Acute kidney injury is associated with impaired cognition and chronic kidney disease in a prospective cohort of children with severe malaria

Abstract

Background: Acute kidney injury (AKI) is a recognized complication of pediatric severe malaria, but its long-term consequences are unknown.

Methods: Ugandan children with cerebral malaria (CM, n = 260) and severe malaria anemia (SMA, n = 219) or community children (CC, n = 173) between 1.5 and 12 years of age were enrolled in a prospective cohort study. Kidney Disease: Improving Global Outcomes (KDIGO) criteria were used to retrospectively define AKI and chronic kidney disease (CKD). Cognitive testing was conducted using the Mullen Scales of Early Learning in children < 5 and Kaufman Assessment Battery for Children (K-ABC) second edition in children ≥ 5 years of age.

Results: The prevalence of AKI was 35.1%, ranging from 25.1% in SMA to 43.5% in CM. In-hospital mortality was 11.9% in AKI compared to 4.2% in children without AKI (p = 0.001), and post-discharge mortality was 4.7% in AKI compared to 1.3% in children without AKI (p = 0.030) corresponding to an all-cause adjusted hazard ratio of 2.30 (95% CI 1.21, 4.35). AKI was a risk factor for short- and long-term neurocognitive impairment. At 1 week post-discharge, the frequency of neurocognitive impairment was 37.3% in AKI compared to 13.5% in children without AKI (adjusted odds ratio (aOR) 2.31 [95% CI 1.32, 4.04]); at 1-year follow-up, it was 13.3% in AKI compared to 3.4% in children without AKI (aOR 2.48 [95% CI 1.01, 6.10]), and at 2-year follow-up, it was 13.0% in AKI compared to 3.4% in children without AKI (aOR 3.03 [95% CI 1.22, 7.58]). AKI was a risk factor for CKD at 1-year follow-up: 7.6% of children with severe malaria-associated AKI had CKD at follow-up compared to 2.8% of children without AKI (p = 0.038) corresponding to an OR of 2.81 (95% CI 1.02, 7.73). The presenting etiology of AKI was consistent with prerenal azotemia, and lactate dehydrogenase as a marker of intravascular hemolysis was an independent risk factor for AKI in CM and SMA (p < 0.0001). In CM, AKI was associated with the presence and severity of retinopathy (p < 0.05) and increased cerebrospinal fluid albumin suggestive of blood-brain barrier disruption.

Conclusions: AKI is a risk factor for long-term neurocognitive impairment and CKD in pediatric severe malaria.

Keywords: Acute kidney injury; Child; Chronic kidney disease; Cognition; Malaria; Mortality.

Fig. 1

Flow chart of the study population. Children with a creatinine on study enrollment were included in the study and had their kidney function re-assessed at 1-year follow-up

Fig. 2

Mortality associated with acute kidney injury and across stages of AKI over 2-year follow-up. Kaplan-Meier plots showing 2-year survival in children with severe malaria based on the presence of KDIGO-defined acute kidney injury (AKI) status at admission (a) or the severity of AKI based on KDIGO-defined AKI stage (b). The break in the horizontal axis separates the first 7 days of follow-up (where the majority of in-hospital deaths occur) from the period of outpatient follow-up. Testing used the log-rank Mantel-Cox test in (a) and log-rank test for trend across stages of AKI (b)

Fig. 3

Association between acute kidney injury and neurocognitive recovery in children following severe malaria. Bar graphs showing the frequency of neurocognitive impairment in children with cerebral malaria (a) or severe malarial anemia (b). Neurocognitive impairment was defined as a gross deficit on the neurologic exam or an age-adjusted z score more than two standard deviations below the mean. Data analyzed using Pearson’s chi-square, *p < 0.01, **p < 0.01, ***p < 0.0001. c. Odds ratio of neurocognitive impairment (95% CI) associated with acute kidney injury from logistic regression models, 1 week post-discharge and 1- and 2-year follow-up. Multivariable-adjusted models included child age, sex, height- and weight-for-age z score, parental education, child schooling, an assessment of enrichment in the home environment, socioeconomic status, disease severity on presentation, the presence of coma, and number of seizures during hospitalization and parenteral antimalarial treatment (quinine vs. artemisinin derivative). *p < 0.05, **p < 0.01, ***p < 0.0001 following adjustment for multiple comparisons