Peripheral Neuropathy in Diabetes

Peripheral Neuropathy in Diabetes: A Clinical Approach

Dr. S.M.Sadikot.
Hon. Endocrinologist,
Jaslok Hospital and Research Centre,
Mumbai 400026

The relationship between diabetes and the peripheral nerves has been known for quite a long time. Rollo is credited as being the first to record this association almost two hundred years ago. In fact, most of the observers in the 19th Century felt that diabetes was caused by the involvement of the nervous system through some unknown mechanism. It was only in 1864 that Michael de Calvi first suggested that diabetes may be the cause of, rather than be caused by, neuropathy. He also recorded the occurrence of pain in the distribution of the sciatic nerve and loss of peripheral sensations. Later, Bouchard noted the loss of tendon reflexes, and Pavy and Althus provided a further account of the nerve involvement. As late as 1945, Rundles first definitively ascribed diabetes to be the cause of peripheral nerve involvement and gave the first comprehensive description.

Estimates of the prevalence of diabetic neuropathy vary widely with studies showing a range from 0% to as much as 93%. This is due to the fact that the criteria used for the estimation of prevalence are not standardised and also to the differences in the duration of diabetes before one tests for the presence of neuropathy. Pirart's classic 25 year prospective study of 4400 unselected patients would give the best idea about the prevalence. He defined neuropathy as the loss of the ankle reflex and/or patellar reflex plus diminished vibration sense irrespective of other clinical signs and symptoms of nerve involvement. Neuropathy was detected at the time of diagnosis of diabetes in 12% of the patients, mostly in those with mild Type 2 diabetes. These were patients in whom it is the most difficult to rule out pre-existing hyperglycemia. In the series reported by Pirart, prevalence increased linearly with the duration of diabetes, reaching about 50% of the patients after 25 years of diabetes. He also reported that patients with retinopathy and/or nephropathy were more likely to have evidence of neuropathy.

Our own study CINDI carried out by DiabetesIndia showed a prevalence of around 15% in patients who were evaluated within three months of initial diagnosis.

In Type I patients in whom the onset is usually easy to pinpoint, Eng has reported that it is quite rare for neuropathy to be detected within the first five years of diabetes.

Although various classifications of diabetic neuropathy have been published, none of them have gained wide acceptance. From a review of the different classifications, I have compiled a clinical classification, which is arbitrary but would help in understanding the various clinical manifestations of diabetic peripheral neuropathy. It should be clear that the different categories are not absolutely independent and patients may show the signs and symptoms of more than one of the classified categories.

This is the commonest form of painful neuropathies having a relatively acute onset. It is extremely painful and like all diabetic neuropathies, the pain is more at night, a diagnostic pointer to its diabetic roots. Curiously, it tends to develop about three weeks after a stressful episode like surgery, heart attack, ketosis, severe infection etc. It may develop after the initiation of insulin therapy especially in those patients in whom the blood glucose is dropped very rapidly and more so if this is accompanied by frequent and/or severe episodes of hypoglycemia. A severely curtailed diet may also play a role, especially in those patients who are already slightly undernourished. Typically, this form a diabetic neuropathy generally lasts for about 6-18 months and then disappears.

Motor involvement is commoner with this neuropathy in comparison to that associated with the gradual onset distal symmetrical polyneuropathy. The motor involvement also tends to be more severe. Depression and loss of appetite is seen so often that this should be considered as a part of the clinical picture in this type of neuropathy.

This commonly seems to attack men in their 60's with a history of diabetes of a relatively short duration. It is associated with a severe pain and like other painful diabetic neuropathies, it may have a time relationship with a stressful situation. Diabetic plexopathy is typically associated with painful weakness and atrophy of the pelvic girdle and thigh musculature. The weakness commonly affects the ilio-psoas and the quadriceps muscles with a relative sparing of the hip extensors and the hamstrings. Thigh flexion and the stability of the knee is affected although the clinical picture does have a resemblance to femoral neuropathy, the associated weakness of the gluteii, thigh adductors and the occasional involvement of the distal muscles like the peroneal and the tibial, gives a distinct differential feature in diabetic pelvic plexopathy.

Sensory signs and symptoms may also occur and include paraesthesias and dysthesias of the anterior thigh and anteromedial leg, and may radiate down to the dorsum of the leg. The pain may start from the sacroiliac region and radiate down the back of the leg. The knee jerk is commonly absent, the ankle jerk being present, whilst the planter may be extensor.

The plexopathy is usually unilateral in onset, but the other side may also be affected whilst sometimes involvement of the contralateral side is seen to manifest after the episode has disappeared from the initial side. The shoulder girdle may be affected in rare cases but if bulbar involvement is seen, then a search must be made for another cause. The main differential diagnosis is lumbar nerve root compression, although neoplastic infiltration of the lumbosacral canal, motor neuron disease, myopathy and vasculitis may have to be kept in mind in atypical cases. Indeed laminectomies have been carried out in quite a few of these patients in order to treat nerve root compression! But the lack of mechanical manifestations, the nocturnal exacerbation of pain with failure to respond to bedrest, the extreme muscular wasting and the types of muscle groups involved may help in the correct diagnosis.

Isolated or multiple lesions of the nerves occur more frequently in people with diabetes than in the general population, and this may frequently be seen even in the absence of signs and symptoms of peripheral nerve involvement. The 3rd and the 6th cranial nerves are the ones commonly affected whilst the 4th cranial nerve would rarely be affected alone, if ever. The onset of diplopia is acute and is often seen to be associated with pain behind and above the affected eye. (It is felt that this pain is caused by involvement of the first and second parts of the trigeminal nerve, as the occulomotor nerves are purely motor).

An important diagnostic point in distinguishing this from the involvement of the third nerve in other pathologies such as cavernous sinus thrombosis is that the pupils are spared in diabetic involvement, whilst in other cases pupillary dilation is one of the first features.

Another important differential diagnosis is with temporal arteritis. In this condition, the muscles are involved in a random manner whilst in diabetic third nerve involvement, all the muscles supplied by the nerve are affected.

After the cranial nerves supplying the external ocular muscles, it is the 7th cranial nerve that seems to be the most affected in diabetes. There seems to be a less favorable prognosis for recovery as compared to a similar non-diabetic case. This is in contrast to the clinical picture seen with the third and sixth nerve involvement, where the pain gradually subsides and the recovery is mostly complete in a while. In fact, it should be made clear that many authorities doubt whether the seventh nerve is really affected in diabetes but feel that diabetes could be a coincidental finding. Other upper cranial nerves may be rarely affected, whilst the lower cranial nerves are almost never the site of diabetic cranial mononeuropathy.

Isolated peripheral mononeuropathy is relatively more common in diabetes. The pathological basis may be a vascular lesion, entrapment or trauma to a superficially placed nerve. Any major nerve trunk may be affected and when several are affected at the same time, the entity is called "mononeuritis multiplex".

The signs and symptoms would obviously depend on the site and type of nerve that is involved. They may be purely sensory as when the lateral cutaneous nerve of the thigh is involved, but more commonly, the clinical manifestations are of a mixed type although pure motor nerve involvement is known to occur. The onset is usually acute, but may rarely be insiduous. The acute onset varieties are invariably quite painful, but satisfactory recovery seems to be a part of the clinical picture. The nerves that are most commonly affected are the median, ulnar, radial, femoral, sciatic and the peroneal, (the nerves are not listed in any particular order).

Detailed descriptions about the signs and symptoms associated with involvement of individual nerves are beyond the scope of this book and the reader could refer to any standard text for these descriptions. What is interesting is that many of the nerves involved are those that are capable of being involved in compression and/or entrapment pathologies and it is possible that diabetes in some manner makes the entrapment much more plausible.

This is a relatively uncommon form of diabetic mononeuropathy, but its correct diagnosis is tremendously important in order to save the patient from unnecessary therapeutic intervention. When the thoracic or upper abdominal radicals are involved, the clinical picture may definitely mimic an acute chest or abdominal emergency. In fact, patients have been admitted in the intensive care units with a provisional diagnosis of myocardial infarction whilst some have even undergone emergency abdominal laparotomy. A careful examination showing that the pain is limited to a dermatomal distribution (with an EMG of the paraspinal muscles if needed) would have avoided this diagnostic gaffe. The cause for this type of neuropathy is not well known although it is felt that ischemic infarction of the nerve may play an important role. In any case, the patient may be comforted with the fact that, invariably, the pain subsides after a period and that there is no residual damage, although the complete recovery may take upto a few months.

This disorder has been recognised as a clinical entity only in relatively recent times. It resembles the syndrome of painful, acute onset symmetrical neuropathy but is associated with an incredible weight loss which has even been known to be in the range of 100 lbs. in some affected individuals. This often leads to an incorrect tentative diagnosis of malignancy and the patient is extensively investigated for the root of the malignant problem. The irony is that this type of nerve involvement usually disappears within 6-12 months and the patient recovers even as the extensive investigations are still going on!

To date, all the patients with this type of syndrome have been men in their 60's and despite the clinical severity of the neuropathy, most have mild degree of hyperglycemia which often is diagnosed only when they are investigated for the cause of the neuropathy and the weight loss. It could be that the mild degree of hyperglycemia may mislead the physician who is unaware of this clinical entity.

Diabetic autonomic neuropathy usually accompanies the gradual onset type of distal symmetrical polyneuropathy. It may, in some patients, be so mild that it is not even taken into consideration, whilst in some cases the degree of involvement of the autonomic nerves may overshadow all other clinical manifestations of peripheral neuropathy. At the same time, it should be realised that even mild degrees of clinical involvement of the autonomic nerves can give size to significant problems to the patient.

Clinical Manifestations of Autonomic Neuropathy

  • Tachycardia, exercise intolerance
  • Cardiac denervation, painless myocardial infarction
  • Orthostatic hypotension
  • Heat intolerance¬∑ Alterations in skin blood flow
  • Esophageal dysfunction¬∑ Gastroparesis diabeticorum
  • Diarrhea
  • Fecal incontinence
  • Erectile dysfunction¬∑ Retrograde ejaculation
  • Cystopathy
  • Neurogenic bladder
Sweating disturbances
  • Areas of symmetrical anhydrosis
  • Gustatory sweating
  • Hypoglycemia unawareness
  • Hypoglycemia unresponsiveness
  • Decreased diameter of dark-adapted pupil
  • Argyll-Robertson-type pupil

Postural hypotension is the commonest form of cardiac autonomic neuropathy in diabetes, and a postural drop of more than 20mm of Hg. is seen in many diabetic patients. In fact, this drop often is used as a test for the presence of autonomic neuropathy. Interestingly, postural hypotension may be made worse after an insulin injection. It has been shown that significant changes occur in the blood pressure after an insulin injection and that this tends to peak within one to three hours after the injection, so that the greatest effect on the postural drop in systolic blood pressure is seen at these times. At the same time, it should be noted that many patients do have a slight drop in systolic blood pressure but this is not large enough to cause any symptoms.

Diabetics with significant degree of autonomic neuropathy may show a partial or complete cardiac denervation with a fixed pulse rate of about 80-90 which does not change with stress, exercise or sleep as it does in normal people. Cardiac autonomic neuropathy also changes the autonomic tone in the heart and one sees adrenergic supersensitivity as a result. Both, the partial or complete cardiac denervation and the adrenergic supersensitivity are blamed for the greater incidence of coronary artery spasms, painless myocardial infarcts and sudden deaths seen in diabetics.



Resting heart rate > 100 beats/min is abnormal.
Beat-to-beat heart rate variation With the patient at rest and supine (no overnight coffee or hypoglycemic episodes), breathing 6 breaths/min, heart rate monitored by ECG , a difference in heart rate of > 15 beats/min is normal and < 10 beats/min is abnormal.
Heart rate response to standing During continuous EKG monitoring, the R-R interval is measured at beats 15 and 30 after standing. Normally, a tachycardia is followed by reflex bradycardia. The 30:15 ratio is normally > 1.03.

Although, exercise is an important aspect in the management of diabetes, it is not without danger as the capacity to exercise may be severely compromised in a diabetic with significant autonomic neuropathy. We have seen that the heart is relatively incapable of making adjustments to stress and exercise when it may be partially or completely denervated. There may also be associated problems with peripheral vascular adjustments. Normally, when a person exercises, the heart rate increases and there is also an increase in the contractility. Peripherally, changes take place in the vessels such that blood flow is diverted from the viscera to the muscles. In diabetics with a significant autonomic neuropathy, besides the problems in the heart, changes in the autonomic tone may prevent an adequate rerouting of the blood flow thus starving the muscles and decreasing the capacity to exercise. Conversely so much blood may be diverted that the internal organs may be starved of blood and there have been reports that acute renal shutdown has occurred when a diabetic has undergone sudden and severe exercise. This should not be taken to mean that exercise should be eschewed by all diabetics, but points to the need for greater pre- exercise evaluation and also for a graded and well prescribed exercise therapy in the management of diabetes.

Diarrhoea is one of the most disturbing manifestations of autonomic nerve involvement, and in its most severe form can be quite disabling, both, from a physical as well as a social viewpoint. The attacks of diarrhoea may be preceeded by abdominal rumbling and discomfort. There is a significant increase in the nocturnal frequency of stools and there may be such a severe degree of urgency that this may be accompanied by fecal incontinence. The stools are usually thin and watery but may be steatorrhic. The patient passes stools, often in the range of 25-30 times a day. These attacks are intermittent and may be followed by periods of normality or even by episodes of severe constipation.

In fact, some studies have noted constipation as the most common manifestation of diabetic autonomic neuropathy.

One of the most emotionally disabling problem is that the patient is unaware about when he may get an attack of diarrhoea and this clouds his social and economic outlook to the possible exclusion of all else. The diarrhoea is multifactorial in its causation and may reflect hypermotility from decreased sympathetic inhibition, bacterial overgrowth, pancreatic insufficiency, diabetic sprue and bile salt malabsorption.

Oesophageal motility disorders have been reported and whilst they may cause dysphagia, the more significant problem seems to be related with the decrease in the tone of the lower oesophageal sphincter allowing regurgitation and the sense of retrosternal fullness and discomfort. Problems with gastric emptying and decreases in acid production have also been reported, as has dysfunction in the contractility of the gall bladder causing it to dilate, although it has not been shown that this increases the risk of infection or stone formation.

Bladder dysfunction is commonly seen in diabetic neuropathy. In the early stages, the patient may complain of a diminished frequency of urination and this may be associated with nocturia. Later, the problem of decreased frequency of urination increases, and this is accompanied by a difficulty in emptying the bladder, incontinence and overflow dribbling. Residual urine in the bladder leads to increased urinary infection.

Impotency is one of the most troubling aspects of diabetic neuropathy and is present in more than 50% of male diabetics; retrograde ejaculation may also be seen. Neurogenic sexual dysfunction is present also in about 25-30% of female diabetics. Inspite of such a high prevalence and the great emotional and social trauma that it causes, this is one aspect that is often ignored by most of us.

This is unfortunate as today, there is so much that can be done for erectile dysfunction. I have discussed this aspect in separate chapter.

Sweating disturbances may be seen in diabetics with autonomic neuropathy. This may be manifest as abnormal and bizzare sweating patterns, the commonest being the absence of sweating in the distal parts of the lower extremities. This could be accompanied by compensatory hyperhydrosis on the trunk and face and the patient may show heat intolerance. Drenching night sweats have been described which an unknowing physician could attribute to nocturnal hypoglycemia. The cause for this nocturnal sweating is obscure, but the distal anhydrosis is akin to the distal symmetrical neuropathy and is said to be caused by involvement of the autonomic nerves regulating peripheral vascular flow and tone of the vessels.

Facial sweating at mealtimes is rarer than one thinks, but when present, it is a striking feature of diabetic peripheral autonomic neuropathy. In severe cases, this may be intolerable to the patient. Sweating usually begins moments after starting to chew tasty foods, specially cheese or other foods rich in tyramine. It starts on the forehead and then may spread to the face, scalp, back and occasionally to the shoulders and the upper part of the body. The patient may be compelled to keep a towel with him during mealtimes. There is very little that can be done about this disturbing condition, although anticholinergics have been useful but are often poorly tolerated. The patient may be advised to avoid the foods he know which bring about this phenomenon.

The pupils of the eye are seen to be constricted and this component of autonomic neuropathy, although not clinically significant in itself, is supposed to reflect duration of diabetes, metabolic factors as well as the presence of other diabetic complications.

The ability of the body to stabilise the blood glucose levels during times of substrate deprivation as well as during mealtimes is dependent on certain finely tuned interplays.

One of the most important of these, which goes a long way to warn about low blood glucose levels as well as to counter-regulate the fall in the glucose levels, is the role played by glucagon and the adrenal medulla. These depend to some extent on the functional integrity of the autonomic nervous system and in diabetic autonomic neuropathy, this system is compromised so that the ability of the body to show the signs and symptoms of hypoglycemia are diminished as is the capacity of the body to overcome hypoglycemia and raise the blood glucose levels to an acceptable value.

It is obvious that one should try and achieve an optimal control of diabetes. This would help in preventing the initiation of neuropathy and in the early stages, it has been shown that it may even help in the regression of the neuropathy. Would optimal control of the blood glucose levels help in cases where the neuropathy is well established. Although there is some controversy in this area, it is felt that one should strive for optimal control as this would definitely help in retarding further progression and in quite a few cases there may be improvement in the signs and symptoms of the nerve involvement.

At the same time, I would like to add a word of caution. There have been numerous reports of worsening of the neuropathy when strict control has been sought especially with the use of insulin and insulin "pumps". A closer analysis of these reports has brought out the fact that this worsening was associated with patients undergoing severe and/or frequent episodes of hypoglycemia. This has to be avoided and one should understand that the routine use of these pumps is fraught with too many problems and in fact, even the American Diabetes Association has recommended that these pumps should only be used under very strict guidelines.

Another problem that one sees in clinical practice is that diabetic patients are routinely put on a severely calory restricted diet and given high doses of the anti-diabetic medications in order to bring about a rapid decrease in the high blood glucose levels. Many of our patients are underweight or normal and their nutrition status is borderline. There is no need, and in fact, it may even be detrimental to severely restrict the calories.

Also unless there is a definite indication for a rapid decrease in the blood glucose levels, there is no justification for trying to bring the blood glucose levels down rapidly. Both these factors may be associated with a worsening of the neuropathy. In fact, in quite a significant number of cases, there is a remarkable improvement in the neuropathy (especially in the burning feet type), when patients who were on a very severely calory restricted diet, were put on a high calory diet and the raised blood glucose levels controlled with small judicious doses of insulin with a resultant smooth control of the diabetes and weight gain in the patient.

With regards to diabetic neuropathy per se, there is at present, no specific drug which is routinely available in our therapeutic armamentarium. Aldose reductase inhibitors have been tried abroad with mixed results. The rationale for using these drugs is the proposed role played by the polyol pathway in general, and the aldose reductase enzyme in particular, in the pathogenesis of diabetic neuropathy.

Whilst some initial studies showed a significant effect on diabetic neuropathy, the more recent studies have shown some improvement, especially in the pain parameter and in improving the conduction in the nerve, whilst others have shown either very slight.improvement or none. Thus, the role that would be played by this group of drugs is still open to discussion. At the same time, the consensus evidence is of the opinion that these drugs would show a beneficial effect only if they were started in the initial stages of the diabetic neuropathy, whilst they may not be as useful once the neuropathy has become severe.

The use of drugs such as alpha-lipoic acid, Gamma-linolenic acid and hydroxycobalmin, Vitamins C and E have not been shown to be very useful and would mostly seem to have a placebo like action except in those occasional patients who have a deficiency of these nutrients.

Injections of B1, B6, and B12 are routinely used by many doctors when faced with a patient with diabetic neuropathy. Unless there is manifest evidence of the deficiency of these vitamins in the patient, the injections would be of use only as a placebo. Although quite a significant number of our patients are borderline in so far as the level of nutrition goes, I am sure that most of the vitamin injections that are used for the diabetic neuropathy have no real therapeutic role.

One also has to accept that the management of neuropathies, especially those with a very painful component, is quite difficult. Telling a patient who is in great agony, that the prognosis for his painful neuropathy is quite bright and that he would get relief from his pain in about 12-18 months, is definitely no solace for the patient. During the very acute and painful stage, relative bed rest may be useful.

In acute cases with a significant motor affectation, the strength and mobility of the weakened muscles must be maintained by graded exercises. Special circumstances, like foot drop or a claw hand, may require specially made artificial support up to the time that they recover.

The pain may be so severe that I have seen patients willing to even cut off the affected part in order to get some relief. Due to this level of severity, there may be a tendency to use powerful narcotics analgesics. This should be avoided. The painful condition lasts for about 12-18 months and there is a significant chance that the patient may become addicted to these drugs. Many doctors feel that one may give a short course of these drugs, bring about some relief and then discontinue the drug and substitute it with a milder analgesic which is not addictive. This, in clinical practice does not work in most cases, Usually, the severe pain returns and the patient demands the powerful narcotic analgesic which may have given him some relief before, this viscious cycle definitely predisposing to addiction.

Moreover, the use of narcotic analgesics for treatment of neuropathic pain is controversial. There is no evidence that nonsteroidal anti-inflammatory drugs or other nonnarcotic analgesics have much effect on neuropathic pain, and for a long time it was held that opioids are also ineffective. In fact, that is usually true if these agents are given at standard doses.

Also many of the patients do not seem to get any significant relief even with the use of these drugs and the tendency would then be to either increase the dose of the drug or use even more powerful ones with an increased chance of making the patient addicted. The best compromise is to give relatively high doses of paracetamol and I have given up to 1 gm every four to six hours. There are some patients who do not find the pain very severe during the day but cannot bear it at night and this interferes with sleep. In such cases, the paracetamol may be given only at bedtime. This regimen works in some cases but in others this needs to be accompanied by other drugs like phenothiazines or tricyclic antidepressives, especially given at bedtime. One regimen is to start with very small doses and then titrate the dose against the response. Here again, it is often better to give just a nighttime dose. I use amitryptiline (10-20mg) at bedtime and have found this to give a good reponse. A bed-time dose induces sleep, rarely interferes with daytime wakefullness, especially at these small doses, and tends to break the terrible depression found with the onset of painful neuropathy. Although in some patients, the drug may have to be given round the clock (10mg t.d.s.), most patients do manage with a nighttime dose of antidepressives with non-narcotic analgesia during the day. Unfortunately, tricyclic antidepressants have only modest efficacy, and their use is limited. First, dose titration takes a long time. Second, these agents have a very narrow therapeutic window; beyond a certain dose, most patients experience intolerable adverse effects such as sedation, urinary retention, orthostatic hypotension, or cardiac arrhythmias. In contrast, narcotic analgesics or newer anticonvulsants do not have a dose ceiling. Higher doses of these agents are more likely to cause side effects but will not cause organ damage.

Selected Drugs for the Treatment of Neuropathic Pain Syndromes

Category Examples Comments


Opioids Morphine,30 mg/75 kg po qid Analgesic tolerance is not in-evitable; side effects include constipation

Adjuvant Analgesics

Tricyclic Antidepressants Nortriptyline, start at25 mg/day po and titrate to 150 mg/day Side effects include sedationand mental clouding
Traditional Anticonvulsants Carbamazepine,400 mg po tid Side effects include ataxia
Valproate,750 mg po tid Side effects include ataxia
Novel Anticonvulsants Gabapentin,600 mg po tid Requires titration; side effects include ataxia
Lamotrigine,100 mg po tid Requires titration; side effects include ataxia and rash
Alpha-Agonists Clonidine,0.2 mg/day po Side effects include orthostatichypotension


NMDA Antagonists Dextromethorphan,60 mg po tid Limited experience intreatment of hyperalgesia; side effects include sedation

Systemic Local Anesthetics

Lidocaine,1 mg/kg/hr continuoussubcutaneous infusion Requires close monitoring;limited experience intreatment of hyperalgesia
Mexiletine, 150-300 mg po tid Requires close monitoring;limited experience in treatmentof hyperalgesia

Carbamezapine and phenytoin sodium have been used for the relief of painful dysthesias with very variable results and most patients do not seem to get much relief with the recommended doses. At higher doses, the side effects of the drugs tend to contraindicate the use of the these drugs. Of the two, carbamezapine may be helful in some patients.

The availability of novel antiepileptics renewed interest in this class of agents, especially since some have few side effects.

Gabapentin is one such drug which has shown quite good results. It has been shown to relieve pain and sleep interference and improve mood and some aspects of quality of life The dose needs to be titrated and is usually around 600mg three times a day. In one study, treatment with gabapentin was initiated at 900 mg a day and the dose was titrated weekly to 1,800, 2400, and finally 3,600 mg a day. In clinical practice, titration can proceed much faster, but the need for slower titration should always be kept in mind when the patient has renal disease, as many do. However, gabapentin is very forgiving in this respect; if adverse effects occur, they are uncomfortable rather than dangerous. The range of effective doses is wide. Some patients respond to as little as 300 mg a day, while others require much larger doses. The thing is to start with a small dose and titrate upwards.

It is usually evident within a few days whether gabapentin will be efficacious. Once an effective dose has been achieved, most patients can continue to take that dose. Perhaps 25% of patients will need dose adjustment. In some patients, the effectiveness of gabapentin diminishes within a few months or weeks, even if the dose is increased. The loss of efficacy may be due to the evolution of the disease or the development of tolerance.

The oral antiarrhythmic mexiletine is occasionally given to patients whose pain is not well controlled with other medications. If mexiletine is selected, serum levels should be monitored.

The management of the autonomic neuropathy is mostly symptomatic. Oesophageal manifestations usually respond to metoclopramide. Constipation may be managed by the use of high fiber diets or bulk laxatives. Care must be taken with the use of the latter so as not to push the patient into laxative abuse syndrome. Diabetic diarrhea is a diagnosis of exclusion and can be very difficult to control. Diphenoxylate, loperamide, or clonidine, can be helpful. Small bowel stasis contributes to bacterial overgrowth, causing diarrhea. Antibiotic treatment is recommended for a period of 2 weeks. Bacterial overgrowth does not have to be proven; it is prudent to treat the condition empirically. Doxycycline, amoxicillin, metronidazole, and ciprofloxacin are choices for the treatment of diabetic diarrhea secondary to bacterial overgrowth of the small intestine.

In milder forms of the postural hypotension, the patient would have to be told to be careful whilst getting up and possibly sleep with a couple of pillows under the head. Symptomatic orthostatic hypotension can be very troubling in patients with diabetic neuropathy. Increasing the patient's salt intake, along with use of compression stockings and sleeping with a couple of pillows under the head may help. If these modalities do not improve symptoms, then fludrocortisone may help.

There is very little that can be done about cardiac denervation except to see that the patient avoid undue stress like severe attacks of hypoglycemia, severe exercise, unnecessary cardiac stress tests etc. In case, the patient has to undergo an operation, it may be preferable to avoid general anaesthesia but if this has to be given, the surgeon and the anaesthetist should be forewarned about this problem.

Management strategies for urinary bladder problems include telling the patient to void at frequent fixed time intervals, teaching the patient manoevres to try and get the bladder emptied as much as possible and treating any urinary tract infection early and completely. Occassionally, a patient may be helped by giving a cholinergic drug like bethanechol (10-20mg t.d.s. to q.d.s.) but frequently the side effects preclude the use of this drug. In certain patients, especially women who tend to get recurrent urinary tract infection, it may be worthwhile to put them on a small bedtime dose of an antibiotic to which the previous infection responded. The management of sexual dysfunction is discussed seperately.

Neuropathic ulcers are a major source of morbidity associated with diabetic neuropathy. The best strategy is their prevention which is definitely better than cure. It is essential that all the patients be taught about the importance of, and the need for, constant care of the feet. It is also imperative that all the patients, each time that they are seen by the diabetic specialists, should have the feet closely examined. Foot care is also discussed in a seperate section.

Finally, it would be worthwhile to reiterate the importance of an early diagnosis of diabetic neuropathy, the need for excellent foot care and lastly, but probably, the most important, the desirability of a smooth and optimal control of diabetes.