MG is an antibody-mediated autoimmune disorder that targets functional proteins on the postsynaptic membrane of the NMJ. Its primary pathophysiological mechanism is impaired neuromuscular transmission, manifesting clinically as fluctuating weakness and fatigability in skeletal muscles (1, 2). Epidemiological studies reveal a bimodal age distribution in MG onset. With an aging population, late-onset MG (LOMG, onset age >50 years) has become increasingly prevalent in clinical settings (3, 4).

Notably, T2DM is one of the most common comorbidities in patients with LOMG (5). Recent meta-analyses have demonstrated a significantly higher prevalence of T2DM among MG patients compared to the general population (6, 7). A case–control study further supports a strong association between T2DM and an increased risk of MG, suggesting that T2DM may serve as a potential risk factor for MG onset (8). This indicates a potential pathophysiological link between the two conditions that extends beyond mere coincidence. The underlying mechanisms of this association are multifaceted and not fully elucidated, potentially involving chronic hyperglycemia and insulin resistance-induced factors such as systemic low-grade inflammation, immune cell dysfunction, accumulation of advanced glycation end products, the effects of glucocorticoid therapy and reduced physical activity (9–14). A deeper understanding of the MG-T2DM comorbidity holds significant clinical value for optimizing patient management, improving prognosis, and identifying new therapeutic targets.

Neurophysiological studies, particularly Repetitive Nerve Stimulation (RNS) (15–18), are fundamental in the diagnosis and assessment of MG, providing objective evidence of “impaired neuromuscular transmission”. However, current clinical research has primarily focused on the electrophysiological characteristics of typical MG, with limited systematic investigation into potential electrophysiological alterations in MG patients with the important metabolic context of comorbid T2DM. Given that T2DM itself can cause Diabetic Peripheral Neuropathy (DPN) (19, 20) and may impact NMJ structure and function, the electrophysiological phenotype in the comorbid state is likely more complex, yet data in this area are scarce. Based on this research context, we designed and conducted a retrospective clinical study to explore the clinical and neurophysiological features of MG co-existing with T2DM.

This retrospective study compared the clinical and electrophysiological characteristics of MG patients with and without comorbid T2DM. The key finding is that MG patients with T2DM displayed a characteristic L-shaped decrement pattern during LFRNS at a specific nerve-muscle site (facial nerve–mentalis muscle). This decrement correlated with fasting blood glucose levels, diabetes duration, and clinical symptom severity. Additionally, MG patients with T2DM were more prone to peripheral nerve conduction abnormalities.

Diabetes has been proposed as a potential risk factor for the onset of MG (8), though the underlying mechanisms remain incompletely understood. Chronic hyperglycemia may induce immune dysregulation, including impaired Ca²⁺signaling and aberrant immune responses under high-glucose conditions (26–28). Advanced glycation end-products can activate the receptor for AGEs pathway, leading to the release of pro-inflammatory cytokines and proliferation of autoimmune-related cell subsets like CD4(+)CD28(−) T cells, contributing to MG pathogenesis (29, 30). Our group’s prior animal studies further showed that a T2DM environment can drive the differentiation and activation of follicular helper T cells, exacerbating humoral immune abnormalities and creating a predisposition for MG development (12, 13). Thus, routine screening for glucometabolic status is recommended for patients with LOMG.

In the present study, the observed positive rate of Antibody positivity in MG exceeded the commonly reported ranges in previous literature. The notably high overall antibody positivity rate observed in our center may be attributed to several factors, including the adoption of highly sensitive detection methods (cellular immunofluorescence assay and cell-based assay) and the routine screening for a broad array of antibodies, including Acetylcholine Receptor Antibody (AChR-Ab), Muscle-Specific Kinase Antibody (MuSK-Ab), Low-Density Lipoprotein Receptor-Related Protein 4 Antibody (LRP4-Ab), Titin Antibody (Titin-Ab) and Ryanodine Receptor Antibody (RyR-Ab). Additionally, potential differences in patient enrollment criteria may also have influenced these findings. To explore the morphological characteristics of LFRNS, this study focused on the facial nerve (recorded at the mentalis muscle) and the accessory nerve (recorded at the trapezius muscle) as the observation targets. This selection was guided by methodological considerations: traditional facial nerve recordings are often hindered by technical variability due to the low amplitude and instability of responses in commonly targeted muscles under repetitive stimulation. In contrast, the mentalis muscle (innervated by the mandibular branch of the facial nerve) and the trapezius muscle (innervated by the accessory nerve) produce more stable and reproducible CMAP waveforms during repetitive electrical stimulation. By using these nerves as experimental models, we aimed to reduce technical variability, improve the reliability of RNS morphological data, and strengthen the validity and persuasiveness of our conclusions.

The most salient electrophysiological finding in this study is the morphological difference in decrement curves. Classic MG typically presents with a negative decrement, where the CMAP amplitude shows significant recovery at the 6th or 7th stimulus during RNS, forming a “U-shaped curve” (31) In the present study, however, we observed an “L-shaped” LFRNS curve in facial nerve-mentalis recordings from MG patients with comorbid T2DM-characterized by an initial amplitude decline after stimulation, followed by a plateau without the typical U-shaped recovery. This pattern is rare in classic postsynaptic MG but resembles disorders with presynaptic or pan-synaptic involvement, such as Lambert-Eaton myasthenic syndrome (LEMS) and amyotrophic lateral sclerosis (ALS) (23, 32–34). It is suggested that the distinct trends of U-shaped and L-shaped curves reflect different underlying mechanisms of NMJ dysfunction. In ALS, the impairment may involve the entire NMJ structure, including both presynaptic and postsynaptic components (35, 36), while in LEMS, it primarily relates to a defect in acetylcholine release, where the physiological reduction in release induced by RNS cannot be compensated by reserve pools (34, 37). This suggests that the mechanism of neuromuscular transmission impairment in MG with T2DM may be more complex.

A plausible explanation is the combined effect of DPN and DM-related injury on the NMJ. MG patients with T2DM in this study exhibited peripheral nerve conduction abnormalities, including prolonged DML, reduced CMAP amplitude, slowed conduction velocity, and prolonged F-wave latency in the lower limbs, consistent with the electrophysiological profile of DPN. In diabetic patients, axonal damage may lead to impaired axoplasmic transport, which can result in disrupted transport of acetylcholine from its site of synthesis in the cell body to the axonal terminals, ultimately causing decreased acetylcholine storage in synaptic vesicles. Hyperglycemia and its downstream effects, such as oxidative stress, mitochondrial dysfunction, and microangiopathy, collectively contribute to axonal and myelin damage in peripheral nerves (38–40). At the NMJ level, both animal and clinical studies indicate that diabetes can alter axonal morphology, NMJ ultrastructure, and motor function (41, 42). Specific ultrastructural abnormalities, including T-tubule disruption, decreased synaptic vesicle number, and mitochondrial swelling with irregular cristae, have been associated with neuromuscular transmission deficits (43–45). Such presynaptic and/or axonal-level damage, along with NMJ ultrastructural alterations, may interact with or superimpose upon the inherent postsynaptic acetylcholine receptor deficiency in MG, collectively contributing to the impaired recovery characteristic of the “L-shaped” decrement pattern. However, this study did not employ systematic DPN clinical grading or advanced assessments, making comparative RNS analysis stratified by DPN severity unfeasible. This limitation warrants investigation in future stratified studies.

It should be emphasized that this distinct “L-shaped” pattern was primarily observed in the facial nerve-mentalis muscle configuration in the present study. The relatively low amplitude in this region may introduce technical artifacts, so this finding should be regarded as an electrophysiological phenotype emerging under specific conditions, namely, in cases of MG coexisting with T2DM and limited to a particular detection site. It does not represent a general characteristic of MG or diabetic neuromuscular pathology and should not be overgeneralized.

Correlation analysis revealed that the degree of low-frequency RNS decrement in the facial nerve was positively correlated not only with the QMG score, reflecting MG severity, but also with fasting blood glucose levels and the duration of diabetes. This further supports the hypothesis that a hyperglycemic environment may exacerbate neuromuscular transmission dysfunction. However, it must be clearly recognized that, due to the cross-sectional design of this study, these correlations do not establish causality. Poor glycemic control and MG disease activity may interact and serve as confounding factors. For instance, severe MG may reduce activity and corticosteroid use, thereby affecting glycemic control; conversely, chronic hyperglycemia may worsen neuromuscular symptoms through the mechanisms described above. This association highlights the clinical importance of blood glucose monitoring and management in MG patients, though exact causal pathways require clarification through prospective studies.

This study has several limitations. First, the retrospective single-center design may introduce selection bias. Second, the limited sample size, particularly in the T2DM subgroup, may affect the power of certain subgroup analyses. Third, as noted earlier, the lack of systematic assessment and stratification of DPN complicates disentangling its influence on electrophysiological outcomes. Fourth, although standardized procedures were followed, RNS recordings from the facial nerve-mentalis muscle are technically challenging and may introduce variability. Fifth, the cross-sectional nature of the data precludes inference of temporal or causal relationships between glycemic parameters and MG characteristics. Finally, the absence of healthy controls or a T2DM-only (without MG) control group limits our understanding of the specificity of the observed phenomena. Future studies could employ more sensitive techniques, including single-fiber electromyography combined with prospective designs, larger samples, and systematic neuropathy evaluation, to validate and extend these findings.

In summary, from an electrophysiological perspective, MG patients with comorbid T2DM not only exhibit peripheral nerve conduction abnormalities but may also demonstrate a distinctive “L-shaped” LFRNS decrement pattern at the facial nerve. This pattern likely reflects the complex interplay at the NMJ between hyperglycemia-related metabolic neuropathy and autoimmune postsynaptic membrane disease. The observed correlations between blood glucose levels, diabetes duration, and the degree of RNS decrement underscore the importance of monitoring glycemic metabolism in managing MG patients. Our findings provide new insights into how metabolic factors may modulate the electrophysiological phenotype of autoimmune neuromuscular disorders, though the underlying mechanisms and clinical implications require further investigation.