Abnormal Cognitive Aging

Abnormal cognitive aging includes unexpected changes in cognition or memory functions. Changes can present either acutely or gradually. Within the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) by the American Psychiatric Association (APA) [1], abnormal cognitive profiles are divided into Mild Neurocognitive Disorder and Major Neurocognitive Disorder (also known as dementia).

Diagnostic criteria for Mild Neurocognitive Disorder are based on either self-reported, or objective evidence of, mild cognitive decline from a previous level of performance [1]. Cognitive changes can occur within any cognitive domain, including complex attention, executive functioning, learning and memory, language, perceptual–motor, or social cognition. Furthermore, the changes in cognition do not interfere with the capacity for independent functioning, cannot occur exclusively in the context of delirium, and cannot be explained by any other mental disorder [1]. The commonly used term mild cognitive impairment (MCI) represents a prodromal state of cognitive change prior to the development of dementia, or most traditionally Alzheimer’s disease (AD). MCI is subsumed under the DSM-5 Mild Neurocognitive Disorder diagnosis [Reference Petersen, Smith, Waring, Ivnik, Tangalos and Kokmen2].

Diagnostic criteria for Major Neurocognitive Disorder, or dementia, by contrast, involve evidence of significant decline from a previous level of performance in one or more of the previously mentioned cognitive domains. Such decline may be evident from: (1) concerns expressed by the individual or a knowledgeable informant; or (2) objective and substantial impairment in cognitive performance on neuropsychological testing that did not occur in the context of delirium [1]. Contrary to MCI, cognitive impairment must interfere with the individual’s ability to function independently. Major Neurocognitive Disorder is further organized into three severity levels, based on the stage of functional impairment: mild, moderate, and severe. Major Neurocognitive Disorder, mild, refers to subtle difficulties with instrumental activities of daily living such as housework and managing money. Major Neurocognitive Disorder, moderate, describes impairment with feeding, dressing, and other basic activities of daily living. Major Neurocognitive Disorder, severe, refers to a state of total reliance on others to complete basic and instrumental activities of daily living.

Considering Abnormal Cognitive Aging Profiles in Preoperative Settings

Although it is often assumed that the neurotoxic effects of sedative/hypnotic anesthetic agents are fully reversible, there appears to be diminished capacity for returning to their baseline in older adults [Reference Cottrell and Hartung3]. Various factors contribute to this age-related increased vulnerability, including changes in the aging brain such as increased free radical buildup (due to fewer scavenging agents), accumulation of prion-like proteins, increased brain atrophy, and diminished oxidative phosphorylation [Reference Cottrell and Hartung3]. Individuals with preoperative mild or major neurocognitive disorders/neurodegenerative disorders present with additional challenges regarding increased sensitivity to pharmaceuticals acting on the central nervous system (CNS) [Reference Burton, Nicholson and Hall4].

Understanding brain–behavioral–anesthetic interactions, particularly for those with mild/major neurocognitive disorder and neurodegenerative diseases, is especially prudent, given our current population challenges. By 2050, people aged 65 and older will reach 1.6 billion worldwide [Reference He, Goodkind and Kowal5]. Our healthcare system will face greater numbers of individuals with early- to late-stage AD [Reference Hebert, Scherr, Bienias, Bennett and Evans6] and other neurodegenerative disorders who require procedures with anesthesia due to comorbid health-related conditions (e.g., cardiac) or who desire operations for improved quality of life (e.g. joint replacement) [Reference Williams, Wolford and Bercovitz7]. Low- and high-risk operations are performed annually on more than half a million individuals aged 65 and older, and with population aging, this number will likely increase [Reference Williams, Wolford and Bercovitz7]. Meanwhile, many older adults may be relatively unaware of, or even deny, their cognitive impairments or neurodegenerative symptoms.

Preoperative markers of mild to major neurocognitive disorder predict postoperative cognitive complications, including delirium and mortality [Reference Oresanya, Lyons and Finlayson8, Reference Price, Garvan, Hizel, Lopez and Billings9]. Individuals with AD and related dementias, as well as other neurodegenerative disorders, such as Parkinson’s disease (PD), have greater risks in perioperative settings [Reference Aminoff, Christine and Friedman10]. For example, cognitive and psychiatric deficiencies associated with dementia interfere with a patient’s ability to comply with medical recommendations and navigate complex environments independently [Reference Ala, Simpson, Holland, Tabassum, Deshpande and Fifer11]. Furthermore, cholinergic system response to anesthesia [Reference Pratico, Quattrone and Lucanto12] and inflammatory responses [Reference Whittington, Planel and Terrando13] are hypothesized to be atypical for individuals with AD [Reference Baranov, Bickler and Crosby14, Reference Silbert, Evered, Scott and Maruff15] and PD [Reference Bohnen and Albin16]. Preoperative indicators of cognitive reserve and brain integrity (e.g., reduced entorhinal thickness, greater white matter disease burden, larger ventricular size) predict differences in intraoperative anesthesia EEG responses in the brain [Reference Giattino, Gardner and Sbahi17, Reference Hernaiz Alonso, Tanner and Wiggins18] and pre-/postfunctional and structural brain changes [Reference Huang, Tanner and Parvataneni19–Reference Hardcastle, Huang and Crowley21].

Older adults with mild or major neurocognitive disorder involving AD or PD profiles also present with increased anesthetic challenges due to alterations in physiology that affect both the pharmacokinetics and pharmacodynamics of drugs [Reference Price, Levy and Tanner22, Reference Roberts and Lewis23]. Interestingly, mechanisms underlying general anesthesia (GA) have reciprocally been suggested to cause neurodegenerative changes similar to those seen in AD. A review article by Vutskits et al. (2016) detailed experiments showing underlying mechanisms that suggest a role of GA exposure in both elderly and postnatal brains leading to lasting morphofunctional CNS changes [Reference Vutskits and Xie24]. For example, various in vitro and in vivo rodent experiments have elucidated mechanisms that link GA-induced apoptosis to increased generation and aggregation of amyloid β [Reference Xie, Culley and Dong25–Reference Dong, Zhang and Zhang27]. Anesthetics have also been linked to increased tau hyperphosphorylation [Reference Le Freche, Brouillette and Fernandez-Gomez28], another dominant pathologic feature of AD. This effect was exacerbated in Mapt transgenic mouse models encoding the gene for tau [Reference Whittington, Virag and Gratuze29]. These in vitro and in vivo experiments found that dexmedetomidine increased tau phosphorylation and aggregation in mice, as well as impaired spatial reference memory.

Another characteristic pathologic finding in AD is the degeneration of the cholinergic CNS. Elderly rats exposed to repeated intravenous (IV) anesthesia with pentobarbital exhibited altered cholinergic binding, relative to controls, due to reduced cerebral cortex cholinergic receptors [Reference Hanning, Blokland, Johnson and Perry30]. Postoperative elevations in AD pathology following anesthesia have also been demonstrated in humans [Reference Palotas, Reis and Bogats31]. One such experiment revealed elevated cerebrospinal fluid (CSF) levels of tau and amyloid β_1–42, a form of amyloid β consistently implicated in AD pathophysiology, 1 week after coronary artery bypass graft (CABG) surgery under GA [Reference Palotas, Reis and Bogats31]. These findings suggest that anesthesia may induce neurologic changes in the elderly, resembling those found in AD, with important implications for a patient population with increased susceptibility not only to neurocognitive diseases, but also to postoperative cognitive impairment.

Studies in individuals diagnosed with PD have shown that these individuals may, in fact, be more sensitive to gamma-aminobutyric acid (GABA)-activating agents and that this mechanism may play a role in anesthetic exacerbation of motor symptoms. In a study by Calon et al. (2002), postmortem observation of individuals with PD demonstrated increased GABA_A receptor concentration in the globus pallidus internus (GP_i) of dyskinetic PD compared to nondyskinetic individuals [Reference Calon and Di Paolo32]. There is also evidence suggesting that anesthetics may affect neuronal activity of the subthalamic nucleus (STN), as propofol was found to inhibit the STN in individuals with advanced PD under consideration for deep brain stimulation implantation [Reference Raz, Eimerl, Zaidel, Bergman and Israel33]. Because the STN has been demonstrated to play a role in inhibiting corticothalamic projections responsible for movement [Reference Browndyke, Berger and Harshbarger20], decreased output from the STN facilitates movement, as evidenced by individuals with STN lesions who exhibit ballism and choreiform movements [Reference Krauss, Borremans, Nobbe and Mundinger34]. Taken together, these studies suggest possible mechanisms underlying an increased risk of anesthesia-induced dyskinesia in PD. Furthermore, a study by Jamsen et al. (2014) found that individuals diagnosed with PD undergoing total knee or hip replacement surgery had prolonged hospitalization, increased risk of recurrent dislocation in the first postoperative year, and poor long-term prognosis, compared to their counterparts without PD who underwent the same procedures [Reference Jamsen, Puolakka, Peltola, Eskelinen and Lehto35]. Previous work has demonstrated that relative to healthy controls, individuals diagnosed with PD show a significant decline in cognitive function after anesthesia and surgery, compared to baseline functioning [Reference Price, Levy and Tanner22]. Retrospective studies and single case reports have shown that individuals with neurodegenerative disorders, such as PD, have greater rates of postoperative delirium and cognitive decline [Reference Newman, Sodhi and Dalton36].

Nevertheless, individuals with mild and major neurocognitive disorders showing prodromal to severe signs of AD and other neurodegenerative disorders receive preoperative care that is comparable to their cognitively intact counterparts, without taking into account their additional needs [Reference Silbert, Evered, Scott and Maruff15]. For these reasons, the benefit–risk ratios for procedures can be misrepresented to individuals with compromised cognition, inadequately identifying the potential for surgical anesthesia to exacerbate neurodegeneration, and thereby precluding informed decision-making [Reference Arora, Gooch and Garcia37]. Insight into a patient’s preoperative cognitive status in medical settings can identify individuals who may benefit from prerehabilitation, additional preoperative support when preparing for surgery, and increased monitoring during and after surgery [Reference Calkins38–Reference Prizer and Zimmerman40]. A discussion on preexisting cognitive vulnerabilities may help patient–caregiver dyads make informed healthcare decisions [Reference Arias, Bursian, Sappenfield and Price41].

For many of these reasons, there has been a call to action for preoperative identification of mild and moderate neurocognitive disorders, awareness of AD and related dementias, and improved understanding of brain mechanisms of change to facilitate options for intervention [Reference Crosby, Culley and Hyman42]. Anesthesiologists who are knowledgeable in cognitive–behavioral profiles can therefore improve patient outcomes, appreciating how individuals with predisposing cognitive impairment require fewer precipitating factors (such as anticholinergics, anesthesia, pain, infection) for delirium onset [Reference Marcantonio43]. For theoretical discussions on this topic, a review published in 1993 on the “threshold theory” summarizes how individuals remain at a critical threshold until various “stressors” overwhelm their remaining cognitive reserve capacity and accelerate symptom manifestation of disease [Reference Satz44].

The effects of anesthesia on older adults and the pharmacologic vulnerabilities of the aging brain are also topics that are increasingly surfacing within the anesthesiology community, including the American Board of Anesthesiology (ABA) Initial Certification in Anesthesiology. Trainees are expected to demonstrate knowledge of geriatric anesthesia, including pre- and postjunctional mechanisms of the cholinergic system, the neurotransmitter being heavily implicated in AD pathology. There are also sections addressing neurologic consequences of anesthesia, including confusion, delirium, cognitive dysfunction, and failure to emerge from anesthesia.

Identifying Characteristics of Preoperative Cognitive Disorders

Anesthesiologists can identify characteristics of preoperative cognitive disorders, even if a disorder is not formally diagnosed on the medical chart. The Society for Perioperative Assessment and Quality Improvement (SPAQI), in collaboration with experts in preoperative neuropsychology, anesthesiology, and geriatric medicine, drafted a two-part statement to detail the rationale for evaluating cognition preoperatively, and provided a succinct list of instruments designed to screen cognitive functioning in fast-paced settings [Reference Arias, Wiggins and Urman45, Reference Wiggins, Arias and Urman46]. Part I introduces the most common neurodegenerative disorders affecting older adults and lists the common signs and symptoms associated with each of those disorders. Part II of this statement: (1) describes factors that must be considered when selecting cognitive screening tools; (2) summarizes a review of the literature on existing cognitive screening tools and lists available cognitive screening tools with demonstrated utility within primary care and preoperative settings; and (3) provides a workflow diagram to assist clinicians with decision-making [Reference Arias, Wiggins and Urman45].

Cognitive screening tests assess selected neurocognitive domains and are interpreted using prespecified cutoff scores [Reference Block, Johnson-Greene, Pliskin and Boake47]. While administration of a cognitive screening tool does not substitute for a thorough neuropsychological assessment, this approach is highly desirable in clinical environments with high throughput. A handful of cognitive screening instruments have been used in preoperative settings. The average administration time for cognitive screening tools previously used in preoperative settings is 5 minutes, and administration time ranges from 1 to 10 minutes.

Studies have shown that these screenings identify symptoms of mild to major neurocognitive disorders in 19–33% of older adults scheduled for elective surgical procedures [Reference Amini, Crowley and Hizel48, Reference Culley, Flaherty and Reddy49] and predict cognitive and noncognitive complications after surgery and anesthesia [Reference Price, Garvan, Hizel, Lopez and Billings9]. Most screenings include variants of short traditional neuropsychological/neurologic tools, that is, three-word memory, counting backwards from 100 by 7, spelling the word ‘WORLD’ backwards such as in the Mini-Cog [Reference Borson, Scanlan, Chen and Ganguli50] or the Mini-Mental State Examination [Reference Folstein, Robins and Helzer51], and clock drawing to command and copy [Reference Libon, Malamut, Swenson, Sands and Cloud52]. Routine inclusion of preoperative cognitive screening for the risk of dementia and delirium is in its infancy. We are only beginning to learn how to integrate educational and cognitive screening into busy clinical settings, and to appropriately report and implement cognitive/memory screening data for perioperative management. Very few hospitals include cognitive screening of older adults as a routine component of their preoperative evaluation. Even fewer hospitals collaborate with neuropsychologists and geriatricians for more integrated pre- and postoperative care.

The University of Florida Health System has established a cognitive screening program for adults aged 65 and older with a planned surgical procedure [Reference Wiggins, Hernaiz, Fahy, Price, Libon, Lamar, Swenson and Heilman53]. Individuals who present to a presurgical clinic and “fail” a preadmission cognitive/memory screen are immediately referred to a team of neuropsychologists and geriatric medicine physicians [Reference Arias, Bursian, Sappenfield and Price41, Reference Block, Johnson-Greene, Pliskin and Boake47]. The purpose of this follow-up assessment is to identify modifiable risk factors for negative postoperative outcomes, including anticholinergic burden or polypharmacy, to provide the anesthesia–surgical team with potential considerations regarding the patient’s cognitive impairment, and to provide recommendations for inpatient geriatric medicine follow-up and closer inpatient and home-based delirium monitoring for both the patient and the care team. Individuals at risk of negative postoperative cognitive outcomes are then educated (along with their families when available) on perioperative risks, thus improving the quality of patient-centered care (e.g., see [Reference Arias, Bursian, Sappenfield and Price41] and [Reference Hamlet, Pasternak, Rabai, Mufti, Hernaiz Alonso and Price54]). Patient reports are available to the anesthesiology and surgery team prior to the surgical procedure, so that anesthesiologists and surgeons can tailor their care based on the patient’s preoperative profile, when possible. For individuals deemed to be at elevated risk and who are inpatient following surgery, perioperative follow-up teams monitor for delirium and inform the patients and families about the possibility of longer hospitalization and discharge to a nursing facility, rather than home.

This form of integrated care provides a more patient-centered approach but does also require close collaboration between multiple medical providers from the patient’s care team, including those from anesthesiology, neuropsychology, and geriatric medicine. Prospective studies across various surgical types are needed to further address and understand the interactions and relationships between preoperative cognitive impairment and the type/depth of anesthesia, invasiveness of the surgical procedure, biomarker changes, and patient outcomes. Addressing cognition within the preoperative environment is relevant to predicting risk or documenting the baseline cognitive status.

Until the efficacy for such programs is shown across multiple hospitals, cognitive screening by anesthesiologists is a viable option. In addition, cognitive screening training workshops have been provided at various International Anesthesia Research Society meetings. We encourage a review of the resources listed in published articles, to contact the American Society of Anesthesiologists’ Brain Health Initiative, or to contact the University of Florida’s Perioperative Cognitive Anesthesia Network (PeCAN) program for more information on available screening resources.