The contribution of kidney disease to cognitive impairment in patients with type 2 diabetes

Abstract

Purpose of review:

This review focuses on the relationships between diabetes, cognitive impairment and the contribution of kidney disease.

Recent findings:

We review the independent contributions of parameters of kidney disease, including albuminuria, glomerular filtration, bone/mineral metabolism and vitamin D synthesis, on cognitive performance in patients with diabetes. Potential pathophysiologic mechanisms underlying these associations are discussed highlighting gaps in existing knowledge. Finally, effects of the dialysis procedure on the brain and cognitive performance are considered. Emphasis is placed on novel non-invasive screening tools with the potential to preserve cerebral perfusion during hemodialysis and limit cognitive decline in patients with diabetic ESKD.

Summary:

Patients with type 2 diabetes and advanced chronic kidney disease suffer a higher prevalence of cognitive impairment. This is particularly true in patients with diabetes and end-stage kidney disease (ESKD).

Introduction

Type 2 diabetes (T2D) and chronic kidney disease (CKD) are independent risk factors for cognitive dysfunction. Approximately 43.5% of patients with T2D develop clinically significant diabetic kidney disease (DKD), manifesting excessive proteinuria and/or reduced kidney function (low estimated glomerular filtration rate [eGFR]).(1) These phenotypes are distinct processes; some patients present with heavy albuminuria prior to loss of kidney function while others progress to advanced nephropathy with minimal proteinuria. Low-level albuminuria, termed micro-albuminuria, is common and levels typically vary over time. The majority of patients with microalbuminuria either stabilize or go into remission; they do not progress to macro-albuminuria (overt DKD). Microalbuminuria is a harbinger of adverse cardiovascular disease (CVD) events, including myocardial infarction and stroke. It is considered a marker of generalized endothelial dysfunction, not kidney disease per se. The excess mortality in patients with T2D is primarily related to the presence of frank nephropathy (overt proteinuria or reduced kidney function).(2) In order to improve outcomes in patients with DKD, it remains critical to consider the effects of hyperglycemia and the diabetic milieu on the cerebral vasculature, cognitive performance and risk of stroke.

Incidence rates of DKD and proteinuria differ markedly between individuals with European, Hispanic, African and Asian ancestry. It is uncertain whether this reflects differences in environmental factors (diet, access to medical care, achievement of optimal glycemic control, etc.) or whether biologic (inherited) factors play a role. Few studies have focused on effects of DKD on the brain. Data from the Third National Health and Nutrition Examination Survey (NHANES III) revealed that among persons 20–59 years old in the general population, moderately reduced eGFR (30–59 ml/min/1.73m²) was associated with poorer performance in visual attention, learning and concentration.(3) The effect of confounders such as age, gender, race and diabetes were considered in this large population-based study. This article reviews relevant studies that have assessed risk for altered cognitive performance with presence of kidney disease phenotypes in patients with T2D.

Effects of subclinical kidney disease on the brain

Several studies have assessed effects of T2D on the brain; many did not consider presence or severity of nephropathy or proteinuria. Subtle changes in executive function are often present in patients with T2D; these can progress to mild cognitive impairment and dementia. It is unclear whether effects predominantly relate to Alzheimer’s Disease-like pathology, cerebral small vessel disease (cSVD) from longstanding poorly controlled blood sugar, or their combination.

The “Chicago Health and Aging Project” (CHAP) reported cognitive function declined at equal rates in African Americans and European Americans with incident T2D; however, African Americans with prevalent diabetes had faster rates of decline.(4) The “Atherosclerosis Risk In Communities” (ARIC) study also found that processing speed and verbal fluency declined faster in African Americans with T2D, compared to European Americans; however, ARIC results may have been confounded by longer T2D durations in African Americans.(5,6) Neither CHAP or ARIC considered data on eGFR or proteinuria; hence, the longer T2D duration and higher risk for DKD in African Americans could have impacted results.

Other reports show similar effects of T2D on cognitive decline in African Americans and European Americans.(7,8) A recent “Diabetes Heart Study Memory IN Diabetes” (DHS MIND) report in 520 African Americans (termed AA-DHS MIND) and 684 European Americans with T2D from DHS MIND who had cerebral magnetic resonance imaging (MRI) and cognitive testing found that poorer quality of education and greater cSVD, detected based on greater white matter lesion (WML) volumes, accounted for much of the difference in cognitive performance between populations.(9) Therefore, we believe the bulk of current evidence supports similar effects of T2D on the brain in African Americans and European Americans given similar qualities of education and equivalent durations (and control) of hyperglycemia, hypertension, hyperlipidemia and obesity. Controversy persists as to whether African Americans with T2D face higher risk for DKD than European Americans. If true, differential susceptibility to DKD could impact the effects of T2D on the brain. Apolipoprotein L1 gene (APOL1) kidney risk variants have adverse effects on nephropathy progression rates in DKD and non-diabetic CKD; these are limited to populations with recent African ancestry.(10) However, the randomized Action to Control Cardiovascular Risk in Diabetes (ACCORD) study reported lower rates of incident DKD in African Americans with T2D than European Americans, without a difference in rate of eGFR decline, proteinuria or kidney failure.(11)

Independent from eGFR, albuminuria (including microalbuminuria) was consistently associated with impaired cognitive performance in DKD from the “Cardiovascular Health Cognition Study”, “Ongoing Telmisartan Alone and in Combination With Ramipril Global End Point Trial” (ONTARGET), “Telmisartan Randomized Assessment Study in ACE-intolerant Subjects With Cardiovascular Disease” (TRANSCEND) Study and housebound elderly individuals receiving home healthcare.(12–14) Results from ACCORD further reveal that cognitive decline in patients with T2D at high risk for CVD is significantly associated with urine albumin to creatinine ratio (UACR), rising albuminuria and higher cystatin C, a measure of kidney function.(15,16)

In contrast to the aforementioned large clinical trials and population-based studies, smaller numbers of patients with T2D and advanced kidney disease have had measures of cognitive performance and brain imaging. A report in 757 Chronic Renal Insufficiency Cohort (CRIC) participants, 50.6% with diabetes and more than 83% with an eGFR <60 ml/min/1.73m², revealed higher levels of high sensitivity C-reactive protein, fibrinogen and interleukin 1β associated with increased risk of impaired attention.(17) In contrast, higher levels of TNF-α were associated with lower risk of impaired executive function. The authors concluded that serum inflammatory markers (e.g., systemic inflammation) likely underlie the complex relationship between CKD and cognitive performance. A subsequent CRIC report revealed that anemia associated with CKD did not relate to progressive cognitive decline.(18)

Sonoda M et al. performed a cross-sectional analysis in 143 patients with T2D lacking dementia or prior stroke; participants underwent cerebral MRI, modified mini-mental state exam (3MS) and an eGFR.(19) In multiple regression models adjusting for potential confounders, eGFR was positively associated with performance on 3MS, inversely associated with plasma homocysteine concentration and did not associate with measures of cSVD. Association of eGFR with 3MS remained significant when each measure of cSVD was included in the model; however, the relationship disappeared after homocysteine was included in place of cSVD. Higher plasma homocysteine concentration is a marker for elevated risk of vascular disease (coronary and carotid artery) and is associated with declining cognitive function and dementia in populations without diabetes. A mediation analysis revealed nearly significant (p=0.06) mediation of homocysteine on cognitive function, whereas cSVD had no meaningful effect. The authors concluded plasma homocysteine may be a potential mediator of the DKD relationship with cognitive performance. Given the cross-sectional design and small sample, replication of this finding will be necessary. It was noteworthy that kidney transplant recipients in the “Folic Acid for Vascular Outcome Reduction in Transplantation” (FAVORIT) study having high homocysteine levels and randomized to B vitamin treatment had improvement in homocysteine levels, processing speed and memory.(20) Not all FAVORIT participants had T2D, but results could have clinical importance when considered with results of Sonoda et al.

The “Action for Health in Diabetes Trial” (Look-AHEAD) prospectively studied whether intensive lifestyle intervention (ILI) defined as weight loss focused on healthy eating and increased physical activity, compared to diabetes support and education, could reduce rates of cognitive impairment. Not only was there no effect of ILI on decline in cognitive performance, subgroups with higher body mass index (≥40) and CVD randomized to ILI tended to have poorer cognitive performance over time.(21) Kidney function and proteinuria were not assessed. We are not aware of studies assessing lifestyle intervention for effects on cognitive decline in patients with DKD.

The community-based DHS and AA-DHS MIND studies provide some insight into differences in the association of eGFR, albuminuria and vascular disease with cognitive function and cerebral anatomy between African Americans and European Americans living with T2D. Participants in these studies underwent intensive phenotyping for kidney disease, subclinical CVD, cognitive performance and cerebral anatomy using brain MRI. In AA-DHS MIND, participants had a mean (SD) age of 60.2 (9.7) years and 62.7% were female. Mean diabetes duration was 14.3 (8.9) years, HbA1c 8.2% (2.2%), eGFR 86.0 (23.2) mL/min/1.73 m², and UACR 155.8 mg/g (542.1 mg/g; median 8.1 mg/g). Those with an eGFR <60 mL/min/1.73 m² or UACR >30 mg/g had smaller gray matter (GM) and higher WML volumes. Higher UACR was significantly associated with higher WML volume and brain atrophy, and smaller GM and hippocampal white matter (WM) volumes. In contrast, higher eGFR associated with larger hippocampal WM volumes. Consistent with higher WML volumes, AA-DHS participants with reduced eGFR or UACR >30 mg/g had significantly poorer processing speed and working memory. These findings were independent from glycemic control.(22) Additional analyses revealed that higher UACR was associated with poorer performance on the 3MS (p=0.014), Montreal Cognitive Assessment (MoCA; p=0.0089), digit symbol coding (p=0.0004), Stroop performance time (p=0.003), Stroop errors (p=0.032), and Stroop interference (p=0.026). Higher eGFR was associated with better performance on digit symbol coding (p=0.0071). Even in those with relatively preserved kidney function, albuminuria and eGFR were significantly associated with cognitive performance in African Americans with T2D. Subclinical atherosclerosis was inversely associated with GM volume in this population and kidney risk variants in the APOL1 gene were associated with reduced cerebral WML volume and increased GM volume.(23, 24) In contrast, brain structure and cognitive performance were less impacted by albuminuria or eGFR in European Americans with T2D from DHS MIND.(25) This may suggest differential effects of albuminuria and eGFR on the brain in individuals with T2D based upon ancestry. Additional studies will be necessary to clarify this question. A meta-analysis in 584 African Americans with T2D from the AA-DHS MIND and ACCORD MIND studies who had a cerebral MRI and cognitive testing assessed relationships between brain anatomy and cognitive performance.(26) Smaller GM and increased WML volumes were associated with poorer performance on tests of global cognitive and executive function. This suggests that volume of WML and GM atrophy associate with cognitive performance, independent from diabetes-related factors in African Americans with T2D.

Metabolomics, diabetes and cognitive decline

Metabolomics is increasingly used to develop novel diagnostic biomarkers for prediction, detection and treatment of T2D, CKD and cognitive decline. In the setting of T2D, plasma and serum metabolomics have implicated elevated branched chain amino acid (BCAA) levels in insulin resistance (27–29) and prediction of overt disease.(30) These findings were supported by transcriptomic analysis in skeletal muscle where reduced expression of genes involved in BCAA metabolism was observed.(31) In CKD, Chen et al. used untargeted metabolomics in a cohort of 2155 participants, inclusive of patients with CKD, to identify 5-methoxytryptophan (5-MTP), an endogenous tryptophan metabolite with anti-inflammatory activity.(32) 5-MTP levels strongly correlate with clinical markers of kidney disease and inversely correlated with CKD progression. In addition to the plasma and serum metabolome, metabolomics studies in CKD have also included urine. Sharma et al. profiled patients with T2D and identified 12 of 13 associated metabolites involved in mitochondrial metabolism.(33) Towards an improved understanding of cognitive decline, Mapstone et al. used blood-based metabolomics and validated a set of ten lipids that predict conversion to mild cognitive impairment or Alzheimer’s disease with over 90% accuracy.(34) Large-scale population-based studies in serum have also implicated several metabolites based on validated cognitive tests and development of overt disease (i.e., dementia). In the ARIC cohort, docosapentaenoate (DPA, 22:5 n-6; long-chain omega-6 polyunsaturated fatty acid found in vegetable oil) was associated with processing speed (Digit Symbol Substitution Test, DSST). In addition, N-acetyl-1-methylhistidine, a histidine metabolite significantly associated with genetic variation in a CKD associated gene N-Acetyltransferase 8 gene (NAT8), was significantly associated with verbal memory (Delayed Word Recall Test, DWRT) and 4-androstene-3beta, 17 beta-diol disulfate 1, involved in androgen steroid metabolism, remained associated with dementia after adjustment for diabetes status.(35, 36) Future metabolomics studies aimed at investigating the shared and unique genomic architecture among these related phenotypes in relevant tissues are warranted. Table 1 summarizes the contributions of non-dialysis-dependent CKD to cognitive impairment in patients with T2D.

Table 1.

Contributions of non-dialysis kidney disease to cognitive impairment in patients with type 2 diabetes

	Summary	References
Albuminuria	Independent of eGFR, albuminuria and urine albumin to creatinine ratio were associated with impaired cognitive performance in patients with diabetes and kidney disease.	Barzilay JI, Gao P, O’Donnell M, Mann JF, Anderson C, Fagard R, et al. Albuminuria and decline in cognitive function: The ONTARGET/TRANSCEND studies. Arch Intern Med. 2011; 171(2): 142–150. Murray AM, Barzilay JI, Lovato JF, Williamson JD, Miller ME, Marcovina S, et al. Biomarkers of renal function and cognitive impairment in patients with diabetes. Diabetes Care. 2011; 34(8): 1827–1832. Weiner DE, Bartolomei K, Scott T, Price LL, Griffith JL, Rosenberg I, et al. Albuminuria, cognitive functioning, and white matter hyperintensities in homebound elders. Am J Kidney Dis. 2009; 53(3): 438–447.
Systemic inflammation and metabolomic markers	Higher levels of serum inflammatory markers (high sensitivity C-reactive protein, fibrinogen, and interleukin 1β) were associated with increased risk of impaired executive function in patients with diabetes and kidney disease. Metabolomics studies identified 5-MTP, NAT8, and other novel diagnostic biomarkers associated with cognitive decline in DKD.	Kurella Tamura M, Tam K, Vittinghoff E, Raj D, Sozio SM, Rosas SE, et al. Inflammatory markers and risk for cognitive decline in chronic kidney disease: The CRIC Study. Kidney Int Rep. 2017; 2(2): 192–200. Sonoda M, Shoji T, Kuwamura Y, Okute Y, Naganuma T, Shima H, et al. Plasma homocysteine and cerebral small vessel disease as possible mediators between kidney and cognitive functions in patients with diabetes mellitus. Sci Rep. 2017; 7(1): 4382. Bressler J, Yu B, Mosley TH, Knopman DS, Gottesman RF, Alonso A, et al. Metabolomics and cognition in African American adults in midlife: the atherosclerosis risk in communities study. Transl Psychiatry. 2017; 7(7): e1173.
Cerebral structural change and endothelial dysfunction	Patients with diabetes and reduced eGFR or elevated urine albumin to creatinine ratio had lower cerebral gray matter and higher WML volumes, which were associated with significantly poorer processing speed and working memory.	Sink KM, Divers J, Whitlow CT, Palmer ND, Smith SC, Xu J, et al. Cerebral structural changes in diabetic kidney disease: African American-diabetes heart study MIND. Diabetes Care. 2015; 38(2): 206–212. Murea M, Hsu FC, Cox AJ, Hugenschmidt CE, Xu J, Adams JN, et al. Structural and functional assessment of the brain in European Americans with mild-to-moderate kidney disease: Diabetes heart study – MIND. Nephrol Dial Transplant. 2015; 30(8): 1322–1329. Mogi M, Horiuchi M, Neurovascular coupling in cognitive impairment associated with diabetes mellitus. Circ J. 2011; 75(5): 1042–1048. Bogush M, Heldt NA, Persidsky Y. Blood brain barrier injury in diabetes: Unrecognized effects on brain and cognition. J Neuroimmune Pharmacol. 2017; 12(4): 593–601. Freedman BI, Divers J, Whitlow CT, Bowden DW, Palmer ND, Smith SC, et al. Subclinical atherosclerosis is inversely associated with gray matter volume in African Americans with type 2 diabetes. Diabetes Care. 2015; 38(11): 2158–2165.
Ancestry	Differential susceptibility to DKD could impact the effects of T2D on the brain. Apolipoprotein L1 gene (APOL1) kidney risk variants were associated with reduced cerebral WML volume and increased GM volume In African Americans with type 2 diabetes, albuminuria and eGFR were significantly associated with cognitive performance even in patients with relatively preserved kidney function. In contrast, brain structure and cognitive performance were less impacted by albuminuria or eGFR in European Americans with type 2 diabetes.	Freedman BI, Gadegbeku CA, Bryan RN, Palmer ND, Hicks PJ, Ma L, et al. APOL1 renal-risk variants associate with reduced cerebral white matter lesion volume and increased gray matter volume. Kidney Int. 2016; 90(2): 440–449. Hughes TM, Sink KM, Williamson JD, Hugenschmidt CE, Wagner BC, Whitlow CT, et al. Relationships between cerebral structure and cognitive function in African Americans with type 2 diabetes. J Diabetes Complications. 2018; 32(10): 916–921.

Diabetic end-stage kidney disease, cerebral anatomy and cognitive performance

Data from the cross-sectional “Choices for Healthy Outcomes in Caring for End Stage Kidney Disease (ESKD)” revealed diabetes as a co-morbidity in 54% of patients with ESKD and diabetic ESKD as the reported cause of nephropathy in 47%.(37) T2D and ESKD independently confer an increased risk of cerebrovascular disease and cognitive impairment. Accumulation of advanced glycosylation end-products triggers vascular endothelial dysfunction, this may produce more brittle and permeable cerebral blood vessels.(38) The pro-inflammatory milieu of ESKD creates vascular endothelial dysfunction and impaired cerebral blood flow, leading to clinical and subclinical cerebrovascular disease. While there are limited data regarding the combined impact of T2D and ESKD on neuro-cognition, the impact of ESKD and dialysis on cerebrovascular disease and cognitive impairment is well studied.

Cognitive impairment is highly prevalent in patients with ESKD treated with chronic dialysis.(39, 40) Moderate to severe cognitive impairment is present in up to 70% on hemodialysis and up to 60% on peritoneal dialysis.(41–43) T2D may increase the risk of cognitive impairment in patients with ESKD. A cross-sectional analysis of the “Dialysis Outcomes and Practice Patterns Study” (DOPPS) found diabetes was independently associated with diagnosed dementia (OR 1.49, [95% CI 1.23–1.8]) among patients with ESKD.(44) Diagnosed dementia was in turn associated with increased risk of death (RR 1.48, [95% CI 1.32–1.66]) and withdrawal from dialysis (RR 2.01, [95% CI 1.57–2.57]). A multicenter cross-sectional study on cognitive function in patients receiving peritoneal dialysis found that patients with patients with ESKD and diabetes who had evidence of microvascular disease based on retinopathy had a significantly increased risk of global cognitive impairment (RR 2.09, [95% CI 1.11–3.92]), impaired executive function (RR 2.89, [95% CI 1.55–5.37]), and impaired immediate memory (RR 2.16, [95% CI 1.15–4.06]).(45) In contrast, the cross-sectional “Reasons for Geographic and Racial Differences in Stroke” (REGARDS) study found that T2D did not affect the association between severe kidney disease measured by eGFR and cognitive impairment.(46) T2D can lead to microvascular disease in the systemic and cerebral vasculature, contributing to the increased risk of cognitive impairment among patients with ESKD and diabetes. In the brain, microvascular disease leads to damaged vascular endothelial tight junctions, disrupted blood-brain barrier, and increased oxidative stress; these can produce cerebral structural disease.(47–49)

Beyond effects of T2D, ESKD exposes the cerebral vasculature to direct neuronal toxicity from uremic toxins and chronic inflammatory mediators, accelerated vascular calcification, and cerebral endothelial dysfunction. For ESKD patients receiving chronic dialysis, impaired cerebral perfusion (including both hypo-perfusion and hyper-perfusion) during dialysis is common.(50,51) This is particularly evident in patients on hemodialysis. Repeated bouts of impaired cerebral perfusion can lead to vascular damage, changes in cerebral structure and blood flow.(52, 53) These changes are associated with worsened processing speed and executive function on cognitive testing.(54, 55) Non-invasive cerebral hemodynamic monitoring by transcranial Doppler (TCD) in patients with ESKD on chronic dialysis found significant declines in intra-dialytic cerebral blood flow. Declines in cerebral blood flow were significantly associated with lower global and executive function, as well as progression of burden of cerebral white matter disease.(56) Efforts to reduce the rate of ultrafiltration and maintain cardiac and cerebral perfusion, such as weight-based ultrafiltration rates, are often employed.(57)

Mineral metabolism may be an independent contributor to cognitive decline in ESKD. Fibroblast growth factor 23 (FGF-23) acts as a negative feedback regulator for activated vitamin D and rises with declining kidney function. Elevated FGF-23 levels have been associated with adverse cardiovascular outcomes, death and cardio-embolic stroke.(58) A cross-sectional study of FGF-23 in patients on hemodialysis receiving comprehensive neurocognitive testing found higher FGF-23 levels were independently associated with lower memory scores, including after adjustment for measures of mineral metabolism.(59) A cross-sectional study of 255 patients within the “Dialysis and Cognition Study” found higher 25-hydroxyvitamin D (25[OH]D) levels were associated with better performance on tests of executive function.(60) It is unclear whether FGF-23 and 25(OH)D serve as markers of disease severity or are bystanders in the causal pathways producing cognitive decline in ESKD or whether they have specific effects in patients with diabetes.

Diabetic ESKD and risk of stroke and stroke-specific cognitive impairment

Patients with ESKD on dialysis are at higher risk of stroke; rates vary from 10 to 33 per 1,000 patient-years.(61) In addition, strokes often occur during the dialysis procedure.(62) The higher prevalence of kidney disease-related atrial fibrillation and large vessel CVD contribute significantly to these rates.(63, 64) In the population-based United States Renal Data System (USRDS) including 8,920 patients on chronic dialysis, Seliger et al. reported T2D was associated with an increased risk of incident stroke (HR 1.36, [95% 1.04–1.78]).(65) In a nationwide study of over 10,000 Taiwanese patients, Wang et al. found T2D conferred an increased risk of both ischemic stroke (HR 2.25 [95% CI 1.82–2.77]) and hemorrhagic stroke (HR 1.46, [95% CI 1.36–2.28]).(66)

Dialysis may pose a particularly high risk of stroke in patients with ESKD and diabetes.(61, 62, 67) Individuals with T2D, particularly older patients, are more likely to have autonomic neuropathy which may impair systemic and cerebral perfusion during ultrafiltration (fluid removal during dialysis).(68,69) In an analysis of 20,979 patients on hemodialysis aged >67 years, stroke rates peaked during the first 30 days after dialysis initiation, irrespective of dialysis modality.(61) Another single-center study showed that 39% of ischemic and 35% of hemorrhagic strokes occurred during or within 30 minutes of the dialysis procedure.(62) The presence of T2D was a significant risk factor for stroke in both of these studies.

T2D and kidney disease may complicate recovery after a stroke. Mechanistic studies have shown that T2D impedes white matter recovery by suppressing oligodendrogenesis and promoting a shift of microglia and macrophages toward a pro-inflammatory state.(70, 71) Though little is known regarding the specific impact of diabetes on post-stroke cognitive impairment in patients with ESKD, post-stroke patients with ESKD in general have poorer functional outcomes and higher mortality. In a large cohort of Medicare beneficiaries with ischemic stroke, patients with ESKD had an increased risk of 1-year post-stroke re-hospitalization (HR 2.04; 95% CI 1.9–2.18) and 1-year post-stroke mortality (HR 2.65; 95% CI 2.49–2.81).(72) The longitudinal “Tel Aviv Brain Acute Stroke Cohort” (TABASCO) study, which followed 431 post-stroke patients without cognitive impairment, found reduced creatinine clearance was a significant predictor for development of cognitive impairment two years after the index stroke (OR 2.01; 95% CI 1.03–3.92).(73) A subsequent TABASCO report found patients with T2D and reduced creatinine clearance had twice the risk for post-stroke cognitive decline compared with either presence of T2D or impaired creatinine clearance alone and nearly four times the risk of cognitive decline compared to patients lacking both conditions.(74) Table 2 summarizes the contributions of dialysis-dependent ESKD to cognitive impairment in patients with T2D.

Table 2.

Contributions of dialysis-dependent kidney disease to cognitive impairment in patients with type 2 diabetes

	Summary	References
Intra-dialytic impaired cerebral perfusion	In patients with ESKD receiving chronic dialysis, impaired cerebral perfusion (including hypo-perfusion and hyper-perfusion) is common during dialysis. Declines in cerebral blood flow during dialysis are significantly associated with lower global and executive function, as well as progression of burden of cerebral white matter disease	Davenport A. What are the causes of the ill effects of chronic hemodialysis? Balancing risks: blood pressure targets, intradialytic hypotension, and ischemic brain injury. Semin Dial. 2014; 27(1): 13–15. Findlay MD, Dawson J, Dickie DA, Forbes KP, McGlynn D, Quinn T, et al. Investigating the relationship between cerebral blood flow and cognitive function in hemodialysis patients. J Am Soc Nephrol. 2019; 30(1): 147–158. Kurella Tamura M, Vittinghoff E, Hsu CY, Tam K, Seliger SL, Sozio S, et al. Loss of executive function after dialysis initiation in adults with chronic kidney disease. Kidney Int. 2017; 91(4): 948–953. Murray AM, Bell EJ, Tupper DE, Davey CS, Pederson SL, Amiot EM, et al. The brain in kidney disease (BRINK) cohort study: Design and baseline cognitive function. Am J Kidney Dis. 2016; 67(4): 593–600.
Mineral metabolism	Mineral metabolism may be an independent contributor to cognitive decline in patients with ESKD. Cross-sectional studies in patients on hemodialysis found higher FGF-23 levels were independently associated with lower memory scores, while higher 25-hydroxyvitamin D levels were associated with better performance on tests of executive function.	Drew DA, Tighiouart H, Scott TM, Lou KV, Fan L, Shaffi K, et al. FGF-23 and cognitive performance in hemodialysis patients. Hemodial Int. 2014; 18(1): 78–86. Shaffi K, Tighiouart H, Scott T, Lou K, Drew D, Weiner D, et al. Low 25-hydroxyvitamin D levels and cognitive impairment in hemodialysis patients. Clin J Am Soc Nephrol. 2013; 8(6): 979–986.
Cerebral structural change	The pro-inflammatory milieu of ESKD creates vascular endothelial dysfunction and impaired cerebral blood flow, while accumulation of advanced glycosylation end-products in diabetes may lead to more brittle and permeable blood vessels. These disease-associated changes contribute directly to clinical and subclinical cerebrovascular disease In patients on chronic dialysis, type 2 diabetes is associated with an increased risk of incident stroke as well as intra-dialytic ischemic and hemorrhagic stroke.	Murray AM, Seliger S, Lakshminarayan K, Herzog CA, Solid CA. Incidence of stroke before and after dialysis initiation in older patients. J Am Soc Nephrol. 2013; 24(7): 1166–1173. Sanchez-Perales C, Vazquez E, Garcia-Cortes MJ, Borrego J, Polaina M, Gutierrez CP, et al. Ischaemic stroke in incident dialysis patients. Nephrol Dial Transplant. 2010; 25(10): 3343–3348. Lee M, Saver JL, Chang KH, Liao HW, Chang SC, Ovbiagele B. Low glomerular filtration rate and risk of stroke: meta-analysis. BMJ. 2010; 341: c4249.
Post-stroke cognitive decline	Post-stroke patients with ESKD have poorer functional outcomes and higher mortality Patients with type 2 diabetes and reduced kidney function have twice the risk for post-stroke cognitive decline compared with patients who have only type 2 diabetes or only reduced kidney function. They have nearly four times the risk of cognitive decline compared to patients who lack these conditions	El Husseini N, Fonarow GC, Smith EE, Ju C, Sheng S, Schwamm LH, et al. Association of kidney function with 30-day and 1-year post-stroke mortality and hospital readmission. Stroke. 2018; 49(12): 2896–2903. Auriel E, Kliper E, Shenhar-Tsarfaty S, Molad J, Berliner S, Shapira I, et al. Impaired renal function is associated with brain atrophy and post-stroke cognitive decline. Neurology. 2016; 86(21): 1996–2005. Ben Assayag E, Eldor R, Korczyn AD, Kliper E, Shenhar-Tsarfaty S, Tene O, et al. Type 2 diabetes mellitus and impaired renal function are associated with brain alterations and post-stroke cognitive decline. Stroke. 2017; 48(9): 2368–2374.

Effects of kidney transplantation on cognitive performance in diabetes

Kidney transplantation is the optimal treatment for ESKD. Current recipients are not only older but also more likely to have pre-transplant T2D than in prior eras.(75, 76) The impact of T2D on cognitive outcomes following kidney transplantation is an evolving area of research. In a longitudinal prospective study of cognitive performance in 90 patients with ESKD, diabetes was not a determinant of poorer long-term cognitive performance in patients that were transplanted or not transplanted.(77) However, significantly fewer patients with diabetic ESKD received a kidney transplant in this report. A cross-sectional study using the MoCA to assess cognitive function change found no association of MoCA score with diabetes nor diabetic ESKD on post-kidney transplant cognitive outcomes.(78) Results could have been influenced by a selection bias of healthier individuals with diabetic ESKD receiving a transplant and younger mean age at transplantation. In contrast, a study of over 40,000 patients greater than 55 years of age with ESKD receiving a kidney transplant found recipients with diabetes had increased risk of developing post-transplant dementia after 10-year follow-up (HR 1.64, 95% CI 1.51–1.78).(79) The association between dementia and death-censored allograft loss also differed by diabetes status. Older patients with diabetic ESKD who developed dementia after transplantation were at increased risk for death-censored allograft failure, compared to non-diabetic and diabetic ESKD patients without dementia.

Conclusions

DKD and diabetic ESKD appear to have a unique and significant impact on cognitive performance. In patients with DKD, higher albuminuria, lower eGFR and markers of systemic inflammation likely contribute to the complex relationship between nephropathy and cognitive impairment. It is unknown whether there is a unique pathology of cognitive decline among patients with DKD and diabetic ESKD, and whether the effects on the brain more closely reflect Alzheimer’s disease-like pathology or cerebral small vessel disease. Among patients with diabetic ESKD, FGF-23 and vitamin D concentrations may play independent roles in cognitive decline. Potential benefits of vitamin D supplementation or antagonism of the FGF-23 pathway in the treatment and/or prevention of cognitive decline in these populations require further study. There is limited information regarding post-stroke cognitive decline and its pathophysiology in patients with diabetes and ESKD. A patient’s ancestry may affect susceptibility to DKD, as well as effects of albuminuria and eGFR on the brain. An improved understanding of the impact of ancestry on cognitive decline in DKD could improve screening and prevention in susceptible populations. Preventive and treatment measures for cognitive decline in DKD and patients with both diabetes and ESKD are evolving. Among patients with diabetes and ESKD treated with chronic dialysis, non-invasive monitoring of cerebral hemodynamics by transcranial Doppler appears to hold great promise for screening and adaptation of the dialysis prescription to limit cerebral hypo-perfusion and associated cognitive decline.(56, 80, 81) The brain and cerebral circulation in patients with diabetes and ESKD have chronically been exposed to hyperglycemic and uremia. Reducing the rates of ultrafiltration (volume removal) in patients on hemodialysis and more frequent prescription of home hemodialysis and daily/nightly peritoneal dialysis have the potential to limit cerebral hypo- and hyper-perfusion during renal replacement therapy.