Elsevier

Medical Hypotheses

Volume 77, Issue 1, July 2011, Pages 140-143
Medical Hypotheses

Hypothesis for the pathophysiology of delirium: Role of baseline brain network connectivity and changes in inhibitory tone

https://doi.org/10.1016/j.mehy.2011.03.048Get rights and content

Abstract

Normal brain function is facilitated by a highly organized and interconnected structure allowing complex integration of sensory information and motor responses. The acute confusional state of delirium is characterized by a fluctuating disturbance in consciousness, arousal level and cognition-memory; as such, delirium represents a failure in the integration and appropriate processing of information. The pathogenesis of this cognitive disintegration is unclear; herein a hypothesis is proposed that delirium results from an acute breakdown in network connectivity within the brain. The hypothesis predicts that the extent to which the network connectivity breaks down is dependent on two factors: (i) the baseline connectivity within the brain and (ii) the level of inhibitory tone. Baseline connectivity is the connectivity of neural networks within the brain before the precipitating insult provoking delirium. Many non-modifiable risk factors for delirium influence baseline connectivity such as age, cognitive impairment, dementia and depression. Precipitant events that provoke delirium (modifiable risk factors) are hypothesized to further, and acutely, breakdown network connectivity by increasing inhibitory tone within the brain. Modifiable risk factors include inflammation, metabolic abnormalities, sleep deprivation and medication such as benzodiazepines. An important role for GABAergic neurotransmission is implicated in increasing the inhibitory tone to produce delirium. This theory accounts for the various forms of delirium, hypoactive, hyperactive and mixed. The form of delirium that ensues will depend upon how and which networks breakdown (dependent on both the individual’s baseline network connectivity and the degree change in inhibitory tone produced).

Introduction

Delirium is an acute confusional state characterized by fluctuating consciousness, arousal level and cognitive function that is prevalent in both seriously ill and perioperative patients. Delirium affects 14–24% of hospital admissions, 15–53% of postoperative patients, and 70–87% of critical care patients and is associated with increased long-term morbidity and mortality [1], [2], [3]. Despite the prevalence of this condition we remain relatively ignorant of the pathogenesis of delirium. In this article a tentative hypothesis is described that ties together several pathogenic strands for future inquiry.

The brain is a highly organized and interconnected structure functioning to allow complex integration of sensory information and motor responses. The acute confusional state of delirium represents a variable failure in the integration and appropriate processing of sensory information and motor responses. Specifically consciousness, arousal level and memory are disturbed. Below a hypothesis is proposed that two important factors determine a subject’s vulnerability to delirium:

  • 1.

    Baseline network connectivity: many of the non-modifiable risk factors for delirium affect baseline network connectivity within the brain including age, pre-existing cognitive impairment, dementia and depression.

  • 2.

    Inhibitory tone: the modifiable risk factors for delirium, including inflammation, metabolic abnormalities, sleep deprivation and medication (notably GABAergic medication) enhance the risk of delirium by increasing inhibitory tone within the brain. Differing insults may precipitate delirium by affecting inhibitory tone in different ways including upregulation of GABAA receptors, increased synthesis of endogenous GABA agonists and stimulation by exogenous GABA agonists.

The theory herein predicts that an acute upregulation of inhibitory tone within the brain acts to further disrupt network connectivity in vulnerable patients (predisposed by reduced baseline connectivity) precipitating delirium. The baseline level of network connectivity represents the ease in which network connectivity is broken down. Put another way, if there is already reduced baseline network connectivity then a smaller increase in inhibitory tone is required to breakdown the network than if baseline connectivity is normal. The degree change in inhibitory tone is the amplitude to which network connectivity is broken down. Therefore a large change in inhibitory tone will produce a big change in network connectivity. This is represented by the equation below:Risk of delirium=Δinhibitory tone×1/baseline network connectivity=k(ΔGABAAagonist activity)×(ΔGABAAreceptor expression)baseline network connectivity

By affecting separate neural networks to differing degrees different phenotypes of the delirium syndrome may become evident: hyperactive, hypoactive and mixed. This would likely be dependent both on the manner in which inhibitory tone changed and the level of baseline connectivity between different brain regions. For example, differing effects on arousal networks may explain hyperactive versus hypoactive delirium. Hallucinations may occur in some patients due to alterations in frontal corticothalamic and limbic networks. Similarly a breakdown in network connectivity in frontal-limbic systems may precipitate confusion due to memory impairment. Beyond the acute changes of delirium, a long-term alteration in connectivity may explain the persistent cognitive dysfunction of intensive care unit patients. A more detailed presentation of the hypothesis is presented below focussing on how consciousness, arousal level and memory may be disturbed. Following this further detail is provided to account for how non-modifiable risk factors contribute to baseline network connectivity and precipitant-modifiable risk factors accounts for changes in inhibitory tone.

Section snippets

Pathogenic framework: delirium a disorder of consciousness

Delirium can be considered a disorder of consciousness as experience is altered; in the delirious state delusions and hallucinations may occur as a patient experiences their own, alternate reality. Recent evidence supports the proposition that consciousness relates to a state of integrated information [4] and that a breakdown of the integration of information is associated with reduced consciousness [5], [6]. (A surrogate of the integration of information, effective connectivity, has been used

Pathogenic framework: delirium a disorder of arousal

Delirium is also a disorder of attention where arousal states may fluctuate significantly. With aging, people more frequently transition between awake and sleep states. It is proposed that orexin neurotransmission stabilizes the wake–sleep states [11]. In the elderly the increased fluctuation in arousal states may be related to reduced orexin neurotransmission within the aging brain [12]. This may make the aging brain intrinsically predisposed to fluctuations in arousal. Changes in inhibitory

Pathogenic framework: delirium a disorder of memory

Delirium is also a disorder of cognition-memory that may relate to both the impairment of consciousness and arousal but may also be dependent on direct effects on the corticolimbic system. Both retrieval of long-term memories and short-term memory formation appears impaired leading to disorientation. Intertwined with impaired attention and consciousness, the impaired memory formation likely produces a state of acute confusion. GABAA agonist drugs exert profound effects on memory encoding and

Specific role of non-modifiable risk factors

Age [18], pre-existing cognitive impairment [19], depression [20] and dementia [21], [22] are all associated with reduced functional connectivity in the human brain. With aging various explanations have been advanced for this finding, notably effects on white matter integrity, gray matter volume and neurotransmission. White matter integrity is of particular interest as it represents axonal connections between brain regions and thus may represent an anatomical correlate of functional

Inflammation and metabolic abnormalities

Central nervous system inflammation is common in acute illness from responses to infection, trauma, cardiovascular disease or metabolic disorders. Inflammation drives an upregulation in the expression of GABAA receptors both in the brain and peripherally [26], [27]. Preliminary evidence suggests that inflammation may drive a simultaneous increase in the expression of glutamic acid decarboxylase, GAD, that leads to an increase in GABA synthesis [27] and to the synthesis of other endogenous

Limitations of the theory

A theoretical framework for the pathophysiology of delirium is difficult to construct due to: (i) the heterogeneity in the manner of patient presentation, (ii) the number of diverse risk factors and (iii) the poor understanding of the pathophysiology to date. Nonetheless the theory herein ties together several essential strands of delirium: disturbance of consciousness, arousal and memory as well as addressing the role of the important risk factors. Nonetheless this hypothesis has avoided

Predictions

This hypothesis requires both preclinical and clinical evaluation prior to changes in clinical practice. Nonetheless this hypothesis would support the use of non-GABAergic sedative and anesthetic agents to reduce the burden of critical care and postoperative delirium. Through testing of this hypothesis novel therapeutics such as GABAA antagonists may also be developed to prevent delirium. Finally this theory advocates focusing on modifiable risk factors that increase inhibitory tone to prevent

Conclusion

To summarize, a hypothesis is proposed that delirium occurs when inhibitory tone rises within the brain to break down network connectivity within the brain. Non-modifiable risk factors, such as age, predispose to delirium as baseline network connectivity is reduced. Acutely, modifiable risk factors increase inhibitory tone to further breakdown network connectivity.

Conflict of interest

RDS has acted as a consultant for Air Liquide on the clinical applications of xenon.

Acknowledgment

RDS is supported by the Medical Research Council, United Kingdom. The author is grateful to Prof. Mervyn Maze (University of California San Francisco) for helpful advice on an early version of the manuscript.

References (37)

  • E.W. Ely et al.

    Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU)

    JAMA

    (2001)
  • G. Tononi

    An information integration theory of consciousness

    BMC Neurosci

    (2004)
  • M. Massimini et al.

    Breakdown of cortical effective connectivity during sleep

    Science

    (2005)
  • F. Ferrarelli et al.

    Breakdown in cortical effective connectivity during midazolam-induced loss of consciousness

    Proc Nat Acad Sci USA

    (2010)
  • D. Balduzzi et al.

    Qualia: the geometry of integrated information

    PLoS Comput Biol

    (2009)
  • S.K. Esser et al.

    Breakdown of effective connectivity during slow wave sleep: investigating the mechanism underlying a cortical gate using large-scale modeling

    J Neurophysiol

    (2009)
  • P. Pandharipande et al.

    Lorazepam is an independent risk factor for transitioning to delirium in intensive care unit patients

    Anesthesiology

    (2006)
  • C.B. Saper et al.

    Hypothalamic regulation of sleep and circadian rhythms

    Nature

    (2005)
  • Cited by (0)

    View full text