Acute Kidney Injury Is Associated With Subtle but Quantifiable Neurocognitive Impairments

Jessica A. Vanderlinden; Joanna S. Semrau; Samuel A. Silver; Rachel M. Holden; Stephen H. Scott; J. Gordon Boyd

Disclosures

Nephrol Dial Transplant. 2022;37(2):285-297. 

In This Article

Results

Enrollment

Participants with AKI were enrolled from May 2018 through March 2020 (Figure 1). During this time, 94 patients were screened, 64 were considered eligible and 43 consented. Most ineligible patients were discharged from the clinic or were deemed unsuitable by the physician due to a psychiatric diagnosis. Of the eligible patients, 11 were not interested and 10 were undecided at the time they were approached. Of the 43 patients that consented, 15 were not able to undergo any testing due to coronavirus disease 2019 (COVID-19) pandemic restrictions, 5 underwent one assessment only, 4 withdrew consent and 3 no longer wished to return to clinic. Thus, 16 participants underwent both assessments. The active control group was obtained from patients participating in a separate study examining the neurocognitive consequences of cardiac surgery. Patients in this group were enrolled between April 2014 and December 2019 (Figure 1). During this time, there were 272 eligible patients of which 91 consented to take part in the study. Of those that consented to the study, 88 performed their assessment prior to their cardiac surgery. From this cohort, a convenience sample of 21 patients was selected based on their lack of pre-existing kidney disease.

Figure 1.

CONSORT-style diagram demonstrating patient flow in the study.

Patient Characteristics

Clinical and demographic variables for the AKI cohort and the active controls are summarized in Table 1. When comparing these groups, demographics such as ethnicity, handedness, comorbidities of diabetes and hypertension, and highest reported education were not significantly different. Unsurprisingly, there were significant differences with regards to Cr and eGFR at the time of the first assessment, as active controls were selected on the absence of pre-existing kidney disease. The active control group was also younger than the AKI cohort. However, both RBANS and Kinarm assessments are normalized to age, facilitating direct comparison between the two groups.

Demographics specific to the patients with AKI were as follows: 6 (29%) had pre-existing kidney disease, with an average baseline Cr of 104 μmol/L (SD 31 μmol/L). During their episode of AKI, 12 patients (57%) required dialysis and had a mean highest Cr of 716 μmol/L (SD 438 μmol/L). About 86% (18/21) had KIDGO Stage 3 AKI, with the remaining three patients having either Stage 2 (2/21, 10%) or Stage 1 (1/21, 5%). The primary cause of AKI consisted of: sepsis (4/21, 19%), rhabdomyolysis (4/21, 19%), cardiovascular related (4/21, 19%), pneumonia (4/21, 19%) and obstruction (3/21, 14%). By the first assessment, nine patients (43%) had recovered to ≤25% of their baseline serum Cr. During the first assessment, the mean (SD) Cr and eGFR were 139 μmol/L (60 μmol/L) and 48.3 mL/min/1.73 m2 (19 mL/min/1.73 m2), respectively. For the second assessment, the average (SD) Cr and eGFR were 152 μmol/L (58 μmol/L) and 41 mL/min/1.73 m2 (18 mL/min/1.73 m2).

Kinarm Quantifies More Impairment Than RBANS in Patients With AKI

The mean time between the episode of AKI and first assessment, and the first and second assessments was 6.8 and 4.6 months, respectively (Table 1). Regarding the neuropsychological battery, little impairment was detected using the standardized clinical assessment (RBANS) at both assessments for the patients with AKI when compared with healthy control data (Table 2 and Figure 2A). At the first assessment, one AKI participant was defined as having global impairment using the composite Total Scale Score. This did not persist at the second assessment. With regards to the individual domains, participant performance is summarized in Table 2. Specifically, at the first assessment, few participants were defined as impaired in the domains of visuospatial (3/20), delayed memory (3/20), attention (2/20) and immediate memory (1/20). At the second assessment, impaired performance was detected in the domains of visuospatial (2/16) and delayed memory (1/16) only (Table 2).

Figure 2.

Compared to the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS, A), Kinarm (B) detected more differences in neurocognitive functioning between patients with acute kidney injury (AKI, blue and purple bars) and the active control group (cardiac, pink bars). In both graphs the dashed line represents the 95th percentile cut off for normative performance (RBANS: <75.25, Kinarm: >1.96). In Figure 2A performance below the red line indicates impairment, whereas in Figure 2B anything above indicated impairment. * indicates performance was significantly different by <0.05, ** indicates significantly different by <0.01, and *** indicates significance <0.001. APM, arm position matching; VGR, visually guided reaching; RVGR, reverse VGR; OH, object hit; OH+A, OH and avoid; BonB, ball on bar; SS, spatial span; TMTA/B, trail making test A/B.

When comparing patients with AKI versus our active control group, there was no difference in mean scores for the Total Scale Score (Figure 2A). The only difference noted in RBANS performance was in the domain of immediate memory (P < 0.05), where the performance on the second assessment by the AKI cohort was significantly better than their performance at baseline (Figure 2A).

In comparison with the RBANS, the Kinarm categorized a higher proportion of patients with AKI as impaired at both assessments using age-matched healthy controls (Table 2 and Figure 2B). At the first assessment, reverse visually guided reaching (11/21), visually guided reaching (10/21) and trail making test form A (10/20) categorized the highest proportion of patients with AKI as impaired (scores >1.96) (Table 2). These tasks specifically measure attention, visuomotor and executive function (Table 2). The other Kinarm tasks: trail making test form B (5/20, attention, visuomotor and executive function), object hit (4/16, attention and visuomotor), object hit and avoid (4/17, attention, visuomotor and executive function), ball on bar (3/13, visuomotor and motor range and coordination), arm position matching (1/18, proprioception) and spatial span (1/15, working memory) demonstrated varying degrees of impairment at the first assessment (Table 2). The number of patients that were classified as impaired decreased at the second assessment; however, impairments were still detected on 8/9 tasks when compared with the healthy age-matched controls. These tasks included: reverse visually guided reaching (8/16), trail making test forms A + B (6/15), visually guided reaching (5/16), object hit (4/13), object hit and avoid (3/12), ball on bar (2/12) and arm position matching (1/13) (Table 2). Spatial span was the only task that did not categorize any patients with AKI as impaired on the second assessment (Table 2).

When comparing Kinarm performance between the patients with AKI and the active control group, significant performance differences were also detected on 4/9 tasks (Figure 2B). Performance differences between the two groups at the first assessment were seen on the tasks of: arm position matching (P < 0.05), reverse visually guided reaching (P < 0.05), visually guided reaching (P < 0.01) and trail making test form A (P < 0.001) (Figure 2B). During the second assessment, only one task (object hit) was significantly different (P < 0.05) between the active control group and the patients with AKI (Figure 2B).

Neurocognitive Performance Mirrors Severity of AKI in Kinarm Tasks Measuring Visuomotor and Attention

The above data suggest that, at least for some Kinarm tasks, patients with AKI are more impaired at their first assessment but recover some neurocognitive function upon repeat testing. To determine whether or not neurocognitive performance mirrored the recovery of renal function, the cohort was subdivided on the basis of whether or not they recovered to ≤25% of their baseline serum Cr by the time of their first assessment. When compared with the active control group, there were no differences in RBANS scores with respect to kidney function recovery at either assessment time point (Figure 3A). In contrast, performance on the Kinarm tasks of visually guided reaching (visuomotor) and trail making test form A (visuomotor and attention) were significantly worse for patients with AKI who had not recovered kidney function, compared with active controls (Figure 3B). This significant difference was seen at the first assessment only (Figure 3B). For patients that had recovered their kidney function, performance on the Kinarm tasks were similar to the active control group, with the exception of trail making test form A where they performed significantly worse than the active controls, but again at the first assessment only (Figure 3B).

Figure 3.

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and Kinarm assessments for both cohorts with regards to whether the patients with AKI recovered ≤25% of their kidney function or not by their first assessment. 3A. Performance on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Note the dashed line at 75.25 represents the 95th percentile cut off and anything below the line indicates impaired neurocognitive performance. 3B Performance on the Kinarm. Note the dashed line at 1.96 represents the 95th percentile cut off and anything above this line indicates impaired neurocognitive performance. * denotes a significant difference in score with a p-value<0.05, ** indicates a p-value <0.01. APM, arm position matching; VGR, visually guided reaching; RVGR, reverse VGR; OH, object hit; OH+A, OH and avoid; BonB, ball on bar; SS, spatial span; TMTA/B, trail making test A/B.

The need for dialysis was also used as an indicator for the severity of AKI. There was no significant difference in performance between the active control participants and the participants with AKI, regardless of the need for dialysis on any of the sections of the RBANS (Figure 4A). When comparing performance on the Kinarm assessment, only the task of trail making test form A demonstrated significant differences between the groups (Figure 4B). Patients with AKI performed worse than the active control patients at the first assessment, regardless of the need for dialysis (Figure 4B). This difference did not persist at the second assessment.

Figure 4.

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and Kinarm assessments for both cohorts with regards to whether the patients with AKI required dialysis or not during their hospitalization. 4A. Performance on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Note the dashed line at 75.25 represents the 95th percentile cut off and anything below the line indicates impaired neurocognitive performance. 4B Performance on the Kinarm. Note the dashed line at 1.96 represents the 95th percentile cut off and anything above this line indicates impaired neurocognitive performance. * denotes a significant difference in score with a p-value<0.05, ** indicates a p-value <0.01. APM, arm position matching; VGR, visually guided reaching; RVGR, reverse VGR; OH, object hit; OH+A, OH and avoid; BonB, ball on bar; SS, spatial span; TMTA/B, trail making test A/B.

processing....