Subscribe to eTOC



Like many a neurologist, I cannot imagine a better supporting opinion than that of Dr. C. Miller Fisher, who has contributed so greatly to our neurological knowledge-base. So, I am pleased at his welcome of the re-examination of the DSM definition of dementia, and also appreciate that similar concerns regarding DSM have been raised during the past decade.

As Drs. Growdon, Miller, and Narrow also point out, there seems to be increasingly broad agreement that memory impairment should not be a sine qua non for diagnosis of dementia. In Korsakoff's syndrome, there are often executive or other involvements beyond simply memory, thus sometimes satisfying various multi-domain definitions of dementia.

Whether a single deficit such as aphasia should, if severe enough, satisfy the definition of dementia is of course an important semantic and nosologic issue – for example, many of us would classify the disorder primary progressive aphasia as a dementia.

The custom in the past has indeed been to reserve the term dementia – for some broader impairment of intellect resulting in loss of functional abilities – which is why dementia might be required to comprise impairment of at least two domains, rather than a single domain. However, often with a severe enough deficit in one domain, brain injury is sufficiently diffuse to directly, or through disconnection, cause significant deficits in other domains – thus, the issue may ultimately devolve into one of severity of affliction in each domain. This is already increasingly surfacing, as we consider the entity mild cognitive impairment, which is recognized by neither DSM nor ICD9/10.

Lawrence S. Honig, MD, PhD

Columbia University, New York, NY