This cross-sectional analytical study was conducted over six months, from February 1 to July 31, 2024. It was carried out in the outpatient metabolic and endocrine disease unit and the clinical neurophysiology laboratory of the Neurology Department at Brazzaville University Hospital.

The study population included all T2DM patients followed at the outpatient metabolic and endocrine disease unit. Participants were recruited exhaustively. Inclusion criteria were age ≥ 18 years, documented T2DM, available glycosylated hemoglobin (HbA1c) value, ENMG assessment, and signed informed consent. Patients with a known cause of peripheral neuropathy (e.g., chronic alcohol use, HIV infection, malignancy, systemic diseases, chemotherapy, neurotoxic drugs), foot deformities, ulcerations, or amputation preventing ENMG, and those with cognitive or psychiatric disorders impairing consent and comprehension were excluded.

Data were collected using a standardized questionnaire. Information included sociodemographic data, medical and diabetes history, and treatment adherence assessed by the Morisky Medication Adherence Scale. Neurological evaluation included DN4 screening for neuropathic pain, measurement of the ankle-brachial index (ABI) via Doppler after 30 minutes of supine rest, and ENMG.

Examinations were scheduled according to patient availability and conducted under optimal technical conditions (controlled skin temperature at 37°C). The following nerves were explored: median and ulnar nerves (motor and sensory components) in the upper limbs; common fibular, tibial (motor), and sural (sensory) nerves in the lower limbs. Standardized techniques were used for nerve stimulation and recording. Needle EMG was not performed due to its invasive nature. Motor parameters included distal/proximal latencies and amplitudes, conduction velocity, and F-wave latency. Sensory parameters included sensory latency, amplitude, and conduction velocity. Interpretation followed University Hospital of Brazzaville standards and ENFS/PNS guidelines [11] [12] .

DPN was categorized as electro-clinical (presence of clinical signs and ENMG abnormalities), clinical (presence of clinical signs only), subclinical (presence of ENMG abnormalities only), or none (absent of both clinical and ENMG findings).

According to the recommendations of the American Diabetes Association (ADA), diabetes control was defined as reaching a glycosylated hemoglobin (HbA1c) level of ≤7%. The ABI was interpreted as follows [13] : >1.30, non-compressible arteries or medial arterial calcification (MAC); 0.90 - 1.30, normal; <0.9, peripheral arterial occlusive disease (PAOD).

The data were entered into Microsoft Excel and analyzed using Jamovi v2.6.17. Categorical variables were expressed as frequencies and percentages, and continuous variables were expressed as the mean ± standard deviation or the median [interquartile range (IQR)], depending on the results of the Shapiro-Wilk test. Categorical comparisons used the Chi-square or Fisher’s exact test. Continuous data comparisons used a Student’s t-test or a Mann-Whitney test. Multivariate logistic regression was performed to identify factors associated with DPN. Variables with p < 0.20 in the univariate analysis were included in the multivariate model. Model fit was evaluated using the Hosmer and Lemeshow test. The significance threshold was set at p < 0.05.

All participants provided written informed consent prior to being included in the study. The study protocol was reviewed and approved by the Ethics and Research Committee of the Faculty of Health Sciences at Marien Ngouabi University. All procedures were conducted in accordance with the principles outlined in the Declaration of Helsinki. Patients diagnosed with DPN were referred to the neurology outpatient unit for appropriate management, including optimization of glycemic control, treatment of neuropathic pain, lifestyle advice, and prevention of complications, such as foot ulcers. These patients were subsequently monitored as part of their routine diabetes care.

The authors declare that they have no conflicts of interest related to this study.

Of the 220 patients with type 2 diabetes mellitus (T2DM) selected, 102 completed clinical and electrophysiological investigations, yielding a participation rate of 46.3%. The remaining 118 patients did not undergo ENMG because they missed their appointment.

DPN was diagnosed in 81 patients, representing a prevalence of 79.4%.

The mean age of patients with DPN was 59.75 ± 9.87 years, ranging from 38 to 82 years old. Of these patients, 56 were women (69.1%) and 25 were men (30.9%), resulting in a male-to-female ratio of 0.44.

Table 1 summarizes occupational status, socioeconomic and education levels.

The mean BMI was 28.2 ± 5.7 kg/m² (range: 16.8 - 44.1). Patients were evenly distributed among normal weight, overweight, and obese categories (33.3% in each).

The median duration of diabetes was 20 months (approximately 1.7 years), and the mean was 56.4 ± 71.5 months (approximately 4.7 ± 6 years). The range was 1 - 384 months.

The ABI was normal in 59.3% of cases and abnormal in 40.7%, with PAOD predominating (63.6%) over MAC (36.4%).

*Others: policeman (n = 2), brewer (n = 1), orderly (n = 1), driver (n = 1), electrician (n = 1), nurse (n = 1) and pastor (n = 1).

Associated complications included nephropathy (12.3%), stroke (8.6%), coronary artery disease (3.7%), and retinopathy (2.5%).

Regarding treatment, 61.7% were on oral antidiabetics (OADs), 19.8% were on insulin, and 18.5% were on both. Treatment adherence was high in 49.4% of patients, moderate in 42% and low in 8.6%.

The mean HbA1c was 8.7% ± 2.1% (range: 5.1-15%). Glycemic control (HbA1c ≤ 7%) was achieved by 19.6% (n = 20) of patients.

DPN was clinically suspected in 66 patients (81.5%).

Previous paresthesia was reported by 25.9% of patients. Eighty-one-point five percent of patients had current neuropathic symptoms, including paresthesia (100%), cramps (28.8%), and neuropathic pain (21.2%). Symptoms had persisted for more than six months in 92.4% of patients.

Objective clinical signs were present in 33 patients (40.7%). These signs included the absence of osteotendinous reflexes (63.6%), thermal hypoesthesia (63.6%), amyotrophy of the thenar eminence or leg (9.1%), and vibratory hypoesthesia (3%).

ENMG confirmed DPN in 54 patients (66.7%). Of the 66 cases suspected clinically cases, ENMG confirmed 81.8%. Twelve patients (18.2%) had no ENMG abnormalities, which suggests small fiber involvement. Conversely, 18,5% of DPN cases were subclinical, presenting only with ENMG abnormalities.

There was no significant association between subjective symptoms and ENMG abnormalities (p = 0.109).

Table 2 presents DPN types identified by ENMG.

DPN types were significantly associated with diabetes duration (p = 0.001), but not with age, BMI, or HbA1c levels. See Table 3 for detailed results.

Univariate analysis revealed associations with age (p = 0.009), diabetes duration (p = 0.034), right ABI (p = 0.040), and the Morisky adherence scale (p = 0.002) (see Table 4 ).

DPN: diabetic peripheral neuropathy, BMI: body mass index, HbA1c: glycated hemoglobin.

DPN: diabetic peripheral neuropathy, PAOD: peripheral arterial occlusive disease, MAC: medial arterial calcification. *Mean ± standard deviation.

Five covariates with a p < 0.20 were included in the multivariate logistic regression: age, diabetes duration, right ABI, Morisky score, and sociodemographic level. With 81 events (patients with DPN), the events-per-variable ratio was 16.2, meeting accepted methodological standards. The multivariate logistic regression analysis identified advanced age (OR = 1.069 [95% CI: 1.008 - 1.134]; p = 0.027) and longer diabetes duration (OR = 1.034 [95% CI: 1.002 - 1.066]; p = 0.034) as independent predictors. The other variables were excluded during stepwise selection due to a lack of statistical significance, suggesting that their effects were confounded or not independently associated with the outcome.

The 79.4% prevalence rate observed in our study is higher than the rates reported in INTERPRET-DD study (26.7%) and an African meta-analysis (46%) [14] [15] . The INTERPRET-DD study included 2733 patients with T2DM from 14 countries worldwide. The African meta-analysis involved 33 studies with 269,691 participants. The estimated prevalence in Central Africa was the lowest, at 35.9%. This difference may be explained by the use of ENMG, which identified 18.5% of subclinical cases.

Similar rates were reported in the neurological functional explorations unit at Limoges University Hospital (87.3%) and among elderly patients undergoing ENMG in Tunis (89.4%) [16] [17] . Other studies using clinical criteria report lower prevalence rates ranging from 14.3% to 44% [18] [19] . In Mali, a prevalence of 69.8% was reported when autonomic neuropathies were included [20] . At Brazzaville University Hospital, 93.13% of patients with diabetic foot syndrome had DPN [10] .

This variability can be attributed to diagnostic methods, population characteristics, and care quality. However, one study reported no significant geographical variation [21] .

Consistent with its known association in diabetics, carpal tunnel syndrome (87%) was the most frequent form [22] - [25] . The pathophysiological mechanisms include endoneural glucose accumulation, ischemia, and compression due to musculoskeletal alterations [23] [24] [26] [27] .

The second most common form was polyneuropathy (37%), which was mainly sensorimotor, axonal, and diffuse, aligning with existing literature [28] - [32] . Oxidative stress appears to be the common pathway for nerve damage induced by hyperglycemia [33] - [35] .

The prevalence of neuropathic pain (21.2%) was lower than in other studies (36.8% to 56.7%) [36] [37] . Radiculopathies were rare and difficult to distinguish from common sciatica [38] - [40] .

The 18.5% rate of subclinical DPN supports the use of systematic ENMG at diagnosis. Up to 50% of DPN may be asymptomatic, which exposes patients to insensitive foot lesions [41] . The shorter median onset time of seven months underscores the early manifestation of these forms compared to clinical forms, which have median onset times of 12 months for small fibers and 42 months for large fibers. Furthermore, clinical symptoms did not influence ENMG results.

Several factors have been linked to the development of DPN in patients with T2DM [42] [43] . Consistent with prior studies, age and diabetes duration were significantly associated with DPN [14] [18] [21] [44] . Other known risk factors include poor glycemic control, smoking, sedentary behavior, obesity, cardiovascular disease, microalbuminuria, severe ketoacidosis, and retinopathy [14] [18] [19] [21] [42] - [44] . However, in our study, neither ABI abnormalities nor HbA1c level were significantly associated with DPN, despite 40.7% of participants having ABI abnormalities and the mean HbA1c level being 8.7%.

The prevalence of DP increases with diabetes duration, rising from 7% in the first year to 50% after 20 years [45] . This is consistent with reports linking DPN to diabetes durations of over 15 years [31] . The main mechanisms involved in nerve degeneration are chronic oxidative stress and metabolic pathway overload induced by hyperglycemia [46] . Thus, while treatment adherence was not associated with DPN occurrence in the multivariate analysis, it was significantly lower among patients with DPN in the univariate analysis. This reflects the increased risk of poor glycemic control associated with DPN development. Furthermore, these patients had slightly higher mean HbA1c levels, and nearly half of them were taking oral antidiabetics medication.

This study has several strengths. First, the systematic use of ENMG enabled the detection of both clinical and subclinical forms of DPN. Second, the rigorous exclusion of alternative causes of neuropathy ensured diagnostic specificity. Conducted in a real-world hospital setting, the study also identified key associated factors, such as age and diabetes duration. However, the cross-sectional design limits causal interpretation, and the relatively small sample size and single-center setting may affect generalizability. Only 102 out of 220 eligible patients completed the full protocol, potentially introducing selection bias. Additionally, this study did not include specific tests for assessing small fibers, such as quantitative sensory testing or skin biopsies, due to the limited availability of these techniques in our setting. Consequently, some cases of small fiber neuropathy may have gone undetected, particularly in patients with normal ENMG results. This limitation may have led to an underestimation of DPN prevalence in the early stages or in its sensory form. Additionally, the lack of longitudinal follow-up for treatment outcomes limits our ability to assess treatment effectiveness.