The role of predictive coding in the pathogenesis of delirium

Medical Hypotheses 103 (2017) 71–77

Contents lists available at ScienceDirect

Medical Hypotheses journal homepage: www.elsevier.com/locate/mehy

The role of predictive coding in the pathogenesis of delirium J.M. FitzGerald Department of Paediatric Surgery, Leeds General Infirmary, Leeds Teaching Hospital Trust NHS, UK

a r t i c l e

i n f o

Article history: Received 24 August 2016 Accepted 21 April 2017

a b s t r a c t Delirium and dementia represent an emerging global crisis in healthcare. Attempts have been made to identify the pathognomonic feature that would make delirium stand out from dementia but unfortunately the global neural dysfunction of both disorders has made the establishment of a direct measurement difficult. Modern conceptualisations of delirium have been influenced by the assessment tools used to assess, detect, and analyse its complex and transient nature. Recent publication of the DSM-V criteria for delirium has marginally altered the previous DSM-IV criteria with a focus upon inattention with vague terms such as consciousness downplayed. Such an alteration has been found to be restrictive and thus impact upon delirium case identification. Although these findings are approximating the empirical state of delirium as measured by validated instruments, a more refined neuroscientifically informed phenomenological framework is required in order to enhance the theoretical understanding of delirium assessment and resolve these challenges. One such application is the predictive coding (PC) model, also known as the hierarchical Bayesian inference model, to interpreting delirium pathophysiology. Therefore, the aims of this paper are to 1) propose the hypothesis that delirium pathophysiology can be explained in terms of the PC model, 2) support this hypothesis by applying this model to current methods of assessing delirium phenomenology, particularly attention, and 3) outline a future programme of research to test many of the parameters of this application. Ó 2017 Elsevier Ltd. All rights reserved.

Introduction Delirium and dementia represent an emerging global crisis in healthcare [1,2]. Historically delirium has been described as a disorder of consciousness with patients experiencing a ’clouding of consciousness’, and has been contrasted to the typical retention of lucidity in dementia [3–5]. However, contemporary attempts have been made to identify the pathognomonic feature that would make delirium stand out from dementia but unfortunately the global neural dysfunction of both disorders has made the establishment of a direct measurement difficult [6,7]. The robust rating scale for delirium phenomenology, the Revised Delirium Rating Scale-Revised-98 (DRS-R98), combined with longitudinal and factor analytical methods have been used to confirm the hypothetical existence of a three domain theory for the phenotype of delirium. This three domain theory has been conceptualised as being composed of 1) general cognition, 2) higher level cognition and 3) circadian integrity Fig. 1 [9–11]. However, disparities of up to a third of delirium status attribution, namely full syndromal delirium (FSD) or subsyndromal delirium (SSD), by validated instruments such as the Confusion assessment method (CAM) and the DRSR98 have indicated that the neurobehavioural interface of what

E-mail address: [email protected] http://dx.doi.org/10.1016/j.mehy.2017.04.015 0306-9877/Ó 2017 Elsevier Ltd. All rights reserved.

can be diagnosed as delirium requires more focused research [12]. Moreover, recent publication of the DSM-V criteria for delirium has marginally altered the previous DSM-IV criteria with a focus upon inattention with vague terms such as consciousness downplayed [13]. Such an alteration has been found to be restrictive and thus impact upon delirium case identification. Indeed, recent work using pooled data sets and retrospective study designs have indicated that there is a varied concordance (30–89%) between DSM IV and DSM-V attributed cases. Such a discrepancy was due to challenges in interpreting key phenomenological features such as orientation, acute onset and fluctuating course [12]. Although these findings are approximating the empirical state of delirium as measured by validated instruments, a more refined neuroscientifically informed phenomenological framework is required in order to enhance the theoretical understanding of delirium assessment and resolve these challenges. One such application is the predictive coding (PC) model, also known as the hierarchical Bayesian inference model, to interpreting delirium pathophysiology. The PC model has been applied to neurocognitive disorders including schizophrenia, as well as global brain states such as waking consciousness and dreaming in an effort to expand their respective theoretical frameworks [13,14]. According to the PC model, perception and attention are bound together with neural networks that integrate top down processing connections (execu-

72

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77

Higher order cognion

General cognion

Circadian integrity

Fig. 1. Three domain theory of delirium phenomenology.

tive cognition) with bottom up sensory input. The hierarchical neural network in which perception results from is the interface between sensory driven analysis and top-down expectation to produce prediction errors Fig. 2. The errors then serve as part of the process to update and enhance the conditional expectations, in an effort to reduce the gap in predictive errors [15,16]. In a state of health, the brain/psyche optimises this complex interface in order to minimise prediction error and ensure that the subject remains orientated and operating at an appropriate cognitive functioning level. In contrast to this state, features of neurocognitive failure e.g. inattention, disorientation, delusions, and hallucinations dominate and are the result of the breakdown in the integrity of this critical system [13,17]. Such domains are integral features of the phenomenology of delirium [7]. Therefore, the aims of this paper are to 1) propose the hypothesis that delirium pathophysiology can be explained in terms of the PC model, 2) support this hypothesis by applying this model to current methods of assessing delirium phenomenology, particularly attention, and 3) outline a future programme of research to test many of the parameters of this application (see Table 1). Overview of the predictive coding model of the brain/psyche The PC model of CNS activity has been proposed to be congruent to more far ranging theories of physics, namely the free energy principle, which when applied to biological systems enables them to resist the otherwise inevitable trajectory towards disorder [18]. In the context of the brain/psyche, the improbability of sensations in conflict with the model of the environment generated by the brain is referred to as surprise or also known as self-information.

Top-down expectaons

Percepon Sensory driven analysis

Fig. 2. Perception as interface between top-down expectations and sensory driven analysis.

Surprise averaged over time is believed to be congruent with entropy, and by minimising this, biological systems such as the brain can resist the restraints of the second law of thermodynamics. The brain can measure free energy in terms of sensory driven analysis and interoceptive mechanisms. Therefore, by resisting the impact of free energy, the organism can operate within the world in an organised and typical manner. Updating prior beliefs into posterior beliefs through minimising prediction error (free energy), the brain can minimise this by changing predictions to match sensory driven analysis or vice versa [see Fig. 3] [16]. This operation is a dimension of the common feature of all biological systems, namely homeostasis [18]. When applied to schizophrenia and psychotic features, the PC model highlights how false inferences are at the core of inappropriate perceptions and beliefs [13]. Such disorders in cognitive processing are congruent with the neuromodualtion of excitation at post synaptic sites particularly situated in cortical lamina [19,20]. Cortical pyramidal neurons have been proposed to be critical to the encoding of the predictive error modulatory process, with superficial neurons signalling predictive errors via extrinsic feedforward ascending connections and deep neurons signalling predictions via intrinsic feedback descending connections towards the superficial pyramidal neuronal population [15,16]. There is also the relative influence of predictive errors to consider, both in terms of their respective gain and weight. This aspect is attributed to the intrinsic connectivity that orchestrates the gain of neuronal groups involved in signalling prediction error. This output is then proposed to correspond to mathematical models of precision and hence confidence regarding to the information processes within the PC system (see Fig. 1). Precision in this sense is attributed to the optimal signalling of prediction errors by the post synaptic gain of neurons involved in this process. This process is a key component of attention and sensory processing, wherein the selectively enabled convection of precise information is performed [16,21]. More generally it has been proposed that attention is the inference about precision (the uncertainty) of the causes of sensory input, while perception is the inference about the causes themselves [16].

Delirium and phenomenology Delirium is a complex phenomenological entity that is as unique as the psyche that experiences it. Investigating the phenomenology of delirium is based upon a wide variety of perspectives from quantitative (e.g. DRS-R98) to qualitative (e.g. semistructured interviews), and from objective (e.g. bioelectronics detection and neuroimaging) to observer rated scales and clinical interviews with the patient. Not surprisingly there is immense overlap between these categories of investigation. For example, in semi-structured interviews of the delirious experience, patients consistently report disturbed features of reality including, blurring between the boundary of reality and dreaming, reversal of daytime and night time, clouding of consciousness, and a feeling of disintegration [22]. In The phenomenology of perception, Merleu-Ponty explains that experience of the external world is not merely a result of perceived images (called ‘‘perceptum,” ‘‘what is seen”) and thus based solely upon sensory detection. The subject that perceives these images is bound with the experience (called ‘‘percipiens,” ‘‘he who is seeing”). The perceptum and percipiens are both aspects of the same seamless phenomena, due in part because the construct of subject is based upon sensation, perception and the cohesion of the experienced body i.e. the ego [23]. The human ego is an agency that gives coherency and masks the underlying fragmentary nature of the body [24]. This phenomenon can be identified with the individuals’ experience as resulting from the multitude of underlying neuropsychological processes that exist

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77

73

Table 1 Programme for future studies. 1. Apply advanced statistical methods (e.g. MRM) to patient populations at risk of developing delirium in order to understand the trajectory of individual features of delirium phenomenology. 2. Integrated investigations of qualitative (semi-structure interviews and bedside clinical interviews) and quantitative methods (e.g. CAM, DRS-R98 and bioelectronics methods) of analysing delirium phenomenology. 3. Development and application of systematised algorithms for generating delirium status attribution e.g. sub syndromal delirium and prodromal delirium. 4. Apply detailed neuroscientific evidence to the domain theory of delirium phenomenology. 5. Explore understudied domains of delirium phenomenology, such as circadian integrity and executive cognition through novel methods.

Changing condional expectaons

Changing sensory driven analysis Fig. 3. Updating prior beliefs into posterior beliefs through minimising prediction error (free energy), the brain can minimise this by changing predictions to match sensory driven analysis or vice versa.

beyond the threshold of consciousness. The perception of the external world is not merely a reproduction of objective findings but is shaped by the subject as it is sensed. This is further reinforced if one considers that cognitive processes are not isolated from emotions and affect but are bound with them as part of optimal behavioural functioning [23]. Understanding the pathogenesis of delirium must also be based upon a comprehensive account of the neural substrate that serves such a phenomenology. In PC, the active processing of conditional expectations with sensory encoded stimuli from the environment illustrates that experience of the world is not merely passive but contingent upon the integration between the cortical feedback networks which convey predictions and feedforward networks convey residual error. Congruent to these processing systems is the hypothesis that feedback connections have the capacity to suppress information which is predicted by the executive cognition [15,16]. In the context of delirium, it has been proposed that disturbances in baseline connectivity between robust cognitive processing systems is a key component to the global neural dysfunction of delirium [25]. This is congruent with the constellations of neural structures that are linked together and collectively implicated in being the neural substrate of delirium. The default mode network (DMN), the neural substrate that serves the resting state of the human psyche i.e. its virtual sensorium, has been found to be significantly altered in the delirious state [26]. The ego gives coherency as part of the emergent properties of executive cognitive functions that it has evolved to serve. The concept of the core self which is composed of features such as, affective regulation, motivation systems, and behavioural patterns have all been implicated as a coherent integrative function of the cortical midline structures and the subcortical midline structures of the mammalian brain [27]. The connection between the cortical and subcortical midline structures and the mirror neuron systems have been proposed to be integrated by a multitude of neural pathways based upon critical relay centres [28]. Examples of critical relay centres are the insular cortex and the anterior cingulate cortex (ACC) which functions as a cortical structure for interoceptive encoded processes and relays signalling from the mirror neuron system to the cortical and subcortical midline structures [29]. Taken together one can

propose that the integrity of these systems based upon the degree of optimised connectedness upon which they operate may reveal the neural substrate that gives rise to the phenomenology of delirium [25]. Moreover, detailed accounts of how these general principle apply to other dimensions of delirium phenomenology, such as consciousness and inattention will further reinforce these propositions.

Delirium and consciousness Delirium in its most pronounced state is a disorder of consciousness [30]. Indeed, both the ICD-10 and DSM V retain the role of consciousness in delirium, but both are less explicit in their use of the term consciousness or its role in delirium phenomenology [12,31]. Consciousness like delirium can be a graded phenomenon and disturbances can be measured both qualitatively and quantitatively [32,33]. A scientific account of consciousness must be based upon an evolutionary framework, which rejects both dualism and the modern form of dualistic theory, epiphenomenalism. Both primary and secondary orders of consciousness are fundamental to understanding the phenomenology of delirium states. Primary consciousness is composed of sensation transformed into perceptual images it is an abstract, virtual and multimodal based system. It has not only the function of being a mode for sensation but can integrate this state with memory to add a temporal dimension to awareness, one that encompasses the present and immediate past [34]. This form of consciousness is itself composed of two components as a consequence of selective attention, with focal awareness is the perceptual experience of attention by the ego and peripheral awareness which is the phenomenal penumbra of this focus [35]. The neural substrate for primary consciousness is based on large systems. The integrated axis of the brainstem and limbic system sets out physiological drives for the organism and encodes a largely implicit foundation for experiencing the external world. The cerebral cortex thalamic axis further encodes more diffuse memory systems and binds these with an awareness of the present. Edelman posits that it is the combination of perceptual awareness with conceptual memory that gives us the primary order of consciousness [34,36]. As part of the development towards an adult psyche, secondary consciousness emerges as a vast mode of psychological being, whereby the functions of metacognition, language and culture are born. Systems of communication exist as part of all animal behaviour as evidenced by the very simple range of signs, cues and stereotypical behaviours [37]. However, the most characteristic form of human communication is indeed language which is finitely complex and can reflect our capacity for self-reflection and orientation [38,39]. Orientation is not a solitary function and thus an attempt at understanding the underlying neural and psychological aspects of it require an understanding of what is the composition of the phenomenology of orientation. It is composed of three broad based systems, the arousal system (often referred to as the level of consciousness), attention and memory [40]. Orientation can be considered as a compound psychological function

74

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77

based upon attention and the autobiographical self. The autobiographical self is the aspect of the ego that is responsible for processing and encoding stimuli pertaining to one’s identity and one’s environment [41]. This aspect of identity has to be intimately bound with both explicit and implicit memory systems in order to retrieve and reproduce information about one’s body, one’s name, memories about one’s life etc. [42–44]. Studies investigating the neural correlates of these functions have been based on the theory of the ’self-referent effect’, in other words the differentiation between psychological processing between others and one’s self [45]. The neural structures implicated in the optimum functioning of this aspect of the ego are cortical midline structures. These structures include the medial prefrontal cortex (MPFC), anterior cingulate cortex (ACC), and posteromedial cortices (PMCs) [46]. Other research has highlighted the role PC model has not only for consciousness as a phenomenon, but how dreaming as a part of consciousness operates [14]. This is particularly useful for the present discussion given the parallels between the phenomenology of delirium and dreaming that have been commented by modern neuroscientific accounts [47]. Recent work into bioelectronic measures of delirium phenomenology have characterised delirium as a state of pathological wakefulness i.e. of being in oneiric (dream) state while being awake [48]. The abstract nature of dreaming highlights the pivotal enhanced mode of primary consciousness and diminished mode of secondary consciousness, with its virtual experience, enhanced cognitive associations and limited mode of self-agency [49]. The qualitative change has been attributed to shifts in neuromodulation from the amine-ergic systems diminishing influence and dopaminergic and cholinergic dominating influence. In REM sleep in particular, PC still operates, but minimises free energy largely in the absence of sensory inputs. Application of PC to sleeping/dreaming further integrates the role sleep has in cognitive processing, the growth of consciousness, and memory formation [14]. The characteristic disturbances in consciousness that patients with delirium experience parallel the distinction between primary and secondary consciousness in dreaming both with a number of distinctions that highlight the fact that delirium is a pathological state of neurocognitive impairment. The hyper dopaminergic state is consistently attributed to the pleasure seeking primary processing and is believed to be implicated in the wish aspect of dreaming. In delirium secondary consciousness functions such as language, metacognition and orientation are all significantly disturbed in longitudinal phenomenological profiles of delirious patients and all three functions constellate the alienation patients report after they recover from a delirious episode [50]. The neuromodulatory (hypo-cholinergic and hyperdopaminergic) disturbances of delirium may serve as the neurochemical basis of such phenomenological disturbances [40].

Virtual sensorium in delirium The generation of a virtual representation of the interface between the external world as transduced by the senses and reprocesses by the CNS is founded upon the primary and secondary modes of consciousness and the perceptual and conceptual systems that reside within them. The primary mode of consciousness functions as a mediator for the perceptual abstraction of the real world. While the secondary mode of consciousness binds this with a much wider experience encoded within metacognitive domains of executive cognition. Indeed visuospatial manipulation requires at least some pivotal role and influence by the processes of executive cognition in particular attention [51,52]. Attention is consistently found to be a key feature of delirium phenomenology and has a strong association with other features in delirium according

to the data [53]. However deficits in attention also exist in chronic conditions such as dementia with Lewy bodies and late stage Alzheimer’s disease [54,55]. Attention is dominated by two aspects, the first being external stimuli and the second being internal goals/bias [56]. Selective attention is a cognitive faculty that enables the filtering out of irrelevant sensory information to further enhance our capacity to detect relevant information. There is a link between prediction error and exogenous (top-down) attention, whereby, prediction errors are thought of as triggers for exogenous attention. When top-down processing is disturbed, possibly due to critical illness impacting upon baseline connectivity, predictions become increasingly less precise, which results in increased prediction errors [57,21]. In terms of endogenous (bottom up) attention, there is also empirical evidence which suggests a link between PC and this form of attention. In this context, diminished influence of bottom up signalling upon perception is proposed to result in decreased functional connectivity or decreased bottom up modulation of attentional processes [58,59]. The subject is then inundated with sensory driven analysis that constitutes overwhelming surprises and stimulation, i.e. delirium. However, attention is not a singular function but is based upon the arousal system of consciousness and also is a function of the interface between executive cognition and the network of unconscious complexes/schemas. According to Beck, schemas are defined as ‘‘A structure for screening, coding, and evaluating the stimuli that impinges on the organism. On the basis of the matrix of schemas, the individual is able to orient himself in relation to time and space, and to categorize and interpret experiences in a meaningful way” [60]. Subsequent theoretical developments have integrated findings form neuroscience to expand the theory of schemas to include its multimodal representation [61]. Many attempts at investigating complexes from neurophysiological perspective were performed during the 1980’s using Quantitative ElectroEncephaloGraph (QEEG) techniques [62]. The faculties of attention and orientation are guided by unconscious processes in the form of schemas. Recent studies have reported that the process of attention was significantly improved when a complex was activated. Subject performance on implicit learning tasks also improved when a complex was activated [63]. Such a finding confirmed Jung’s reports that complexes had a directional influence upon attention [64]. More complicated still, attention is modulated by the locus of working memory, a form of attention directed memory [65]. Therefore, working memory can be conceived as being a component of ego-consciousness which enables the binding together of transient functions of working memory e.g. visuospatial sketchpad, the phonological loop and the episodic buffer. These transient functions are also working in concert with more established systems such as long term memory, language and visual semantics [66]. The ego being the most integrative and dominant complex attempts to execute its functions via the bi-directional input from the complexes. However, given the overlap and high density integration of these complexes, the difference between the ego and the other complexes in the healthy individual must be minimal. Indeed studies conducting memory performance tasks and measuring the difference between self and other have demonstrated that there are reduced differences when the other is closer to one’s self e.g. friends, family etc. [67,68]. Regions dedicated to this task have been associated with the Prefrontal cortex and cortical midline structures (e.g. MPFC and PMC) [69,70]. Neuroimaging studies have reported that CMSs have activation patterns that reflect this. In particular, the medial prefrontal cortex (MPFC) exhibits similar neural activity when traits for self and a close other are being tested. The reverse has also been demonstrated [71]. It has also been proposed that the MPFC is differentially activated by self and other, with the most ventral areas more active for self and

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77

more dorsal areas more active for other [72]. Delirium has been proposed as a transient state of CNS integrity breakdown, such breakdown in these systems reflects the neural substrate of delirium phenomenology [7]. The degree of integrity between these systems of PC, is significantly disrupted by critical illness, pro inflammatory mechanisms, prior neurocognitive impairment, and reduced neurocognitive reserve. Sensory driven analysis is not altering prediction errors and hence a significant mismatch between the systems occurs. Such disconnect explains the association between the severity of delirium, its outcomes, and its phenomenology. For example, in hyperactive delirium, there is more florid manifestations of features e.g. delusions, hallucinations, while with hypoactive delirium there is significant cognitive impairment, where there is a clear lack of narrative structure to psychotic features [8].

Revisiting delirium assessment Cognitive efficiency within the PC model is a central part of the pathology of cognitive failure and therefore measures of this efficiency would indicate in clear terms the progress and remission of such a failure. Attentional deficits are a critical component to the phenomenology of delirium and inattention has been categorised in terms of focussing, shifting and sustaining attentional processes [31]. Rudimentary bedside assessment of these aspects can be understood in terms of how the patient orientates to the presentation of stimuli by the examiner including stimuli emanating from the examiner(e.g. their voice when giving simple commands or questions) [73]. However, attention is not solely a simple scanning reflex of the psyche but is a convergent function of the unconscious psyche (motivational drives guided by the schemas) and ego consciousness (the virtual sensorium mediated by working memory and visuospatial processing). Visuospatial processing and working memory are well recognised functions linked to the domain of executive cognition and assessing age related declines in cognition in general [74]. In elderly populations, patients with Alzheimer’s disease have been shown to have significant deficits in visuospatial processing and working memory [75]. Classical methods of assessing the virtual sensorium include the clock drawing test, and although has high utility for differentiating those who have cognitive impairment from healthy controls, it does not have any apparent discriminatory capacity between dementia and delirium [76]. The term ‘dysexecutive syndrome’ has been used to account for this age related and pathological dysfunction between these functional systems as measured by spatial span tests [66]. Assessing visuospatial processing can be done using the spatial span subtest derived from the Wechsler Memory scale which states that its purpose as a measure that ’‘taps an examinee’s ability to hold a visual spatial sequence of locations in working memory and then reproduce the sequence’ [77]. The spatial span forward (SSF) and spatial span backwards (SSB) are based upon measures of the integrity of the virtual sensorium of the individual [78]. In the context of delirium it has been found that the SSF may be useful in differentiating delirium from dementia due to the preserved short term features of certain forms of dementia. It has been proposed that the SSF emphasises attention over the faculty of working memory and hence is capable of enabling examiners differentiate delirium from dementia [73]. More recent work has proposed the use of the Edinburgh Delirium Test Box, which is a series of tests aimed at measuring sustained visual attention and has been shown to discriminate delirium from dementia and cognitively robust cohorts [79]. An abbreviated form of the cognitive test for delirium (CTD) which uses two (recognition memory for pictures and visual attention span) out of a total of nine items has demonstrated both reliability and validity in dis-

75

crimination for delirium [80]. Advances in bioelectronic technology have yielded encouraging results with the introduction of eye tracking technology, smart phone based testing applications, and oculomotor analysis to further develop means of detecting delirium and its severity [53,81,82]. Although delirium is considered a unitary syndrome there is a significant gap in the literature discussing the trajectory of individual features over the course of delirium [7]. Modern conceptualisations of delirium have been influenced by the assessment tools used to assess, detect and analyse its complex and transient nature [83,84]. Studies analysing the phenomenology of delirium have previously been based on cross sectional methods and have thus presented a static picture of delirium. Given that delirium is a fluctuating and often reversible condition, an accurate analysis of its phenomenology must take into account these key features. Therefore, accurate characterisation of its nature must be based on longitudinal analysis of it [85]. Methods for analysing longitudinal data in delirium research have been identified with the generalised estimating equation (GEE) method particularly suited for studying inter subject differences i.e. population specific patterns. Whereas mixed effects modelling (MRM) has been identified as highly applicable to elucidating the phenomenological profile for individual patients [85,86]. The relationship between biological and clinical factors has been explored through a combination of GEE and MRM [85]. Examples of MRM application have been the relationship of S100B protein to delirium status, and the identification of delirium as a state of broad based neural dysfunction (7, [87]) Summary and recommendations In prodromal, sub/full syndromal delirious states, there is an acute onset, qualitative and quantitative breakdown in the functional integrity of the CNS. Such a breakdown is marked by a compensatory reconfiguration of the remnants of primary and secondary consciousness as experienced by the typical virtual sensorium the unitary syndrome of delirium. This unitary syndrome is the result of the reconfigured experiences of reality as abstractly encoded by the neural network of complexes and the retained perceptual system of the psyche. Such a constellation of experiences shapes motivation and hence guides the attention and orientation systems through the interface of ego consciousness. A breakdown at this interface is detected as the heterogeneous phenomenology that is pronounced in delirium, typically inattention and disorientation. Taken together this elucidates the historically qualitative descriptions of delirium as a clouding of consciousness. Applying the PC model to delirium enables researchers to identify at what point the insults of critical illness interact with the baseline connectivity of the human CNS with a view to developing either treatment strategies or better detection methods. Therefore, having a neuroscientifically informed foundation for the phenomenology of delirium may further enhance researchers to refine tests that measure the breakdown in the patient’s virtual sensorium and further enable them to distinguish between other co-morbid conditions such as dementia and depression. Conflicts of interest The author declares no conflict of interest. References [1] Witlox J, Eurelings LS, de Jonghe JF, Kalisvaart KJ, Eikelenboom P, Van Gool WA. Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementiaa meta-analysis. JAMA 2010;304(4):443–51.

76

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77

[2] Fick DM, Steis MR, Waller JL, Inouye SK. Delirium superimposed on dementia is associated with prolonged length of stay and poor outcomes in hospitalized older adults. J Hosp Med 2013;8(9):500–5. [3] Greiner GFC. Der Traum und das fi eberhafte Irreseyn. Ein physiologischpsychologischer Versuch; 1817 [Altenburg und LeipzigF. A. Brockhaus]. [4] Lipowski ZJ. Deliriumhow its concept has developed. Int Psychogeriatr 1991;3:115–20. [5] Hogan RE, Kaiboriboon K. The ‘‘dreamy state”John Hughlings-Jackson’s ideas of epilepsy and consciousness. Am J Psychiatry 2003;160:1740–7. [6] Morandi A, Rogers BP, Gunther ML, Merkle K, Pandharipande P, Girard TD, et al. The relationship between delirium duration, white matter integrity, and cognitive impairment in intensive care unit survivors as determined by diffusion tensor imaging. Crit Care Med 2012;40(7):2182. [7] Leonard M, Adamis D, Saunders J, Trzepacz P, Meagher D. A longitudinal study of delirium phenomenology indicates widespread neural dysfunction. Palliat Support Care 2013:1–10. [8] Trzepacz PT, Meagher DJ. Neuropsychiatric aspects of delirium. In: Yudofsky SC, Hales RE, editors. American psychiatric publishing textbook of neuropsychiatry. Washington, DC: American Psychiatric Publishing; 2009. p. 445–517. [9] Franco JG, Trzepacz PT, Mejía MA, Ochoa SB. Factor analysis of the Colombian translation of the delirium Rating Scale (DRS), revised-98. Psychosomatics 2009;50(3):255–62. [10] Kean J, Trzepacz PT, Murray LL, Abell M, Trexler L. Initial validation of a brief provisional diagnostic scale for delirium. Brain Inj 2010;24(10):1222–30. [11] Meagher DJ, Morandi A, Inouye SK, Ely W, Adamis D, Maclullich AJ, et al. Concordance between DSM-IV and DSM-5 criteria for delirium diagnosis in a pooled database of 768 prospectively evaluated patients using the delirium rating scale-revised-98. BMC Med 2014;12(1):164. [12] European Delirium Association. The DSM-5 criteria, level of arousal and delirium diagnosisinclusiveness is safer. BMC Med 2014;12(1):141. [13] Adams RA, Stephan KE, Brown HR, Frith CD, Friston KJ. The computational anatomy of psychosis. Front Psychiatry 2013;4. [14] Hobson JA, Hong CCH, Friston KJ. Virtual reality and consciousness inference in dreaming. Front Psychol 2014;5. [15] Bastos AM, Usrey WM, Adams RA, Mangun GR, Fries P, Friston KJ. Canonical microcircuits for predictive coding. Neuron 2012;76(4):695–711. [16] Friston K. Predictive coding, precision and synchrony. Cogn Neurosci 2012;3 (3–4):238–9. [17] Adams, R. A., Brown, H. R., & Friston, K. J. (2015). Bayesian inference, predictive coding and delusions. AVANT Vol. V, No. 3, 51. [18] Ashby WR. Principles of the self-organizing dynamic system. J Gen Psychol 1947;37(2):125–8. [19] Stephan KE, Baldeweg T, Friston KJ. Synaptic plasticity and dysconnection in schizophrenia. Biol Psychiatry 2006;59(10):929–39. [20] Harrison PJ, Lewis DA. Neuropathology of Schizophrenia. 2nd ed. Schizophrenia; 2011. p. 310–25. [21] Feldman H, Friston KJ. Attention, uncertainty, and free-energy. Front Hum Neurosci 2010;4. [22] Partridge JS, Martin FC, Harari D, Dhesi JK. The delirium experiencewhat is the effect on patients, relatives and staff and what can be done to modify this? Int J Geriatr. Psychiatry 2013;28(8):804–12. [23] Merleau-Ponty M. Phenomenology of perception (transSmith C). London: Kegan Paul; 1962. [24] Lacan J. Some Reflections on the Ego. Int J Psycho-Anal 1953;34:11. [25] Sanders RD. Hypothesis for the pathophysiology of deliriumRole of baseline brain network connectivity and changes in inhibitory tone. Med Hypotheses 2011;77(1):140–3. [26] Choi SH, Lee H, Chung TS, et al. Neural network functional connectivity during and after an episode of delirium. Am J Psychiatry 2012;169:498–507. [27] Araujo HF, Kaplan J, Damasio A. Cortical midline structures and autobiographical-self processes an activation-likelihood estimation metaanalysis. Front Hum Neurosci 2013;7. [28] Brooks SJ, Savov V, Allzen E, Benedict C, Fredriksson R, Schiöth HB. Exposure to subliminal arousing stimuli induces robust activation in the amygdala, hippocampus, anterior cingulate, insular cortex and primary visual cortexa systematic meta-analysis of fMRI studies. Neuroimage 2012;59(3):2962–73. [29] Rizzolatti, G., & Fabbri-Destro, M. (2013). The mirror mechanismUnderstanding others from the inside. Understanding Other MindsPerspectives from developmental social neuroscience, 264. [30] Chester JG, Beth Harrington M, Rudolph JL. Serial administration of a modified Richmond Agitation and Sedation Scale for delirium screening. J Hosp Med 2012;7(5):450–3. [31] World Health Organization. International statistical classification of diseases and related health problems, 10th revision. Geneva: World Health Organization; 1993. [32] Gunther ML, Jackson JC, Ely EW. Loss of IQ in the ICU brain injury without the insult. Med Hypotheses 2007;69:1179–82. [33] Balduzzi D, Tononi G. Qualiathe geometry of integrated information. PLoS Comput Biol 2009;5:e1000462. [34] Edelman GM. Wider than the skythe phenomenal gift of consciousness. Yale University Press; 2004. [35] Johanson M, Revonsuo A, Chaplin J, Wedlund JE. Level and contents of consciousness in connection with partial epileptic seizures. Epilepsy Behav 2003;4:279–85.

[36] Edelman GM. Naturalizing consciousnessa theoretical framework. Proc Natl Acad Sci 2003;100(9):5520–4. [37] Searcy WA, Nowicki S. The evolution of animal communicationreliability and deception in signaling systems. Princeton University Press; 2005. [38] Jackendoff R. Foundations of languageBrain, meaning, grammar, evolution. Oxford University Press; 2002. [39] Pinker S. The stuff of thought. London: Allen Lane; 2007. [40] Trzepacz PT. Update on the neuropathogenesis of delirium. Dement Geriatr Cogn Disord 1999;10:330–4. [41] Damasio, A. (2011) Self Comes to mindconstructing the conscious brain (Vintage). [42] Cabeza R, St. Jacques P. Functional neuroimaging of auto-biographical memory. Trends Cogn Sci 2007;11:219–27. [43] Platek SM, Wathne K, Tierney NG, Thomson JW. Neural correlates of self-face recognition:an effect-location meta-analysis. Brain Res 2008;1232:173–84. [44] Tacikowski P, Brechmann A, Marchewka A, Jednoróg K, Dobrowolny M, Nowicka A. Is it about the self or the significance? An fMRI study of self-name recogni- tion. Soc Neurosci 2011;6:98–107. [45] Rogers TB, Kuiper NA, Kirker WS. Self-reference and the encoding of personal information. J Pers Soc Psychol 1977;35:677–88. http://dx.doi.org/10.1037/ 0022-3514. 35.9.677. [46] Northoff G, Heinzel A, deGreck M, Bermpohl F, Dobrowolny H, Panksepp J. Selfreferential processing in our brain– a meta-analysis of imaging studies on the self. Neuroimage 2006;31:440–57. [47] Hobson, J. A. (1999). Dreaming as deliriumHow the brain goes out of its mind. Mit Press. [48] Jacobson SA, Dwyer PC, Machan JT, Carskadon MA. Quantitative analysis of rest-activity patterns in elderly postoperative patients with deliriumsupport for a theory of pathologic wakefulness. J Clin Sleep Med 2008;4(2):137. [49] Hobson JA. REM sleep and dreamingtowards a theory of protoconsciousness. Nat Rev Neurosci 2009;10(11):803–13. [50] Grover S, Shah R. Distress due to delirium experience. Gen Hosp Psychiatry 2011;33(6):637–9. [51] Curtiss G, Vanderploeg RD, Spencer J, Salazar AM. Patterns of verbal learning and memory in traumatic brain injury. J Int Neuropsychol Soc 2001;7 (5):574–85. [52] Dobbs BM, Dobbs AR, Kiss I. Working memory deficits associated with chronic fatigue syndrome. J Int Neuropsychol Soc 2001;7(3):285–93. [53] Tieges Z, Brown LJ, MacLullich AM. Objective assessment of attention in deliriuma narrative review. Int J Geriatr Psychiatry 2014;29(12):1185–97. [54] Metzler-Baddeley C. A review of cognitive impairments in dementia with Lewy bodies relative to Alzheimer’s disease and Parkinson’s disease with dementia. Cortex 2007;43(5):583–600. [55] Kolanowski AM, Fick DM, Yevchak MAM, Hill MNL, Mulhall MPM, McDowell MJA. Pay attention!the critical importance of assessing attention in older adults with dementia. J Gerontol Nurs 2012;38(11):23. [56] Noudoost B, Chang MH, Steinmetz NA, Moore T. Top-down control of visual attention. Curr Opin Neurobiol 2010;20(2):183–90. [57] Happé F, Frith U. The weak coherence accountdetail-focused cognitive style in autism spectrum disorders. J Autism Dev Disord 2006;36(1):5–25. [58] Zhang W, Luck SJ. Feature-based attention modulates feedforward visual processing. Nat Neurosci 2009;12(1):24–5. [59] Liu T, Yuan Z, Sun J, Wang J, Zheng N, Tang X, et al. Learning to detect a salient object. IEEE Trans Pattern Anal Mach Intell 2011;33(2):353–67. [60] Beck AT. Depressioncauses and treatment. Philadelphia, PA: University of Pennsylvania Press; 1967. [61] James IA, Reichelt FK, Freeston MH, Barton SB. Schemas as memoriesimplications for treatment. J Cogn Psychother 2007;21(1):51–7. [62] Petchkovsky L, Petchkovsky M, Morris P, Dickson P, Montgomery D, Dwyer J, et al. FMRI responses to Jung’s Word Association Testimplications for theory, treatment and research. J Anal Psychol 2013;2013(58):409–31. [63] Shin Joong-Sun Lee, O-S Han, B-YR. The influence of complexes on implicit learning. J Anal Psychol, 502005175–90. [64] Jung CG. Studies in word association. Collected Works, Vol. 2. Trans R.F.C. Hull, Routledge1904. [65] Cowan, N. in Models of Working MemoryMechanisms of Active Maintenance and Executive Control (eds Miyake, A. & Shah, P.) 62–101 (Cambridge Univ. Press, New York, 1999). [66] Baddeley A. Working memorylooking back and looking forward. Nat Rev Neurosci 2003;4(10):829–39. [67] Markus HR, Kitayama S. Culture and the selfimplications for cognition, emotion, and motivation. Psychol Rev 1991;98:224–53. [68] Symons CS, Johnson BT. The self-reference effect in memorya meta-analysis. Psychol Bull 1997;121:371–94. [69] Parvizi J, van Hoesen GW, Buckwalter J, Damasio A. Neural connections of the posteromedial cortex in the macaque. Proc Natl Acad Sci USA 2006;103:1563–8. [70] Hagmann P, Cammoun L, Gigandet X, Meuli R, Honey CJ, Wedeen VJ, et al. Mapping the structural core of human cerebral cortex. PLoSBiol 2008;6:e159. [71] Ochsner KN, Beer JS, Robertson ER, Cooper JC, Gabrieli JDE, Kihsltrom JF, et al. The neural correlates of direct and reflected self-knowledge. Neuroimage 2005;28:797–814. [72] Amodio DM, Frith CD. Meeting of mindsthe medial frontal cortex and social cognition. Nat Rev Neurosci 2006;7:268–77. [73] Hall RJ, Meagher DJ, MacLullich AM. Delirium detection and monitoring outside the ICU. Best Pract Res Clin Anaesthesiol 2012;26(3):367–83.

J.M. FitzGerald / Medical Hypotheses 103 (2017) 71–77 [74] Van der Linden (1998). The relationships between working memory and longterm memory. Comptes Rendus de l Acadamie des Sciences, Serie III, Sciences de la vie (Paris), 321(2–3), 175–177. [75] Thompson JM, Hamilton CJ, Gray JM, Quinn JG, Mackin P, Young AH, et al. Executive and visuospatial sketchpad resources in euthymic bipolar disorderimplications for visuospatial working memory architecture. Memory 2006;14(4):437–51. [76] Adamis D, Morrison C, Treloar A, et al. The performance of the clock drawing test in elderly medical inpatientsdoes it have utility in the identification of delirium? J Geriatr Psychiatry Neurol 2005;18:129–33. [77] Bo J, Borza V, Seidler R. Age-related declines in visuospatial working memory correlate with deficits in explicit motor sequence learning. J Neurophysiol 2009;102:2744–54. [78] Wilde NJ, Strauss E, Tulsky DS. Memory span on the Wechsler Scales. J Clin Exp Neuropsychol 2004;26:539–49. [79] Brown LJ, Fordyce C, Zaghdani H, et al. Detecting deficits of sustained visual attention in delirium. J Neurol Neurosurg Psychiatry 2011;82:1334–40. [80] Hart RP, Best AM, Sessler CN, et al. Abbreviated cognitive test for delirium. J Psychosom Res 1997;43:417–23.

77

[81] Exton C, Leonard M. Eye tracking technologya fresh approach in delirium assessment? Int Rev Psychiatry 2009;21(1):8–14. [82] van der Kooi AW, Rots ML, Huiskamp G, Klijn FA, Koek HL, Kluin J, et al. Delirium detection based on monitoring of blinks and eye movements. Am J Geriatr Psychiatry 2014. [83] Adamis D, Sharma N, Whelan PJP, Macdonald AJD. Delirium scalesa review of current evidence, 14(5)2010, 543–555. [84] Adamis D, Slor CJ, Leonard M, Witlox J, de Jonghe JF, Macdonald AJ, et al. Reliability of delirium rating scale (DRS) and delirium rating scale-revised-98 (DRS-R98) using variance-based multivariate modelling. J Psychiatr Res 2013;47(7):966–71. [85] Adamis D. Statistical methods for analysing longitudinal data in delirium studies. Int Rev Psychiatry 2009;21:74–85. [86] Liu LC, Hedeker D. A mixed-effects regression model for longitudinal multivariate ordinal data. Biometrics 2006;62(1):261–8. [87] van Munster BC, Korse CM, de Rooij SE, Bonfrer JM, Zwinderman AH, Korevaar JC. Markers of cerebral damage during delirium in elderly patients with hip fracture. BMC Neurol 2009;9(1):21.