NDT Advance Access originally published online on June 4, 2006
Nephrology Dialysis Transplantation 2006 21(9):2543-2548; doi:10.1093/ndt/gfl275
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author [2006]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
Original Articles: Dialysis and Transplantation
Correlates and outcomes of dementia among dialysis patients: the Dialysis Outcomes and Practice Patterns Study
1 Division of Nephrology, Department of Medicine, University of California, San Francisco, San Francisco, CA and 2 University Renal Research and Education Association, Ann Arbor, MI, USA
Correspondence and offprint requests to: Manjula Kurella, MD, MPH, Department of Medicine Research University of California, San Francisco, UCSF Laurel Heights Suite 430, 3333 California Street, San Francisco, CA 94118-1211, USA. Email: manjula.kurella{at}ucsf.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Recent studies suggest a high prevalence of cognitive impairment and dementia in persons with end-stage renal disease (ESRD), yet risk factors for dementia and its prognostic significance in persons with ESRD remain unclear. The goals of this study were to determine the prevalence, correlates and dialysis-related outcomes of dementia in an international sample of haemodialysis patients.
Methods. We analysed data collected from a cohort of 16 694 patients in the Dialysis Outcomes and Practice Patterns Study. Dementia was defined as a diagnosis of dementia documented in the medical record. We used logistic regression to determine the baseline correlates of dementia and Cox proportional hazards models to determine the relative risk (RR) of death and dialysis withdrawal for patients with dementia, while adjusting for a number of confounding factors.
Results. Overall, 4% of the cohort had a recorded diagnosis of dementia. In the cross-sectional analyses, risk factors for dementia in the general population including age, black race, low educational attainment, cerebrovascular disease and diabetes, as well as modifiable uraemia-related factors, including markers of malnutrition and anaemia, were independently associated with dementia. After adjustment for a number of confounding factors, dementia was associated with an increased risk of death [RR 1.48, 95% confidence interval (CI) 1.32–1.66] and dialysis withdrawal (RR 2.01, 95% CI 1.57–2.57).
Conclusions. Dementia is associated with adverse outcomes among ESRD patients. Dialysis providers should consider instituting routine screening for cognitive impairment among elderly patients in order to identify those at risk for associated adverse outcomes.
Keywords: dementia; end-stage renal disease; epidemiology
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Dementia is a major cause of death and disability among elderly individuals in the general population [1]; and contemporary studies suggest that individuals with end-stage renal disease (ESRD) have a 2–7-fold higher prevalence of cognitive impairment and dementia compared with the general population [2–4]. However, unlike other comorbid conditions recorded in the Centers for Medicare and Medicaid Services Medical Evidence Form (the ‘2728’), no previous national or international efforts have been undertaken to estimate the burden of dementia or to identify its risk factors in ESRD. In the general population, age, cerebrovascular disease and cardiovascular risk factors including diabetes, hypertension, hyperlipidaemia and smoking are risk factors for dementia [5]. Among persons with ESRD, a number of uraemia-related factors have been implicated as risk factors for cognitive impairment, including an insufficient dose of dialysis, anaemia and aluminium exposure [6–8], although these findings have not been confirmed in recent studies [3,4].
Despite the strong association of dementia with morbidity and mortality in the general elderly population, there are limited data regarding the association of dementia with outcomes in ESRD. In particular, no studies to date have examined the association of dementia with the risks of death and dialysis withdrawal. Using the data from a large international sample of haemodialysis patients, we sought to examine the prevalence and correlates of dementia among persons with ESRD across multiple geographic regions. We also aimed to determine whether dementia was independently associated with all-cause mortality and dialysis withdrawal.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The Dialysis Outcomes and Practice Patterns Study (DOPPS) is an international prospective study of haemodialysis practice patterns and related outcomes. The design of DOPPS has been previously published [9]. Random samples of 20–40 haemodialysis patients (including incident and prevalent patients) from national representative haemodialysis facilities in seven countries, the US, Japan, France, Italy, Germany, Spain and the UK, were selected for participation in the DOPPS. Data collection began in 1997 in the US, 1998 in Europe and 1999 in Japan, with outcome ascertainment available through 2002 in the US, 2000 in Europe and 2001 in Japan. Patients were replaced as they left participating facilities for reasons of death, transplantation, change in treatment modality, recovery of renal function or transfer to another facility. Institutional review boards in each country approved the study, and all patients gave informed consent according to national requirements. The present analyses include 16 694 patients with complete baseline demographic data and non-missing dementia diagnosis.
Definition of physician-diagnosed dementia and covariates
Study coordinators at each facility abstracted data from the medical record using a standardized data collection instrument. Demographic factors, comorbid conditions such as cardiovascular disease and diagnosed depression, years on dialysis (‘vintage’), medication use and laboratory measures were recorded at the time of study entry. Dementia was defined as a diagnosis of known or suspected dementia listed in the medical record. All the patients were tracked up to the end of the follow-up period for death and dialysis withdrawal.
Statistical analyses
Continuous variables were expressed as mean ± SD and were compared using Student's t-test. Categorical variables were expressed as proportions and compared using the chi-squared test. We estimated the prevalence of dementia in the overall sample and within the subgroups by age and country. We used logistic regression to determine the association [expressed as odds ratios (OR) with 95% confidence intervals (95% CI)] among a variety of patient characteristics and dementia. Model building proceeded as follows. We first conducted unadjusted analyses to determine the association of demographic, clinical and laboratory variables with dementia. Multivariable models were adjusted for country of residence and effects of facility clustering, in addition to demographic, clinical and laboratory variables. Variables that were not significantly associated with dementia were added back to the model individually to evaluate for residual confounding, defined as a change in the parameter estimate >10%. We included interaction terms to test for effect modification by age, vintage and region with other clinical characteristics. We used Cox proportional hazards models to evaluate the relative hazard, or relative risk (RR) of mortality and dialysis withdrawal in relation to dementia. We constructed two models for these analyses. In our primary model, we adjusted for demographic characteristics, comorbid diagnoses and laboratory values a priori hypothesized to be associated with dementia, death/withdrawal or both. In a second model, we also adjusted for functional status variables that may be in the ‘causal pathway’ (in other words, potential manifestations of, rather than risks for, dementia, including cachexia, the ability to eat independently and the ability to transfer independently). This latter model would provide the most conservative estimates of the RR attributed directly to dementia itself. To assess whether geographic factors materially influenced our results, we stratified the analyses by continents. Analyses were conducted using SAS version 8.2 (SAS Institute, Cary, NC, USA).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patient characteristics
The mean age of the study sample was 60 ± 15 years, 57% were male, 19% were black and 67% had graduated high school. The overall prevalence of diagnosed dementia in the study cohort was 4%. Despite a similar age distribution among the DOPPS countries, the prevalence of diagnosed dementia varied 6-fold, with the lowest prevalence observed in Europe (1–3%), and the highest prevalence in the US (6%) (Table 1). The prevalence of diagnosed dementia increased substantially with age (Figure 1). Below the age of 60, the prevalence of diagnosed dementia ranged from 1 to 2%. Over the age of 60, the prevalence of diagnosed dementia nearly doubled with each decade of age, with prevalence rates ranging from 3% for ages 60–69 to 18% for ages 90 and above.
|
|
Clinical correlates of dementia
A number of clinical characteristics were associated with diagnosed dementia (Table 2). In addition to older age, black race, fewer years of education, cerebrovascular disease, diabetes, cancer, diagnosed depression, the absence of residual renal function, markers of malnutrition and low haemoglobin concentration were independently associated with diagnosed dementia. The associations between select clinical characteristics and diagnosed dementia varied by DOPPS country (interaction P-value <0.05). Older age was more strongly associated with diagnosed dementia in Japan than in Europe or the US. For example, relative to subjects aged <60 years, persons aged
60 years had a 9-fold increased odds of diagnosed dementia in Japan (OR 9.45, 95% CI 3.44–25.95) and a 3-fold increased odds of diagnosed dementia in Europe and the US (OR 2.68, 95% CI 2.10–3.43). Similarly, depression was more strongly associated with diagnosed dementia in Japan (OR 12.99, 95% CI 4.41–38.46) than in Europe or the US (OR 2.36, 95% CI 1.95–2.87). The prevalence of diagnosed dementia was similar between incident (n = 2553) and prevalent (n = 14 141) ESRD patients (4% vs 4%, P = 0.43), and incident status did not modify the association of diagnosed dementia with other clinical characteristics.
|
In exploratory analyses, we examined the association of diagnosed dementia with haemoglobin concentration by quartiles to evaluate for a ‘dose-response’ relation. The ORs were relatively uniform among patients within the lower three quartiles: haemoglobin <9.2 g/dl (OR 1.59, 95% CI 1.06–2.38), haemoglobin
9.2 g/dl and <10.3 g/dl (OR 1.48, 95% CI 1.00–2.17), and haemoglobin 10.3 g/dl and <11.3 g/dl (OR 1.40, 95% CI 0.98–2.00), relative to patients with a haemoglobin 11.3 g/dl. Erythropoietin use, age and sex did not significantly modify these associations. There was no significant association among smoking status, blood pressure, cholesterol, parathyroid hormone, dialysis dose or aluminium concentration and diagnosed dementia. Because aluminium levels were not routinely measured in DOPPS, we also determined potential exposure to aluminium through aluminium-containing binders or inadequate water treatment. The use of aluminium-containing binders was slightly less common among persons with dementia (3% vs 4%, P = 0.06), and <1% of DOPPS facilities used inadequate water treatment practices that might result in aluminium exposure.
Prospective association of diagnosed dementia with dialysis outcomes
Figures 2 and 3 demonstrate the age-adjusted survival curves for an all-cause mortality and dialysis withdrawal, respectively, stratified by diagnosed dementia. After adjusting for the clinical characteristics listed in Table 2, diagnosed dementia was associated with an increased risk of death (RR 1.48, 95% CI 1.32–1.66) and dialysis withdrawal (RR 2.01, 95% CI 1.57–2.57) (Table 3). When we further adjusted for functional status, the association between diagnosed dementia and the risk of death was attenuated, but remained statistically significant. The associations of dementia with all-cause mortality and dialysis withdrawal were similar among incident and prevalent dialysis patients. For example, diagnosed dementia was associated with a 1.8-fold increased risk of death (RR 1.80, 95% CI 1.27–2.56) among incident patients and a 1.4-fold increased risk of death (RR 1.45, 95% CI 1.28–1.64) among prevalent dialysis patients. To assess whether the grouping of patients from different countries affected our results, we performed a stratified analysis by DOPPS region. The results of these analyses were not qualitatively different. (Note: there were no dialysis withdrawals recorded in Japan.)
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In this large, internationally representative sample of haemodialysis patients, diagnosed dementia was associated with an increased risk of death and dialysis withdrawal, independent of a number of confounding factors. Risk factors for dementia in the general population, such as age, race, educational attainment, diabetes and cerebrovascular disease, as well as several modifiable uraemia-related factors, were associated with diagnosed dementia among patients with ESRD, suggesting potential avenues for risk reduction.
In DOPPS, the prevalence of diagnosed dementia was generally similar to that reported in large studies of the general elderly population in which objective measures of cognitive function were utilized. For example, the prevalence of dementia among 65–75-year-olds in the US is estimated at 5–9%, and quadruples with each decade of age [10]. It is estimated that physician recognition and documentation of dementia underestimates the true prevalence of dementia by 17–65% [11,12]. Thus, if the providers of ESRD care underestimate (or under-document) the presence of dementia at a similar frequency to general practitioners, then it is likely that the true prevalence of dementia is substantially higher among ESRD patients compared with the general population. Despite a similar age distribution, the estimated prevalence of diagnosed dementia varied across DOPPS countries. While some of the regional variations in dementia prevalence might be attributable to subtle differences in case mix at dialysis initiation or to the effects of dialysis and other therapies thereafter, European providers may also be less likely to diagnose or document dementia compared with their US counterparts. Some of these variations may also be attributable to differences in genetic or acquired risk factors across populations. For example, Japanese patients have high rates of stroke, and while absolute prevalence rates of dementia in Japan are similar to other industrialized nations, the ratio of vascular dementia to Alzheimer's dementia in Japan is much higher [13].
These results suggest that several non-modifiable risk factors for dementia in the general population are also risk factors for dementia in ESRD patients. Age is the single most important risk factor for dementia in the general population, so it is not surprising that age was strongly associated with dementia in the current study. The differential influence of age on dementia risk in Japan vs Europe and the US could be due to differences in underlying genetic or acquired risk factors (such as APOE genotype) or differences in ascertainment. Cerebrovascular disease was associated with an almost 3-fold increased risk of the diagnosed dementia, consistent with the study by Fukunishi et al. [2], who reported an increased incidence of vascular dementia among Japanese ESRD patients. Among modifiable cardiovascular risk factors, only diabetes was significantly associated with the diagnosed dementia in the current study. As is the case for the prediction of cardiovascular events in ESRD patients, the absence of an association among hypertension, blood pressure and cholesterol and diagnosed dementia may reflect residual confounding by malnutrition, or the effects of known and unknown ‘non-traditional’ risk factors for cardiovascular disease.
A number of ‘uraemic’ factors have been implicated in the pathogenesis of cognitive impairment among ESRD patients, but have not been well-studied as risk factors for dementia, per se. Erythropoietin therapy ameliorates the symptoms of uraemic encephalopathy among ESRD patients [6]; however, anaemia is not a consistent risk factor for cognitive impairment and dementia in the general elderly population. Moreover, contemporary cross-sectional studies of ESRD patients have found no relation between haemoglobin level and the presence of cognitive impairment. In the current study, anaemia, defined as a haemoglobin concentration <10 g/dl, was associated with a 25% increased risk of dementia. Exploratory analyses suggested a threshold of 
11 g/dl rather than a dose–response relation, though this finding should be interpreted with caution. Since most haemodialysis programmes target haemoglobin concentrations above 10–11 g/dl, it is unknown whether the anaemia itself or other factors (e.g. inflammation or other comorbid conditions such as gastrointestinal bleeding) preventing achievement of the haemoglobin target are responsible for the increased risk of dementia associated with lower haemoglobin concentrations.
The presence of residual renal function was also associated with a significantly lower risk of diagnosed dementia, suggesting a role for middle-molecule neurotoxins or cumulative effects of the dialysis process, such as oxidative stress, in the pathogenesis of dementia. Conversely, this finding might simply reflect the inability to ascertain residual renal function among persons with dementia. Diagnosed dementia was associated with lower body weight and lower serum albumin concentration, both potentially reflecting the presence of inflammation, which has been determined to be a risk factor for dementia in the general population [14]. Finally, we found no association between exposure to aluminium and diagnosed dementia, suggesting that dementia as a consequence of aluminium toxicity is likely to be extremely rare among contemporary haemodialysis patients.
Depressive symptoms are associated with poorer performance on tests of cognitive function [15]. Conversely, depressive symptoms are reported in up to half of patients with dementia [16], and several prospective studies have suggested that treated depression or depressive symptoms are risk factors for the development of dementia [17,18]. Similar to these studies in the general population, we found that diagnosed depression was associated with a 2-fold increased risk of diagnosed dementia, although further studies are needed to understand the complex relation between depression, cognitive impairment and dementia in ESRD.
Relatively few studies have examined the outcomes associated with dementia and cognitive impairment in ESRD. Kimmel et al. [19] found that Medicare recipients with ESRD had higher hospitalization rates for dementia compared with Medicare recipients with other chronic illnesses. Sehgal et al. [3] reported that cognitive impairment, measured by the Mini-Mental Status Exam, was associated with an increased number of hospital days and greater staff time after the termination of a dialysis session. Also using data from DOPPS, Mapes et al. [20] demonstrated that a 10-point lower score on the Kidney Disease Quality of Life Cognitive Function subscale was associated with a 2% increased risk of hospitalization and 5% increased risk of death. Our study extends these findings by demonstrating that diagnosed dementia is associated with an increased risk of death and dialysis withdrawal. Collectively, these data underscore the adverse outcomes and increased utilization of health care resources associated with cognitive impairment and dementia, emphasizing the importance of communicating prognosis and addressing treatment goals and advance directives with patients and families.
Given these findings, dialysis providers should consider instituting routine screening for cognitive impairment among elderly patients in order to identify those at risk for poor outcomes who may benefit from an improved management. For example, two classes of medications are now approved for the treatment of mild or moderate dementia. While the benefits of pharmacological therapy are only modest and have not been specifically tested in ESRD patients, many more drugs are in development that may also prove to be efficacious. In addition, several non-pharmacological interventions to manage behavioural, psychosocial, functional and caregiver needs have demonstrated utility in the management of dementia. Dialysis care providers may find a multidisciplinary approach involving geriatricians and/or geriatric psychiatrists may affect aspects of ESRD management, such as improved medication and dietary adherence. Finally, diagnosis allows a discussion of prognosis, placement options and end-of-life care that may enhance understanding, and ultimately reduce patient and caregiver stress.
This study has several limitations. First, the diagnosis of dementia was based on review of the medical record and not on formal cognitive function testing, and as previous studies suggest, ascertainment of dementia based on physician diagnosis is likely to underestimate the true number of cases of dementia and significant cognitive impairment. Thus, under-ascertainment of dementia cases or differential misclassification may have biased the results. Second, the analyses of baseline characteristics with dementia were cross-sectional, and we did not examine the association of clinical characteristics with incident dementia. Finally, we were unable to assess the contribution of several novel risk factors for dementia in the general population, such as homocysteine and inflammatory cytokines, since these markers were not measured in DOPPS.
In summary, we have identified a number of modifiable factors associated with diagnosed dementia in ESRD patients, suggesting avenues for therapeutic intervention in future studies. Given the anticipated increase in the number of elderly persons with ESRD and the adverse outcomes associated with dementia, prospective studies to define the optimal methods for detection, prevention and treatment of dementia are urgently needed in this high-risk population.
|
Acknowledgments |
|---|
M.K. was supported by the ASN-ASP-Junior Development award in Geriatric Nephrology, funded through Atlantic Philanthropies, the American Society of Nephrology and the John A. Hartford Foundation. G.M.C was supported by NIH–NIDDK RO1 DK58411 and NIH–NIDDK RO1 DK01005. The Dialysis Outcomes and Practice Patterns Study (DOPPS) is supported by research grants from Amgen and Kirin without restrictions on publications.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Brookmeyer R, Gray S, Kawas C. Projections of Alzheimer's disease in the United States and the public health impact of delaying disease onset. Am J Public Health 1998; 88: 1337–1342
[Abstract/Free Full Text] - Fukunishi I, Kitaoka T, Shirai T, Kino K, Kanematsu E, Sato Y. Psychiatric disorders among patients undergoing hemodialysis therapy. Nephron 2002; 91: 344–347[CrossRef][ISI][Medline]
- Sehgal AR, Grey SF, DeOreo PB, Whitehouse PJ. Prevalence, recognition, and implications of mental impairment among hemodialysis patients. Am J Kidney Dis 1997; 30: 41–49[ISI][Medline]
- Kurella M, Chertow GM, Luan J, Yaffe K. Cognitive impairment in chronic kidney disease. J Am Geriatr Soc 2004; 52: 1863–1869[CrossRef][ISI][Medline]
- Knopman D, Boland LL, Mosley T et al. Cardiovascular risk factors and cognitive decline in middle-aged adults. Neurology 2001; 56: 42–48
[Abstract/Free Full Text] - Grimm G, Stockenhuber F, Schneeweiss B, Madl C, Zeitlhofer J, Schneider B. Improvement of brain function in hemodialysis patients treated with erythropoietin. Kidney Int 1990; 38: 480–486[ISI][Medline]
- Teschan PE, Bourne JR, Reed RB, Ward JW. Electrophysiological and neurobehavioral responses to therapy: the National Cooperative Dialysis Study. Kidney Int 1983; 13: S58–S65
- Rozas VV, Port FK, Rutt WM. Progressive dialysis encephalopathy from dialysate aluminum. Arch Intern Med 1978; 138: 1375–1377[Abstract]
- Pisoni RL, Gillespie BW, Dickinson DM, Chen K, Kutner MH, Wolfe RA. The Dialysis Outcomes and Practice Patterns Study (DOPPS): design, data elements, and methodology. Am J Kidney Dis 2004; 44: 7–15[CrossRef][Medline]
- Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch Neurol 2003; 60: 1119–1122
[Abstract/Free Full Text] - Chodosh J, Petitti DB, Elliott M et al. Physician recognition of cognitive impairment: evaluating the need for improvement. J Am Geriatr Soc 2004; 52: 1051–1059[CrossRef][ISI][Medline]
- Valcour VG, Masaki KH, Curb JD, Blanchette PL. The detection of dementia in the primary care setting. Arch Intern Med 2000; 160: 2964–2968
[Abstract/Free Full Text] - Farrer LA. Intercontinental epidemiology of Alzheimer disease: a global approach to bad gene hunting. JAMA 2001; 285: 796–798
[Free Full Text] - Yaffe K, Lindquist K, Penninx BW et al. Inflammatory markers and cognition in well-functioning African-American and white elders. Neurology 2003; 61: 76–80
[Abstract/Free Full Text] - Rubin EH, Kinscherf DA, Grant EA, Storandt M. The influence of major depression on clinical and psychometric assessment of senile dementia of the Alzheimer type. Am J Psychiatry 1991; 148: 1164–1171
[Abstract/Free Full Text] - Li Y, Meyer JS, Thornby J. Depressive symptoms among cognitively normal versus cognitively impaired elderly subjects. Int J Geriatr Psychiatry 2001; 16: 455–461[CrossRef][ISI][Medline]
- Modrego PJ, Ferrandez J. Depression in patients with mild cognitive impairment increases the risk of developing dementia of Alzheimer type: a prospective cohort study. Arch Neurol 2004; 61: 1290–1293
[Abstract/Free Full Text] - Jorm AF, van Duijn CM, Chandra V et al. Psychiatric history and related exposures as risk factors for Alzheimer's disease: a collaborative re-analysis of case-control studies. EURODEM Risk Factors Research Group. Int J Epidemiol 1991; 20 [Suppl 2]: S43–S47
- Kimmel PL, Thamer M, Richard CM, Ray NF. Psychiatric illness in patients with end-stage renal disease. Am J Med 1998; 105: 214–221[CrossRef][ISI][Medline]
- Mapes DL, Lopes AA, Satayathum S et al. Health-related quality of life as a predictor of mortality and hospitalization: the Dialysis Outcomes and Practice Patterns Study (DOPPS). Kidney Int 2003; 64: 339–349[CrossRef][ISI][Medline]
Accepted in revised form: 20. 4.06
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. L. Wingard, L. B. Pupim, M. Krishnan, A. Shintani, T. A. Ikizler, and R. M. Hakim Early Intervention Improves Mortality and Hospitalization Rates in Incident Hemodialysis Patients: RightStart Program Clin. J. Am. Soc. Nephrol., November 1, 2007; 2(6): 1170 - 1175. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||