- Email: [email protected]
Temporal effects of dialysis on cognitive functioning in patients with ESRD
Temporal Effects of Dialysis on Cognitive Functioning in Patients With ESRD Mark A. Williams, MA, Allan H. Sklar, MD, Richard G. Burright, PhD, and Peter J. Donovick, PhD ● Background: Although dialysis has been shown to improve cognitive deficits resulting from uremia, little is known about potential temporal variations in cognitive measures between hemodialysis treatments. Methods: We studied dialysis patients for possible fluctuations in attention and memory by using repeatable neuropsychological measures. Twenty patients undergoing hemodialysis on a thrice-weekly dialysis schedule were assessed at intervals of 1, 24, and 67 hours after the last weekly hemodialysis session. For purposes of comparison, we also studied 10 patients on continuous ambulatory peritoneal dialysis (CAPD) therapy at similar intervals. Results: CAPD patients showed overall stable performance on measures of attention and memory. However, hemodialysis patients showed significant changes (P < 0.05) in auditory memory for both immediate and delayed recall, with the greatest impairment occurring 67 hours postdialysis. Hemodialysis patients also showed a significant change (P < 0.05) in attention between the second and third test periods, ie, 24 to 67 hours postdialysis, with the greatest impairment seen at the 67-hour assessment time. No relationship was found between performance on measures of cognition and levels of postdialysis fatigue in hemodialysis patients. Conclusion: CAPD patients showed cognitive stability, whereas hemodialysis patients showed temporal fluctuations in cognitive performance. Am J Kidney Dis 43:705-711. © 2004 by the National Kidney Foundation, Inc. INDEX WORDS: End-stage renal disease (ESRD); attention; memory; neuropsychology; hemodialysis (HD).
U
REMIA WAS characterized as far back as 1839 as being primarily a neurobehavioral syndrome.1 In the modern era, neuropsychological testing has shown that uremic patients show significant impairment on measures of attention and memory.2 Patients with end-stage renal disease (ESRD) who are undergoing hemodialysis consistently perform significantly better on neuropsychological measures of memory and attention approximately 24 hours after their last hemodialysis treatment compared with nondialyzed uremic subjects.3 Previous research found no clear neuropsychological deficits in patients at 24 hours after hemodialysis treatment compared with medical controls.4 In one of the few studies to address cognitive fluctuations in patients with ESRD dialyzed long term, researchers concluded that despite significant daily changes in serum levels of toxic substances retained in uremia, there was little evidence to suggest that hemodialysis patients undergo daily fluctuations in cognitive and sensory-motor functioning.5 However, these investigators did not examine hemodialysis patients at their most vulnerable period of uremia; that is, after a weekend without dialysis. Previous research examining reaction time and visual attention (ie, digit symbol subtest of the Wechsler Adult Intelligence Scale-Revised) did not show a relationship between performance and time since
dialysis treatment.6 However, this study examined only 6 participants and did not comprehensively examine measures of memory that have since shown vulnerability in patients with ESRD.5,7,8 Research comparing the effects of hemodialysis with those of continuous ambulatory peritoneal dialysis (CAPD) on cognitive functioning found patients undergoing hemodialysis, when tested at times ranging from 60 minutes before to 30 minutes after the initiation of a dialysis treatment, had significantly greater cognitive deficits compared with matched patients on CAPD therapy at their routine monthly clinic visit on measures of attention, learning, and visualspatial orientation.9 These results suggest that CAPD may provide for greater stability in restoFrom the Environmental Neuropsychology Laboratory, State University of New York at Binghamton; and the Department of Medicine and Neuropsychology Unit, United Health Services Hospitals, Johnson City, NY. Received August 4, 2003; accepted in revised form December 5, 2003. Supported in part by The Southern New York Neuropsychological Group. Address reprint requests to Peter J. Donovick, PhD, Department of Psychology, PO Box 6000, Binghamton University, Binghamton, NY 13902-6000. E-mail: donovick@ binghamton.edu © 2004 by the National Kidney Foundation, Inc. 0272-6386/04/4304-0013$30.00/0 doi:10.1053/j.ajkd.2003.12.031
American Journal of Kidney Diseases, Vol 43, No 4 (April), 2004: pp 705-711
705
706
WILLIAMS ET AL
ration of cognitive abilities that have been compromised by the uremic state. However, interpretations of results of this investigation are limited because interdialytic periods were not specified and sequential neuropsychological testing was not conducted. To date, the studies described provide some evidence for the efficacy of hemodialysis in optimizing cognitive functioning at a 24-hour point in time after hemodialysis, but fall short of examining any changes that may occur in cognitive performance throughout the week. This issue has particular relevance today in view of the interest in quotidian hemodialysis.10 The major goal of the current study is to examine potential temporal fluctuations in memory and attention in patients with ESRD during the longest interdialytic period of the hemodialysis cycle. Hemodialysis patients’ performances on cognitive measures were examined 1, 24, and 67 hours after the last weekly hemodialysis session. These times were selected to provide performance measures at various points during the hemodialysis weekend period. Results of cognitive performance in hemodialysis patients were compared with those obtained in a group of patients with ESRD on continuous dialysis therapy, ie, CAPD patients, tested at the same intervals. In addition, we examined both hemodialysis and CAPD performances on neuropsychological measures compared with available norms. Finally, we evaluated the potential influence of fatigue, which is commonly experienced by patients after a hemodialysis treatment, on cognitive performance. METHODS
Participants Thirty patients with ESRD (15 men, 15 women) from a community hospital dialysis facility in upstate New York volunteered to participate in the study. Twenty participants were on hemodialysis therapy and 10 participants were on CAPD therapy. Inclusion criteria included the following: patients with ESRD aged 18 years and older; completion of at least 3 months of maintenance dialysis therapy before study entry; all hemodialysis and CAPD participants were required to have stable urea clearance with a Kt/V greater than 1.2 and Kt/V greater than 2.0, respectively; and a hematocrit greater than 30% for at least 3 months. Patients on the earliest hemodialysis shift of the day or with a history of alcoholism, brain injury, dementia, or psychosis were not considered for participation in the study. It must be noted that no patient on the first shift was employed or attended
school. All participants received preliminary auditory and visual acuity screens, along with a screen for color blindness to assess for any sensory deficits that may impair performance on the neuropsychological measures used in this study. Written informed consent was obtained from all patients, and the study was approved by both the Binghamton University Human Subjects Research Review Committee and the United Health Services Hospitals’ Human Assurance Committee.
Neuropsychological Tests Administered Dodrill Stroop. The Stroop measure is one of the most commonly used tests of selective attention.11 Part I of the Dodrill Stroop test requires the participant to read words that are color names printed in incongruent colors. Part II requires naming the color of the ink in which the color names are printed. Studies have consistently shown that the speed of reading color names invariably is faster than naming the ink color. The Stroop Effect score is obtained by subtracting the time to complete Part I from the time to complete Part II. For the purpose of our repeated design, equivalent forms of the Dodrill Stroop12 were used. Rey Auditory-Verbal Learning Test. The Rey AuditoryVerbal Learning Test (RAVLT) was used to evaluate memory functioning.13 The RAVLT is a word-list learning test. This measure consists of lists of 15 words that are read 1 at a time to the participant during a series of 5 trials. The RAVLT provides a measure of both immediate and 20-minute delayed recall, as well as recognition of words presented. We used alternate forms14 for each test time.
Intelligence Measure Kaufman Brief Intelligence Test. The Kaufman Brief Intelligence Test (K-BIT) is a brief, individually administered, psychometric test that provides a measure of verbal and nonverbal intelligence.15 It has been used in a variety of clinical and nonclinical populations.16 K-BIT scores correlate moderately well with other measures of intelligence, such as the Wechsler Adult Intelligence Scale-Revised.17
Symptoms of Depression Measure Beck Depression Inventory-II. The Beck Depression Inventory-II (BDI-II) is a 21-item measure of the depth or severity of depression that was derived from both clinical and research sources.18 The BDI-II was administered to examine for the presence and possible influence of depression symptoms on cognitive performance.
Protocol Repeated measures of the Stroop and RAVLT were administered to the 20 hemodialysis patients to examine possible fluctuations in cognitive performance as a result of time since dialysis. For comparative purposes, the same repeated measures also were administered to 10 CAPD patients at intervals equivalent to those of the hemodialysis group. For the purpose of this study, we designated evaluation session times as follows. For hemodialysis patients; T1 is 1 hour after the Friday or Saturday (ie, preweekend) hemodialysis
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
session, T2 is 24 hours after T1, and T3 is 67 hours after T1, ie, immediately before the Monday or Tuesday hemodialysis session. For CAPD patients, a baseline time for T1 was established arbitrarily. Once T1 was established for CAPD patients, their evaluation interval hours were identical to those of the hemodialysis group. Repeated measures of the neuropsychological tests were administered at T1, T2, and T3. One-time assessments of intelligence and symptoms of depression were administered at T2. This interval was selected because of previous studies indicating dialysis patients are cognitively efficient at 24 hours after dialysis treatment. All testing was conducted in a quiet room proximal to the dialysis unit. To evaluate the potential impact of postdialysis fatigue on cognitive performance, we had each patient record their level of fatigue on an hourly basis throughout the 67-hour study period. Patients indicated level of fatigue on a 5-point Likert scale ranging from 1, no fatigue, to 5, overwhelming fatigue (sleeping).
Statistical Analysis Data were analyzed using SPSS 10.0 for Windows statistical software (SSPS Inc, Chicago, IL).19 Analysis included descriptive data, 2-factor mixed analysis of variance (ANOVA), chi-square analysis, and paired-sample t-tests, as well as independent-samples t-tests. Performance on psychological measures by the hemodialysis and CAPD groups were examined in a 2 ⫻ 3 mixed design ANOVA. Standard scores were calculated to determine how hemodialysis and CAPD group performances compared with a sample of age-, sex-, and/or education-matched individuals without ESRD. All data are expressed as mean ⫾ SE. P of 0.05 was used to establish statistical significance.
RESULTS
Characteristics of study participants are listed in Table 1. Although the average hemodialysis patient was older than the average CAPD patient by nearly 10 years, an independent-samples ttest analysis indicated the difference was not statistically different. In addition, when we compared older (ⱖ50 years) with younger (ⱕ49 years), the hemodialysis and CAPD groups did not differ significantly in age distribution of participants. Chi-square analyses indicated that the groups did not differ in terms of level of education and income. Average time on dialysis therapy was almost 30 months longer for the hemodialysis group compared with the CAPD group. However, an independent-samples t-test analysis indicated the groups did not differ significantly in dialysis vintage. In addition, we compared patients who had received dialysis treatment for 30 months or longer with those receiving dialysis treatment for less than 30 months. We found no difference in vintage distribution be-
707 Table 1.
Characteristics of Study Participants Hemodialysis (n ⫽ 20)
Age (y) Women Men Education ⱕGrade 12 High school diploma/GED Some college College degree Ethnicity Caucasian African American Other Income ($) ⬍10,000 10,001-19,999 20,000-39,999 ⬎40,000 Duration of dialysis (mo) Cause of ESRD Diabetes Hypertension Glomerulonephritis Polycystic kidney disease Other Mean IQ (K-BIT) Mean depression symptoms (BDI-II)
CAPD (n ⫽ 10)
54.6 ⫾ 2.9 10 10
45.1 ⫾ 4.8 5 5
1 10 8 1
2 3 4 1
18 2 0
8 1 1
6 7 3 4 65.4 ⫾ 13.1
4 4 2 0 36.1 ⫾ 7.2
6 4 3 2 5 104.8 ⫾ 1.8 11.9 ⫾ 1.8
5 2 0 1 2 102.2 ⫾ 4.0 17.4 ⫾ 3.0
NOTE. Values expressed as mean ⫾ SE or number of patients, unless noted otherwise.
tween the hemodialysis and CAPD groups. Furthermore, both groups were equivalent and within an average range on intellectual functioning measured by the K-BIT. On average, CAPD and hemodialysis patients endorsed depressive symptoms in the minimal to mild range on the BDI-II measure. The groups did not differ significantly in the degree of depressive symptoms they endorsed. Dodrill Stroop An interaction between dialysis group and test time was shown for the Stroop Effect. Specifically, the hemodialysis group showed a significant decline in performance at T3 compared with T2 (t ⫽ ⫺2.65; df ⫽ 19; P ⬍ 0.05) on the Stroop Effect. Conversely, the CAPD group showed a significant improvement in performance at both T2 and T3 compared with T1 (t ⫽ 2.32; df ⫽ 9; P ⬍ 0.05 and t ⫽ 2.65; df ⫽ 9; P ⬍ 0.05,
708
respectively; Fig 1). Both the hemodialysis and CAPD groups performed within 1 SD of a normative sample11 on the Stroop Effect across all 3 time points. RAVLT There was a significant difference between groups on immediate recall. The hemodialysis group produced fewer words at each consecutive time, with the least number of words generated during T3. In comparison, the CAPD group showed a stable level of performance across all 3 test times (Fig 1). The hemodialysis group had significantly less immediate recall of words at T3 than the CAPD group (t ⫽ ⫺3.36; df ⫽ 29; P ⬍ 0.005). Results were similar for Delayed Recall, in which the hemodialysis group produced significantly fewer words than the CAPD group (t ⫽ ⫺3.58; df ⫽ 29; P ⬍ 0.005). For purposes of comparison, a normative sample of adult performance on the RAVLT was used.14 Mean performance of hemodialysis patients showed a progressive decline in performances on the immediate, delayed, and recognition tasks of the RAVLT compared with a normative sample of individuals of equivalent age, gender, and IQ. On average, hemodialysis patients performed within 1 SD of the normative sample on immediate and delayed recall at T1. However, at T2, the hemodialysis group performance decreased below 1 SD of the normative sample. By T3, the hemodialysis group on average was performing nearly 2 SDs below the normative mean on both immediate and delayed recall. Baseline impairment and decline were even greater for the recognition task in the hemodialysis group. They progressed from nearly 2 SDs below the normative mean at T1 to more than 4 SDs below the normative sample at T3. In comparison, CAPD patients generally earned scores within 1 SD across all 3 test periods. The 1 exception was the recognition task for CAPD patients. On this measure, CAPD patients showed, on average, scores greater than 2 SDs below a normative sample during the T3 test period. Fatigue Rating As shown in Fig 1, hemodialysis patients reported significantly more fatigue immediately after their dialysis treatments than at later points
WILLIAMS ET AL
in time in the interdialytic period. Levels of postdialysis fatigue did not correlate temporally with cognitive performance. As would be expected, CAPD patients had a stable fatigue profile throughout an identical period. DISCUSSION
Although dialysis has been shown to improve cognitive deficits resulting from uremia, little is known about potential temporal variation in cognitive measures between hemodialysis treatments. We studied dialysis patients for possible fluctuations in cognitive performance. We examined 20 patients undergoing thrice-weekly hemodialysis at intervals of 1, 24, and 67 hours postdialysis. For purposes of comparison, we also studied 10 peritoneal dialysis patients at similar intervals. The 2 groups of patients did not differ significantly from one another in age, gender, education, or level of depression, and both were in the average range of intelligence. On the Dodrill Stroop test, the hemodialysis group clearly showed a decrease in selective attention at T3 compared with their performance at T2. This is in striking contrast to the CAPD group, who clearly showed a pattern of sequential improvement in selective attention over the 3 test times, measured by the Stroop. This pattern of performance suggests the CAPD group may have benefited from a practice effect in their performance. Conversely, the hemodialysis group showed no indication of benefiting from this possible practice effect. Their performance significantly declined over time. This would suggest hemodialysis patients’ ability to learn is negatively effected in comparison to CAPD patients. On measures of memory using the RAVLT, the hemodialysis group again showed their most impaired performance at the T3 test time on measures of both immediate and delayed recall. Some caution should be applied on the interpretation of these memory measure results. Given that test time comparisons within groups were nested in a within-subjects repeated-measures design, the decline in ability for the hemodialysis group to recall words may be caused by proactive interference on the memory measure. That is, with increase in exposure to words to recall over the 3 trials, recall becomes more difficult.
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
Fig 1. Mean ⴞ SE for fatigue scores, Stroop Effect, and total words generated in the immediate and delayed recall of the RAVLT for the hemodialysis and CAPD groups at 1, 24, and 67 hours after hemodialysis and after baseline for the CAPD group. An increase in scores for both fatigue and Stroop Effect reflects greater impairment, whereas an increase in number of words produced in both the immediate and delayed recall reflects less impairment.
709
710
Even given the possibility of the proactive interference confound, it is interesting to note that the CAPD group showed no indications of being affected by it. This suggests that the hemodialysis group is more sensitive to the effect, if it is occurring. Results show that the hemodialysis group generally performed worse at 67 hours compared with the 1- and 24-hour postdialysis test times. These findings are consistent with a previous investigation showing a relationship between time since dialysis and visual evoked potential latency; the shortest latencies typically occurred at the 24-hour postdialysis period and steadily increased during the 42-hour and 66-hour intervals after the hemodialysis session.6 The 24-hour postdialysis test time is the period typically used when examining cognitive functioning in hemodialysis patients. These results suggest that capturing cognitive performance for this group at only the 24-hour test time does not reflect the actual cognitive difficulties that hemodialysis patients may be experiencing at the end of a weekend without dialysis. In comparison, CAPD patients, who are undergoing continuous dialysis, showed an overall stable performance in cognitive tasks across all 3 test times. It would be interesting to compare our findings with those observed in a group of patients undergoing emerging forms of daily or nocturnal hemodialysis therapies. What is responsible for the deterioration in cognitive performance in hemodialysis patients with the passage of time from the previous dialysis treatment? Our data suggest that fatigue does not account for the deterioration in mental acuity in these patients. Rather, the fluctuation of psychometric measures in temporal correlation with hemodialysis treatments suggests that increasing accumulation of toxic uremic metabolites in the interdialytic period is involved in cognitive fluctuations. It should be noted that although vintage time between the hemodialysis and CAPD group was not statistically significant, the 30-month increase in vintage time for the hemodialysis group could have clinical significance. This increased vintage time may contribute to increased susceptibility to cognitive assault in the hemodi-
WILLIAMS ET AL
alysis group. Future investigations need to address this issue. We believe we have shown fluctuations in cognitive performance in hemodialysis patients, determined by the Stroop and RVALT psychometric measures. These results suggest that hemodialysis patients might be progressively compromised in the cognitive domains of attention and functions of memory as the interdialytic period proceeds. The question remains whether these deficiencies reach a point of clinical significance. The impact of these cognitive limitations on compliance with diet, fluid restriction, and medication use, as well as ability to make complex decisions needed to engage in adequate self-care behaviors, remains to be determined. Cognitive functioning occurs in the context of a whole human being, and the impact of cognitive deterioration or fluctuation will be unique to each individual who endures it. ACKNOWLEDGMENT The authors thank the dialysis patients who participated in this study; Dr Raymond G. Romanczyk and Dr Carolyn M. Pepper for support, guidance, and feedback throughout the project; and Purvi D. Shah and Myriam J. Sollman for help in data collection.
REFERENCES 1. Addison T: On the disorders of the brain connected with diseased kidneys. Guy’s Hosp Rep 4:1-7, 1839 2. Hart RP, Pederson JA, Czerwinski AW, Adams RL: Chronic renal failure, dialysis, and neuropsychological function. J Neuropsychol 5:301-312, 1983 3. Ryan JJ, Souheaver GT, DeWolfe AS: Halstead-Reitan test results in chronic hemodialysis. J Nerv Ment Dis 16:311314, 1981 4. Pliskin NH, Yurk HM, Ho LT, Umans JG: Neurocognitive function in chronic hemodialysis patients. Kidney Int 49:1435-1440, 1996 5. Ratner DP, Adams KM, Levin W, Rourke BP: Effects of hemodialysis on the cognitive and sensory-motor functioning of the adult chronic hemodialysis patient. J Behav Med 6:291-311, 1983 6. Lewis EG, O’Neil WM, Dustman RE, Beck EC: Temporal effects of hemodialysis on measures of neural efficiency. Kidney Int 17:357-363, 1980 7. Souheaver GT, Ryan JJ, DeWolfe AS: Neurological patterns in uremia. J Clin Psychol 38:490-496, 1982 8. Osberg JW, Meares GJ, McKee DC: Intellectual functioning in renal failure and chronic dialysis. J Chronic Disord 35:445-457, 1982 9. Wolcott DL, Wellisch DK, Marsh JT, Schaeffer J, Landsverk J, Nissenson AR: Relationship of dialysis modal-
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
ity and other factors to cognitive functions in chronic dialysis patients. Am J Kidney Dis 12:275-284, 1988 10. Lindsay RM, Leitch R, Heidenheim AP: The London daily/nocturnal hemodialysis study—Study design, morbidity, and mortality results. Am J Kidney Dis 42:S5-S12, 2003 (suppl 1) 11. Dodrill CB: A neuropsychological battery for epilepsy. Epilepsy 19:611-623, 1978 12. Sacks TL, Clark CR, Pols R, Geffen LB: Comparability and stability of performance on six alternate forms of the Dodrill-Stroop Color-Word Test. Clin Neuropsychol 5:220225, 1991 13. Rey A: L’examen Clinique en psychologie. Paris, France, Presses Universitaires de France, 1964 14. Geffen G, Moar KJ, O’Hanlon AP, Clark CR, Geffen LB: Performance measures of 16 to 86 year old males and
711
females on the auditory verbal learning test. Clin Neuropsychol 4:45-63, 1990 15. Kaufman AS, Kaufman NL: Kaufman Brief Intelligence Test Manual. Circle Pines, MN, American Guidance Service, 1990 16. Donovick PJ, Burright RG, Burg JS: A screen for IQ in six diverse populations. J Clin Psychol Med Settings 3:131-140, 1996 17. Eisenstein N, Engelhart CI: Comparison of the KBIT with short forms of the WAIS-R in a neuropsychological population. Psychol Assess 9:57-62, 1997 18. Beck AT, Steer RA, Brown GK: Beck Depression Inventory Manual (ed 2). San Antonio, TX, Psychological Corp, 1996 19. SPSS Inc: Statistical Package for the Social Sciences, version 10.0. Chicago, IL, SSPS Inc.
U
REMIA WAS characterized as far back as 1839 as being primarily a neurobehavioral syndrome.1 In the modern era, neuropsychological testing has shown that uremic patients show significant impairment on measures of attention and memory.2 Patients with end-stage renal disease (ESRD) who are undergoing hemodialysis consistently perform significantly better on neuropsychological measures of memory and attention approximately 24 hours after their last hemodialysis treatment compared with nondialyzed uremic subjects.3 Previous research found no clear neuropsychological deficits in patients at 24 hours after hemodialysis treatment compared with medical controls.4 In one of the few studies to address cognitive fluctuations in patients with ESRD dialyzed long term, researchers concluded that despite significant daily changes in serum levels of toxic substances retained in uremia, there was little evidence to suggest that hemodialysis patients undergo daily fluctuations in cognitive and sensory-motor functioning.5 However, these investigators did not examine hemodialysis patients at their most vulnerable period of uremia; that is, after a weekend without dialysis. Previous research examining reaction time and visual attention (ie, digit symbol subtest of the Wechsler Adult Intelligence Scale-Revised) did not show a relationship between performance and time since
dialysis treatment.6 However, this study examined only 6 participants and did not comprehensively examine measures of memory that have since shown vulnerability in patients with ESRD.5,7,8 Research comparing the effects of hemodialysis with those of continuous ambulatory peritoneal dialysis (CAPD) on cognitive functioning found patients undergoing hemodialysis, when tested at times ranging from 60 minutes before to 30 minutes after the initiation of a dialysis treatment, had significantly greater cognitive deficits compared with matched patients on CAPD therapy at their routine monthly clinic visit on measures of attention, learning, and visualspatial orientation.9 These results suggest that CAPD may provide for greater stability in restoFrom the Environmental Neuropsychology Laboratory, State University of New York at Binghamton; and the Department of Medicine and Neuropsychology Unit, United Health Services Hospitals, Johnson City, NY. Received August 4, 2003; accepted in revised form December 5, 2003. Supported in part by The Southern New York Neuropsychological Group. Address reprint requests to Peter J. Donovick, PhD, Department of Psychology, PO Box 6000, Binghamton University, Binghamton, NY 13902-6000. E-mail: donovick@ binghamton.edu © 2004 by the National Kidney Foundation, Inc. 0272-6386/04/4304-0013$30.00/0 doi:10.1053/j.ajkd.2003.12.031
American Journal of Kidney Diseases, Vol 43, No 4 (April), 2004: pp 705-711
705
706
WILLIAMS ET AL
ration of cognitive abilities that have been compromised by the uremic state. However, interpretations of results of this investigation are limited because interdialytic periods were not specified and sequential neuropsychological testing was not conducted. To date, the studies described provide some evidence for the efficacy of hemodialysis in optimizing cognitive functioning at a 24-hour point in time after hemodialysis, but fall short of examining any changes that may occur in cognitive performance throughout the week. This issue has particular relevance today in view of the interest in quotidian hemodialysis.10 The major goal of the current study is to examine potential temporal fluctuations in memory and attention in patients with ESRD during the longest interdialytic period of the hemodialysis cycle. Hemodialysis patients’ performances on cognitive measures were examined 1, 24, and 67 hours after the last weekly hemodialysis session. These times were selected to provide performance measures at various points during the hemodialysis weekend period. Results of cognitive performance in hemodialysis patients were compared with those obtained in a group of patients with ESRD on continuous dialysis therapy, ie, CAPD patients, tested at the same intervals. In addition, we examined both hemodialysis and CAPD performances on neuropsychological measures compared with available norms. Finally, we evaluated the potential influence of fatigue, which is commonly experienced by patients after a hemodialysis treatment, on cognitive performance. METHODS
Participants Thirty patients with ESRD (15 men, 15 women) from a community hospital dialysis facility in upstate New York volunteered to participate in the study. Twenty participants were on hemodialysis therapy and 10 participants were on CAPD therapy. Inclusion criteria included the following: patients with ESRD aged 18 years and older; completion of at least 3 months of maintenance dialysis therapy before study entry; all hemodialysis and CAPD participants were required to have stable urea clearance with a Kt/V greater than 1.2 and Kt/V greater than 2.0, respectively; and a hematocrit greater than 30% for at least 3 months. Patients on the earliest hemodialysis shift of the day or with a history of alcoholism, brain injury, dementia, or psychosis were not considered for participation in the study. It must be noted that no patient on the first shift was employed or attended
school. All participants received preliminary auditory and visual acuity screens, along with a screen for color blindness to assess for any sensory deficits that may impair performance on the neuropsychological measures used in this study. Written informed consent was obtained from all patients, and the study was approved by both the Binghamton University Human Subjects Research Review Committee and the United Health Services Hospitals’ Human Assurance Committee.
Neuropsychological Tests Administered Dodrill Stroop. The Stroop measure is one of the most commonly used tests of selective attention.11 Part I of the Dodrill Stroop test requires the participant to read words that are color names printed in incongruent colors. Part II requires naming the color of the ink in which the color names are printed. Studies have consistently shown that the speed of reading color names invariably is faster than naming the ink color. The Stroop Effect score is obtained by subtracting the time to complete Part I from the time to complete Part II. For the purpose of our repeated design, equivalent forms of the Dodrill Stroop12 were used. Rey Auditory-Verbal Learning Test. The Rey AuditoryVerbal Learning Test (RAVLT) was used to evaluate memory functioning.13 The RAVLT is a word-list learning test. This measure consists of lists of 15 words that are read 1 at a time to the participant during a series of 5 trials. The RAVLT provides a measure of both immediate and 20-minute delayed recall, as well as recognition of words presented. We used alternate forms14 for each test time.
Intelligence Measure Kaufman Brief Intelligence Test. The Kaufman Brief Intelligence Test (K-BIT) is a brief, individually administered, psychometric test that provides a measure of verbal and nonverbal intelligence.15 It has been used in a variety of clinical and nonclinical populations.16 K-BIT scores correlate moderately well with other measures of intelligence, such as the Wechsler Adult Intelligence Scale-Revised.17
Symptoms of Depression Measure Beck Depression Inventory-II. The Beck Depression Inventory-II (BDI-II) is a 21-item measure of the depth or severity of depression that was derived from both clinical and research sources.18 The BDI-II was administered to examine for the presence and possible influence of depression symptoms on cognitive performance.
Protocol Repeated measures of the Stroop and RAVLT were administered to the 20 hemodialysis patients to examine possible fluctuations in cognitive performance as a result of time since dialysis. For comparative purposes, the same repeated measures also were administered to 10 CAPD patients at intervals equivalent to those of the hemodialysis group. For the purpose of this study, we designated evaluation session times as follows. For hemodialysis patients; T1 is 1 hour after the Friday or Saturday (ie, preweekend) hemodialysis
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
session, T2 is 24 hours after T1, and T3 is 67 hours after T1, ie, immediately before the Monday or Tuesday hemodialysis session. For CAPD patients, a baseline time for T1 was established arbitrarily. Once T1 was established for CAPD patients, their evaluation interval hours were identical to those of the hemodialysis group. Repeated measures of the neuropsychological tests were administered at T1, T2, and T3. One-time assessments of intelligence and symptoms of depression were administered at T2. This interval was selected because of previous studies indicating dialysis patients are cognitively efficient at 24 hours after dialysis treatment. All testing was conducted in a quiet room proximal to the dialysis unit. To evaluate the potential impact of postdialysis fatigue on cognitive performance, we had each patient record their level of fatigue on an hourly basis throughout the 67-hour study period. Patients indicated level of fatigue on a 5-point Likert scale ranging from 1, no fatigue, to 5, overwhelming fatigue (sleeping).
Statistical Analysis Data were analyzed using SPSS 10.0 for Windows statistical software (SSPS Inc, Chicago, IL).19 Analysis included descriptive data, 2-factor mixed analysis of variance (ANOVA), chi-square analysis, and paired-sample t-tests, as well as independent-samples t-tests. Performance on psychological measures by the hemodialysis and CAPD groups were examined in a 2 ⫻ 3 mixed design ANOVA. Standard scores were calculated to determine how hemodialysis and CAPD group performances compared with a sample of age-, sex-, and/or education-matched individuals without ESRD. All data are expressed as mean ⫾ SE. P of 0.05 was used to establish statistical significance.
RESULTS
Characteristics of study participants are listed in Table 1. Although the average hemodialysis patient was older than the average CAPD patient by nearly 10 years, an independent-samples ttest analysis indicated the difference was not statistically different. In addition, when we compared older (ⱖ50 years) with younger (ⱕ49 years), the hemodialysis and CAPD groups did not differ significantly in age distribution of participants. Chi-square analyses indicated that the groups did not differ in terms of level of education and income. Average time on dialysis therapy was almost 30 months longer for the hemodialysis group compared with the CAPD group. However, an independent-samples t-test analysis indicated the groups did not differ significantly in dialysis vintage. In addition, we compared patients who had received dialysis treatment for 30 months or longer with those receiving dialysis treatment for less than 30 months. We found no difference in vintage distribution be-
707 Table 1.
Characteristics of Study Participants Hemodialysis (n ⫽ 20)
Age (y) Women Men Education ⱕGrade 12 High school diploma/GED Some college College degree Ethnicity Caucasian African American Other Income ($) ⬍10,000 10,001-19,999 20,000-39,999 ⬎40,000 Duration of dialysis (mo) Cause of ESRD Diabetes Hypertension Glomerulonephritis Polycystic kidney disease Other Mean IQ (K-BIT) Mean depression symptoms (BDI-II)
CAPD (n ⫽ 10)
54.6 ⫾ 2.9 10 10
45.1 ⫾ 4.8 5 5
1 10 8 1
2 3 4 1
18 2 0
8 1 1
6 7 3 4 65.4 ⫾ 13.1
4 4 2 0 36.1 ⫾ 7.2
6 4 3 2 5 104.8 ⫾ 1.8 11.9 ⫾ 1.8
5 2 0 1 2 102.2 ⫾ 4.0 17.4 ⫾ 3.0
NOTE. Values expressed as mean ⫾ SE or number of patients, unless noted otherwise.
tween the hemodialysis and CAPD groups. Furthermore, both groups were equivalent and within an average range on intellectual functioning measured by the K-BIT. On average, CAPD and hemodialysis patients endorsed depressive symptoms in the minimal to mild range on the BDI-II measure. The groups did not differ significantly in the degree of depressive symptoms they endorsed. Dodrill Stroop An interaction between dialysis group and test time was shown for the Stroop Effect. Specifically, the hemodialysis group showed a significant decline in performance at T3 compared with T2 (t ⫽ ⫺2.65; df ⫽ 19; P ⬍ 0.05) on the Stroop Effect. Conversely, the CAPD group showed a significant improvement in performance at both T2 and T3 compared with T1 (t ⫽ 2.32; df ⫽ 9; P ⬍ 0.05 and t ⫽ 2.65; df ⫽ 9; P ⬍ 0.05,
708
respectively; Fig 1). Both the hemodialysis and CAPD groups performed within 1 SD of a normative sample11 on the Stroop Effect across all 3 time points. RAVLT There was a significant difference between groups on immediate recall. The hemodialysis group produced fewer words at each consecutive time, with the least number of words generated during T3. In comparison, the CAPD group showed a stable level of performance across all 3 test times (Fig 1). The hemodialysis group had significantly less immediate recall of words at T3 than the CAPD group (t ⫽ ⫺3.36; df ⫽ 29; P ⬍ 0.005). Results were similar for Delayed Recall, in which the hemodialysis group produced significantly fewer words than the CAPD group (t ⫽ ⫺3.58; df ⫽ 29; P ⬍ 0.005). For purposes of comparison, a normative sample of adult performance on the RAVLT was used.14 Mean performance of hemodialysis patients showed a progressive decline in performances on the immediate, delayed, and recognition tasks of the RAVLT compared with a normative sample of individuals of equivalent age, gender, and IQ. On average, hemodialysis patients performed within 1 SD of the normative sample on immediate and delayed recall at T1. However, at T2, the hemodialysis group performance decreased below 1 SD of the normative sample. By T3, the hemodialysis group on average was performing nearly 2 SDs below the normative mean on both immediate and delayed recall. Baseline impairment and decline were even greater for the recognition task in the hemodialysis group. They progressed from nearly 2 SDs below the normative mean at T1 to more than 4 SDs below the normative sample at T3. In comparison, CAPD patients generally earned scores within 1 SD across all 3 test periods. The 1 exception was the recognition task for CAPD patients. On this measure, CAPD patients showed, on average, scores greater than 2 SDs below a normative sample during the T3 test period. Fatigue Rating As shown in Fig 1, hemodialysis patients reported significantly more fatigue immediately after their dialysis treatments than at later points
WILLIAMS ET AL
in time in the interdialytic period. Levels of postdialysis fatigue did not correlate temporally with cognitive performance. As would be expected, CAPD patients had a stable fatigue profile throughout an identical period. DISCUSSION
Although dialysis has been shown to improve cognitive deficits resulting from uremia, little is known about potential temporal variation in cognitive measures between hemodialysis treatments. We studied dialysis patients for possible fluctuations in cognitive performance. We examined 20 patients undergoing thrice-weekly hemodialysis at intervals of 1, 24, and 67 hours postdialysis. For purposes of comparison, we also studied 10 peritoneal dialysis patients at similar intervals. The 2 groups of patients did not differ significantly from one another in age, gender, education, or level of depression, and both were in the average range of intelligence. On the Dodrill Stroop test, the hemodialysis group clearly showed a decrease in selective attention at T3 compared with their performance at T2. This is in striking contrast to the CAPD group, who clearly showed a pattern of sequential improvement in selective attention over the 3 test times, measured by the Stroop. This pattern of performance suggests the CAPD group may have benefited from a practice effect in their performance. Conversely, the hemodialysis group showed no indication of benefiting from this possible practice effect. Their performance significantly declined over time. This would suggest hemodialysis patients’ ability to learn is negatively effected in comparison to CAPD patients. On measures of memory using the RAVLT, the hemodialysis group again showed their most impaired performance at the T3 test time on measures of both immediate and delayed recall. Some caution should be applied on the interpretation of these memory measure results. Given that test time comparisons within groups were nested in a within-subjects repeated-measures design, the decline in ability for the hemodialysis group to recall words may be caused by proactive interference on the memory measure. That is, with increase in exposure to words to recall over the 3 trials, recall becomes more difficult.
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
Fig 1. Mean ⴞ SE for fatigue scores, Stroop Effect, and total words generated in the immediate and delayed recall of the RAVLT for the hemodialysis and CAPD groups at 1, 24, and 67 hours after hemodialysis and after baseline for the CAPD group. An increase in scores for both fatigue and Stroop Effect reflects greater impairment, whereas an increase in number of words produced in both the immediate and delayed recall reflects less impairment.
709
710
Even given the possibility of the proactive interference confound, it is interesting to note that the CAPD group showed no indications of being affected by it. This suggests that the hemodialysis group is more sensitive to the effect, if it is occurring. Results show that the hemodialysis group generally performed worse at 67 hours compared with the 1- and 24-hour postdialysis test times. These findings are consistent with a previous investigation showing a relationship between time since dialysis and visual evoked potential latency; the shortest latencies typically occurred at the 24-hour postdialysis period and steadily increased during the 42-hour and 66-hour intervals after the hemodialysis session.6 The 24-hour postdialysis test time is the period typically used when examining cognitive functioning in hemodialysis patients. These results suggest that capturing cognitive performance for this group at only the 24-hour test time does not reflect the actual cognitive difficulties that hemodialysis patients may be experiencing at the end of a weekend without dialysis. In comparison, CAPD patients, who are undergoing continuous dialysis, showed an overall stable performance in cognitive tasks across all 3 test times. It would be interesting to compare our findings with those observed in a group of patients undergoing emerging forms of daily or nocturnal hemodialysis therapies. What is responsible for the deterioration in cognitive performance in hemodialysis patients with the passage of time from the previous dialysis treatment? Our data suggest that fatigue does not account for the deterioration in mental acuity in these patients. Rather, the fluctuation of psychometric measures in temporal correlation with hemodialysis treatments suggests that increasing accumulation of toxic uremic metabolites in the interdialytic period is involved in cognitive fluctuations. It should be noted that although vintage time between the hemodialysis and CAPD group was not statistically significant, the 30-month increase in vintage time for the hemodialysis group could have clinical significance. This increased vintage time may contribute to increased susceptibility to cognitive assault in the hemodi-
WILLIAMS ET AL
alysis group. Future investigations need to address this issue. We believe we have shown fluctuations in cognitive performance in hemodialysis patients, determined by the Stroop and RVALT psychometric measures. These results suggest that hemodialysis patients might be progressively compromised in the cognitive domains of attention and functions of memory as the interdialytic period proceeds. The question remains whether these deficiencies reach a point of clinical significance. The impact of these cognitive limitations on compliance with diet, fluid restriction, and medication use, as well as ability to make complex decisions needed to engage in adequate self-care behaviors, remains to be determined. Cognitive functioning occurs in the context of a whole human being, and the impact of cognitive deterioration or fluctuation will be unique to each individual who endures it. ACKNOWLEDGMENT The authors thank the dialysis patients who participated in this study; Dr Raymond G. Romanczyk and Dr Carolyn M. Pepper for support, guidance, and feedback throughout the project; and Purvi D. Shah and Myriam J. Sollman for help in data collection.
REFERENCES 1. Addison T: On the disorders of the brain connected with diseased kidneys. Guy’s Hosp Rep 4:1-7, 1839 2. Hart RP, Pederson JA, Czerwinski AW, Adams RL: Chronic renal failure, dialysis, and neuropsychological function. J Neuropsychol 5:301-312, 1983 3. Ryan JJ, Souheaver GT, DeWolfe AS: Halstead-Reitan test results in chronic hemodialysis. J Nerv Ment Dis 16:311314, 1981 4. Pliskin NH, Yurk HM, Ho LT, Umans JG: Neurocognitive function in chronic hemodialysis patients. Kidney Int 49:1435-1440, 1996 5. Ratner DP, Adams KM, Levin W, Rourke BP: Effects of hemodialysis on the cognitive and sensory-motor functioning of the adult chronic hemodialysis patient. J Behav Med 6:291-311, 1983 6. Lewis EG, O’Neil WM, Dustman RE, Beck EC: Temporal effects of hemodialysis on measures of neural efficiency. Kidney Int 17:357-363, 1980 7. Souheaver GT, Ryan JJ, DeWolfe AS: Neurological patterns in uremia. J Clin Psychol 38:490-496, 1982 8. Osberg JW, Meares GJ, McKee DC: Intellectual functioning in renal failure and chronic dialysis. J Chronic Disord 35:445-457, 1982 9. Wolcott DL, Wellisch DK, Marsh JT, Schaeffer J, Landsverk J, Nissenson AR: Relationship of dialysis modal-
TEMPORAL COGNITIVE EFFECTS OF DIALYSIS
ity and other factors to cognitive functions in chronic dialysis patients. Am J Kidney Dis 12:275-284, 1988 10. Lindsay RM, Leitch R, Heidenheim AP: The London daily/nocturnal hemodialysis study—Study design, morbidity, and mortality results. Am J Kidney Dis 42:S5-S12, 2003 (suppl 1) 11. Dodrill CB: A neuropsychological battery for epilepsy. Epilepsy 19:611-623, 1978 12. Sacks TL, Clark CR, Pols R, Geffen LB: Comparability and stability of performance on six alternate forms of the Dodrill-Stroop Color-Word Test. Clin Neuropsychol 5:220225, 1991 13. Rey A: L’examen Clinique en psychologie. Paris, France, Presses Universitaires de France, 1964 14. Geffen G, Moar KJ, O’Hanlon AP, Clark CR, Geffen LB: Performance measures of 16 to 86 year old males and
711
females on the auditory verbal learning test. Clin Neuropsychol 4:45-63, 1990 15. Kaufman AS, Kaufman NL: Kaufman Brief Intelligence Test Manual. Circle Pines, MN, American Guidance Service, 1990 16. Donovick PJ, Burright RG, Burg JS: A screen for IQ in six diverse populations. J Clin Psychol Med Settings 3:131-140, 1996 17. Eisenstein N, Engelhart CI: Comparison of the KBIT with short forms of the WAIS-R in a neuropsychological population. Psychol Assess 9:57-62, 1997 18. Beck AT, Steer RA, Brown GK: Beck Depression Inventory Manual (ed 2). San Antonio, TX, Psychological Corp, 1996 19. SPSS Inc: Statistical Package for the Social Sciences, version 10.0. Chicago, IL, SSPS Inc.