# Persistent immune, coagulation and cardiac dysregulation are correlated with later post-discharge mortality in children with severe malnutrition

## TL;DR

Children with severe malnutrition who die months after hospital discharge show ongoing immune and heart issues that could be targeted for treatment.

## Contribution

The study identifies biomarkers of persistent immune and cardiovascular dysfunction linked to later mortality in children recovering from severe malnutrition.

## Key findings

- Elevated inflammatory markers like IL-10, IL-15, and IFN-α2 are associated with increased risk of later mortality.
- Coagulation markers such as fibrinogen and histidine-rich glycoprotein are elevated in children who later die.
- Cardiovascular proteins like angiotensinogen and tropomyosin are also linked to later mortality.

## Abstract

Children with complicated severe malnutrition (CSM) face high mortality after hospital discharge, yet the underlying mechanisms remain poorly understood. While early post-discharge mortality (< 2 months) has been linked to a sepsis-like inflammatory profile measured at discharge, it is unclear whether this relationship persists (later mortality; 2–6 months post-discharge). This study investigated whether immune, inflammatory, and endothelial dysfunction at 2 months post-discharge are associated with later mortality in children recovering from CSM.

We conducted a case–control study nested within a randomised placebo-controlled trial of daily co-trimoxazole in HIV-negative children aged 2–59 months with CSM in four Kenyan hospitals. Cases were children who died between 2 and 6 months post-discharge; controls were survivors frequency-matched by sex, site, and trial arm. Plasma cytokines, chemokines, endothelial markers, and untargeted proteomics were measured at discharge and 2 months post-discharge. Conditional Cox regression, adjusted for age, sex, site, mid-upper arm circumference (MUAC), and randomisation arm, was used to identify biomarkers associated with later mortality.

Cases were younger (had a median of 7 vs. 11 months), had longer hospital stays (14 vs. 10 days), and showed lower anthropometry (MUAC = 10.7 vs. 12.0 cm) and lower haemoglobin (9.7 vs. 10.6 g/dL) at 2 months post-discharge (all p < 0.05). Mortality 2–6 months post-discharge was associated with elevated inflammatory mediators (e.g. IL-10 [hazard ratio, HR: 1.47, 95% confidence interval, CI: 1.00–2.14], IL-15 [1.65, 95% CI: 1.08–2.51], IFN-α2 [1.51, 95% CI: 1.02–2.23]), acute phase proteins, apolipoproteins and coagulation markers, including fibrinogen, histidine-rich glycoprotein (1.40, 95% CI: 1.01–1.94), protein C inhibitor (SERPINA5, 1.50, 95% CI: 1.07–2.08), SERPINA10 (1.42, 95% CI: 1.02–1.99), and ADAMTS13 (0.41, 95% CI: 0.24–0.70). Additionally, cardiovascular and muscle-related proteins such as angiotensinogen (1.46, 95% CI: 1.03–2.08), α- and β-tropomyosin (0.68, 95% CI: 0.48–0.98), PI16 (0.72, 95% CI:0.54–0.97), and zyxin (0.61, 95% CI: 0.40–0.92) were elevated in cases.

Later mortality in children recovering from CSM is associated with persistent immune activation, a sepsis-like phenotype involving multiple systems. These findings suggest that children at risk of later mortality may benefit from biomarker-guided interventions initiated at discharge.

The online version contains supplementary material available at 10.1186/s12916-026-04647-9.