Treating Diabetes with Insulin

11th January 1922 is a red letter day in the history of diabetes. Culminating a summer of hard work, epoch making experiments, many failures and finally, what could be “success”, Frederick Banting and Charles Best were convinced that they had isolated insulin. Now it had to be proved that what they had isolated would work in humans.

Leonard Thomson, was a 14 year old boy with insulin dependent diabetes. On 11th January 1922, the first injection of insulin, as prepared by the two doctors, was given to Leonard. Subsequently, other patients too were injected with insulin by Banting and Best. It would be worth while to quote the results in their own words. Writing in the Canadian Medical Association Journal, Banting summarised his findings as follows:

“Following the production of what appears to be a concentrated internal secretion of the pancreas and the demonstration of its physiological activity in animals, and under careful control, its relatively low toxicity, we are presenting a preliminary report on the pharmacological activity of this extract in human diabetes mellitus. Clinical observation at this juncture would appear to justify the following conclusions: (1) Blood sugar can be markedly reduced even to normal values. (2) Glycosuria can be abolished. (3) The ace- tone bodies can be made to disappear from the urine. (4) The respiratory quotient shows evidence of increased utilisation of carbohydrates. (5) A definite improvement is observed in the general condition of these patients and in addition the patient themselves report a subjective sense of well-being and increased vigor for a period following the administration of these preparations”.

Euphoria accompanied the availability of insulin. It was hoped that we would be able to prevent the acute and also the chronic long-term complications of diabetes. To a certain extent, the hopes have proved true. There is no doubt that insulin therapy has allowed a number of people with diabetes to live a much longer and better life than it would otherwise have been possible.

In 1982, a banquet was held in New York attended largely by insulin dependent diabetics. The Chief Guest, one of the first individuals who had received insulin from the hands of Banting himself put matters in the right perspective, ” If today, this hall resounds to the vibrant voices of living beings, and not to the ghoulish wailings of our ghosts, it is entirely due to the great courage and spirit of two radicals who had the confidence, some would call it obstinacy, to believe in themselves, even in the face of adversity. If the world does make progress, it is due to such free thinkers, rather than those who would rather follow the straight, safe and well trodden path.”

At the same time, we must accept that the initial euphoria that “diabetes had been defeated” was, and is, misplaced. Whilst it has been possible to prevent, or adequately manage acute complications of diabetes like ketoacidosis, we have failed to prevent or treat some of the long term complications of diabetes.

Whilst it would be naïve to feel that the longterm complications of diabetes are so simple that they would have a single etiology, there is enough evidence to show that good control of diabetes can prevent, to a large extent, and definitely alleviate most of these dreaded complications affecting the eye, kidneys, nerves etc. Then why is that in spite of having insulin in our therapeutic armamentarium, we are unable to offer optimal control and thereby prevent these very complications? Could it be that we have been unable to optimise the use of insulin thus far?

The most important stumbling block to the optimal use of insulin seems to be the inablity to define and understand the precise role that insulin therapy should play in our management of diabetes. Very often, patients who should preferably be on insulin are treated with massive doses of oral agents without effective control of the blood glucose levels, whilst many obese Type 2 patients, who have a fair amount of endogenous insulin, and should be managed with diet and exercise are administered excessive doses of insulin in an attempt to bring about control. I think that it is this “abuse” of insulin with all its attendant problems that prevents us from offering our diabetics the best possible “use” of insulin therapy and as a consequence, optimal management of their diabetes.

So which patients should receive insulin therapy?

Patients Who Should Receive Insulin Therapy 1 ) It is obvious that insulin therapy is mandatory for all Type 1 patients;

2) Any diabetic with significant chronic complications like coronary artery disease, cerebrovascular disease, peripheral artery disease, neuropathy, retinopathy, nephropathy etc. should receive insulin;

3) Any diabetic with an acute problem like severe infections, injuries etc., should be given insulin;

4) All diabetics with tuberculosis should always be given insulin;

5) Most diabetics undergoing surgery may need to be stabilised on insulin;

6) Pregnant diabetics, if not tightly controlled with diet and exercise should be given insulin to ensure tight control;

7) Any patient, even if well controlled on oral agents, who shows even the slightest evidence that may contraindicate the use of oral agents. For example, a liver disorder should entail stopping of the tablet and the patient should be shifted over to insulin;

8) All underweight patients and those with malnutrition related diabetes should be given a high calorie diabetic diet and covered with insulin;

9) Patients with INSULIN-REQUIRING diabetes, even though they are not prone to ketosis, should be identified and their management supplemented with insulin to get the best possible control;

I really feel that once a decision is made that a person would do better with insulin therapy, it becomes essential for the doctor to be able to convince the patient that insulin injections are the best for the patient. No compromise should be entertained about administering it. It is true that many of our patients have a great aversion to taking insulin and overcoming this resistance is quite a difficult problem. Unfortunately, many doctors give in to this resistance. Although, they may offer many excuses for this, I am quite skeptical of most of these. I feel that the real reason is that they are afraid of losing the patient to some other doctor who will give the patient oral tablets. I think that this attitude is indefensible. If the patient needs insulin, then there is no compromise in so far as the doctor is concerned. Having treated quite a few patients with diabetes, a substantial number of whom are on insulin, I feel that the major contributing factor to the initial patient resistance is the totally false and anecdotal concepts that they have about the problems and dangers associated with insulin therapy.

Simple explainations to clear these misconceptions, accompanied with a certain amount of firmness and occasionally a threat about the potential loss of eye, limb and/or life, is often sufficient to overcome the initial patient resistance, Later, once patients are well stabilised on insulin, they feel so much better that they often opt to continue the insulin therapy. A few patients benefit by being hospitalised,. especially in a diabetic unit. Seeing other patients receive, or take their own, injections of insulin, somehow seems to make the concept of taking insulin injections much more acceptable. In these diabetic units, they also see patients with foot problems which may have needed amputation, patients with vision loss etc., and this too helps in their resolve to take insulin. I have known many “adamant” patients accept insulin injections after a short period of hospitalisation.

At the same time, it is essential to add a word of caution.

It is widely felt that high insulin levels in the blood may be an important factor in causing associated problems such as obesity, hypertension, lipid disorders and atherosclerosis (Syndrome X). Thus, one should guard against a misplaced enthusiasm to use insulin in all Type 2 patients unless they fall into one of the categories listed above.

As we have discussed in the section dealing with oral agents, the hyperglycemia in many Type 2 patients is mainly due to peripheral resistance to the action of insulin. In such patients the optimal treatment would be to use drugs insulin ” sensitisers” such as glitazones and/or metformin. Even in those who also have a decreased secretion of insulin from their beta cells, it may be worth while to see if the use of small doses of insulin “secretagogues” would optimize the control.

It is only when a rational use of oral agents fails to provide the desired levels of control that one would add insulin to the therapy under normal circumstances.

There should be NO misplaced enthusiasm to use insulin in all diabetic patients. Of course, the pendulum should not completely swing the other way. Those who need insulin MUST be given it!

After deciding that a patient will need to take insulin, the next step is to decide which insulin one will use.

Such a great plethora of insulins are available for clinical use that there is often some confusion about which insulin to use. In this context, it is interesting to note that even 10 years back more than 43 different varieties of insulin were available in the U.S.A. and that these were made by just 3 companies!

But the picture much more clear, if one realizes that when choosing the insulin, there are two major areas. The first, which in my opinion is relatively minor, is the species of insulin one will use. The more crucial area is the “time of activity” characteristics of the insulin.

Species of Insulin It is really unfortunate that there has been so much controversy generated regarding the species of insulin to be used. I think that this is a relatively minor matter, and the time, money and effort spent in generating this controversy could easily have been put to better use.

The insulins available for routine clinical use are the beef, porcine and human insulins. Basically this means that the beef or Bovine insulins are got from beef pancreas, the porcine insulin is got from pig pancreas whilst the “human” insulin is manufactured through genetic engineering.

Today, all insulins available are of the pure variety and the contamination is not an issue.
Beef insulins differ from human insulin in three amino acids, whilst porcine insulin differs from human insulin in only one amino acid. Thus, porcine amino acids are less immunogenic than beef insulins. At the same time, it must be mentioned that this does not significantly affect the efficacy of the bovine or porcine insulins in most patients.
Human insulins are replacing the other insulins in most developed countries, but its cost is the inhibiting factor to its widespread acceptability in most developing countries.

I personally feel that the decision about the species of insulin to be used should be a matter best left to the patients and their doctors. At the same time, many experts feel that that there are certain circumstances where it may be better to use the “human” variety of insulins.

Patients who should preferably use human insulins are: a) All patients who are on beef or porcine insulins and manifest resistance due to the presence of antibodies;

b) Patients requiring intermittent therapy, i.e. patients with gestational diabetes, those undergoing major surgery, patients with acute infections, etc., who otherwise may be controlled on diet, with or with out OHA’s, should use human insulins.

c) Patients who require very large doses of beef or porcine insulins (>80 units/day), may benefit with change over to human insulins.

Time-Activity Characteristics

This is an area of major importance and a thorough understanding of the time of activity character of the insulin which is to be used is essential for the correct initiation and more importantly, of the adjustment of the insulin doses.

What does one mean by time-activity characteristics. In simple terms it implies knowledge about when, after being injected, a particular insulin would start showing its action, the time when peak activity would be seen and also the time when the activity of the insulin would cease.

In India, presently, the following insulins are available:

Rapid acting

This is seen in “special” types of insulin which are called insulin analogs. Human insulin has been further modified to alter the time of its activity. Presently only Lyspro is available in India, although insulin aspart should soon be available for clinical use.

Short acting;

Short-acting (regular) insulin usually reaches the blood within 30 minutes after injection. It peaks 2 to 4 hours later and stays in the blood for about 4 to 8 hours.

Intermediate acting;

Intermediate-acting (NPH and lente) insulins reach the blood 2 to 6 hours after injection. They peak 4 to 14 hours later and stay in the blood for about 14 to 20 hours.

Intermediate-acting insulins include lente and NPH. Insulin preparations with a predetermined proportion of NPH mixed with regular, such as 70% NPH to 30% regular, or a 50/50 mix are called intermediate acting for purposes of classification, although their activity characteristics would be different from either only NPH/Lente or only Regular insulins.

Long acting;

Long-acting (ultralente) insulin takes 6 to 14 hours to start working. It has no peak or a very small peak 10 to 16 hours after injection. It stays in the blood between 20 and 24 hours.


However, each person responds to insulin in his or her own way. That is why onset, peak time, and duration are given as ranges.

Fortunately, the picture becomes much more clear, if it is realised that from a practical and clinical viewpoint, insulins can be divided into two main groups, depending on their time course of action. These are the “short-acting (SAI)” and the “intermediate acting (IAI)” insulins. Within these two main groups, the insulins may differ in their source of origin and other details, but for all practical purposes, they are similar in their mode and duration of action.

The rapid acting insulin analogues are very useful for use in special circumstances and whilst they can be used in place of the SAI’s, this has not yet become standard practice. Longer acting insulins like Ultralente are very rarely used in India. 



Traditionally, one starts with a dose of an intermediate acting insulin (IAI) given before breakfast; I am often asked whether there is any formula whereby one could calculate the dose that should be given. Let me make it quite clear at the outset that there is no such magic formula! When I was an undergraduate in medical college, I was taught that one should start with a dose of intermediate acting insulin which was one tenth the level of the fasting blood glucose. In other words, if the FBG was 250 mg%, then one should give 25 units of the IAI. This just does not work and frankly could lead to quite serious problems for reasons which I shall discuss below.

So, how much insulin should be given? I usually start with a small dose of IAI, unless there are adequate reasons for lowering the blood glucose levels very rapidly. In practice, this words out to an initial dose of an IAI of about 8-12 units. There are many valid reasons for starting with this small dose.

There is no method by which we can judge the sensitivity of the patient to insulin. Although it has been said that the thin patients are more sensitive to insulin than normal weight or overweight patients, this is just a generalisation. There is a great amount of difference in the sensitivity to insulin. Even patients with the same weight and similar blood glucose levels would tend to differ in the response that they show to a certain dose of insulin. As we are dealing with patients and not generalities, it would be prudent to start with a small initial dose, judge the response in the patient and adjust the dose accordingly. This will allow the patients to escape from the hazards of hypoglycemia especially if the patient turns out to be very sensitive to insulin injections and a large dose of insulin has been given. In fact, one of the commonest emergencies many of us are called to treat are patients who have gone into severe hypoglycemia when they have been given large doses of insulin due to the fact that the fasting blood glucose levels were very high !

I have discussed how patients may differ in their individual characteristic with regards to the time course of action of the insulin. There may be “early”, normal or “late” activators. It is not possible to forecast how a patient would react. A high initial dose may cause the patient to have hypoglycemic reactions at odd hours, which would be quite problematic. A small initial dose would not only protect the patient from these vagaries, but also allow us to judge the type of activator that the patient is.

To sum up, I would once again like to emphasise that the dose of insulin is determined empirically and that it is always better to start with a small dose and adjust according to the response, unless there are relevant reasons for the rapid lowering of the raised blood glucose levels.

After I start the patient on 8-12 units of IAI, given before breakfast at about 8 a.m., I ask the patient to do home blood glucose monitoring. In the absence of home blood glucose monitoring, I ask the patient to estimate his blood glucose levels every 3-4 days and adjust the doses according to these reports. Many patients do not like testing their blood every 3-4 days. It is worthwhile to explain to them that this is required only till the blood glucose levels are normalised. This small explaination makes the patient much more willing to accept intensive testing of blood in the initial stages until his blood glucose have stabilised to acceptable levels. Of course, IDDM patients, pregnant women etc., may have to continue to keep a close and frequent check in order to maintain good control.

Before we discuss the adjustments in the treatment, it would be worthwhile to be very clear in our minds about the aims of the treatment. We would like the patient to have “acceptable” blood glucose levels throughout the 24 hours of the day. This means that the fasting blood glucose levels, the premeal and the postmeal blood glucose levels as also the levels in the interim period should all be stabilised at these acceptable values. It goes without saying that this should not be accompanied by frequent and/or severe episodes of hypoglycemia. Once, this aim of the treatment is clear, the adjustments in the doses of insulin will be much easier to understand.

I review the patient every week until the patient is stabilized on the insulin therapy if I am seeing the patient in the clinic situation. I do not feel that any person with diabetes, unless he has some complication, should be admitted into hospital just for controlling his diabetes. Even if a patient is admitted, I do not think that the insulin doses should be changed more frequently than every 2 days.

Let us now discuss the various types of response that we would tend to see to the initial dose and our management of these responses.


A few of the patients would show during the followup that their fasting and postmeal blood glucose levels are within normal, acceptable limits. This means that the patient is controlled and the dose of insulin should be continued. I try and decrease the dose by about 2 units and see whether the control is still acceptable. If this is so, then I continue to gradually decrease the dose of the IAI (about 2 units at a time) until I find that the blood glucose values are higher than acceptable. The dose is then increased by about 2 units and the patient called back after 15 days to see about his control. If he continues to remain in control, then he is called for a routine followup after 6-8 weeks. I would like to make it clear that this scenario is rarely seen. SCENARIO 2

The more usual response is that both, the fasting and the post lunch blood glucose levels continue to remain high and that the urine, if tested, shows glucosuria throughout the day. In such cases, the dose of the IAI is gradually increased (again by about 2-4 units at a time) and the patient examined every week. This would be repeated until either, the patient shows adequate control with the fasting and the post meal blood glucose levels within acceptable limits and no glucosuria throughout most of the day, or one of the following scenarios would occur.


Whilst, the fasting blood values would be within acceptable limits, the post lunch blood glucose values would be higher than normal. This is often accompanied by high post breakfast blood glucose values, whilst the evening or the pre-dinner values are acceptable. This is quite commonly seen. The action of the IAI normally, starts after 3-4 hours, peaks about 8-12 hours and lasts for 18-24 hours. Therefore, if the injection of the IAI is taken before breakfast, say around 8-8.30 a.m., then the peak activity would occur in the evening or around the pre-dinner time. This peak activity could cause the blood glucose values at this time to be acceptable. As the activity of the IAI is supposed to start after 3-4 hours and then gradually increase, it may not show sufficient activity to cover the post breakfast and the post lunch levels which would occur within 1-2 hours and 5-6 hours after the injection when the activity of the IAI would just be gradually starting.

If we were to increase the dose of the IAI, so that it would show stronger activity after 5-6 hours and may help in correcting the raised post lunch blood glucose levels, but then it is quite possible that at the time of its peak activity, the action of the IAI would be strong enough to cause hypoglycemia which would be manifest in the evening or just before dinner. “Late” activators could go into hypoglycemia later in the night, possibly in sleep. Therefore, it may not be possible to increase the dose of the IAI in order to control the post lunch blood glucose values, without exposing the patient to hypoglycemia later.

The answer to this problem is to add a small amount of short acting insulin (SAI) along with the dose of the IAI, in the morning injection. The rational for this is that the SAI would help in controlling the post lunch values! At first sight, this may appear doubtful. The action of SAI starts within about 30 minutes of the injection, peaks around 1-2 hours and lasts for 4-6 hours. How could the addition of SAI under these circumstances, given before breakfast help to control the post lunch values? Clinically, this works!

It could help through various mechanisms. It could act on the glucose levels in the post breakfast period and lower them so that the pre-lunch blood glucose values would be lower than before. This would lead to a lowering of the post lunch blood glucose values. As an example, let us suppose that with the prescribed diet, the rise in the level of the blood glucose after lunch would be in the vicinity of 80mg%. If with only IAI, the pre-lunch blood glucose values would be 150%, then the values seen after lunch would be 230mg%. But, if the addition of SAI causes the pre-lunch blood glucose levels to be about 100mg% (this would be possible due to the fact that the early activity of the SAI would tend to lower the blood glucose values seen after breakfast and also before lunch), then the addition on the lunch increase of 80mg% would cause the post lunch blood glucose levels to be about 180mg%. This would be a definite improvement over 230mg%!

More importantly, when the patient takes SAI regularly for a while, the time characteristics of its activity change somewhat. It has been shown that activity of the SAI starts slightly later, peaks later and also lasts longer. Therefore, activity of the SAI would still be present to a significant extent after lunch, even if the injection is taken before breakfast. It should be remembered that in India, breakfast is usually taken at about 8-8.30 a.m., whilst lunch is at about 1 p.m. Therefore, even with a time action lasting 6 hours some of the effect of the pre-breakfast SAI would be seen in the post lunch period, With the prolongation of the time of activity of the SAI with prolonged use, this effect would be all the more significant!

Before adding the short acting insulin (SAI) to the morning injection of the intermediate acting insulin (IAI), I check the blood glucose levels of’ the patient in the evening and also just before dinner, i.e. between 4-8 p.m. If these levels are seen to be in the “well-controlled” range; then I slightly decrease the dose of the IAI whilst adding the SAI. This precaution is necessary to protect the patient from the hazards of hypoglycemia occurring at these times. If this is not done, the decrease in the afternoon blood glucose levels, and consequently the evening blood glucose levels, brought about by the SAI, added to the low in the blood glucose levels brought about by the peak activity of the morning IAI would make the patient prone to episodes of hypoglycemia.

Often, this judicious use of an SAI along with the IAI brings about adequate control in a number of patients, with acceptable fasting and post prandial blood glucose levels and the absence of glucosuria (in a patient with a normal renal threshold for glucose).


A situation which is less frequently seen than the one described above, is one where the morning, post lunch, and even the evening and pre-inner blood glucose levels are within acceptable limits. But, the fasting blood glucose levels are consistently above normal and acceptable values. We have seen that the action of the IAI lasts for around 24 hours and therefore an increase in the dose of the IAI could help, in theory, to control the raised fasting glucose levels. But, this would in clinical practice, lead to unwanted hypoglycemia during the day, as we have seen that in this case, except for the fasting values, the blood glucose levels are well controlled during the day. Therefore, if one is sure that one is not dealing with the Somogyi phenomenon or the Dawn phenomenon, then the treatment strategy would be to add a small dose of IAI to be taken by the patient before dinner, or at bedtime. This is basically putting the patient on a type of Multiple Dose Regimen (MDR).

This strategy would help in normalising the raised fasting levels. The peak activity of the IAI is normally seen after 8-12 hours. As dinner is usually taken around 8-8.30 p.m. the peak activity of this would be seen the next morning in time to help in normalising the fasting levels. Early activators who show a peak activity after 6-8 hours would be able to take the injection around bedtime such that the peak activity would occur in the early morning hours (and help normalise the raised fasting blood glucose levels) rather than in the middle of the night.

When adding an evening dose of IAI, I decrease the morning dose of the IAI injection. This would protect the patient from hypoglycemia. We have seen that the morning dose of the insulin is sufficient to control the blood glucose throughout the day, but the waning of its overnight activity allows the blood glucose to rise to high fasting level. We have also seen that adding more insulin would further depress the blood glucose levels and therefore, we cannot increase the morning dose of insulin. But, when the evening dose of insulin is given, it must not be forgotten that some of its activity will be present for 24 hours and would add to the insulin effect even during the day. As the blood glucose levels are already well controlled, this further insulin effect would tend to bring the blood glucose down further and possible depress them to hypoglycemic levels! Therefore, even though the dose of the morning injection is reduced the additive activity of both the morning and evening insulin injections would bring about adequate control during the day and the night.


This scenario is quite the same as that seen in scenario 4. Here, the addition of the evening dose of IAI helps in normalising the fasting blood glucose levels, But, in this case, the post dinner blood glucose levels tend to be higher than acceptable. The treatment strategy here is to add a small dose of SAI to be taken before dinner. This would help in keeping the post dinner blood glucose levels down, whilst the two injections of IAI (taken at morning and evening) would, both, keep the blood glucose levels in the well-controlled range throughout the rest of the 24 hour period.

Although, I have divided the response of the patient into these 5 distinct scenarios and have discussed the treatment strategies for each, it should be clear that one would rarely see such clearly demarcated responses. What we see in practice is often a permutation and combination of the scenarios that we have discussed. Obviously, the treatment strategies would also need to be adjusted to meet these variations,

But from the basic principles of dose adjustment seen in the 5 scenarios and an understanding of the time course of activity of the two groups of insulin, it should be quite easy to evolve a specific strategy to combat most of the various responses that patients would tend to show to the initial dose of the intermediate acting insulin.

It is fortunate that the vast majority of the patients that we see in clinical practice have a significant amount of endogenous insulin secretion and are therefore much more easier to manage with these conventional methods of giving insulin. One modification, or “short-cut” of this conventional method which I have found to be quite useful in clinical practice is to start with a small mixture of the short acting and the intermediate acting insulin given before breakfast. The response to this is judged after 3-4 days. Depending on this, the dose is adjusted and the patient retested after a few days. When a stage is reached that the patient is receiving about 25-30 units of insulin in the morning injection, I do not increase the dose any further, I examine the patient for the presence of any conditions which may be hampering the blood glucose control. Some of the common conditions which reduce the insulin action include, infections, obesity, insulin receptor and post-receptor defects, ketosis, destruction of insulin at the site of injection, true immune mediated resistance, hormonal conditions associated with excess of cortisol, Growth hormone and thyrotoxicosis, associated use of drugs which increase glucose intolerance. If present, this is obviously treated. In the absence of any such associated condition, I add a small dose of an insulin sensitiser such as a glitazone or metformin. If the postprandial blood glucose levels are in the very high range, acarbose would be an helpful addition. I usually would not add a sulfonylurea to the insulin, as these, especially the older ones, basically try and increase the endogenous insulin secretion. This is not too important once we are injecting insulin! It is often seen that the addition of this small dose of oral agent tends to bring about quite an acceptable control!

I am often asked why I stop increasing the dose after about 25-30 units and add an oral agent in a patient with Type 2 diabetes. This is absolutely empirical and based on the fact that the amount of insulin secreted by a normal person during the day is about 35-40 units. I also feel that when a dose of insulin taken as on injection reaches these levels, further increases rarely help, but the patient may benefit from splitting the dose into two smaller injections taken before breakfast and before dinner. Many patients would be extremely reluctant to take the second shot of insulin and in quite a few of these, the addition of small dose of an oral agent to the morning injection of insulin does bring about control. It goes without saying that IDDM patients and pregnant women who need insulin are not given the oral agents but they may need to split the injection or even go in for MDR.


Insulin therapy does not end with the act of injection itself. If the injected insulin is to be effective, it must get absorbed from the subcutaneous site, enter the bloodstream and thus reach the insulin receptors where it would exert its activity. Unfortunately, not much attention is paid to this aspect inspite of the fact that it plays a crucial intermediate role in our quest for optimal insulin therapy.

In clinical practice, one often comes across patients who show an odd or unexpected reaction to the insulin. This response may take several forms. Some patients show little, if any, response to insulin in spite of the fact that the insulin is potent, injected correctly and at doses where one would expect to find at least some response. There are patients who show erratic timing of the activity of insulin. As an example, a patient may show peak activity to an injection of the intermediate acting insulin after 8-12 hours, as is normally expected. But often he shows activity which is in keeping with those who are “early” activators and at other times, the time of peak activity would classify him as a “late” activator! Now a person does not change his characteristic activity pattern suddenly or often. There are patients who in spite of a correct dosage of insulin continue to show a high 2-3 hour post prandial blood glucose levels and then go into hypoglycemia at a later period, say after 5-7 hours! One also comes across patients who are well controlled on a certain dose of insulin and are fairly stable. The patient may suddenly go into hypoglycemia even though the routine daily dose of insulin has been injected. Patients with Type 1 diabetes may even relapse into ketosis.

Faced with such patients, our first reaction is to put the blame squarely on the patient. One feels that the patient has not followed his diet regime or has made a mistake in the injection, either in the dosage or techniques. I do not deny that this may be true in quite a few patients, but I feel that when one is confronted with a patient who often manifests such bizarre responses, one should consider that the response could be due to factors that may affect the absorption of insulin from the site of injection, thereby varying the amount of insulin that enters the bloodstream and consequently, show a varying response. Since, we want the patient to show a predictable and consistent response, it is imperative that we understand the reason for this varying absorption rates.

The basic time lag between the injection of insulin and the appearance of that insulin in the blood stream is that the commercial insulins which are available are hexameric. This means that six molecules of insulin are “joined” together. These have to disperse into monomeric ( single molecule of insulin) or dimeric ( two molecules of insulin) in the extravascular region before the insulin can be absorbed into the blood stream.

The differences in the rate of absorption is due to the differing milieu in which the hexameric molecule has to disperse. All the factors which we used to make use of to alter the rate of absorption basically acted through changing the rate of dispersion of the hexameric insulin. Now with the availability of insulin analogues which are in the monomeric or dimeric state and therefore almost instantly absorbed into the blood stream after being injected, the manipulations which we used to do in the past are no longer necessary and now it even seems laughable about some of the things we used to recommend to the patient!

The best solution for a patient showing erratic absorption is to change them over to the insulin analogues.



Many would not consider this as a complication of insulin therapy, thinking of this as more of an excessive side effect of the therapy. Hypoglycemia is discussed in a seperate section.


Insulin does cause salt and water retention. Thus, patients may complain of edema, rapid weight gain and a feeling of bloating. The edema is usually mild and usually disappears after a few weeks. Occassionally, it may be more severe and may require the use of a diuretic. In patients with cardiac problems, hypertension and renal problems, this salt and water retention may create additional problems and some modification may be required in their treatment of these conditions.

Allergy Allergic reactions to insulin have now become very rare now that the “pure” varieties of the insulin are available and due to the increasing use of human insulins.

All the same, they do occur and the treating doctor should be aware of this. Unfortunately, when we think of allergic reactions, we usually refer to the full blown systemic reaction.

Allergic reactions to insulin may be local or systemic. One should be aware that a patient may exhibit a local reaction even with the first injection although it is more common to be manifest after a few injections have been taken.

The local reaction may be “immediate” (manifesting within 15 minutes to two hours of the injection) of “delayed” (seen within 6-24 hours of the injection). Some patients may exhibit both the immediate and the delayed reactions. The allergic lesions may be characterised by local pruritus, erythema, and indurated areas which may be from 1-5 cms. in diameter. The lesions may gradually increase in intensity for up to a day and then gradually subside within a few days.

Some patients show a true systemic form of insulin allergy. In some patients, severe systemic reactions are preceeded for a few days during which the severity of the localised reactions seems to increase significantly. Such an occurence should be taken as a warning that a severe systemic reaction is in the offing and adequate steps should be taken to avoid this. Fortunately, such severe reactions occur very rarely (about 0.1% of patients receiving the newer insulins).

When one is faced with a mild form of localised insulin allergy, one should, first make sure that the insulin is being injected subcutaneously and not intradermally, and is at room temeperature. The patient may be allergic to the alcohol used to disinfect the skin. More importantly, some patients are allergic to the to the protamine contained in the longer acting insulins, although this is quite rare, and such a patient may tolerate the lente insulins better than the NPH insulins. In case the patient is shown to be allergic to protamine, it should be very clearly mentioned in his case records, especially if the patient is to undergo any cardiovascular surgery as a large amount of protamine is often used post operatively.

If in spite of this, the patient still continues to show an allergic reaction, the next step would be to shift the patient over to the rDNA derived human insulins.

As with the case of localised allergy, it is more prudent to prevent the systemic allergy from occuring rather than in trying to manage it once this has occurred! Interrupted insulin therapy, especially with the animal insulins should be avoided and such patients should be preferably treated with rDNA derived human insulins. But if a patient taking animal insulins does show systemic reactions, then again it would be better to shift to the rDNA derived human insulins.

This may help in the occasional patient. The reason why we see such a rare positive response, is that systemic insulin allergy is due to the presence of antibodies and these are known to be directed against the insulin molecule itself rather than against any impurity contained in the insulin. Therefore, once the antibodies have been formed and the allergic reaction is manifest, a simple change over to the human insulin may not be effective. As I have shown before, the beef and the porcine insulin molecule differs from the human insulin molecule by three and one amino acid respectively. Whilst this discrepancy may have initiated the antibody formation, once the antibodies are formed, they would be directed even against the other parts of the insulin molecule! Thus, removing the source of the antibody formation would not stop the allergic reaction but may help in decreasing further progression.

When faced with a seemingly intractable problem of insulin allergy in a patient with diabetes in whom insulin therapy is mandatory, the only option left before us is to desensitise the patient. This is best left in expert hands.

Insulin Resistance True insulin resistance is usually defined as a situation in which the patient receives more than 200 units of insulin per day for two or more consecutive days in order to try and achieve a control of the blood glucose levels. In my opinion, this definition is quite arbitrary, and as the amount of insulin secreted by the pancreas in a normal person is about 35-40 units per day, it is logical to assume the presence of some degree of insulin resistance when OPTIMALLY administered insulin doses exceed a total of around 50-60 units daily, especially in a “compliant” patient.

I have purposely tried to highlight the point about optimal administration of the insulin, I have seen numerous patients who were receiving around 100-120 units of insulin in a once a day dose and showed a poor control. When many of them are shifted over to a twice a day regimen (both times being given a mix of the short and intermediate acting insulins) they show an excellent control with the daily total insulin administration of around 40-50 units! Surely these patients are not really resistant although one could say that they manifest “iatrogenic” resistance!

In any case, whenever we talk about insulin resistance, we need to be quite clear in our minds as to what exactly we are referring to. From a purely theoretical viewpoint, true insulin resistance is a condition where there is an immunological barrier to insulin action. This is due to the presence of specific antibodies that interfere with the action of insulin so that massive doses are required before any therapeutic response is seen.

At the same time, there are many other causes and conditions, which for varied reasons, manifest an antagonism to insulin action such that more than normal doses of insulin are needed for a response. These conditions cannot be considered to be causes of “true” insulin resistance as there is no immunological barrier to insulin action involved.

These cause “relative” insulin resistance and from a purely clinical view, they constitute the vast majority of cases where we find the insulin requirements to be much larger than normal. I would therefore classify insulin resistance into: True insulin resistance, caused by immunological mechanisms and relative insulin resistance caused by a variety of factors which manifest as increased insulin requirements, but which do not have an immunological basis.

As my approach is more clinical, and as cases of relative insulin resistance are so much more common and important, I would like discuss a few of the more common and important factors first. I feel that it is imperative to reiterate that the increased insulin requirements are in spite of optimally administered doses in a compliant patient.


It is quite well know that in the presence of any infection, diabetic control deteriorates and the insulin doses may have to be increased quite a bit. Con- versely, when faced with a clinical situation wherein the control deteriorates without an obvious cause, or where it is difficult to control the blood glucose levels in spite of optimal management, and relatively large doses are required, it is imperative that a thorough search be made for the presence of any obvious or occult infection. In my experience, the most common infection in our country would be active tuberculous infections. The association of diabetes with tuberculosis is frequent enough to justify routine ruling out of tuberculosis not only when faced with a case with “difficult” control but in all cases! The other common problem is caused by urinary tract infection which may not cause symptoms or signs that would bring it into clinical diagnostic consideration. Thus, a routine urine examination is also mandatory to rule out a urinary tract infection.


Type 2 patients who are overweight and especially those with a raised waist to hip ratio often manifest resistance to the action of insulin. It is well documented that in most of these patients, the pancreas secretes more than average insulin. Yet they show a poor control and the situation may not become better even with the injection of additional doses of insulin. In such patients the problem seems to lie with the peripheral insulin receptor and the best management for such patients would be to try to optimise the weight of the patient with diet, exercise and the possible use of a small amount of sensitisers such as metformin and/or the glitazones.


Another relatively common cause of relative insulin resistance is the presence of an increase in the activity of those hormones that have an antagonistic effect to that of insulin. These are hormones like glucagon, growth hormone, cortisol and thyroid hormones. From a clinical viewpoint, the commonest condition that causes an increase in insulin requirements in so far as these hormones are concerned is thyrotoxicosis. This conditions is not as rare as one may imagine especially in the young. I have seen patients whose weight loss has been attributed to the diabetes itself and whose diarrhea has been thought of as due to amebiasis or even autonomic neuropathy! When the thyrotoxicosis is controlled, one sees a good diabetic control with small doses of insulin.


I have already discussed the factors associated with the bioavailability of insulin from the site of the injection. It is obvious that if the injected insulin is not properly absorbed from the site of injection, it would not reach the receptors in sufficient amounts to be effective, thus creating a relative insulin resistance. Besides this, in the rare patient, the increased insulin requirement may be due to an increased amount of degradation of insulin at the site of injection.

Finally let us come to the problem of true insulin resistance. This is due to an immune based mechanism. All patients who receive insulin therapy, especially with the older conventional insulin, do show a presence of antibodies, but the titers of these antibodies are in such a low range that this seldom causes any clinical problem. These titers are usually in the range of 10 units per liter of serum. In patients showing a true insulin resistance, these titers may range from 100 to as high as 50,000 units per liter of serum.

Beef insulin differs from human insulin in three amino acids and the porcine insulin differs from human insulin in one amino acid. Thus, beef insulin would have a greater tendency to give rise to the antibodies as compared to porcine insulin. But once again I would like to make it clear that in most instances these antibodies do not reach a significant level. Higher antibody levels can also be seen in people who receive intermittent therapy especially with the older insulins. The reason why the antibody titers assume a clinical significance in only a relatively few patients is not quite clear but may be due to differences in immune responsiveness of individual patients.

Although true insulin resistance can be managed in specialized centers it is a complex matter and the best way to avoid this problem is to use the newer insulins which are now available, and also by avoiding needless intermittent insulin therapy.

Lipodystrophy and scar formation

Whilst injecting insulin, one should invariably rotate the site of the injection so that no area of the body about 3 cms. in diameter, should receive the injection more than once every three to four weeks. One of the complications of not following this rule, is that when the insulin is repeatedly injected into the same area, the skin and subcutaneous tissue may become thickened and scarred with the formation of insulin lumps. As injections in this scarred area are relatively painless, the patient keeps injecting the insulin into this area. This exposes the patient to bizzare reactions as the absorption of insulin from such a site is delayed and wholly unpredictable.

Insulin lipodystrophy comprises both, lipoatrophy as well as lipohypertrophy. The insulin induced lipoatrophy is basically a loss of subcutaneous fat at the site of the insulin injections. Insulin induced lipoatrophy may not seem to very important from a purely clinical viewpoint but it may give rise to considerable cosmetic disfigurement and many of the younger patients would rather discontinue the insulin therapy than accept these unsightly blemishes.

It has been shown that rotating the site of the injection, using a slightly longer needle so that the injection goes deep into the subcutaneous tissue, or even in- tramuscularly, are some of the simple manoevers to avoid lipoatrophy, as is the routine use of the newer and purer insulins. In fact, if the newer insulins are injected into the site of lipoatrophic areas, these will fill out again! This is due to the new deposition of fat in the lipoatrophic areas and may take about 3-4 weeks. The areas which have thus filled out should get injections of insulin every three to four weeks or else it is possible that these areas may lose some of the fat again.

Of course, now that only the pure monocomponent insulins are available for use, one will see less and less of lipoatrophy.

Lipohypertrophy is much more rare than lipoatrophy. It basically means that the subcutaneous fat cell at the site of the insulin injection undergo hypertrophy and is presumably a manifestation of the lipogenic action of insulin. One factor that predisposes to lipohypertrophy is the repeated injection of insulin at the same site. Once slight hypertrophy develops, the patient may continue to inject his insulin at the same site as this is less painful than other normal areas. Since lipohypertrophy is related to the inherent lipogenic action of insulin, it can occur even with the newer insulins. The best management of established lipohypertrophy would be to avoid using that site for the injection for a long time in order to allow the increased fat to resolve by itself.

Source : Diabetes India


About Anuj Agarwal

I am a professional with varied working experience of more than 18 years in India, USA and Middle East. My areas of work and interest are Information Technology, Corporate Social Responsibility, Health and Wellness. Please feel free to contact me through the blog or write me at
This entry was posted in Diabetes at and tagged , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s