OVERVIEW: What every practitioner needs to know

Are you sure your patient has type 1 diabetes mellitus? What are the typical findings for this disease?

Diabetes mellitus is a disorder of the metabolic homeostasis controlled by insulin, resulting in abnormalities of carbohydrate and lipid metabolism. Type 1 diabetes is caused by the autoimmune destruction of the insulin-producing beta cells of the pancreas. This results in an absolute insulin deficiency. Although there has been a marked rise in the percentage of cases of diabetes mellitus in children and adolescents caused by type 2 diabetes mellitus in the past one to two decades, type 1 diabetes remains the most common etiology of diabetes mellitus in prepubertal children.

The most common symptoms are polydipsia and polyuria. Nocturia, re-emergence of bed-wetting, or a frequent need to leave class in school to use the bathroom are typical complaints that suggest polyuria.

Other common symptoms at presentation of type 1 diabetes mellitus are polyphagia and weight loss.

Continue Reading

The above symptoms are typically present for less than a month; less commonly, the symptoms may be present for several months.

Patients with diabetes mellitus will have an elevated plasma glucose level; a level greater than 200 mg/dl (11.1 mmol/L) in a child with symptoms of diabetes is diagnostic of diabetes mellitus (see below). Glucosuria on urinalysis will often be seen in patients with diabetes mellitus, but an absence of glucosuria does not exclude the diagnosis of diabetes mellitus, and glucosuria can be caused by other conditions (see below). Urine ketones on urinalysis in a child with suspected diabetes mellitus raises the possibility that the child may be developing diabetic ketoacidosis.

What are the findings of a child with type 1 diabetes mellitus presenting in diabetic ketoacidosis?

Children with diabetic ketoacidosis present with nausea, vomiting, and lethargy, as well as with signs of dehydration. Some patients will develop abdominal pain that is severe enough to suggest appendicitis or other intraabdominal process. There may be a fruity odor to the breath from expired acetone.

With severe acidosis, the patient will develop the deep labored breathing of Kussmaul respirations.

As the DKA worsens, the lethargy can progress to severe obtundation and coma.

On questioning, these children will usually also have a history of the classic symptoms of diabetes (polydipsia, polyuria, polyphagia, and weight loss). Diabetic ketoacidosis should be suspected in all children presenting with vomiting and lethargy, but particularly in the absence of fever and diarrhea.

What other disease/condition shares some of these symptoms?

  • Type 2 diabetes mellitus

  • Maturity-onset diabetes of youth

  • Neonatal diabetes

  • Diabetes insipidus

  • Viral gastroenteritis (for a child presenting in diabetic ketoacidosis)

What caused this disease to develop at this time?

Symptoms of diabetes occur when the capacity to produce insulin is insufficient to maintain metabolic homeostasis. It is felt that this point is not reached until greater than 80% of the beta-cells are destroyed, a process that probably takes months to years.

When the serum glucose level rises above approximately 180 mg/dl (10 mmol/L), the kidneys are unable to resorb all the glucose being presented to them. This results in glucosuria, which stimulates an osmotic diuresis and the symptom of polyuria. Polydipsia is then triggered to maintain euvolemia.

Insulin deficiency also results in increased lipolysis and protein breakdown. This, along with the calories lost with glucosuria, results in weight loss in spite of polyphagia.

With profound insulin deficiency there is decompensation into ketoacidosis. The severe insulin deficiency allows for excessive lipolysis and the generation of ketoacids in the liver from the liberated fatty acids. As these ketoacids accumulate they cause an ileus. This ileus results in nausea, and later, vomiting, both of which impair fluid intake.

The impaired fluid intake in the presence of the glucosuric osmotic diuresis produces dehydration. The dehydration impairs the ability to excrete glucose and acids, worsening the hyperglycemia and acidemia. The worsening hyperglycemia increases the osmotic diuresis, setting up a cycle of worsening dehydration and worsening hyperglycemia.

The dehydration also stimulates the release of stress hormones including cortisol and catecholamines, both of which further stimulate lipolysis and hepatic glucose release, further worsening the metabolic derangements.

Medical stress, including simple viral illnesses, results in insulin resistance. Because of this, a child may present with diabetes mellitus during a viral illness or other medical stress. In this situation, the illness resulted in an earlier presentation of the diabetes mellitus, because the increased insulin needs imposed by the insulin resistance exceeded the capacity for insulin production sooner than if the insulin needs remained stable.

What laboratory studies should you request to help confirm the diagnosis? How should you interpret the results?

  • A plasma glucose level of greater than 200 mg/dl (11.1 mmol/L), in the presence of the classic symptoms of diabetes mellitus, confirms the diagnosis of diabetes mellitus.

  • A hemoglobin A1c (HbA1c) greater than 6.5% is also presumptive confirmation of a diagnosis of diabetes mellitus.

  • Confirmation of the diagnosis of type 1 diabetes mellitus is generally straightforward, with glucose and hemoglobin A1c levels unambiguously elevated above the diagnostic thresholds.

  • Diabetes mellitus can also be diagnosed based on a fasting plasma glucose level greater than or equal to 126 mg/dl (7 mmol/L), or a plasma glucose level greater than or equal to 200 mg/dl (11.1 mmol/L) obtained 2 hours after ingestion of an oral glucose load (the oral glucose tolerance test, OGTT). If this testing is performed in the absence of the classic symptoms of diabetes mellitus, the diagnosis requires confirmation with repeat testing on a subsequent day. Because type 1 diabetes mellitus does not have a prolonged asymptomatic period, it is rarely appropriate to do such testing in an asymptomatic child. In cases where fasting plasma glucose testing or the OGTT is performed to investigate a possible diagnosis of diabetes mellitus, a diagnosis other than type 1 diabetes mellitus (such as MODY or type 2 diabetes mellitus) may be more likely.

  • DKA is present if the child has hyperglycemia (glucose greater than 300 mg/dl (16.7 mmol/L)), along with acidosis (venous pH 7.3 or less; serum bicarbonate 15 mEq/L or less), and significant ketonuria or ketonemia.

How do you differentiate stress-induced hyperglycemia from type 1 diabetes mellitus?

Occasionally, a child will be found to have hyperglycemia during an acute illness. While this could represent the situation of a viral illness accelerating the presentation of type 1 diabetes mellitus (see above), it may also represent stress-induced hyperglycemia. Stress-induced hyperglycemia is most often relatively mild, but can result in plasma glucose levels above 300 mg/dl (16.7 mmol/L). In addition, transient hyperglycemia is common in children treated with beta-agonists and glucocorticoids for asthma.

A lack of a history of the classic symptoms of diabetes mellitus preceding the acute illness makes it more likely the hyperglycemia is due to stress hyperglycemia rather than diabetes mellitus.

In most cases, a normal hemoglobin A1c is sufficient to demonstrate the acute nature of the hyperglycemia and to exclude a diagnosis of type 1 diabetes mellitus.

Positive diabetes autoantibodies, especially the presence of multiple diabetes autoantibodies, suggests an increased risk of the patient developing type 1 diabetes mellitus in the future, although progression to diabetes mellitus from stress-induced hyperglycemia is very rare.

What other diseases/conditions present with glucosuria?

In type 1 diabetes mellitus, glucosuria will occur when the plasma glucose level exceeds approximately 180 mg/dl (10 mmol/L). Thus, in a child with the classic symptoms of diabetes mellitus, glucosuria suggests a diagnosis of diabetes mellitus, although this should be confirmed with plasma glucose testing.

Renal tubular dysfunction can lead to glucosuria in the absence of hyperglycemia. This will occur in Fanconi syndrome, as an isolated congenital disorder, or due to an acquired tubular dysfunction.

The lack of hyperglycemia and/or a normal hemoglobin A1c excludes the diagnosis of diabetes mellitus when glucosuria is noted.

Would imaging studies be helpful? If so, which ones?

There are no imaging studies that help in diagnosing diabetes mellitus.

Confirming the diagnosis

  • The diagnosis of diabetes mellitus is made based on documenting hyperglycemia through measurement of the plasma glucose level or by measurement of the HbA1c (see above).

  • Once a diagnosis of diabetes mellitus is made, one should consider whether it is a case of type 1 diabetes mellitus, or another type of diabetes mellitus.

  • In some cases, the definitive determination of the type of diabetes the patient has cannot be made at the time of presentation. In such cases, the type of diabetes mellitus present is determined over time by the clinical course, including follow-up testing (such as measurement of the c-peptide level two or more years after the diagnosis of diabetes mellitus, to quantify the amount of ongoing endogenous insulin production).

  • A diagnosis of something other than type 1 diabetes mellitus is very unlikely in a child with all of the following at presentation:

    a prepubertal child that is older than 1 year of age

    diabetes mellitus diagnosed based on a random plasma blood sugar of greater than 200 mg/dl (11.1 mmol/L) after presenting with the classic symptoms of diabetes mellitus over a period of days to a few months

    there is no acute illness that could be causing stress hyperglycemia

    the child is not being treated with medications that can cause hyperglycemia (such as glucocorticoid)

    positive diabetes autoantibodies support the diagnosis of type 1 diabetes mellitus, but are probably unnecessary in such patients.

When should a diagnosis of neonatal diabetes/monogenic diabetes be considered?

Neonatal diabetes/monogenic diabetes mellitus should be considered in any infant less than 12 months of age who is diagnosed with diabetes mellitus.

  • Monogenic diabetes is much less likely in infants presenting after 6 months of age, but should be considered up through 12 months of age.

When should a diagnosis of MODY (maturity onset diabetes of youth) be considered?

MODY should be considered in a child diagnosed with diabetes who has a family history suggesting an autosomal dominant inheritance pattern of early-onset diabetes mellitus.

Type 2 diabetes mellitus has a strong inheritance pattern that can suggest autosomal dominant inheritance. However, in families with type 2 diabetes, the individuals with diabetes will generally be obese.

MODY should be considered in a lean patient with evidence of significant endogenous insulin secretion more than 2-3 years after diagnosis (i.e., a detectable C-peptide level [>200nmol/l] with the glucose level > 144 mg/dl (8 mmol/L).)

MODY2 should be considered in a patient identified with mild hyperglycemia (i.e., fasting glucose level 100-150 mg/dl (5.6-8.3 mmol/L)) in the absence of the classic symptoms of diabetes mellitus or of risk factors for type 2 diabetes mellitus.

When should a diagnosis of type 2 diabetes mellitus be considered?

In most pubertal or postpubertal children diagnosed with diabetes mellitus, it is appropriate to consider the diagnosis of type 2 diabetes as well as type 1 diabetes. Type 2 diabetes is particularly likely in an obese child of minority racial/ethnic background. In contrast, a Caucasian child diagnosed with diabetes who is not overweight or obese is unlikely to have type 2 diabetes.

A presentation with ketoacidosis does not exclude a diagnosis of type 2 diabetes mellitus. However, insulin is needed for the treatment of diabetic ketoacidosis regardless of the underlying type of diabetes. In addition, children with type 2 diabetes presenting with ketoacidosis will likely require ongoing treatment with insulin, at least for some period of time.

If diabetes mellitus is diagnosed based on testing done in a child because of risk factors for type 2 diabetes, the presumptive diagnosis is type 2 diabetes mellitus.

How is type 1 diabetes mellitus differentiated from type 2 diabetes mellitus?

When it is appropriate to consider whether a child with diabetes mellitus has type 1 diabetes versus type 2 diabetes, diabetes autoantibodies should be measured.

The diabetes autoantibodies that can be measured are:

islet cell antibodies (ICA)

antibody to glutamic acid decarboxylase (GAD-65)

antibody to insulin

these antibodies may develop in any patient treated with insulin. Therefore, positive insulin antibodies are not useful in differentiating type 1 versus type 2 diabetes if they are measured more than 1-2 weeks after a patient is treated with insulin

antibody to insulinoma-associated protein 2 (IA-2; also known as ICA512)

antibody to zinc transporter 8 (ZnT8A)

One or more positive diabetes autoantibodies supports a diagnosis of type 1 diabetes mellitus

Each of the diabetes autoantibodies are positive in 50-90% of patients at diagnosis

Negative tests for three or more diabetes autoantibodies supports the diagnosis of type 2 diabetes mellitus

less than 5% of patients with type 1 diabetes will be negative for three or more tests for diabetes autoantibodies

Positive diabetes autoantibodies may be found in up to 36% of youth with what otherwise appears to be type 2 diabetes mellitus

It is unclear what the appropriate categorization of diabetes type in these patients is: whether they are obese, insulin resistant patients with type 1 diabetes, or whether they are type 2 diabetes patients with a low level of beta cell autoimmunity.

Most of these patients will require insulin treatment.

In patients with type 2 diabetes treated with oral agents, positive diabetes autoantibodies indicates an increased risk of a need for insulin treatment in the near future in order to maintain glycemic control.

In a patient with presumed type 2 diabetes who is initiated on treatment with insulin, the presence of autoantibodies indicates a need for caution and careful consideration before discontinuing insulin.

If you are able to confirm that the patient has type 1 diabetes mellitus, what treatment should be initiated?

All patients with type 1 diabetes mellitus are treated with subcutaneous insulin. Treatment for the child presenting in DKA is presented in a separate section. For children not presenting in DKA, treatment should also be started the same day as the diagnosis is made, in order to avoid the risk that the child develops DKA awaiting initiation of treatment. In the vast majority of patients, insulin treatment for type 1 diabetes will be a basal/bolus regimen. In this regimen, a long acting insulin is given to provide for basal (fasting) insulin needs, and a rapid-acting insulin is given to provide prandial insulin needs.

Use of an insulin pump to deliver continuous subcutaneous insulin infusion also provides a basal/bolus regimen. In this case, the basal insulin needs are provided by the basal rate of insulin delivered by the pump; the basal rate can be set to vary throughout the day.

Prandial doses for patients using an insulin pump are calculated the same as those for patients using subcutaneous injections. Pumps provide a number of advantages over injection therapy (including the need to insert the infusion catheter subcutaneously only once every 2-3 days rather than a minimum of 4 subcutaneous injections with injection therapy). However, logistically pump therapy is not feasible to be started at the time of diagnosis. In addition, insulin pumps occasionally fail, in which case injection therapy must be used until a working pump is obtained. Therefore, it is important that all patients and families with type 1 diabetes know and be comfortable using injection treatment.

To calculate insulin doses, the first step is to calculate the total daily insulin dose. In an established patient with type 1 diabetes, after there is complete loss of ß-cell function, total daily insulin needs are approximately 0.75-1 unit per kilogram of body weight (Puberty is associated with a physiologic insulin resistance, so that insulin requirements in adolescents are higher – typically 1-1.2 U/kg/day).

However, at the time of diagnosis, the patient with type 1 diabetes still has some remaining insulin production. Therefore, initial insulin requirements are lower. In addition, the initial insulin dose chosen is intentionally somewhat low to avoid the acute risk of hypoglycemia. As the patient’s response to these initial doses is determined, the insulin doses can be increased to achieve the desired glycemic control.

Starting insulin dose

For children less than 2 years of age, a typical starting dose of insulin is 0.5 U/kg/day

For older children the typical starting dose is 0.6-0.7 U/kg/day.

Basal insulin needs

Basal insulin needs are approximately 50% of the total daily insulin needs:

-Insulin glargine insulin detemir are used as the basal insulin as these insulins have no significant peaks in their action.

-Insulin glargine is typically given once a day, and can be given at any time of day (but then is given at the same time of day for a given patient).

-Insulin detemir has a shorter half life than insulin glargine, and is therefore generally given twice a day (at breakfast and at bedtime).

-Neither insulin glargine nor insulin detemir should be mixed with another insulin (i.e., they should not be mixed in the syringe with rapid-acting insulin in order to inject them together).

Bolus insulin dose

Rapid-acting insulin is used for the bolus insulin. The three rapid-acting insulins (aspart, glulisine, lispro) are equivalent in their pharmacokinetics.

The bolus insulin doses have two roles. First, they provide the insulin needed to meet the requirements of food intake. Second, the bolus dose provides additional insulin to correct for hyperglycemia.

While euglycemia is the goal of insulin therapy in diabetes, this is never perfectly achieved. Various factors result in significant variability in glucose levels, even in response to apparently identical situations of insulin doses, activity, and food intake.

To calculate a bolus dose the patient or caregiver calculates the insulin dose needed based on the amount of food to be eaten, measures the patient’s blood sugar level and calculates the dose of insulin needed to correct for hyperglycemia (if present), and adds these two amounts to give the total bolus dose (see example below).

The onset of action of the rapid-acting insulins is approximately 15 minutes, with a peak action occurring at 30-60 minutes, and a duration of 3-4 hours. To optimally match the pharmacodynamics of these insulins, the rapid-acting insulin is injected at the start of the meal. However, this requires an accurate prediction of the amount to be eaten during that meal. For younger children in particular, this may not be possible, and injection of insulin at the start of the meal would then have the risk of over-dosing the insulin and causing hypoglycemia. In such cases, the insulin is injected immediately after the meal is completed.

Insulin to carbohydrate ratio

The amount of insulin needed at a meal is predominantly affected by the carbohydrate content of the meal. The actual ratio of insulin needed per carbohydrate consumed will vary from patient to patient. Typically, infants and toddlers will need 1 unit of insulin for every 20 to 40 grams of carbohydrate consumed, school-age children may need 1 unit for every 10 to 20 grams of carbohydrate consumed, while adolescents may need 1 unit for every 5 to 7.5 grams of carbohydrate consumed. (These doses are typical for a child with established diabetes – the requirement at the time of diagnosis will be for less insulin due to ongoing endogenous insulin production.)

To calculate an initial insulin to carbohydrate ratio, an empiric formula can be used whereby the grams of carbohydrate covered by a unit of insulin is equal to 450 divided by the total daily insulin dose.

Correction dose

An initial correction dose can be calculated using an empiric formula whereby 1 unit of insulin will lower the blood glucose (in mg/dl) level by 1800 divided by the total daily insulin dose. This will give the amount the blood sugar will decrease (in mg/dl) in response to a unit of insulin (in the absence of food intake). (To convert glucose level in mg/dL to mmol/L, divide by 18; 180 mg/dL = 10 mmol/L.)

This correction dose is given when the preprandial blood sugar is above the target glucose level. For older preschool and school age children, correction doses may be given starting for blood sugars greater than 150 mg/dl. For infants and toddlers, correction doses may not be given unless the blood sugar is somewhat higher, perhaps greater than 180 mg/dl. This is due to a greater concern for hypoglycemia in these children, due in large part to the greater difficulty identifying hypoglycemia in these children. In addition, the threshold for adding the correction dose of insulin is dependent on the fall in blood sugar expected with the smallest correction dose that can be given. For example, if the patient is using an insulin syringe calibrated to give insulin in 0.5 unit increments, and has a correction dose of 1 unit per 200 mg/dl (0.5 units per 100 mg/dl), then the typical correction would start with giving 0.5 units only once the blood sugar level is above 200 mg/dl, as this extra 0.5 units would be expected to decrease the blood sugar level by 100 mg/dl (e.g., 201 mg/dl down to 101 mg/dl).

In children in whom the bolus dose is given after the meal because of an inability to accurate predict the amount to be eaten at the meal, the bolus dose is based on the carbohydrates eaten calculated after the meal is completed but the correction dose is still based on the blood sugar obtained just prior to the meal.

Example calculation of initial insulin dose:

A 10-year-old girl weighing 27 kg presents with polydipsia and polyuria of 2 weeks’ duration. She is not in DKA.

Total daily dose (TDD) = 0.6 x 27 kg = 16.2 units

Basal insulin = 50% of TDD = 0.5 x 16.2 = 8.1

Use 8 units of insulin glargine, given at bedtime

Insulin:carbohydrate ratio = 450 ÷ TDD = 450 ÷ 16.2 = 27.7

Use 1 unit per 25 grams of carbohydrate

Correction dose = 1800 ÷ TDD = 1800 ÷ 16.2 = 111

Use 1 unit per 100 mg/dl for blood sugars that are above 200 mg/dl


Use 0.5 units per 50 mg/dl for blood sugars that are above 150 mg/dl

This child’s usual breakfast contains 55 grams of carbohydrate. Her blood sugar one morning is 212 mg/dl.

Bolus dose:

55 grams of carbohydrate ÷ Insulin:carb ratio = 55÷25 = 2.2; give 2 units


212 mg/dl is two 50 mg/dl increments above 150 mg/dl; give 1.0 unit


3.0 units of rapid-acting insulin before breakfast this morning

Evaluating and adjusting insulin doses

The goal of the basal insulin (glargine or detemir) is to maintain glucose levels steady during fasting. A rising blood glucose level during a fasting period (i.e., overnight), indicates a need to increase the basal insulin dose. Note, however, that unless fasting blood glucose levels fall below normal, a falling blood glucose level during fasting in a patient recently diagnosed with diabetes may represent the effect of the remaining endogenous insulin production, rather than indicating too high a dose of basal insulin.

The goal of the insulin to carbohydrate ratio is to provide sufficient insulin so that the glucose level after a meal is the same as the level before the meal (“after” means after full absorption of the meal and after the majority of the action of the rapid-acting insulin; i.e., 3 hours after the insulin injection). Thus, if the blood sugar level consistently rises from within the target range before meals to above the target range after meals, more insulin should be given for carbohydrate ingestion (e.g., a change from 1 unit of insulin per 10 grams of carbs to 1 unit of insulin per 7 grams of carbs). The insulin to carb ratio is best assessed by evaluating the glucose levels after a meal when the glucose level before the meal was within the target range, since at these meals, no correction dose is given, isolating the effect of the insulin to carb ratio.

The role of the correction dose is to bring any glucose level that is above the target range down into the target range (Note that this is the only dose whose role is to bring the blood glucose level into the target range). The correction dose can be directly tested by treating a fasting blood sugar that is above the target range with insulin and measuring the fall that occurs over the subsequent 2-3 hours (while maintaining the fast). The correction factor is the fall in the blood glucose level divided by the units of insulin given.

Practically, however, the usual approach to evaluating the correction dose is to evaluate the glucose response to the rapid-acting bolus given with meals. This requires that the insulin to carbohydrate ratio first be determined to be correct. If glucose levels then remain above the target range when measured 3 hours after bolus doses given with meals when the blood sugar is above the target range, more insulin needs to be given as a correction dose (e.g., a change in the correction dose from 1 unit per 50 mg/dl to a dose of 1 unit per 40 mg/dl).

Because of day to day and dose to dose variations, insulin doses are usually not adjusted based on single blood glucose levels, but rather based on the pattern of glucose levels identified over a period of days or weeks. The exception might be at the initiation of insulin treatment at the time of diagnosis, when early adjustments may be appropriate based on fewer blood glucose measurements.

What is the honeymoon period?

At the time of diagnosis with type 1 diabetes, the autoimmune destruction of β-cells is incomplete, with approximately 10-20% of β-cell function remaining. Chronic hyperglycemia impairs β-cell function. Thus, the chronic, unrecognized hyperglycemia that exists prior to the diagnosis of diabetes being made further impairs β-cells function at the time of diagnosis beyond that of the immune destruction. This metabolic impairment of β-cells resolves as insulin treatment is initiated.

This temporary improvement can result in some degree of clinical remission. Many patients will have a decrease in their requirement for exogenous insulin, with a small minority able to maintain normal or near normal glucose levels without insulin treatment. This “honeymoon period” of improved β-cell function is temporary, as the remaining β-cells continue to be lost through the ongoing immune destruction. The duration of the honeymoon period is typically a few months, but it can vary from weeks to years, with the duration generally being inversely related to the age at presentation.

What is the target range for blood glucose levels?

Because subcutaneous delivery of insulin cannot match the pharmacodynamics of endogenous insulin secretion, patients with type 1 diabetes are not able to achieve completely normal blood sugar levels. The target for blood sugar levels is shown below. While the levels listed below are guidelines, the true target range is the lowest glucose levels that can be achieved without causing hypoglycemia (or with an acceptable frequency and severity of hypoglycemia).

Even patients with excellent glycemia control will not have all of their blood glucose measurements in the target range. Getting 75% of blood glucose levels into the target range would be excellent control.

Recommended target ranges: PUBMED:26696687

Before Meals:

90-130 mg/dl (5.0-7.2 mmol/L)


90-150 mg/dl (5.0-8.3 mmol/L)

Spilt-mixed insulin regimen

Prior to the advent of rapid-acting and basal insulin analogues, insulin regimens for type 1 diabetes comprised 2 to 3 injections a day utilizing short-acting regular insulin and intermediate-acting NPH (neutral protamine Hagedorn) insulin. These have a greater risk of hypoglycemia because the peak action of these insulin does not correspond to a time of peak insulin need (food intake).

This regimen requires that food intake match the insulin regimen, in contrast to the basal/bolus regimen where insulin doses match the food intake. Thus, using this regimen, meals should be taken at the same time and should have the same amount of carbohydrate each day. Split-mixed regimen using rapid-acting insulin in place of regular insulin decrease the risk of between-meal hypoglycemia associated with regular insulin use, but continue to have the risk of hypoglycemia from the non-physiologic peak of NPH.

In the split-mixed regimen, regular (or rapid-acting) insulin plus NPH is given at breakfast and dinner. Because NPH can be combined in a single syringe with either regular or rapid-acting insulins, this regimen requires only two injections a day. A variation of the split-mixed regimen is to divide the evening dose into an injection of regular (or rapid-acting) insulin at dinner and an injection of NPH at bedtime (this was often necessary because the duration of NPH action is not always sufficient to persist from dinner through to breakfast).

The dose of insulins used for a split-mixed regimen when initiating treatment in a newly diagnosed patient with diabetes again starts with estimating the total daily insulin dose as 0.5-0.7 U/kg/day (see above). 2/3 of this dose is given at breakfast; 2/3 of this breakfast dose is NPH, 1/3 is regular (or rapid-acting). 1/3 of the total daily dose is given in the evening; 1/2 of this evening dose is NPH, 1/2 is regular (or rapid-acting).

For example, for the 27 kg 10-year-old, the total daily dose is again 16 units (0.6 x 27).

Breakfast dose = 2/3 x 16 = 10.6

Breakfast NPH = 2/3 x 10.6 = 7.1 (give 7 units)

Breakfast regular (or rapid acting) insulin = 1/3 x 10.6 = 3.5 units (give 3.5 units)

Evening dose = 1/3 x 16 = 5.3

Evening NPH = 1/2 x 5.3 = 2.67 (give 2.5 units)

Evening regular (or rapid-acting) insulin = 1/2 x 5.3 = 2.67 (give 2.5 units)

After initiating the above insulin regimen, doses are adjusted based on monitored blood glucose levels (preprandial, at bedtime, and middle of the night). Generally, the breakfast NPH dose is adjusted based on the pre-dinner blood glucose, the breakfast regular (or rapid-acting) insulin dose is adjusted based on the pre-lunch blood glucose, the evening NPH dose is adjusted based on the overnight and pre-breakfast blood glucoses, and the evening regular (or rapid-acting) insulin dose is adjusted based on the pre-bedtime blood glucose.

What are the adverse effects associated with each treatment option?

No current insulin regimen can perfectly match the insulin requirements needed to maintain euglycemia. This results in blood sugar levels that are frequently above and below the target range.

Before meal target: 90-130 mg/dL (5.0-7.2 mmol/L); before bedtime target: 90-150 mg/dl (5.0-8.3 mmol/L).

The goal of insulin management in type 1 diabetes is to minimize both the magnitude and the frequency of out-of-range glucose levels. Minimizing the amount of hyperglycemia is important because this decreases the risk of long-term microvascular diabetes complications (retinopathy, neuropathy, and nephropathy). Acutely, however, hyperglycemia has no associated morbidity other than polyuria, unless the hyperglycemia becomes extreme as in DKA.

Hypoglycemia can result in the loss of consciousness and seizure. While the vast majority of even severe episodes of hypoglycemia have no apparent long-term morbidity, each episode does have the risk of complication resulting in morbidity and mortality. Injury can occur when hypoglycemia causes a loss of consciousness. This is particularly true during specific activities such as swimming and driving. And as with seizures from other causes there is a risk of hypoxic injury during hypoglycemic seizures.

Up to 6% of deaths in individuals with type 1 diabetes may be due to hypoglycemia. In addition, approximately 6% of deaths in patients with type 1 diabetes are individuals found “dead in bed” with no apparent explanation for their death with a higher proportion of deaths in younger patients with diabetes ascribed to this phenomenon. Hypoglycemia is one of the hypothesized causes for the “dead in bed” phenomenon. There is also data that suggests that recurrent severe hypoglycemia has a negative effect on children’s cognitive function. Thus, iatrogenic hypoglycemia is a consequence of insulin therapy in type 1 diabetes mellitus, and the risks associated with hypoglycemia limits the ability to prevent hyperglycemia in type 1 diabetes mellitus.

Among the most important initial skills that patients and families must be taught at the initiation of insulin treatment is the ability to recognize and treat hypoglycemia.

What blood glucose level defines hypoglycemia? What are the signs and symptoms of hypoglycemia?

Signs and symptoms of hypoglycemia occur due to the effects of cerebral glucose deficiency (neuroglycopenic symptoms) and to the effects of catecholamines that are released as a counterregulatory response to the falling glucose level (adrenergic signs and symptoms).

Hypoglycemia is defined as a plasma glucose <70 mg/dl 3.9 mmol/L). Catecholamine release is triggered as blood glucose falls below this level. Catecholamine release can also be stimulated at higher glucose levels if the level is declining rapidly, or in those with chronic hyperglycemia. Conversely, catecholamine release may be blunted, and only occur at a lower glucose level in those with chronic hypoglycemia (causing hypoglycemic unawareness).

The adrenergic signs and symptoms include pallor, sweating, trembling, palpitations, and anxiety. Neuroglycopenic symptoms occur at lower glucose levels than adrenergic signs and symptoms, typically at blood glucose levels less than 50-55 mg/dl (2.8-3.1 mmol/L). The neuroglycopenic symptoms include hunger, headache, lightheadedness, drowsiness, and confusion. As discussed above, with severe hypoglycemia coma, seizures and, very rarely, death can occur.

How is hypoglycemia treated?

Mild to moderate hypoglycemia – when the child is conscious and able to cooperate – is treated by ingestion of 10-15 grams of glucose:

4 ounces of juice or non-diet soft drink

5-8 Lifesavers® hard candies

2-3 hard candies (e.g., Jolly Rancher®)

4 packets (teaspoons) of table sugar

If the child is conscious, but is not cooperative, a glucose gel can be squeezed into the patients mouth. Note that glucose is not absorbed in the mouth – it must be swallowed to be absorbed. Therefore, this approach is not appropriate for an unconscious or seizing patient. Tubes of glucose gel marketed specifically for treatment of hypoglycemia contain 15g of glucose. Tubes of cake decoration gel can also be used.

It is important that hypoglycemia be treated with a source of pure carbohydrate, as protein and fat ingestion slows gastric emptying which would delay the correction of hypoglycemia (therefore, candy containing chocolate, for example, is not appropriate treatment of hypoglycemia.)

For severe hypoglycemia – when the child is having a seizure or is unconscious – glucagon is given as an intramuscular or subcutaneous injection. Patients should have a glucagon emergency kit with them at all times. These kits contain 1 mg of glucagon as well as a syringe for injection. For children less than 20 kg, 0.5 mg (1/2 of the emergency vial) is given; for larger children the full 1 mg vial is injected.

What are the possible outcomes of type 1 diabetes mellitus?

Type 1 diabetes mellitus is a lifelong disease. It does not go away, and there currently is no cure. However, our current treatment and management of type 1 diabetes allows children to grow to be healthy adults. While the proper management of type 1 diabetes requires numerous tasks each and every day, individuals with diabetes can participate in any activities and can pursue any goals that they would have if they did not have diabetes. Life expectancy is shortened for those with type 1 diabetes, but with current therapies, this shortening is quite small. For individuals born between 1965-1980, life expectancy is only 1.4-4.9 years less than that of the general population.

What causes this disease and how frequent is it?

  • The prevalence of type 1 diabetes in children 0-19 years of age in the United States is approximately 2 per 1000.

  • The incidence of type 1 diabetes varies substantially across the globe. Rates in China and parts of South America are near 1 per 100,000 person-years, while rates above 30 per 100,000 person-years occurs in Sweden, Finland and Sardinia.

  • The incidence of type 1 diabetes also varies across racial and ethnic groups within the United States. For youth 1-9 years of age:

    The incidence in non-Hispanic whites is now greater than 20 per 100,000 children per year

    The incidence in African American and Hispanic youth is approximately 15 per 100,000 children per year

    The incidence in Asian and Pacific Islander youth is approximately 7 per 100,000 children per year

    The incidence in Native American (Navajo) youth is below 5 per 100,000 children per year.

  • The incidence of type 1 diabetes mellitus is increasing worldwide, at a rate of approximately 3% per year:

    For US youth ages 0-14 years, the incidence of type 1 diabetes has increased from approximately 16 per 100,000 per year in the 1980s to 1990s to over 27 per 100,000 per year in 2002-2005.

  • There are two peaks in the age-specific incidence of type 1 diabetes mellitus: the largest is at 10-14 years of age, with a smaller peak in earlier childhood.

  • There is some seasonal variation in the onset of type 1 diabetes mellitus, with fewer cases presenting during the summer months.

What causes type 1 diabetes mellitus?

Type 1 diabetes mellitus is caused by the autoimmune destruction of the insulin-producing β-cells in the pancreatic islets of Langerhans. This results in an absolute deficiency of insulin. Symptoms become evident when approximately 80% of the β-cells are destroyed. Over time, nearly all of the β-cells are destroyed.

Type 1 diabetes is felt to occur when an environmental trigger initiates the autoimmune response in a genetically susceptible individual.

What is known about the genetics of type 1 diabetes mellitus?

There is an increased risk of type 1 diabetes in relatives of individuals with type 1 diabetes. In spite of this, however, only 10-20% of individuals with type 1 diabetes have family members with type 1 diabetes (See Figure 1).

Figure 1.

Risk of developing type 1 diabetes for the general population and for individuals with a relative with type 1 diabetes

The strongest contributor to the genetic risk of type 1 diabetes is the major histocompatibility complex (MHC), and specifically the class II DR and DQ antigens. This risk association is complicated by the extensive allelic variation of these genes, as well as by the genetic linkage between MHC genes.

Up to 90% of individuals with type 1 diabetes carry either the DR3/DQ2 or the DR4/DQ8 haplotypes. Thus only 10% of individuals with type 1 diabetes lack both of these haplotypes, while almost 60% of the general population will lack both of these haplotypes. However, while 2% of the population carry the high risk DR3-DR4/DQ8 genotype, 95% of these individuals will not develop type 1 diabetes.

Another locus conferring risk for type 1 diabetes is the variable number of tandem repeats located upstream of the insulin and IGF2 genes.

What are the environmental triggers for type 1 diabetes mellitus?

The concordance for type 1 diabetes in monozygotic twins is only 30-40%, indicating that something other than genes are responsible for the development of type 1 diabetes. This could include differences in the immune system, such as differences in T- and B-cell receptor gene rearrangements. However, another possibility is that environmental factors contribute to the development of type 1 diabetes.

A number of viruses have been implicated in the pathogenesis of type 1 diabetes:

Up to 20% of children born with congenital rubella infection will develop type 1 diabetes.

The role of other viruses in the pathogenesis of type 1 diabetes is less certain. There is, however, accumulating evidence of an association of enterovirus infections with type 1 diabetes. However, whether this association indicates a causative role of enteroviruses in the pathogenesis of type 1 diabetes, and the mechanism of such a role is not known.

Other factors have been proposed as environmental triggers for type 1 diabetes based on association studies including nitrosoureas, gliadin, and cow’s milk/bovine serum albumin. As for enteroviruses, the role of these agents in the pathogenesis of type 1 diabetes remains unproven.

The rodenticide Vacor causes an insulinopenic diabetes associated with islet cell antibodies. While Vacor-induced diabetes is a rare cause of diabetes, like congenital rubella infection and viruses, this finding demonstrates the potential for environmental toxins to contribute to the pathogenesis of type 1 diabetes.

What acute complications might you expect from type 1 diabetes mellitus?

The acute complications of type 1 diabetes are diabetic ketoacidosis and hypoglycemia. Hypoglycemia occurs in response to treatment with insulin, resulting from a mismatch of insulin action with insulin needs. DKA reflects the severe metabolic decompensation that results from severe insulin deficiency.

DKA is the most common cause of morbidity and mortality in children with type 1 diabetes. Early identification and treatment is key to avoiding severe DKA and its attendant risks.

Intercurrent illnesses, particularly infections, increase insulin needs. If these increased needs are not met, blood glucose levels will rise, and lipolysis and ketogenesis will begin. If the increased insulin needs remain unmet, decompensation into DKA can occur. Therefore, it is important to identify developing ketosis and to prevent this decompensation. Ketones in blood and urine can be measured by patients at home. This should be done at times of illness or persistent significant hyperglycemia (i.e., blood glucose >250 mg/dl (13.9 mmol/L)). Once identified, aggressive treatment with insulin is given to reverse the ketosis. While intercurrent illness is a precipitant of DKA, in children DKA is often caused by the inappropriate omission of insulin.

What are the long-term complications of type 1 diabetes mellitus?

Diabetes mellitus causes injury to the microvascular circulation. This results in tissue and organ damage, leading to the diabetic microvascular complications of retinopathy, nephropathy, and neuropathy. Diabetes mellitus also increases the risk of atherosclerotic vascular disease, the macrovascular complication of diabetes. Because of this, blindness, renal failure, and amputations are possible long-term complications of diabetes, and patients with diabetes have an increased risk of strokes and heart attacks.

Both the microvascular and macrovascular complications of type 1 diabetes mellitus are related to the hyperglycemia that persists even with treatment of the disease. The risk of these complications increases with the duration of the disease, and with higher levels of glycemia.

Some late adolescents who had an early onset of diabetes may show early evidence of complications such as non-proliferative retinopathy, microalbuminuria without loss of renal function, or clinically silent changes in nerve conduction. Fortunately, however, it is very rare for a child to have significant diabetic microvascular or macrovascular complications as these generally take decades to appear, even in patients with poorly controlled diabetes. However, children with type 1 diabetes have decades of life ahead of them with the disease, so it is important that glycemic control be maximized to minimize their risk of developing these complications as they age.

The hemoglobin A1c (HbA1c) is directly proportional to the average blood glucose level that was present during the lifespan of the red blood cell – and thus reflects the blood glucose level over the prior 3 months. An improvement in glycemic control that results in a decrease in HbA1c of 1% (reflecting a decrease in the average blood glucose level of 30-35 mg/dl (1.7-1.9 mmol/L)) decreases the risk of microvascular complications by 20-50%. There is no threshold for this effect; lowering the average blood glucose level always decreases the risk of microvascular complications. The challenge, however, is that lowering the average blood glucose level increases the acute risk of hypoglycemia.

How can type 1 diabetes mellitus be prevented?

There are a number of interventions that have been proposed as means to prevent the development of type 1 diabetes mellitus. These have been based on either observations and findings regarding the epidemiology of type 1 diabetes, or on theories of immune regulation. However, as yet there remains no proven prevention strategy.

Breastfeeding has been associated with a decreased incidence of type 1 diabetes, although it is not known for certain whether this is due to a protective effect of breastmilk per se or due to the effects of delayed exposure to cow’s milk proteins (or other complex dietary components). Even in the absence of definitive studies, given the overall health benefits of breastfeeding, for those families seeking recommendations, it is reasonable to recommend breastfeeding as a possible means to decrease the risk of type 1 diabetes.

Immune therapies have been investigated in attempts to prevent type 1 diabetes in individuals at increased risk of the disease. These have included treatment with the diabetes autoantigens insulin and GAD. While some studies have had preliminary data suggesting benefit, definitive studies have either not shown benefit, or have yet to be completed. Thus, while there are ongoing attempts to identify a therapy that could reduce the risk of type 1 diabetes in individuals with increased risk, no such treatment has been found yet.

The other approach to altering the immune destruction of β-cells in type 1 diabetes has been to use immunomodulators in patients with newly diagnosed diabetes to attempt to preserve the β-cell function that remains at the time of diagnosis. Such “secondary prevention” would be of benefit to patients with type 1 diabetes, as optimal glycemic control is more successfully achieved in the presence of residual β-cell function. In addition, if such a treatment were safe and effective, it could, at least in theory, be used prior to diagnosis in an at-risk individual to preserve sufficient β-cell function to prevent the development of diabetes.

Treatments that have been investigated include those directed at T-lymphocytes, such as monoclonal anti-CD3 antibody and anti-thymocyte globulin, as well as those directed against B-lymphocytes (Rituximab, a CD20 monoclonal antibody). While some studies have suggested that such an approach may be possible, no such treatment with a clear lasting benefit has yet been identified.

What is the Evidence?

“Standards of medical care in diabetes. 11”. Children and Adolescents. vol. 39. 2016. pp. S86-93. (Standards of care for children and adolescents with diabetes mellitus from the American Diabetes Association. The guidelines include grading of the evidence for the recommendations.)

Danne, T, Bangstad, H-J, Jarosz-Chobot, P, Mungaie, L. “Insulin treatment in children and adolescents with diabetes”. Pediatric Diabetes. vol. 15. 2014. pp. 115-34. (These are treatment guidelines for type 1 diabetes from the International Society for Pediatric and Adolescent Diabetes, and includes grading of the evidence of the recommendations.)

“2. Classification and diagnosis of diabetes”. Diabetes Care. vol. 39. 2016. pp. S13-S22. (This report outlines the diagnostic criteria for diabetes mellitus as well as the classification scheme for the different types of diabetes mellitus.)

Jackson, CC, Albanese-O’Neill, A, Butler, KL, Chiang, JL. “Diabetes care in the setting: a position statement of the American Diabetes Association”. Diabetes Care. vol. 38. 2015. pp. 1958-1963.

Siminerio, LM, Albanese-O’Neill, A, Chiang, JL, Hathaway, KJ. “Care of young children with diabetes in the child care setting: a position statement of the American Diabetes Association”. Diabetes Care. vol. 37. 2014. pp. 2834-2842. (These reports give recommendations for the management of diabetes mellitus in children in the school and day care setting.)

“The effect of intensive diabetes treatment on the development and progression of long-term complications in IDDM: the DCCT”. New England Journal of Medicine. vol. 329. 1993. pp. 977-986. (The Diabetes Control and Complications Trial (DCCT) was the seminal study proving that intensive diabetes treatment that lowers the average blood sugar decreases the risk of microvascular complications of diabetes.)

“The effect of intensive diabetes treatment on long-term complications in adolescents with IDDM: the DCCT”. Journal of Pediatrics. vol. 125. 1994. pp. 177-188. (The data from the DCCT regarding adolescents in the trial indicating that, as with adults, intensive diabetes treatment that lowers the average blood sugar decreases the risk of microvascular complications of diabetes.)

Ongoing controversies regarding etiology, diagnosis, treatment

One approach to cure type 1 diabetes is to use transplantation to replace the destroyed β-cells. Such transplantation requires ongoing immunosuppression in order to prevent the immune destruction of the transplanted β-cells. In patients already exposed to the risks and side-effects of immunosuppression for a renal transplant, subsequent or concurrent pancreas transplant is an appropriate consideration. However, assumption of the risks of immunosuppression is not felt to be appropriate for the vast majority of patients with type 1 diabetes.

Pancreas only transplants could be considered in patients with recurrent, severe acute complications of diabetes (hypoglycemia, ketoacidosis) with a consistent failure to prevent these acute complications with usual diabetes treatments. Because pediatric patients will not develop diabetes nephropathy sufficient to require a kidney transplant, the issue of kidney-pancreas transplant does not arise in pediatric patients. In addition, the risks of transplantation and life-long immunosuppression are not felt to be appropriate for pediatric patients with type 1 diabetes, even those with recurrent severe acute complications. In such patients, poor compliance is very likely a significant contributor to their acute complications, and would be a contraindication to transplantation.

Injection of isolated cadaveric islets into the portal vein is a less invasive approach to islet-only transplantation. As with pancreas transplantation, islet transplantation requires lifelong immunosuppression. Survival of islet function in this approach remains significantly lower than that of whole pancreas transplantation. This technique remains experimental, and is currently restricted to adults with recurrent, severe hypoglycemia.

Another experimental avenue of investigation for a cure of type 1 diabetes is the use of stem cell therapies to replace the lost β-cells. Such an approach would also require an intervention to prevent the destruction of these cells by the underlying immune process present in those with type 1 diabetes. As for other approaches to cure diabetes, this approach remains experimental.

Jump to Section