Promising predictors of diabetic peripheral neuropathy in children and adolescents with type 1 diabetes mellitus

Aims: To assess the utility of the Michigan neuropathy screening instrument (MNSI), lipid profile, serum neuron specific enolase (NSE), and serum heat shock protein 27 (HSP 27) to the prediction of diabetic peripheral neuropathy (DPN) in children and adolescents with type 1 diabetes (T1DM).

Methods: In this case-control study, 50 children (mean age 13.7±3 years) diagnosed with T1DM for at least five years were enrolled and evaluated through complete neurological examination, MNSI, and nerve conduction study (NCS). Additionally, HbA1c, lipid profile, serum NSE, and serum HSP 27 levels were measured for patients and controls, i.e., 50 age and sex matched healthy children.

Results: Based on NCS criteria, the prevalence of DPN was 24%. Within this cohort, statistically significant lower conduction velocity for the posterior tibial and sural nerves, as well as a prolonged latency period for the common peroneal and sural nerves was seen. In this neuropathic cohort: older age, earlier age of diabetes onset, longer disease duration, higher total cholesterol, triglycerides, LDL cholesterol, HbA1c, serum NSE, and HSP27 levels were observed. ROC analysis (AUC: 0.955, with 75% sensitivity and 94.74% specificity) determined that a score of ≥1.5 for the MNSI examination was the best cut-off point for differentiating between neuropathic and non-neuropathic patients. However, the MNSI questionnaire’s cut-off point of ≥5 had an AUC of 0.720, 75% sensitivity, and 63% specificity, with improved overall instrument performance when combining both scores. Regarding blood biomarkers, serum NSE had greater sensitivity and specificity in discriminating neuropathic patients than HSP27 (92% and 74% versus 75% and 71%, respectively). Regression analysis revealed a substantial dependency for MNSI and serum NSE in predicting DPN in patients.

Conclusions: In this study of 50 children, diagnosed with T1DM for ≥5 years, MNSI and serum NSE are promising predictive tools for DPN in children and adolescents with T1DM, even when they are asymptomatic. Poor glycaemic control and lipid profile changes may play a critical role in the development of DPN in these patients, despite conflicting results in various studies.

Comments.  DPN in children and adolescents with T1DM is now being increasingly recognised as a growing issue. However there remains considerable heterogeneity of data in incidence and prevalence (3-62%) mainly due to which racial and ethnic differences and the variability of methods used in characterizing and diagnosing neuropathy. The American Diabetes Association (ADA) recommended that all patients with T1DM should be assessed for DPN 5 years following diagnosis, and at least annually after that. In my opinion, this is first study assessing DPN in children and adolescents with T1DM combining both MNSI, HbA1c, lipid profile, HSP27, and NSE, using NCS as the gold standard to define DPN. Equally importantly all the patients were devoid of neuropathic symptoms making the diagnosis subclinical. The limitations, as is obvious, include smaller numbers as well challenges of performing MNSI and NCS in young children.

So, to conclude, some thoughts for NEUROdiab colleagues to consider:

1. Are our measures of small fibre neuropathy which have consistently shown to be more sensitive than large fibre methods in detecting DPN validated for use in children and adolescent? E.g. cornea in children have different thickness, curvature, composition, mechanical strength as well as cell differentiation and adhesion.

2. Are there any prospective studies correlating NSE and HSP with other measures of neuropathy? My Medline search did not show anything.

Sanjeev Sharma

Reference. Abo Hola AS, Abd El Naby SA, Allam ET, Gab Allah AA, Hammad DA. Promising predictors of diabetic peripheral neuropathy in children and adolescents with type 1 diabetes mellitus. Ital J Pediatr. 2024 Oct 14;50(1):215. doi: 10.1186/s13052-024-01774-y. PMID: 39402605; PMCID: PMC11479551.

https://ijponline.biomedcentral.com/articles/10.1186/s13052-024-01774-y