A 58-year-old female with a history of coronary artery disease was found in a disoriented and mildly agitated state one morning, wandering outside her apartment after she failed to show up at work. All laboratory studies were negative. Head computed tomography (CT) revealed a relatively small right anterior pontine infarct just inferior to the fourth ventricle. Her confusional state cleared within a week, but for some time she continued to show significant disturbances in attentional function, in spatial relations, especially spatial synthesis, and in other forms of nonverbal or novel cognitive processing, along with quite poor and easily disrupted working memory. The atten-tional and working memory disturbance was significantly worse in the morning for uncertain reasons. She became depressed, though this was successfully treated with an antidepressant that possessed both serotonergic and noradrenergic properties. This type of antidepressant was chosen over a selective serotonin reuptake inhibitor (SSRI) with the hope that it would better improve right-hemisphere arousal presumably disrupted in the context of the right pontine reticular formation cerebrovascular accident (CVA).
Some 3 months later, the patient was brought to the hospital with congestive heart failure, with associated acute renal insufficiency blood urea nitrogen (BUN > 90). At this time she also showed a moderate confusional state, with marked disorientation to the environment, obvious disturbance in her ability to sustain coherent attentional processes or task frameworks, and mild agitation. This second delirium was clinically and phenomenologically virtually indistinguishable from the first, with the exception of some greater degree of fatigue. It is possible that the previous structural insult had left a residual disruption of right-hemisphere arousal, lowering the threshold for confusional states. In the context of toxic metabolic or neuromodulatory disturbances, this woman was likely more vulnerable to delirium. Unlike many patients with confu-sional states, this woman did not evidence any sign of baseline dementia, nor did she evidence significant prodromal stage cognitive declines from very early Alzheimer's disease (AD). Given her vascular history, however, she was deemed at risk for such a neurodegenerative process and followed at yearly intervals.
Delirium may be the most commonplace disturbance of consciousness encountered by psychiatric clinicians, as well as by other physicians, and is virtually ubiquitous on medical services in general hospitals. Its quite commonplace nature contrasts with a curious neglect within both clinical neuroscience and consciousness studies of the disorder, as relatively little attention has been paid to understanding the underlying neural and neurodynamic foundations for delirium and confusional states. Delirium is most classically associated with toxic-metabolic disturbances of a wide variety, or neuro-modulatory disruptions secondary to psychotropic medicines, often superimposed upon and degrading a baseline dementia of the Alzheimer's type. In terms of neuromodula-tory disruptions, it is most typically associated with the effect of anticholinergics, but it is also commonly seen in dopamine precursor loading, from the effects of opiates, and from gamma-aminobutyric acid (GABA) agonist effects of various medicines, including benzodiazepines and anticonvulsants. There is substantial clinical/anecdotal evidence that thresholds for anticholinergic deliriums (and deliriums from virtually all etiologies) are significantly lowered by preexisting AD, even very early stage AD, where there are only fairly modest cognitive deficits. This may possibly be due to early involvement in AD of the cholinergic basal forebrain, although this correlation between lowered thresholds for confusional states and cholinergic deprivation of the forebrain is also not empirically established in AD. As a wide variety of dementing disorders progress into their middle and late stages, the distinction between baseline dementia and confusional states gradually disappears, as patients progressively lose working memory integrity and task/behavioral organization. Diffuse Lewy body disease, more recently appreciated as a disorder distinct from classical Parkinson's disease, often produces a chronic confusional state after only 2 to 3 years of cognitive and behavioral declines, possibly due to its extensive disruption of numerous brainstem neuromodulatory systems, including ACh, DA, and NE (see chapter on neurodegenerative disorders).
Although older concepts of delirium emphasized sensory alterations, perceptual illusions and hallucinations (not unlike some descriptions of schizophrenia, contributing to the tendency to misconstrue delirium as a psychotic disorder), more recent concepts emphasize the core attentional derailment, and an associated collapse of the integrity of working memory. In delirium and confusional states, the normal segue of working memories collapses. Normally, working memories must in some sense "father their successors" and thus show a coherent trajectory. Instead, patients with delirium severely derail (losing task set) even when engaged in simple tasks, and working memory is highly vulnerable to interference. Patients with confusional states often become quite tangential, and language becomes increasingly fragmented (as confusional state worsens) due to the progressive failure of semantic working memory. Patients often fail to register simple information from the immediate environment, in direct proportion to the severity of the confusional state. They frequently cannot shift or maintain a focus of attention in an adaptive manner. Gross behavioral disorganization ensues, particularly as the degree of confusional state worsens, often accompanied by agitation (in nonobtunded confusional states). Confusional states sometimes show transient and disorganized paranoid ideation that can lead to misdiagnosis of the delirium as a psychosis. In misdiagnosed cases, subsequent treatment with neuroleptics (which often reduce the agitation), seem to confirm the psychosis, sometimes with unfortunate concomitant failures to identify reversible toxic-metabolic or neuromodulatory etiologies.
A major theoretical challenge remains for the clinical neurosciences to develop more heuristic concepts of delirium (like those of Mesulam, 2000), integrating the dozens of classical metabolic etiologies with less frequent structural lesion correlates. Case study 1 shows that structural insults to crucial reticular systems in the right hemisphere may generate a disorganized state virtually indistinguishable from classical toxic-metabolic encephalopathies. In this case, the structural insult was a right pontine CVA, leading to a disruption (possibly differentially although this is not known) of pontine noradrener-gic, serotonergic, and cholinergic nuclei in the reticular portions of the anterior pons, and subsequently disrupting thalamocortical arousal and integration. However, the more classical structural correlates for CVAs in the right hemisphere would be the heteromodal portions of the parietal lobe, particularly the inferior parietal lobe, as well as dorsolateral prefrontal regions, the thalamus and basal ganglia, and the cingulate.
Thus, one must emphasize that there exist many pathways to the clinical presentation of delirium, and our case represents only one lesion and one metabolic correlate. Delirium also shows us that the coherent organization of working memory can be derailed by direct lesioning of the classical working memory systems in dorsolateral prefrontal cortex, closely related heteromodal systems in parietal lobe essential to attentional function, or by insults to thalamic, basal ganglia, and reticular support systems for these heteromodal cortical areas. Working memory in turn can be conceptualized as an important index of attentional function, the residue of what attentional mechanisms "capture" within a global workspace. The common toxic metabolic processes generating delirium do so presumably by disrupting the gamma band activity required to functionally instantiate these complex distributed networks, perhaps particularly their corticocortical aspects (see section on neurodynamics). Such gamma activity may be more physiologically demanding and thus more vulnerable to toxic-metabolic problems. Delirium also illustrates the close, intrinsic relationship of attentional processes to higher executive processes and organized purposeful behavior, as these two fundamental functional envelopes of consciousness are both affected in direct proportion to the severity of the confusional state. Indeed, the selection, maintenance and updating of working memories is a central executive task for the attentional systems of the brain.
Lastly, but not trivially, delirium presents a little appreciated comment on a very controversial point in consciousness studies, concerning the neuropsychological substrates for feeling states, the manifestation of emotion in consciousness. There are three basic arguments coming from behavioral neuroscience on how this might happen: (1) LeDoux (1996) argues that working memory comes to represent the otherwise unconscious changes associated with emotional activity, and this happens largely in dorsolateral prefrontal cortex; (2) Rolls (1999) suggests that language representation underpins feelings; (3) Panksepp and Damasio suggest that feelings emerge from largely subcortical dynamics, with Panksepp (1998) emphasizing the interactions between PAG, superior colliculus, and pontine motor systems, while Damasio (1999) emphasizes changes in the proto-self-systems, particularly somatosensory cortices remapping various changes in bodily state, generated by the interaction with an emotionally charged object or person. In the nonstuporous deliriums, emotion appears consistently disinhibited, despite the devastation of working memory, and in the severe deliriums, devastation of even coherent language. This argues for the subcortical view of emotion, suggesting that the collapse of working memory and higher cognitive functions does not prevent emotion from entering a very disorganized conscious state. Dysphoric emotion and highly agitated states appear particularly released and commonplace, although deliriums, from some ILN lesions, and other atypical deliriums, will sometimes show manic features. One might readily argue that most individuals would be expected to become dysphoric and agitated when faced with the collapse of basic cognitive functions including even perceptual integrity and stability, and that such a state is intrinsically frightening. Thus, any careful review of the clinical phenomenology of delirium does not support the notion that working memory or language are necessary for prototypical feeling states, and instead suggests that those states sit underneath cognition, underneath working memory, and other higher cortical functions.
Delirium remains very much a promising area for the neuroscientific study of consciousness. Yet, it is a curiously neglected "orphaned child" within clinical neuroscience, having achieved this status despite its virtually ubiquitous presence in general hospitals and nursing homes. Although the treatment of delirium has always emphasized the mitigation of the offending etiologies, research into interventions that might mitigate attentional and executive collapse in chronic confusional states or improve recovery has almost no initiative within the fields of psychiatry and neurology. Given that confusional states (particularly in their more serious forms) devastate capacities for independent functioning, the lack of any comprehensive theory or interest in ongoing empirical research is most puzzling.
Was this article helpful?