Integrating assessment of cognitive status in elderly cardiovascular care

Abstract Cardiovascular clinicians tend to pay little attention to issues related to cognition, and yet those caring for older adults will encounter a variety of conditions that may lead to cognitive impairment. Most commonly, these include cardiovascular disease‐specific conditions such as cerebrovascular disease or heart failure, but may also include neurodegenerative conditions, mood disorders, medication side effects and polypharmacy, and nutritional deficiencies and metabolic derangements among others. This review presents evidence supporting the importance of assessing cognitive status in older adults with cardiovascular disease, and suggests a practical approach to assessment and management of cognitive impairment in this population when it is found. Special attention is paid to the importance of collaboration between cardiovascular and geriatric specialists, and the value it may bring to patients.


| INTRODUCTION
Cardiovascular clinicians who care for older adults tend to pay little attention to issues related to cognition and cognitive impairment in this patient population. Reasons for this vary and include a lack of awareness about the high prevalence of cognitive impairment in older adults with cardiovascular disease and its association with poor outcomes, a lack of knowledge about how to assess for cognitive impairment, a perception that assessment is lengthy and cumbersome, awkwardness about telling patients they may have problems with cognition, and a lack of understanding of how to manage cognitive impairment when it is identified. This review presents evidence supporting the importance of assessing cognitive status in older adults with cardiovascular disease, and suggests a practical approach to assessment and management of cognitive impairment in this population when it is found.

| COGNITIVE DOMAINS AND DISORDERS
In general, a human being's performance is driven by cognitive and noncognitive skills or "domains." The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) defines six key domains of cognitive function-complex attention, executive function, learning and memory, language, perceptual-motor function, and social cognition-each with its own subdomains ( Figure 1). 1 The DSM-5 further categorizes cognitive disorders within three separate syndromes, each with a range of possible etiologies, including delirium, mild neurocognitive disorder, and major neurocognitive disorder (or dementia). Noncognitive domains include dexterity, situational awareness, professionalism, compassion, integrity, team work, and resilience-their discussion is beyond the scope of this review.
Mild neurocognitive disorder, a condition commonly found in patients with heart failure (HF), is defined in DSM-5 as having evidence of modest cognitive decline from a previous level of performance in Both the authors contributed equally to this article. one or more of the six cognitive domains noted above. To meet criteria for mild neurocognitive disorder, these deficits should not interfere with capacity for independence in everyday activities, and should not be explained by delirium or another mental disorder such as major depression or schizophrenia. Major neurocognitive disorder or dementia, on the other hand, manifests in more severe abnormalities of cognition, whereby the impairment interferes with activities of daily living.
The prevalence of cognitive impairment without dementia has not been clearly defined, though population estimates exist. In the Aging, Demographics, and Memory Study, a longitudinal study of older adults who were representative of the United States population, the derived estimated national prevalence rates of cognitive impairment were approximately 22% in individuals age 71 years or older (16% among ages 71-79 years, 29% among ages 80-89, and 39% among ages ≥90). 2

| COGNITIVE IMPAIRMENT IN CARDIOVASCULAR PRACTICE
Cardiovascular clinicians caring for older adults will encounter a variety of conditions that may lead to cognitive impairment. Most commonly, these include cardiovascular disease-specific conditions such as cerebrovascular disease or HF, but may also include neurodegenerative conditions (Alzheimer's disease, Lewy-Body disease, Parkinson's disease, and frontotemporal degeneration), mood disorders (depression and anxiety), medication side effects/polypharmacy, and nutritional deficiencies and metabolic derangements among others. The mechanism by which the later three medical conditions result in return to normal cognition may be delirium.

| HEART FAILURE
The prevalence of cognitive impairment in HF cohorts is estimated to be 43%, but with a high variability ranging between 30% and 55%. 5 Reasons for this variability include variation in screening tools used between studies, heterogeneity in HF cause and severity, and differences in inpatient vs outpatient venues of patient care whereby acute illness may cause or exacerbate cognitive problems. Another increasingly recognized reason for variability in estimates of cognitive impairment prevalence is that the burden of cognitive impairment increases cumulatively with HF disease course. 6,7 In healthy older adults in their 70s, cognitive decline with age is extremely slow, perceptible only on a time scale of decades. 8 On the other hand, incident HF is associated with a sudden acceleration in prevalence of cognitive impairment. In the Atherosclerosis Risk in Communities Study, cognitive decline over a 15-year period was greater in participants with HF as compared to those without, even after adjustment for other comorbid conditions. 6 These observations suggest that pathophysiological mechanisms related to HF drive cognitive impairment over time.
The pathophysiology of cognitive impairment in older adults with HF is complex, and often independent of aging. A position paper published by the Heart Failure Association of the European Society of Cardiology in 2018 proposed a novel systematic order of pathophysiological principles, feedback signals, and categories of functional impairment to describe heart and brain interactions in this population. 9 The underlying principle of this proposed system is the bidirectionality of (a) the failing heart affecting cerebral function and (b) neuronal signals impacting the myocardium. 9 The five pathophysiological categories proposed The confluence of factors which leads to cognitive impairment in patients with HF was termed by Havakuk et al, 10 the "cardiocerebral F I G U R E 3 Systematic overview of heart and brain interactions in heart failure (HF). Adapted from Reference 9 F I G U R E 2 Mechanisms underlying vascular cognitive impairment. Adapted from Reference 4 syndrome," defined as "a state of cognitive impairment of undefined cause in HF patients, beyond the one anticipated in age-matched controls, typically accompanied by anatomic brain changes." Magnetic resonance imaging (MRI) in this population typically shows white matter hyperintensities, particularly periventricular, and gray matter atrophy, particularly involving the hippocampus and frontal cortex.
Laboratory studies typically show high levels of neurohormones and inflammatory markers (IL-6, TNF-alpha, cortisol, epinephrine). Other medical and neurological conditions that can cause cognitive dysfunction are simultaneously ruled out. 10 The most common cognitive domain abnormalities among patients with HF are learning and memory, complex attention, and executive function. Examples commonly recognized by HF clinicians include a patient who repeats themselves in conversation often within the same conversation (learning and memory), a patient who cannot keep track of a short list of items (learning and memory), a patient who needs to rely on others to plan activities of daily living or make decisions related to self-care or medication management (executive function), a patient who has increased difficulty in environments with multiple stimuli including conversation (complex attention), and a patient who has difficulty holding new information in mind such as reporting what was just said (complex attention).
Cognitive impairment in older adults with HF is associated with poor outcomes. Among patients hospitalized for acute decompensated HF, cognitive impairment has been associated with increased readmission risk in multiple studies. 11  Among community-based older adults with HF, cognitive impairment has complex cross-sectional associations with physical frailty, sleep disorders, and mood disorders including depression, anxiety, and hopelessness. [14][15][16][17] Cognitive impairment has been linked to a variety of poor outcomes including worse health-related quality of life, increased spousal/caregiver distress, increased disability, worse cardiovascular outcomes at 180 days, and increased mortality risk. 18,19 A common link, which has immense practical implications for clinicians caring for older adults with HF, is that cognitive impairment in this population is associated with worse self-management, self-care, and self-confidence. 20 A high index of suspicion is warranted as patients with cognitive impairment are less likely to seek assistance. 20

| When should clinicians screen for cognitive impairment?
Little is known about the ideal settings and time points during which screening for cognitive impairment in older adults with cardiovascular disease is indicated. Based on our observations and practice, we can adversely affect health and contribute to substantial impairments during the early recovery period, an inability to fend off disease, and susceptibility to mental error." 21 Among older adults hospitalized for HF cognitive impairment is associated with post-discharge readmission and mortality risk. 12,13 For these reasons, screening for cognitive impairment in the inpatient setting is important. The ideal timing to conduct this screening is unknown, but is likely most valuable when conducted as part of discharge planning, approximately 2 to 3 days prior to the patient's transition from hospital to community. 12 In one study of cognitive impairment and postdischarge outcome in older adults with HF, there was effect modification by venue of discharge, whereby patients with cognitive impairment discharged to a nursing facility had longer time to hospital readmission or mortality as compared with those discharged home (Figure 4). 12 This suggests that for patients with cognitive impairment structured postdischarge programming offered at nursing facilities may protect from worse outcomes. Randomized clinical trials are needed to assess whether this observation is valid. Further, it is unknown whether structured home care programs specifically designed for older adults with HF and cognitive impairment can improve outcomes.

| Screening for cognitive impairment: Tools and approaches
There is no formal consensus about which tools should be used to assess cognition in older adults who have cardiovascular disease. We believe that the following three tools/approaches are of value Subjective memory complaints in older adults are important but are often missed or overlooked in most clinical encounters, both in specialist and primary care settings. [23][24][25][26] Clinicians should take note of subjective short-term memory changes-especially those reported by family members-as it has long been recognized that these observations may have a high level of specificity for pathological cognitive changes including dementia. Patients themselves may not be able to fully recollect the true extent of their short-term memory changes, or may withhold this information from their clinicians fearing loss of independence, self-esteem, or the possibility of more testing. Focusing on short-term memory changes is important as long-term memory is preserved in most dementias.
The Mini-Cog, a three-word recall and clock-drawing test, is a validated screening tool for cognitive impairment. 27 It is ultrashort and takes less than 3 minutes to complete, making it is ideal for use in busy clinical settings. 28 The Mini-Cog, and especially abnormal clock drawing, make cognitive impairment visible even if it has not yet become apparent to patients and families. While a positive Mini-Cog finding does not "rule in" dementia, it enables appropriate navigation toward referral to a specialist such as geriatricians for further cognitive evaluation. The Mini-Cog can also be easily integrated into the clinical workflow of medical assistants' vital screening and triage process, ahead of the cardiovascular clinician's arrival into the exam room.
IADLs include using the telephone, grocery shopping, cooking, housekeeping, laundry, managing medications and finances, and driving. Progressive compromise of these tasks may be related to underlying cognitive impairment, and is associated with loss of functional independence. 29 Arguably, the IADL components that are most dependent on cognitive skills are medication self-management and ability to handle finances. 30 The Lawton IADL scale 22

| Barriers to cognitive status assessment
Despite the fact that screening for cognitive impairment among older adults with cardiovascular disease is intuitive, valuable, and easily accomplished, little research exists about barriers to implementation.
In one survey of family physicians and general practitioners from Canada, only 24% of practitioners routinely screened patients for cognitive impairment, although 82% believed screening was needed. 37 The most common perceived barriers to screening were lack of time (82% of respondents; each respondent could list more than one bar-