Major and minor neurocognitive disorders in DSM-5: The difference between the map and the terrain

‘If the map differs from the terrain, believe the terrain’. Norse proverb.

DSM Digest

In day-to-day clinical practice, the categories of major and minor neurocognitive disorders (NCDs) in DSM-V (American Psychiatric Association, 2013a) seem to bear little resemblance to the conditions formerly described as respectively, dementia and mild cognitive impairment. Evidence of cognitive decline must comprise decrement in one or more cognitive domains as ascertained by subjective report or observation, and documentation of impaired performance by neuropsychological or other quantified clinical assessment. The key distinction of major neurocognitive disorders (née dementia) is of interference in independence in activities of daily living (ADLs), as opposed to minor disorders, in which they do not.

DSM-5 then cleaves both these two categories into an array of subtypes: Alzheimer’s disease, frontotemporal lobar degeneration, Lewy Body disease, vascular disease, traumatic brain injury, substance/medication misuse, HIV infection, prion disease, Parkinson disease, Huntington disease, other conditions, combinations and unspecified. All specifics of diagnosis are deferred to classification by subtype, as is whether behavioural disturbance is present, and severity (mild-moderate-severe).

Perhaps the subtypes are useful, although with the caveat that the usefulness of attempting to classify minor NCDs by subtype would seem potentially futile as these represent nascent rather than fully manifest conditions. Especially with mild NCD (née mild cognitive impairment), the research evidence has been conflicting, with multiple criteria and consequent different outcomes of progression or non-progression. To then attempt to subtype mild NCD seems to be overly optimistic of certainty in uncertain terrain.

Examining the subtype criteria, as noted by a previous commentary on DSM-5 (Carroll, 2013), biomarkers are generally missing. The criteria still seem rigidly focused upon primarily clinical observations of behaviour and cognitive impairment, without sufficient specification of patterns of neuropsychological impairment, neuropsychiatric features, neuroimaging or other clinical biomarkers. In international clinical practice, with all the possible variations of access to diagnostic testing and cost implications, this restriction to clinical criteria is partially understandable. Such constraints do not necessarily preclude cognitive profiles, mapping of neuropsychiatric features (depression, apathy and similar) or structural neuroimaging. Some may argue that clinicians do not have access to tests for biomarkers, and that even with access, requirements for such investigation are expensive and will impede diagnosis and thereby reimbursement of care. While reimbursement dependent on diagnostic criteria are admittedly a consideration within the framework of the DSM-5, we assert that as a tool for diagnosis the DSM-5 must have clinical utility that maps to the current understanding of the diagnostic features of the conditions, both simple and complex. It has been argued that requiring more clinical features in criteria will delay diagnosis, however, this can be handled by staging criteria, such as those used in Parkinson’s disease.

For Alzheimer NCD, the diagnostic criteria are almost entirely non-specific: insidious decline in one or more cognitive domains, with the concession of memory and learning decrement as a key feature. More specific neurological (CSF amyloid beta-42 biomarkers) neuropsychological (relative preservation of executive function), neuropsychiatric (depression as prodrome) and neuroimaging (hippocampal atrophy, Pittsburgh compound B amyloid imaging) criteria are absent (Sonnen et al., 2008).

In fronto-temporal lobar NCD, there is a separation into a behavioural variant and a language variant, eschewing the usual clinical classification of whether the language variant (née primary progressive aphasia) represents a fluent or non-fluent aphasia which can be useful as the non-fluent form may be quite severe, whereas the fluent form may masquerade as Alzheimer NCD (Rabinovici and Miller, 2010). Despite burgeoning neuroimaging research into these fronto-temporal lobar NCD showing that neuroimaging can aid in diagnosis (Rabinovici and Miller, 2010), DSM-5 offers a concession that frontal or temporal lobe atrophy may be evident on neuroimaging for probable fronto-temporal lobar NCD .

Similarly, the categories of Huntington and Parkinson NCD are mostly non-specific, except that the Huntington disease is based upon clinical diagnosis (motor dysfunction etc. detailed in the explanatory material) or risk, based on family/genetic history. There are specific descriptions in the explanatory material for each NCD in DSM-5 (diagnostic features, associated features supporting diagnosis, prevalence, prevalence, development and course, diagnostic markers, functional consequences, and differential diagnosis). It is unclear why the reasonable and helpful explanatory material was not generally incorporated fully into the diagnostic criteria. For example, both Huntington disease and Parkinson disease NCD commonly manifest prominent dysexecutive cognitive symptoms, and can be differentiated on motor dysfunction and neuropsychiatric features, as well as response to treatment (Ross and Tabrizi, 2011, Pagonabarraga and Kulisevsky, 2012). Similarly, neuroimaging (especially caudate atrophy) is very useful in diagnosis of Huntington disease, but not included as a criterion, and SPECT may be useful in Parkinson disease diagnosis (Ross and Tabrizi, 2011, Pagonabarraga and Kulisevsky, 2012).

Substance/medication-induced NCDs criteria are non-specific, comprising mainly of exhortations to exclude intoxication, delirium and other causes other than the putative substance. The lack of specific criteria seems to render this category relatively useless, as it does not even include sub-criteria for alcohol-related NCD, although the explanatory notes do include some detail on alcohol and methamphetamines.

In contrast, the more specific clinical criteria for Lewy Body NCD may a suggest a way forward with such diagnostic categories, including fluctuant cognition, visual hallucinations, Parkinsonism, and supplementally, REM sleep behaviour disorder plus neuroleptic sensitivity. However, even this category could be improved by including fall, autonomic dysfunction and frequent delusions, as noted in the explanatory section following the criteria (American Psychiatric Association, 2013a). The continuum of the synucleionopathy and associated neuropsychiatric manifestations as the basis of an overlap of Lewy Body NCD with Parkinson NCD could be further emphasised (McKeith et al., 2004).

The categories of NCD in DSM-5 appear designed for classification of conditions towards recording health usage data, rather than as a useful framework for clinical application, or indeed research criteria. A review of the FAQ’s for clinicians on the DSM-5 website leaves the reader with the impression that the DSM-5 is primarily a tool for clinicians to use in their interactions with insurers and for other ‘health transactions’(American Psychiatric Association, 2013b), although the DSM-5 manual states the main purpose is for diagnosis, case formulation and treatment planning (American Psychiatric Association, 2013a). Failure to include specific neurological, neuropsychological, neuropsychiatric, neuroimaging and the burgeoning CSF biomarker information is ultimately unhelpful for the clinician for diagnosis, case formulation and treatment planning, and potentially unhelpful for research diagnoses. It is not clear why more specific criteria, including the explanatory material appended to each set of diagnostic criteria could not be integrated in the overall framework, at least including simple neuropsychological, neuropsychiatric and neuroimaging investigations. The concepts of major and minor NCD are useful to conceptualise a continuum of cognitive impairment due to neurodegeneration or brain injury, but attempting to specify whether mild NCD is due to a specific cause seems premature. The map provided by DSM-5 seems very different from the terrain of cognitive impairment/dementia seen by clinicians and researchers. Unlike the hapless tourists whose GPS navigation leads them to a cliff’s edge, we continue to believe the terrain.