Uncomplicated general anesthesia in the elderly results in cognitive decline: Does cognitive decline predict morbidity and mortality?

https://doi.org/10.1016/j.mehy.2006.08.030Get rights and content

Summary

Elderly surgical patients constitute a unique surgical group. They require special consideration in order to preempt the long term adverse effects of anesthesia. This paper examines the proposition that general anesthesia causes harm to elderly patients with its impact being felt long after the anesthetic agents are cleared from the body. One complication, Postoperative Cognitive Decline (POCD), is associated with the administration of anesthesia and deep sedation. Its’ occurrence may herald an increase in morbidity and mortality. Based on both human and animal data, this paper outlines a unitary theoretical framework to explain these phenomena. If this hypothesis proves to be correct, anesthesiologist should consider regional rather than general anesthesia for equivalent surgical procedures to reduce POCD and consequently achieving superior patient outcome.

Introduction

Improvements in surgical and anesthetic methods are factors that have contributed to the increasing number of surgeries performed on elderly patients [1]. Of this cohort, patients of 85 year or older represent the fastest growing sector [2], [3]. It is now possible for such patients to safely undergo even the most invasive and extensive procedures. Although advances in perioperative care have significantly decreased perioperative mortality, these rates still remains high among the elderly. The crude perioperative mortality rate in the general population is 1.2% compared with 8.4% in patients over 90 years. These numbers are significantly worse when one considers major surgery only [4]. The postoperative mortality rates among elderly patients can be attributed to co-morbid conditions [5] including cognitive dysfunction.

The effect of cognitive dysfunction on postsurgery outcome in the elderly is not well understood. A more puzzling phenomenon is the Postoperative Cognitive Dysfunction (POCD), which affects elderly patients long after the anesthetic agents have been cleared from the body.

The implications of POCD are far reaching. Its occurrence significantly affects the patient immediately following surgery. Also the existence of this phenomenon also implies that the peri-operative mortality and morbidity period should be extended beyond the accepted 30 days. It is therefore important for the perioperative care team to understand this problem because of the emotional impact on the families and the financial consequences to the society.

We aim to develop a theoretical model that will generate further research. This will help us obtain an understanding of the etiology of this problem and help us to develop interventions designed to prevent this important complication.

In the first part of this paper, we present current knowledge about cognitive changes in the general elderly population and POCD. The following section presents a unified theoretical hypothesis that explains the etiology of POCD and a model to validate the hypothesis. We hypothesize that the administration of deep sedation and/or anesthesia can result in a number of neurophysiologic changes that produce POCD. When it occurs, POCD may be indicative of serious physiological alterations that could herald increased morbidity and mortality for the patient.

Cognition is the mental processes of perception, memory, and information processing. These mental processes are necessary for daily living and allow for acquisition of knowledge and problem solving. Cognition should not be confused with intelligence, which is the capacity to reason and plan.

Normal aging processes affect both the central and peripheral nervous systems [6]. This normal age related cerebral changes typically begin in middle age [7]. The cerebral changes constitute of reduction of the complexity of neuronal connections; reduction of the synthesis of neurotransmitters; increase of the postsynaptic degradation of neurotransmitters [8]. These changes reduce the capacity of the brain of an elderly person to integrate multiple neural inputs.

It is possible that old persons show no or minimal cognitive decline [9], however this form of age related deterioration progresses with advancing age [10]. Worsening of age related cognitive decline has been strongly associated with survival time [11]. Another form of cognitive deterioration has been identified that is not age related.

This can be characterized as representing a transitional state between the cognitive changes of normal aging and very early dementia. However, it is a distinct entity from these two conditions. The term is used to describe the non-surgical elderly population. Often, the initial cognitive changes in MCI are subtle [12].

MCI has several possible outcomes including: resolution [13], stabilization, or conversion to dementia and subsequent death. In fact it and is becoming increasingly recognized as a risk factor for the degenerative form of dementia termed Alzheimer disease (AD) [14], [15], [16]. It has been suggested that MCI may represent an early stage of dementia on a continuum of deterioration [14], [17]. When memory loss is the predominant symptom of MCI it is termed “amnestic MCI” and it is this that is thought to be a prodromal state of AD [18].

Rates of evolution from MCI to dementia have been quoted between 7% and 20% per year and up to 45% within 3 years [19], [20]. At the stage of diagnosis of dementia the risk of death increases. It is interesting to note that in all three forms of cognitive deterioration, the risk of death increases with worsening of the cognitive deterioration [21].

The incidence of MCI is twice as common as the more established dementia [22]. MCI is associated with increased morbidity [23] (i.e., diabetes [24], [25] and hypertension [26]) and mortality [27], [28], [29]. This association holds even after controlling for socio-demographic characteristics, co-morbidity, and functional limitation. Patients who suffer from a more severe degree of MCI have a shorter survival time [30], [31], [32].

Whereas MCI is a problem of the general geriatric population, POCD is unique to those who have undergone surgery.

POCD is a neurocognitive disorder defined as a “deterioration of intellectual function presenting as impaired memory or concentration” [33]. This condition has been well described in cardiac [34] and, to a lesser extent, in non-cardiac surgical patients [35]. Often, POCD is subtle in its presentation requiring neuropsychological testing before and after surgery to make the diagnosis [36].

Because of its generally subtle presentation, POCD is frequently overlooked. The disorder can affect isolated domains of cognition such as verbal memory, visual memory, language comprehension, visuo-spatial abstraction, attention, or concentration. POCD can have a significant impact on the patients’ life. If severe, it can limit the ability of the patient to live independently.

POCD is not uncommon in the elderly surgical population [32], [37], [38] and the estimated incidence is 15–25% of cases [39], [40], [41]. The incidence may be related to the type of surgery, medications received [42], and preexisting medical conditions [43]. POCD can last for weeks or months after surgery.

The disorder constitutes a risk factor for poor postoperative outcome [44], [45]. If patients have POCD, then at one week postoperatively it is more likely that the symptoms have gotten worse than improved. In fact, patients who exhibit POCD one week after surgery are more likely to suffer from some cognitive problem one to two years postoperatively [46]. After discharge from hospital, these patients encounter a reduction in their daily living activities and quality of life [35]. They are almost three times as likely to suffer further cognitive decline one to two years after surgery [46].

It has been estimated that 10% of elderly patients exhibit POCD three months after surgery. Few of those exhibiting POCD at 3 months also had the diagnosis at the 1-week postoperative test. Those patients with POCD at 3 months continued to deteriorate after surgery [47].

POCD has been associated with an increase in surgical morbidity and mortality [46]. Its avoidance and treatment therefore represents one of the greatest challenges for the perioperative physician dealing with an elderly surgical population.

The acute manifestation of POCD is delirium. It commonly occurs early during the postoperative course. The disorder is characterized by an acute temporary, variable disturbance of consciousness, attention, cognition, and perception [48]. Delirium is characterized by waxing and waning in its intensity even through a given day. Like MCI and POCD its occurrence is predictive of poor patient outcome. It is related with longer duration of hospital stay [49], higher rates of institutionalization, poor social adjustment [50], and mortality rates of almost 30% [51].

We can see from the discussion so far that the phenomena of MCI and POCD are similar in presentation. They differ in their context. As discussed previously, MCI occurs in the non-surgical population and POCD following an operative procedure. The subtle decline in cognitive functioning seen in both necessitates the use of a sensitive neuropsychological test battery for the detection.

Whether or not general anesthesia (GA) increases the possibility of developing long term cognitive dysfunction remains unclear. The investigations to date have had significant methodological problems.

Within the realm of geriatric anesthesia this issue forms part of the regional anesthesia (RA) versus GA debate. This is an old debate. It focuses on the superiority of one approach over another. It has produced few clear answers. Given its inconclusiveness the possibility still exists that the type of anesthesia administered could possibly affect the incidence of POCD. Probably one of the commonest surgical procedures that the elderly undergo is repair of a fractured hip. This is characterized by a high incidence of postoperative cognitive problems [52]. It has been postulated that elderly patients subjected to GA for such procedures may more frequently experience immediate postoperative cognitive problems compared to those who received a regional technique [53]. This latter observation being explained in terms of RA is based on blocking nerve impulses supplying a certain region of the body rather than the wider blocking of central nervous activity associated with GA. However, within the literature insufficient evidence exists to support or rule out significant clinically important differences between the techniques. Neither can conclusions be drawn concerning mortality nor other outcomes comparing RA vs. GA [53], [42].

Despite these observations it is still commonly believed that POCD and delirium are less common after RA than after GA [54]. Only one small trial [55] identified a statistically significant decline in cognitive outcome associated with general anesthesia. This trial, however, had significant methodological flaws. It used no formal test of cognitive function. When these patients were assessed following discharge from hospital there was no difference in cognitive outcome between regional and general anesthesia.

The inability to show a statistical difference in studies comparing GA with RA may be accounted for by the use of a significant amount of sedation in association with the latter technique [56]. It is possible that many of these studies to date do not really compare RA and GA. Rather they compare GA with RA supplemented with some degree of deep sedation.

There is some evidence to indicate that postoperative cognitive changes in elderly patients receiving RA with sedation are self-limited and patients recover while still in the early postoperative period. However, no significant difference was reported in the incidence of cognitive dysfunction six months after either GA or RA anaesthesia in elderly patients [58].

Data suggests that the type of anesthesia, GA or RA, does not affect the magnitude or pattern of postoperative cognitive dysfunction [42], [57]. Again, there are methodological and study-design issues present in many of these studies such as the use of a small sample size [58], limitation to one type of surgery (i.e., fractured hips) [59], and lack of a control group who did not undergo surgery.

Exposure to anesthetic agents is one possible cause of POCD in elderly patients. Anesthetic agents affecting the release of central nervous system neurotransmitters. Therefore, they could potentially damage memory process, particularly in elderly surgical patients.

The brain is the target organ for anesthetic drugs. It has been commonly believed that the effects of such agents do not outlive their pharmacological action. In this model, the brain is restored to its previous state once the agent is removed. However, there is increasing evidence that concept is not true. It seems that long term neurological changes can follow administration of anesthetic drugs and the brain may be particularly vulnerable in the elderly. Data supporting this position is presented in this section.

Animal studies suggest that even short exposure to a volatile agent can alter proteins in the brain. This continues to be true even after the drug has been removed from the body [60]. There may be persistent cognitive impairment in aged rats, followed for weeks to months after general anesthesia [61], [62], [63].

In humans an association has been noted between cumulative deep hypnotic time, and 1-year postoperative mortality. Like in animals this suggests that intra-operative anesthetic management may affect outcomes over longer time periods than previously appreciated [64], [65]. One suggestion has been that there may be an inflammatory mechanism underlying these findings [66].

For a very long time nitrous oxide (N2O) has been widely used as an inhalational anesthetic. There is some evidence that this agent may be a non-competitive N-methyl-d-aspartate (NMDA) receptor antagonist. Similar to other NMDA antagonist, N2O has been shown to produce in clinically relevant concentrations neurotoxic effects in rat brain tissue at high concentrations [67], and in an age-dependent manner [68], [69]. This effect seems to be independent of hypoxia and occurring predominately in the cerebral cortex. Whether this neurotoxic response reported in the experimental setting may be reflected clinically as a cognitive change such as delirium is an important unanswered question. To date no increase in the incidence of delirium has been demonstrated with the use of N2O [70].

In the past, the pathogenesis of AD has also been associated with the administration of GA [71]. However, in most studies neither anesthesia nor surgeries are commonly associated with long term cognitive disturbance [72], [56], [73]. AD has also been linked to such perioperative events as hypoxia [60] and hypocapnia [74]. However, other studies found no connection with perioperative hypoxemia and hypotension [75].

The accumulation of the [beta]-amyloid protein (A[beta]), manufactured by the orderly cleavage of the amyloid precursor protein (APP) by [beta]—and [gamma]-secretase, is a central pathogenic event in AD. It has been shown using multidimensional NMR spectroscopy that volatile anesthetics specifically interacts with Abeta40 and Abeta42 peptide. Halothane induces structural alternation of Abeta peptide from soluble monomeric alpha-helical form to oligomeric beta-sheet conformation, which may hasten the onset of Alzheimer’s disease. Abeta42 is more prone to oligomerization compared to Abeta40 in the presence of halothane [76].

AD is also associated with apoptosis [77]. It has been demonstrated that clinically relevant concentrations of isoflurane can induce apoptosis, alter APP processing and increase A[beta] levels in human H4 neuroglioma cells. Therefore, in addition to inducing apoptosis, isoflurane affects biosynthesis of a protein implicated in the pathogenesis of AD [78].

Is there a genetic basis for POCD? The apolipoprotein e4 allele is associated with poorer cognitive and neurological outcomes after brain injury and stroke, and to be a risk factor for the development of AD [79]. However, to date this gene has not been associated with the incidence of POCD [80].

Isoflurane has been shown to alter gene expression in the amygdala. The limbic system, especially the amygdala, has long been considered to be directly implicated in anxiety and this effect may play a role in the development of POCD [81].

So far we have seen that in both the surgical and non-surgical populations cognitive changes are associated with poor clinical outcomes. Assuming that some ‘common’ mechanism links these cognitive changes with adverse events then this leads to the development of several significant questions:

  • Do MCI and POCD represent the same entity, simply differing in clinical context?

  • In the elderly; is surgery associated with increased incidence of POCD?

  • In the elderly, does the data presented support the premise that individuals with POCD are at increased risk to die?

  • Does POCD represent an early manifestation of a cognitive or software problem indicative of future hardware problems?

  • If this premise holds then, what factors are responsible for the association between POCD and the increased morbidity/mortality?

  • We have seen that GA via a so far unidentified mechanism may results in a number of neurological changes that manifest themselves as cognitive dysfunction. Is it possible that these psychological changes are manifestations of wider changes in the body that result in increased morbidity and mortality in the postoperative period?

  • Is increased mortality solely caused by age or is health in general including mental factors accountable for the greater mortality?

  • If this phenomenon is more common in the elderly? Are we doing them a disservice by administrating heavy sedation or general anesthesia?

If these positions hold true then they could serve as the basis to build a unique unitary mechanism. This would link cognitive dysfunction in elderly patients and the observed increased rate of adverse events. Such a paradigm is represented by Diagram 1

The paradigm being developed here is based on the observation that the nervous system in the elderly undergoes a number of structural and functional changes. These alterations have functional implications, making the occurrence of decline in cognitive function more likely. Superimposed on this we have the administration of sedative and/anesthesia drugs. In this susceptible elderly individual these medications result in a central neurological change. This has two manifestations. In the short term it gives rise to altered cognitive function. However, in the longer term this central disorder has peripheral physiological manifestations. This would account for the observed increase in morbidity and mortality seen in those patients with cognitive disorders.

This construct, if it holds would predict that the outcome for patients undergoing a given procedure using RA (with no sedation) (Diagram 2) would be superior to an individual undergoing the same procedure utilizing GA (Diagram 3).

Based on the assumptions developed here it is possible to draw an analogy between POCD and an old automobile. In both the car as well as the human, changes can be very subtle. Consider what would happen if instead of putting oil in the car one fills it up with water. Prior to the development of mechanical problems the engine can emit an unusual noise, such as a slight knocking sound. This ‘noise’ may become louder and louder. The greater the ‘noise’ in the car the more likely a serious mechanical problem will occur and after a variable period the car may break down. In the elderly patient who receives anesthesia for surgery. Initially they could develop POCD slight cognitive changes. This could be followed after several months by an either a disease state or death. This ‘noise’ may become chronic just becoming louder and louder, this represents the transition of MCI to more significant cognitive problems such as dementia. Occasionally the problem may resolve.

In order to test the above hypothesis we suggest a three stage strategy:

Charts should be reviewed for all patients who underwent non-cardiac surgery. Patients would be segmented into RA and GA groups. Data including prevalence and incidence of postoperative complications including POCD and mortality would be compared within the same age groups for patients that did not have any surgery and those patients who had surgery using either GA or RA. This approach would help identify if there was a real relationship between POCD and outcome.

This would incorporate a very simple experimental design. Two groups of animals would be used. The first would be exposed for a prolonged period to a general anesthetic and the second would be a control. Both groups would be exposed to learning tasks using a maze. Differences between the groups would be noted. Animals would be euthanized and histological specimens of cerebral cortex to look at structural and biochemical differences in the brain between the groups (i.e., nucleosomal DNA fragmentation is characteristic of apoptotic nuclei. Necrotic cell death as contrasted to apoptosis can be detected by looking at LDH release. Mitochondrial activity that reflects cell viability can be measured using a dye test). The idea of this stage would be to identify if cognitive changes occurred in the animals and if there was some biological basis for this model.

This would be the ‘gold standard’. The idea of the initial two stages is to possibly indicate a mechanism underlying the problem, identify the severity of the problem, the incidence of post anesthesia cognitive deterioration and mortality rates. Data generated would allow calculation of the study population size. This large long term study will follow three matched groups of patients; those having non-cardiac surgery using GA, those having non-cardiac surgery using RA, and a control non-intervention group not having any surgery. All groups will be followed for cognitive function as well as morbidity and mortality data.

Section snippets

Conclusion

This paper looked at the phenomenon of cognitive decline observed in the elderly population after anesthesia. It attempted to construct a model that explains the increased mortality and morbidity that has been suggested to occur as a result of being exposed to surgery under anesthesia. A possible mechanism based on the long term alteration of neurotransmitters following GA but less following RA was suggested. A methodology to develop and test this theoretical construct was proposed.

References (83)

  • J.M. Leung et al.

    Nitrous oxide does not change the incidence of postoperative delirium or cognitive decline in elderly surgical patients

    Br J Anaesthesiol

    (2006)
  • J.T. Moller et al.

    Long-term postoperative dysfunction in the elderly: ISPOCD 1 study

    Lancet

    (1998)
  • A.K. Raina et al.

    Apoptotic promoters and inhibitors in Alzheimer’s disease: Who wins out?

    Prog Neuropsychopharmacol Biol Psychiatry

    (2003)
  • G.M. Teasdale et al.

    Association of apolipoprotein E polymorphism with outcome after head injury

    Lancet

    (1997)
  • Y. Kojima et al.

    Postoperative outcome among elderly patients after general anesthesia

    Acta Anaesthesiol Scand

    (2006)
  • J.L. Djokovic et al.

    Prediction of outcome of surgery and anesthesia in patients over 80

    JAMA

    (1979)
  • D.A. Etzioni et al.

    Elderly patients in surgical workloads: a population-based analysis

    Am Surg

    (2003)
  • R.A. Audisio et al.

    Preoperative assessment of surgical risk in oncogeriatric patients

    The Oncologist

    (2005)
  • R.J. Ackermann et al.

    Surgery in nonagenarians: morbidity, mortality, and functional outcome

    J Fam Pract

    (1995)
  • J.C. Morris et al.

    The neurology of aging: normal versus pathologic change

    Geriatrics

    (1991)
  • H. Creasey et al.

    The aging human brain

    Ann Neurol

    (1985)
  • J.A. Severson

    Neurotransmitter receptors and aging

    J Am Geriatr Soc

    (1984)
  • G. Berg et al.

    Longitudinal changes in three brief assessments of SDAT

    J Am Geriatr Soc

    (1987)
  • L. Parretti et al.

    Memory impairment in the elderly. A three year follow up

    Arch Gerontol Geriatr

    (1985)
  • D.J.H. Deeg et al.

    The association between change in cognitive function and longevity in Dutch elderly

    Am J Epidemiol

    (1990)
  • P. Maruff et al.

    Subtle memory decline over 12 months in mild cognitive impairment

    Dementia Geriatric Cognitive Disorders

    (2004)
  • P. Alexopoulos et al.

    Do all patients with mild cognitive impairment progress to dementia?

    J Am Geriatr Soc

    (2006)
  • A. Kluger et al.

    Patterns of motor impairment in normal aging, mild cognitive decline, and early Alzheimer’s disease

    Gerontol B Psychol Sci Soc Sci

    (1997)
  • B. Winblad et al.

    Mild cognitive impairment beyond controversies, towards a consensus: Report of the International Working Group on mild cognitive impairment

    J Intern Med

    (2004)
  • R. Petersen et al.

    Mild cognitive impairment as a clinical entity and treatment target

    Arch Neurol

    (2005)
  • J.S. Hui et al.

    Rate of cognitive decline and mortality in Alzheimer’s disease

    Neurology

    (2003)
  • R.C. Petersen

    Conceptual issues

  • S. Larrieu et al.

    Incidence and outcome of mild cognitive impairment in a population-based prospective cohort

    Neurology

    (2002)
  • K. Ritchie et al.

    Classification criteria for mild cognitive impairment: a population-based validation study

    Neurology

    (2001)
  • R.C. Roy

    Anesthetic implications of the rectangular survival curve

  • J.C. Morris et al.

    Cognitive impairment represents early stage Alzheimer disease

    Arch Neurol

    (2001)
  • R.S. Tilvis et al.

    Predictors of cognitive decline and mortality of aged people over a 10-year period

    J Gerontol A Biol Sci Med Sci

    (2004)
  • E.W. Gregg et al.

    Is diabetes associated with cognitive impairment and cognitive decline among older women?

    Arch Intern Med

    (1999)
  • D.A. Bennett

    Diabetes and change in cognitive function

    Arch Intern Med

    (2000)
  • R.J. Glynn et al.

    Current and remote blood pressure and cognitive decline

    JAMA

    (1999)
  • T. Ostbye et al.

    Mortality in elderly Canadians with and without dementia: a 5-year follow-up

    Neurology

    (1999)
  • Cited by (0)

    1

    Dr. Paniagua is supported in part by funds from the Division of State, Community, and Public Health, Bureau of Health Professions (BHPr), Health Resources and Services Administration (HRSA), Department of Health and Human Services (DHHS), under 1 K01 HP 00118-01 and Geriatrics Academic Career Award (GACA), from The Department of Health and Human Services, Health Resources and Services Administration. The information or content and conclusions are those of the author and should not be construed as the official position or policy of, nor should any endorsements be inferred by the Bureau of Health Professions, HRSA, DHHS or the U.S. Government.

    View full text