The dissertation work is devoted to determining the influence of glycemic variability (GV) on heart rate and blood pressure (BP) in patients with diabetes mellitus (DM) type 1 and 2 and the study of heart rate variability (HRV) dynamics on the background of carbohydrate metabolism correction based on analysis of data, obtained by simultaneous registration of electrocardiogram, BP monitoring and continuous glucose monitoring (CGM). According to a survey of 125 patients with DM, including 72 and 53 patients type 2 DM patients with optimal and insufficient glycemic control and its comparison with the relevant control groups, it was found that patients with type 1 and 2 type DM had a significant decrease in HRV with sympathetic predominance. We determined that in patients with type 1 DM, the indicators of GV influenced HRV and by 45.50 % stipulated the latter, with the greatest influence on total power (TP) and power in high frequency range (HF) for the 24-hours, daytime and nighttime periods, R=0, 89 (χ2=280.83; p=0.002). In patients with type 2 DM, the glycemic fluctuations, namely glycemic range (GR) and standard deviation from mean glycemia (SD), stipulated HRV by 40.11 % and most significantly influenced the characteristics of standard deviation of all of the RR intervals (SDNN) and the ratio of low frequency power to high frequency power (LF/HF), R=0, 97 (χ2=198.92; p<0.001). The moment of the hypoglycemic episode is accompanied by significant QTc lengthening, an increase in the ratio of LF to HF power (LF/HF), and a decrease in HRV frequency characteristics. In patients with type 1 DM LF/HF ratio increased by 2 times, and QTc duration increased by 5.71 %. In patients with type 2 DM LF/HF ratio increased by 3.3 times, and QTc duration increased by 11.06 %. Disturbances of BP circadian were observed in 56.94 % of patients with type 1 DM and 80.00 % of patients with type 2 DM with predominance of pathological profiles of “non-dipper” and “night-peaker” types. In patients with type 1 DM, GV indicators had an effect on ambulatory blood pressure monitoring (ABPM) indices and determined them by 50.52 % with the greatest influence on the diastolic pressure area index (DPAI24) and the diastolic pressure time index (DPTI24), R=0.97 (χ2=519.64; p<0.001). In patients with type 2 DM, glycemic fluctuations stipulated ABPM indicators by 60.55 %, with the greatest influence on the diurnal characteristics of systolic pressure area index (SPAI24) and systolic pressure time index (SPTI24), R=0.94 (χ2=254.89; p<0.001). A strong inverse correlation between the level of glycated hemoglobin (HbA1c) and the incidence of hypoglycemia in the sample of patients with type 1 DM (rs= -0.83; p<0.001) was determined. In patients with type 2 DM, hypoglycemia was associated with the presence of genetically engineered insulins and sulphonylureas in the scheme of sugar-lowering therapy (88.89 % and 56.25 % of all patients treated with the appropriate group of drugs). The dynamics of HRV 6 months after modification of the sugar-lowering therapy was positive in both samples of patients with diabetes, namely: the decrease in relative hypersympaticotonia occurred in both samples: in patients with type 1 DM daily LF/HF decreased by 25.7 %, in patients with type 2 DM, the decrease in daily LF/HF was 42.9 %. Changes in ABPM indices after 6 months were determined by a significant increase in the diurnal indices (DI) of systolic blood pressure (SBP) and diastolic blood pressure (DBP) in both samples, which increased the proportion of patients with a “dipper” profile. In patients with type 1 DM, the proportion of “dipper” increased from 44.62 % to 79.27 % in DI SBP and from 47.43 % to 74.39 % in DI DBP. In patients with type 2 DM – from 25.75 % to 37.74 % in DI SBP and from 35.84 % to 49.06 % in DI DBP, respectively. In both samples indices of pressure load significantly decreased. However, in patients with type 1 DM, a more significant decrease was related to the DBP indicators (DPTI24 and DPAI24). In patients with type 2 DM, a decrease in SBP load prevailed and a decrease in SBP and DBP variability was observed. In the course of the study, mathematical models of HRV prediction were developed on the basis of GV indices separately for type 1 and type 2 DM. In the study we developed mathematical models for predicting HRV after modification of antihyperglycemic therapy based on HRV separately for type 1 and 2 DM, with have good performance in ROC-analysis: AUC 0.739 (95.0% CI 0.622-0.836; p = 0.001) for type 1 DM and AUC 0.717 (95.0% CI 0.559-0.843; p = 0.001) for type 2 DM. The highest value for HRV prediction in patients with type 1 DM was found in hypoglycemia (sensitivity 90.0%, specificity 45.45%, AUC = 0.677) and high glucose fluctuations (SD) with sensitivity 90.0%, specificity 59.09%, AUC = 0.708. In contrast, only hypoglycemia is predictor of HRV is patients with type 2 DM, with sensitivity 60.0% and specificity 83.33%, AUC = 0.717.
