Background: During and after exercise there is a twofold stimulation of glucose transporter type 4 in type 1 diabetes mellitus (T1DM) caused by an exogenous insulin injection and muscle contraction. In case of no insulin reduction before exercise and/or an additional carbohydrate supplement patients with T1DM would become hypoglycemic. There are no specific data dependent on individual exercise intensities and strategies of bolus insulin reduction. The aim of the present study is to avoid hypoglycemia during and after exercise and to calculate a critical time where blood glucose would decrease to 50% of baseline by a pre exercise standardized insulin regimen during standardized cycle ergometer exercises dependent on different modes, intensities and energy expenditures as well as counter regulatory hormones.

Methods/design: Participants consist of 7 male subjects aged 18 to 35 years with T1DM (fasting C-peptide (≤ 0.3 nmol/l), HbA1c ≤ 64 mmol/mol). After a four-weeks basal insulin adaptation of ultralong-acting basal insulin (Novo Nordisk) each subject will perform a maximal incremental cycle ergometer exercise test with a reduction of short time insulin of 40% to determine the individual aerobic/anaerobic threshold called first/second lactate turn point (LTP1, LTP2). Four constant load and three high intensity intermittent cycle ergometer exercise tests (30min) will be performed at 5% Pmax below and above LTP1 and LTP2 (at 5% above LTP2 no high intensity intermittent cycle ergometer exercise test) with a reduction of short time insulin doses (standardized breakfast-four hours before each test) of 25% at 5% Pmax below LTP1 and 50% at 5% Pmax above LTP1 and 75% at 5% Pmax below and above LTP2 with one week interception between testing. 24h before testing subjects will be fitted with a continuous glucose monitoring system for at least 48h. Primary outcomes include blood glucose decrease from baseline during constant load tests, energy expenditure during testing, 24h CGMS measurements and hormones catecholamines, cortisol, glucagon and insulin growth factor 1. Secondary outcome measurements include heart rate, lactate and respiratory gas exchange values.

Discussion: If the results confirm the hypothesis, this study could be one of the first recommendations for T1DM patients, how to adapt bolus insulin dose individually before defined exercises without the risk hypoglycemia during and after exercise.

1.1. Pathophysiology of type 1 diabetes mellitus

Type 1 diabetes mellitus (T1DM) is an metabolic disease with insulin deficiency and a dysfunctional release of counter regulatory hormone glucagon. In the pathogenesis of T1DM autoantibodies to insulin (IAA), Glutamic Acid Decarboxylase Antibody (GADA) , islet antigen-2 (IA-2A) and zinc transporter 8 (ZnT8A) can be detected. In absence of endogen produced insulin the translocating glucose transporters (GLUT) cannot be activated (excepted insulin insensitive GLUT-2 in liver, β-cells of pancreas, kidneys and small intestine and GLUT-4 while muscle contraction) (Jensen & Richter 2012). The GLUT translocate to the cell surface and import glucose into the cell for metabolism or glycogen storage by glucose-6-phosphatase (G-6-P). For that reason in patients with T1DM exogenous insulin must be injected by intensified insulin therapy or insulin pump therapy. This exogenous insulin docks on the insulin receptor, the insulin receptor substrate (IRS) reacts with G-proteins (guanine nucleotide-binding proteins) on the phosphatidylinositide-3 kinase (PI-3 kinase) and activates GLUT. Furthermore without insulin there is a deficient carbohydrate metabolism while glycogen synthesis, glycogenolysis, gluconeogenesis and glycolysis.

Dependent on dysfunctional insulin release there is also failure in counter regulatory glucagon reaction. In T1DM is no decrease of insulin release while hypoglycemia and thus no increase of α-cell glucagon secretion. So in T1DM there are less endogenous opportunities to avoid hypoglycemia.

1.2. Physiology of type 1 diabetes mellitus in sports

It's already known that in T1DM and non type 1 diabetes mellitus (nT1DM) during and after physical activity the glucose utilization increases. During exercise glucose degradation rate in blood plasma increases by enhanced translocating GLUT-4 through skeletal muscles contractions caused by an increase of calcium (Ca2+), adenosine triphosphate (ATP) turnover, AMP-activated protein kinase (AMPK), founding member of a family of proteins sharing a 180- to 200-amino acid (TBC1D1/4), endothelial nitric oxide synthase (eNOS), p38 mitogen-activated protein kinases (p38 MAPK) and sucrose nonfermenting AMPK-related kinase (SNARK).

In place of endogenous insulin in nT1DM and exogenous insulin in T1DM muscle contractions activate the GLUT-4 during exercise and as a result glucose consumption increases. In contrast to T1DM patients in healthy subjects endogenous insulin secretion decreases during exercise according to mode, intensity and duration. Therefore in T1DM is a twofold GLUT-4 increase and a high risk of hypoglycemia during exercise.

Furthermore the risk of hypoglycemia exists also post exercise. Several studies have shown that one bout of exercise increases insulin sensitivity for hours which often causes difficulties in insulin regimen.

1.3. Adaptation of carbohydrate intake and insulin injection before an exercise in type 1 diabetes mellitus

Based on these physiological aspects guidelines describe how to reduce the risk of hypoglycemia during and after exercise. However, these recommendations are usually very general and unspecific. Several studies have shown, that the regular insulin dose before an exercise leads to hypoglycemia due to the above mentioned mechanisms. Therefore patients with T1DM have to increase carbohydrate intake and/or to reduce insulin injection dose before and after performing exercise The mode of exercise is an influencing factor, which impacts blood glucose levels during and especially after exercise. Some studies demonstrated, that high intensity intermittent exercises lead to a lower decrease in blood glucose level than constant load exercise However, there is a lack in the guidelines to which amount to reduce insulin or increase carbohydrate intake related to different exercise intensities and modes.

2. Pilot study

Objective: To perform sports and exercise in patients with type I diabetes they need to reduce insulin dose or increase carbohydrate intake to avoid hypoglycemia. There is a lack in the guidelines how to reduce insulin or increase carbohydrate intake related to different exercise intensities and modes. The purpose of the present study was to investigate the relationship between the blood glucose concentration and the intensity and the duration of the work load applying a standardized pre exercise glucose and insulin regimen. As a hypothesis we expect an energy expenditure and intensity dependent decrease of blood glucose concentration which offers the opportunity to calculate a critical time.

Research Design and Methods: One trained male subject with type I diabetes (age: 25 years; weight: 72 kg; high: 1.8 m; maximal oxygen uptake (VO2max): 55.4 ml.kg-1.min-1; insulin: NovoRapid/Levemir; C-peptide positive, HbA1c 48 mmol.mol-1) performed 4 hours after the last insulin/carbohydrate supplement, with a reduction of short time insulin to 40%, a maximal incremental cycle ergometer exercise test (40 W start; 20 W.min-1 increments) to determine the first (LTP1) and the second lactate turn point (LTP2) by means of computer based linear regression break point analysis. Three phases of energy supply can be detected and separated by LTP1 and LTP2. The LTP1 is characterized as the first increase in blood lactate concentration above baseline. The LTP2 is detected as the second abrupt increase in lactate between LTP1 and Pmax which defines the highest constant workload to give still a lactate steady state. Four constant load ergometer exercise tests (30 min) were performed at 5% Pmax below and above LTP1 and LTP2 with a reduction of short time insulin doses of 25% at 5% Pmax below and above LTP1 and 50% at 5% Pmax below LTP2 and 75% at 5% Pmax above LTP2 according to the literature. Heart rate and gas exchange variables were determined continuously, blood lactate concentration (La) and blood glucose concentration were determined at rest, at the end of every workload step, every 5 min in constant load ergometer exercise as well as during 3 and 6 min of active and passive recovery.

Results: Linear declines of blood glucose were found in all tests. At 5% < LTP1 glucose decreased from 191 mg/dl to 149 mg/dl, 5% > LTP1 from 131 md/dl to 96 mg/dl, 5% < LTP2 from 169 mg/dl to 91 mg/dl and 5% > LTP2 from 187 mg/dl to 144 mg/dl (early stop because of work load acidosis above LTP2). The decline of the blood glucose concentrations were calculated by a linear interpolation to 50% rest of the baseline value without any other supply of carbohydrates. time critical: 5 % < LTP1 - 77 min; 5 % > LTP1 - 31 min; 5% < LTP2 - 31 min; 5% > LTP2 - 29 min.

Conclusion: The results show a linear reduction of glucose concentration in relation to the intensity and duration of the work load. We suggest that it is possibly to calculate a critical time for a certain glucose threshold in type I diabetes patients to avoid hypoglycemia during sports and exercise.

3. Study goal and hypothesis

We do know that patients with T1DM must adapt insulin and carbohydrate intake before exercise, but we don't know the specific individual amount. Therefore we have shown in our pilot study a model how to adapt insulin injection dose (while constant basal insulin use) with respect to standardized intensities and how to calculate a critical time where glucose level in T1DM patients with insulin therapy would fall into hypoglycemia during sports and exercise. Now we have to verify those data with a group of subjects and additional hormonal markers while using a ultralong-acting basal insulin (Novo Nordisk).

The aim of the present study is to calculate a critical time by a pre exercise standardized insulin regimen during standardized cycle ergometer exercises dependent on different modes, intensity, duration and energy expenditure as well as counter regulatory hormones. As a hypothesis we expect no hypoglycemia during the cycle ergometer exercises dependent on an standardized pre exercise standardized insulin regimen in relation to the intensity, duration and the energy expenditure of the exercise and no post exercise hypoglycemia.

If the results confirm the hypothesis, this study could be the first recommendation for T1DM patients, how to individually adapt insulin dose before defined exercises without hypoglycemia during and after exercise.

4. Methods

4.1. General study design

This project is a proof of concept study. After the recruitment by the Medical University of Graz (Division of Endocrinology and Metabolism), at the first screening visit control of the inclusion and exclusion criteria will be done and subjects will give their signed informed consent. Before testing a 4-week adaptation of ultralong-acting basal insulin (Novo Nordisk) will be done. Each subject will perform a maximal incremental cycle ergometer exercise test to determine exercise markers to prescribe constant load and intermittent exercise tests. Four constant load and three intermittent ergometer exercise tests (30 min) will be performed at 5% Pmax below and above LTP1 and LTP2 with a reduction of short time insulin doses (breakfast-four hours before each test) of 25% at 5% Pmax below LTP1 and 50% at 5% Pmax above LTP1 and 75% at 5% Pmax below and above LTP2. 24 h before testing subjects will be fitted with a continuous glucose monitoring system (CGMS) for at least 48 h. At 5% Pmax above LTP2 there will be no intermittent ergometer exercise test. The same reduction of bolus insulin before testing while be done 15min after testing.

4.2. Subjects recruitment

Inclusion criteria:

• Subjects must give their signed and dated informed consent before any trial-related activities. Trial-related activities are any procedure that would not have been performed during normal management of the subject
• Male subjects with type 1 diabetes with duration ≥12 month
• Age ≥ 18 to ≤ 35 years, both inclusive
• HbA1c ≤ 64 mmol/mol
• Fasting C-peptide (≤ 0.3 nmol/l)
• Treatment with intensified insulin therapy or insulin pump therapy
• No diabetic long term complications
• No other physical and/or mental disease

Exclusion criteria:

• Previous participation in this trial
• History of any illness or disease that, in the opinion of the Investigator might confound the results of the trial
• Use of drugs, which may interfere with the interpretation of trial results or are known to cause clinically relevant interference with insulin action, glucose utilisation, or recovery from hypoglycemia
• Current addiction to alcohol or substances of abuse as determined by the investigator
• Known or suspected allergy to trial products or related products
• Any condition that the investigator feels would interfere with the trial participation or evaluation of data

Testing day inclusion criteria:

• 48h before testing no hypoglycemia
• 24h before testing no alcohol

Testing day exclusion criteria:

• Illness on and/or before testing day
• Low glucose level immediately before testing (< 80 mg/dl)
• Defect CGMS
• Incorrect time of bolus insulin injection (4 hours before testing)
• Mental incapacity, unwillingness, or language barriers precluding adequate understanding or co-operation
• Incorrect amount of bolus insulin injection (last before testing)

4.3. Screening visit

Subjects will receive a subject's number in ascending order. The following will be assessed and recorded in the case report form (CRF):

• Subjects get informed of study design and will give informed consent
• Assessment of inclusion and exclusion criteria
• Demography
• Abuse of drugs, alcohol and smoking habits
• Diagnosis of diabetes
• Body measurements: height (m), weight (kg) and body mass index (BMI)
• Determination of total daily dose (TDD) (through diabetes diary. 4.4. Adaptation of ultralong-acting basal insulin (Novo Nordisk) Insulin is indicated to be administered once-daily subcutaneous at the same time every day. In T1DM ultralong-acting basal insulin (Novo Nordisk)have to be combined with bolus insulin to cover mealtime requirements, so there will be no adaptation in bolus insulin amount. The dosage of insulin the ultralong-acting basal insulin (Novo Nordisk) will be adjusted every three days individually, with a starting dosage of 70% of the TDD of insulin.

4.5. Calculation of carbohydrate factor and correction factor

To determine the amount of carbohydrate exchanges, which increases blood glucose level, we will calculate the carbohydrate factor (CarbF) and the glucose correction factors (CorrF).

The carbohydrate factor indicates how many grams of carbohydrates 1 unit of insulin covers and the correction factor measures how far glucose concentration will fall per 1 unit of insulin.

4.6. Install of CGMS and edition carbohydrate supplement

Subjects will be fitted with the CGMS 24h before exercise testing. A hypoglycemic alarm function will be set at 80 mg/dl to avoid hypoglycemia below baseline (70 mg/dl). Standard glucose measurements will be done in a usual way to calibrate the CGMS. Also carbohydrate supplement (Fortimel complete) will be handed out for the standardized breakfast with the individual amount calculated for each test.

4.7. Maximal incremental cycle ergometer exercise test

Four hours after the last insulin/carbohydrate supplement with a reduction of short time insulin to 40% and constant basal insulin (calculated amount of carbohydrate intake and dose of insulin injection of the CarbF and CorrF), subjects will perform a maximal incremental cycle ergometer exercise test (40 W start; 20 W.min-1 increments) to determine the first (LTP1) and the second lactate turn point (LTP2). Three phases of energy supply can be detected and separated by LTP1 and LTP2. The LTP1 is characterized as the first increase in blood lactate concentration above baseline. The LTP2 is detected as the second abrupt increase in lactate between LTP1 and Pmax which defines the highest constant workload to give still a lactate steady state. Heart rate and gas exchange variables will be determined continuously, blood lactate concentration and blood glucose concentration will be determined at rest at the end of every workload step, every 5 min in constant load ergometer exercise as well as during 3 and 6 min of active and passive recovery. We decided to choose intensities for the constant load tests and the high intensity intermittent exercise tests related to defined phases of energy supply representing common activities. 5% < LTP1 is like daily activity at low intensity. 5% > LTP1 is like fast walking at low to moderate intensity. 5% < LTP2 is equivalent to moderate to high intensity activity or sports below the maximal lactate steady state. 5% > LTP2 is a high intensity activity with no more lactate steady state. To simulate sports games we decided to do those high intensity intermittent exercise tests, which are very similar to this type of physical activity.

4.8. Constant load test

In all constant load tests heart rate and gas exchange variables will be determined continuously, blood lactate concentration and blood glucose concentration will be determined at rest, at the end of every workload step, every 5 min in constant load ergometer exercise as well as during 3 and 6 min of active and passive recovery. Gas exchange variables will be processed to calculate the energy expenditure and the distribution of glucose and fat metabolism (glucose and fat g/min). Furthermore, at the start, in the middle and at the end of exercise venous will be drawn to determine adrenaline, noradrenaline, cortisol, glucagon and somatropin. Every 10 minutes blood glucose will be measured additionally from fingertip blood samples to determine the actual blood glucose level to provide hypoglycemia during the exercises and to calibrate CGMS. After the tests glucose concentration will be measured autonomously four times every full hour to avoid post exercise hypoglycemia also by fingertip measures.

The constant load ergometer exercises tests will be performed for 30 min at 5% Pmax below and above LTP1 and for 30min at 5% Pmax below and above LTP2 with a standardized reduction of short time insulin and constant basal insulin (calculated amount of carbohydrate intake and dose of insulin injection of the CarbF and CorrF).

4.9. High intensity intermittent exercise tests

In all high intensity intermittent exercise tests heart rate and gas exchange variables will be determined continuously, blood lactate concentration and blood glucose concentration will be determined at rest, at the end of every workload step, every 5 min in constant load ergometer exercise as well as during 3 and 6 min of active and passive recovery. Gas exchange variables will be processed to calculate the energy expenditure and the distribution of glucose and fat metabolism (glucose and fat g/min). Furthermore at the start, in the middle and at the end of exercise adrenaline, noradrenaline, cortisol, glucagon and somatropin will be determined from venous blood samples. Every 10 minutes, blood glucose will be measured additionally from fingertip blood samples to determine the actual blood glucose level to provide hypoglycemia during the exercises and to calibrate CGMS. After the tests glucose concentration will be measured autonomously four times every full hour to avoid post exercise hypoglycemia.

The high intermittent exercise tests will be performed for 30 min at 5% Pmax below and above LTP1 and for 30min at 5% Pmax below LTP2 with a standardized reduction of short time insulin and constant basal insulin (calculated amount of carbohydrate intake and dose of insulin injection of the CarbF and CorrF). The high intermittent exercise tests will be performed with the same mean work load like in the constant load exercise tests which can be calculated by the following equation:

Intervals will be set at Pmax from 20 sec interspersed by active recovery at 80 % power of LTP1 whereas the recovery time will be calculated.

Measurements 4.9.1. Anthropometric data and TDD

At the screening visit we will determine body measurements: height (m), weight (kg) and body mass index (BMI) and TDD (U).

4.9.2. Blood glucose, gas exchange, heart rate variables and lactate

In all tests blood glucose and lactate will be measured form blood samples from the earlobe and heart rate and gas exchange variables will be determined continuously. Blood lactate concentration and blood glucose concentration will be determined at rest, at the end of every workload step, every 5 min in constant load ergometer exercise as well as during 3 and 6 min of active and passive recovery. Lactate and glucose concentration will be determined by system EKFDiagnostic, GER). Heart rate will be determined through a 12-lead ECG (ZAN, AUT) and also by Polar (Polar Electro, FIN). Gas exchange variables will be determined by an open spiro-ergometry-system (ZAN, AUT). LTP1 and LTP2 will be assessed by means of computer based linear regression break point analysis Pro Sport (AUT). Additional fingertip glucose measures will be performed by the subjects own glucose measurement device.

4.9.3. Continuous glucose monitoring system (CGMS)

CGMS sensor will be inserted into subcutaneous tissue (postural-lateral abdominal region) 24h prior to each test for 48h, to get a permanent glucose recording pre, during and post exercise and to minimize the risk of hypoglycemia. Blood glucose readings will be stored in the memory of the monitor, which is connected to the sensor. Glucose data will be processed computer based 4.9.4. Hormones

Adrenaline, noradrenaline, cortisol, glucagon and somatropin will be determined from venous blood samples obtained from a cubital vene at the start, in the middle and at the end of each exercise test. Adrenaline and noradrenaline will be quantified by DRGDiagnostics, USA and glucagon by ICNDiagnostics, USA. Somatropin will be measured using CLIASiemens Healthcare Diagnostics, USA and cortisol by CENTAUR Siemens Healthcare Diagnostics, USA).

5. Statistics

A repeated-measures ANOVA design (number of repetitions = 4) was calculated a priori with a medium effect size of 0.5 and an alpha-error of 0.05 based on pilot data. Correlations between repetitions were assumed to equal 0.5. With a sample size of 6 patients the achieved power (beta-1) is greater than 0.95 and is therefore appropriate for a high risk study. In case of dropout we decided to do our study with 7 subjects.

Results will be assessed by analysis of variance (ANOVA) for repeated measures with the paired or unpaired t test, Wilcoxon's rank.-sum test for paired data, or Friedman's repeated measures by ANOVA on ranks when applicable (GraphPad Prism Software version 4.0, USA).

Inclusion Criteria:

• Subjects must give their signed and dated informed consent before any trial-related activities. Trial-related activities are any procedure that would not have been performed during normal management of the subject
• Male subjects with T1DM with duration ≥12 month
• Age ≥ 18 to ≤ 35 years, both inclusive
• HbA1c ≤ 64 mmol/mol
• Fasting C-peptide ≤ 0.3 nmol/l
• Treatment with intensified insulin therapy or insulin pump therapy
• No diabetic long term complications
• No other physical and/or mental disease

Exclusion Criteria:

• Previous participation in this trial
• History of any illness or disease that, in the opinion of the Investigator might confound the results of the trial
• Use of drugs, which may interfere with the interpretation of trial results or are known to cause clinically relevant interference with insulin action, glucose utilization, or recovery from hypoglycemia
• Current addiction to alcohol or substances of abuse as determined by the investigator
• Known or suspected allergy to trial products or related products
• Any condition that the investigator feels would interfere with the trial participation or evaluation of data