Dehydration is a reliable predictor of impaired cognitive status. Objective data, using tests of cortical function, support the deterioration of mental performance in mildly dehydrated younger adults. Dehydration frequently results in delirium as a manifestation of cognitive dysfunction. Although, the occurrence of delirium suggests transient acute global cerebral dysfunction, cognitive impairment may not be completely reversible. Animal studies have identified neuronal mitochondrial damage and glutamate hypertransmission in dehydrated rats. Additional studies have identified an increase in cerebral nicotinamide adenine dinucleotide phosphate-diaphorase activity (nitric oxide synthase, NOS) with dehydration. Available evidence also implicates NOS as a neurotransmitter in long-term potentiation, rendering this a critical enzyme in facilitating learning and memory. With ageing, a reduction of NOS activity has been identified in the cortex and striatum of rats. The reduction of NOs synthase activity that occurs with ageing may blunt the rise that occurs with dehydration, and possibly interfere with memory processing and cognitive function. Dehydration has been shown to be a reliable predictor of increasing frailty, deteriorating mental performance and poor quality of life. Intervention models directed toward improving outcomes in dehydration must incorporate strategies to enhance prompt recognition of cognitive dysfunction.
Thirst is a potentially unpleasant sensation that also serves as a strong driving force capable of influencing complex human behavior. In addition, the hedonic qualities associated with assuaging thirst are perceived as a reflection of the mitigation of the undesirable effects of fluid deprivation. Thus, although the precise pathophysiologic mechanisms have yet to be unraveled, the complexity of man's response to thirst and fluid deprivation indicates sophisticated cognitive processing.
The clinical effects of severe dehydration on cognitive function highlight the importance of maintaining optimal hydration status. Such effects are generally the result of profound hypovolemia and subsequent cerebral hypoperfusion (Figure 1). In contrast, the cognitive manifestations of mild dehydration have not been fully explored. The paucity of available data precludes detailed and focused exploration of the pathophysiological mechanisms that underlie disruption of cognitive function in less severe dehydrated states.
Available evidence indicates the increased susceptibility of older adults to dehydration and the resulting complications (Silver, 1987; Rolls & Phillips, 1990). Dehydration in older adults has been shown to be a reliable predictor of increasing frailty, progressive deterioration in cognitive function and an overall reduction in quality of life (Warren et al, 1994; Miller et al, 1998). These adverse long-term outcomes of dehydration in the elderly mandate further research into the pathophysiology and manifestations of mild dehydration in order to facilitate early recognition and prompt intervention. In addition, the increased susceptibility of older adults to the complications of dehydration provides a sensitive cohort in which the manifestations of subtle degrees of dehydration may be readily observed.
Available studies have identified several domains of cognitive function that are affected by dehydration (Sharma et al, 1986; Gopinathan et al, 1988). Gopinathan et al (1988) examined changes in cognitive function in 11 healthy adult subjects under varying degrees of dehydration induced by a combination of water restriction and heat stress. Resulting data indicated a significant correlation between cognitive dysfunction and severity of dehydration. Subjects exhibited progressive impairment in mathematical ability, short-term memory and visuomotor function once 2% body fluid deficit was achieved (Gopinathan et al, 1988). Cian et al (2000) demonstrated impaired long-term memory following dehydration resulting from heat stress. In a later study, Cian et al (2000) also demonstrated impairment in short- and long-term memory, visuospatial function, perceptive discrimination and reaction time in dehydrated subjects. The subjects studied also demonstrated increased subjective perception of fatigue following prolonged dehydration, validating the clinical association of hydration status with quality of life (Cian et al, 2001).
In the absence of an operational definition of cognition, several working models of cognitive awareness have been developed, based on unification of physiological, psychological and philosophical concepts. Of these, the most convincing is Barr's global workspace theory (Barr, 1993). This is based on the concept that cognition is of limited capacity. Thus, cognitive processes perceived as working in parallel are actually functioning in competition with each other. Selected activities are presumed to dominate cognitive awareness by harnessing executive function more effectively than other competing processes. Utilizing Barr's theory, Cohen attempts to explain the effect of dehydration on cognitive function in a simplistic, albeit, rather plausible manner (Cohen, 1983). Based on the assumption that complex tasks require increased attention, Cohen proffers that acute stressors such as dehydration compete for executive attention and awareness with parallel processes occurring in other cognitive domains, thereby compromising overall cognitive performance.
In the light of the obvious complexity of the neurobiological mechanisms involved in cognition, it is unlikely that a unitary explanation will emerge to account for cognitive dysfunction in dehydrated states. Current research trends driven by hypotheses based on the integration of cellular and hormonal theories, as explanations for cognitive dysfunction in dehydrated states, are perhaps more appropriate (Table 1).
The activation of the renin–angiotensin system in response to hypohydration, resulting in an increase in arginine vasopressin (AVP), is well documented. Early studies suggest that prostaglandin E (PGE) plays a pivotal role in the modulation of AVP release in hypohydrated states (Leskell, 1976; Weitzman & Kleeman, 1979). In animal studies, PGE has been shown to augment the release of AVP in response to injection of hypertonic saline (Yamamoto et al, 1976). Studies seeking to identify other biological indicators of hydration status have shown an increase in serum cortisol levels with dehydration. Achieving euhydration, in such cases, is associated with normalizing of the serum cortisol levels (Francesconi et al, 1989). The results of the aforementioned studies favor the hypothesis that cognitive dysfunction resulting from mild dehydration may result from the central effect of alterations in the hormonal profile. However, review of more recent studies reveals conflicting results. Animal studies indicate that although corticosteroid hormones may not influence passive learning, hypercortisolemia tends to worsen active learning and compromise short-term memory (Vedhara et al, 2000).
The deleterious effects of pharmacological steroids on cognitive function are well recognized. However, Newcomer et al (1999) also demonstrated a reduction in verbal memory following the administration of exogenous cortisol at physiological doses. In contrast, Buchanan and Lovallo (2001), examining the effect of exogenous cortisol administration on human memory, demonstrated enhancement of memory specifically associated with emotionally arousing information. Studies have also demonstrated an association between hypocortisolemia and enhanced short-term memory, although there was no correlation with verbal memory (Vedhara et al, 2000). The theory of a hormonal basis for cognitive dysfunction resulting from dehydration is further challenged by the results of several studies identifying a positive effect of cerebral DDAVP on memory (Beckwith et al, 1990). Elaborate cell culture studies have also identified increased neurite length and bifurcation points following exposure to vasopressin receptor agonists. These findings challenge the role of peripheral AVP in the genesis of cognitive dysfunction resulting from dehydration. It remains unclear however as to whether there is a linear correlation between peripheral and central levels of AVP. Nevertheless, the identification of vasopressin-induced neurotrophism warrants further research into the effect of hormones on cognitive function in dehydrated persons (Chen et al, 2000).
Additional evidence also implicates central neurotransmitters in the genesis of cognitive dysfunction in dehydrated persons. Animal studies reveal an increased density of nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-diaphorase) in forebrain circumventricular structures of rats following a 72-h period of dehydration (Ciriello et al, 1996). Notably, cellular and histochemical studies indicate that NADPH-diaphorase and nitric oxide synthase (NOS) share identical ultrastructural locations (Calka et al, 1994). These data indicate a likely role for NOS as a mediator in central homeostatic mechanisms regulating fluid balance. Nitric oxide (NO) has gained increasing recognition as a critical neurotransmitter molecule. Available data on ultrastructural enzyme location reveal that NOS is present in most parts of the brain and plays a crucial role as either a retrograde messenger or a paracrine factor in facilitating long-term potentiation of memory (Salemme et al, 1996). Animal studies support the role of NO as a central diffusible messenger in facilitating learning and memory. Additional studies identifying reduced NO production in older rats suggest that NO plays a role in the genesis of age-related memory impairment (Noda et al, 1997). Based on these data, aging may result in a relative reduction in the compensatory increase of NO with dehydration, thereby explaining the increased susceptibility of older adults to delirium following relatively mild degrees of dehydration.
Current literature permits only hypothetical formulation of the role of neurotransmitters in the cognitive manifestations of dehydration. Similarly, the cellular mechanisms involved in the cognitive manifestations of dehydration are yet to be objectively elucidated. Experimental models examining neuronal vulnerability to different acute insults suggest alternative theories of dehydration-related neurotoxicity.
Regardless of the initiating insult, neuronal death tends to be initiated by pathological membrane depolarization that results in a critical accumulation of intracellular calcium (Lee et al, 1999). The precise chain of events as it relates to the particular insult is still unclear. In addition, the observation that neuronal subtypes differ in their degree of vulnerability to acute insults complicates this area of study even further (Haddad & Jiang, 1993). However, neuronal subtypes that are more vulnerable are more likely to be affected by milder insults. The striatum is one example of an extremely vulnerable region that serves as a convenient experimental model for the study of cellular response to injury (Calabresi et al, 2000). This has led to the development of several interesting theories. The energy deprivation theory is based on the effect of acute cellular injury in disrupting mitochondrial function. The resultant loss of ATP-dependent ionic segregation and subsequent failure to maintain the normal ionic gradient triggers inappropriate membrane depolarization. Consequently, voltage-dependent calcium channels are activated resulting in intracellular calcium accumulation and eventually neuronal death (Martin et al, 1994; Dimagl et al, 1999; Lee et al, 1999).
Excitotoxicity resulting from glutamate hypertransmission is another favored theory that may account for the neuronal response to acute insults (Choi & Rothmann, 1990; Obrenovitch & Urenjjak, 1994). Two categories of glutamate receptors have been identified, namely, ionotropic glutamate (iGlu) receptors linked to cation channels and metabotropic glutamate (mGlu) receptors that are coupled to second messenger systems. The latter group of receptors is further divided into three subgroups. Subgroup 1 promotes intracellular calcium mobilization, while subgroups 2 and 3 are negatively coupled to adenylate cyclase activity. Acute insults result in sustained glutamate-mediated excitatory activation of both iGlu and mGlu receptors, resulting in the activation of an intracellular metabolic pathway that leads to increased enzymatic activity of phospholipase C and increased accumulation of intracellular calcium (Chen et al, 1996; Tallaksen-Greene & Albin, 1996; Bernard et al, 1997) (Figure 2).
Glutamate hypertransmission may also provide the link between cellular dehydration and altered cellular energetics. Häusinger et al (1993) working with isolated cells showed that cellular dehydration triggers increased protein catabolism. Increased tissue liberation of glutamine, which is the most abundant free amino acid, may result from this process. Data indicating a reduction in the intracellular concentration of glutamine following tissue injury supports this theory (Rennie et al, 1989).
Increased elaboration of cytokines and subsequent stimulation of neuroendocrine activity are central features of the metabolic response to acute insult (Hill & Hill, 1998). However, the specific effect of cytokines in the mediation of the cognitive response to dehydration is still unclear. Data implicating tumor necrosis factor (TNF) and interleukin-1 (IL-1) as mediators in the acute phase response complicating thermal injuries may prove relevant to heat induced dehydration (Mester et al, 1994). However, available evidence indicates that systemic cytokine release fails to account for all the observed responses to acute insult and tissue injury. With the exception of IL-6, systemic cytokine elaboration cannot be consistently identified in patients with acute injury (Hill & Hill, 1998). Tissue-specific cytokine elaboration may prove to be a more attractive hypothesis. Animal studies have identified TNF receptors in the murine brain and elaborate IL-1 nerve fibers within the hypothalamus of rats. Rat astrocytes have also been shown to produce of TNF in vivo (Lieberman et al, 1989; Kinouchi et al, 1991; Hill et al, 1996). Additional studies have implicated several cytokines, namely IL-1, 2,6 and granulocyte–macrophage CSF in hippocampal neurotransmitter modulation (Bianchi et al, 1998). Thus, although the roles of cytokines and cytokine receptors are yet to be fully elucidated, existing data justifies further research in this area.
Homeostasis is a complex, yet well integrated, set of physiological responses designed to oppose deviation from the norm. In the majority of cases, such responses are well coordinated and result in a rapid restoration of normal function following the activation of a chain of appropriate compensatory responses. However, in a few cases, especially in the presence of multiple coexisting insults, an excessive homeostatic response may occur that paradoxically triggers damaging processes. Thus, although physiological derangement resulting from fluid loss is very often treated rather simplistically with replacement therapy, it is likely that high-risk patients may benefit from metabolic manipulation in order to combat the threat of inappropriate activation of homeostatic responses.
Although the cellular and hormonal events associated with the neuronal response to dehydration are yet to be fully elucidated, there is an abundance of clinical evidence implicating dehydration as a common precipitant of acute confusion (Levkoff et al, 1988; Hoffman, 1991; Francis & Kapoor, 1992; Murray et al, 1993; Bianchi et al, 1998; Mentes et al, 1998). Delirium is a common manifestation of dehydration that clearly reflects the global impact of dehydration on cerebral function. However, the three areas of the brain most vulnerable to the effects of dehydration are the reticular activating system, which subserves attention and wakefulness; the autonomic structures that regulate psychomotor and regulatory functions; and the cortical and mid-brain structures that are responsible for thought, memory and perception (Neelon & Champagne, 1992).
The syndrome of delirium is characterized by the transient nature of its occurrence and the potential reversibility inherent in this diagnosis. Thus, clinicians often assume that the identification and treatment of dehydration with appropriate replacement therapy should result in complete resolution of any associated cognitive dysfunction. However, the lack of data demonstrating a firm correlation between objective indices of tissue hydration and cognitive function challenges the notion that dehydration presents exclusively as acute confusion and is inherently a relatively benign and reversible condition.
Neelon and Champagne's (1992) models for the patterns of onset of acute confusion may be extended to incorporate parallel changes in the patterns of onset of hydration levels (Figure 3). Particularly within vulnerable subgroups of the population such as the elderly, this model may enhance awareness and detection of the complete spectrum of cognitive abnormalities that may result from fluid depletion.
Within the older population, cognitive impairment has been shown to herald the onset of functional decline in dehydrated patients. This course of events may be attributed to the hierarchical sequence of the domains of impact of dehydration, namely, cognitive function, task processing, functional decline and quality of life. This sequence further underscores the importance of the development and use of an appropriate intervention model, based not only on the simplistic and defensive theory of replacement, but also on more long-term anticipatory strategies.
An objective approach to the clinical detection and management of dehydration is hampered by the paucity of data defining the pathophysiological and cellular mechanisms underlying this disease process. The syndrome of dehydration is relatively unique, in that meaningful research must incorporate parallel exploration of quantitative indices of hydration status and outcome measures. The development and validation of risk assessment and monitoring tools is a critical component of this process. In addition, research into adjunctive measures aimed at limiting adverse effects of dehydration is warranted. Evidence of neurotoxicity resulting from dehydration suggests that traditional fluid replacement may be inadequate as sole therapy for cognitive dysfunction arising from dehydration. Further research into the use of innovative therapies such as prostaglandin inhibitors, NO modulators and antibodies targeting specific cytokines may prove useful.