the different severities of the patients involved. Based on MMSE scores, at least one half of the patients in that study were more severely impaired (median AD MMSE 10.5, DLB 9.0) than any of the patients in our study (minimum MMSE 13). That study does help confirm that constructional abilities worsen as the dementia worsens, for as we mentioned in our results,1 none of the additional cases of AD and DLB in our brain bank with initial MMSE scores less than 13 had acceptable pentagon copies.

    Finally, albeit a minor detail at variance with Della Sala and colleagues, we did report that the DLB group copied the pentagons significantly less accurately than the AD group (p=0.002 using the original criteria).


    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.