Which fibre first? Subtyping nerve damage in type 1 diabetes mellitus

Aims: To determine the optimal combination of diagnostic tests for confirmed diabetic polyneuropathy (DPN) in identifying specific neuropathy subtypes and relate these subtypes to DPN status and severity.

Methods: In this prospective cross-sectional cohort study, 221 adults [controls n=51; type 1 diabetes mellitus (T1DM) n=170] were assessed using clinical examination, point-of-care nerve conduction, questionnaires and skin biopsies for intra-epidermal nerve fibre density (IENFD) analysis. DPN severity was categorised using the Toronto Clinical Neuropathy Score (TCNS). Diagnosis of DPN followed the Toronto Diabetic Neuropathy Expert Group criteria. DPN was further subdivided into non-painful (NP-DPN) and painful (P-DPN) based on the NeuPSIG grading system, with P-DPN defined by an average pain intensity score of ≥4. Three diagnostic models were used to identify the predominant DPN subtypes: small fibre (SFN), mixed fibre (MFN), and large fibre (LFN) neuropathy. Model 1 required at least one abnormal result in small or large fibre measures, while MFN required one abnormal result in both. Models 2 and 3 raised the threshold to two and three abnormal results, respectively. These diagnostic models were applied across all groups and subgroups to evaluate classification frequency, with multiple-group comparisons and correlation analyses conducted. The performance of Model 2 was also compared to the ‘Besta Criteria’ for SFN to assess its efficacy.

Results: Participants with NP-DPN were older, had longer diabetes duration (p=0.009), and lower HbA1c (p=0.006) compared to those with P-DPN, who were more frequently smokers (p=0.035). Participants with P-DPN had significantly higher PainDETECT, Michigan Neuropathy Screening Instrument, and TCNS scores compared to NP-DPN (all p<0.001). SFN and LFN were more prevalent in NP-DPN and P-DPN relative to controls (p<0.001), with no significant difference between NP-DPN and P-DPN groups. Neuropathy severity correlated with increased HbA1c, worsening IENFD and nerve conduction tests (p<0.001). As severity increased, MFN became more prevalent, while isolated LFN and SFN presentations decreased. No significant differences between NP-DPN and P-DPN were observed in the Besta model or Model 2 (adapted), with consistent classification patterns across groups.

Conclusions: In the mild stages of DPN, either small or large nerve fibres are more vulnerable to damage. MFN predominates in T1DM DPN, with no clear distinction between painful and non-painful presentations. For SFN diagnosis in T1DM, Model 2 based on two abnormal tests had the best diagnostic performance.

Comments. The measures proposed in Model 2, together with IENFD, are effective for detecting nerve fibre abnormalities in the clinical setting, with SFN identified in 6/48 (12.5%) participants with T1DM and without DPN. This study offers valuable data for T1DM and supports the use of a clinical model for detecting SFN, MFN and LFN in this population. Previous research in T2DM identified distinct subgroups of thermal and mechanical hyperalgesia, which progress to sensory loss (Tsilingiris, D et al. Diabetes 2023; 73:135-146). This progression may also be relevant to T1DM, where sensory loss and large fibre involvement may signify critical stages of irreversible axonal damage (Mooshage CM et al Diabetologia 2024; 67:275-289). As the condition advances, MFN becomes a predominant finding, serving as a key clinical diagnostic measure. Therefore, early intervention using personalised, multi-factorial approaches, which reflect these varying mechanisms is imperative.

Jamie Burgess

Reference. Karlsson P, Sjogaard MB, Schousboe K, Mizrak HI, Kufaishi H, Staehelin Jensen T, Randel Nyengaard J, Hansen CS, Yderstræde KB, Buhl CS. Assessment of neuropathy subtypes in type 1 diabetes. BMJ Open Diabetes Res Care. 2024 Jul 18;12(4):e004289. doi: 10.1136/bmjdrc-2024-004289. PMID: 39025795; PMCID: PMC11261698.

https://drc.bmj.com/content/12/4/e004289.long