Around 10% of diabetics in India have onset below 30 years of age. Youth-onset diabetes in developing countries is a heterogenous set of disorders. Till recently, information on occurrence, chnico-pathological and biochemical features as well as etiology of youth onset diabetes in India were not coherent. Though in 1985, a WHO group of experts classified youth-onset diabetes in developing countries into different categories, that classification has recently been questioned and is now under review. Meanwhile type1 diabetes as seen in the West being more clearly defined immunologically and sub-categories are emerging on the basis of natural history and immune markers.

At the AlIMS we operate a "diabetes in the young" clinic. Patients with age of onset of diabetes <30 years are registered in this clinic and provided state of the art diabetic care free of cost. The clinic has registry of 1033 with an average of 110 fresh patients registered each year. Catchment area of the clinic spreads beyond confinement of Delhi and includes States like Uttar Pradesh, Rajasthan, Haryana, Bihar, Jammu and Kashmir, West Bengal, Assam and Orissa. A recent analysis of the patients attending clinic showed that 60%, of them were socioeconomically poor with monthly income of parents up to Rs.2,000. In our clinic it is common to have patients with type1 diabetes mellitus presenting with ketoacidosis as a result of interrupted insulin therapy due to financial constraints. Young patients with ketosis-resistant diabetes present in states of extreme wasting due to chronically poor glycemic control.

The study subjects comprised of 132 consecutive patients with diagnosis of diabetes less than 30 years of age who attended the "Diabetes in the Young" Clinic on the AIIMS during the 2 year period spanning 1995-1997. The mean age was 25±8.0 and mean duration 5.5±6.0 years. 49 patients were females while 83 were males.

Clinical characteristics, insulin requirement, ketosis proneness on insulin withdrawal, fami1v history of diabetes, nutritional status and radiological and ultrasonographical evidences of pancreatic calculi were assessed. C-peptide was measured at 0 and 6 mts. post glucagons (1 mg IN). The daily insulin dose requirements were assessed by close monitoring of blood glucose with administration of purified insulin before the three principal meals to achieve ideal glycemic status throughout the day (fasting blood glucose < 120 mg% and post meal < 140 mg%). The patients were categorized as type 1, type 2 and fibrocalculous pancreatopathy (FCPD). Type1 diabetes was diagnosed if ketosis developed on insulin withdrawal. Fibrocalculous pancreatopathy was diagnosed on the basis of pancreatic calculi on plain X-ray abdomen or ultrasonography. Diagnosis of type 2 diabetes was made if subjects achieved glycemic control without insulin and with normal or above normal body mass indices (BMI). Diabetes as part of an autoimmune polyglandular syndrome was considered if there was a co-existing clinically overt autoimmune thyroid disorder. No other organ specific autoimmune disorder coexisted with the present group of youth-onset diabetic individuals. Additionally, an entity designated as 'ketosis resistant type' was categorised. These were insulin requiring youth-onset diabetic individuals, who were resistant to ketosis on insulin withdrawal and who has low BMI with other features of malnutrition. More recently, an International Workshop (1995) named this entity 'malnutrition modulated diabetes (MMDM)'. They did not have seen in type1 and autoimmune polyglandular groups respectively. The fibrocalculous pancreatopathy group showed GAD 65 plus IA-2 autoantibody positivity in 14.2% and lone GAD 65 autoantibody positivity in 7.1%. Twenty six percent and 60% respectively of the type 1 and autoimmune polyalandular syndrome groups had thyroid nucrosomal autoantibody positivity. Type 1 showed significantly less C-peptide response to glucagons when compared to the ketosis-resistant and autoinimune polyglandular syndrome groups. The controls and type 2 diabetic individuals tested negative for tested cell autoimmunity markers.

Seventy patients (43 males and 27 females, mean (SD) age 17.9±6.1 years, mean duration of Bovine insulin therapy 5.1 ± 5.4 years, were studied.

Radiobinding assays for specific insulin binding was carried out by using a 121 tracer of human insulin prepared in our laboratory using Chloramine-T technique. The mean and specific binding of 100 healthy subjects without any family-history of diabetes or other autoimmune disease was 0.9±1.1%. The insulin antibody titres observed in patients was expressed in the assay precision unit, SD score.

Briefly, all the patients treated with bovine insulin showed high titres of insulin antibodies with SD score ranging from 5.1 to 42. No significant difference was observed in the mean SD score of insulin antibodies in the three diabetic groups. Insulin antibodies SD score and its affinity did not show significant relationship with daily insulin dose and HBAI at admission. Only 27±7% variations in daily insulin dose requirement were accounted for by total insulin binding power. There was a significant inverse relationship between insulin antibody SD score and duration of insulin therapy (r = -0, 4172, p<0.0004).

There was no relationship between insulin antibody titres and daily insulin dose requirement. In fact despite presence of high insulin antibody titres, daily insulin dose requirement remained within physiological range in all three types of diabetics studied. The major variation in insulin dose requirement is thus governed by factors other than insulin antibody.

Overall, these results indicate that bovine insulin related antibody response does not result in any clinically significant impact in terms of daily insulin requirement for diabetes control.