Diabetic peripheral neuropathy (DPN) is a prevalent and progressive microvascular complication of diabetes.¹ Early detection and prevention are essential in reducing DPN-related morbidity and mortality. To do this, risk factors of DPN must be identified. According to previous epidemiologic studies^2,3 and intervention studies⁴, age, diabetes duration, smoking status, and components of metabolic syndrome including obesity are well-known risk factors for DPN.

There are five components of metabolic syndrome: central obesity, high blood pressure, low high-density lipoprotein (HDL) cholesterol, high triglycerides, and hyperglycemia.⁵ Patients with diabetes already have hyperglycemia, but other modifiable risk factors such as hypertension, dyslipidemia, or central obesity can be addressed. Central obesity with insulin resistance is a key pathophysiologic factor of metabolic syndrome, so assessment of the relationship between obesity and insulin resistance should be investigated. The Anglo-Danish-Dutch study of Intensive Treatment of Diabetes in Primary Care (ADDITION) study⁶ and Cooperative Health Research in the Region of Augsburg (KORA) study⁷ consistently demonstrated that general and central obesity increased the risk of DPN. However, these well-designed prospective observational studies did not assess body composition other than measurement of waist circumference (WC). Furthermore, the ADDITION and KORA studies enrolled patients that were older than 60 and 65 years, respectively. Body composition is influenced by menopause (or andropause for men)^8,9 and age-related sarcopenia.¹⁰ Therefore, we enrolled young to middle adults to minimize the effect of aging on body composition. This study investigated the association between body composition, components of metabolic syndrome, and DPN in middle-aged adults with type 2 diabetes mellitus (T2DM). We hypothesized that body fat was highly associated with the presence of DPN.

Subjects

Subjects diagnosed with T2DM were recruited. This is a subgroup analysis of an ongoing prospective observational study. The study plans to enroll 100 subjects annually, and there will be 500 final participants with T2DM. The study has been designed to investigate reliable tools and discover biomarkers for DPN. The current analysis was performed using subgroup data for the initial 2-year enrolled participants. Inclusion criteria of current analysis included age from 20 to 55 years and no change of antidiabetic medications in the previous 3 months. Exclusion criteria were other causes of neuropathy such as heavy alcohol consumption, renal dysfunction (estimated glomerular filtration rate less than 50 mL/min/1.73 m²), any history of cancer, and exposure to neurotoxic agents. The study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (IRB No. B-1903/526-101). All participants provided written informed consent.

Assessment of DPN

DPN was diagnosed using the Michigan Neuropathy Screening Instrument (MNSI)¹¹, which was validated in Korean patients with diabetes.¹² The MNSI is composed of two components, a 15-item self-administered questionnaire and a comprehensive physical examination. DPN was diagnosed when the MNSI-questionnaire (MNSI-Q) was ≥7 or MNSI physical examination was ≥4.

Anthropometric and body composition measurements

Body weight, height, WC, and blood pressure were measured by a single trained research nurse. All anthropometric examination was performed in light clothing with bare feet. Body mass index (BMI) was calculated as body weight (kg)/height (m²). WC was obtained at the middle of the lower rib margin and iliac crest. Systolic and diastolic blood pressure (SBP/DBP) were measured after a 10-minute rest using an electronic blood pressure meter (UA-1020 device; A&D Co., Tokyo, Japan). Total body fat mass, muscle mass, and visceral fat area were measured by bio-impedance analysis (InBody770; InBody, Seoul, Korea).

Biochemical analysis

Participants visited the Endocrine Clinic after fasting overnight. Blood glucose levels were measured by the hexokinase method, and glycosylated hemoglobin (HbA1c) levels were obtained by high-performance liquid chromatography (Bio-Rad, Hercules, CA, USA). Fasting insulin levels were measured by immunoradiometric assay (DIAsource, Nivelles, Belgium). Homogeneous enzymatic assays and glycerol-3-phosphate oxidase peroxide methods were used to measure HDL cholesterol, low-density lipoprotein cholesterol, and triglyceride. Liver and renal function tests were performed using the protocol of the central laboratory of Seoul National University Bundang Hospital. Homeostatic model assessment for insulin resistance (HOMA-IR) was calculated as an indicator of insulin resistance using the following formula: plasma insulin (μIU/mL)×plasma glucose (mg/dL)/405.¹³ Insulin resistant status was defined as when HOMA-IR level was higher than 3.0, according to a previous study.¹⁴

Statistical analysis

Data are shown as mean±standard deviation or number with percentage. Means were compared between subjects with and without DPN using Student t-test or the Mann-Whitney U-test. Differences in sex, history of smoking, alcohol, hypertension, dyslipidemia, and antidiabetic medications were tested by chi-square test. Spearman’s correlation analysis was performed to evaluate the association between neuropathy examination results, metabolic syndrome components, and body composition. Univariate and multivariate logistic regression models were used to estimate the association of clinical parameters and presence of DPN using odds ratios (ORs) with 95% confidential intervals (CIs). Statistical analysis was performed using IBM SPSS version 22.0 (IBM Corp., Armonk, NY, USA).

In this study, data from 65 subjects with T2DM was analyzed. Among 65 subjects, 44.6% were diagnosed with DPN. There was no difference in age or sex between subjects without DPN and those with DPN. BMI was higher in subjects with DPN than in subjects without DPN (26.5±4.3 kg/m² vs. 24.1±3.2 kg/m², P=0.011). SBP tended to be higher in the DPN (+) group, but lipid profiles were comparable between groups. HOMA-IR was significantly increased in the DPN (+) group compared to the DPN (−) group. There was a numerically longer duration of diabetes in the DPN (+) group than the DPN (−) group, and subjects with DPN were more likely to be treated with insulin than subjects without DPN (Table 1).

Table 2 shows body composition data. WC, fat mass, fat percent, and visceral fat area were significantly increased in the DPN (+) group than the DPN (−) group. However, lean body mass was not different between groups.

MNSI-physical examination (MNSI-PE) scores were positively correlated with HbA1c (rho=0.261, P=0.035), SBP (rho=0.300, P=0.017), and HOMA-IR (rho=0.310, P=0.017). However, these parameters did not show significant association with MNSI-Q. Among body composition parameters, total obesity and abdominal obesity were positively correlated with MNSI-PE (Table 3).

Univariate regression analysis revealed that HOMA-IR (OR, 1.561; 95% CI, 1.079–2.258; P=0.018), WC (OR, 1.086; 95% CI, 1.018–1.158; P=0.013), BMI (OR, 1.210; 95% CI, 1.033–1.417; P=0.018), and visceral fat area (OR, 1.018; 95% CI, 1.003–1.035; P=0.022) were associated with the presence of DPN. Significance was maintained after adjusting for age, sex, diabetes duration, and smoking status (Table 4).

When subjects were categorized according to their BMI and insulin resistance status, lean subjects with lower HOMA-IR were less likely to have DPN (18.5%). In contrast, the majority of subjects who were obese and had higher insulin resistance also had DPN (66.7%) (Fig. 1).

In this observational study that included young to middle-aged adults with T2DM, 44.6% of subjects were diagnosed with DPN according to MNSI. Subjects with DPN showed higher insulin resistance and a greater degree of general and abdominal obesity. Furthermore, the physical examination score of subjects with DPN was significantly associated with HOMA-IR, BMI, WC, and visceral fat area.

Among the five components of metabolic syndrome, central obesity was the most important risk factor for DPN in subjects with diabetes. In our study, the parameters of central obesity, WC, and visceral fat area were highly correlated with HOMA-IR (rho= 0.510 for WC, P<0.001; rho=0.366 for visceral fat area, P=0.008). In addition, the statistically significant association between DPN and visceral fat area diminished after adjusting for HOMA-IR (data not shown). Therefore, insulin resistance might be a key regulator between obesity and DPN. Insulin resistant status represents low grade inflammation¹⁵, which influences endothelial dysfunction¹⁶ and microvascular complications.^17,18 The KORA study proved that biomarkers of subclinical inflammation were associated with DPN progression.¹⁹ Our study provides evidence of the pathophysiologic role of abdominal obesity and insulin resistance in DPN. However, hyperglycemic-euglycemic clamp studies or oral glucose tolerance tests at multiple time points are needed to measure insulin resistance more accurately.

Previous intervention studies showed that body weight loss decreased the incidence of DPN.^4,20 Although secondary effects due to improved hyperglycemia after body weight loss cannot be ruled out, body weight management is an important intervention strategy for DPN. Studies have already shown the beneficial effect of bariatric/metabolic surgery on cardiac and sudomotor dysfunction.²¹ However, we do not know whether other treatment modalities related to body weight gain, such as insulin therapy²², might contribute to DPN progression. In this study, a numerically higher proportion of subjects with DPN were treated with insulin than subjects without DPN. Further study is necessary to investigate the impact of insulin-induced weight gain on DPN progression.

Correlation analysis showed significant correlations only between obesity parameters and MNSI-PE. In contrast, there was no association between obesity parameters and MNSI-Q. Therefore, the only objective signs of DPN were related to metabolic parameters. Further investigation is needed to demonstrate the association between DPN and subjective symptoms. In line with our findings, physical examination might be more useful in clinical practice than questionnaire-based evaluation in subjects with obesity.

Previously, Won et al.²³ reported that the prevalence of DPN was 33.5% in Korean subjects with T2DM. The previous study enrolled subjects from both secondary and tertiary hospitals, while we enrolled subjects from a single tertiary hospital that patients were referred to by primary care physicians. Furthermore, DPN criteria were not the same between previous and current studies. The previous study diagnosed DPN when both the MNSI-Q (≥3) and 10-g monofilament test were abnormal, while we diagnosed DPN using the MNSI-Q (≥7) and/or physical examination. The 10-g monofilament test is recommended to detect loss of protective sensation²⁴, rather than to diagnose DPN, so we did not adopt the 10-g monofilament test to diagnose DPN. In summary, differences in the study population and diagnostic criteria are likely related to the different prevalence of DPN in different studies. Further multicenter prospective observational studies are necessary to confirm our findings.

This study has several limitations. First, a relatively small sample was analyzed. Accordingly, no significant association was identified between DPN and its traditional risk factors, such as age and diabetes duration. The current analysis was performed in a subgroup sample from a whole cohort population, and further confirmative analysis is mandatory to confirm the study findings. Second, electrophysiologic study was not used to confirm DPN. However, some proportion of DPN cannot be diagnosed by nerve conduction studies (NCSs) because small fiber neuropathy is often not detected by NCSs. MNSI has been validated for the diagnosis of DPN and is widely used in many epidemiologic studies. Third, other symptom scores were not analyzed. Fourth, middle aged people were enrolled. Therefore, the association between DPN and body composition should be tested in elderly patients. The strengths of our study include assessment of abdominal obesity using not only anthropometric measurement but also bio-impedance analysis, although accuracy of bio-impedance analysis is limited with respect to measuring visceral fat area.²⁵

In conclusion, total obesity and abdominal obesity were associated with DPN, and insulin resistance might mediate obesity and DPN. The current study provided important insight into DPN as a clinical parameter in obesity intervention studies.

The authors declare no conflict of interest.