Although many people attribute the first clinical description of diabetic peripheral neuropathy to Rollo at the end of the eighteenth century, it was Marchall de Calvi in France who recognized the true nature of the condition in 1864.¹⁰ Later, Charcot extended these observations and described (initially in syphilis) the neuroarthropathy that now is named after him.¹¹ Davies-Pryce, a surgeon working in Nottingham, England, was the first to recognize the link between diabetic neuropathy and foot ulceration.¹² It was not until the twentieth century, however, that autonomic neuropathy in diabetes was first reported.¹³

Definitions and Classification

Although previous classifications have been based on pathologic and causative considerations, it has become increasingly clear that, as is discussed below, causative mechanisms resulting in neuropathy are multiple and complex, so a clinical or descriptive classification of the neuropathies is favored.5,17 Even in this area, a number of classifications exist. Examples include the purely clinical descriptive classification proposed by Boulton and Ward¹⁵ (Table 27-1) and that based on potential reversibility together with clinical description^5,14 (Table 27-2).

Table 27-1

Descriptive Clinical Classification of Diabetic Neuropathies

Polyneuropathy	Mononeuropathy
Sensory	Cranial
Chronic sensorimotor
Acute sensory
Autonomic	Isolated peripheral
Proximal motor	Mononeuritis multiplex
Truncal	Truncal

From Boulton AJ, Ward JD: Diabetic neuropathies and pain. J Clin Endocrinol Metab 16:917–931, 1986.

Table 27-2

Classification of Diabetic Neuropathies Based on Potential Reversibility

Adapted from Boulton AJ, Malik RA, Arezzo J, Sosenko JM: Diabetic somatic polyneuropathies. Diabetes Care 26:1458–1486, 2004, and Thomas PK. Classification, differential diagnosis and staging of diabetic peripheral neuropathy. Diabetes 46(suppl 2):S54–S57, 1997.

A simple definition as to what constitutes diabetic neuropathy was agreed on at an international consensus meeting on clinical diagnosis and management: “The presence of symptoms and/or signs of peripheral nerve dysfunction in people with diabetes after the exclusion of other causes.”¹⁷ The exclusion of other causes is particularly important, as was emphasized by the baseline data from the Rochester Diabetic Neuropathy Study, in which 5% of patients had a nondiabetic cause for their neuropathy.¹⁸

For research, epidemiologic, and clinical trial purposes, a more detailed definition that includes subclinical neuropathy is required.¹⁹ The San Antonio consensus defined diabetic neuropathy as “a demonstrable disorder either clinically evident or subclinical, occurring in the setting of diabetes without nondiabetic causes, including manifestations in the somatic and/or autonomic parts of the peripheral nervous system.”²⁰ The Rochester Diabetic Neuropathy Study established a paradigm for clinical trial design.^18,19 The following were assessed: (1) neuropathic symptoms (neuropathy symptom score [NSS]), (2) neuropathic deficits (neuropathy impairment score), (3) sensorimotor nerve conduction velocity, (4) quantitative sensory tests, and (5) autonomic function tests. The minimum criteria for a diagnosis of neuropathy required two or more abnormalities among the listed criteria, at least one being 3 or 5. Staging was as follows: N0 = no neuropathy, minimum criteria unfulfilled; N1 = asymptomatic neuropathy (NSS = 0); N2 = symptomatic neuropathy; N3 = disabling neuropathy.

Epidemiology

The quality and even quantity of epidemiologic data on diabetic neuropathy remain poor for a number of reasons, including inconsistent definitions, poor ascertainment, lack of population-based studies, and failure to exclude nondiabetic neurologic disease.5,16,17 Most studies report on chronic sensorimotor or autonomic neuropathies,²¹ so this section focuses on these two types. However, despite these problems, diabetic neuropathy undoubtedly is very common, possibly the most common of the late complications of diabetes.

The larger reports of the prevalence of chronic sensorimotor neuropathy published over the last 15 years are summarized in Table 27-3. Of three clinic-based studies from Europe (enrolling more than 2000 patients), a remarkable similarity was noted in prevalence, which ranged from 22.5% to 28.5% for symptomatic neuropathy^22–24; it is reassuring that a more recent population-based survey from Germany reported a prevalence almost identical to those from the clinic-based survey.²⁵ Other population-based studies showed an even higher prevalence, suggesting that at least half of older, type 2 diabetic patients had significant neuropathic deficits and therefore must be considered as being at high risk for insensitive foot ulceration.²⁶ Because only a minority of patients in the population-based studies were symptomatic, most cases of neuropathy would be missed if a careful clinical neurologic examination were not performed. The largest study, a community-based survey from the northwestern United Kingdom, reported the prevalence of a moderate or severe neuropathic deficit to be 22.4% of 9710 diabetic patients.²⁷ Most studies include patients with both type 1 and type 2 diabetes; it must be remembered that neuropathy may be present at diagnosis in type 2 diabetes, as was demonstrated by the UKPDS,³ which reported a prevalence at diagnosis of 13%.

Table 27-3

Prevalence of Diabetic Peripheral Sensorimotor Neuropathy

Certain prospective studies have assessed risk factors for the development of neuropathy. The DCCT ² and UKPDS³ demonstrated a clear relationship between poor glycemic control and the development of neuropathy. In addition to glycemic control, Adler and coworkers²⁸ identified height, age, and alcohol intake as significant risk factors for neuropathy in a study of U.S. veterans. Other studies have identified ischemic heart disease, smoking, and diabetes duration as being independently related to neuropathy.²¹

Autonomic neuropathy has been the subject of fewer epidemiologic investigations, and the results are less consistent than those for somatic dysfunction. In the Eurodiab Type 1 diabetes study, abnormal autonomic function tests (AFTs) were found in 36% of subjects, with cardiovascular risk factors such as cigarette smoking, triglycerides, and diastolic blood pressure showing strong associations with abnormal tests.²⁹ In prospective studies, the DCCT found mixed results in the association between glycemic control and the 5-year cumulative incidence of autonomic neuropathy.² It is surprising, however, that glycemic control was found to be a significant risk factor for deterioration in only one autonomic function test in one study.²

Clinical Features

Focal and Multifocal Neuropathies

Several characteristic focal and multifocal neuropathies, none of which are unique to the diabetic patient, occur in diabetes; together, they account for no more than 10% of all neuropathies. Most of these tend to occur in older, type 2 patients, and the prognosis generally includes recovery of the deficits (partial or complete) and also of the pain that is frequently present. The rapid onset of symptoms and signs in most cases, together with the focal nature of the deficits, is suggestive of a vascular origin. Exclusion of nondiabetic causes is particularly important in these neuropathies; in contrast, any nondiabetic patient with these presentations should be screened for diabetes.

Cranial Mononeuropathies

Acute isolated third, fourth, and sixth nerve palsies occur more commonly in patients with vascular risk factors, including diabetes mellitus, hypertension, hypercholesterolemia, or coronary artery disease.³⁰ Diabetic ophthalmoplegia (third nerve palsy) is the most common, and may be of relatively rapid onset, presenting with pain in the orbit, diplopia, and ptosis. Exclusion of other causes is important; in a study of 66 patients with acute isolated ocular motor mononeuropathies, magnetic resonance imaging (MRI) or computed tomography (CT) demonstrated that 14% of patients had a range of other possible causes, which included brain stem and skull base neoplasms, brain stem infarcts, aneurysms, demyelinating disease, and pituitary apoplexy.³⁰ Furthermore, although these neuropathies traditionally have been believed to be due to acute ischemia within the nerve, Hopf and Guttmann³¹ provided evidence for microinfarcts within the third nerve nuclei.

Isolated and Multiple Mononeuropathies

Numerous nerves are prone to pressure damage in diabetes; by far the most common of these is the median nerve, because it passes under the flexor retinaculum resulting in carpal tunnel syndrome (CTS). In the Rochester Diabetic Neuropathy Study, 30% of patients had EP evidence of median nerve compression, although only fewer than 10% had characteristic symptoms.¹⁸ Recently in the Fremantle Diabetes Study, 1284 patients with type 2 diabetes without a history of CTS surgery were followed over 9.4 ± 3.7 years. The incidence of CTS surgery was 4.2 times greater than in the general population, and significant independent determinants included a higher body mass index (BMI), taking lipid-lowering medication, and, it is interesting to note, being in a stable relationship.³² Furthermore, CTS has been found to be three times more common and of greater electrophysiologic severity in patients with metabolic syndrome when compared with those without metabolic syndrome.³³ Other, less frequently seen entrapment neuropathies may involve the ulnar nerve, the lateral cutaneous nerve of the thigh (meralgia paresthetica), the radial nerve (wristdrop), and the peroneal nerve (footdrop). Occurring in isolation, most of the above (except footdrop) carry a good prognosis with recovery, although surgical decompression may be required. However, increasing reports have described severe bilateral ulnar neuropathy occurring in the presence of long-standing diabetes and other complications—a very different picture from the isolated focal mononeuropathies. Moreover, in one series,³⁴ most cases demonstrated mainly axonal damage due to probable ischemia rather than compression, so surgical decompression would not be beneficial. Mononeuritis multiplex simply describes the occurrence of more than one isolated mononeuropathy in an individual patient.

Truncal Neuropathies

Truncal neuropathy typically is characterized by pain that occurs in a dermatomal band-like distribution around the chest or abdomen. The pain may be severe and may have characteristics of both nerve trunk pain and dysesthesias, typically experienced in mononeuropathies and sensory polyneuropathies, respectively. Thus, the patient may experience dull, aching, boring pain together with burning discomfort or allodynia, and the differential diagnosis includes shingles and spinal root compression. EP investigation, including needle electrode electromyography, is useful and can be diagnostic; it should be performed in any patient who is suspected of this diagnosis. Truncal neuropathies occasionally may present with motor manifestations, typically a unilateral bulging of abdominal muscles that usually is associated with pain, as described earlier (Fig. 27-1). Again, electrodiagnostic studies help to secure the diagnosis, and the natural history for symptoms and signs is good, with recovery the rule.³⁵

FIGURE 27-1 Diabetic truncal polyradiculopathy presenting as a bulge in the right abdominal wall secondary to muscle weakness. (From Boulton AJM, et al: Diabetic thoracic polyradicaloneuropathy presenting as an abdominal swelling. Br Med J 289:798–799, 1984.)

Proximal Motor Neuropathy

Typically affecting older, male, type 2 diabetic patients, proximal motor neuropathy (amyotrophy) presents with pain, wasting, and weakness in the proximal muscles of the lower limbs, either unilaterally or with asymmetrical bilateral involvement. In addition, a distal symmetrical sensory neuropathy occurs, and weight loss of as much as 40% of premorbid body mass may occur.³⁶ However, a series of neuropathologic studies have provided some interesting insights into the pathogenesis of this condition.^37,38 Thus in biopsies of the intermediate cutaneous nerve of the thigh, asymmetrical axonal loss within and between nerve fascicles suggests an ischemic process; an increased incidence of segmental demyelination and remyelination also occurs. In addition, however, a unique mononuclear cell (CD4⁺, CD8⁺) and macrophage infiltrate is noted around epineurial and perineurial vessels with endoneurial hemorrhage.³⁷ Previously, no specific treatment other than improving glycemic control and physiotherapy had been advocated, and in most cases, recovery was gradual but at times protracted. On the basis of the immunopathologic findings, immunosuppression has been advocated as a therapeutic option.³⁹ However, controlled clinical trials of this intervention have not been undertaken, and given that the natural history of this condition is improvement with time, the results of the open trials are difficult to interpret.

Chronic Inflammatory Demyelinating Polyneuropathy

A demyelinating neuropathy that meets the electrophysiologic criteria for chronic inflammatory demyelinating polyneuropathy (CIDP) has been increasingly recognized to occur more commonly in patients with both type 1 and type 2 diabetes.⁴⁰ The clinical picture is of a symmetrical, predominantly motor polyneuropathy that has a progressive course, with proximal and distal weakness in the lower limbs and reduced reflexes. Electrophysiologic, clinical, cerebrospinal fluid, and histologic criteria for the diagnosis are well described, although not all may be necessary in individual cases.⁴⁰ Because patients with CIDP might respond to immunomodulatory therapy, it is important to distinguish this condition from other diabetic neuropathies, particularly proximal motor neuropathy. Therefore, CIDP should be suspected in neuropathic diabetic patients in the following cases:

1. A predominance of motor signs involves proximal or distal lower limb muscles.

2. After some years of distal sensory neuropathy, a motor neuropathy develops with progressive symptoms and signs.

3. A patient is diagnosed with proximal motor neuropathy (amyotrophy).

A recent study has shown that diabetic patients with CIDP present with a higher frequency of autonomic dysfunction and electrophysiologic evidence of associated axonal loss, which may explain a poorer response to treatment with oral prednisolone 1 mg/kg/day with or without azathioprine 1 to 2 mg/kg over 6 months.⁴¹

Symmetrical Neuropathies

Autonomic Neuropathy

The autonomic nervous system, which controls a wide range of bodily functions, can be damaged in diabetes with a variety of manifestations, most commonly, cardiovascular, urogenital, gastrointestinal, thermoregulatory, and sudomotor function.⁴²

Cardiovascular: Cardiac autonomic neuropathy manifests initially as an increase in heart rate secondary to vagal denervation, followed by a decrease due to sympathetic denervation; finally, a fixed heart rate, which responds only minimally to physiologic stimuli, supervenes, bearing similarities to the transplanted heart and suggesting almost complete denervation. Postural hypotension, defined as a 20 mm Hg and 10 mm Hg drop in systolic and diastolic blood pressures, respectively, occurs as a consequence of impaired vasoconstriction in the splanchnic and cutaneous vascular beds due to efferent sympathetic denervation. Twenty-five percent of children display some degree of cardiac autonomic dysfunction on diagnosis of type 1 diabetes,⁴³ and an abnormality in the expiration-inspiration ratio has been reported in up to 28% of patients with impaired glucose tolerance.⁴⁴ Parasympathetic dysfunction is present in 65% of type 2 diabetic patients 10 years after diagnosis, and combined parasympathetic-sympathetic neuropathy is present in 15.2%.⁴⁵

Gastrointestinal: Autonomic neuropathy of the gastrointestinal system manifests as an abnormality in motility, secretion, and absorption through derangement of both extrinsic parasympathetic (vagus and spinal S2 to S4) and sympathetic, as well as intrinsic enteric, innervation provided by Auerbach’s plexus. Clinically, patients present with two major problems: diabetic gastroparesis, manifest by nausea, postprandial vomiting, and alternating nocturnal diarrhea and constipation.⁴² It is fortunate that the diagnosis and treatment of these abnormalities represent an extremely difficult clinical problem in only a minority of diabetic patients.

Erectile Dysfunction

Erectile dysfunction (ED) in diabetes is usually of multifactorial origin, although in most series, autonomic neuropathy is a major contributory factor.38,46 In the 4-year study of Veves and coworkers,⁴⁷ neuropathy was the principal cause of ED in 27% of newly presented patients with ED and a contributory cause in a further 38%. Cholinergic and noncholinergic noradrenergic neurotransmitters mediate erectile function by relaxing the smooth muscle in the corpus cavernosum; the ED resulting from autonomic dysfunction is usually progressive but of gradual onset and progression.⁴² Other features include occasional retrograde ejaculation, although some ejaculation and orgasm are maintained. Because of the multiple contributory factors to most cases of ED in diabetes, careful assessment of each case is essential. Consideration of other potential causes, including vascular disease, other medications, local problems such as Peyronie’s disease, and psychological factors, is essential before therapeutic approaches are considered.

Bladder Dysfunction

Bladder dysfunction is well recognized as a consequence of autonomic neuropathy in some patients; this “cystopathy” is usually the result of neurogenic detrusor muscle abnormality. In extreme cases, gross bladder distention may occur with abdominal distention and overflow incontinence.

Sweating Abnormalities

Abnormalities of sweating are common but often neglected symptoms of autonomic neuropathies.⁴⁴ Most common is reduced sweating in the extremities, particularly the feet, which is a manifestation of sympathetic dysfunction. The sweat gland has a complex peptidergic and cholinergic innervation, and neuropeptide immunoreactivity (especially for vasoactive intestinal polypeptide) is low in sudomotor nerves.

In contrast to dry feet, some patients complain of drenching truncal sweating, particularly at night. Gustatory sweating, which is profuse sweating in the head and neck region caused by eating certain foods, is a highly characteristic symptom of diabetic autonomic neuropathy and is common in patients with nephropathy; it is “cured” by renal transplantation.⁴⁹

Distal Sensory Neuropathy

The clinical presentation of distal sensory neuropathy, the most common of all the diabetic neuropathies, is extremely variable, ranging from severely painful (positive) symptoms at one extreme to the completely painless variety that may present with an insensitive foot ulcer.⁵ It is a diffuse symmetrical disorder that mainly affects the feet and lower legs in a stocking distribution but rarely also involves the hands in a glove distribution. As the disease progresses, some motor dysfunction (including small muscle wasting and sensorimotor neuropathy) usually occurs, together with abnormalities of AFTs.

The onset of sensory neuropathy is usually gradual, with the insidious appearance of symptoms that may be intermittent in the early stages. However, an acute sensory neuropathy is recognized with rapid onset of painful symptoms. In this latter type, which often follows a period of severe metabolic instability or may be precipitated by a sudden improvement in control (“insulin neuritis”),³⁶ the symptoms are usually severe, whereas few if any clinical signs may be noted, and quantitative testing may be normal. Recently, a similar, predominantly small-fiber neuropathy, often with severe painful symptoms, has been observed in patients with impaired glucose tolerance (IGT).⁵⁰

Neuropathic symptoms may be difficult for the patient to describe but typically fall into a recognizable pattern, ranging from severely painful (or positive) at one extreme, with burning pain, stabbing, and shooting sensations; uncomfortable temperature sensations; and paresthesias, hyperesthesias, and allodynia; to mild or “negative symptoms,” such as decreased pain sensation, deadness, and numbness. Symptoms fluctuate with time but tend to be extremely uncomfortable, distressing, and prone to nocturnal exacerbation with bedclothes hyperesthesias.

A symptom complex that has been recognized only recently as a relatively common complaint in neuropathy is that of postural instability; diabetic neuropathic patients report more falls, and unsteadiness (secondary to disturbances in proprioception) should be added to the list of neuropathic symptoms: it often may result in depression.⁵¹ Studies have confirmed this phenomenon, showing that neuropathic patients sway more when quantitatively assessed with Romberg’s test.^9,52

Although neuropathic symptoms are predominantly if not exclusively sensory, in many cases the signs are both sensory and motor, with sensory loss in a stocking distribution, together with minor degrees of small muscle wasting and occasional weakness. The ankle reflex usually is reduced or absent, and the skin in the dorsal and especially plantar surfaces may be dry, owing to associated sympathetic autonomic dysfunction. Because some neuropathic patients may be asymptomatic, it is essential that all diabetic patients have their feet examined on a regular basis.¹⁷

Small-Fiber Neuropathy

Some confusion has been expressed among authorities about definitions of diabetic neuropathy. Some believe that there exists a specific small-fiber neuropathy with neuropathic pain, sometimes together with autonomic dysfunction but few signs. This shares many similarities with acute sensory neuropathy, but symptoms tend to be more persistent.6,31 However, this simply may represent an early stage in the development of chronic sensorimotor neuropathy.⁵³ These painful sensory neuropathies should not be confused with hyperglycemic neuropathy, which may occur in newly diagnosed patients and is characterized by rapidly reversible abnormalities of nerve function and, occasionally, transient symptoms.⁵

Natural History of Chronic Distal Sensory Neuropathy

The natural history of neuropathy is poorly understood, and few worthwhile studies have been published. It generally was believed that neuropathic symptoms waxed and waned but persisted for years; however, in a prospective study, Benbow and coworkers ⁵⁴ reported that most patients reported improved symptoms during this time, although progressive deterioration was noted in quantitative sensory testing (QST). Thus, improvement in symptoms must not be equated with parallel improvement in nerve function.^5,35

A recent community follow-up study of patients with painful neuropathy reported that although symptoms resolved in a minority, they tended to persist in most of those followed for 5 years.⁵⁵ Controversy still exists as to which sensory modality is affected first, although it generally is accepted that small-fiber dysfunction is present early in the course of neuropathy.³¹ However, no doubt exists that gradual loss of nerve function in diabetic patients is more rapid than that in age-matched nondiabetic subjects; this rate of loss is related to the level of glycemic control.^1–4 One consequence of this progressive diminution of nerve function is an increasing risk for insensitive foot ulceration; progressive loss of large- or small-fiber function is associated with an increasing risk for foot ulceration.⁵⁶

Measures of Neuropathy

The diagnosis and staging of neuropathy are important not only for day-to-day clinical practice, but also for the conduct of clinical protocols to assess its origin and natural history and to test newly proposed treatments. As was stated previously, definitions and classifications of neuropathy are available for both clinical practice ¹⁷ and clinical trials.²⁰ The Peripheral Nerve Society has issued a consensus statement on measures used to assess efficacy in controlled trials of new therapies for diabetic neuropathy⁵⁷; the use of composite scores of nerve function was advocated in this and other reports.⁵⁸ In this section, potential measures for clinical diagnosis or follow-up of patients in clinical trials are discussed.

Clinical Symptoms

Accurate recording of symptoms is essential for both clinical practice and trials of new medications. It is important to record patients’ descriptions of their complaints verbatim; the physician must not attempt to interpret or translate patients’ symptoms into medical terminology. A number of instruments developed to quantify neuropathic symptoms might aid in diagnosis and in longitudinal studies.59,60 The McGill Pain Questionnaire, which consists of descriptors of symptoms from which patients select those that best describe their experience, when applied to diabetic neuropathy, was found to be a sensitive measure.⁵ The recently validated “NeuroQol” instrument combines a neuropathic symptom score with an assessment of quality of life.⁶⁰

The neuropathy symptom score (NSS) and its derivatives, the neuropathy symptom profile (NSP) and neuropathy symptom change scores (NSC), are perhaps the most commonly used measures in clinical trials.19,57,58 The NSS is a standardized list of questions and neuropathic symptoms that is applied by a trained individual in a standardized manner. A simplified NSS has been used for epidemiologic studies and can be applied in clinical practice for patient follow-up. It can be administered in a few minutes and scores typical symptoms with additional weighting for nocturnal exacerbation.^22,27

Clinical Signs

Simple clinical observation may identify a neuropathic foot; evidence might include small muscle wasting, clawing of toes, prominent metatarsal heads, dry skin and callus (secondary to sympathetic dysfunction), and bony deformities secondary to Charcot’s neuroarthropathy.

Two simple instruments can be used in clinical practice or for clinical trial assessment. First, Feldman and coworkers ⁶¹ developed the Michigan Neuropathy Screening Instrument (MNSI); this two-step program is used for diagnosis and staging of neuropathy. The MNSI consists of a 15-question yes/no symptom questionnaire that is supplemented by a simple clinical examination. Patients with an abnormal score on the MNSI are referred for quantitative sensory testing (QST) and electrophysiology (EP). Second, the simplified neuropathy disability score (NDS) is a simple clinical examination that sums abnormalities of reflexes and sensory assessment; it has been used in clinical practice and in epidemiologic studies.^22,27 The original NDS was developed by Dyck and colleagues at the Mayo Clinic for the detailed structured assessment of neurologic deficits secondary to neuropathy.^18,19,58 This technique is reproducible if performed by trained and experienced physicians and is being used in a number of ongoing trials of new therapies for diabetic neuropathy.

Quantitative Sensory Testing

QSTs assess patients’ ability to detect a number of sensory stimuli and offer the advantage that they directly assess the degree of sensory loss at the most vulnerable site: the foot.⁶² However, these tests are complex psychophysiologic tests that also rely on patients’ responses and therefore cooperation and concentration. Moreover, an abnormal finding does not necessarily confirm that the abnormality lies in the peripheral nerve; it might lie anywhere in the afferent pathway. QSTs vary in complexity; the simpler instruments can be used in day-to-day clinical practice, whereas the more sophisticated instruments usually are used for more detailed assessment and for follow-up assessments in clinical trials. Some of the more commonly used techniques are now discussed briefly.