Close relationship between β2‐microglobulin and type 2 diabetes‐associated peripheral neuropathy

ABSTRACT

Aims/Introduction

This study aimed to elucidate the correlation between β2‐microglobulin (β2‐MG) and peripheral neuropathy in type 2 diabetes mellitus (T2DM).

Materials and Methods

Diabetic peripheral neuropathy (DPN) was defined on the nerve conduction study (NCS). The general information and relevant clinical data of patients with T2DM were retrospectively analyzed using logistic regression to identify independent risk factors. The discriminative ability of individual and combined risk factors was evaluated using the area under the receiver operating characteristic curve (AUC).

Results

A total of 1,219 patients with T2DM who received medical care at Dongyang People's Hospital between January 2023 and December 2023 were included in this study. Of the 954 patients with abnormal NCS findings, 899 exhibited sensory abnormalities, whereas 492 exhibited motor abnormalities. Multivariate analysis showed that DPN was positively correlated with age (P < 0.001), β2‐MG (P = 0.022), uric acid (P = 0.028), and male sex (P < 0.001), but negatively correlated with albumin (P < 0.001). The combined discriminative power of these risk factors was 0.754, with β2‐MG showing the highest individual diagnostic performance (AUC = 0.710).

Conclusions

T2DM patients with advanced age, elevated β2‐MG and uric acid levels, and hypoalbuminemia have higher DPN possibilities; follow‐up confirmation with NCS may facilitate the early diagnosis and management of peripheral neuropathy.

Abbreviations

ADA

American Diabetes Association

AUC

area under the receiver operating characteristic curve

DPN

diabetic peripheral neuropathy

HbA1c

glycated hemoglobin

NCS

nerve conduction study

T2DM

type 2 diabetes mellitus

β2‐MG

β2‐microglobulin

INTRODUCTION

Diabetes is a global health challenge, affecting 529 million people in 2021 and is projected to affect over 1.3 billion people worldwide by 2050 ¹ . Diabetic peripheral neuropathy (DPN) is a prevalent chronic complication of diabetes. Over 11% of patients present with peripheral neuropathy during diabetes diagnosis, with more than half of them having asymptomatic diabetic peripheral neuropathy. Furthermore, DPN is associated with 50–75% of lower limb amputations in patients with diabetes ² , ³ , ⁴ . Therefore, DPN poses a major risk to increased hospitalization and mortality in patients with diabetes, representing a serious global public health concern.

In clinical practice, DPN is primarily assessed and diagnosed through clinical symptoms, targeted physical examination, neurological impairment assessment scales, and nerve conduction study (NCS). However, these approaches have notable limitations. For example, neurological impairment assessment scales are used for the quantitative evaluation of DPN prevalence and grading ⁵ ; nevertheless, their accuracy is influenced by the subjective perception of patients. It cannot objectively reflect the severity of the condition and is better suited for patients with symptomatic DPN. NCS is the gold standard for DPN diagnosis. However, its widespread use in epidemiological research and diabetic clinics is impeded by several factors, including patient discomfort from electrical stimulation, the demand for skilled operators, prolonged procedure times, labor‐intensive requirements, and high cost ⁶ . Therefore, a simple, safe, and rapid method for the early prediction and diagnosis of ADPN is warranted.

Under normal physiological settings, β2‐microglobulin (β2‐MG), a low‐molecular‐weight protein, is constitutively expressed at low levels. This protein has been extensively studied in malignant tumors, and its elevated expression has been associated with tumor proliferation and invasion ⁷ . In patients with diabetes, β2‐MG has been proposed as a marker for concurrent microangiopathy. Because DPN is a form of microangiopathy, macrophages control its onset by facilitating β2‐MG ⁸ , ⁹ . Therefore, this study analyzes the correlation between β2‐MG and DPN and proposes that β2‐MG may serve as a screening marker for early‐stage DPN. Hence, further research should be conducted to determine its predictive value in DPN.

MATERIALS AND METHODS

Data collection

Patients with diabetes who received medical care at Dongyang People's Hospital between January 2023 and December 2023 were assessed for eligibility based on the following eligibility criteria:

Inclusion criteria

Patients diagnosed with type 2 diabetes mellitus (T2DM) according to the American Diabetes Association (ADA) criteria for the diagnosis of diabetes (2024 edition) ¹⁰ .

Exclusion criteria

(1) Other causes of peripheral neuropathy, including genetic factors, drugs, surgery, and other metabolic diseases; (2) lumbar diseases and sequelae of cerebrovascular diseases; (3) severe organ dysfunction; (4) acute complications of diabetes, pregnancy, malignant tumor, trauma, concurrent acute or chronic inflammatory diseases (e.g., foot ulcer, infection), and psychiatric disorders; (5) thyroid abnormalities; (6) current use of methotrexate, carbamazepine, vitamin B6, folic acid, or antibiotics; (7) other diseases that cause β2‐MG elevation, such as infection or renal insufficiency.

Definition and collection of variables

Data were collected from inpatients with T2DM based on the Larger Health scientific research platform of Dongyang People's Hospital. Collected data included general information (name, age, sex, disease duration, and pre‐existing conditions) and clinical markers (glycated hemoglobin, blood lipids, C‐peptide levels, and NCS data) at admission. The diagnostic criteria for DPN and the interpretation of NCS findings were based on the ADA criteria for diagnosing diabetic peripheral neuropathy (2018 edition) ¹¹ .

Statistical analysis

Continuous variables are presented as median (interquartile range), and categorical variables as counts (percentages). Variables with more than 20% missing data were excluded, whereas those with less than 20% missing data were imputed using multiple imputation by chained equations. The complete function was used to generate a complete dataset. Groups were defined based on the dependent variable, and univariate analysis was conducted using the twogrps function from the CRC package in R (version 4.5.0). The possible linear correlations between β2‐microglobulin and continual variables were evaluated by Spearman test and visualized by ggplot2 packages in R. Variables with P‐values <0.05 were included in the multivariate logistic regression analysis. Independent risk factors were identified using backward regression analysis and multicollinearities between any two continual variables were evaluated by variance inflation factor in SPSS (version 26.0). The discriminative power of the risk factors, both individual and combined, was evaluated using the area under the receiver operating characteristic curve (AUC).

RESULTS

General patient information

A total of 1,219 patients with T2DM were included, comprising 441 women and 778 men. Among them, 265 patients had normal NCS findings, whereas 954 had abnormal results, including 899 with sensory abnormalities and 492 with motor abnormalities (Table 1).

Table 1.

Distribution of involved nerves in patients with diabetes

Sensory	Motor
	No		Yes
No	265	82.80%	55	17.20%
Yes	462	51.40%	437	48.60%

β2‐MG levels were positively associated with age, creatinine, uric acid, and urea with Spearman correlation coefficient between 0.3 and 0.6 (P < 0.001). In contrast, a negative correlation existed between β2‐MG levels and albumin (Spearman correlation coefficient −0.347, P < 0.001, Figure 1).

Figure 1.

Spearman correlation analysis between β2‐microglobulin and continual variables.

Risk factors for peripheral sensory nerve injury

Univariate analysis suggested that all 21 variables, except for glycated hemoglobin (HbA1c), high‐density lipoprotein cholesterol, neutrophils, and C‐reactive protein, were associated with peripheral sensory nerve injury in patients with T2DM (Table S1, all P < 0.05).

Multivariate logistic regression analysis suggested that peripheral sensory nerve injury was positively associated with age (OR = 1.05, 95% CI: 1.038–1.061, P < 0.001), β2‐MG levels (OR = 1.145, 95% CI: 1.026–1.279, P = 0.016), uric acid levels (OR = 1.002, 95% CI: 1.000–1.003, P = 0.047), and male sex (OR = 2.458, 95% CI: 1.827–3.308, P < 0.001), but negatively associated with albumin levels (OR = 0.924, 95% CI: 0.888–0.962, P < 0.001). No significant association was observed with alanine aminotransferase levels (OR = 0.998, 95% CI: 0.996–1.000, P = 0.108) (Table 2). The combined discriminative ability of these risk factors yielded an AUC of 0.755 (Figure 2), with β2‐MG showing the strongest individual predictive performance (AUC = 0.710).

Table 2.

Independent risk factors related to peripheral sensory neuropathy in patients with diabetes

Variables	OR (95% CI)	P
Age	1.05 (1.038, 1.061)	<0.001
Gender (male)	2.458 (1.827, 3.308)	<0.001
Beta‐2‐microglobulin	1.145 (1.026, 1.279)	0.016
Alanine aminotransferase	0.998 (0.996, 1)	0.108
Uric acid	1.002 (1, 1.003)	0.047
Albumin	0.924 (0.888, 0.962)	<0.001

Figure 2.

ROC curves of the independent risk factors related to peripheral sensory neuropathy in patients with T2DM.

Risk factors for peripheral motor nerve injury

Univariate analysis identified 11 variables that significantly differed between patients with and without peripheral motor nerve injury (Table S2, P < 0.05). Multivariate logistic regression analysis further indicated that peripheral motor nerve injury was positively associated with age (OR = 1.015, 95% CI: 1.005–1.024, P = 0.002), β2‐MG levels (OR = 1.083, 95% CI: 1.032–1.138, P = 0.001), uric acid levels (OR = 1.001, 95% CI: 1.000–1.003, P = 0.041), urea levels (OR = 1.148, 95% CI: 1.082–1.217, P < 0.001), and HbA1c levels (OR = 1.085, 95% CI: 1.031–1.142, P = 0.002), but negatively associated with albumin (OR = 0.937, 95% CI: 0.908–0.968, P < 0.001) and creatinine levels (OR = 0.991, 95% CI: 0.987–0.995, P < 0.001) (Table 3). The combined discriminative ability of these risk factors for motor nerve injury yielded an AUC of 0.672 (Figure 3), with β2‐MG showing the strongest individual predictive performance (AUC = 0.616).

Table 3.

Independent risk factors related to peripheral motor neuropathy in patients with diabetes

Variables	OR (95% CI)	P
Age	1.015 (1.005, 1.024)	0.002
Beta‐2 microglobulin	1.083 (1.032, 1.138)	0.001
HbA1c	1.085 (1.031, 1.142)	0.002
Uric acid	1.001 (1, 1.003)	0.041
Albumin	0.937 (0.908, 0.968)	<0.001
Creatinine	0.991 (0.987, 0.995)	<0.001
Urea	1.148 (1.082, 1.217)	<0.001

Figure 3.

ROC curves of the independent risk factors related to peripheral motor neuropathy in patients with T2DM.

Risk factors for peripheral neuropathy

Of the 16 variables identified as significant in the univariate analysis (Table S3), only five remained independently associated with DPN in the multivariate analysis (Table 4). In detail, peripheral neuropathy was positively associated with age (OR = 1.045, 95% CI: 1.033–1.057, P < 0.001), β2‐MG levels (OR = 1.182, 95% CI: 1.024–1.364, P = 0.022), uric acid levels (OR = 1.002, 95% CI: 1.000–1.003, P = 0.028), and male sex (OR = 1.904, 95% CI: 1.376–2.634, P < 0.001), whereas negatively associated with albumin (OR = 0.917, 95% CI: 0.878–0.958, P < 0.001). The combined discriminative ability of these risk factors yielded an AUC of 0.754 for detecting peripheral neuropathy (Figure 4), with β2‐MG demonstrating the strongest individual predictive performance (AUC = 0.710).

Table 4.

Independent risk factors related to peripheral neuropathy in patients with diabetes

Variables	OR (95% CI)	P
Age	1.045 (1.033, 1.057)	<0.001
Alcohol consumption	1.586 (0.969, 2.595)	0.066
Gender (male)	1.904 (1.376, 2.634)	<0.001
Beta‐2‐microglobulin	1.182 (1.024, 1.364)	0.022
Alanine aminotransferase	0.998 (0.996, 1)	0.054
Uric acid	1.002 (1, 1.003)	0.028
Albumin	0.917 (0.878, 0.958)	<0.001

Figure 4.

ROC curves of the independent risk factors related to peripheral neuropathy in patients with T2DM.

DISCUSSION

Peripheral neuropathy is a prevalent complication in patients with T2DM. In this study, peripheral sensory nerve injury was associated with age, β2‐MG levels, uric acid levels, male sex, and albumin levels. Peripheral motor nerve injury was related to age, β2‐MG, uric acid, urea, HbA1c, albumin, and creatinine levels. Peripheral neuropathy was associated with age, β2‐MG, uric acid, and albumin levels. Collectively, these indicators may allow for the early identification of peripheral neuropathy in patients with T2DM and support subsequent disease confirmation using NCS.

Diabetic nephropathy and DPN share many pathogenic pathways. β2‐MG contributes to renal impairment and serves as an early and sensitive marker for diabetic nephropathy ¹² . Our study demonstrated a positive correlation between serum β2‐MG levels and the incidence of DPN, encompassing both peripheral sensory and motor nerve injuries. Moreover, β2‐MG exhibited the highest discriminative power for detecting DPN and its subtypes. These results highlight a strong association between β2‐MG and DPN. Elevated plasma β2‐MG levels significantly reduce nerve conduction velocity ¹³ . This effect is attributed to oxidative stress and tissue ischemia in DPN, which cause neuronal injury and apoptosis, promote the release of proinflammatory cytokines, recruit macrophages, and stimulate β2‐MG release from circulating monocytes. To the best of our knowledge, this is the first study to demonstrate that β2‐MG may serve as a useful indicator for the early detection of DPN. However, it remains unknown whether early interventions targeted at reducing β2‐MG levels in patients with ADPN might delay disease progression. It was not possible to evaluate the correlation between β2‐MG levels and DPN progression because the research population was not followed up with.

In addition to β2‐MG, uric acid, age, and albumin were identified as independent risk factors for peripheral sensory and motor neuropathy. Although positive correlations between β2‐MG and uric acid and age were found in original data, there were no multi‐collinearities between enrolled continual variables for DPN in the logistic regression (Table S4). Furthermore, the elevated β2‐MG levels could be partially explained by impaired renal functions (higher levels of creatinine and urea and a lower level of albumin), but its contribution in identifying DPN remains still significant in the final regression analysis. Therefore, there might be additional mechanisms between enrolled variables and the pathogenesis of DPN. Uric acid induces oxidative stress and causes vascular endothelial cell injury. Together with microangiopathy, it has been implicated in DPN ¹⁴ . Increasing age is a risk factor for peripheral neuropathy, potentially because of age‐related mitochondrial dysfunction and reduced metabolic capacity, leading to the increased production of reactive oxygen species and subsequent damage to peripheral nerves ¹⁵ . Moreover, older adults often present with comorbidities, such as hypertension and hyperlipidemia, which can indirectly contribute to nervous system damage. Furthermore, the use of concomitant medications for chronic conditions may lead to drug‐induced neurotoxicity. Here, hypoalbuminemia was associated with a higher incidence of DPN. Albumin is a key antioxidant in the blood, and a decrease in serum albumin levels can exacerbate oxidative stress, promote chronic inflammation, and accelerate nerve fiber apoptosis ¹⁶ . Additionally, hypoalbuminemia increases endoneurocapillary permeability, leading to endoneural edema and exacerbated neurological ischemia and hypoxia ¹⁷ .

Apart from these common risk factors, male sex was identified as an independent risk factor for peripheral sensory nerve injury, whereas increased urea, elevated HbA1c, and decreased creatinine levels were independent risk factors for peripheral motor nerve injury in T2DM. Hyperglycemia can damage capillary endothelial cells, causing reduced nerve perfusion, hypoxia, and the accumulation of metabolic toxins ¹⁸ . It impairs small sensory peripheral nerves that depend on microvascular supply. In addition to a higher prevalence of drinking and smoking, which both further compromise microvascular integrity, men may have a higher incidence of sensory nerve damage attributed to lower estrogen levels, which reduce microvascular protection ¹⁹ , ²⁰ . Elevated urea and decreased creatinine levels reflect abnormalities in protein‐energy metabolism, which impair the energy supply to peripheral neurons. Compared with sensory nerves, peripheral motor nerves are more susceptible to these metabolic disturbances ²¹ , ²² . High urea concentrations can directly destabilize motor nerve cell membranes, whereas reduced creatinine levels indicate malnutrition and diminished muscle mass, potentially exacerbating insulin resistance and decreasing neurotrophic support ²³ . Finally, one noteworthy observation in patients with peripheral motor nerve injury was high HbA1c. Unlike peripheral sensory nerves, which are mostly unmyelinated or hypomyelinated, peripheral motor nerves have thick myelin sheaths. Hyperglycemia promotes the accumulation of advanced glycation end products in myelin and axons, directly impairing neurostructural protein function and exacerbating oxidative stress. This cascade activates the polyol pathway, causing osmotic imbalance, Schwann cell edema, and apoptosis, and the subsequent disruption of myelin integrity in peripheral motor nerves ²⁴ , ²⁵ .

This study has several limitations, including a small sample size, limited data, and a cross‐sectional design that makes it impossible to determine causality and might introduce bias. Moreover, other risk factors that might be related to DPN are not available in this study, including heavy alcohol consumption ²⁶ , duration diabetes ²⁷ , BMI ²⁸ , and obesity ²⁹ . Also, the included cases in this study seem severer regarding the high percentage of abnormality in peripheral sensory nerve injury, which might impair the genericity of our findings for other studies. Due to missing information of clinical symptoms for DPN, the association between abnormal NCS findings and symptoms could not be established in this study. Therefore, larger, multicenter, randomized studies are required to confirm these findings.

CONCLUSIONS

β2‐MG is closely related to the possibility of DPN in patients with T2DM. DPN screening should be prioritized for patients with T2DM who have advanced age, elevated β2‐MG and uric acid levels, and hypoalbuminemia. Subsequent confirmation with NCS may be a valuable strategy for the early diagnosis and management of peripheral neuropathy.

DISCLOSURE

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Approval of the research protocol: The study protocol was approved by the ethics committee and the institutional review board of the Dongyang People's Hospital (2024‐YX‐072) and conducted in accordance with the Declaration of Helsinki. Due to the retrospective nature of the study and the use of anonymized clinical data, the committee waived the need for individual informed consent. All patient data were de‐identified and analyzed anonymously to protect confidentiality.

Informed consent: N/A.

Registry and the registration no. of the study/trial: N/A.

Animal studies: N/A.

Supporting information

Table S1. Univariable analysis of factors related to peripheral sensory neuropathy in patients with diabetes.

Table S2. Univariable analysis of factors related to peripheral motor neuropathy in patients with diabetes.

Table S3. Univariable analysis of factors related to peripheral neuropathy in patients with diabetes.

Table S4. Multicollinearity test between continual variable in the multivariate logistic regression analysis by variance inflation factors.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.