Advances in Diabetes Mellitus in Children

   Puneet Kumar

Diabetes mellitus (DM) is the 2nd most common chronic disease in pediatric age group. Last 25 years have seen rapid strides in deeper understanding of almost every aspect of the disorder, resulting in significant changes in the way it is managed and ultimately in improved outcomes. This compilation aims to update practicing pediatricians with these changes. This is not an exhaustive coverage of diabetes in children, as the focus remains on the changes that have been there recently. Secondly, the developments that are clinically relevant have been reviewed in greater detail. The format of FAQs has been purposely chosen, considering the time constraints of a busy practioner.

Q. 1. Why has there been a change in nomenclature/ classification of Diabetes Mellitus?
The older terms of “insulin dependent diabetes mellitus” (IDDM)/”juvenile onset diabetes mellitus” and “non-insulin dependent diabetes mellitus” (NIDDM)/”juvenile onset diabetes mellitus” have been discarded, since it is now clear that about 50% cases of type-1 diabetes mellitus (T1DM) present in older age groups and some cases of type-2 diabetes mellitus (T2DM) do present in pediatric age group. Further, many cases of T2DM are “insulin dependent” in later stages.

The classification of DM has been changing overtime with clearer understanding of underlying etipathogenesis. The latest classification as per American Diabetes Association is listed in Table-1.

Table-1. Etiologic classification of Diabetes Mellitus

Category of diabetes


Type 1 diabetes (destruction of beta cells with absolute insulin deficiency)

1A: Autoimmune (95% of T1DM cases)

1B: Idiopathic

Type 2 diabetes (due to combination of insulin resistance with relative insulin deficiency in varying proportions)

2A: Typical

2B: Atypical

Other specific types of diabetes

(these account for 4% of all DM cases)

1.     Genetic defects of beta cell function: Neonatal DM (transient/ permanent), MODY (with 11 subtypes involving distinct genes), Wolfram syndrome, Roger syndrome, Mitochondrial DM

2.     Genetic defects in insulin action: Type A insulin resistance, Leprechaunism, Rabson-Mendenhall syndrome, Lipoatrophy syndromes

3.     Disorders of exocrine pancreas: FCPD, cystic fibrosis, chronic pancreatitis, autoimmune pancreatitis, hemosiderosis, beta-thalassemia, alpha-1-antitrypsin deficiency, pancreatectomy

4.     Endocrinopathies: Cushing syndrome, hyperthyroidism, pheochromocytome, ets.

5.     Drug or chemical induced: High dose steroids, vincristime, azathioprine, alpha-interferon, diazoxide, beta-blockers, thiazides, pentamidine, olanzapine

6.     Infections: Congenital rubella, CMV and others

7.     Uncommon forms of immune mediated diabetes: Stiff person syndrome, APS-1, IPEX syndrome, type B insulin resistance

8.     Miscellaneous genetic syndromes sometimes associated with diabetes: Down, Turner, Prader-Willi and Laurence-Moon-Biedl syndromes, myotonic dystrophy, Friedreich ataxia

Gestational diabetes



While T1DM accounts for most cases of diabetes mellitus in children, following red-flag signs should alert the pediatrician towards the possibility of other types of diabetes mellitus, since the management is significantly different in different types of DM:

·      Obese adolescent with acanthosis nigrans, hypertension and/or features of PCOD and/ or strong family history of T2TM.

·      Very low insulin requirement with normal C-peptide after 2-3 years of DM.

·      Mild, non-progressive course with HbA1c always remaining <7%.

·      Onset of DM < 6 months of age.

·      Absence of pancreatic antibodies at outset.

·      History of DM in a last 3 generations (autosomal dominant inheritance).

·      History of recurrent abdominal pain/history suggestive of exocrine pancreatic insufficiency.

·      Preceding chronic ailment in a patient diagnosed with DM.

·      Patient on medications known to cause insulin deficiency or resistance.

·      Patient with syndromic  (dysmorphic) features or lipoatrophy.

·      Deafness or optic atrophy in patient/family.

·      Lean patient with very high insulin requirement (especially if associated acanthosis nigricans and hyperandrogenism).

·      History/findings suggestive of endocrinopathies.

If there is a doubt regarding classification of a patient, pancreatic antibodies [antibodies to glutamic acid decarboxylase (GAD65), islet cell antibodies (ICA), Insulin autoantibodies (IAA) and antibodies to insulinoma associated antigen-2 (IA-2)] and C-peptide levels can be tested.

About 50% of neonatal diabetics can be managed with sulfonylureas with excellent glycemic control and overall quality of life, all patients who are diagnosed with DM in age group <6 months of age (and possibly below 12 years of age) must be tested for specific mutations. This also allows for genetic counseling for families.

Q. 2. Is the incidence of Diabetes Mellitus in children rising?

A. Incidence of T1DM is steadily increasing in most (but not all) populations and this increase is most marked in countries where incidence of autoimmune disorders has been traditionally low. Secondly, the increase is most marked in younger children (1-5 years age group) and that, in turn, has a multiplier effect on overall prevalence. Overall, there has been about 3% increase in cases each ear since 1990. The reason of increasing incidence remains elusive, since known environmental/ genetic factors fail to explain this. The most plausible hypothesis is “hygiene hypothesis” that states that with reduction in incidence of childhood infections leaves the immune system “less trained” to handle its main task (host defense) that leads to increasing incidence of autoimmune diseases, including autoimmune DM. However, even this remains a “hypothesis” as on date as studies done to prove this have thrown up equivocal results.

The incidence of T1DM varies widely across the globe from 0.1 [per 100,000 children (0-14 years] in Paraguay and 0.5 in Pakistan to 62.3 in Finland. The incidence in India is 3 per 100,000 children in 0-14 year age group.

Incidence of T2DM is also increasing in almost every part of the world and the increase parallels increasing urbanization, sedentary lifestyles, and unhealthy diets, including high sugar intake and thus, increasing obesity.  Since Asians in general seem to develop T2DM at lower body mass index levels than Europeans, but continue to have much lower incidence of T1DM as compared to the West, the proportion of children having T2DM is increasing in Asian countries, with some centers reporting upto 50% of children with DM having T2DM.

Q. 3. What are the recent advances in our understanding of etiopathogenesis and natural history of diabetes mellitus?

A. T1DM occurs in a when a genetically predisposed child is exposed to certain environmental factors that trigger autoimmunity (in 90% cases) or an unknown mechanism that damages beta cells of pancreas. The damage to beta cells is not mediated by autoantibodies, but is T-cell mediated process.

Now, it is clearly known that the genetic predisposition to T1DM is closely linked to HLA class II genes that are involved in immune regulation. While HLA DR3/DQ2 and HLA DR4/DQ8 confer risk, others like HLA DR15/DQ6 have a protective effect. While 50% of genetic susceptibility comes from HLA variants, the other half comes from 40 other genes that are also now well characterized. Despite the fact that T1DM has a large genetic component, 85% of newly diagnosed cases have no family history of T1DM.

As far as environmental factors are concerned, numerous viral infections have been implicated in pathogenesis of T1DM, but strongest evidence is available for congenital rubella infection that results in beta-cell autoimmunity in 70% cases and T1DM in 40% of these in 2nd decade of life. However, postnatal infection with rubella virus or vaccination with live vaccine does not increase the risk of T1DM. Among dietary factors, early introduction of cow’s milk, early introduction of gluten, deficiencies of omega-3 fatty acids/Vitamin D/ Vitamin C/ Vitamin E/ Zinc have all been implicated, but the evidence is lacking. Breastfeeding is considered to have protective effect. Ongoing trials like TEDDY (The Enviornmental Determinants of Diabetes in Young) are likely to shed more light on these factors.

Natural history of T1DM involves 6 stages: 1. Initiation of Autoimmunity; 2. Preclinical autoimmunity with progressive loss of beta cells; 3. Onset of Clinical disease; 4. Transient remission; 5. Established disease; 6. Development of complications.

Research in last 25 years has added some new perspectives in our understanding of the natural history of T1DM:

1.     Precipitating events might occur even in utero

2.     Genetic predisposition probably the key driver or linkage to immune abnormalities

3.     Beyond precipitating, environment might influence entire natural history

4.     Although overall loss of B cells is potentially linear, it could show a relapsing or remitting pattern

5.     Presence of two of more islet autoantibodies might represent asymptomatic T1DM

6.     Increasing glucose excursions as individual approaches symptomatic  onset

7.     Some patients produce low concentrations of c-peptide long after onset

8.     B-cell mass not always zero in long-standing disease  (leaving a window of opportunity of secondary prevention even in long standing cases!)

T2DM has got a stronger genetic component with most cases having a strong family history of T2DM. However, the genetic basis of T2DM is very complex and ill-understood. There is no single identified defect predominates like HLA in T1DM. The currently identified genetic polymorphisms that confer the risk of T2DM are able to explain less than 20% of population risk of T2DM. Shared environmental/ lifestyle factors have a dominant role in the genesis of T2DM. Moreover, epigenetic factors and fetal programming are also involved in pathogenesis of T2DM, but the underlying molecular mechanisms are yet to be understood.

Q. 3. Have there been any changes in the protocol for diabetic Ketoacidosis (DKA)?

A. The Milwaukee protocol (revised in 1988) has been widely used for management of DKA with satisfactory results. However, incidence of cerebral edema is the commonest cause of death during management of DKA. Subsequent recommendations from British Society for Pediatric Diabetes (BSPED) [in 2004 & 2007] and those from International Society for Pediatric and Adolescent Diabetes (ISPAD) [in 2009 and 2014] have fine-tuned the protocol, mainly to reduce the risk of cerebral edema. The main changes that are recommended are:

·      Initial bolus of insulin is not to be given.

·      The initial fluid bolus is 10mL/ kg, unless the child who presents in shock where the dose is 20mL/kg

·      Subsequent fluid infusion has to be 0.9% saline for 4-6 hours, preferably up to 12 hours (earlier this was recommended only for 1 hour). Thereafter, replacement fluid should be 0.45% saline. However, corrected sodium levels need to be calculated repeatedly. Corrected sodium levels increase with fall in blood glucose levels. If it doesn’t happen, 0.9%saline is continued, rather than changing to 0.45% saline.

·      Rate of insulin infusion should be 0.1U/kg/hour (earlier recommendation was 0.05-0.1U/kg/hour).

·      The insulin infusion is started 1-2 hours after fluids have been started (earlier it was recommended to start concurrently with fluids.

·      The fluid deficit is corrected over 48 hours (earlier, it was recommended to correct over 23 hours.

·      The insulin infusion has to continue at the same rate (0.1U/kg/hour) till the resolution of DKA (pH>7.3). (Earlier, it was lowered after correction of hyperglycemia)

Q. 4. How have insulins evolved over the years?

A. Discovery of Insulin in 1921 changed T1DM from being a uniformly fatal disease to a manageable disease. Next improvement was development of “pre-mixed” insulins and availability of disposable syringes that simplified administration of insulin injections. Further, with dawn of recombinant technology, human insulin became available that improved the safety profile of exogenous insulin. However, unlike hypothyroidism, the hormone replacement therapy in DM is far from perfect. Since the exogenous insulin is delivered subcutaneously rather than portal circulation, the delivery to liver is lesser. Secondly, the dose of insulin is based on anticipated diet and activity pattern, rather than prevailing blood glucose levels. Third, the absorption of insulin from injection site varies with activity. Fourth, the exogenous insulins have a lag period, longer half- life and long tail, thus requiring careful structuring of food intake and activity to maintain euglycemia. The last of these issues has been tackled with development of insulin analogues in which the structure of insulin is altered so as to have a desired time action profile. The rapid acting analogues (Lispro, Aspart and Glulisine) have a very rapid onset of action. Hence, there is no need to wait for 30 minutes after injection (as with human regular insulin). In fact, it can be given immediately after meal: thus making it extremely useful for toddlers and on sick days when food intake after injection is not assured. These analogues also achieve quicker and higher peak level, hence are better in controlling post-meal glycemia. The decline of action is also faster (within 3 hours) with these analogues, hence reducing the risk of delayed hypoglycemia and obviating the need of extra snacks “to feed the insulin”. In view of all these advantages, currently, these analogues are preferred over human regular insulin wherever possible. Clinician has to consider the cost of therapy also while prescribing: while a vial of human regular insulin (400units/vial) costs about D140, one cartridge (300 units/cartridge) costs about D250, one cartridge of rapid acting insulin analogues (300 units/cartridge) cost about D 550.


Long acting (Glargine and Detemir) and ultra-long acting (Degludec) insulin analogues have also been developed to tackle problems associated with pharmacokinetic properties of NPH/ Lente/ ultralente insulins. These analogues give more steady levels (unlike NPH that has a distinct peak) that obviate the need of extra snacks between the meals and reduce chances of nocturnal hypoglycemia. Duration of action of Glargine insulin is 24 hours in most (70-80%) children, hence, once daily dose suffices. In others, twice a day dosing is required. With Detemir, the dosing is twice a day and it has a small peak 3-9 hours after injection, unlike glargine insulin that is virtually “peakless” action.  Degludec licensed only for adults; however some European countries are using it even in younger children. As far as cost is concerned, NPH insulin vial (400U/vial) cost D140, cartridge (300U/cartridge) cost D400 while that of Glargine costs D600 per cartridge.

Q. 5. Which Insulin Regimens are currently recommended for day-to-day management of T1DM?

A. Traditionally, split-mix regimen has been used where the patient injects a mixture of short acting and long acting insulin (most commonly in 30:70 ratio) before breakfast and dinner. However, regular insulin limits postprandial glycemia and causes delayed hypoglycemic effect because of delayed & wide peak and long duration of action. The NPH/ Lente insulin also has limitation of producing a peak at noon and midnight, requiring extra snacks in evening and risk of nocturnal hypoglycemia.

Development of insulin analogues helped formulate basal-bolus regimens of insulin therapy that provided more physiological glycemic control.  In these regimens, long acting insulin analogue is given once or twice a day to regulate hepatic release of glucose in fasting state while 3-4 injections of regular insulin (or preferably rapid acting insulin analogue) are given to control postmeal glycemia. The varying carbohydrate content in the child’s diet and varying exercise pattern does not affect glycemic control with this regimen, if the pre-meal boluses of insulin are carefully adjusted. The grams of carbohydrate for which 1 unit of insulin is needed is obtained by dividing 500 by total daily dose (TDD) of insulin. The results of the landmark Diabetes Control and Complications Trial (DCCT), a large, multicenter, randomized, long-term clinical trial conclusively proved that maintaining near-normal glycemia with minimum glucose level excursions over long term significantly improved long-term outcomes of diabetic patients with over 50% reduction in development and progression of microvascular complications. Hence, this basal-bolus regimen has become standard of care for management of T1DM.

However, such an intensive regimen requires lot of commitment and motivation on part of the patient and family to be successful. Repeated self-monitoring of blood glucose (SMBG) is integral part of this regimen. A fair middle path is to prescribe a 3-injection regimen that involves giving NPH with a rapid acting insulin analogue at breakfast, a dose of rapid acting insulin analogue before dinner and glargine insulin at bedtime. Often even that is not possible in many cases in our setting, and we have to continue with split-mix regimen in many cases.

Q. 6. What is insulin pump therapy? Is it available in India?

A. Also called as Continuous Subcutaneous Insulin Infusion (CSII) device it is a small, battery-powered device that delivers insulin continuously in subcutaneous space. It uses only rapid acting insulin analogue: both for boluses and basal insulin delivery. Basal insulin rates can be variable for different times of the day and even can be temporarily suspended: for example during unscheduled exercise. It automatically calculates and delivers appropriate dose of insulin according to the pre-meal blood glucose level and amount of carbohydrate in the diet: these figures need to be entered at every meal-time. Many pumps also have an additional feature by which even bolus can be given at a variable rate: as one shot or slowly over 0.5-4 hours depending upon glycemic index (fat and fibre content) of meal to be eaten can also be adjusted, if these values are entered into the device. The episodes of severe hypoglycemia are reduced to a great extent by CSII. The degree of glycemic control, even with CSII is heavily dependent on the motivation and commitment of the patient/ family. Blood glucose values are to be checked 4-10 times a day and data needs to be fed at least 4 times a day. The subcutaneous catheter needs to be changed every 2-3 days, and the chances of local infection are always there.  The device is specially recommended for toddlers and those with wide swings in blood glucose values, very high HbA1c values, and those with frequent episodes of hypoglycemia.

It is available and is being used in many centers in India. The cost of the device is D 1.5 lakhs and the cost of consumables is about D 5000/ month.

Q. 7. How has therapeutic monitoring changed with advancements in management of T1DM?

A. The DCCT trial and the follow up EDIC (Epidemiology of Diabetes Interventions and Complications) trial showed that microvascular as well as macrovascular complications of DM can be reduced with intensified regimens and maintaining near-normal glycemia, but such regimens increased the risk of hypoglycemia by upto 3 times and risk of obesity by 2 times as compared to conventional regimens. This necessitated more careful and frequent monitoring, rather than just fasting and post-prandial blood glucose monitoring. Availability of point-of-care glucometers and lancet devices for almost painless bloodletting made frequent self-blood glucose monitoring (SMBG) feasible. Every patient needs to monitor blood glucose levels 2-4 times a day. These can be planned at different times of the day on different days of the week so as to have good understanding of blood glucose levels with minimum testing. An example of such a scheme is presented in Table-2. In addition, urine ketones need to be checked on sick days. A meter to measure blood ketone level is also available: that is more reliable than urine ketone test.

HbA1c measurement (glycosylated Hb levels) reflects average blood glucose level of preceding 2-3 months. It is recommended to get tested 3-4 times a year. While interpreting HbA1c levels, following must be considered: 1. HbA1c levels vary from lab to lab depending upon methodology used. Hence reference values provided by the lab have to be used. 2. In patients with thalassemia or other conditions with raised HbF, HbA1c levels may be spuriously raised, while in patients with sickle cell disease or other conditions with increased turnover of RBCs, the HbA1c levels might be spuriously low. In such patients, fructosamine test can be done that reflects average blood glucose levels of preceding 2-3weeks.

Table-2. Example of scheme of SBGM over a week

For patients on basal-bolus regimen


3 am








































































For patients on split-mix regimen

























































*Comments: Space to document any unscheduled exercise/ food/ stress factor underlying aberration in report.


The target values of blood glucose levels and HbA1c levels vary with age, as tighter control is possible in older children. American Diabetic Association (ADA), in its 2014 recommendation has defined these targets in different age-groups. They are given in Table-3. If pre-meal blood glucose level is high, the additional dose of insulin needs to be calculated. Dividing 1800 by total daily dose of insulin represents the anticipatory drop in blood glucose level (in mg/dL) with 1 unit of insulin.

Table-3. Target HbA1c and blood glucose levels as per ADA 2014 recommendations

Age group

Target HbA1C

Bedtime and 3 am blood glucose

Premeal blood glucose

< 6 years

< 8.5%

110-200 mg/dL

100-180 mg/dL

6-12 years

< 8.0%

100-180 mg/dL

90-180 mg/dL

13-18 years

< 7.5%

90-150 mg/dL

90-130 mg/dL


One of the most important (and often forgotten) aspects of therapeutic monitoring is testing for co-morbidities and complications. All patients of T1DM must be tested for other autoimmune disorders. Hence, TSH and anti-thyroid antibodies, tissue transglutaminase IgA and IgA levels, 21-hydroxylase antibodies are tested in every newly diagnosed case of T1DM. Thereafter, testing for TSH, anti-thyroid antibodies and tissue transglutaminase is recommended on an annual basis. As far as complications are concerned, urine for microalbuminuria, lipid profile and ophthalmic examination (including fundoscopy) is recommended at diagnosis and then annually, starting 5 years after diagnosis in prepubertal child or 2 years after diagnosis in a pubertal child. Regular assessment for growth, development (including pubertal development) and psychological well-being is of utmost importance.

Q. 8. What is Continuous Glucose Monitoring System (CGMS)?

A. As the name suggests, it is a new technology that measures blood glucose levels “continuously” so as to give a better insight into the glycemic control round the clock. In this a tiny, flexible glucose sensor is inserted in subcutaneous space that measures interstitial glucose level every 5 minutes for 3-14 days. First generation of CGMS provided data that could be downloaded only at the end of monitoring period. These are called historical CGMS. Newer generation of CGMS are “real time CGMS (rtCGMS)” that have a small display screen attached wirelessly with the glucose sensor that gives the blood glucose values (converted from interstitial glucose values) on real time basis in form of graphical display of the trend (speed and direction of blood glucose movement). However, even in these rtCGMS, the blood glucose values are not “real time” in real sense, as there is a lag of about 13 minutes between blood glucose and interstitial glucose values and a further lag of 2-5 minutes between processing and display of the data. Hence, it does not directly be used to control insulin administration, but just give overall pattern of glycemic control and fine tune the management. They are very useful for detecting nocturnal hypoglycemia which often gets missed otherwise. Latest CGMS have alarms that can be set at below and above certain blood glucose values. Short term studies have shown improved control of diabetes over conventional methods, when used by motivated and well-informed families.

Q. 9. What is the current status of oral hypoglycemic in pediatric age group?

A.  In pediatric age group, metformin is the only drug that is FDA-approved for management of T2DM. Sulfonylureas are useful in certain forms of Neonatal Diabetes (mutations in KCNJ11 and ABCC8 genes) and other specific forms of diabetes (MODY types 1, 2, 3 and 4). All other classes of oral hypoglycemic drugs (Glitinides, Thiazolidinediones, GLP-1 analogues etc.) that are used in management of T2DM in adults are not yet approved for use in pediatric age group.

Q. 10. What developments are in the pipeline in the field of diabetology?

A. Diabetes mellitus continues to be one of the hot topics of research globally. While the ultimate goal is to prevent it altogether and to cure those who are suffering already, there are many developments that, once take final shape, will definitely improve the prognosis and quality of life of patients this serious, chronic disease. Table-3 list these developments as well as the hurdles/ issues to be taken care of, so that they see the light of the day.

Table-3. Future developments in diabetology



Issues/ hurdles to be taken care of

Inhaled insulin/insulin patches/oral insulin

Treatment of DM would become non-invasive

Bioavailability issues and mode of dose adjustments

Elucidation of dietary and other environmental factors in genesis of diabetes

Tackling those factors would reduce risk of diabetes in genetically predisposed individuals/ communities

Exact pathogenesis of DM ill-understood; large interplay between genetic and environmental factors

Antigen-based immunotherapy to antagonize autoimmunity

Tackling autoimmunity would prevent genesis of diabetes

Complex, ill-understood pathogenesis; trials with parenteral/nasal insulin, GAD and IL-1β inhibition have failed, but those using oral insulin have shown encouraging results

Closed loop insulin pump (connecting CGM with CSII interposed with a microprocessor) (“artificial pancreas”)

Excellent glycemic control with minimum human interface

Removing lags at various stages: between blood glucose and interstitial glucose, time required for sensing and interpreting interstitial glucose by the device, and between CGM and insulin pump, calculation of insulin dose based on data received, between subcutaneous delivery of insulin and action of insulin at periphery and liver, etc.

Islet cell transplantation

Potential to cure diabetes

Obtaining adequate number of islet cells from cadaveric donors, need for lifelong immunosuppression

Cells that can produce insulin (genetic engineering)

Potential to cure diabetes

Control of insulin production based on ambient glucose levels

Stem cell transplantation

Potential to cure diabetes

Regulating growth of stem cells, regulating production of insulin and preventing rejection by immune system

Pancreas transplantation

Potential to cure diabetes

Huge surgical risk, need for lifelong immunosuppression


-Puneet Kumar, Kumar Child Care Clinic, New Delhi



1.     Sroven BM, Jospe N. Diabetes Mellitus in Children. In: Kleigman RM, Stanton BF, St Geme JW, et al (Eds). Nelson Textbook of Pediatrics, 20th edition, 2016, Elsevier Inc: pp2760-2790

2.     Irani A. Classification of Diabetes Mellitus. In: Gupta P, Menon PSN, Ramji S, et al (Eds). PG Textbook of Pediatrics, 1st edition, 2015, Jaypee Brothers:  pp 2374-78.

3.     Irani A. Type-1 Diabetes Mellitus. In: Gupta P, Menon PSN, Ramji S, et al (Eds). PG Textbook of Pediatrics, 1st edition, 2015, Jaypee Brothers:  pp 2378-84.

4.     International Diabetes Federation. IDF Diabetes Atlas, 7th edn. Brussels, Belgium:
International Diabetes Federation, 2015. Available online:
http://www.diabetesatlas.org. Accessed on December 8, 2016.

5.     Ravikumar KG. Acute and Chronic Complications of Diabetes Mellitus. In: Gupta P, Menon PSN, Ramji S, et al (Eds). PG Textbook of Pediatrics, 1st edition, 2015, Jaypee Brothers:  pp 2384-88.