Advances in diagnostic methods for early‐stage diabetic polyneuropathy

Significant advancements have been made in diagnostic methods for early-stage diabetic polyneuropathy. Early and accurate diagnosis of diabetic polyneuropathy is crucial for preventing further complications and enabling timely intervention. Furthermore, there is a need for an objective numerical value to evaluate the early stage of diabetic polyneuropathy. Diabetic polyneuropathy (DPN) is one of the most common and debilitating complications of diabetes mellitus. It is significantly associated with all-cause and cardiovascular mortality, and cardiovascular disease events, even after adjusting for cardiovascular disease risk factors and diabetes status1. DPN affects the peripheral nervous system, and can lead to dysfunctions in sensory, motor and autonomic nervous systems. Diagnosis of DPN is based on a combination of symptoms, ankle reflex and vibrational sensation2. However, this diagnostic procedure is time-consuming, and can only be carried out after DPN has progressed to the middle stage. Although the nerve conduction velocity test is the gold standard for evaluating DPN, it does not show a decline in the early stage of DPN. Early and accurate diagnosis of DPN is crucial for preventing further complications and enabling timely intervention. Furthermore, there is a need for an objective numerical value to evaluate the early stages of DPN. As small nerve fibers are affected in the early stage of DPN, validated measurement of small fiber neuropathy is necessary. Advancements have been made in diagnostic methods for early-stage DPN, enhancing our ability to effectively detect and manage this condition (Figure 1). Small fiber neuropathy, wherein the thinly myelinated Aδ and unmyelinated C fibers are mainly affected, develops in the early stage of DPN. Diagnosis of small fiber neuropathy can be made from skin biopsy samples, where the intraepidermal nerve fiber density (IENFD) is determined through immunostaining of histological skin sections. The combination of skin biopsy and intraepidermal nerve fiber density measurement serves as a diagnostic approach for early-stage DPN. This method involves a small skin biopsy, and assessing the density of intraepidermal nerve fibers. Reduced IENFD serves as a sensitive marker of small fiber neuropathy. Corneal confocal microscoty (CCM) is a non-invasive imaging technique that allows for the visualization of corneal nerve fibers. Studies show that corneal nerve fiber density is reduced in patients with DPN3. CCM is a rapid, non-invasive technique for imaging and quantifying small fiber pathology, and is reportedly as effective as measuring IENFD for the diagnosis of DPN4. CCM has garnered attention as a complementary or alternative method for diagnosing DPN. Data from CCM can be used to analyze the association between risk factors and changes in neuropathy5. A recent study compared the diagnostic ability of IENFD and CCM using machine algorithms to create wide-field image mosaics and quantify nerves in an area 37-fold the size of prior studies to avoid human bias6. In the same participants and at the same time point, the study found no correlation between IENFD and corneal nerve density. Corneal nerve density also showed no correlation with clinical measures of DPN, including neuropathy symptoms and disability scores, nerve conduction studies or quantitative sensory tests. Conversely, IENFD appeared to better reflect the clinical status of DPN. In contrast, a mouse study identified corneal nerve parameters that were more sensitive to DPN7. In this study, the corneal stromal-epithelial nerve penetration site (CSENPS) density, average fiber length/CSENPS, axonal alignment and corneal nerve fiber length were analyzed. At 8 weeks, db/db mice showed a significant increase in density and branch points in the periphery, which dissipated at 16 and 32 weeks. The total number of CSENPS significantly increased in db/db mice relative to controls at 8, 16 and 32 weeks. In contrast, only the alignment of the corneal nerve and the number of CSENPS were significantly increased in mice fed a high-fat diet compared with the controls. These results might suggest the need for further improvements in CCM and image software. The heart rate-corrected QT (QTc) interval is calculated using Bazett's formula: QTc interval = QT interval / square root of the RR interval. A prolonged QTc interval, even within the normal reference range, is reportedly associated with diabetic complications, such as diabetic cardiac autonomic neuropathy and diabetic nephropathy8, 9. Patients with DPN have longer QTc intervals compared with those without10. The area under the curve of the QTc interval was 0.894, greater than the nerve conduction velocity. The optimal cut-off point of the QTc interval for DPN was 428.5 ms, with a sensitivity and specificity of 0.715 and 0.921, respectively (P < 0.001). The coefficient of variation of R-R intervals (CVRR) calculated from 100 R-R intervals at rest and deep breathing recorded on an electrocardiogram has been used as a simple and convenient measure to assess cardiac autonomic function in patients with diabetes. Although the cut-off value of <2% has been used for a decrease of the CVRR at rest, the CVRRs are affected by age. Age-adjusted CVRR might be a more perceptive indicator for autonomic neuropathy. Age-adjusted CVRRs are calculated as follows: CVRR at rest 3.87–0.05 × age (years) and CVRR during deep breathing 8.55–0.09 × age (years)11. It is interesting that age-adjusted CVRR at rest is more strongly associated with other autonomic parameters and sensory symptoms than CVRR during deep breathing. Sympathetic skin response is accessed for the early stage of DPN. Sudomotor function is mainly mediated by unmyelinated sympathetic small-nerve C-fibers. Sudomotor function was assessed using SUDOSCAN 2 (Impeto Medical, Paris, France), a non-invasive device that rapidly measures the electrochemical skin conductivity of sweat glands based on the reverse ion electroosmosis technique induced by a low DC voltage12. The ability of the area under the curve to diagnose DPN was 0.663, and the apical cut-off value was 56 ms. Sudomotor function was also useful for the early diagnosis of peripheral artery disease when combined with the ankle-brachial index in patients with type 2 diabetes13. The sympathetic skin response identified using Synergy Mobile (Oxford, UK) showed that the latency of the sympathetic skin response was significantly higher in patients with type 2 diabetes than in healthy controls14. The sensitivity of the sympathetic skin response for detecting DPN was 83.3%. Furthermore, the sympathetic skin response positively correlated with microalbuminuria. Researchers are actively investigating biomarkers that can serve as diagnostic indicators of DPN. Blood-based biomarkers, such as specific cytokines, growth factors and microribonucleic acids have shown promise for identifying DPN and assessing its severity. These biomarkers have the potential to complement the existing diagnostic methods and enhance early detection. Plasma metabolite profiles showed that the plasma levels of 12 metabolites (dimethylarginine, N6-acetyllysine, N-acetylhistidine, N,N,N-trimethyl-alanylproline betaine, cysteine, 7-methylguanine, N6-carbamoylthreonyladenosine, pseudouridine, 5-methylthioadenosine, N2,N2-dimethylguanosine, aconitate and C-glycosyl tryptophan) were significantly higher in the type 2 diabetes mellitus with DPN group than in the control and type 2 diabetes mellitus without DPN groups15. Receiver operating characteristic curve analyses showed that all 12 metabolites had an area under the curve >0.75, and 5-methylthioadenosine, N6-acetyllysine and aconitate had the strongest discriminatory power of >0.84. Urinary epidermal growth factor, a sensitive biomarker of early diabetic kidney disease, is associated with cardiac autonomic neuropathy. Urinary epidermal growth factor correlated with baseline and longitudinal changes in the low-frequency area: respiratory frequency area ratio (measure of sympathovagal balance) after adjusting for age, sex, body mass index and hemoglobin A1c16. Although these biomarkers are expected to allow an easy and early diagnosis of DPN, further clinical validation is needed. The present article focused on recent attempts to quantify numerical values for evaluating early-stage DPN. A simple, quantifiable evaluation of early-stage DPN would be beneficial for preventing the progression of diabetic neuropathy and facilitating its treatment. This represents a promising development in clinical research on DPN. The author declares no conflict of interest. Keiko Naruse is an Editorial Board member of Journal of Diabetes Investigation. To minimize bias, they were excluded from all editorial decision-making related to the acceptance of this article for publication.