The purpose of this study is to determine whether or not bedtime snacks are necessary to prevent overnight low blood glucose reactions (nocturnal hypoglycemia) in adults with type 1 diabetes who are using intensive insulin therapies, either multiple injections of insulin or insulin pump.

We hypothesize that a bedtime snack is not necessary if the bedtime insulin is very well adjusted and delivered, especially by use of insulin pump which is considered the "gold standard" for overnight blood glucose control.

Nocturnal hypoglycemia is a significant side effect of insulin replacement therapy and can lead to many problems including coma and death in the most extreme cases and deterioration of glycemic control. Strategies to prevent nocturnal hypoglycemia include development of newer insulins and delivery systems and the use of specific bedtime snack compositions. Detection of nocturnal hypoglycemia, as it occurs in everyday life, has been a problem since most episodes go unnoticed and therefore untreated leading to a very dangerous situation. With the recent availability of technology which allows for continuous glucose monitoring , it has been made possible to detect nocturnal hypoglycemia as it occurs in daily life. As well, this technology allows for the possibility of better titration of therapy.

Therefore, the goal of our study was to use continuous glucose monitoring technology as a tool to optimize titration of insulin therapy and as a means to detect nocturnal hypoglycemia as it occurs in daily life. Specifically, we sought to compare the use of insulin pump therapy with no bedtime snacks and using a new type of insulin called an analogue(lispro) versus multiple daily injections of insulin, using the same type of insulin analogue, and with and without bedtime snacks. Therefore, we compared 3 groups: 1) insulin pump with no bedtime snacks; 2) multiple injections with bedtime snacks and 3) multiple injections with no bedtime snacks. The groups were primarily compared for the occurrence and frequency of nocturnal hypoglycemia. Other comparisons included: long term glycemic control (A1c), weight and body composition and quality of life.

Inclusion Criteria:

• Adults with type 1 diabetes for more than 3 years duration and on intensive insulin therapy using multiple daily injections of insulin (MDI) for at least 1 year prior to study.

Exclusion Criteria:

• Pregnant women or intending to become pregnant in next year,
• Significant hepatic or renal disease,
• Active cancer,
• Abuse of alcohol or drugs.