OUP user menu

Delirium, dementia and senility

Raj N. Kalaria, Elizabeta B. Mukaetova-Ladinska
DOI: http://dx.doi.org/10.1093/brain/aws235 2582-2584 First published online: 8 September 2012

Delirium is a neuropsychiatric syndrome characterized by sudden onset and global effects on degrees of consciousness, attention and cognition. It is now widely recognized that delirium is associated with high morbidity and mortality. Increasing age appears the strongest risk factor for delirium (Table 1). It affects up to 40% of all medically ill older adults. Delirium subsequent to surgical procedures, particularly those involving major fractures, cardiovascular grafting or organ transplantation is unpredictable, ranging 9–87% with an average of 40%, depending on the age of the patient, type of intervention and pre-existing vascular disease risks or infections (Cerejeira and Mukaetova-Ladinska, 2011). The presence of delirium complicates outcomes in the majority of ill patients, as demonstrated by longer hospital admission, and impact on daily performance, frequently leading to costly institutional care. In all these situations, it is necessary to ensure that delirium is consistently and correctly diagnosed. Being well qualified to make the diagnosis—as psychiatrist, neurologist or geriatrician—is an important consideration when reporting on the nature or consequences of delirium in any particular study.

View this table:
Table 1

Predisposing and precipitating biological factors associated with delirium

Increasing age
Surgical procedures
    Cardiovascular or aneurysm surgery
    Hip or femoral fracture and surgery
    Organ transplant
Chronic infections
Metabolic disorders
Vascular disease: diabetes mellitus, atrial fibrillation
Ischaemic strokes (single and recurrent)
    Post-stroke dementia
Cognitive impairment
Neurodegenerative disease:
    Alzheimer’s disease
    Frontotemporal lobar degeneration
    Parkinson’s disease
    Prion disease
    Depressive illness
  • Observations derived from several retrospective and prospective studies cited in reviews (MacLullich et al., 2009; Witlox et al., 2010; de Lange et al., 2012). Samples investigated involved community- and hospital-based subjects, including intensive care unit subjects. Varied sample sizes from 34 to 1218 subjects in the 55–85-year age range. Delirium was diagnosed using the Diagnostic and Statistical Manual of Mental Disorders III-R or Diagnostic and Statistical Manual of Mental Disorders IV, Confusion Assessment Method, Delirium Rating Scale or the Delirium Index.

Cognitive changes after episodes of delirium

Over and above the established immediate consequences, there also appears to be a link between delirium and long-term cognitive impairment (MacLullich et al., 2009; Witlox et al., 2010; de Lange et al., 2012). Despite varied sample sizes, study designs, screening methods and origin of patients being studied, there is remarkable consistency of evidence that even a single episode of delirium is detrimental to cognitive function and may more than double the risk of dementia (Witlox et al., 2010). An increased risk of poor cognitive outcome and dementia is particularly noteworthy in older adults who have had a prolonged episode of delirium lasting weeks to months (MacLullich et al., 2009), irrespective of gender, co-morbidity, illness severity or presence of dementia at baseline. Conversely, cognitive impairment is itself one of the recognized risk factors for delirium. The prevalence of delirium is estimated at up to 70% in institutionalized people with dementia (de Lange et al., 2012). Others have observed the incidence of delirium at 25% in individuals diagnosed with Alzheimer’s disease (Fong et al., 2012). Patients with co-morbid conditions characterized by progressive supranuclear palsy and Alzheimer’s disease may exhibit prominent bouts of delirium along with emotional and personality changes (Sakamoto et al., 2009). Delirium is also a frequent occurrence after stroke—episodes being diagnosed in 20% of ischaemic stroke patients aged 55–85 years and associated with dementia at 3 months in approximately half of those affected, with age and severity of stroke injury leading to increased long-term mortality in patients who have experienced delirium (Melkas et al., 2011).

Postoperative cognitive impairment associated with delirium may be transient, but persistent cognitive decline (beyond 3 months after surgical treatments) has also been demonstrated in a number of studies (Hovens et al., 2012). It remains unclear how the surgical manipulation(s) play a role in the long-term cognitive deficits in delirium sufferers. One possibility is that peripheral vascular changes associated with cardiac surgery, organ transplant or orthopaedic procedure impact on cerebral brain perfusion.

The elucidation of putative mechanisms that cause delirium and deciphering how these result in dementia are important if the burden is to be reduced in the long-term. Why should a single episode of delirium in older adults cause or trigger a chronic decline in cognitive function? Is delirium simply one marker of a pre-existing chronic progressive pathological process or the consequence of an acute brain insult that involves reactive responses leading to subsequent specific changes in neural circuits or structure? Is delirium a symptom of reducing brain reserve capacity? The pathological substrates of cognitive impairment or dementia are not necessarily difficult to identify, but characterizing those linked to an episode of delirium is more challenging.

In this issue of Brain, Davis et al. (2012) explore whether delirium is a risk factor for cognitive decline and incident dementia in the oldest old. The Vantaa 85+ community-based study is one of the largest projects of its kind. Given that the oldest old may represent an unusual group of individuals who have accumulated, but survived, various pathologies, it is important to know that, even in a relatively small numbers of cases in whom prior events were largely assessed retrospectively, delirious episodes are evidently a strong risk for cognitive decline. Delirium increases the risk of incident dementia by nearly 9-fold and is associated with a 3-fold worsening of dementia severity and deterioration in global intellectual function. These findings appear robust, despite measures of cognitive decline being limited to the Mini-Mental State Examination scores (Davis et al., 2012). In an attempt to find tangible mechanistic links, the Vantaa 85+ study team also asked whether the burden of pathology is altered in those who exhibit a history of delirium. They tested whether neurofibrillary pathology (Braak stage) in those who came to autopsy is associated with delirium episodes. Perhaps, not surprisingly, there is no relationship between delirium and this Alzheimer’s disease marker or with neuritic amyloid, α-synucleinopathy, presence of infarcts and apolipoprotein allele status. Although one might have expected a marker or two to be identified, a more rigorous pathological screening or even wider survey of cellular markers, e.g. glia, synapses, neurotransmitters or quantification of cerebral atrophy, will be necessary to overcome confounding mixed pathologies and tease out the key substrates in the oldest old.

Pathological substrates of delirium?

Peripheral neurotransmitter chemistry has been described to be altered before, during and after episodes of delirium (Cerejeira et al., 2012). However, there is now accumulating evidence that inflammatory mechanisms are involved in delirium (Cerejeira et al., 2010). In a small neuropathological study, there was an association between delirium and widespread microglial activity, with a 4- to 6-fold increase (as assessed by CD68 and HLA-DR, respectively), astrocytic response and levels of interleukin-6 in older adults (Munster et al., 2011). Plasma markers of an inflammatory response in critically ill delirious patients, with and without pre-existing infection or systemic inflammation, appear to arise from different mechanisms. The proinflammatory cytokine interleukin-8 is associated with delirium in inflamed subjects, whereas increase in the anti-inflammatory cytokine interleukin-10 and amyloid β peptide 1-42/40 are associated with delirium in patients without inflammatory disease (van den Boogaard et al., 2011). The elevated concentrations of amyloid β also correlate with long-term subjective cognitive impairment, suggesting that this may represent the first sign of a subclinical pathological process leading to dementia. Although these preliminary reports need to be confirmed in larger prospectively designed and controlled studies, they provide a basis for further investigation of brain inflammatory mechanisms in delirium. Substrates germane to processes that cause delirium may be identified by producing blood signatures from proteomic or microarray studies before and after episodes in targeted patients. These could then be ideally compared with paired CSF samples. Alternatively or additionally, e.g. PET imaging with the use of translocator protein 7 radioligands could be one way to track progression and severity of neuroinflammation or microglial activation in the living brain of patients with delirium.

Recent, but preliminary, studies make it likely that delirium will be associated with subtle changes in the white matter or brain structure. In a diffusion tensor imaging–magnetic resonance study (Morandi et al., 2012), greater duration of delirium was associated with lower fractional anisotropy in the genu, splenium of the corpus callosum and anterior limb of the internal capsule and was still present at 3 months after hospital discharge. In a parallel study, these authors also reported that longer duration of delirium is associated with smaller superior frontal lobe and hippocampal volumes at discharge (Gunther et al., 2012). Yet to be confirmed in larger studies, these observations may reflect the multiple causes and medical interventions that contribute to the long-term cognitive outcome after delirium.

In contrast to the available studies linking a history of delirium to cognitive decline, there is an overall paucity of research prospectively studying long-term outcomes in people who experience an episode of delirium. The identification of risk factors and neurochemical biomarkers linked to brain pathological changes, aimed at understanding the mechanisms of cognitive decline after episodes of delirium, is much needed.


Our work has been supported by the RCUK Newcastle Centre for Brain Ageing and Vitality, Medical Research Council (UK), Alzheimer’s Research UK, and the Newcastle NIHR Biomedical Research Centre in Ageing and Age Related Diseases, Newcastle upon Tyne Hospitals NHS Foundation Trust.