The association between visceral fat indices and diabetic peripheral neuropathy in type 2 diabetes mellitus: A retrospective cross-sectional study Authors: Nesibe Aydogdu, Emel Sabaz Karakeci Published: 2026 ABSTRACT ObjectivesThe research investigates how type 2 diabetes mellitus (T2DM) patients with diabetic peripheral neuropathy (DPN) relate to their visceral fat indices, which include visceral adiposity index, lipid accumulation product, and atherogenic plasma index (AIP).Patients and methodsThe retrospective study analyzed 562 T2DM patients (324 females, 238 males; mean age: 64±13.3 years; range, 56 to 72 years) between January 2022 and December 2024. Participants were divided into two groups based on their electromyography (EMG) results and Leeds Assessment of Neuropathic Symptoms and Signs pain scale scores (a score of 12 or higher indicated neuropathy). Electromyography-positive patients (n=213) were compared with EMG-negative patients (n=349) for their anthropometric data, biochemical results, and visceral fat measurements. Independent factors that predicted DPN were sought to be identified.ResultsThe EMG-positive patients showed higher values for body mass index, visceral adiposity index, lipid accumulation product, and AIP, as well as triglycerides and hemoglobin A1c levels, compared to EMG-negative patients (p<0.001 and p<0.01). The logistic regression analysis revealed that hemoglobin levels (odds ratio =0.83, 95% confidence interval 0.74-0.94), hemoglobin A1c levels, waist circumference, and AIP served as independent risk factors for DPN.ConclusionThe findings show that visceral fat indices serve as effective biomarkers in identifying T2DM patients at high risk of developing DPN. The combination of these indices with glycemic control monitoring could improve DPN prevention methods. Future research needs to conduct bigger prospective studies to confirm these findings and understand the intricate links between lipid metabolism, visceral fat, and DPN. Diabetes mellitus (DM) is the leading cause of peripheral polyneuropathy.[1] About half of all adults with DM are affected by diabetic peripheral polyneuropathy (DPN), which can lead to foot ulcers and lower extremity amputations.[2] Hereditary distal symmetric polyneuropathy, the most common type of polyneuropathy, represents approximately 75% of cases. Raising awareness of these issues can help with treatment plan diversification and demonstrates the need for comprehensive healthcare for those suffering from diabetes.[2] Hyperglycemia has long been documented as a primary contributor to the pathogenesis of nerve damage in DM. There is an increasing body of evidence showing that disturbances in lipoproteins in DM may also be largely contributory.[3] Hyperlipidemia, as is evidenced with elevated triglyceride (TG) levels, has a strong association with the development of diabetic neuropathy.[4] Hyperlipidemia is associated with elevated levels of oxidized low-density lipoprotein (oxLDL) in the plasma that are recognized by the receptors for oxLDL on the membranes of the neurons and gamma-delta (γδ) T cells. Recognition of oxLDL stimulates cellular nicotinamide adenine dinucleotide phosphate oxidases.[5] Inflammatory characteristics are a relevant component of the development of peripheral neuropathies. This is particularly relevant for patients with elevated cholesterol levels. When a cell undergoes oxidative stress, it may result in a state of inflammation within the peripheral nerves. Lipid products such oxysterols and oxLDL may facilitate an enhanced release of inflammatory markers, contributing to a state of localized inflammation. Moreover, when an individual has elevated cholesterol levels, it can impact overall blood flow to small arteries, which vascularize the nerves leading to less oxygen supply in the nerve area.[6] Recent research utilizing magnetic resonance imaging has identified hypointense lesions in the sciatic nerve of patients with type 2 DM (T2DM). The lesions showed a significant association with increased TG levels and decreased levels of high-density lipoprotein (HDL) cholesterol (HDL-C).[7] Studies originating from experimental conditions provide evidence that neuropathy could result from a common underlying mechanism, independent of glycemia. In high-fat-fed mouse models, it has been postulated that elevated circulating oxLDL could cause oxLDL binding to dorsal root ganglion receptors, potentially contributing to nerve damage.[8] Visceral adiposity is the accumulation of fat in the abdominal cavity around internal organs and is deemed more dangerous than subcutaneous fat due to its association with cardiovascular disease and metabolic syndrome. Visceral fat is metabolically active and is accompanied by inflammation, dyslipidemia, and insulin resistance. Assessment of visceral fat can be completed via the visceral adiposity index (VAI).[9] The VAI is a simple and reliable anthropometric indicator based on body mass index (BMI), waist circumference (WC), and lipid markers such as TGs and HDL-C.[10] The lipid accumulation product (LAP) is a significant parameter indicating visceral fat with associated multiple metabolic risk factors, calculated from a combination of WC and fasting TG levels.[11] The atherogenic plasma index (AIP) is a blood marker for atherosclerosis and is believed to correlate more accurately for TG and HDL-C levels along with the pathogenesis and characterization of dyslipidemia. A recent study has suggested that the AIP is related to prediabetes and obesity and may even be a predictive marker for developing diabetes.[12, 13, 14] All three markers are related to lipid metabolism and provide interesting information on the distribution of lipid abnormalities and fats. Taken together, a comprehensive exploration of each of these markers may add clarity to the complex interactions between metabolic diseases and DPN. The diagnosis of DPN is largely based upon clinical assessments and electromyography (EMG). However, it can be difficult to diagnose DPN, particularly in its early stages with small fiber involvement, or in patients with atypical or mixed symptoms. Although EMG or other electrophysiologic methods can detect large fiber involvement, this does not confirm a diagnosis of DPN in all cases. Consequently, some patients may not receive an accurate diagnosis or the clinical care they require. Current research is exploring the use of glucose and lipid metabolism parameters as additional biomarkers for diagnosis. It is important to highlight that studies investigating the relationship between visceral fat indices and EMG findings are limited. This study aimed to investigate the factors that may influence EMG results in individuals with DPN, with a specific focus on measurements of visceral adiposity. PATIENTS AND METHODS This retrospective study included patients who presented to the outpatient clinic of the Department of Internal Medicine at the Health Sciences University, Fethi Sekin City Hospital between January 2022 and December 2024. The guidelines from the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) were used for reporting.[15, 16] The study cohort included adults aged 30 to 80 years who met all of the following criteria: physician-diagnosed T2DM with a documented medical history, confirmed DPN based on standard diagnostic criteria, neuropathic pain verified by a Leeds Assessment of Neuropathic Symptoms and Signs pain score of 12 or higher, availability of complete EMG and nerve conduction study records from the Department of Physical Medicine and Rehabilitation.[17] Patients diagnosed with type 1 DM, those with impaired cognitive function or insufficient mental capacity, individuals with familial hypercholesterolemia, and patients using lipid-lowering medications such as statins were excluded from the study. Additionally, individuals with alternative causes of neuropathy (e.g., vitamin B12 deficiency, hypothyroidism, chronic kidney, liver, cerebrovascular and acute cardiovascular diseases, and alcohol abuse), those using neurotoxic drugs, hereditary neuropathies, and individuals with missing data were also excluded. A total of 821 patients were initially enrolled in the study. We excluded 220 patients who lacked EMG results and 39 patients with incomplete information. Therefore, the final analysis comprised 562 patients (324 females, 238 males; mean age: 64±13.3 years; range, 56 to 72 years). Written informed consent was obtained from all participants. The study protocol was approved by the Non-Interventional Clinical Research Ethics Committee of Fethi Sekin City Hospital (Date: 05.12.2024, Approval No: 263451305). The study was conducted in accordance with the principles of the Declaration of Helsinki. The recorded data was based on various factors, including anamnesis, physical examination, age, sex, and laboratory results such as hemoglobin A1c (HbA1c), glucose, urea, creatinine, aspartate transaminase (AST), alanine aminotransferase (ALT), HDL, LDL, TGs, albumin, hemoglobin, height, weight, and WC, VAI, AIP, and LAP. For each participant, height, weight, and WC were measured while standing, using a flexible tape measure positioned between the lower edge of the rib cage and the upper boundary of the iliac crest. All measurements were taken by the same staff member using the same scale. Body mass index was determined by taking an individual's weight in kilograms and dividing it by the square of their height measured in meters. The VAI and LAP are used to better understand visceral fat, taking sex into account based on the thoughtful formulas developed by Amato et al.[18] and Kahn.[19] Additionally, AIP, which helps us assess the risk of atherogenic plasma, is determined through the TG/HDL-C ratio. The VAI evaluates visceral fat by combining anthropometric variables, such as BMI and WC, with metabolic variables, such as TG and HDL-C concentrations (mmol/L). High VAI values are related to a greater amount of visceral fat. The LAP is, similarly, another key way to assess abdominal lipid accumulation based on WC and TG concentration. The AIP provides a more specific representation of plasma atherogenicity and may provide a better picture of TG and HDL-C, as well as the development and type of dyslipidemia. The VAI formula for males is defined as: VAI = WC / [39.68 + 1.88 × BMI] × (TG / 1.03) × (1.31 / HDL) For females as: VAI = WC / [36.58 + 1.89 × BMI] × (TG / 0.81) × (1.52 / HDL) AIP = (TG / HDL-C). The LAP formula for males is defined as: LAP = (WC − 65) × TG For females as: LAP = (WC − 58) × TG. All EMGs (Natus UltraPro S100; Natus Medical Incorporated, Middleton, WI, USA) were performed by the same EMG nurse in the physical therapy clinic. The EMG results were interpreted and classified as either normal or abnormal electrophysiologic studies for polyneuropathy by the same specialist. The diagnostic criteria for polyneuropathy recommended by the American Academy of Neurology were utilized. For sensory nerves, an amplitude of the sural nerve <5 μV or a conduction velocity <40 m/sec indicated abnormalities. For motor nerves, an amplitude of the peroneal nerve <2 mV or a conduction velocity less than 40 m/sec were considered abnormal. A diagnosis was confirmed when there were abnormalities in two or more nerves, including at least one sensory nerve.[20] Statistical analysis The analyses were performed using IBM SPSS version 22.0 (IBM Corp., Armonk, NY, USA) and G*Power version 3.1 software (Heinrich-Heine Universität Düsseldorf, Düsseldorf, Germany).[21] In this study, factors were dichotomized into two groups: EMG positivity and negativity. We distinguished between parametric and nonparametric data within the quantitative (continuous) variables. For parametric data, such as hemoglobin, we reported the mean along with the standard deviation. For nonparametric data, we used the median and indicated the first and third quartiles. Categorical data were presented as frequency and percentages. For the univariate analysis, we analyzed the differences between groups using Student's t-test for parametric data and the Mann-Whitney U test for nonparametric data. Additionally, for univariate analysis of categorical data, we used the chi-square test and cross-tabulation. The cutoff values were determined only for the significant continuous data (Table 1). The sensitivity and specificity for EMG positivity were evaluated using the area under the curve (AUC) derived from receiver operating characteristic (ROC) curves. To compare these curves, the DeLong method was employed.[19] The optimal cutoff values for achieving the highest diagnostic accuracy of each analytical parameter were established using the Youden index formula: sensitivity + specificity − 1.[22] During the second phase, both bivariate Pearson and Spearman correlation analyses were utilized to assess the strength of relationships among all continuous predictors. The Pearson chi-square test was applied to examine the strength of associations among all categorical predictors. To prevent collinearity in the regression model, a decision was made regarding which highly correlated covariates to include. When issues of collinearity emerged, the study researchers conferred to decide which variable should be kept based on its clinical importance, reliability of measurement, and accessibility.[23] The interactions among the covariates in the final model were analyzed separately, and multicollinearity was mitigated by eliminating the variable that caused it from the model.[24] A backward Wald logistic regression analysis was used to determine the independent predictors of EMG negativity at the final stage. The odds ratio (OR) and 95% confidence intervals (CIs) were derived from the beta coefficients of variables that exhibited a significant p-value. For all tests, a p-value <0.05 was regarded as statistically significant. A statistical power of over 80% was established to decrease the likelihood of a type 2 error. To compute effect size, Cohen's d is applied for t-tests, Cramér's V for chi-square tests, and Rosenthal's r for Mann-Whitney U tests. RESULTS The demographic data and first-level univariate analysis results are presented in Table 1. Additionally, EMG results were positive for 213 patients, which represents 37.9% of those included in the study. In a univariate analysis of 20 factors with EMG negativity, significant differences were found in 13, where individuals with a positive EMG had significantly higher levels of sex, glucose, HbA1c, height, weight, WC, and TG than those with a negative EMG. The positive EMG group demonstrated significantly lower albumin and hemoglobin levels. Furthermore, individuals with a positive EMG had significantly increased BMI, VAI, AIP, and LAP. Both groups had similar levels of urea, creatinine, AST, ALT, HDL, and LDL (Table 1). In the second analytical phase, we advanced sex, HbA1c, hemoglobin, WC, and AIP to the third stage after confirming that multicollinearity was absent. A backward Wald logistic regression analysis identified these variables as independent predictors in the final model, as detailed in Table 2. Triglycerides alone may not fully reflect the metabolic risk indicated by the AIP, which combines TG and HDL levels. The regression analysis retained the AIP due to its stronger association with DPN. Optimal cutoff values for the identified independent predictors were as follows: Hb <12.2 g/dL, HbA1c >9.1%, WC ≥97 cm, and AIP >3.286 (Table 1). The effectiveness of these cutoff values in diagnosis, including sensitivity, specificity, AUC, and 95% confidence intervals, is illustrated in Table 3 through ROC analysis. DISCUSSION In individuals with T2DM, previously identified risk factors such as extended disease duration, elevated blood sugar levels, high HbA1c, as well as the use of alcohol and tobacco, high blood pressure, and abdominal obesity, might not be adequate to prevent the onset of DPN and are capable of lowering the incidence of peripheral neuropathy by only 5 to 7%. Even with strict management of hyperglycemia, more than 40% of patients may eventually receive a diagnosis of peripheral neuropathy.[2, 25] Studies have shown that disruptions in lipid metabolism, particularly lower levels of HDL-C and increased plasma TG, could be associated with DPN in individuals with DM.[26] In patients experiencing painful DPN, hypertriglyceridemia is noticeably present; however, in those who do not have neuropathic pain, there is no correlation between total cholesterol, TG levels, and plasma LDL cholesterol and DPN.[27] Recent studies indicate that the distribution of visceral fat plays a key role in insulin resistance, DM, and consequent vascular and nerve complications.[28, 29] Some theoretical speculations have been made about diabetic complications related to the accumulation of visceral fat. One hypothesis is that the lipolytic activity of visceral fat cells is greater than subcutaneous fat cells. This increases the likelihood that visceral fat will lead to hyperglycemia, hyperinsulinemia, hypertriglyceridemia, and elevated apolipoprotein B, which could influence microvessels. The area of visceral fat had a considerably larger amount of inflammatory cells, including macrophages. Visceral fat secretes cytokines including interleukin-6, tumor necrosis factor-alpha, plasminogen activator inhibitor-1, and C-reactive protein that can lead to endothelial dysfunction and sparked angiogenesis and vascular permeability.[29] Currently, there are various indices (e.g., VAI, Chinese visceral adiposity index (CVAI), and LAP) that have been created to assess visceral fat accumulation and are used clinically to estimate the risk of cardiovascular disease and DM.[30, 31] An investigation on a possible association with T2DM and metabolic markers of visceral adiposity, namely the VAI and LAP, in individuals with a history of hypertension indicated a notable positive relationship, suggesting that these metabolic markers may be good risk predictors.[32] More recently, a research article indicated that the VAI may be a helpful marker to determine lipid dysfunction while supporting significant associations between VAI and cardiometabolic risk, particularly concerning T2DM.[18] Kahn[19] indicated in their research that LAP may be a useful measure for determining metabolic syndrome and T2DM risk. Conventional anthropometric measures such as BMI only consider weight in relation to height and do not evaluate the amount and location of body fat; LAP offers a more precise method of identifying individuals that may be at risk. They argue for its value for clinicians to assess metabolic health and engaging in proactive discussions with this at-risk group. In the study of Wei el al.,[33] correlations with diabetic retinopathy and different visceral fat measurement indicators (VAI, AIP, and LAP among others) were explored. The authors found significant correlations with higher VAI, AIP, and LAP levels and increased prevalence of diabetic retinopathy, indicating keeping these metabolic markers in the normal range may offer important clinical consideration or a possible method to delay or slow diabetic retinopathy. Chen et al.[34] extensively investigated the connection between diabetic retinopathy, a frequent complication of T2DM, and visceral obesity. They confirmed that visceral obesity is a significant predictor of newly diagnosed diabetic retinopathy. Notably, this relationship was found to be independent of measures of general and abdominal obesity. The authors also analyzed the relationship through several other factors, including age, duration of DM, coexisting hypertension, sex, dyslipidemia, renal function, and BMI, and found that none of these factors altered the association. The authors also considered VAI and LAP as individual measures of the relationship but concluded that neither measure was superior to the other at predicting risk of diabetic retinopathy. Although available epidemiological literature, including the work by Chang et al.,[35] indicate that hyperlipidemia and lipid-lowering agents, such as statins, were not related to DPN in adults with T2DM, this may not rule out visceral adiposity indices as more sensitive markers of metabolic risk. While there are significant methodological differences between the study of relevance and our study, it is possible the relevant study relates both to the methodology of the study and whether statin therapy was effective. Additionally, the fact that we specifically selected participants who had never taken pharmacological therapy may have affected our study. Finally, the complexities of the VAI and LAP measures, which include both measures of body composition and lipid levels, may have contributed to the outcomes. Although EMG offers helpful electrophysiological information for diagnosing DPN, it has varying accuracy depending on location, patterns, and type of nerve injury. This study appears to be the first to examine the impact of these factors on the predictive value of EMG. We found that DPN patients with positive EMG had statistically significant increases in height, weight, WC, TGs, albumin, hemoglobin, glucose, HbA1c, VAI, AIP, and LAP compared to patients who with negative EMG. The findings also showed that elevated hemoglobin with low HbA1c, WC, and AIP decreased the chance of a positive EMG result. Hemoglobin changes are thought to reflect changes in oxygen transport efficiency and, possibly, an effect of the long-term presence of anemia. Although not identified as a separate risk factor in the regression analysis, univariate analysis suggested that VAI and LAP values may be relevant to the EMG positivity. These outcomes suggest that a better risk assessment framework for T2DM patients may be achieved by adding VAI, AIP, and LAP values. This strategy could help identify the patients who are at high risk for peripheral neuropathy. Monitoring visceral fat accumulation appears to be a crucial element of risk assessment, in addition to blood glucose and blood pressure. It may be beneficial to consider visceral obesity indices, which are modifiable healthcare indices and could be useful in identifying patients at risk. Early diagnosis of this condition may allow early intervention and optimal management. This study had several limitations. The study was retrospective in design, and we did not assess for small fiber neuropathy, which is known to develop first before large fiber damage in the course of DPN. In addition, we did not have questionnaires or screening tools to support the diagnosis of DPN. The cross-sectional aspect of the study, accompanied by limited access to comorbid data (e.g., demographics, length of DM, lifestyle, and comorbidities) limited the outcome of the study. In conclusion, this study provides retrospective evidence of a clear link between visceral obesity (using BMI, LAP, VAI, and AIP) and DPN risk. This data emphasizes that the measurement of visceral fat and surplus lipids must be considered when evaluating DPN risk in those with T2DM. We emphasize that continued monitoring of visceral obesity metrics is warranted. Future large-scale, prospective studies are needed to provide a more definitive statement on this association.