CKD is a term that encompasses a wide spectrum of renal dysfunction ranging from mild (which is usually only diagnosed by blood and urine testing) to ESRD, in which kidney function is impaired to such an extent that the patient requires renal replacement therapy in order to prevent potentially fatal complications from the retention of metabolic waste products, salt, and water. This continuum of CKD has been classified into a five-stage system based on the estimated glomerular filtration rate (eGFR) to help assess the severity of the CKD and its potential for progression [11] (Table 36.3). Overall, CKD has increased in prevalence to the point that it is now recognized as a worldwide public health problem [33]. More than 10% of the adult population currently has some degree of CKD [34], and its prevalence increases considerably with advancing age; 20 to 25% in people aged 65 to 74 to nearly 50% in those aged 75 and over [11, 23, 35].

The prevalence of CI/D, like CKD, also increases with age, affecting 10% of the population over 65 [36–38]. Interestingly, CI/D has many similar risk factors to CKD including obesity, diabetes mellitus, hypertension, and dyslipidemia [39–41], suggesting that similar mechanisms may be contributing to its development. Thus, it is not surprising that these two diseases occur concomitantly in many patients, particularly the elderly. However, the prevalence of CI/D in CKD patients far outpaces its prevalence in the non-CKD population [42, 43], suggesting that CKD itself exacerbates the risk of CI/D [10, 44]. Indeed, CKD has been shown to be an independent risk factor for the development and progression of CI/D by both cross-sectional studies and longitudinal studies [22, 24, 39, 40, 44]. Even moderate CKD increases the risk of CI/D by nearly 40% after adjusting for confounders [22, 35, 45, 46]. Moreover, the type of dementia induced by ageing alone compared to ageing plus CKD is different. Alzheimer’s Disease is the most common form of dementia in the elderly [47], whereas vascular dementia accounts for the increased prevalence of CI/D in CKD patients [48]. One explanation for this may be the greater prevalence of both cardiovascular and cerebrovascular risk factors among CKD patients, predisposing them to develop cerebrovascular diseases and consequently CI/D [32, 49, 50].

Patients with CKD of any stage have a greater risk of developing CI/D than the general population [22, 51, 52]. As might be expected, the risk of CI/D increases markedly with worsening kidney function. For every 10 ml/min/1.73m² drop in eGFR, the risk of developing CI/D increases by 11% [35, 53], and the risk of declining memory, language skills, executive functioning, and global cognition increases by 15–25% [45]. While most studies have found that CKD is associated with increased risks of CI/D, and that the more advanced the CKD the greater the risk, there are a few studies that have not confirmed this connection. Neither the 3C study [48], nor the study by Slinin et al. [54] found a consistent correlation between CKD and dementia. However, the patient cohort in these studies tended to be healthier, had a relatively low incidence of CKD at baseline, and had fewer cardiovascular risk factors, complicating direct comparisons with other studies. Overall, the bulk of the evidence indicates that CKD is associated with an increased risk for the development of CI/D and accelerated decline in cognitive function.

ESRD, the last stage of CKD, means that the kidneys are functioning below 15% of their normal function and can no longer support a person’s day-to-day life. In the US, more than 300,000 people are diagnosed with ESRD, and its prevalence is increasing worldwide [33, 55]. As previously mentioned, the rate of CI/D increases together with the severity of kidney disease, thus patients with ESRD have the highest prevalence. In ESRD, the prevalence rate of CI/D ranges between 16 and 38%. This is nearly 3 times greater than in the age-matched general population [22, 24, 25, 56, 57]. In fact, some studies that used more robust diagnostic criteria report even higher levels, up to 87% [24]. The variability in the reported prevalence may in part be due to the patient population studied, but more importantly to study design and the large variability in diagnostic testing and criteria used in the distinct studies. For instance, early studies reported moderate rates of CI/D that probably underestimated the true prevalence because they often excluded older and sicker patients from analysis, and used screening tests of limited sensitivity. Later studies tended to be more inclusive and used more thorough neuropsychological testing, thus increasing the diagnostic sensitivity [39]. Not surprisingly, they found the prevalence to be much higher. The cross-sectional analysis by Murray et al. highlights not only the high prevalence of CI/D in hemodialysis patients, but also the marked disparity between the documented history of CI/D (only 2.9%) and its presence upon more thorough neuropsychological testing; 12.7% had normal cognitive function, whereas 13.9%, 36.1%, and 37.3% had mild, moderate, and severe cognitive impairment respectively [24]. Thus overall, it is clear that the prevalence of CI/D is much higher in the CKD/ESRD population than in the general population, even after adjusting for age and other common risk factors. This implies that renal disease per se is a strong risk factor for the development of CI/D. It remains to be determined how much of this increase in risk is due to the same mechanisms that catalyze the accelerated vascular disease present during renal disease, and how much is due to other mechanisms, including those brought on by the treatment strategies used in CKD/ESRD.

The increase in thromboembolic strokes in CKD/ESRD is also related to the high frequency of dilated cardiomyopathy and arrhythmias [66, 67]. In this respect, atrial fibrillation deserves special attention, as it is the most common cardiac dysrhythmia and is associated with increased cardiovascular morbidity and mortality [68, 69]. Patients with advanced CKD and ESRD are especially prone to it because of (a) structural heart disease including left ventricular hypertrophy, coronary artery disease, and degenerative valvular disease (as a result of accelerated calcifications) [69, 70], (b) increased activity of the sympathetic nervous system and the renin angiotensin system, and (c) rapid shifts of fluid and electrolytes [68, 69, 71, 72]. Its prevalence is 10–20 times higher than in the general population; it is 7% and 13% in peritoneal dialysis patients and hemodialysis patients respectively, according to USRDS data [73]. Smaller studies using more sensitive methods suggest much higher rates [74, 75]. The Monitoring in Dialysis Study recorded patients’ heart rhythms continuously for up to 30 days, and found that 44% of patients experienced at least one episode of sustained atrial fibrillation (>6 min duration), while 85% developed some period of atrial fibrillation during the 30-day follow-up [76]. Thus, while the risks of atrial fibrillation are well recognized, its impact on CI/D in CKD/ESRD is poorly described and probably grossly under-recognized.

The first step in evaluating these patients is to elicit a good history when possible. A proper history should optimally include the family/caregivers in addition to the patient, due to the inherent nature of the cognitive defect and also since the caregivers will be the first to notice any overt changes in the patient’s cognitive symptoms. The history should include a complete physical examination looking for signs of neurodegenerative diseases such as Parkinson’s, tardive dyskinesia, and others, as they can mimic early signs of CI/D [11]. Despite the importance of the history, it is important to recognize that it lacks sensitivity and thus one must have a low threshold to perform additional screening or testing.

The initial evaluation must also assess for the presence of other conditions that can have similar signs and symptoms to those of CI/D, in particular depression and delirium. Indeed, depression is the most common neuropsychiatric disease observed in ESRD patients [13, 124]. The early signs and symptoms of depression are usually indistinguishable from those of early CI/D. It must be ruled out as a reason by performing depression screening tests such as the Beck depression inventory, the five-item Geriatric Depression Scale, or a patient health questionnaire [125, 126]. Similarly, delirium is especially common in older patients who are on multiple medications (such as elderly CKD patients), and must be ruled out by performing the confusion assessment method (a highly sensitive and specific test to detect delirium) [127, 128]. Delirium can be precipitated by a number of causes such as electrolyte disorders (hyponatremia, hypercalcemia, or hypoglycemia), medications (opioids, anti-psychotics, anti-cholinergic and anti-histaminergic drugs) infections, alcohol and drug intoxication, or withdrawal among others [129] (Table 36.5). Once delirium is suspected, the nephrologist must assess the likelihood that the metabolic and chemical abnormalities (a measure of dialysis adequacy) may be contributing to the CI/D. Note that because uremia causes encephalopathy, it mistakenly gets blamed for delirium quite frequently. The physician who is following the patient in a longitudinal manner will be able to determine this with more accuracy.

After eliminating the reversible causes of cognitive impairment, the patient with suspected CI/D undergoes general screening to detect the presence and type of cognitive dysfunction. Since these patients are followed frequently in CKD/dialysis units, ideally they should undergo these screening tests during their clinic visit or just prior to the dialysis session. However, this may be difficult to accomplish in most renal clinics and dialysis units due to logistical reasons, so early referral to specialized services for evaluation using neuropsychological and neurophysiological tests is warranted.

When screening in CKD clinics and hemodialysis is possible, it normally starts with generalized neuropsychological tests. The ones most commonly used in renal clinics/hemodialysis units include the MMSE, the 3MS, Cognitive Capacity Screening Exam (CCSE) and the Kidney Disease Quality of Life Cognitive Function (KDQOL-CF) scale [11, 130, 131]. Their main drawbacks are that they have suboptimal sensitivity and specificity, and limited efficacy in detecting abnormalities in executive function, a very common finding in vascular dementia and ESRD [39, 131, 132]. Moreover, none of these screening tests (other than the KDQOL-CF) have been validated by clinical trials in CKD/ESRD patients with CI/D [11, 39, 131, 132]. It is also important to point out that CKD/ESRD clinics have a large variability of ethnic groups and tend to have a high proportion of patients with limited education, which further limits the effectiveness of the tests, as some rely on the patient’s verbal or mathematical skills [133–135]. The results may also vary depending on when they are administered. For example, they are commonly run on ESRD patients during dialysis (for logistical reasons), yet this is the worst time to run them since cognitive function is at its lowest. One study which looked at the cognitive function 1, 24 and 67 h post-dialysis session in patients on hemodialysis versus peritoneal dialysis found that the hemodialysis patients fared worse after 67 h compared to 1 and 24 h, whereas the peritoneal dialysis patients fared similarly throughout the testing period [136]. Despite these drawbacks, they are valuable in identifying patients that may benefit from referral to a specialist. In our view, any CKD/ESRD patient with abnormal screening or in which the nephrologist has a high suspicion that CI/D may be present, should be referred to a specialist (Fig. 36.2).

Referral to a specialist will trigger a more comprehensive assessment, including in-depth testing to detect specific abnormalities in orientation, memory, intelligence, attention span, depression, and verbal skills (Fig. 36.2). These may include the Wechsler Memory Scale — Third Edition (WAIS-III) to assess personal, temporal, and spatial orientation, estimated verbal IQ test (NAART) to assess intelligence, Simple reaction time (SRT) to assess attention and vigilance, Boston naming test for verbal skills, and Wechsler Memory Scale — Third Edition (WMS-III) for memory [77, 121]. These tests are more sensitive and can thus detect mild to severe cognitive impairment [77, 137]. The specialist can additionally determine whether the patient merits neurophysiological techniques such as EEG and event-related potentials, or specialized imaging.

In general, imaging studies are most useful in determining potential etiologic factors that may need to be addressed. For instance, they can confirm the presence of carotid artery stenosis, or suggest the presence of paroxysmal atrial fibrillation because of embolic strokes. Imaging can also not only uncover the presence of white matter lesions, micro-bleeds, lacunar infarcts, and cerebral atrophy, but can also sometimes surprise us with findings consistent with cerebral edema from dialysis disequilibrium syndrome, osmotic demyelination syndrome, posterior reversible encephalopathy syndrome, infections, and sinus vein thrombosis among others [138–140].

As in non-CKD/ESRD patients, neuroimaging is not used to diagnose cognitive impairment or dementia (these are of course clinical diagnoses); rather, it assists the physician to rule out the structural and functional causes which can contribute to CI/D. Routinely used imaging techniques include computed tomography (CT) and magnetic resonance imaging (MRI) to detect structural abnormalities, and positron emission tomography (PET) to exclude functional abnormalities. The American Academy of Neurology recommends using CT or MRI imaging to identify structural brain lesions leading to cognitive impairment [141]. CT is usually obtained first in patients presenting with cognitive symptoms because it is widely and quickly available. CT can detect lacunar strokes, periventricular white matter lesions, and infarcts, which when present in the dominant cerebral hemispheres and limbic system suggest the presence of vascular dementia [142]. Moreover, it can also help detect silent brain infarcts, intracerebral hemorrhage, micro-bleeds, and cerebral atrophy, a very common finding in CKD/ESRD patients [32, 79, 140]. While using intravenous contrast can increase the sensitivity of CT for certain etiologies, it must be used with care in patients with renal dysfunction because it may precipitate or exacerbate acute kidney injury. It should be noted that the presence of renal failure is not a complete contraindication to the use of contrast; the risk of contrast nephropathy is actually quite low if appropriate prophylactic measures are undertaken, particularly ensuring adequate hydration [143, 144].

MRI is more sensitive than CT scans due to the better tissue contrast, flexibility of the image, and also absence of ionizing radiation [145]. MRI studies in CKD/ESRD patients have revealed increased incidence of large and small vessel strokes, along with white matter lesions (risk factors for cognitive dysfunction). Some of the conventional MRI techniques used to detect structural lesions in CKD patients include T2-diffusion weighted imaging and diffusion tensor imaging [146, 147]. However, these conventional MRI techniques are limited in detecting the subtle vascular changes seen in mild to moderate cognitive impairment in some of these CKD/ESRD patients. Advances in MRI techniques, including voxel-based morphometry, have been developed which can be used to better detect these changes. Using gadolinium as a contrast agent improves the sensitivity, but it should not be used in patients with Stage 4 or 5 CKD (an eGFR of <30 ml/min) because it can instigate nephrogenic systemic fibrosis, a catastrophic complication causing systemic fibrosis [148, 149]. If gadolinium has inadvertently been administered to a patient with advanced CKD or ESRD, dialysis should be considered, as it may reduce the contrast load [150]. Various advances to the MRI technique have been developed in order to improve the imaging of the white matter (diffusion tensor imaging) and grey matter (voxel-based morphometry). These methods have been used to demonstrate that CKD/ESRD with CI/D have white matter fiber abnormalities and/or minute microstructural brain atrophy and lower grey matter volume, including in the insular gyrus when compared to healthy controls [14, 151, 152]. These changes may explain the propensity of these patients to develop CI/D [152, 153]. Other advanced techniques include arterial spin-labeled perfusion (ASL) and blood-oxygen-level-determining (BOLD) MRIs. ASL-MRI has been utilized to study the cerebral perfusion changes in hemodialysis patients and the BOLD-functional MRI (fMRI) has the potential to be used to examine the neural mechanisms linked to the pathogenesis of CI/D in ESRD patients [14, 154]. Overall, CT and MRI are very useful in detecting cerebrovascular lesions. In the absence of such lesions, the chances of a vascular etiology for the dementia are low [155, 156].

While CT and MRI reveal structural abnormalities, functional and metabolic abnormalities can be assessed using the PET imaging technique. PET imaging can detect frontal lobe hypoperfusion and hypometabolism in vascular dementia patients, thereby assisting in differentiating it from Alzheimer’s Disease, which shows a more parieto-temporal pattern of lowered cerebral perfusion [157–159]. Using PET imaging, Kanai et al. found that hemodialysis patients metabolized less oxygen in the brain and have lower regional blood flow compared to normal individuals [160]. These parameters were greatly diminished post-dialysis. These hemodynamic changes have been postulated to play a role in the cognitive dysfunction observed in these patients.

Vascular dementia in CKD is often interlinked with cerebrovascular events. Hence, medications decreasing stroke risk, such as anti-hypertensives, statins, antiplatelet agents, and anticoagulants may be beneficial in CKD-induced CI/D. In addition, smoking and diabetes should be managed as in patients without CKD. We will discuss three main risks in which there are important considerations in this patient population: hypertension, dyslipidemia, and atrial fibrillation.

Hypertension is highly prevalent in patients with CKD/ESRD, and can be refractory to therapy. While most nephrologists agree that lowering blood pressure is important to slow the progression of CKD and to reduce the risk of cardiovascular complications in CKD /ESRD patients, its effect on CI/D is less clear. Observational studies and clinical trials suggest that antihypertensive therapy reduces CI/D, but longitudinal studies have provided inconsistent results [39, 161]. Moreover, there is controversy as to: (a) what is the optimal blood pressure (cohort studies suggest that aggressive control increases the risk of worsening cognition), (b) what are the best agents for achieving this target, and (c) what is the true risk/benefit ratio for CI/D at different blood pressures [162]. Hence, there are no absolute blood pressure targets, and thus management should be individualized according to the patient’s age, severity of albuminuria, and comorbidities [162, 163]. The Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guidelines for the management of blood pressure in CKD suggests a blood pressure threshold of <140/90 mm Hg for CKD patients without proteinuria, and a lower threshold of <130/80 mm Hg in those patients with albuminuria of >30 mg/24 h [164]. The recent SPRINT trial [165] suggests that our target should actually be lower, <120/80 mm Hg in patients with high cardiovascular risk (e.g., CKD patients). However, there are no trials where the main outcome was CI/D (we anxiously await the results of the SPRINT–MIND study). Thus, particular attention must be paid to complaints that may be related to episodic hypotension and changes in neuropsychological symptoms. This is especially true in ESRD patients, who often require more precise adjustments of their antihypertensive regimen because of the additional risk of intradialytic hypotension [166].

Whether there is a preferred class of medication for CI/D in patients with CKD also remains to be determined. Due to the unique microvascular susceptibility seen in both the kidneys and the brain, some studies have suggested that the microvasculature of both organs may benefit from renin–angiotensin system (RAS) blockade, particularly in the presence of albuminuria [138]. Indeed, several studies have shown that RAS blockers prevent renal damage in the kidneys and also the occurrence or recurrence of stroke in the brain [167, 168]. Interestingly, the protective effects of ACE inhibitors on cognitive decline and dementia do not correlate well with their antihypertensive effects [169]. However it should be mentioned that despite their reported beneficial effects, ACE inhibitors might actually worsen cognitive defects, since they inhibit the conversion of Aβ₄₂ (amyloid beta peptide) to the less toxic and amyloidogenic Aβ₄₀ [39]. Angiotensin receptor blockers (ARB) do not have this potential negative effect, and thus theoretically may be a better alternative. Calcium channel blockers (CCB) have also been shown to be protective against dementia in a large European trial [170]. Moreover, the combination of a CCB with an ARB was more effective than that of a diuretic and an ARB [138, 171]. Despite these promising results, more work is needed before we can offer strong recommendations.