A major problem in replacing insulin in type 1 diabetes mellitus is that currently no depot preparation exists that is capable of mimicking the background insulin secretion of the healthy pancreas. All available intermediate or long acting insulin preparations have a peak period profile, excess insulin action at midnight, and insulin waning at dawn occur whenever an insulin is given at supper time. This easily results in hypoglycemia in the early evening hours and in the fasting state. The dawn phenomenon is the combination of an initial decrease in insulin requirements between 2400 and 0300 hours followed by an increase in insulin needs between 0500 and 0800 hours.

Nocturnal hypoglycemia is a fearful condition for three reasons.

1) During sleep the autonomic symptoms may not be potent enough to awaken the patient thus mild hypoglycemia may progress to severe hypoglycemia at a time when external assistance to the patient may not be available.

2) The excessive correction of hypoglycemia at night by spontaneous counterregulation and food ingestion may cause subsequent hyperglycemia.

3) Frequent episodes of unrecognised nocturnal hypoglycemia may induce hypoglycemia unawareness which may lead to severe hypoglycemia at any time day or night. Furthermore for a given plasma insulin concentration hypoglycemia may induce subsequent hyperglycemia at any time.

Strict control of blood glucose levels is therefore essential to avoid these side effects. (10)

Clinical Pharmacology of Human Insulin

Human insulin is used daily by millions of diabetic patients. The biological effect of human insulin is comparable to that of porcine insulin. After subcutaneous injection, pharmacological and clinical studies have showed pharmacokinetic and pharmacodynamic differences between human and animal insulins. These differences can be more pronounced and can be of clinical relevance with intermediate and long acting insulin preparations. Optimal metabolic control can be achieved with either human or highly purified animan insulin preparations provided appropriate insulin replacement strategies are adopted.

Human insulin has made it possible to treat IDDM patients with a hormone that has an ammino acid sequence identical to endogenous insulin. After characterization of the biological activity of human insulin in vitro and in animal studies a series of efficacy and safety trials was performed in humans. Several studies in early years compared the potency of human insulin and animal insulin preparations with respect to their pharmacological properties. Later studies compared human insulin preparations manufactured using different methods.

Most of the literature on human insulin, including proceedings of commercial sponsored symposia, was published 15 - 20 years ago, all non-peer reviewed, and very few papers have passed a peer review system. This disturbing fact may be relevant since pharmacological differences between human insulin and animal insulin may have practical implications for the daily therapy for millions of patients worldwide and possibly up to 80,000 IDDM patients in the UK.

The structure of animal insulin has minor but potentially important differences from human insulin. Porcine insulin differs by one amino acid (alanine instead of threonine at the carboxy-terminal of the beta chainat position B30), and beef insulin differs by two additional alterations of the sequence of the A-chain (threonine and isoleucine on positions A8 and A10 are alanine and valine). There is thus nearly a complete homology between human insulin and porcine insulin in the amino acid sequence. Biosynthetic production of human insulin has been made possible by advances in genetic engineering especially in recombinant DNA technology. In the drug model human insulin shows a more rapid onset and shorter duration of action, along with a lower potency, compared to bovine insulin. When using the intravenous route investigators have come to the conclusion that there is little or no difference in the biological potency of human insulin and animal insulin. This does not appear to apply however to the subcutaneously injected route where clinical studies have shown a difference in absorption properties between human and animal insulin.

It has been suggested that the daily dose of insulin should be reduced by 10-25% when switching from animal insulin to human insulin. Furthermore it has also been suggested that it may be necessary to change the insulin dosage depending on activity. For example an office worker may have to reduce insulin dosage at weekends if undertaking exercise and children may have to alter insulin dosage when participating in summer sporting and camping activities. Furthermore it has also been suggested that an overlapping interaction exists between the metabolic activity of the insulin of the current injection and that of the previous day's injections. This unpredictable accumulation of insulin action can result in prolonged and severe hypoglycemia. (11)

Child Growth and Diabetes Mellitus

Prior to the discovery of insulin linear growth failure was common among children with IDDM. Today significant growth failure is unusual in pediatric diabetes clinical populations. However significant abnormalities in the hypothalamic pituitary-GH axis exists even in normally growing children with IDDM. What happens postpubertally when GH secretation declines is unknown but studies in adults with IDDM suggest that growth hormone levels may remain elevated and insulinlike growth factor levels may be reduced. (12)

Insulin Therapy and Glycemic Control in Pregnancy

Dietary therapy is the keystone of diabetes management. In 1922 Priscilla White, an American physician, devoted her time to the care of children and women with diabetes of juvenile onset. Before the availability of insulin good treatment of diabetes was important with patients seen weekly by both an obstetrician and physician. With the advent of insulin patients were given enough insulin to prevent glycosuria and ketosis with a basic dose of long acting insulin before breakfast supplemented by 3-4 doses of short acting regular insulin before meals and evening snack. Patients were taught to weigh out amounts of food based on a diet plan of 30 kcal/kg body weight. The average 2 hour postprandial blood glucose was found to be 8.3 mM. Close attention to detail resulted in low frequencies of diabetic ketoacidosis in 2% of pregnancies and hypoglycemic coma resulted in 1% with no fetal deaths.

The most important limiting factor in intensified insulin therapy of diabetic women is the maternal CNS danger from hypoglycemic coma which appears to come on more quickly and often without warning signs during gestation. In pregnancy the fetal-placental unit continues to consume glucose and alanine in post absorptive periods, and exogenous insulin may limit alternative fuel sources by restraining lipolysis. Fortunately the fetus seems to be protected from maternal hypoglycemia.

A challenge for the future is to determine the level of diabetic control necessary to produce normally developing offspring, assuming congenital malformations can be prevented. Older controversial data on childhood intellectual function requires renewed attention. In 1969 it was reported by Farquhar of Edinburgh that follow up of 210 children of insulin treated diabetic women 5 children had educational subnormality; 20 had congenital abnormalities and 10% were less than the third percentile for height and greater than 20% had excessive weight by adolescence. (13)