Diabetes Mellitus and Thyroid Disorders
To begin with, the talk of diabetes, I want to give you a definition, a working definition the sort of phenotype of the child - or adult for that matter - with type I or insulin-dependent diabetes. Such patients are insulinopenic by definition. That is to say, they are absolutely dependent on exogenous insulin for survival. Without insulin they are prone to ketoacidosis. They typically, without insulin, would be lean and complain of recent weight loss. The development of the disease actually is rather insidious. It may occur over years, although the presentation phase can be fairly abrupt. You usually see this form of diabetes under age 40, but there are reports of individuals as old as 87-years-of-age developing type I diabetes … for whatever that’s worth.
Now a couple of words about the epidemiology of diabetes in the United States. Now the incidence of the disease - and I will remind you that that’s defined as the number of new cases in a defined population per unit time … we are talking now of under 19-years-of-age - is 16 new cases per 100,000 children per year, or 1.6 per 10,000. The prevalence, which is the number of existing cases in a defined population in a moment of time, is 140 cases per 100,000 or 1.4 cases per thousand children walking around. Of 1,000 children out there, 1.4 of them have diabetes. At a large high school of 3,000 students, 4.2 children with have diabetes, type I. That number actually may be on the rise. The mean age of onset is Junior High School age or so, 11 years in females, 12 ½ years in males. The incidence by gender varies slightly with age, but overall it’s equal.
A word about the pathogenesis. Again there’s probably a three-part explanation to this. It’s not absolute gospel but it’s the working hypothesis; there is a genetic predisposition linked to the HLA locus. That alone does not give you diabetes, or in identical twin pairs when the first one got it the second one would get it, and that only happens 35% of the time. But there appears to be necessary, atop that genetic predisposition, an environmental trigger such as an infection and Coxsackie viral infection has been purported as one of those. It may well be, and it may not be the only one. That combination then appears to trigger an autoimmune reaction which ultimately results in the destruction of the islets known histologically as insulitis. This is a T-cell mediated phenomenon. Although there are measurable, in the serum in most patients at the time of diagnosis - and even before, if you know to look - antibodies which are reflections of the autoimmune process but not the cause of the disease. These antibodies against islet cells, known as ICA, against the insulin molecule itself and against a islet component known as glutamic acid decarboxylase, or GAD. This triad ultimately then leads to beta cell rejection, if you will - because that’s really what’s happening - this as I mentioned is an insidious process for which there is evidence that it can take as much as ten years prior to presentation. And it isn’t until the beta cell mass drops from its original 100% to less than 5-10% of that, that symptoms of diabetes become manifest.
The clinical presentation; polyuria is present in about 75%, weight loss 35%, actual DKA 25% and actual coma 5%. What does this say? This says that patients are not nearly as sick as they used to be when I first started out in this business. When they present it’s found earlier. Either the disease is milder or doctors, parents or teachers are more astute, pick it up earlier, it’s not clear. Not surprisingly there is a correlate to this. The laboratory presentation is also milder, of course hypoglycemia and/or glycosuria will be present in 100%, but ketonuria only 85%, low bicarb under 18 in 60% and a low pH in only 20%.
Now there are other causes of hyperglycemia, however, in childhood besides type I diabetes. We actually see type II diabetes in children now. It is responsible for as many as 16% of new onset patients in childhood with diabetes. It tends to be autosomal dominant in multigenerations, particularly in females. It begins more so in adolescence, if it’s going to affect a child, than younger but we’ve now seen it in children under age 10 and it tends to predominate in high-risk minority populations; including African-Americans, Hispanics and native Americans. Now there’s an unofficial form of hyperglycemia known as type 1.5 diabetes, which seems to be unique to African-Americans. It starts out every which way like type I diabetes with slower DKA, but as time goes on the patients become insulin-independent, or non-insulin dependent. Unlike classic type I there is no HLA association and no evidence of autoimmunity. Lastly, there is a mild form of type II diabetes that is very rare that occurs in families, in children, that’s called MODY, or maturity-onset diabetes of the young. This is not the same as this. It’s totally different.
Now there are yet other causes of hyperglycemia that aren’t necessarily even diabetes at all that may be transient, or may not be, depending on the cause. Now certain drugs are associated with hyperglycemia. I think the most famous one of which would be glucocorticoids used at high doses, because it causes insulin resistance. There’s a list here. Most of these typically don’t cause hyperglycemia often but the list is made complete. I would point out to you the use of pentamidine, particularly in children with HIV infection, can cause either hypo or hyperglycemia and that’s probably the next most important one to be aware of . Now children with hypernatremic dehydration - which we don’t see all that often - manifest hyperglycemia 50% of the time. And the saline can be 152 and the glucose can be 900 and nobody knows why that is. You don’t usually have to treat the glucose, you just have to treat the dehydration and the hypernatremia.
As a prototypic infectious agent associated with hyperglycemia/diabetes, I include here congenital Rubella. This tends to cause a permanent form of hyperglycemia and resembling type I diabetes, but somewhat different. Now there are various genetic disorders and genetic neuromuscular disorders also associated with hyperglycemia that, something that we’ve talked about, and the list is here. I will just tell you because you might ask, this syndrome, while rare, is very interesting and this is an acronym for diabetes insipidus, diabetes mellitus, optic atrophy and sensorineural deafness, or DIDMOA syndrome, also known as Wolfram syndrome. Let me tell you this; when you have diabetes mellitus it’s very hard to diagnose diabetes insipidus on top of that, but I have one such patient.
Now, when the patient actually gets to the acidosis, which again is much less common than in the past, one of the key metabolic events responsible for that - well, everybody knows about insulin deficiency and without insulin of course there will a failure of peripheral tissues to utilize glucose normally and that will result in hyperglycemia, there will also be a breakdown of tissue fat leading to overproduction of fatty acids and ketoacids. Now what’s probably less well appreciated is that when there is insulin deficiency there will also be overproduction of glucagon. Now glucagon comes from the pancreatic alpha cells which have a connection to the beta cells, and when the signaling is abnormal - that is to say that the beta cells stop making insulin - the alpha cells make too much glucagon. What glucagon excess does is it stimulates the liver to make more glucose. That’s what glucose neogenesis means. It means the formation of glucose by the liver from non-glucose substrates such as protein that gets shuttled through there. And glucagon accentuates this process and who needs more glucose from the liver when your tissues in the periphery can’t take it up from insulin deficiency. Glucagon also stimulates glycogenolysis, the breakdown of glycogen, further leading to high glucose. This is a hepatic process and we take advantage of this physiology when we give diabetic patients with hypoglycemia glucagon therapy. To break down their existing glycogen to fix their hypoglycemia. And then glucagon also, through this complicated pathway - which I’m not going to belabor greatly - also contributes to the overproduction of ketoacids. Glucagon excess is probably more important to this whole process than is insulin deficiency, in actuality.
Now for the child with ketoacidosis, like any other presenting patient, you want to start with a history and physical and here are some key points to be aware of; of course, pay attention to the neurological status of the patient, which could be obtunded at the beginning or could become obtunded with treatment, which is a very scary phenomenon. Immediately assess signs of potassium deficiency through simple physical examination findings, such as reflexes, presence of bowel sounds. Obtain a weight, height and body surfaces because your calculation of fluids is going to be based on that, and if you need a formula for calculating body surface area is shown here. Just take the height in centimeters, multiply it by the weight in kilos, divide by 3600 and press the square root function on your calculator. It works at all ages and does not require a ruler. Now certainly assess basically the ABC’s here of impending demise; that is, look for shock, blood pressure, pulse, assess hydration and remember, because of the hyperglycemia driving urine output there can be severe dehydration with ongoing urine _ in DKA, unlike most other situations. And don’t be fooled. Once in awhile the presentation is not really true diabetes. On rare occasions. Salicylism can present with hyperglycemia and acidosis, although it’s more likely to cause hypoglycemia if it affects glucose metabolism and occasionally you can get severe hyperglycemia from trauma and infection, and associated stresses of those entities.
Your immediate approach to the patient from a laboratory perspective is summarized on this slide. You would like to measure a glucose level starting immediately with a bedside measurement by some strip or meter methodology. Ketones either in the blood or in the urine, electrolytes, BUN, and creatinine, and be aware that many of the autoanalyzers that measure creatinine will give you a false elevation of creatinine if there is present elevated concentrations of ketones concomitantly so that for a renal function assessment, the BUN is a better test in this situation. Knowing the magnesium, calcium and phosphorous are also important. You can measure or calculate the osmolality according to this formula. Most people will obtain a urinalysis because it just keeps coming and it’s just nice because you can get a ketone measure most quickly that way. If the patient is sick - and only if the patient is sick - a blood gas would be indicated, and you do not ever need basically to do an arterial blood gas in a child with DKA because you are not looking for problems with exogenation. You can do a venous blood gas because you want to know the pH and there’s not much difference between the arterial venous gradient so far as pH is concerned. Children will tell you - or I will tell you - that they much appreciate that while you are sticking their vein, to do this test rather than then having to blindly poke their artery. It’s very important to obtain a ribbon strip as a means of assessing potassium status.
Now, with all that said, in the last nine years 90% of our newly diagnosed patients with type I diabetes above the age of two have not required most all of that. But rather they come in and go home the same day with no IV and just being educated and taught survival skills related to diabetes. So again, they are not that sick and just because a patient is newly diagnosed with diabetes does not mean any IV’s, blood gases or even hospitalization. If only I could convince my emergency room colleagues of that point.
If the patient, however, is sick and does require all nine yards, so to speak, I will outline to you my approach to that. The first thing you are likely to do to such a patient is give them intravenous fluids, and you determine how much based on the percent dehydration, and always be aware that unlike other forms of dehydration there will be significant ongoing losses, in most cases, again, because of the glucose drive. If shock is present, personally I would get help. But you do things like boluses of normal saline, Foley catheters, NG tubes, etc. Now the usual fluid approach that I employ in this scenario would be to calculate a figure that takes into account maintenance fluids replacement and some ongoing losses, which almost always works out - if you do it this way - to the use of half normal saline at 3 liters per meter squared per day. That takes into account all of this. That’s roughly the same as maintenance plus 10% dehydration, if you do it with the other formula. That can be either given as half over the first eight hours, and then a quarter, a quarter over the next 16. The whole thing over one day. You can divide the whole thing into six shifts over 48 hours. People do this differently and if you are careful it works pretty much no matter what you do. Now be aware that those fluids alone, before you even give insulin, may lower the glucose concentration 100 - 200 mg/dl in an hour just from dilution. Also remember, unlike other forms of dehydration, you pretty much need to add potassium immediately and usually the patient will have peed anyway so you are not going to feel nervous about that, but they often need very high amounts of potassium replacements, somewhere in the range of 60-80 mEq per meter squared per day, which is 2-3 times maintenance and that’s because as you give insulin you drive potassium into the cells. You also need to remember to add D5 if this treatment regimen causes the glucose concentration to drop below 250, and even D10 when less than 150. You’d rather do that than lower the insulin infusion rate because you need the insulin ongoing to restrain ongoing glycolysis and further ketoacid production.
Now what about insulin? If the patient is sick enough to come into the hospital they usually need IV and we usually use either Humalog, which is the ultra short-acting insulin analog, also known as lispro, or human regular insulin and we give that by continuous infusion, and this is sort of a standard cookbook; 1/10 of a unit per kilo per hour, which is a rate studied and proven to lower plasma glucose concentration by about 100 mg/dl per hour and you don’t want to lower it any faster than that because of the osmolality situation. Now in the old days we used to precede this with a bolus of 1/10 of a unit per kilo but this turns out not to make any difference, so we tend not to give it. When you give insulin at .1 unit per kilo per hour, that’s the equivalent of 2.4 units per kilo per day, if you do it over 24 hours, which is a rather hefty dose of insulin; but you have to remember the patients in DKA not only have insulin deficiency, which is the primary problem, but they have superimposed on that, insulin resistance. This of course is secondary to a combination of secondary derangements, including hyper-insulinemia itself, hyper-osmolality, counter regulatory hormone excess, elevated plasma free fatty acids, acidosis, low phosphate and low magnesium.
Now more on insulin. We typically mix the insulin solution. Something like half a unit per milliliter, or 250 units in a 500 mg bag of normal saline and then you piggy-back it into the system. But you must remember that before piggy-backing it that you have to flush the catheter tubing, or the nurse should, with 50 ml of insulin saline solution to saturate the insulin binding sites in the plastic tubing. If you don’t do that, based on the slow infusion rates that the pumps will deliver of this solution, the patient won’t get insulin for seven hours; until you fill up all the dead space in the tubing. You’ll wonder why the patient is not getting better. As I mentioned previously, as the glucose falls you have to add dextrose to the IV fluids rather than lowering the insulin, so as to continue to prevent further lipolysis. The advantages of this IV insulin regime, compared say to intramuscular or subcu, is that there is less hypokalemia, less hypoglycemia. You can control the situation just by a dial, basically. Then once the patient has improved enough so that you are ready to start subcu and continue the IV insulin drip for about 30 minutes after the first subcu dose of regular, or five minutes after the first subcu dose of Humalog so as not to have any periods of lack of insulin administration.
What about bicarbonate? Well, nothing has changed, is the answer. But some people give it, some people don’t. It’s criteria for administration varies quite a bit and includes severe acidosis, typically pH less than 7.1, and in a patient who is huffing and puffing and out of it, most people would still give it. It can be helpful if it there is associated hyperkalemia, as it drives the potassium into the cells. Some people will also use it for severe coma, although that is usually in conjunction with low pH. The optimal rate of replacement has not been studied definitively and suggestions include 3 mEq/kg over 30-60 minutes versus over 12 hours. Again, probably they both work. You just have to… the patient has to be monitored carefully. The risks of giving bicarbonate are important to know. They include hyper-osmolality. This is a very osmolar formulation given to a patient who is already hyper-osmolar. There’s the entity known as paradoxical CSF acidosis where the patient given bicarbonate who is neurologically sound suddenly after a few hours becomes obtundent because of the CO2 that is derived from the bicarbonate, diffuses into the CSF and causes sensorial reduction. You can over-shoot and lower the potassium too much, which is bad. And you can alter the hemoglobin affinity for oxygen, which is also not good. And there is a suggestion, but no proof, that bicarbonate use increases the risk of cerebral edema.
What about phosphate? Should you give phosphate to patients with DKA? Well, it makes sense to avoid hypophosphatemia, which frequently if not present at admission will occur, again because insulin shifts the phosphate into the cells. Hypophosphatemia is not a good thing. When the phosphate level, for example, goes below 1 -which we don’t let happen too often - rhabdomyolysis can occur, dysfunction basically of all the hemochoatic system with hemolysis. You can get myocardial dysfunction, CNS dysfunction, nothing good. There’s also the theoretical concern that if the serum phosphate level falls that that will cause a reduction in something called 23DPG, or 23 diphosphoglycerate, which if reduced there is also a reciprocal tightening of affinity of oxygen to the hemoglobin molecule in someone who is already hypovolemic you don’t want to iatrogenically reduce oxygen delivery to the tissues.
As I mentioned, hypophosphatemia usually occurs after starting insulin, due to shifts into cells. If you choose to give it, the way you do is you take the potassium replacement and give 2/3 of KCL and 1/3 of K phos or 50/50. It turns out, interestingly, that despite all the reasons for doing it that the outcome of the patient is not in any way influenced by this. So I don’t believe it’s mandatory. And if you just give it for the sake of just giving it, there have been patients recorded who have gotten reciprocal hypocalcemia resulting in tetany, so it isn’t necessarily benign.
Now this is the one slide of this type that I’ll show you just to remind you a little bit of the physiology. This is the oxygen/hemoglobin association curve. On the X axis is PO2 and on the Y axis is oxygen saturation. Now if you look at the thick black line, that’s basically where I want you to start and I want you to start with a PO2 here of 40 mmHg. You’ll notice in this scenario that if someone has acidosis, shown here as increased hydrogen ion, that will shift the oxygen/hemoglobin association curve to the right and you’ll have, instead of 25% oxygen delivery - which would normally occur at a PO2 of 40 - you’ll have 40% oxygen delivery. So acidosis helps … with acidosis your body is smart and it releases oxygen and hemoglobin better than normally, which would be beneficial. If you happen to have low 23 diphosphoglycerate that shifts the curve to the left. So as I mentioned earlier, you hold onto oxygen more tightly in this scenario. But because in a untreated patient with DK you probably have this and this, opposing influences, the curve is probably normal. And it isn’t until you disrupt metabolism with your treatment that you could adversely shift the curve to the left at a time when you wouldn’t want to. So basically, do as little meddling as possible.
These are two CT scans, two separate patients with DKA shown at two time points on the left, two studies, are two different patients within about six hours of admission in DKA, and then on the right are the same two patients over time, just prior to discharge. What this shows you in the two left hand CT cuts is the presence of cerebral edema. These were not unusual patients because this was found in six consecutive patients, and the way you can assess that is after treatment the ventricles get slightly larger, and that means there is less edema constricting it. The point of this slide is that probably all patients that come in with DKA have at least a small amount of cerebral edema, and what you don’t want to happen is make one of those patients have a large amount of cerebral edema. Now, the patients that get a large amount are fortunately few and far between, but they all typically will have a terrible outcome, either death or severe neurological handicap. Nobody knows what it is specifically that shifts the patient with a mild cerebral edema - which is practically all of them - to the rare one with major cerebral edema. But it’s been linked to excessive fluid administration, greater than 4 liters per meter squared per day, which is why I have that number 3 liters per meter squared per day mentioned previously. It’s also been suggested, but by no means proven, that it tends to occur that if, during treatment, there is a failure of the serum sodium corrected for prevailing hyperglycemia to rise as the plasma glucose concentration falls with treatment. So that would suggest that there is too dilute fluid being administered so the sodium level doesn’t go up. It stays where it is or drops. And that has been suggested by some as a marker for impending doom, but I don’t think it holds up in all patients. So nothing really does and nobody really knows why that happens.
At any event, let me shift within the subject of diabetes to how we take care of the average patient with diabetes. We base our philosophy on tight glycemic control. The importance of which has been substantiated through the 1994 landmark study known as the Diabetes Control and Complications Trial, in which they compared - mostly in adults, but some adolescents - two groups of individuals followed for a number of years, one of whom was intensively treated and another was conventionally treated, and they found in the patients who were intensively treated that there was a marked reduction in retinopathy by 76%, in nephropathy by 54%, and in neuropathy by 60%. So this underscored kind of what we knew, that keeping the blood sugar under control on a long-term basis is in ones best interest. There was also some improvement in macrovascular outcomes related to coronary, cerebral, and peripheral vascular disease and cholesterol dynamics as well.
Now with the good news there is going to be a little bit of bad, and in the same study, in the patients who were more intensively controlled, there was a two to threefold increase in severe hypoglycemic episodes, defined as episodes requiring the assistance of another, which probably isn’t surprising because you are aiming for normalcy more often so you are more likely to overshoot. Also there was, in the intensively treated group, they tended to gain 3 kilos more of weight than the conventional group, which I could see in adolescence might be a little bit of a deterrent to tight control. Now we have to be careful in that what was known from this study is not necessarily applicable to young children with type I diabetes, or what we think it is. They were not part of the study. The type of care that the intensively treated people got on this nationally funded research study was rather laborious and expensive and probably can’t be exactly applied to all patients, but importantly the regimen you recommend for any given patient needs to be tailored to that patient.
What is conventional insulin treatment as we use it in the clinic setting? Well, that typically can be defined as a split dose, or two injections, containing mixtures in each dose of short and intermediate acting insulin, such as regular NPH or Humalog and NPH. An intensive regimen would be a three-shot regimen which would include a pre-breakfast dose, similar to this, with a mixture of short and intermediate acting insulin and typically at dinner, only short acting insulin, and at bedtime only intermediate acting insulin; perhaps with a small amount of short acting insulin. The reason for this splitting of the second shot is because a lot of people like to eat dinner at 5 o’clock in the afternoon and have breakfast at 8, and a lot of people like, the next morning, to eat breakfast at 8, 9, 10 or 11 o’clock in the morning. So by delaying the NPH in the second shot you have better control with those kinds of schedule quirks. Some people take four injections a day. There’s a lot of different approaches, and there’s a growing use of insulin by pump administration in the United States.
Now how do you utilize insulin properly? Well, a couple of clues. We use Humalog as our short acting insulin most of the time and this is a synthetic analog of regular insulin that’s been around since ’96, and is absorbed faster than regular and therefore people can eat right after they take their insulin, typically, which is what they used to do anyway when they took regular and they’d have bad control. The modifications of this molecule enable it to be absorbed faster, basically. Patients whose blood sugar before meals and doses would normally be told to wait 20-30 minutes from when they took the regular insulin before eating, and now they can take Humalog and eat immediately. If ones blood sugar before a meal and insulin dose is high, you really should wait longer before you eat, from the time you take your shot although not as long with Humalog as one would with regular. You also want to space, if you are on two shots a day, your insulin doses about twelve hours apart so that the NPH is covered the whole day and don’t overlap or underlap. But you can be sure that people don’t always do what they are told. Then patients need to do intensive glucose monitoring and based on the results, make changes that would either be corrective, preventive or anticipatory in their insulin dose as often as necessary on a daily basis.
Glucose monitoring that we recommend; the conventional way is to do like three tests a day, before the two injections, and at bedtime. The more intensive program would involved this pattern plus testing before lunch and even during the night on occasion. Glycosylated hemoglobin we tend to get every three months. Some people do it a little more often. I don’t think it is particularly valuable that way. Just to remind you that glycosylated is a name that is inclusive of various moieties including hemoglobin A1C, which is the one probably most famous, which is specifically … the glucose moiety is down to the N-terminal valent of the beta chain of the hemoglobin molecule. That’s A1C. And what the glycosylated hemoglobin A1C does for you is it gives you an integrated measurement of the degree of hyperglycemia during the antecedent 8-12 weeks prior to the phlebotomy, providing an objective index of mean glucose over an extended period of time. The test is predicated on the fact that there is a non-enzymatic, slow, continuous and irreversible reaction between glucose and hemoglobin within the red blood cell, the half-life of which is 120 days, which is why it gives you a measure of control over this period of time. But you have to remember, it’s not a perfect test. And because it is also influenced by hypoglycemia, patients whose glycosylated hemoglobin is also lower than expected, must presumably be having asymptomatic periods of hypoglycemia offsetting some of their hyperglycemia, which is scary.
Now you can’t do any of this without diet, obviously, and I’d say of all the things with diabetes, diet is the hardest. We recommend that the carbohydrate intake actually be liberalized to about 55-60% of dietary intake, but of course quality is important. We also recommend decreasing intake of saturated fats, which are not good for you. Protein intake should be around .08 gram per kilogram of body weight, or 12-15% of the total caloric intake, and if somebody has chronic renal disease you need to lower the protein intake. Fat should constitute no more than 30% of the total daily calories, of which less than 10% should be saturated, and the cholesterol intake should be restricted to less than 300 mg per day. We encourage the judicious use of free foods and reasonable amounts of artificial sweeteners such as Nutrasweet, to help the patients.
What about exercise? Well, we like exercise, but to avoid exercised-induced hypoglycemia it’s better to use injection sites away from those muscles which will be most active in the exercise, since those muscles will have increased vascular flow and that enhances insulin absorption and action and has the propensity toward hypoglycemia. Having the injection in the abdomen rather than in the thigh. Low insulin may be used before exercise, it’s not routine, or you might need to add a snack in front of that exercise if it were not pre-planned. Now be aware though that exercise has a paradoxical effect on the glucose. Those patients who are chronically under-insulinized with poor glycemic control and ketosis, if they exercise, may get hyperglycemia and enhancement of ketone bodies and perhaps even DKA. This occurs because exercise-mediated muscular uptake of glucose is insulin dependent and if you have relative insulin deficiency those muscles cannot take up the glucose, and that’s why the patients get into trouble.
There are some special issues in pediatric diabetology. Let me remind you that we have a particular management issue with pre-breakfast hyperglycemia, which typically has three possible causes. One is called the dawn phenomenon that refers to the time of day and it thought to represent a period of early morning insulin resistance, perhaps secondary to the delayed effects of growth hormone secretion from earlier in the sleep period. One may also see just a waning effect of the previous dose of the intermediate acting insulin the night before. There just wasn’t enough insulin. Or, what everybody seems to know, as the Somogyi , which is also known as post-hypoglycemic hyperglycemia or rebound. Patients all know this, doctors all know this, and it turns out that it rarely ever happens. Dr. Somogyi was a Hungarian biochemist so I don’t know if he ever saw any patients, but in any event it turns out that in patients who do have the Somogyi phenomenon the blood sugar rarely ever exceeds 180 in the morning and in point of fact, you can have the Somogyi phenomenon when your blood sugar is 100 in the morning. Meaning it could have been 40 during the night, which is a little scary to me.
Shown here are those three examples. This is the overnight period. Here the blood sugar is at the Y axis, shown in green is the Somogyi phenomenon bounce and goes low and then goes high. Shown here in the red is the dawn phenomenon. The blood sugars are good all night and then in the dawn hours it goes up, and then in the blue is; just not enough insulin all night long so you go to sleep at a certain number and the blood sugar just goes steadily up thereafter.
What about hypoglycemia? A couple of words. Severe hypoglycemia can occur in patient with type I diabetes because they lose their ability to counter-regulate over time. It’s a secondary defect in the disease, so they have a loss of glucagon and epinephrine responsiveness to hypoglycemia, which therefore means that they don’t recognize hypoglycemia until they start getting sensorial changes, when it can be too late. This is typically fait accompli by about five years after diagnosis. I mentioned in the DCCT study that there is a heightened risk of hypoglycemia, severe hypoglycemia, with tight control so you have to be careful about using tight control in young children or irresponsible patients.
Finally, just another word, remember that adolescents in particular with diabetes are prone to eating disorders probably more often than are adolescents in general. And I mean here specifically bulimia and they have a unique way in which they can be bulimic. Instead of eating and sticking their fingers down their throat or taking laxatives, they just have to withhold their insulin selectively and they can lose their calories right out in their urine. Let me tell you, patients figure this out.
What about the future? Well, we are looking at new ways of administering insulin; through the nose, through the mouth, through pulmonary routes like asthma medicines. We are looking at new ways of monitoring glucose that does not involve finger pokes. There is a new little machine on the market that does this through a little catheter that’s placed under the skin and connected to a readout machine that can give almost continuous readouts, so we are making some progress there. We are looking at drugs maybe to treat patients who already have complications and make their lives a little easier, and ultimately we are looking for cures for this disease through transplantation, either whole or segmental pancreas or islet cells or encapsulated islets, things like that. Also using molecular biology and gene technology to figure out how to teach cells that have the genetic machinery to make insulin but don’t normally make insulin, to go ahead and make insulin and maybe be transplanted into the patient. This might prove to be a very interesting way to treat this disease.