Association between intraneural blood flow of the sciatic nerve, peripheral neuropathies, and foot ulcers in patients with diabetes

Introduction

Diabetic foot ulcers (DFUs) are defined by a disruption in the epidermis extending into the dermis in individuals with diabetes (1). These ulcers are a common and severely debilitating outcome of poorly controlled diabetes. Among the approximately 537 million people globally with diabetes, 19–34% are expected to develop a DFU during their lifetime (1). Of those affected, 14–24% may require lower extremity amputation, and 10% may succumb within a year of their initial DFU diagnosis (1,2). DFUs also impose a significant financial strain on healthcare systems, accounting for 20–40% of resources allocated to diabetes care (3). Therefore, the early prediction of DFU occurrences is critical for timely intervention, potentially reducing treatment costs and improving the quality of life for patients with diabetes.

Diabetic peripheral neuropathy (DPN) is the primary etiological factor in 86% of DFU cases (4). Early identification of DPN could therefore play a pivotal role in delaying or preventing its progression, thereby decreasing the likelihood of DFU development. DPN, a prevalent complication of diabetes, is marked by the progressive deterioration of nerve function (5), affecting approximately 50% of patients with diabetes (6). Identifying high-risk DPN patients could provide a novel approach to predicting DFUs at an earlier stage.

Nerve ultrasound is a non-invasive, accessible, and cost-effective technique that facilitates direct visualization of peripheral nerve structures and pathological changes (7-10). Its utility in diagnosing DPN, particularly for early detection, has been well-documented (11-17). Recently, power Doppler ultrasound imaging (a variant of ultrasound that can detect slow blood flow in human tissues and organs) has been used in predicting DPN in patients with diabetes by detecting intraneural blood flow (INBF) (18-20). Under normal conditions, peripheral nerves typically exhibit no detectable Doppler signal. However, the presence of INBF indicates pathological changes within the nerve (10,20,21). Previous studies have reported an association between hypervascularity in the median and tibial nerves and DPN in patients with diabetes (19,20). However, no studies have yet explored the use of power Doppler ultrasound to assess INBF in the sciatic nerve (SN_inbf) for predicting DPN in patients with diabetes and DFU in those with DPN. It was hypothesized that SN_inbf detected by power Doppler ultrasound may be correlated with DPN in patients with diabetes and DFU in those with DPN. Thus, this study aimed to examine the relationship between SN_inbf, as detected by nerve power Doppler ultrasound, and the presence of DPN and DFUs in patients with diabetes, using electroneurophysiological examination results as the reference standard for diagnosing DPN. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2773/rc).

Methods

Patient selection

This retrospective cross-sectional study received approval from The First Affiliated Hospital of Wenzhou Medical University (No. 2021-R027-01). The study was registered on Chinese Clinical Trial Registry (www.chictr.org.cn) (NCT No. ChiCTR2100047334). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The requirement for informed consent was waived due to the retrospective nature of the analysis.

The study was conducted on patients with diabetes who underwent sciatic nerve (SN) ultrasound examinations between June 2018 and July 2024. The inclusion criteria encompassed the following: (I) hospitalization for type 2 diabetes mellitus; (II) electromyogram examination due to suspected nerve injury; (III) bilateral lower limb SN ultrasound (including both grey-scale and Doppler imaging) performed by a single radiologist; and (IV) completion of both nerve ultrasound and electromyogram examinations within one week. The exclusion criteria were as follows: (I) presence of trauma or foot deformities; (II) absence of hospitalization; (III) unilateral lower limb SN ultrasound or unclear ultrasound imaging (including grey-scale and power Doppler) or repeated ultrasound examination; (IV) lack of electromyogram examination; (V) electrolyte disturbances such as hypokalemia; or (VI) peripheral neuropathy unrelated to diabetes, including alcohol-related myopathy. The selection process for patients with diabetes in this study is depicted in Figure 1.

Figure 1 Flowchart outlining the criteria for selecting patients with diabetes.

Clinical data collection and grouping

Baseline data, including demographic characteristics, anthropometric measurements, and history of conditions such as lower limb peripheral arterial disease (PAD) were retrospectively extracted from individual medical records, as previously documented (22). Laboratory parameters, including glucose, albumin, glycated hemoglobin (HbA1c), hemoglobin (Hb), creatinine, total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C), were also gathered. Body mass index (BMI) was calculated as weight divided by height squared (kg/m²).

Patients were stratified into two groups: those with diabetes and DPN (DPN group) and those with diabetes without DPN (non-DPN group), based on the presence or absence of DPN. Within the DPN group, patients were further categorized into DFU and non-DFU subgroups, depending on the presence or absence of DFUs.

Type 2 diabetes mellitus was diagnosed according to World Health Organization (WHO) criteria, whereas DPN was diagnosed based on electrophysiological criteria, as previously described (8,23). DFU was defined as a lesion involving the epidermis and at least a portion of the dermis in a diabetic patient, regardless of its location on unilateral or bilateral lower limbs (1). In this study, all patients with DFU were confirmed to have DPN via electromyogram examination.

Meanwhile, SNs of healthy controls (HCs) were taken as normal controls to determine whether SN_inbf can be detected in HC by power Doppler ultrasound. The inclusion criteria for HC were as follows: (I) without diabetes; (II) no other disease causing peripheral neuropathy; (III) bilateral lower limb SN ultrasound (including both grey-scale and Doppler imaging) performed by a single radiologist; (IV) completion of both nerve ultrasound and electromyogram examinations within one week; and (V) the results of electromyogram examinations is within the normal limitations.

Ultrasound imaging and image analysis

Ultrasound

Ultrasound examination of the distal SN, approximately 3 cm proximal to the bifurcation of the tibial and common fibular nerves, was performed with patients in a prone position, with the knee extended and the hip in a neutral position. Standard sonographic equipment (Logiq E9 and E11 ultrasound system; GE Healthcare, Wauwatosa, WI, USA) equipped with a 6–15 MHz high-frequency linear array transducer was utilized (8). The default machine settings (musculoskeletal mode) for gray-scale imaging and power Doppler were applied. All SN imaging was conducted by a radiologist with over six years of experience in musculoskeletal ultrasound, following a previously established protocol (8). The ultrasound probe was consistently maintained perpendicular to the nerve to ensure result reproducibility. Additionally, all examinations were performed under uniform environmental conditions, with the room temperature controlled at 26 ℃ using central air-conditioning to minimize temperature-related effects on vascular and nerve structures.

The nerve’s cross-sectional area (CSA) was measured on transverse images using manual tracing or the ellipsoid method directly over the epineurium. Power Doppler ultrasound was employed in both transverse and longitudinal planes to detect the presence or absence of INBF in the distal SN. During power Doppler usage, the color box was confined to the region of interest, and the gain was incrementally adjusted until noise from surrounding soft tissues was eliminated (8,24). Spectral Doppler, with or without angle correction, was applied when concerns arose regarding artifact-related pixels to verify whether Doppler signals indicated true arterial or venous flow (8,18,25).

Furthermore, ultrasound examinations of the lower limb arteries, including the common femoral, superficial femoral, popliteal, anterior tibial, posterior tibial, and dorsalis pedis arteries, were conducted within one week before or after the SN ultrasound. These examinations adhered to reference criteria for identifying severe PAD (8,26). Severe artery stenosis (70–99% stenosis) in the lower limbs was diagnosed if the flow velocity ratio between the stenosis site and the proximal corresponding artery was ≥4 (8,26,27). Complete occlusion of a lower limb artery was diagnosed when no flow signal was detected in the corresponding artery via duplex Doppler.

Ultrasonographic image analysis

The vascularization status of the SN was evaluated by determining the presence or absence of INBF. Given that diabetes-related nerve damage may affect multiple or isolated nerves in a distal-to-proximal pattern (5), INBF was identified as power Doppler signals demonstrating pulsatile blood flow over at least three cardiac cycles in the distal portion of the SN in any ultrasound plane (or vice versa, as shown in Figure 2) (25). Severe PAD was defined using Doppler ultrasound criteria, specifically as the presence of one or more severe arterial stenoses or occlusions in the lower limbs, to assess whether lower limb arterial ischemia impacts INBF in the SNs. Such ischemic conditions can induce hemodynamic changes in the arteries, thereby elevating the risk of lower limb ischemia.

Figure 2 Longitudinal power Doppler ultrasound scan illustrating the absence (A) and presence (B) of blood flow in the sciatic nerve in patients with diabetes. (A) A 73-year-old man with a 20-year history of diabetes, but without confirmed DPN via electrophysiological examination. Power Doppler ultrasound showed the absence of blood flow in the sciatic nerve (indicated by arrows). (B) A 67-year-old man with a 10-year history of diabetes, with DPN confirmed via electrophysiological examination. Power Doppler ultrasound showed the presence of blood flow (indicated by arrowheads) in the sciatic nerve (indicated by arrows). Spectral Doppler reveals blood flow from a small artery within the sciatic nerve (indicated by the thin arrow). DPN, diabetic peripheral neuropathy.

The CSA of the SN was measured using the ultrasound machine’s built-in calculation software during the nerve ultrasound examinations. Documentation of the CSA, SN_inbf, and the status of lower limb arteries was included in the ultrasonographic reports and corresponding images for each patient. To maintain consistency, the time interval between the nerve ultrasound and electromyogram examinations was limited to 1 week. The radiologist was blinded to the electromyogram results and clinical characteristics, including the presence of DFUs, before analyzing the nerve ultrasonographic images and reports. Consequently, the radiologist was unaware of whether patients with diabetes had been categorized into the DPN or DFU subgroups.

To ensure observer consistency, all images of the included SNs and lower limb arteries were independently reviewed retrospectively by a single author (S.P.C.), who has 12 years of experience in musculoskeletal ultrasound and 17 years of experience in vascular ultrasound.

Statistical analysis

Statistical analyses were conducted using the software SPSS 27.0 (IBM Corp., Armonk, NY, USA) and MedCalc 12.0 (MedCalc Software, Ostend, Belgium). Continuous variables were reported as mean ± standard deviation (SD) or median with interquartile range (IQR), whereas categorical variables were expressed as percentages. Comparisons between groups were made using the unpaired Student’s t-test for normally distributed continuous variables, the Mann-Whitney U test for non-normally distributed continuous variables, and the Chi-squared test or Fisher’s exact test for categorical variables. Statistical significance was established at a threshold of P<0.05. The relationship between the presence of SN_inbf, DPN, and foot ulcers in patients with diabetes was further explored using binary logistic regression. Potential confounding factors (age, sex, duration of diabetes, smoking status, drinking status, and BMI) as well as candidate variables with a P value <0.1 in univariate analysis were included in the multivariate model to analyze the association between SN_inbf presence and DPN or DFU. We initially conducted binary logistic regression analysis using two approaches: the stepwise mode with MedCalc statistical software and the backward mode with SPSS software. This analysis was aimed at identifying significant independent factors associated with DPN or DFU. To validate our findings and reduce multicollinearity, we subsequently employed the bootstrap method, utilizing 1,000 samples in SPSS software to increase the patient sample size. The results obtained from the bootstrap method were considered the final outcomes for the logistic regression analysis.

The diagnostic performance of SN_inbf for identifying DPN in patients with diabetes and DFU in patients with DPN was assessed using the area under the receiver operating characteristic curve (AUROC). The accuracy of SN_inbf in diagnosing DPN or DFU was reported as AUROC values with 95% confidence intervals (CIs). Differences in the AUROC values for diagnosing DPN or DFU between SN_inbf alone, atherosclerosis (or severe PAD) alone, and the combination of SN_inbf with atherosclerosis (or severe PAD) were evaluated using the Z-test.

Results

Comparison of clinical and biochemical characteristics and ultrasound findings of the SN between patients with and without DPN (group analysis)

Clinical and biochemical characteristics in patients with and without DPN

The baseline clinical and biochemical characteristics of the diabetic patient group are detailed in Table 1. The study included 113 patients with diabetes, comprising 71 males and 42 females, with an average age of 63.81±12.31 years (range, 29–90 years). The participants were divided into two groups: 82 patients in the DPN group and 31 in the non-DPN group. No significant differences were found between the DPN and non-DPN groups in terms of age, sex distribution, and BMI (all P>0.05). However, patients with DPN exhibited lower estimated glomerular filtration rates, reduced blood albumin, Hb, TC, TG, and LDL-C levels, a longer duration of diabetes, and a higher prevalence of atherosclerosis, diabetic retinopathy, diabetic nephropathy, and severe PAD compared to those without DPN.

Conventional and power Doppler ultrasound findings of the SN

Table 1 also outlines the ultrasound findings of the SN in patients with and without DPN. Among the 113 patients with diabetes, 53.1% (60/113) showed detectable blood flow in the SNs. Specifically, SN_inbf was observed in 55 (67.1%) patients with DPN—37 had bilateral involvement, nine in the right SN, and nine in the left. In contrast, only 5 (16.1%) patients without DPN showed SN_inbf, with three cases involving both nerves, one in the right nerve, and one in the left. Univariate analysis revealed that the presence of SN_inbf was significantly higher in the DPN group compared to the non-DPN group [67.1% (55/82) vs. 16.1% (5/31); P=0.001] (Figure 3). However, no significant difference in the CSA of the SNs was observed between the two groups (P>0.05).

Figure 3 The bar charts compare SN_inbf between patients with DPN and those without DPN among patients with diabetes (A) and between patients with DFU and those without DFU among DPN patients (B). (C-E) Additionally, a 72-year-old man with a 3-month history of a left foot ulcer and a 10-year history of diabetes, confirmed to have DPN via electrophysiological examination, underwent a power Doppler ultrasound scan. (A) The bar chart shows that DPN patients have higher SN_inbf than non-DPN patients. (B) Patients with both DFU and DPN exhibit increased SN_inbf compared to non-DFU patients with DPN. (C,D) Intraneural blood flow in the left sciatic nerve (indicated by white arrowheads and arrows). (E) Confirmed blood flow from an artery within the left sciatic nerve using spectral Doppler (indicated by the white arrow). DFU, diabetic foot ulcer; DPN, diabetic peripheral neuropathy; SN_inbf, intraneural blood flow of sciatic nerve.

The association among the presence of SN_inbf, atherosclerosis, and DPN using adjusted logistic regression and using ROC curve for diagnosing DPN

The selected variables, including BMI, diabetes duration, atherosclerosis, severe PAD, glucose, estimated glomerular filtration rate, albumin, Hb, TC, TG, LDL-C, and SN_inbf, were subjected to adjusted logistic regression analysis following univariate assessment (Table 2). In the multivariate analysis, the stepwise method revealed that SN_inbf [odds ratio (OR): 7.399; 95% CI: 1.959–27.944; P=0.001], atherosclerosis (OR: 8.304; 95% CI: 2.323–29.685; P<0.001), TC (OR: 0.648; 95% CI: 0.428–0.980; P=0.023), and the duration of diabetes (OR: 1.008; 95% CI: 1.000–1.016; P=0.045) were identified as independent significant factors associated with DPN. In the backward method, a significantly higher prevalence of SN_inbf was found in patients with DPN compared to those without it (OR: 7.399; 95% CI: 1.959–27.944; P=0.003). Additionally, individuals with DPN exhibited a higher prevalence of atherosclerosis than those without DPN (OR: 8.304; 95% CI: 2.323–29.685; P=0.001). TC was also determined to be an independent significant factor associated with DPN (OR: 0.648; 95% CI: 0.428–0.980; P=0.040). Further analysis using the bootstrap sampling method confirmed that only SN_inbf and atherosclerosis were independent significant factors associated with DPN (P=0.004 and P=0.003, respectively) (Table 2).

Receiver operating characteristic (ROC) curve analysis demonstrated that the AUROC values for SN_inbf, atherosclerosis, and the combination of SN_inbf and atherosclerosis were 0.755 (95% CI: 0.665–0.831, P<0.0001), 0.791 (95% CI: 0.705–0.862, P<0.0001), and 0.712 (95% CI: 0.619–0.793, P<0.0001), respectively, for diagnosing DPN in patients with diabetes. Table 3 provides a summary of the diagnostic indices for SN_inbf, atherosclerosis, and their combination in diagnosing DPN. When using SN_inbf to diagnose DPN in patients with diabetes, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were 67.07% (55/82), 83.87% (26/31), 91.7% (55/60), 49.1% (26/53), and 71.7% (81/113), respectively. Atherosclerosis alone was shown to be the most accurate method for diagnosing DPN, whereas the combination of SN_inbf and atherosclerosis yielded a PPV of 95.2% for DPN diagnosis. When comparing AUROC values for diagnosing DPN, a statistically significant difference was observed between the AUROC for atherosclerosis and that for the combination of SN_inbf and atherosclerosis (P=0.028); however, no significant difference was found between the AUROC values for SN_inbf and atherosclerosis (P=0.495) (Table 4; Figure 4).

Figure 4 ROC curves for SN_inbf, atherosclerosis, and SN_inbf combined with atherosclerosis for the differential diagnosis of DPN from diabetes. AUROCs were 0.755 (95% CI: 0.665–0.831, P<0.0001) for SN_inbf, 0.791 (95% CI: 0.705–0.862, P<0.0001) for atherosclerosis, and 0.712 (95% CI: 0.619–0.793, P<0.0001) for SN_inbf combined with atherosclerosis. The circle, square, and triangle in the figure represent the points obtained maximal Youden index for diagnosis DPNs in patients with diabetes using the presence of atherosclerosis, SN_inbf, and SN_inbf combined with atherosclerosis, respectively. AUROCs, areas under the ROC curve; CI, confidence interval; DPN, diabetic peripheral neuropathy; ROC, receiver operating characteristic; SN_inbf, intraneural blood flow of sciatic nerve.

Comparison of clinical, biochemical characteristics, and ultrasound findings of the SN between patients with DFUs and those without DFUs but with DPN (subgroup analysis)

Clinical and biochemical characteristics in patients with DFUs and those without DFUs but with DPNs

Among the 82 patients with DPN, 21 were categorized into the non-DFU subgroup, whereas the remaining 61 comprised the DFU subgroup. Within the DFU subgroup, 55 patients had ulcers in unilateral lower limbs (30 on the left side and 25 on the right), and six patients presented with bilateral ulcers. Among 61 patients with DFUs, 34.4% (21/61) were purely neuropathic, whereas 65.6% (40/61, 40 patients also with PAD) were neuroischemic. According to Wagner’s grading system, 4 cases were classified as grade I, 32 as grade II, 11 as grade III, and 14 as grade IV. Table 5 presents the baseline clinical and biochemical characteristics of patients with DPN. Age, sex distribution, and BMI were comparable between those with and without DFU (all P>0.05). However, patients with both DFU and DPN exhibited a longer duration of diabetes, a higher prevalence of severe PAD, and lower levels of estimated glomerular filtration rate, blood albumin, Hb, and HDL-C compared to those with DPN but without DFU.

Conventional and power Doppler ultrasound findings of the SN and prediction of DFUs

Table 5 also presents the ultrasound findings of the SNs in patients with DPN, comparing those with and without DFU. Among these patients, 48 (78.7%) with DFU exhibited SN_inbf, with 35 cases involving bilateral SNs, six in the right nerve, and seven in the left. In contrast, only 7 (33.3%) patients without DFU showed SN_inbf, with two cases bilaterally, three in the right nerve, and two in the left. Patients with DFU demonstrated a significantly higher presence of SN_inbf compared to those without DFU (P=0.0285) (Figures 3,5).

Figure 5 A 64-year-old man with a one-week history of a left foot ulcer (A, white arrow) and a 20-year history of diabetes mellitus. Electromyogram examination confirmed DPN. Power Doppler ultrasound in a transverse scan revealed intraneural blood flow (white arrowheads) in both the left (B) and right (C) sciatic nerves (white arrows). (D) Longitudinal power Doppler ultrasound also demonstrated intraneural blood flow (black arrow) in the right sciatic nerve (white arrows). (E) Spectral Doppler further confirmed blood flow from a small artery within the right sciatic nerve (white arrow). DPN, diabetic peripheral neuropathy.

Within the DPN group, patients with DFU had a smaller CSA of the SNs in both lower limbs compared to those without DFU. However, no significant difference in CSA was found between the left and right lower limbs in patients with and without DFU (P=0.315 and P=0.223, respectively).

The association between the presence of SN_inbf, severe PAD and DFU using adjusted logistic regression, and using the ROC curve for diagnosing DFU

The selected variables, including diabetes duration, severe PAD, diabetic retinopathy, estimated glomerular filtration rate, albumin, Hb, HDL-C, CSA of the SN, and SN_inbf, were analyzed using adjusted logistic regression following univariate analysis. In the multivariate analysis, both the stepwise and backward methods indicated that patients with DFU had a significantly higher rate of SN_inbf compared to those without DFU, with an OR of 12.255 (95% CI: 2.881–52.135; all P<0.001). Furthermore, individuals with DFU exhibited a higher prevalence of PAD than those without DFU, with an OR of 14.889 (95% CI: 3.159–70.175; all P<0.001). Additionally, diabetic retinopathy emerged as another significant independent factor associated with DFU, with an OR of 6.530 (95% CI: 0.157–27.167; P=0.010). A further analysis using the bootstrap sampling method showed that only SN_inbf and PAD were independent significant factors associated with DFU (all P=0.001) (Table 6).

ROC curve analysis revealed that the AUROC values for SN_inbf, severe PAD, and the combination of SN_inbf and severe PAD were 0.727 (95% CI: 0.617–0.819, P=0.0001), 0.733 (95% CI: 0.623–0.824, P<0.0001), and 0.754 (95% CI: 0.647–0.843, P<0.0001), respectively, for predicting DFU in patients with DPN.

Table 7 summarizes the diagnostic indices for SN_inbf, severe PAD, and their combination in predicting DFU. When using SN_inbf to predict DFU occurrence in patients with DPN, the sensitivity, specificity, PPV, NPV, and accuracy were 78.69% (48/61), 66.67% (14/21), 87.3% (48/55), 51.9% (14/27), and 75.6% (62/82), respectively. When employing severe PAD as a predictor, the sensitivity and specificity were 65.6% (40/61) and 80.95% (17/21), respectively. The combination of SN_inbf and severe PAD achieved 100% specificity and 100% PPV for diagnosing DFU. However, when comparing the AUROC values for diagnosing DFU, no statistically significant differences were found among the AUROC values for SN_inbf, severe PAD, and their combination (all P>0.05) (Table 8; Figure 6).

Figure 6 ROC curves for SN_inbf, severe PAD, and SN_inbf combined with severe PAD for the differential diagnosis of DFU from DPN. AUROC 0.727 (95% CI: 0.617–0.819, P=0.0001) for SN_inbf, 0.733 (95% CI: 0.623–0.824, P<0.0001) for severe PAD, and 0.754 (95% CI: 0.647–0.843, P<0.0001) for SN_inbf combined with severe PAD. The square, circle, and triangle in the figure represent the points obtained maximal Youden index for diagnosing DFUs in patients with DPNs using the presence of SN_inbf, severe PAD, and SN_inbf combined with severe PAD, respectively. AUROC, areas under the ROC curve; CI, confidence interval; DFU, diabetic foot ulcer; DPN, diabetic peripheral neuropathy; PAD, peripheral arterial disease; ROC, receiver operating characteristic; SN_inbf, intraneural blood flow of sciatic nerve.

Whether SN_inbf can be detected in healthy control individuals by power Doppler ultrasound (HC analysis)

A total of 19 HCs (38 SNs; mean age, 58.68±15.47 years; 11 women) were included in the present study. No significant differences in sex and age were found between the patients with diabetes (mean age, 63.81±12.31 years; 42 women) and HCs (P=0.108 and P=0.187, respectively). Among those HCs (38 SNs), no SN_inbf was detected by power Doppler ultrasound.

Discussion

In the present study, SN_inbf was not detected by power Doppler ultrasound in HCs but SN_inbf was detected in patients with diabetes and with DPN. This finding aligns with previous studies (10,20), which indicated that SNs typically exhibit no detectable Doppler signal under normal conditions and detectable SN_inbf indicates pathological changes within SN. Furthermore, in the present study, SN_inbf was not only significantly and positively associated with DPN in patients with diabetes but also with foot ulcers in those with DPN. Subsequent adjusted logistic regression analysis confirmed that SN_inbf serves as an independent factor positively linked to DPN in patients with diabetes and to DFUs in those with DPN. Additionally, ROC curve analysis demonstrated that SN_inbf had high sensitivity and specificity for identifying DPN in patients with diabetes and predicting DFUs in those with DPN. Notably, the combination of SN_inbf and severe PAD achieved the highest specificity (100%) and PPV (100%) for diagnosing DFU. To our knowledge, these associations and findings are reported here for the first time.

Few studies have focused on INBF in patients with diabetes and DPN. Previous research has identified the presence of INBF in the median and tibial nerves of patients with and without DPN, but not in the SN (12,19,20). Moreover, our study reported a higher rate of blood flow detection within the SNs [53.1% (60/113)] compared to previous studies [28% (21/75), P=0.010] (20), which might be attributed to differences in the nerve distributions assessed.

The mechanisms underlying the associations between INBF and diabetes or DPN are not yet fully understood. Endoneurial microangiopathy, including capillary basement membrane thickening, pericyte degeneration, and endothelial cell hyperplasia—resulting from chronic hyperglycemia in patients with diabetes—has been recognized as a fundamental pathological change leading to diabetic nerve impairment (20,28). These pathological changes can cause intraneural ischemia, which may induce alterations in intraneural microcirculation, impair neural perfusion, and trigger compensatory hypervascularity in patients with diabetes (12,19,20). Compared to non-DPN-diagnosed diabetic individuals, those with DPN may exhibit more pronounced vessel hypervascularity in response to endoneurial ischemia, potentially due to longer diabetes duration (5,20). These vascular alterations in nerves affected by microangiopathy in patients with diabetes or those with DPN can be more readily detected by power Doppler ultrasound.

Subgroup analysis in our study identified SN_inbf as an independent factor positively associated with DFUs in patients with DPN. Specifically, a higher prevalence of SN_inbf was observed in patients with both DFU and DPN compared to those with DPN but without DFU. The elevated SN_inbf in DFU patients with DPN may be attributed to several factors: initially, DPN primarily affects small fiber nerves such as the sural nerve, leading to symmetrical paraesthesia, sensory loss, and pain hypersensitivity in the early stages (1). As the disease progresses, DPN can extend to larger nerves, including the tibial and common peroneal nerves. Ultimately, in advanced stages, DPN may impact the largest nerve in the human body, the SN (5).

DPN is a microvascular complication characterized by damage to the nerve microvasculature, including endothelial dysfunction, capillary abnormalities, thickening of the basement membrane, reduced endoneurial perfusion, and ischemia (5,29). These pathological changes in the SN may cause ischemia, potentially leading to DFU in patients with DPN. The increased compensatory INBF in the SN, detectable by power Doppler ultrasound, may be a response to these ischemic conditions. This is supported by previous research indicating that nerve hypervascularity is common in individuals with moderate to severe neuropathy (20), aligning with our findings of increased SN hypervascularity in DFU patients with DPN compared to those without DFU. In summary, the SN, being the largest nerve in the human body, is typically affected in the later stages of DPN. Moreover, nerve blood flow is more commonly observed in individuals with moderate to severe neuropathy (20). Therefore, the presence of SN_inbf may indicate advanced or late-stage DPN, which could predispose patients to DFU. However, further studies are required to confirm this hypothesis. Moreover, previous studies (30-32) have shown that diabetes and DPNs are linked to injuries in the SN in animal models, which involve changes in blood flow within the nerve. These findings from animal models support the current study’s observation that patients with DPN experience SN injuries, which can be detected as changes in INBF using power Doppler ultrasound. However, to the best of our knowledge, very few studies have investigated the relationship between DFU and changes in INBF of the SN in animal models. Additionally, the underlying causes of DFUs can be classified as purely neuropathic (35%), purely ischemic (15%), and mixed neuroischemic (50%). PAD is a significant factor associated with DFUs. PAD involves the narrowing or blockage of blood vessels in the lower extremities, leading to decreased blood flow and chronic limb ischemia, resulting in DFUs. When patients have DPN alongside PAD, the risk of developing DFUs increases due to ischemic conditions and nerve injury in the lower extremities. In the present study involving 61 DFU cases, 34.4% (21/61) were classified as purely neuropathic, whereas 65.6% (40/61, including those with PAD) were identified as neuroischemic. Since purely ischemic DFUs were not included in our study, the incidence of neuroischemic DFUs (65.6%) was higher than the previously reported figure of 50%. However, this finding aligns with earlier reports indicating that mixed neuroischemic DFUs are the most common type. Additionally, the 34.4% incidence of purely neuropathic DFUs in our study is consistent with earlier reports (35%), suggesting that our distribution of DFU cases mirrors the broader population distributions found in the existing literature (1).

Furthermore, the study demonstrated that using SN_inbf yielded high AUROC values (0.755 for diagnosing DPN in patients with diabetes and 0.723 for diagnosing DFU in patients with DPN). The sensitivity and specificity for diagnosing DPN were 67.1% and 83.9%, respectively, whereas for diagnosing DFU, they were 76.7% and 68.2%, respectively. These results indicate that detecting SN_inbf via nerve power Doppler ultrasound could be effectively applied in clinical practice for identifying DPN in patients with diabetes or predicting DFUs in those with DPNs.

Beyond its diagnostic utility, SN_inbf was also identified as an independent predictor for DPN, whereas atherosclerosis and severe PAD were independently associated with DPN and DFU, respectively, in subgroup analyses. Given that hyperlipidemia is a primary cause of atherosclerosis, our findings align with previous studies (5) which suggest that hyperlipidemia can lead to reduced nerve perfusion and axonal degeneration, resulting in nerve microangiopathy and subsequent DPN. In our study, atherosclerosis demonstrated higher accuracy for predicting DPN, and the combination of SN_inbf with atherosclerosis yielded a 95.2% PPV for diagnosing DPN. However, SN_inbf and atherosclerosis showed similar diagnostic effectiveness for predicting DPN. Additionally, subgroup analysis confirmed that ischemia (severe PAD) and neuropathy (SN_inbf) are key underlying factors in the development of DFU (1). Although previous studies have not clarified the extent of lower limb artery ischemia or nerve impairment required to cause DFU, our findings suggest that a combination of SN_inbf and severe PAD achieves 100% specificity and 100% PPV for diagnosing DFU. Further validation of these findings is necessary.

These results have significant clinical implications. First, detecting SN_inbf through nerve power Doppler ultrasound shows great promise as a predictive tool for identifying DPN in patients with diabetes and predicting DFUs in those with DPN. Second, SN_inbf can be used to monitor patients undergoing treatment for DFUs or DPN. If the previous presence of SN_inbf is not detected using power Doppler ultrasound in patients with DPN or DFU after treatment, it may indicate that the treatment has been effective. Conversely, the remaining presence of SN_inbf might suggest that the treatment is ineffective. Third, SN_inbf emerges as a reliable biomarker for assessing diabetic neuropathy, particularly in identifying individuals at high risk for developing DFUs. It may also serve as a potential target for therapeutic interventions aimed at preventing the progression from DPN to DFUs.

This study possesses several strengths. First, all patients underwent electromyogram examinations, providing a more accurate diagnosis of DPN compared to clinical diagnosis alone. Second, power Doppler ultrasound, known for its high sensitivity in detecting slow blood flow, is widely utilized and can be easily applied in clinical settings to predict DPN in patients with diabetes or DFUs in those with DPN. Finally, all ultrasound examinations and analyses of the SN were conducted by an experienced radiologist, potentially enhancing the consistency and reliability of the results.

This study had several limitations. First, it is a retrospective analysis conducted at a single institution, which introduces the potential for bias in data collection. The study focused specifically on inpatients with diabetes who underwent both bilateral SN ultrasound and electromyography. Therefore, the findings may not represent the broader population of patients with diabetes. Consequently, the effectiveness of SN_inbf for diagnosing DPN in outpatient settings, as well as in patients with DFU who also have DPN, still needs to be verified. Additionally, more multicenter retrospective or prospective studies with larger patient populations are necessary to confirm our results. Second, measurements were limited to the distal portion of the SN rather than the entire nerve. However, given that the SN is the largest nerve in the human body and typically the last to be affected by DPN in the lower limbs, focusing on the distal portion may still yield valuable insights into the extent of nerve involvement. Third, the study dichotomized PAD in the lower limbs as either present or absent. Nevertheless, since all patients with DFU had confirmed DPN through electrophysiological examination, the potential impact of PAD on DFU might be minimized within this group. Fourth, echogenicity changes in the SNs were not utilized to predict DFU in patients with DPN. Although echogenicity changes have been established as reliable indicators for predicting peripheral neuropathies, with most affected nerves showing such alterations (10), their utility in predicting DFU in DPN patients may be limited. Fifth, the observed high specificity (100%) for diagnosing DFU using the combination of SN_inbf and severe PAD may indicate potential overfitting in our regression models. This could be due to the retrospective selection of cases and the relatively small sample size. Such factors could limit the generalizability of our findings, so it is essential to interpret the diagnostic performance metrics with caution. Sixth, although all HCs underwent electromyogram examinations that fell within normal limits, the relatively small control population (only 19 cases) may limit its reliability. A larger control population may be needed to confirm the present finding. Seventh, we did not assess inter-rater agreement in the ultrasound imaging analysis to ensure observer reliability. Nonetheless, the observer has over 17 years of experience with power Doppler ultrasound, which helps to ensure the accuracy of the imaging analysis results. We will conduct this comparative analysis with larger patient populations in further study. Additionally, the diagnosis of SN_inbfs often requires high-frequency ultrasound probes and high-level ultrasound machines, as they indicate the microcirculation of the SN (21). When untrained or inexperienced operators perform power Doppler ultrasound, there is a risk of failing to diagnose SN_inbf. However, learning to use power Doppler ultrasound can be straightforward. Training skilled ultrasound operators can help to address the current limitations and shortage of qualified power Doppler ultrasound technicians in primary care settings. This can improve the ability to predict DPN or DFU in patients with diabetes. Lastly, due to the retrospective design of our study and the small patient population, we did not compare the cost-effectiveness and ease of use of SN ultrasound with those of other diagnostic methods, such as neural electrophysiology. We will conduct this comparative analysis with larger patient populations in further study.

This study is the first to highlight the potential of SN_inbf as a predictor of DPN in patients with diabetes, particularly in forecasting DFU occurrence among those with DPN. This finding confirms the feasibility of using SN ultrasound for DFU diagnosis in DPN patients, opening new avenues for non-invasive imaging assessments of DFU in clinical settings. Further multicenter retrospective or prospective studies with larger patient populations are needed to confirm the efficacy of power Doppler ultrasound in evaluating SN_inbf for diagnosing DPN in patients with diabetes, particularly in predicting DFU development in those with DPN.

Conclusions

Detectable SN_inbf by power Doppler ultrasound indicates pathological changes within SN. Furthermore, the presence of SN_inbf is positively correlated not only with DPN in patients with diabetes but also with DFU in those with DPN. SN_inbf demonstrated high sensitivity and specificity for diagnosing DPN in patients with diabetes and DFU in patients with DPN. Remarkably, combining SN_inbf with severe PAD achieved 100% specificity and PPV for diagnosing DFU. SN ultrasound, particularly power Doppler ultrasound, offers a valuable method for evaluating SN_inbf, making it a potential tool for diagnosing DPN in patients with diabetes and DFU in those with DPN in clinical practice.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2773/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2773/dss

Funding: This research was supported by the Science and Technology Bureau Project of Wenzhou (No. Y2023114).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2773/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This retrospective cross-sectional study received approval from The First Affiliated Hospital of Wenzhou Medical University (No. 2021-R027-01). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The requirement for informed consent was waived due to the retrospective nature of the analysis.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. McDermott K, Fang M, Boulton AJM, Selvin E, Hicks CW. Etiology, Epidemiology, and Disparities in the Burden of Diabetic Foot Ulcers. Diabetes Care 2023;46:209-21. [Crossref] [PubMed]
2. Raja JM, Maturana MA, Kayali S, Khouzam A, Efeovbokhan N. Diabetic foot ulcer: A comprehensive review of pathophysiology and management modalities. World J Clin Cases 2023;11:1684-93. [Crossref] [PubMed]
3. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract 2010;87:4-14. [Crossref] [PubMed]
4. Faruque LI, Wiebe N, Ehteshami-Afshar A, Liu Y, Dianati-Maleki N, Hemmelgarn BR, Manns BJ, Tonelli M. Effect of telemedicine on glycated hemoglobin in diabetes: a systematic review and meta-analysis of randomized trials. CMAJ 2017;189:E341-64. [Crossref] [PubMed]
5. Eid SA, Rumora AE, Beirowski B, Bennett DL, Hur J, Savelieff MG, Feldman EL. New perspectives in diabetic neuropathy. Neuron 2023;111:2623-41. [Crossref] [PubMed]
6. Feldman EL, Nave KA, Jensen TS, Bennett DLH. New Horizons in Diabetic Neuropathy: Mechanisms, Bioenergetics, and Pain. Neuron 2017;93:1296-313. [Crossref] [PubMed]
7. Singh KP, Singh P, Gupta K. Reference values for the cross-sectional area of the normal sciatic nerve using high-resolution ultrasonography. J Ultrason 2021;21:e95-e104. [Crossref] [PubMed]
8. Chen SP, Ye TT, Hong J, Zhu H. Evaluation of Sciatic Nerve Stiffness Using Shear Wave Elastography in Patients with Unilateral Diabetic Foot Ulcers. Diagnostics (Basel) 2023;13:547. [Crossref] [PubMed]
9. Cocco G, Ricci V, Corvino A, Pacini P, Boccatonda A, Naňka O, Sensi SL, Caulo M, Delli Pizzi A. Ultrasound Imaging of the Sciatic Nerve. Ultraschall Med 2023;44:e263-73. [Crossref] [PubMed]
10. Pitarokoili K, Gold R, Fisse AL. Nerve ultrasound for the diagnosis and follow-up of peripheral neuropathies. Curr Opin Neurol 2023;36:373-81. [Crossref] [PubMed]
11. Dhanapalaratnam R, Issar T, Poynten AM, Milner KL, Kwai NCG, Krishnan AV. Diagnostic accuracy of nerve ultrasonography for the detection of peripheral neuropathy in type 2 diabetes. Eur J Neurol 2022;29:3571-9. [Crossref] [PubMed]
12. Huang H, Wu S. Application of High-Resolution Ultrasound on Diagnosing Diabetic Peripheral Neuropathy. Diabetes Metab Syndr Obes 2021;14:139-52. [Crossref] [PubMed]
13. He Y, Xiang X, Zhu BH, Qiu L. Shear wave elastography evaluation of the median and tibial nerve in diabetic peripheral neuropathy. Quant Imaging Med Surg 2019;9:273-82. [Crossref] [PubMed]
14. Wang C, Wang H, Zhou Y, Zhang S, Huang M. Evaluation of the clinical value of shear wave elastography for early detection and diagnosis of diabetic peripheral neuropathy: a controlled preliminary prospective clinical study. BMC Musculoskelet Disord 2022;23:1120. [Crossref] [PubMed]
15. Tandon A, Khullar T, Maheshwari S, Bhatt S, Narang S. High resolution ultrasound in subclinical diabetic neuropathy: A potential screening tool. Ultrasound 2021;29:150-61. [Crossref] [PubMed]
16. Huang H, Tang C, Li M, Huang J, Li Y, Wu S. Sensitivity and specificity of high frequency ultrasound score (DCEC) in diabetic peripheral neuropathy. J Diabetes Metab Disord 2022;21:1459-67. [Crossref] [PubMed]
17. Song Y, Zhang Y, Zhang Y, Hu B. Two-dimensional ultrasound and two-dimensional shear wave elastography on femoral and saphenous neuropathy in patients with type 2 diabetes mellitus. Front Neurol 2022;13:996199. [Crossref] [PubMed]
18. Yao B, Evans KD, Roll SC. Assessing the Potential for Error in Investigating Intraneural Vascularity: A Need for a Standardized Imaging Protocol. J Diagn Med Sonogr 2023;39:549-59. [Crossref] [PubMed]
19. Issar T, Walker S, Arnold R, Poynten AM, Endre ZH, Krishnan AV. Peripheral nerve morphology and intraneural blood flow in chronic kidney disease with and without diabetes. Muscle Nerve 2022;65:603-7. [Crossref] [PubMed]
20. Borire AA, Issar T, Kwai NC, Visser LH, Simon NG, Poynten AM, Kiernan MC, Krishnan AV. Sonographic assessment of nerve blood flow in diabetic neuropathy. Diabet Med 2020;37:343-9. [Crossref] [PubMed]
21. Ricci V, Ricci C, Cocco G, Gervasoni F, Donati D, Farì G, Özçakar L. Histopathology and high-resolution ultrasound imaging for peripheral nerve (injuries). J Neurol 2022;269:3663-75. [Crossref] [PubMed]
22. Hong J, Hu X, Liu W, Qian X, Jiang F, Xu Z, Shen F, Zhu H. Impact of red cell distribution width and red cell distribution width/albumin ratio on all-cause mortality in patients with type 2 diabetes and foot ulcers: a retrospective cohort study. Cardiovasc Diabetol 2022;21:91. [Crossref] [PubMed]
23. Tang XF. Electromyographic findings in 310 normal Chinese. Zhonghua Yi Xue Za Zhi 1984;64:91-6.
24. Gamil AM, Shalaby MH, Shehata KA, El Deeb AM. Value of Grayscale and Power Doppler High-Resolution Ultrasound in Assessment of Patients with Clinically Suspected Carpal Tunnel Syndrome. J Ultrasound Med 2020;39:1155-62. [Crossref] [PubMed]
25. Borire AA, Visser LH, Padua L, Colebatch JG, Huynh W, Simon NG, Kiernan MC, Krishnan AV. Utility of maximum perfusion intensity as an ultrasonographic marker of intraneural blood flow. Muscle Nerve 2017;55:77-83. [Crossref] [PubMed]
26. Elgzyri T, Ekberg G, Peterson K, Lundell A, Apelqvist J. Can duplex arterial ultrasonography reduce unnecessary angiography? J Wound Care 2008;17:497-500. [Crossref] [PubMed]
27. Gao M, Hua Y, Zhao X, Jia L, Yang J, Liu B. Optimal Ultrasound Criteria for Grading Stenosis of the Superficial Femoral Artery. Ultrasound Med Biol 2018;44:350-8. [Crossref] [PubMed]
28. Cameron NE, Eaton SE, Cotter MA, Tesfaye S. Vascular factors and metabolic interactions in the pathogenesis of diabetic neuropathy. Diabetologia 2001;44:1973-88. [Crossref] [PubMed]
29. Quattrini C, Jeziorska M, Boulton AJ, Malik RA. Reduced vascular endothelial growth factor expression and intra-epidermal nerve fiber loss in human diabetic neuropathy. Diabetes Care 2008;31:140-5. [Crossref] [PubMed]
30. Kappelle AC, Biessels GJ, Van Buren T, Erkelens DW, De Wildt DJ, Gispen WH. Effects of nimodipine on sciatic nerve blood flow and vasa nervorum responsiveness in the diabetic rat. Eur J Pharmacol 1993;250:43-9. [Crossref] [PubMed]
31. Sameni H, Panahi M. The Effect of Co-administration of 4-Methylcatechol and Progesterone on Sciatic Nerve Function and Neurohistological Alterations in Streptozotocin-Induced Diabetic Neuropathy in Rats. Cell J 2011;13:31-8.
32. Khan I, Preeti K, Kumar R, Khatri DK, Singh SB. Activation of SIRT1 by silibinin improved mitochondrial health and alleviated the oxidative damage in experimental diabetic neuropathy and high glucose-mediated neurotoxicity. Arch Physiol Biochem 2024;130:420-36. [Crossref] [PubMed]