# Molecular cross-talk via extracellular vesicles for the characterization of young subjects with type 1 diabetes unravels new potential markers of insulin resistance and double diabetes

## TL;DR

This study uses molecular signals in blood to identify new markers for insulin resistance in children with type 1 diabetes and obesity.

## Contribution

A combined omics approach reveals EVs and specific metabolites as potential markers for insulin resistance in pediatric double diabetes.

## Key findings

- EVs from insulin-resistant children show proteins linked to fatty acid suppression and liver damage.
- Oleoylcarnitine and linoleoylcarnitine levels are significantly higher in insulin-resistant children.
- Palmitoleoylcarnitine and oleoylcarnitine best distinguish insulin-resistant from non-resistant children.

## Abstract

Insulin resistance (IR) is commonly calculated using a simple mathematical formula, the eGDR (estimated Glucose Disposal Rate), but in the paediatric type I diabetes (T1DM) population this value has provided contrasting information. We aimed to provide a clearer metabolic “fingerprint” in children with “double diabetes”, focusing on the molecular cross-talk mediated by extracellular vesicles (EVs).

Paediatric patients were classified based on the eGDR value in: insulin-resistant (T1DM+, eGDR < 8 mg/Kg/min, n = 29) and non-insulin-resistant (T1DM-, eGDR > 8 mg/Kg/min, n = 35). Venous blood collected from them, and 30 healthy controls was used to obtain dried blood spots (DBS) for AAs and ACs analysis by FIA-MS/MS and for EV by a patented flow cytometry method. Then, EVs were subjected to shotgun proteomics analysis by LC-MS/MS.

Our data showed that T1DM + EVs were packaged with proteins involved in fatty acid metabolism suppression through STAT3 inhibition and related to possible liver damage. ACs on DBS samples corroborated these data, demonstrating a significant increase in oleoylcarnitine (C18:1), linoleoylcarnitine (C18:2), and myristoylcarnitine (C14) in T1DM+. The combination of clinical and metabolic data led to the identification of a statistical model with an out-of-bag error of 0.115%, demonstrating that palmitoleoylcarnitine (C16:1) and C18:1 are the metabolites that best distinguish children with T1DM + from T1DM- ones. C16:1 correlated significantly with eGDR (p = 0.0023).

Combined “omics” approach allowed us to identify a new metabolic “photograph” in a complex context involving diabetes complications related to obesity and IR in a paediatric population that is not yet fully characterized, identifying EVs as well-organized and functionalized shuttles.

The online version contains supplementary material available at 10.1186/s13098-025-02042-7.