Classification of the Dementias

Richard Camicioli, Kenneth Rockwood

Overview

Dementia is a critical public health problem worldwide, especially as populations age. As a consequence, physicians dealing with a range of other age-associated problems can expect to see such problems in patients with dementia. Dementia diagnosis is evolving. Here, we review recent developments in the syndromic approach to dementia, and then highlight aspects of its differential diagnosis.¹ Dementia generally has been defined in terms of multifocal cognitive impairment sufficient to impair function and typically is distinguished from cognitive decline of lesser severity and impact. Since the seventh edition of this book was published in 2010, several changes in dementia conceptualization have been proposed. These include a new lexicon and several new sets of diagnostic criteria,²^–⁶ as well as growing support for the claim that “subjective cognitive impairment,” even “subjective memory complaints,” are important risks for developing dementia.⁷ Especially important have been the proposals to diagnose what previously were understood as dementing illnesses, notably Alzheimer disease, without dementia being present. Even so, the new criteria must be considered in light of potential challenges raised by prospective, community-based autopsy studies. As detailed next, pathologic changes consistent with pathologically defined entities, including Alzheimer disease (plaques and tangles), Parkinson disease (α-synuclein/Lewy bodies), and cerebrovascular disease (large and small vessel strokes, white matter changes), can be seen in patients with cognitive decline short of dementia. Moreover, such pathologic changes often overlap in individuals. Furthermore, these studies show that some individuals can have clinical dementia without clear pathologic markers.

Limitations of Current Classification

The syndrome of dementia remains largely as it always has been. The fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders has renamed dementia as a “Major Neurocognitive Disorder.”⁸ Although there are many important details for research,⁹ including an attempt to provide a common language for the description of overlapping aspects of dementia, delirium, and depression, the new criteria require validation and appear to have little practical impact on how to proceed. Awaiting more widespread adoption of this inclusive term, we will stick with dementia in this chapter. It remains the case that, as they age, most people find that their memory is not as good as it used to be. Many fewer actually have dementia.¹⁰ Where to draw the line between memory complaints and dementia—and how many lines to draw—is controversial, highlighting the artificiality of the definition as a purely discrete entity.¹¹ In addition, other aspects of behavior, such as neuropsychiatric problems (including depression) and mobility difficulties, precede dementia and are associated with cognitive decline, highlighting the restrictive nature of current classifications based mainly on cognitive criteria.¹² Nevertheless, making a diagnosis of the dementia syndrome in relation to the severity of cognitive impairment and a clinically important impact on activities of daily living remains practical. It recognizes that care needs are present or anticipated, and it assists in linking cognitive and behavioral syndromes to pathology.

Against this traditional approach is the view that waiting until the disease is clinically evident is to wait too long. This desire to get at the dementing illnesses before their clinical onset motivates the new criteria (especially the proposal for preclinical diagnosis), which hold that to do so is to wait past a time when intervention might alter the clinical course of dementia, perhaps even preventing it entirely. We will consider this, and the prospect of disease-modifying drugs, later.

Turning to other limitations of the how dementia is approached as a syndrome, the issue of the onset of clinically important change is another area of potential controversy. Delirium is considered a syndrome of acute onset with fluctuating attention and an identifiable cause. Delirium does not always reverse.¹³ Indeed, persisting delirium is now recognized as a state that not uncommonly merges with dementia, likely reflecting shared elements of pathophysiology in relation to neurodegenerative and inflammatory processes.¹⁴ Whether this is because the delirium causes the chronic, persistent, cognitive impairment or simply because the delirium occurred in the setting of dementia (or predementia), or whether many delirium symptoms that fall short of meeting delirium criteria are signaling an at-risk state for which our current descriptions are inadequate, is not clear.¹⁵ Acute onset of dementia can occur and does not always indicate delirium.

Especially in the emerging era of presymptomatic diagnosis of dementing illnesses, the issue of how best to consider multiple causes requires consideration anew. An important challenge is the realization that the specificity of cause, symptoms, and neuropathology is much fuzzier than had been appreciated, especially in very old people.¹⁶ Dementia chiefly happens not just in older adults but, as is becoming clear from an emerging body of evidence, it arises largely in people who have many health deficits.¹⁷^–²⁰ Despite many patients having distinct symptomatic profiles in relation to the type of dementia that they have, as age increases, overlap of symptoms is common.²¹ This overlap also extends to the underlying neuropathology, something that especially has been revealed by prospective, community-based studies. The fact that autopsy series in community-dwelling older adults show that many have mixed pathologies²² (up to three quarters in some series) has implications, especially for our understanding of mechanisms.²³ Interpretations vary as to how often dementia arises as a result of complex constellations of underlying neuropathology versus the simple fact of cumulative damage.²⁴ Whether neuropathology under these circumstances will retain its iconic “gold standard” status or whether it will be viewed best as an aspect of construct validation (another factor to be considered) is debatable.²⁵ Other consequences arise. From a clinical standpoint, DSM-V modifications have been argued to be most in line with this new reality and are held to provide clinicians with a common language that can thereby be used to describe overlapping symptoms and set these in the context of distinct profiles.²⁶ Against this background, it remains to be seen how much sense it makes to view patients with single-gene mutations in early-onset dementia as offering “proof of concept” for disease-modifying therapy that might avoid late life dementia in people with a panoply of causes. Very early attempts at combining biomarkers (e.g., within cerebrospinal fluid [CSF] or across modalities) showed that dementia tests are not exempt from trade-offs in sensitivity and specificity.²⁷

An additional challenge to the interpretation of the many causes that give rise to dementia in older people is overlapping pathology, which may be more the rule in older adults where Alzheimer pathology can overlap with vascular or Lewy bodies and other age-related changes. Furthermore, each of these pathologic markers can be present in patients who are apparently cognitively intact. Further challenging the present understanding is that older people can decline cognitively without obvious brain changes other than atrophy, at least as detectable by current techniques.

Dementia Definitions

General criteria for dementia (and its congeners) include the DSM-V criteria and International Classification of Diseases, 10th revision (ICD-10) criteria; each features the requirements of both cognitive impairment (in more than one domain of cognition) and functional impairment. Memory impairment is specified as a domain for ICD-10. Competing sets of criteria for vascular dementia have been supplanted by a proposal from the Society for Vascular Cognitive Impairment (VASCOG).²⁸ Current dementia with Lewy bodies criteria include the presence of distinctive clinical features, such as REM sleep behavior disorder, and laboratory findings such as decreased dopamine transporter binding.²⁹^,³⁰ Even so, although the criteria have increased in specificity, this has come at a cost of less sensitivity; interestingly, one proposed remedy has been the use of composite scores, reflecting a general theme of paying attention to both quantitative and qualitative aspects of dementia diagnosis, especially in late life. The definition of Parkinson dementia has been addressed,³¹ although its main distinction in relation to Lewy body dementia (how long motor symptoms preceded cognitive ones) is artificial. Criteria for frontotemporal lobar degeneration that predict pathologic findings have also been put forward.³² Recent criteria for the behavioral variant of frontotemporal dementia have been validated in autopsy series, with individual items showing moderate-substantial interrater reliability.³³ Similarly, subtypes of primary progressive aphasias, which have varied pathologies, have been related to autopsy findings, albeit with imperfect clinical-pathologic correlation.³⁴

Risk Factors

Age

Age is the most potent risk factor for most dementias,³⁵ including Alzheimer disease, dementia with Lewy bodies, and Parkinson disease with dementia and vascular cognitive impairment/dementia. Since age is associated with each of these, overlapping pathology is common.²²^,²³^,³⁶ Despite the obvious relationship with age, the now obsolete NINCDS-ADRD criteria placed age limits, with a lower limit of 40 years of age and upper limit of 90 years of age, for a diagnosis of Alzheimer disease.³⁷ This highlights why those criteria needed updating: dementia appears to affect most people aged 90 years and older and to arise from multiple causes.¹⁶ In familial Alzheimer disease, onset in the 30s is not unheard of and even though more unusual disorders such as childhood metabolic disorders become more common, Alzheimer remains an important consideration in younger cases. Although a number of genetic and metabolic conditions cause dementia in younger people, these will not be covered.³⁸

Frontotemporal dementias (FTDs) are a group of clinically defined syndromes associated with dementia, typically with onset younger than age 65 and with recognized variants.³²^,³³^,³⁴^,³⁹ However, many patients with FTD syndromes, including frontotemporal lobar degeneration, progressive nonfluent aphasia, semantic dementia, corticobasal ganglionic degeneration, and frontotemporal dementia with motor neuron disease, have onset in old age, where clinical features such as memory problems may be confused with Alzheimer disease, which itself can have focal presentations, with prominent executive dysfunction, apraxia, or visuospatial deficits.⁴⁰^,⁴¹ Additional distinct pathologic entities, including argyrophilic grain disease⁴² and hippocampal sclerosis,⁴³ have been recognized as entities that overlap phenotypically with more common dementias. Given their predilection for medial temporal lobe involvement, not surprisingly, memory impairment is evident, which leads to confusion with Alzheimer disease.⁴⁴ Their distinction from other pathologic disorders such as the FTD syndromes is not clear.

Family History and Genetics

Young-onset dementia is more often associated with a strong family history consistent with autosomal dominance. For example, familial Alzheimer disease and FTD typically occur in the younger age range, but late-onset individuals are not that unusual in practice. Patients with frontotemporal dementias are more likely to have a family history.⁴⁵ Gene carriers with common mutations associated with frontotemporal dementia, including progranulin, MAPT, and C9orf72, can have a long prodromal phase, leading to confusion with psychiatric disorders or mild cognitive impairment states.⁴⁶ The range in age of onset can vary within a family with a genetic family history, and sometimes family history is not known. In Alzheimer disease, mutations in the presenilin 1 (PSEN1) gene predominate among patients who are younger than 60 years of age but can have a range in age of onset.⁴⁷ A family history, with dementia occurring in more than one first-degree relative, is not uncommon, even in old age. A history of more than one first-degree relative is more common in young-onset dementia than in those with onset in the oldest old age range (>80 years). Although the apolipoprotein E4 (ApoE4) allele is a risk factor for Alzheimer dementia, it does not account for all of the increased risk associated with family history. It now has been established that several genes are implicated in late-onset Alzheimer disease, where their roles typically are small but act in combination, including with APOE loci variants.⁴⁸^,⁴⁹

Gender is a factor that affects the differential diagnosis on an epidemiologic basis, with men more likely to have vascular dementia and Lewy body dementia and women more likely to have Alzheimer disease.⁵⁰ More than a genetic factor, gender influences behavior (i.e., exposure to environmental risk factors) and hormonal levels, which could influence dementia risk. Moreover, women may be at greater risk for autoimmune disorders.

Psychiatric Disorders

Neuropsychiatric symptoms may precede a dementia diagnosis and are more common in people with mild cognitive impairment than in cognitively intact older people.⁵¹ Neuropsychiatric symptoms can increase the risk of progression from a mild cognitive impairment state to dementia.⁵² Depression has been identified as a risk factor for dementia, and a depression syndrome of dementia has been identified.⁵³ Even so, disentangling the two can be complicated. Reactive depression can occur as a consequence of a dementia diagnosis. Given that depression is common in older adults, occurrence of dementia with depression might occur by chance. However, common causes (i.e., cerebrovascular disease, dementia with Lewy bodies, and Alzheimer disease) can lead to both depression and dementia, and several studies have shown that depression precedes dementia.⁵⁴

Neuropsychiatric symptoms are common in all of the dementias. Psychosis, specifically visual hallucinations, is a core criterion for dementia with Lewy bodies and commonly occurs in Parkinson disease, where symptoms of psychosis can precede dementia or occur in the setting of dementia.⁵⁵^,⁵⁶ Delusions, notably paranoia, are commonly seen in adults who have Alzheimer disease, and impaired insight and judgment, as well as other behavioral problems, are central symptoms of frontotemporal dementia and are commonly seen in adults who have vascular dementia.⁵⁷

Other Risk Factors

As noted, the overall level of health is an important risk for dementia: in older adults, dementia is strongly associated with more health deficits and with frailty (defined either as a syndrome or as a state).¹⁷^–²⁰ Note also that, especially considering frailty as a syndrome, some of the neuropathologic markers of dementia are also associated with phenotypic features of frailty.⁵⁸ Cardiovascular disease–related risk factors (i.e., hypertension, diabetes, smoking, hypercholesterolemia) increase the risk of late life dementia.⁵⁹ It is not clear if vascular risk factors cause damage via stroke or ischemia that lowers cognitive reserve or if they incite a pathologic cascade that leads to accelerated cognitive decline. Related risk factors are educational attainment and physical activity level, which have been associated with dementia risk but are confounded by socioeconomic status, which in turn affects general health, especially cardiovascular health.⁶⁰

Head trauma is another fairly consistently identified risk factor for late life dementia, which may similarly affect brain function.⁶¹ Chronic traumatic encephalopathy is receiving increased attention as a potentially preventable cause of dementia.⁶² Both acute cerebrovascular events and head injury in older people can be associated with global cognitive decline and can accelerate the course of an established dementia. Data regarding mild cognitive impairment are less clear.

Dementia Associated with Other Disorders

Numerous disorders can cause dementia because of effects on brain function. Approximately 7% to 10% of dementias can come on suddenly, often indicating a cerebrovascular event or other medical problem.⁶³ When onset is abrupt, it is possible that the onset marks a delirium, and investigations and interventions should proceed accordingly. Although the differential diagnosis of delirium can overlap with that of dementia, and people with dementia are at risk for delirium, the two should be considered distinct. The disorders discussed in this section can give rise to prolonged illness that can lead to confusion with degenerative disorders. In some cases they are treatable and can potentially be cured, although this is unusual.⁶⁴^,⁶⁵ Strictly speaking, such disorders are among issues that should be excluded for a patient to meet criteria for a degenerative dementia; however, sometimes they can cause irreversible change in the absence of a neurodegenerative disorder.

Alcohol/Drugs/Toxins

Alcohol is an important covert cause of dementia worldwide. Abuse of alcohol is underrecognized in general, and it should be screened for.⁶⁶ Direct effects of alcohol on central nervous system function can produce syndromes such as Wernicke encephalopathy (nystagmus, restricted extraocular movements, ataxia) and Korsakoff syndrome (with persistent executive and memory dysfunction), which are recognizable causes of cognitive impairment since they often present in association with other examination features in the setting of alcohol abuse. This clinical picture can also occur in the absence of alcohol exposure with severe malnutrition where similar imaging and clinical changes are present,⁶⁷ with involvement of the medial thalami and the periventricular region of the third ventricle, the periaqueductal area, the mammillary bodies, the tectal plate, and, in rare cases, the dorsal medulla.⁶⁸ It can be difficult to separate effects of chronic exposure of alcohol from other effects, such as effects on vascular risk and brain injuries, for which people with alcohol abuse are at increased risk. Associated lifestyle habits (e.g., smoking) and socioeconomic status can also compound risk for dementia in a population prone to alcohol abuse. Lower levels of intake of alcohol are reported to be protective, but it is difficult to separate this from other factors.⁶⁹

Although alcohol is a common cause of hepatic disease, including cirrhosis, hepatic dysfunction from any cause can lead to an encephalopathy and cognitive impairment. Although in principle reversal of liver damage will lead to normalization of cognitive function, this may not always be the case. Direct effects of toxins such as manganese accumulation and unmeasured substances may be responsible for neurologic dysfunction in the setting of hepatic dysfunction.⁷⁰

Medications for various conditions have been associated with chronic cognitive impairment. Particular drugs include corticosteroids, anticholinergic medications, benzodiazepines, psychiatric medications, and antiepileptic medications, among others. Drugs should be discontinued if they might be temporally associated with cognitive decline. Aging is associated with pharmacokinetic and pharmacodynamic changes that might lead to a chronic medication becoming toxic without clearly being temporally associated with cognitive decline. Accidental overdosing due to forgetfulness can also contribute to amplified effects of medications.⁷¹

Other toxins, especially metals such as aluminum, mercury, bismuth, and lead, have been associated with cognitive dysfunction. Toxins affecting cholinergic or mitochondrial function (e.g., pesticides) or those causing white matter damage (e.g., solvents) can be related to chronic cognitive impairment. Metals such as copper, zinc, and iron are involved in normal cellular processing and have been examined in relation to dementia.

Autoimmune and Inflammatory Disorders

Autoimmune and inflammatory causes are rare but treatable causes of dementia in older adults.⁷² Vasculitis can cause dementia through ischemic damage. All forms of vasculitis (e.g., small vessel or large vessel) can lead to progressive neurologic dysfunction. There are often clues to the diagnosis, such as systemic complaints, headache, and sometimes evidence for an associated autoimmune disease. Large vessel vasculitides such as giant cell (temporal) arteritis, Churg-Strauss syndrome, polyarteritis nodosa, Wegener granulomatosis, Behçet syndrome, and Sjögren syndrome often show evidence for systemic disease and inflammatory markers (increased C-reactive protein and sedimentation rate). Serologic investigations can identify vasculitis associated with other autoimmune disorders; however, primary central nervous system vasculitis can cause progressive cognitive decline unassociated with other markers.⁷³ Angiography and brain biopsy may be necessary to diagnose inflammatory disorders causing dementia, although evidence for an inflammatory process can often be identified by examining spinal fluid (elevated cell count, increased immunoglobulin synthesis, elevated protein level). A sedimentation rate or C-reactive protein can be associated with a chronic inflammatory process of any cause, but also specifically with giant cell arteritis. Immunoglobulins are elevated in inflammatory disorders, but monoclonal elevation can be associated with multiple myeloma.

Limbic encephalitis on an autoimmune basis can present as an acute or a subacute dementia. It can be paraneoplastic or non-paraneoplastic. Imaging findings can mimic prion disease, but distinct changes, including high signal change on magnetic resonance imaging (MRI), are important to recognize. As with other autoimmune dementias, nonspecific CSF changes can be evident. Specific serum or CSF abnormalities can confirm the diagnosis. Practically, such patients should be examined and investigated for the presence of a neoplasm, recognizing that a purely autoimmune process may cause a clinically indistinguishable syndrome and that these disorders are treatable. Thus care must be taken to not simply rely on the presence of antibodies but to consider the level of the antibody titers and the possibility of false positive results.⁷⁴ Inflammatory disorders can also lead to a hypercoagulable state, which can be seen in association with anticardiolipin or antiphospholipid antibodies, sometimes with evidence for a lupus anticoagulant. Sneddon syndrome is characterized by livedo reticularis associated with hypercoagulability, which typically occurs in younger people but can also occur in older people.⁷⁵ Antinuclear antibodies, extractable nuclear antibodies, antiphospholipid or anticardiolipin antibodies or a lupus anticoagulant, as well as physical examination findings, may point to an autoimmune or collagen vascular disorder. Nonvasculitic autoimmune meningoencephalitis is a relatively recently recognized cause of dementia, sometimes, but not necessarily, associated with collagen vascular disorders such as Sjögren syndrome, antiphospholipid antibody syndrome, systemic lupus erythematosus, mixed connective tissue disease, or Hashimoto encephalitis.⁷⁶^,⁷⁷ In some patients brain biopsy may be warranted, which may be justifiable given that it is treatable. Drugs, notably nonsteroidal antiinflammatory medications, can be associated with a meningoencephalitis, which can mimic a subacutely progressive dementia.⁷⁸ Sarcoidosis is a great mimicker among inflammatory disorders. It can have ocular, central, and peripheral nervous system involvement. Since it can present at any age, it should be considered in the differential diagnosis of dementia in older people.⁷⁹