31. Tentori F, Hunt WC, Rohrscheib M et al. Which targets in clinical practice 40.

40. Palmer SC, Hayen A, Macaskill P et al. Serum levels of phosphorus, para-
guidelines are associated with improved survival in a large dialysis organ-
ization? J Am Soc Nephrol 2007; 18: 2377–2384
32. Voormolen N, Noordzij M, Grootendorst DC et al. High plasma phosphate
as a risk factor for decline in renal function and mortality in pre-dialysis
patients. Nephrol Dial Transplant 2007; 22: 2909–2916
33. Covic A, Kothawala P, Bernal M et al. Systematic review of the evidence
underlying the association between mineral metabolism disturbances and
risk of all-cause mortality, cardiovascular mortality and cardiovascular events
in chronic kidney disease. Nephrol Dial Transplant 2009; 24: 1506–1523
34. Bradbury BD, Fissell RB, Albert JM et al. Predictors of early mortality
among incident US hemodialysis patients in the Dialysis Outcomes and
Practice Patterns Study (DOPPS). Clin J Am Soc Nephrol 2007; 2: 89–99
35. Noordzij M, Korevaar JC, Dekker FW et al. Mineral metabolism and mor-
tality in dialysis patients: a reassessment of the K/DOQI guideline. Blood
Purif 2008; 26: 231–237
36. Wald R, Sarnak MJ, Tighiouart H et al. Disordered mineral metabolism in
hemodialysis patients: an analysis of cumulative effects in the Hemodialysis
(HEMO) Study. Am J Kidney Dis 2008; 52: 531–540
37. Lacson E, Jr, Wang W, Hakim RM et al. Associates of mortality and hospi-
talization in hemodialysis: potentially actionable laboratory variables and
vascular access. Am J Kidney Dis 2009; 53: 79–90
38. Robinson BM, Port FK. International hemodialysis patient outcomes com-
parisons revisited: the role of practice patterns and other factors. Clin J Am
Soc Nephrol 2009; 4 Suppl 1: S12–S17
39. Larsson TE, Olauson H, Hagstrom E et al. Conjoint effects of serum calcium
and phosphate on risk of total, cardiovascular, and noncardiovascular mor-
tality in the community. Arterioscler Thromb Vasc Biol 2010; 30: 333–339
thyroid hormone, and calcium and risks of death and cardiovascular dis-
ease in individuals with chronic kidney disease: a systematic review and
meta-analysis. JAMA 2011; 305: 1119–1127
41. Binder H, Sauerbrei W, Royston P. Comparison between splines and frac-
tional polynomials for multivariable model building with continuous cov-
ariates: a simulation study with continuous response. Stat Med 2013; 32:
2262–2277
42. Malloy EJ, Spiegelman D, Eisen EA. Comparing measures of model selec-
tion for penalized splines in Cox models. Comput Stat Data Anal 2009; 53:
2605–2616
43. Govindarajulu US, Malloy EJ, Ganguli B et al. The comparison of alterna-
tive smoothing methods for fitting non-linear exposure-response rela-
tionships with Cox models in a simulation study. Int J Biostat 2009; 5:
Article 2. doi: 10.2202/1557-4679.1104
44. Strasak AM, Umlauf N, Pfeiffer RM et al. Comparing penalized splines and
fractional polynomials for flexible modelling of the effects of continuous
predictor variables. Comput Stat Data Anal 2011; 55: 1540–1551
45. Tentori F, Wang M, Bieber BA et al. Recent changes in therapeutic ap-
proaches and association with outcomes among patients with secondary
hyperparathyroidism on chronic hemodialysis: the DOPPS study. Clin J
Am Soc Nephrol 2015; 10: 98–109
46. Torres A, Lorenzo V, Hernandez D et al. Bone disease in predialysis, hemo-
dialysis, and CAPD patients: evidence of a better bone response to PTH.
Kidney Int 1995; 47: 1434–1442
Received for publication: 18.12.2014; Accepted in revised form: 17.3.2015

ORIGINAL ARTICLE
Nephrol Dial Transplant (2015) 30: 1551–1559
doi: 10.1093/ndt/gfv213
Advance Access publication 11 June 2015

Effect of a single dialysis session on cognitive function in CKD5D

patients: a prospective clinical study

Sabrina M. Schneider1,2, Anne K. Malecki2,3, Katrin Müller3, Robby Schönfeld3, Matthias Girndt4,
Peter Mohr4, Marcus Hiss2, Heike Kielstein5, Kristin Jäger5 and Jan T. Kielstein2
1
Department of Neurology, Philipps-University Marburg, Marburg, Germany, 2Department of Nephrology and Hypertension Hannover Medical
School, Hannover, Germany, 3Department of Psychology, Martin-Luther University Halle-Wittenberg, Halle (Saale), Germany, 4Department of
Internal Medicine II, Martin-Luther University Halle-Wittenberg, Halle (Saale), Germany and 5Department of Anatomy, Martin-Luther
University Halle-Wittenberg, Halle (Saale), Germany

Correspondence and offprint requests to: Jan T. Kielstein; E-mail: kielstein@yahoo.com

dialysis (CKD5D)]. Aside from structural permanent damage,

A B S T R AC T
Background. Cognitive function declines in parallel to the
decrease in glomerular filtration rate, best epitomized by the
markedly reduced cerebral performance in patients undergoing
maintenance haemodialysis [chronic kidney disease stage 5
there seems to be a reversible part of low cognitive performance.
The potential effect of a single dialysis session on cognitive
function remains still elusive. The aim of the study was to assess
cognitive function using a widespread test battery and avoiding
excluding effects of circadian variations.

© The Author 2015. Published by Oxford University Press 1551

on behalf
Downloaded from of ERA-EDTA. All rights reserved.
https://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
by guest
on 03 January 2018
 Methods. Twenty-eight medically stable CKD5D patients
(age: 54.9 ± 13.2 years, dialysis vintage: 46.2 ± 51.0 month)
at two tertiary care centres with outpatient dialysis units
were enrolled. Cognitive testing was always performed twice
within 24 h, 1 h prior to haemodialysis (T1pre-dialysis) as well
as 19 h after the end of dialysis (T2post-dialysis) including as-
sessment of memory, attention and concentration, executive
functioning, word fluency and psychomotor speed by using
a well-validated neuropsychological test battery. Patients
were randomized into two groups. One group was examined
before (T1pre-dialysis) and after (T2post-dialysis) Dialysis Session
1. The other group was first examined the day after Dialysis
Session 1 (T2post-dialysis) and then before Dialysis Session 2
(T1pre-dialysis) in order to exclude potential learning effects.
Twenty age-matched subjects with normal excretory renal
function were used for comparison.
Results. Neuropsychological testing found that the CKD5D
performed significantly worse on measures of alertness, atten-
tion, working memory, logical and visual memory, word flu-
ency and executive functions compared with non-CKD
subjects. No differences in short-term memory, selective atten-
tion, as well as problem-solving and planning were found be-
tween CKD5D patients and non-CKD subjects. A single
haemodialysis session led to a significant improvement in logic-
al (Rivermead Behaviour Memory Test story: P < 0.001) and
ORIGINAL ARTICLE
visual memories [Rey-Osterrieth Complex Figure Test
(RCFT) memory quotient: P < 0.001], psychomotor speed
[Trail Making Test (TMT) B: P = 0.020], activity planning (ex-
ecutive functions) (RCFT copy/points deduction: P < 0.001)
and concentration (TMT A: P < 0.001).
Conclusion. Our data demonstrate improvements in memory
functions, executive functions and psychomotor abilities after a
single dialysis session, pointing to a reversible component of
low cognitive performance in CKD5D.
Keywords: neuropsychological tests, uraemic toxins
INTRODUCTION
Treatment of uraemia by chronic haemodialysis can lead to a
marked improvement in neurocognitive function as already
reported by Scriber et al. in 1960 [1]. The first two chronic
haemodialysis [chronic kidney disease stage 5 dialysis
(CKD5D)] patients who showed ‘increased fatigability, muscle
cramps, irritability and lethargy’ before the initiation of treat-
ment improved to a point that ‘neither patient has yet shown
the relentless loss of weight and the mental deterioration’ [1].
Since this first description, a burgeoning body of evidence has
accumulated that cognitive function is indeed severely impaired
in patients with CKD5D as recently reviewed by Bugnicourt
et al. [2]. Rare cases of severe uraemia, which is fortunately
only rarely seen in industrialized countries, vividly illustrate
the reversibility of decreased cognitive performance by dialysis
[3]. In the setting of maintenance dialysis, the effect of a single
haemodialysis on cognitive function is controversially dis-
cussed. One study could not find any effect of a single dialysis
on an insensitive test like mini mental state examination [4].
Downloaded from1552
https://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
by guest
on 03 January 2018
Other authors reported that the removal of uraemic toxins by
haemodialysis leads to an improvement in cognitive processing,
but mainly if the severely impaired cognitive function 1 h after
dialysis is taken as a comparator [5]. Also, the potential effect of
an increase in dialysis intensity on cognitive function is contro-
versially discussed. While one pilot study in 12 patients showed
improved general cognitive efficiency 6 months after switching
from thrice weekly to nocturnal haemodialysis [6], the results
from the frequent haemodialysis network nocturnal trial
showed no effect of an increased dialysis intensity on cognitive
performance, measured by the Trail Making Test (TMT) B only
[7]. Hence, the aim of our study was to investigate the effect of
a single dialysis session on a large variety of neurocognitive
functions in CKD5D patients using a large, state-of-the-art
test battery instead of simple screening tests.
METHODS
Study participants
The study was approved by the local Ethics Committee (ref-
erence # 2012-30). All patients provided informed consent.
Patients were treated at the dialysis units of the two participat-
ing tertiary care centres in Hannover and Halle. Eligible
CKD5D patients were approached if they met the following cri-
teria: (i) aged 18 years or more, (ii) no clinical history of neuro-
logical disorders (e.g. stroke and Parkinson’s disease) and (iii)
being fluent in spoken German. Of the 43 approached patients,
31 agreed to take part in the study. Out of those, three declined
to perform the second testing session. All dialysis patients (n =
28) were dialysed thrice weekly. The non-CKD group consisted
of 20 subjects recruited from the community, with normal renal
function and self-reported health, confirmed by laboratory data
obtained from the primary care physician within 12 months of
testing, who also met the same inclusion criteria.
Measures
Sociodemographic characteristics and clinical measures.
Demographic, clinical data and medication details were
collected from patients’ medical records as well as during
the assessment. Furthermore, routine clinical chemistry
and complete blood count results were recorded on the day
of T1pre-dialysis.
Neuropsychological assessments. To assess various cogni-
tive domains, we designed a neuropsychological test battery
measuring memory, attention, executive functions and psycho-
motor functions. All participants were tested twice. Cognitive
testing was always performed twice within 24 h, 1 h prior to
haemodialysis (T1pre-dialysis) as well as 19 h after the end of dia-
lysis (T2post-dialysis) including assessment of memory, attention
and concentration, executive functioning, word fluency and
psychomotor speed by using a well-validated neuropsycho-
logical test battery. Patients were randomized into two groups.
One group was examined before (T1pre-dialysis) and after
(T2post-dialysis) Dialysis Session 1. The other group was first ex-
amined the day after Dialysis Session 1 (T2post-dialysis) and then
before Dialysis Session 2 (T1pre-dialysis) in order to exclude
S.M. Schneider et al.
 potential learning effects that have been well knwon [8, 9].
(Supplementary data, Table S1). Tests were performed in a sep-
arate room and quiet environment. All assessments were con-
ducted by researchers who had received training on
administering these tests. With all participants, we used the
standardized and validated test versions in German. The neuro-
psychological test battery is summarized in Table 1. A more ex-
tensive description can be found in Supplementary data, File 1.
Statistical analyses
Continuous variables were characterized using mean ± SD
or percentages and compared using t-tests. Independent sam-
ple t-tests were used to evaluate the differences between
non-CKD subjects and CKD5D patients. Frequencies are pre-
sented for categorical variables. Pearson’s chi-square tests
were performed to compare sociodemographic characteristics
between CKD5D patients and non-CKD cohort. For each
cognitive test, raw scores were converted to age- and
education-standardized T-scores (the equivalent to Z-scores
in samples sizes of <30 with an arbitrarily assigned mean
and SD) obtained from the test manuals, which were used
to evaluate performance. For Rivermead Behaviour Memory
Test (RBMT) story, Rey-Osterrieth Complex Figure Test
(RCFT) points deduction T-scores were not available, so we
used raw scores for statistical analyses. A subject’s individual
of variance. The statistical analyses were performed using
SPSS version 17.0 for Windows. All statistical tests were two
sided. Differences were considered statistically significant at
P < 0.05.
R E S U LT S
Subject characteristics
Twenty-eight CKD5D patients were recruited for the study.
The mean age was 54.9 ± 13.1 years (range = 27–79). The non-
CKD group of 20 participants had a mean age of 54.6 ± 14.0
years (range = 32–80). Eleven (39.3%) of the CKD5D patients
and 13 (65%) non-CKD subjects were female. Socio-
demographic and medical characteristics of the study partici-
pants are shown in Tables 2 and 3.
Baseline test results of CKD5D patients compared with
healthy non-CKDs
Baseline scores (T1pre-dialysis) identified significant differ-
ences between non-CKD subjects and CKD5D patients
(Table 4a). Patients showed a poorer performance in tests re-
garding attention, long-term memory and working memory
functions.
T-scores visualizing the time needed to complete TMT B

ORIGINAL ARTICLE
test performance was considered impaired if it was >1 SD (≙ were significantly higher in the CKD5D group, which reflects
T-scores <40) below the mean of the norms. The influence of a poorer performance. For TMTs A and B, we found 18
gender on the test results was calculated by using an analysis CKD5D patients (64.3%) for each of the tests, who received a

Table 1. Neuropsychological assessments used in this study

Test Description Task Parameter

Attention TAP [10] Subtests: –Reaction (pushing a button) to a –Reaction time (in ms)
–Alertness→measures every day attention visual stimulus
–Go/NoGo→measures ability to suppress undesired
responses
–Computer-based test battery
TMT A [11] –Measures attention, visual scanning and processing –Connections of randomly arranged –Processing time (in s)
speed numbers from 1 to 25
Memory Wechsler Memory Subtests: –Memorization of a set of numbers –Number of correct
Scale-Revised (WMS-R) –Digit span forward→measures short-term memory in the correct serial order sets of numbers
[12] –Digit span backward→measures working memory
RBMT [13] –Subtest: story measures logical memory –Retrieval of a short story that has –Number of correct
–Alternate forms for T2post-dialysis been read out to the subjects facts
–Immediate and delayed recall
(30 min)
RCFT [14] –Measures visual memory (and executive function) –Copy and recall of a complex figure –Number of correct
–Quality of drawing = RCFT copy (delayed recall after 3 and 30 min) items of the figure
–Visual memory function = RCFT memory quotient
–Logical activity planning = RCFT points deduction
–Alternate forms for T2post-dialysis
Executive TMT B [11] –Measures mental flexibility and psychomotor speed –Connecting numbers and letters in –Processing time (in s)
function an alternate sequence
BADS [13] –Subtests key search and zoo map measure the –Key search: creating logical –Points for logical
capability to plan behaviour and logical activity strategies to search an imaginative strategies
planning key on a huge meadow
–Zoo map: finding the way through a
labyrinth
RWT [15] –Measures verbal fluency –Retrieval of words from a certain –Number of words in
–Categories: semantic, divergent semantic, lexical and category within 1 min 1 min
divergent lexical

N e uhttps://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
Downloaded from rocognitive improvement by a single dialys is 1553
by guest
on 03 January 2018
 Table 2. Characteristics of study sample: Part 1
CKD5D patients (n = 28) Non-CKD subjects (n = 20) P

M ± SD M ± SD
Age (years) (M ± SD) 54.9 ± 13.2 54.6 ± 14.0 0.941
Gender (F/M) 11/17 13/7 0.079
Dialysis vintage (months) 46.2 ± 51.0
Education 0.040
9th grade (%) 35.7 5
10th grade (%) 35.7 60
>12th grade (%) 28.6 35
Laboratory results Reference | KDIGOa
Haemoglobin (g/dL) 10.02 ± 2.02 15.02 ± 1.79 12.0–16.0 | <10.0 to <11.5a
Leukocytes (10³/μL) 7.75 ± 4.21 7.35 ± 3.88 4.4–11.3
Albumin (g/dL) 33.00 ± 5.83 35–52
Ferritin (μg/L) 774 ± 635 27–365 | >500
MCV (fl) 91.80 ± 8.50 83.27 ± 5.79 80–100
MCHC (g/dL) 31.84 ± 4.13 31.37 ± 3.75 31–37
PTH (ng/L) 425.15 ± 587.22 15–65 | 150–300a
TSH (mU/L) 2.67 ± 4.19 0.27–4.2
S_Potassium (mmol/L) 5.27 ± 0.84 4.17 ± 1.25 3.6–5.4
S_Sodium (mmol/L) 138.1 ± 4.1 135–145
S_Calcium (mmol/L) 2.22 ± 0.21 2.15–2.60
Phosphate (mmol/L) 1.50 ± 0.47 0.83–1.67
Creatinine (μmol/L) 640.6 ± 300.1 82.7 59–104
eGFR mL/min CKD-EPI 98.5 ± 17.3 >60
Urea (mmol/L) 16.63 ± 9.04 2.8–7.2
S_Uric acid (μmol/L) 351.3 ± 116.2 200–420
S_GOT (U/L) 20.61 ± 10.51 <35
S_GPT (U/L) 22.77 ± 32.22 <45
ORIGINAL ARTICLE

GGT (U/L) 58.49 ± 59.62 <55

S-Protein 61.39 ± 9.14 65–80
Iron (μmol/L) 8.10 ± 3.81 11–25
Transferrin satur. (%) 15.47 15–45
a
KDIGO (Kidney Disease Global Outcome Initiative) recommendation if applicable.
MCV, mean corpuscular/cell volume; MCHC, mean corpuscular/cellular hemoglobin concentration; PTH, Parathyroid hormone; TSH, Thyroid stimulating hormone.

score of >1 SD below the mean, meeting the definition for cog-
nitive impairment. In the non-CKD cohort, there were only
three (15%) participants (TMT A) and four (20%) participants
(TMT B) with impairment in both tests. Other significant dif-
ferences between the groups are discovered on Tests of Atten-
tional Performance (TAP) (attention). Fourteen (50%) CKD5D
patients and four (20%) non-CKD subjects were cognitively
impaired. While 14 (50%) CKD5D patients had an impaired
performance on digit span backwards (working memory),
there was only 1 (5%) individual in the non-CKD cohort. Mea-
sures of short-term memory (forward digit span) and selective
attention (TAP Go/NoGo) did not differ between groups (Fig-
ure 1). Comparisons between non-CKD subjects and CKD5D
patients also revealed significant differences regarding particu-
lar results in executive performance. Regensburg Word Fluency
Test (RWT) produced a consistent effect between the groups
regarding every subtest T-score. The strongest impairment
was found in the subtest divergent lexical with 16 (57%)
CKD5D patients, who received a score clearly more than 1
SD below the mean, compared with 5 (25%) of the non-CKD
group. Subtests from Behavioural Assessment of Dysexecutive
Syndrome (BADS) inventory did not differentiate between the
groups. The executive assessment being included in the RCFT
could mark a significant difference between our two groups.
The results of the independent t-test are presented in Table 4a.
Pre-dialysis (T1pre-dialysis)/and post-dialysis
(T2post-dialysis) tests
Assessing the effect of a single dialysis session by a pre–post-
comparison of test results showed an improvement in attention
( processing speed/visual scanning) and long-term memory
functions (Table 4b). This was best epitomized by criteria for
mild cognitive impairment in the TMT A test. While at pre-
dialysis (T1pre-dialysis) 64% fulfilled the criteria for cognitive im-
pairment, this was only true for 25% of CKD5D patients at
T2post-dialysis, i.e. the day after dialysis. Other functions of atten-
tion (like alertness and selective attention), short-term and
working memory did, however, not differ between T1pre-dialysis
and T2post-dialysis (Figure 2).
Cognitive shifting (TMT B) improved significantly after a
single dialysis session regarding normed T-scores. Interestingly,
the same was true for the non-CKD group. Results of the TMT
B were the only test results that also improved in the non-CKD
group, pointing to a potential learning effect in this test. To ac-
count for this potential learning effect, the study design of the
CKD5D patients was tailored accordingly (see above). While
64% of patients fulfilled the impaired performance before dia-
lysis (T1pre-dialysis), this decreased to 46% at T2post-dialysis, i.e. the
day after dialysis. In contrast to attention and cognitive shifting,
neither call of semantic and lexical words in an alternating way
nor productive thinking (RWT) improved after dialysis. The

Downloaded from1554
https://academic.oup.com/ndt/article-abstract/30/9/1551/2460056 S.M. Schneider et al.
by guest
on 03 January 2018
 Table 3. Characteristics of study sample—Part 2
DISCUSSION
CKD5D patients
(n = 28)
The aim of the study was a detailed examination of the cognitive
M ± SD performance level of adults with CKD5D. The neuropsychologic-
Weight (kg) 75.6 ± 19.1 al test battery consisted of tests measuring memory, attention and
Ultrafiltration (mL) 1608 ± 1346 executive functions. Compared with the non-CKD group, we
Dialysis time (h) 4.07 ± 0.51 found a poorer performance level in CKD5D patients. More spe-
Dialysis vintage (months) 46.2 ± 51.0
Heart rate (bpm) 81.9 ± 15.7 cifically, it should be noted that significant differences in perform-
Diastolic blood pressure (mmHg) 80.3 ± 14.4 ance regarding logical and visual long-term memories (tested
Systolic blood pressure (mmHg) 130.0 ± 20.8 with RBMT story or RCFT), working memory (digit span back-
Kt/Vurea 1.34 ± 0.12 wards), processing speed (TMT A) and alertness (TAP), as well as
Pre-dialysis HCO3/dialysis bath 22.2 ± 2.7/33.9 ± 2.8
HCO3
executive functions such as cognitive shifting (TMT B), activity
Pre-dialysis K+/dialysis bath K+ 5.27 ± 0.84/3.77 ± 0.37 planning (RCFT points deduction) and verbal fluency (RWT)
N %
could be detected between the two study groups at T1pre-dialysis.
Medication Besides the cross-sectional comparison between CKD5D pa-
AT1-blocker 14 50 tients before dialysis and non-CKD subjects, we aimed to iden-
Beta-blocker 17 60.7 tify the effect of a single dialysis session on cognitive functions.
EPO 16 57.1 After dialysis session (T2post-dialysis), patients showed an im-
Diuretics 11 39.3
Vitamin D 13 46.4
proved performance in logical and visual long-term memories
Statins 4 14.3 (RBMT story and RCFT), processing speed/visual scanning
PPI 20 71.4 (TMT A), cognitive shifting (TMT B) and activity planning
Calcium antagonists 12 42.9 (RCFT points deduction) when compared with the test results
Phosphate binders w/Ca 5 17.9 before a dialysis session (T1pre-dialysis). There was neither a per-
Phosphate binders w/o Ca 10 35.7
Intravenous iron 12 42.9
formance change regarding short-term and working memories

Benzodiazepines 4 14.3
Antidepressants 6 21.4
Total number of tablets per day 28 9.29 ± 2.84
slight improvement of T-scores regarding working memory was
not significant comparing normed T-scores within both
groups.
Capability to plan behaviour and to refer to logical activity
planning, examined by the subtests key search and zoo map,
was not influenced by dialysis. Patients’ ability to visual plan-
ning was assessed by the quality of the complex figure they had
to copy. Interestingly, the objective scoring improved signifi-
cantly comparing pre-dialysis with post-dialysis (RCFT copy
and RCFT points deduction). All patients showed an average
to above average copy performance at T2post-dialysis. Consider-
ing strong ceiling effects for these test results, a chi-square test
was conducted in order to reveal correlation between quality of
copying and group membership. Chi-square test shows a rela-
tionship between group membership and test results during
the first point of testing [RCFT (copy): χ 2 = −3.15, P = 0.003].
For the second point of testing, this correlation is much weaker
[RCFT (copy): χ 2 = −1.88, P = 0.07], supporting the notion
that executive functioning in CKD5D is improved after a single
dialysis session (Figure 3). Similar results could be found in the
deduction of points during figure copying [T1pre-dialysis: RCFT
( points deduction): χ 2 = 2.37, P = 0.02/T2post-dialysis: RCFT
( points deduction): χ 2 = 1.59, P = 0.118]. All results of the de-
pendent t-tests are summarized in Table 4b. Gender did not
influence the increase in performance comparing pre- and
post-dialysis cognitive function in the CKD5D patients (data
not shown).
ORIGINAL ARTICLE
(forward and backward digit span) and attention (TAP) nor in
verbal fluency and planning behaviour (BADS). Despite the im-
provements in some cognitive areas, it should be noted that the
level of cognitive function in CKD5D was still below the per-
formance level of the non-CKD group. With the exception of
the TMT B, there was no difference between the two time points
of testing (which were 24 h apart) in non-CKD subjects. The
results in TMT B did improve significantly pointing to a pos-
sible learning effect, underlining the importance of the study
design we did choose for the CKD5D patients to minimize
the impact of learning effects.
CKD5D patients showed a stable performance on the for-
ward digit span, which is thought to reflect unimpaired short-
term memory functions. The available alertness of dialysis pa-
tients lasts to keep up with those short-term memory functions.
Therefore, the performance in these tests remains on an average
level. However, learning and memorizing information over a
longer period of time require more than short activation of at-
tention. Higher brain actions are involved, which are apparently
reduced in patients with CKD. Some specific abilities are more
vulnerable to moderate CKD than others. Our findings support
the hypothesis that cognitive impairment in CKD5D patients is
reversible by haemodialysis. Patients are more likely able to
memorize, concentrate and manage situations a day after
dialysis.
A chronic disease, like CKD, requires a patient’s adherence to a
complex pharmacological regimen and to specific dietary restric-
tions. A lower cognitive function level is a major barrier as it im-
pairs self-management, which is even more important in patients
undergoing dialysis treatment. In CKD5D, errors in medication
and diet due to decreased cognitive performance can have dele-
terious consequences. As pointed out by Elias et al., the measures
of everyday cognitive abilities relevant to patient understanding of

N e uhttps://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
Downloaded from rocognitive improvement by a single dialys is 1555
by guest
on 03 January 2018
 Table 4. Mean and SD of T scores before dialysis (T1pre-dialysis) and the morning after dialysis (T2post-dialysis)

Group a) before dialysis (T1pre-dialysis) b) follow-up—the morning after

dialysis (T2post-dialysis)

M ± SD P valuea M ± SD P valueb
Memory
RBMT—storyc CKD5D 1.71 ± 0.50 0.004 1.93 ± 0.18 <0.001
Non-CKD 1.96 ± 0.11 1.90 ± 0.35 0.210
RCFT—MQ CKD5D 49.82 ± 7.80 0.001 57.93 ± 4.87 <0.001
Non-CKD 58.00 ± 8.09 60.10 ± 6.95 0.131
Digit span forward CKD5D 53.39 ± 10.14 0.474 55.57 ± 12.63 0.141
Non-CKD 55.55 ± 10.28 56.50 ± 9.70 0.712
Digit span backward CKD5D 45.32 ± 12.21 0.002 46.68 ± 11.64 0.540
Non-CKD 55.55 ± 8.15 53.80 ± 9.84 0.229
Attention
TMT A CKD5D 38.89 ± 5.94 0.001 43.86 ± 8.13 <0.001
Non-CKD 50.45 ± 8.00 52.20 ± 8.33 0.090
TAP alertness
With signal CKD5D 40.50 ± 7.21 0.001 41.57 ± 8.01 0.307
Non-CKD 48.05 ± 8.19 50.25 ± 7.30 0.214
Without signal CKD5D 40.68 ± 8.29 <0.001 42.96 ± 9.51 0.083
Non-CKD 50.55 ± 7.86 50.80 ± 7.63 0.868
TAP Go/NoGo CKD5D 47.71 ± 12.82 0.111 49.89 ± 11.78 0.225
Non-CKD 53.05 ± 8.48 52.85 ± 8.02 0.851
Executive functions
TMT B CKD5D 38.79 ± 7.59 <0.001 41.54 ± 9.25 0.020
Non-CKD 49.35 ± 9.52 55.30 ± 10.44 0.001
RCFT copy CKD5D 58.04 ± 13.60 0.003 66.25 ± 8.99 0.001
Non-CKD 70 ± 12.09 73.90 ± 8.64 0.175
RCFT points deductionc CKD5D 2 ± 2.48 0.022 0.25 ± 0.70 <0.001
ORIGINAL ARTICLE

Non-CKD 0.55 ± 1.36 0.00 ± 0.00 0.086

BADS key searchc CKD5D 12.32 ± 4.09 0.599 12.57 ± 3.70 0.588
Non-CKD 12.90 ± 3.16 13.50 ± 2.82 0.280
BADS zoo mapc CKD5D 14.14 ± 1.88 0.991 14.71 ± 1.30 0.062
Non-CKD 14.15 ± 2.28 14.75 ± 2.92 0.209
RWT lexical CKD5D 41.29 ± 7.88 0.005 43.57 ± 10.68 0.133
Non-CKD 48.20 ± 8.33 48.65 ± 10.54 0.767
RWT divergent lexical CKD5D 38.61 ± 7.85 0.001 41.21 ± 12.31 0.119
Non-CKD 47.50 ± 9.78 49.20 ± 11.80 0.427
RWT semantic CKD5D 43.46 ± 11.40 0.005 47.36 ± 10.02 0.030
Non-CKD 52.70 ± 9.42 54.10 ± 8.73 0.504
RWT divergent semantic CKD5D 44.25 ± 9.22 <0.001 47.11 ± 8.43 0.122
Non-CKD 54.15 ± 7.65 55.20 ± 8.29 0.581
a
P value of independent t-tests between the groups (patient group versus non-CKD).
b
P value of dependent t-tests within the groups (Tpre versus Tpost).
c
absolute numbers (raw scores)

the disease and treatments need to be outlined, studied and im-
plemented [16]. Our data indicate that the best time for doctor–
patient communication is the day after a dialysis session. Patients
seem to be more able to listen carefully, comprehend, memorize
and to follow instructions of nurses, dieticians and physicians. In
real life, this time point, which would require an additional trip to
the dialysis unit, seems difficult to implement.
Our results support earlier studies that have found an asso-
ciation between decreased kidney function and lower cognitive
performance over a whole range of renal function [17]. More-
over, only a few studies have examined the influence of a single
dialysis session on cognitive functioning. Among these, Har-
ciarek et al. [18] and Lux et al. [19] have been the only two stud-
ies, which have tested specific cognitive areas. Those studies
were different from our study in the following respects. The
sample size in the study of Lux was quite small, only including
12 participants. Harciarek and colleagues tested 20 CKD5D
patients, using a state-of-the-art test battery to examine learn-
ing and memory functions, yet in this study detailed tests of at-
tention were not performed.
All studies point to a cognitive improvement after dialysis, or
more specifically both studies as well as our study have found an
increase in psychomotor performance, executive functions,
learning and memory.
We wish to point out several important limitations of our
study. First, the recruitment of patients underlies a positive se-
lection bias. Clinically stable patients voluntarily participated in
the study. Some of the envisaged patients did not consider
themselves ‘good’ enough to participate because of the fear of
performing badly. Furthermore, patients with severe cognitive
impairment were less likely to participate. These facts lead to
the assumption of an underestimation of the true severity
of cognitive decline in CKD5D patients. Second, the generaliz-
ability of our study is limited by the small sample size. Although

Downloaded from1556
https://academic.oup.com/ndt/article-abstract/30/9/1551/2460056 S.M. Schneider et al.
by guest
on 03 January 2018
 F I G U R E 1 : Memory tested pre- (T1pre-dialysis) and post-dialysis (T2post-dialysis) with the following tests: (A) forward digit span, (B) backward digit

ORIGINAL ARTICLE
span, (C) RMBT and (D) RCFT. Data are visualized as box and whisker plots (horizontal bars indicate median values).

F I G U R E 2 : Attention tested pre- (T1pre-dialysis) and post-dialysis (T2post-dialysis) with the following tests: (A) TMT A, (B) TAP Go/NoGo, (C) TAP
Alertness with signal and (D) TAP Alertness without signal. Data are visualized as box and whisker plots (horizontal bars indicate median values).

N e uhttps://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
Downloaded from rocognitive improvement by a single dialys is 1557
by guest
on 03 January 2018
 ORIGINAL ARTICLE

F I G U R E 3 : Executive function tested pre- (T1pre-dialysis) and post-dialysis (T2post-dialysis) with the following tests: (A) TMT A, (B) RCFT copy, (C)
RCFT points deduction, (D) BADS—key search, (E) BADS—zoo map and (F–I) RWT. Data are visualized as box and whisker plots (horizontal
bars indicate median values).

patients underwent an extensive test battery, we limited the tests
for specific cognitive areas to executive functions like the BADS
(key search and zoo map) for performance of complex executive
functions. Third, controls exhibited a higher education than
CKD5D patients, which had, however, no effect on the evalu-
ation of the influence of a single dialysis on the applied test bat-
tery, which was the main study aim.
Does dialysis result in transitory improvement or longer-term
improvement? If so, is this related to the dose of renal replacement
therapy? Are changes in blood pressure (during dialysis) related to
cognitive function decline, as described for patients with uncon-
trolled hypertension [20]? Can early intervention in patients with
mild to moderate CKD reduce the decline in cognitive perform-
ance? Should a rapid decline in cognitive performance prompt the
start of renal replacement therapy? These remain open questions
that urgently need to be studied.
cognitive impairment could be detected, which means that the
results were at least 1 SD below the mean for age- and education-
matched controls. In addition, this study showed that a single
dialysis session improves several aspects of cognitive function,
pointing to a reversible component of decreased cognitive per-
formance in CKD5D. These data support that the knowledge
of cognitive function level and the temporary improvement of
cognitive functions are exceedingly important for the progress
of the disease. Treatment of cognitive deficits should start early
in the beginning of a kidney disease to ensure the best psycho-
logical and nephrological care geared to specific patient needs.
FUNDING

The study was funded by intramural support.

CONCLUSION

In summary, our findings demonstrate that among adults with

CKD5D, lower cognitive performance is common. Compared S U P P L E M E N TA R Y D ATA
with non-renal subjects, we found a decreased performance in
long-term memory. Even more severely affected were executive Supplementary data are available online at http://ndt.oxford
functions, as well as attention. In some of the tests, even a mild journals.org.

Downloaded from1558
https://academic.oup.com/ndt/article-abstract/30/9/1551/2460056 S.M. Schneider et al.
by guest
on 03 January 2018

C O N F L I C T O F I N T E R E S T S TAT E M E N T
None declared.
AUTHORS’ CONTRIBUTIONS
S.S., A.M. and K.M. conducted the neurocognitive testing. J.T.
K. was the treating physician of the patient reported. S.S., A.M.,
K.M., H.K., J.T.K. and R.S. evaluated the test results. All of the
authors have participated in the discussion and in writing of the
submitted manuscript. All authors read and approved the final
manuscript.
(See related article by Elias et al. Improved cognitive perform-
ance after a single dialysis session: where do we go from here?
Nephrol Dial Transplant 2015; 30: 1414–1417.)
REFERENCES
1. Scribner BH, Buri R, Caner JE et al. The treatment of chronic uremia by
means of intermittent hemodialysis: a preliminary report. Trans Am Soc
Artif Intern Organs 1960; 6: 114–122
2. Bugnicourt JM, Godefroy O, Chillon JM et al. Cognitive disorders and de-
mentia in CKD: the neglected kidney-brain axis. J Am Soc Nephrol 2013;
24: 353–363
3. Schneider S, Malecki AK, Boenisch O et al. Cognitive function at 2443
mumol/l creatinine. BMC Nephrol 2012; 13: 86
4. Dahbour SS, Wahbeh AM, Hamdan MZ. Mini mental status examination
(MMSE) in stable chronic renal failure patients on hemodialysis: the effects
of hemodialysis on the MMSE score. A prospective study. Hemodial Int
2009; 13: 80–85
5. Williams MA, Sklar AH, Burright RG et al. Temporal effects of dialysis on cog-
nitive functioning in patients with ESRD. Am J Kidney Dis 2004; 43: 705–711
6. Jassal SV, Devins GM, Chan CT et al. Improvements in cognition in pa-
tients converting from thrice weekly hemodialysis to nocturnal hemodialy-
sis: a longitudinal pilot study. Kidney Int 2006; 70: 956–962
N e uhttps://academic.oup.com/ndt/article-abstract/30/9/1551/2460056
Downloaded from rocognitive improvement by a single dialys
by guest
on 03 January 2018
7. Rocco MV, Lockridge RS, Jr, Beck GJ et al. The effects of frequent nocturnal
home hemodialysis: the Frequent Hemodialysis Network Nocturnal Trial.
Kidney Int 2011; 80: 1080–1091
8. Falleti MG, Maruff P, Collie A et al. Practice effects associated with the
repeated assessment of cognitive function using the CogState battery at
10-minute, one week and one month test-retest intervals. J Clin Exp
Neuropsychol 2006; 28: 1095–1112
9. Pyc MA, Rawson KA. Why testing improves memory: mediator effective-
ness hypothesis. Science 2010; 330: 335
10. Zimmermann P, Sprengelmeyer R, Fimm B et al. Cognitive slowing in
decision tasks in early and advanced Parkinson’s disease. Brain Cogn
1992; 18: 60–69
11. Reitan RM. The relation of the trail making test to organic brain damage.
J Consult Psychol 1955; 19: 393–394
12. Cohen J. Wechsler Memory Scale performance of psychoneurotic, organic,
and schizophrenic groups. J Consult Psychol 1950; 14: 371–375
13. Baddeley A. Measuring memory. Psychopharmacol Ser 1988; 6: 12–22
14. Tupler LA, Welsh KA, Asare-Aboagye Y et al. Reliability of the Rey-
Osterrieth Complex Figure in use with memory-impaired patients. J Clin
Exp Neuropsychol 1995; 17: 566–579
15. Wutzler A, Becker R, Lammler G et al. The anticipatory proportion as an
indicator of language impairment in early-stage cognitive disorder in the
elderly. Dement Geriatr Cogn Disord 2013; 36: 300–309
16. Elias MF, Dore GA, Davey A. Kidney disease and cognitive function.
Contrib Nephrol 2013; 179: 42–57
17. Davey A, Elias MF, Robbins MA et al. Decline in renal functioning is asso-
ciated with longitudinal decline in global cognitive functioning, abstract
reasoning and verbal memory. Nephrol Dial Transplant 2013; 28:
1810–1819
18. Harciarek M, Biedunkiewicz B, Lichodziejewska-Niemierko M et al. Cog-
ORIGINAL ARTICLE
nitive performance before and after kidney transplantation: a prospective
controlled study of adequately dialyzed patients with end-stage renal dis-
ease. J Int Neuropsychol Soc 2009; 15: 684–694
19. Lux S, Mirzazade S, Kuzmanovic B et al. Differential activation of
memory-relevant brain regions during a dialysis cycle. Kidney Int 2010;
78: 794–802
20. Crichton GE, Elias MF, Dore GA et al. Measurement-to-measurement
blood pressure variability is related to cognitive performance: the Maine
Syracuse study. Hypertension 2014; 64: 1094–1101
Received for publication: 29.8.2014; Accepted in revised form: 10.4.2015
is 1559