The occurrence of Lewy bodies has a prevalence rate of 2%–9% in elderly people1 and dementia with Lewy bodies (DLB) accounts for 12%–27% of cases previously diagnosed as dementia of the Alzheimer type (DAT).2 3 The core features of DLB are fluctuating cognition with pronounced variation in attention and alertness, recurrent visual hallucinations, and spontaneous parkinsonian signs; probable DLB requires two of these features. There is considerable overlap between DLB and DAT,4 but there have been only a few comparative neuropsychological studies. Various neuropsychological issues were addressed in the papers by Lambon Ralphet al 5 (this issue, pp 149–156) and Calderonet al 6 (this issue, pp157–164) who disclosed some valuable insights that merit closer inspection. Clinicians and researchers will also find a useful tabulation of recent findings in the paper by Lambon Ralphet al 5.

Previous studies have suggested that visuoperceptual problems are salient in DLB, but this evidence came from measures that represent a complex of abilities. The papers here report that basic figure-ground discrimination was worse in one DLB sample, whereas the other DLB sample instead had problems identifying silhouettes of real versus non-real objects. More complex visual tasks produced similar deficits in both DLB and DAT groups. Perhaps the most interesting finding was that the DLB groups in both studies showed marked impairments when identifying fragmented letters. This task has minimal cognitive load, and was unaltered in the DAT samples, so it may be especially promising for differential diagnosis and treatment evaluations.

Attention may be a second area of weakness in DLB3 which, together with the related areas of working memory and executive function, influences adaptive functioning and performance on formal tests. Calderon et al 6 have confirmed that patients with DLB show widespread difficulties in this domain. Whereas patients with DAT showed set shifting, letter fluency, and selective attention deficits, the DLB group had additional problems in sustained and divided attention tasks.

The third contribution made by these two papers concerns long term episodic memory and semantic memory, two major hallmarks of DAT. One important finding is that delayed recall represents one of the apparently few areas in which patients with DAT have a disproportionately greater weakness than their DLB counterparts, even though the patients with DLB do show substantial deficits on recall and recognition tasks. Category fluency and picture naming were also substantially but equally impaired in both dementias, so semantic memory itself does not distinguish these two disorders. The more marked visuoperceptual problems in patients with DLB seem, however, to exacerbate their semantic memory performance in some tests when presentation uses the visual modality.

Calderon et al 6 also make the interesting point that visual hallucinations in DLB may be related to the combination of impaired visuoperception and fluctuating attention. Cholinergic deficits are more profound in DLB and this too may be associated with both attentional difficulties and hallucinations. These ideas, and the various lines of evidence presented, will undoubtedly guide future research on the behavioural and neurobiological sequelae of the DLB syndrome.