Growth after pediatric and neonatal acute kidney injury: a meta-analysis

Abstract

Background: Acute kidney injury (AKI) occurs commonly in critically ill children. The impact of AKI on pediatric growth outcomes has been sparsely described.

Objective: To compare growth in children with a history of AKI compared to those without AKI. We hypothesized that children with AKI would have worse growth compared to those without AKI.

Data sources: A convenience sample of existing prospective and retrospective cohorts of children with AKI who had already collected or were able to collect data on growth parameters before and after an episode of AKI.

Study eligibility criteria: There are < 5 studies in the published literature on growth in children with AKI. These investigators were contacted, and additional studies were added by contacting primary investigators of studies of childhood AKI in which data on growth parameters was able to be collected.

Participants and interventions: Children from existing cohorts evaluating AKI (exposure) during childhood. Each included cohort had previously received local IRB approval per institutional guidelines. As our study was a meta-analysis and only used cohort-level data, no IRB approval was required for this report.

Study appraisal and synthesis methods: Growth parameters (length and weight z-scores) before and after an episode of AKI were compared using a meta-means analysis. MOOSE guidelines were used. Data were pooled using a random-effects model. Hedges g was calculated, and Higgins I² statistic was used to define variability due to between-cohort heterogeneity.

Results: We included 3,586 children from 17 existing cohorts of AKI in various populations, including infants, children with cardiac disease, solid organ transplant and critically ill children without cardiac disease with follow-up from 12 months to 11 years after AKI. At most distant follow-up, those with AKI had lower length z-score than those without AKI (mean difference -0.37 [95%CI -0.52, -0.22, p < 0.001]) and lower weight z-score (mean difference of -0.29 [95%CI -0.43, -0.15, p < 0.001]). This difference was most striking in infants, as those with AKI had impaired growth (both length z-score and weight z-score) after AKI compared to those without AKI.

Limitations: The analysis included only a convenience sample of observational cohorts of children, study selection could have been biased, and we did not evaluate the relationship between decreased kidney function (e.g., chronic kidney disease) after AKI in these cohorts and its relationship to poor growth.

Conclusions and implications of key findings: This meta-analysis found that children with AKI have impaired growth after AKI. These findings were most striking in infants. We suggest focusing on growth outcomes in both clinical care and research investigating the impacts of AKI.

Systematic review registration number: NA.

Keywords: Acute kidney failure; Acute kidney injury; Development; Growth; Neonates.

A higher resolution version of the Graphical abstract is available as Supplementary information

Fig. 1

Forest plot of standardized mean differences and 95% CIs for growth z-scores for children and infants. Panels A and B show growth z-scores at the time of or prior to an AKI episode with length for age z-score (Panel A) and weight-for-age z-score (Panel B) in the 14 studies with baseline growth parameters. Panels C and D show growth z-scores at the time of most distant follow-up with length for age z-score (Panel C) and weight-for-age z-score (Panel D) for all 17 studies

Fig. 2

Length-for-age z-score following episode of AKI for neonates and infants (A), children with cardiac disease (B), children with non-kidney solid organ transplant (C), and critically ill non-cardiac disease (D) depicting meta-mean z-score in months following episode of AKI. In this figure, data points beyond 48 months are displayed at 48 months

Fig. 3

Weight-for-age z-score following episode of AKI for neonates and infants (A), children with cardiac disease (B), children with non-kidney solid organ transplant (C), and critically ill non-cardiac disease (D) depicting meta-mean z-score in months following episode of AKI. In this figure, data points beyond 48 months are displayed at 48 months