This study evaluates the effect of different doses of metformin on the function of endothelium in people with pre-diabetes. One group of the patients will receive metformin in dose: 1500 mg, the second one will receive 3000 mg/day.

The parameters from healthy volunteers will be taken only at the study beginning to compare the test results with the parameters from patients with pre-diabetes. This group will be not treated with metformin (no intervention)

In addition to exercise and diet, metformin is a medicine used to treat diabetes mellitus (DM) and pre-diabetic (pre-DM) conditions. This drug improves insulin sensitivity in both groups of patients (DM and pre-DM) and as a result gives a reduction in blood glucose.

However, studies also confirm the effect of metformin on the reduction of cardiovascular risk in patients with diabetes (UKPDStudy), regardless of the hypoglycaemic effect.

The precise mechanism of action of the drug in this field is not clear. There is also no data on whether similar effects apply to patients with pre-diabetes. Although we know that this group of patients is also characterized by increased cardiovascular risk.

One of the most important substances that are involved in vasodilation is nitric oxide (NO). Impairment of its secretion is an important signal of endothelial damage and is connected with cardiovascular complications.

The impact of metformin on endothelial function in pre-diabetes patients is not known. We do not know the effect of the drug dose and its different serum concentration on the secretion of the dilators of the vessels, which are associated with endothelial function.

The study involves patients with pre-diabetes who meet the inclusion criteria, have no contraindication to participate in the study (see exclusion criteria), and give their written consent after reading "Information for the patient".

The conditions (IFG, IGT) for participation in the study are confirmed by a diabetologist (principal investigator ) based on fasting glucose and OGTT (oral glucose tolerance test) with 75 mg of glucose.

Metformin will be given in an increasing dose in accordance with the test protocol.

After reaching a one-week treatment with a dose of 3 x 500 mg, patients will be assigned to group A -a continuation of a dose of 3 x 500 and group B- increase dose to 3 x 1000mg. Randomization depends on the identification number (ID). The patients with an even, second number in the PESEL (identification number) will be randomized to the A group, the patients with the second, odd number in the PESEL will be randomized to the group B. Here as an example: PESEL: 60010102823-" 0" is as an even number so the patient will be randomized to the group A; PESEL: 61010102823- "1" is an odd number- the patient will be randomized to the group B. This PESEL number (ID) in Poland is given to every person shortly after birth and the researchers have no influence on it.

In the final stage also patients from group B (metformin: 3 x 1000 mg), will back to the treatment dose of metformin 3 x 500 mg- to show the relationship between the dose, serum concentration of the metformin and its effect on the secretion of the measured substances.

The healthy volunteers will be not treated with metformin. Patients will be reminded by phone about the increase in metformin dose as well as on control visits.

The lack of possible treatment with proper doses of metformin due to poor drug tolerance will move the patient from group B to group A if such dose (3 x 500 mg) is well tolerated. The patient will be excluded from the study if no compliance or lack of contact with the patient will be recorded during the treatment period or if metformin dose: 3 x 500 mg will be characterized with bad tolerance.

Information about: age, gender, BMI, cardiovascular risk factors, the current basic lab-tests, and pharmacotherapy will be recorded.

The blood samples for: plasma metformin level and listed substances will be collected for patients with pre-diabetes as described below (see diagram). The basic parameters as well as NO indirect products concentrations will be assessed for healthy volunteers only once-at the beginning of the study.

Test 1: NO(0) (indirect products) for patients before treatment start

3 weeks increasing dose of metformin to the final dose: 1500mg/day

3 x 500 mg (1500/day)- the dose is reached

3 weeks treatment 1500mg/day

Tests 2: NO1(1500) and metformin concentration

Randomization B:3 weeks increasing dose A: 6 weeks continuation with of metformin to the final a dose 3 x 500 mg

dose 3 x 1000mg

3 x 1000mg (3000mg/day)- the dose is reached 3 weeks treatment 3000mg/day Tests 3: NO2(1000) or NO3(500) and metformin concentration for both groups

3 weeks treatment 1500mg/day for both groups

Test 4: NO4 and metformin concentration

Statistical analysis:

For statistical analysis, the Statistica 12 program will be used (StatSoft Polska Sp. z o.o. www.statsoft.pl).

The cut-off point for statistical significance (p) was determined at 0.05. To determine the statistical significance will be used tests compliant with the distribution of variables and data character (Student's t-test, Mann-Whitney test, chi-square test, Kruskal-Wallis ANOVA test). For analysis taking into account the duration of the study and determine the impact of relevant variables to achieve the appropriate concentration of nitric oxide will be used Cox proportional hazard regression. The goal of the optimal determination cut-off point for predictors will use ROC curves. Logistic regression was used to determine the independent predictors to obtain the desired concentration of nitric oxide. In order to determine the correlation, it will be used correlation of order Spearman's rank correlation coefficient or Pearson correlation coefficient.

Inclusion Criteria for treated groups ( A or B):

• age: 40-65 years;
• pre-diabetic status based on fasting plasma glucose (FPG) and / or OGTT;
• without metformin before;
• without ischemic heart disease in history;
• without a stroke in a history;
• without PAOD (peripheral arterial occlusive disease) in a history;
• without active cancer in a history

Exclusion Criteria for treated groups (A or B):

• age <40 or >65;
• diabetes;
• taking metformin before study;
• active cancer;
• history of macro-angiopathy (ischemic heart disease, stroke or TIA, PAOD);
• serious gastrointestinal disease that may affect metformin tolerance;
• renal failure with GFR<45 ml/min/1.73m2;
• alanin transaminase > 3 x ULN

Inclusion criteria for healthy volunteers:

• age: 40-65 years;
• no carbo-hydrates disturbances (based on fasting plasma glucose (FPG) and/or OGTT);
• without metformin before;
• without ischemic heart disease in history;
• without a stroke in a history;
• without PAOD (peripheral arterial occlusive disease) in a history;
• without active cancer in a history

De-identified individual participant data for all primary outcome will be made available.

the results will be available after the publication of the study

data will be available after giving permission by the Investigator

group C was a comparator only for baseline parameters

15 weeks

Other- important gastrointestinal side effects which did not allow the prescribed dose to be continued

We didn't differentiate pts according to sex or gender. Before the study was finished we assessed the correlation between metformin concentration and BMI/ body mass. It was prepared as an additional analysis based on 20 cases (not published yet).

At the beginning of the study, patients are assigned a number by a nurse and are assigned to group A or B according to ID. When analyzing the results, the researchers only knows the numbers of blood samples.