Background: Based on international evidence, current management of people with T1DM on intensive insulin therapy (IIT) use algorithms based on the meal carbohydrate content (MCC) to calculate the prandial insulin dose. Typically, these calculations do not take into account the protein or fat content of the meal. There is a lack of clinical advice for optimal management of high protein/fat meals due to a paucity of evidence regarding the impact of protein/fat on glycaemic control.

Objective: To determine the mean glucose excursion from fasting (measured by continuous glucose monitoring, CGMS) at each 30 minute interval over the 8 hour postprandial period for each test condition. Protein effects will be looked at in a separate parallel study in Australia.

Hypothesis: The fat content of a meal will cause a dose-response change in the postprandial glucose concentration in children with T1DM.

Research Design and Methods: Randomised cross-over study involving thirty patients. Inclusion criteria: T1DM >1 year, aged 8-18 years, with HbA1c <8% and BMI <91st centile, on intensive insulin therapy. Participants will be given a test meal on 6 consecutive nights in random order; 4 test meals varying in fat content, and one 20g carbohydrate test meal with zero fat given as control meal. A CGMS will be used to assess glucose responses at 5 minute intervals for 8 hours after test meal consumption. The relationship between the fat loads in the test meals and the mean change in postprandial glucose concentration will be analysed and described.

Conclusions: This study will determine whether fat causes dose dependent response in glucose concentrations leading to refining the guidelines and possible adjustment of insulin doses for the fat content of a meal.

1. Aims of the project To determine the postprandial glucose dose-response curves response to varying fat amounts by studying various parameters (glycaemic excursion, rate of glucose level increase, area under the curve, percent time in target glucose range, maximal glucose excursion, time to maximal glucose excursion and time until the glucose level returns to fasting concentration) provided by continuous glucose monitoring over a 6 day study period with a view to provide data to calculate an accurate insulin dosing schedule to account for varying dietary fat ingestion.

   Objective 1 To define the impact of varying quantities (dose) of fat on the post-prandial glucose concentrations i.e. determine the relationship between the fat load in the test meals and the mean change in postprandial glucose concentrations.

   Objective 2 To compare the impact of the varying fat quantities in the test meals each with a control carbohydrate snack (20g) without extra insulin.

2. Protocol Participants will be contacted daily for one week prior to the study to review their overnight and fasting glucose concentrations and if necessary, adjust their insulin doses. The study will be carried out during a week long period for each patient. Participants will be given test meals over 6 consecutive days in random order, with 5 test meals varying in lipid content and one carbohydrate (20g) test meal as comparator. A continuous glucose monitoring system (CGMS) will be inserted on the day of the first test meal and will be removed at the completion of the sixth night.

   To define the impact of lipid on the post-prandial glucose concentration, participants will consume a standardised meal at 18:00hours and receive a standard insulin bolus by injection or pump (using standard wave bolus if on insulin pump) based on the meal carbohydrate content. At 22:00 hours the participants will consume a lipid test meal dose (varying fat content of 2.5, 12.5, 25, 37.5, 50g) or carbohydrate snack 20g. The calculations of the fat doses have been based on the Pankowska algorithm [3] which recommends additional insulin for every 100 kcal fat or protein. There will be randomised order of test meals.

   Test meals will be formulated to be palatable, consumed within 5 minutes and to contain negligible carbohydrate, protein and fibre. A dietician will design the test meals in conjunction with the co-investigator in detail to ensure ease of formulation and practical administration.

   The lipid dose in test meals will be tested for acceptability and palatability to ensure adherence to protocol. The order of the test meal types/carbohydrate snacks will be random (computer generated randomisation). The comparator 20g carbohydrate snack has been included to assess the participant's typical glucose response to 20g of carbohydrate. No insulin will be given for the lipid or carbohydrate snacks. Participants will fast until breakfast.

3. Outcomes Glucose concentrations will be measured at the start of the test meal and at 5 minute intervals over the 8 hour postprandial period. Therefore, the data will consist of repeated measurements on the same individual within test meal loads as well as the same individuals across test meal loads.

The primary outcome of interest will be the Area Under the Curve (AUC) of glucose concentrations during the post prandial 8 hours. Linear regressions, within a generalised linear mixed model framework, will be used to test for a dose response in AUCs with increasing size of fat means.

Inclusion Criteria:

• Diagnosed with T1DM for more than 1 year
• HbA1c less than 8%
• BMI less than 91st centile
• on intensive insulin therapy

Exclusion Criteria:

-