Diabetic polyneuropathy (DPN) is the earliest and most common complication of diabetes mellitus (DM) type 1 and 2 disrupting the quality of life, resulting disability, high risk of foot ulcers and amputations (diabetic foot syndrome, DFS). DPN is also associated with higher mortality and health care costs [1, 2]. In elderly patients with DM type 2, DPN increases the risk of falls, fractures, hospitalization and mortality in some cases [3, 4]. DPN can develop at the early stages of carbohydrate metabolism disorders. Moreover, distal sensory polyneuropathy can also occur in patients with metabolic syndrome without carbohydrate metabolism disorders.

Timely diagnosis of DPN is extremely important for early therapy of disease with higher effectiveness. The purpose of this review is to systematize data on diagnosis of DPN for primary care physicians.

Introduction

DPN is currently defined as a distal symmetric sensory-motor polyneuropathy with lesions of long nerve fibers (length-dependent polyneuropathy). Polyneuropathy develops because of metabolic and microvascular disorders following chronic hyperglycemia and cardiovascular risk factors. Clinical symptoms may be absent in some patients. DPN is characterized by damage to peripheral nervous system and diagnosed if other potential causes of polyneuropathy are excluded [5, 6].

Painful course of DPN is conditionally distinguished depending on presence or absence of neuropathic pain syndrome. Painful diabetic polyneuropathy not only worsens the quality of life. It is also one of the leading causes of hospitalizations in Russia that undoubtedly increases cost of treatment of these patients [7].

DPN develops in patients with chronic hyperglycemia and metabolic abnormalities (activation of polyol pathway, accumulation of glycation products, oxidative stress, dyslipidemia) and is associated with major cardiovascular risk factors. Microvasculature changes are typical and universal for diabetic retinopathy, nephropathy and polyneuropathy. Importantly, microvascular disorders appear before polyneuropathy. At the same time, there is a clear correlation in course and development of these microvascular complications. Diabetic retinopathy and/or nephropathy significantly increases the risk of DPN. Interrelated and synergistically acting mechanisms are involved in pathogenesis of all forms of DPN. Trigger of these processes is hyperglycemia [8]. Experimental studies confirm multifactorial model of DPN pathogenesis. Oxidative stress, metabolic disorders and inflammation are predominant factors leading to damage to nerve structures [9–12]. The risk of DPN is associated with glycemia, height, smoking, blood pressure, weight and dyslipidemia [13, 14]. Development of DPN after diagnosis of DM depends on polymorphism of antioxidant enzyme genes [15].

Epidemiology

Diabetic distal symmetric sensory-motor polyneuropathy is the most common form of diabetic neuropathy accounting for about 75% of all diabetic neuropathies [16, 17].

According to the State Register of Diabetes Mellitus, incidence of DPN significantly varies in patients with DM type 1 and 2 in the Russian Federation. In 2017, DPN was detected in 33.6% of patients with DM type 1 and 18.6% of patients with DM type 2. In different regions of Russia, these values ranged from 0.1 to 67.2% for DM type 1 and from 0.1 to 42.4% for DM type 2 [18].

According to large-scale epidemiological studies, the true prevalence of DPN is much higher. Incidence of DPN is 10-15% among patients with DM type 2 de novo [19, 20] and 50% in patients with more than 10-year anamnesis of DM type 2 [21, 22]. Incidence of DPN in patients with DM type 1 and more than 20-year anamnesis is at least 20% [23-26]. DPN is much more common in patients with DM type 1 admitted to the hospital. The number of patients with DPN is significantly increased in more than 10-year anamnesis of diabetes mellitus [27].

Painful DPN is detected in 20% of adults with diabetes and at least 30% of patients with DPN [28]. Potential painful DPN with predominant lesion of thin nerve fibers in 10-30% of patients with impaired glucose tolerance should be also considered [29-34].

Clinical manifestations

DPN is heterogeneous regarding clinical course and lesion of peripheral nervous system. Different onset, course and clinical manifestations of polyneuropathy are possible depending on primary lesion of thick or thin nerve fibers. It was also proposed to distinguish typical diabetic distal symmetric sensory-motor polyneuropathy and atypical diabetic polyneuropathy [6, 35]. These gradations ("typical" and "atypical") of DPN are not entirely successful. The American Diabetes Association (2017) classified DPN into 3 options: primary lesion of thin nerve fibers, thick nerve fibers, mixed lesion (the most common one) [28]. DPN with primary involvement of thick nerve fibers is extremely rare.

Depending on pain syndrome, painful and painless forms of DPN are distinguished. Painful DPN implies pain as predominant symptom (neuropathic pain) of disease. Depending on presence or absence of positive neuropathic symptoms (pain, burning, numbness, paresthesia), subclinical and symptomatic stages of DPN are distinguished [6]. Subclinical form is diagnosed in patients without complaints and neurological impairment. However, ENMG changes in 2 nerves of the legs and obligatory lesion of n. suralis are diagnosed. Identification of this form ensures early therapy of DPN.

Typical DPN has a certain phenotype: 1) neuropathic symptoms are always distal and symmetrical; 2) primary lesion of legs with delayed involvement of hands; 3) primary lesion of sensitive fibers and delayed involvement of motor fibers; 4) primary impairment of superficial sensitivity (thin fibers) and delayed violation of proprioception (thick fibers); 5) gradual progression of DPN over several years without sudden acute impairment.

DPN phenotype is extremely important, since at least 10% of polyneuropathies in patients with diabetes are not diabetic. In this case, phenotype of polyneuropathy usually differs from DPN [5].

The main feature of DPN is symmetrical impairment of distal sensitivity (mainly superficial) in legs. Sensory impairment can range from mild to severe disorders with complete loss of sensation.

Painless variant is characterized by slow development, minimal symptoms without pain and gradual progression of sensory-motor deficit. The most common complaints are feet numbness and impaired sensitivity. Objective examination reveals various sensitive disorders, reduction or absence of tendon reflexes. Distal muscle weakness occurs later. The patient does not stand well and walks on the heels.

Painful DPN can be acute or chronic. Chronic painful DPN is characterized by an undulating course with remissions and aggravations. Duration of pain syndrome usually exceeds 3 months. Diurnal rhythm of symptoms is typical for neuropathic pain (intensification in the evening and at night). The nature of pain is very diverse: burning, acute pain, tingling, painful numbness, and convulsions. In majority of cases, absence or impairment of sensitivity and reflexes is observed along with these symptoms.

Typical sensory-motor or predominantly sensory polyneuropathy with a mixed pattern of damage to different types of nerve fibers is the most common (over 90% of all clinical variants of DPN). Acute sensory DPN with damage to thin nerve fibers is a rare variant (atypical DPN). At least two clinical variants are possible: acute painful sensory polyneuropathy of thin nerve fibers (Ellenberg syndrome, diabetic neuropathic cachexia) and acute painful neuropathy induced by the treatment of diabetes mellitus (iatrogenic, insulin neuritis). The first type is the most typical for patients with DM type 1. There is usually no correlation with glycemic parameters. This polyneuropathy is characterized by severe pain. Acute, burning pain, allodynia, hyperesthesia, convulsions, sleep disturbance following pain are typical. Concomitant depression and significant decrease in body weight (> 25%) are common. Involvement of autonomic nervous system is typical: resting tachycardia, orthostatic hypotension, disturbance of circadian blood pressure rhythm, sudomotor dysfunction (night sweating not associated with hypoglycemia), hyper- or less often hypomotor dysfunction of gastrointestinal tract (diarrhea, constipation), erectile dysfunction in men, vaginal dryness in women. Selective damage to thin nerve fibers underlies these symptoms, and electroneuromyography (ENMG) finds no significant abnormalities. This variant requires the mandatory participation of a neurologist and combined analgesic therapy in most cases. Pain usually disappears within 6-12 months of therapy without subsequent recurrence. The second variant is provoked by significant decrease in blood glucose and can develop in DM type 1 and 2. Symptoms are similar and often more stable. Acute pain and autonomic symptoms within 8 weeks after significant decrease in glycemia are typical. HbA1c reduction by 2-3% within 3 months increases the risk of thin-fiber DPN by 20%, decrease > 4% — by more than 80% [36, 37].

Thin fibers can be primarily affected in all types of carbohydrate metabolism disorders (metabolic syndrome, impaired glucose tolerance, diabetes mellitus) that results early impairment of pain and temperature sensitivity [38]. Moreover, polyneuropathy is often determined by lesion of only thin nerve fibers in patients with impaired glucose tolerance. In patients with diabetes mellitus, a combined lesion of thin and thick nerve fibers is observed that will determine clinical picture and diagnostic approaches to DPN (Table 1) [6, 39, 40].

Table 1. Clinical signs and diagnosis of lesion of different nerve fibers in diabetic polyneuropathy [6, 40]

Thin-nerve neuropathy is often undetectable for standard electrophysiological examination. Therefore, other methods should be used to quantify peripheral thin-nerve dysfunction.

Diagnosis

In clinical practice, the diagnosis of DPN is usually based on typical positive neuropathic symptoms (complaints of patients), neurological examination with identification of negative neuropathic symptoms (impaired sensitivity, reflexes, and muscle strength) and ENMG data if this method is available.

DPN screening is indicated for patients with DM type 2, DM type 1 with more than 5-year anamnesis and prediabetes with neuropathic symptoms. Subsequent annual examination is required [28].

Diagnostic criteria of typical DPN (with mixed lesion of nerve fibers) [6, 16, 41]:

1. Possible DPN — positive neurological symptoms (numbness, stitching, cutting pain, paresthesia, burning in toes, feet and legs) or negative neuropathic symptoms (distal symmetric impairment of sensitivity or clear decrease/absence of tendon reflexes).

2. Probable DPN — symptoms and signs of neuropathy (≥ 2 signs).

3. Confirmed DPN — symptoms and/or signs of polyneuropathy (as described above) combined with ENMG disorders.

4. Subclinical DPN — no symptoms or signs of neuropathy, but there are ENMG abnormalities.

Definitions 1, 2 and 3 are recommended for diagnosis in daily clinical practice, definitions 3 and 4 — for clinical research. Diagnosis of damage to thin nerve fibers is recommended in patients with symptoms/signs of DPN and no ENMG abnormalities: quantitative sensory testing (threshold of pain and temperature sensitivity on the feet), punch biopsy of shin and/or foot skin (analysis of intraepidermal density of thin nerve fibers), confocal microscopy of thin corneal nerves [6, 16, 41].

Diagnostic criteria of painful DPN with selective damage to thin nerve fibers (atypical DPN) [42]:

1. Possible DPN with selective damage to thin nerve fibers (symptoms and/or signs of damage to thin nerve fibers).

2. Probable DPN with selective damage to thin nerve fibers (symptoms and/or signs of damage to thin nerve fibers and normal conduction through n. suralis).

3. Confirmed DPN with selective damage to thin nerve fibers (symptoms and/or signs of damage to thin nerve fibers, normal or abnormal conduction through n. suralis and abnormalities in quantitative sensory testing of feet or reduced intraepidermal density of thin nerve fibers on the foot/lower leg according to punch biopsy).

Complaints, anamnesis and physical examination

Resting pain (shooting, burning, painful sensation of cold) in the feet with aggravation in the evening and/or at night, paresthesia, numbness, chilliness of the feet, cramps in leg and feet muscles are typical [28, 43]. Anamnesis is extremely important, especially for an individualized qualitative assessment of pain. The last aspect is valuable not only to suspect neuropathic pain, but also diagnose its cause. It is necessary to invite the patient to describe his pain in as much detail as possible. You should also find out localization, spread, intensity and duration of pain. In addition, comprehensive survey will allow identifying of spontaneous and evoked components of pain syndrome, such as dysesthesia and paresthesia, allodynia, etc. Importantly, patients are not always able to fully and accurately describe their pain. In this case, physician should help the patient to choose the words that most fully and clearly describe pain [44].

Dry skin, hyperkeratosis, muscle atrophy, typical deformities of fingers (hammer toe) are observed. It should also be determined whether pain is localized in the neuroanatomical area typical for DPN.

It is extremely important to diagnose negative neuropathic symptoms, primarily sensitivity disorders.

Peripheral sensitivity assessment

Peripheral sensitivity assessment should be started from distal parts of the legs with gradual proximal movement of stimulus (blunt needle prick, cold, warm stimulus) to determine the level of impairment of one or another type of sensitivity. Assessment of ≥ 2 types of sensitivity increases sensitivity and specificity of clinical diagnosis of DPN [28].

According to the American Diabetes Association (ADA) guidelines, assessment of temperature and/or pain sensitivity (function of thin nerve fibers), vibration sensitivity using a 128 Hz tuning fork and tactile sensitivity using 10 g monofilament (function of thick nerve fibers) is recommended in addition to analysis of anamnesis and complaints for primary diagnosis of DPN [28].

Primary clinical diagnosis is sufficient in most patients. Extended diagnosis including ENMG, confocal corneal microscopy, punch biopsy with assessment of intraepidermal density of thin nerve fibers and quantitative sensory testing is possible in complex cases requiring differential diagnosis or confirmation of the diagnosis. Neurologic examination should include functional assessment of thick and thin nerve fibers. Assessment of thick nerve fibers includes motor function (muscle strength of flexion/extension of the feet, toes, big toes, atrophy of small muscles of the feet) and sensory functions (vibration sensitivity — 128 Hz tuning fork, tactile sensitivity — 10 g monofilament Semmes-Weinstein). Biotensiometry may be used for vibration sensitivity testing (abnormal value > 25 V) [45]. The techniques specified in the appendix are applied to assess peripheral sensitivity [28, 46-52].

There are many methods for tactile sensitivity assessment with various sensitivity and specificity. Standard techniques with a 10 g monofilament are the most acceptable. However, one should remember that assessment of DFS risk and diagnosis of DPN require different approaches [50, 51]. For the diagnosis of DPN, it is not recommended to perform a monofilament test on the plantar surface [53]. It should be remembered that test with a 10 g monofilament has limited diagnostic value and ensures diagnosis of only severe DPN in patients with high risk of DFS and amputations [54, 55]. Annual assessment of tactile sensitivity is recommended for analysis of the risk of DFS and amputations [56, 57].

A test with a 10 g monofilament is performed in 4 points to identify the risk of DFS: plantar surface in projection of the distal heads of the 1^stt, 3rd and 5th metatarsal bones and plantar surface of the first toe [43]. Ipswich test may be applied if monofilament is not available (see Appendix) [58].

Application. Diagnostic tests for assessing peripheral sensitivity

Different protocols for examining the feet are recommended in diabetic patients with risk of diabetic foot syndrome. Three-minute protocol is the simplest and shortest one (Table 2). This protocol can be recommended for all primary care physicians dealing with DM patients [28, 59].

Table 2. Three-minute protocol of foot examination [55]

Diagnostic questionnaires and scales

The validated diagnostic questionnaires including the Michigan Neuropathy Screening Instrument (MNSI), Neuropathy Disability Score (modified version, (NDSm), Neuropathy Symptom Score (NSS) are recommended for primary diagnosis of DPN [60 — 62].

MNSI (Table 3) was developed to facilitate early diagnosis of diabetic polyneuropathy with high sensitivity and specificity. This instrument consists of a questionnaire and examination results. Overall index ≥ 7 determines higher probability of DPN and survey is recommended. Lower cut-off values (≥ 4 scores for questionnaire and > 2 scores for examination) may be advisable for more accurate diagnosis [63 — 65].

Table 3. Michigan Neuropath Screening Instrument (MNSI) [63]

Neuropathy Disability Score modified (NDSm) (Table 4) includes assessment of different types of sensitivity and ensures functional estimation of thin and thick nerve fibers. Interpretation: ≥ 3 scores — suspected polyneuropathy (3—5 scores — mild polyneuropathy, 6—8 scores — moderate polyneuropathy, 9—10 scores — severe polyneuropathy) [19, 62, 66-68]. Total score ≥ 6 significantly increases the risk of DFS.

Table 4. Modified Neuropathy Disability Score

The Neuropathy Symptom Score (NSS) ensures assessment of pain severity, but it is not a tool for diagnosis of polyneuropathy (Table 5) [69 — 71]. Total score of 3—4 points corresponds to mild manifestations, 5—6 scores — moderate painful polyneuropathy, 7—10 scores — severe painful polyneuropathy.

Table 5. Neuropathy Symptom Score

One can objectify the diagnosis of DPN using questionnaires and scales. DPN is possible if NDSm score > 3 or MNSI > 2.5. Painful DPN is probable in NDSm score > 3 or MNSI > 2.5 plus NSS > 3 [62].

The DN4 questionnaire can be used to identify neuropathic pain [72, 73]. Total score ≥ 4 can indicate neuropathic pain that requires further clinical confirmation (Table 6).

Table 6. DN4 questionnaire for diagnosis of pain type

Visual analogue scale (VAS) or numerical rating scale (NRS) is recommended to assess pain severity in all patients with painful DPN before therapy and throughout the follow-up period (Table 7) [74, 75].Visual analogue scale (VAS) or numerical rating scale (NRS) is recommended to assess pain severity in all patients with painful DPN before therapy and throughout the follow-up period (Table 7) [74, 75].

Table 7. Visual analogue scale and numeric rating scale

Electrophysiological tests or neurological survey are not mandatory for DPN screening and necessary in patients with atypical symptoms, as well as for clarification of etiology and differential diagnosis of polyneuropathy. Indications for differential diagnosis are the following situations [76, 77]:

— polyneuropathy with less than 5-year history of DM type 1;

— severe polyneuropathy without nephropathy and/or retinopathy;

— asymmetry of symptoms and neurological impairment;

— predominance of motor symptoms over sensitive ones;

— primary lesion of the upper limbs;

— fast progression of symptoms and neurological impairment;

— chronic alcohol use;

— previous medical and toxic factors (metformin intake for more than 3 years and more than 2 g/day; cytostatics, chemotherapy, work with heavy metals);

— family history of polyneuropathy;

— other neurological syndromes outside of DPN.

ENMG of peripheral nerves can be recommended for patients with diabetes to confirm the diagnosis, especially in atypical symptoms of nervous system lesion, to identify rare forms of peripheral nerve damage in diabetes or concomitant neurological pathology [6, 60, 63].

Confocal corneal microscopy of thin nerves and/or quantitative sensory testing and/or assessment of intraepidermal density of nerves are advisable for patients with diabetes and suspected lesion of unmyelinated nerve fibers in difficult diagnostic cases [38, 42, 60, 61, 78, 79]. These exams are carried out in specialized centers.

Conclusion

Diagnostic approaches to early detection of DPN, proposed today by the global medical community, differ significantly. These features result highly variable incidence of diagnosis of this disease. The diagnostic algorithms for primary diagnosis of DPN described in this manuscript are important for various physicians. The proposed algorithms will make it possible to objectify the diagnosis and improve the timeliness and quality of medical care in patients with DPN.

The authors declare no conflicts of interest111.