Carbohydrate metabolism disorders in patients with heart failure: data from the local registry

Aim. To study the prevalence of carbohydrate metabolism disorders in patients with heart failure (HF) hospitalized in the city HF center.Material and methods. According to the local registry, the study sequentially included 183 patients (99 men and 84 women) hospitalized in the Nizhny Novgorod city HF center from September 1, 2019. The examination and treatment were carried out in accordance with the current clinical guidelines. In the first 48 hours after hospitalization, the concentration of the N-terminal pro-brain natriuretic peptide, soluble stimulating growth factor 2 (sST2), neutrophil gelatinase-associated lipocalin, cystatin C, blood creatinine was determined. The glomerular filtration rate was calculated using the СKDEPI equation. To assess the carbohydrate metabolism disorders, all patients were studied for fasting plasma glucose, glycated hemoglobin (HbA1c) and fructosamine. Statistical data processing was carried out using the R statistics package (R Core Team (2019)).Results. The incidence of carbohydrate metabolism disorders among patients withdecompensated HF was 75,89%, including previously diagnosed type 2 diabetes in 31,25%, newly diagnosed dysglycemia in 44,64% of patients. Less than one fourth of patients had normal parameters of carbohydrate metabolism according to HbA1c, fructosamine and fasting plasma glucose. The severity of carbohydrate metabolism disorders was significantly correlated with the severity of HF according to the following criteria: 6-minute walk test, HF functional class, sST2 level, and some parameters of cardiac remodeling. Among the criteria used for carbohydrate metabolism disorders, the HbA1c level was most closely associated with the criteria for HF severity.Conclusion. Carbohydrate metabolism disorders in HF patients are widespread and underdiagnosed during routine examination. The interrelation of carbohydrate metabolism parameters and indicators of HF severity is rationale for active detection of dysglycemia in these patients in order to potentially influence the prognosis.

Statistical data processing was carried out using the statistical package R [5]. To assess the normal distribution of a quantitative trait, the Shapiro-Wilk test was used, as well as visual assessment of the distribution shape. Descriptive statistics for quantitative features are presented as a median (1st quartile; 3rd quartile), and for nominative features -as a percentage. In assessing the statistical significance level of differences in subgroups, the Mann-Whitney U test was used, and the χ 2 test or the exact Fisher test for small subgroup sizes was used to analyze the frequency differences. In the case of multiple comparisons, the Beniamini-Hochberg multiple comparison correction was applied. Linear regression methods are used in the construction of regression models. The critical level of null hypotheses significance was assumed to be p<0,05.

Results
The prevalence of dysglycemia in the examined cohort using the criteria HbA 1c , fructosamine and fasting plasma glucose was 75,89%, including 31,25% of patients with previously diagnosed T2DM and 50 patients (44,64%) with first-time dysglycemia. Only 24,11% of patients had normal indicators of carbohydrate metabolism. Among patients with newly diagnosed dysglycemia, only one indicator was deviated from the norm in 26 of 112 patients (23,21%). In 24 (21,43%), 2 indicators were deviated from the norm.
For further analysis, the patients were divided into the following 2 groups: without carbohydrate metabolism disorders and with dysglycemia, including patients with a previously established diagnosis of T2DM. The patients were divided into groups with and without dysglycemia depending on the level of glycolized hemoglobin (Table 1), fructosamine (Table 2), fasting plasma glucose (Table 3).
When divided by HbA 1c (Table 1), the obtained data analysis revealed that patients with dysglycemia are younger than patients without carbohydrate metabolism disorders. Patients with carbohydrate metabolism disorders compared with patients with normoglycemia were statistically significantly more likely to belong to the III-IV functional class (FC) of CHF, less often had II FC of CHF, and had lower 6MWT distance indicators. The main The relevance of studying the problem of carbohydrate metabolism disorders in patients with chronic heart failure (CHF) derived from high prevalence of type 2 diabetes mellitus (T2DM) and prediabetes in patients with CHF, common pathogenesis mechanisms and mutual negative impact on the quality of life and prognosis of patients. The number of patients with T2DM and CHF increases annually both in the Russian Federation and worldwide [1].
Goal: to study the prevalence of carbohydrate metabolism disorders in patients with CHF hospitalized in the city HF center (the State Bud getary Institution of Healthcare of Nizhny Novgorod Region of the City Clinical Hospital No. 38 of the city of Nizhny Novgorod), as well as the relationship between indicators of glycemic status and CHF severity.

Material and methods
The study was carried out in accordance with the Good Clinical Practice standards and the principles of the Helsinki Declaration. The study protocol was approved by the Ethics Committee of FSBEI HE PRMU of the Ministry of Health of the Russian Federation. All participants received written informed consent before enrollment.
The local registry included 183 consecutive patients with CHF of any etiology aged 18 years and older (99 men and 84 women). All patients were treated for CHF decompensation in the Nizhny Novgorod city HF center. The patients were examined and treated in accordance with the existing clinical practice guidelines [2].
Patients underwent echocardiography (EchoCG) on the Vivid3 device (Austria, 2007) by transthoracic method according to the standard protocol with a single-crystal phased sensor SP3-8. In the first 48 hours after hospitalization, the concentration of the N-terminal fragment of the brain natriuretic peptide precursor (NT-proBNP), soluble stimulating growth factor expressed by gene 2 "soluble suppression of tumorigenicity-2" (sST2), neutrophil gelatinaseassociated lipocalin (NGal), cystatin C, and blood creatinine was determined, and the glomerular filtration rate was calculated using the formula CKD-EPI. All patients were examined for fasting plasma glucose, glycated hemoglobin (HbA 1c ), and fructosamine levels. Dysglycemia in this study was understood as T2DM and prediabetes [3]. Carbohydrate metabolism disorder (T2DM and prediabetes) was verified in accordance with the clinical recommendations "Algorithms of specialized medical care for patients with diabetes mellitus" [4]. The NT-proBNP concentration in blood serum was determined by an enzyme immunoassay Note: * -significance of differences between groups 1 and 2. body of patients in both groups were patients with preserved ejection fraction, but there was a tendency to higher values of ejection fraction in patients with normoglycemia. When analyzing the EchoCG parameters in patients with dysglycemia, there was a statistically significant increase in end-diastolic dimension of the left ventricle (EDDlv). Despite the absence of inter-group differences in NT-proBNP, patients with dysglycemia had a statistically sig nificantly increased sST2. When dividing patients by glycemic status based on the fructosamine level (Table 2), as in the division by HbA 1c , patients with dysglycemia had lower 6MWT values, more often referred to III-IV CHF FC, and the frequency of IV CHF FC was 2 times higher than in normoglycemia. Signs of pulmonary hypertension were found statistically significantly more often in the dysglycemia group and there was a tendency to increase in EDDlv.
The main clinical and laboratory-instrumental characteristics of patients with CHF, depending on glycemic status when divided by fasting plasma glucose level, are presented in Table 3. Patients with dysglycemia were statistically significantly younger, had a statistically significantly greater left ventricular dilatation according to the results of EDDlv and end-systolic dimension of the left ventricle. In the dysglycemia group, there was a tendency to increase Note: * -significance of differences between groups 3 and 4. the NT-proBNP level and a significant increase in the sST2 level, a highly sensitive C-reactive protein, and the NGal level.
The most significant differences in the cri teria for the severity of CHF, such as 6MWT, prevalence of FC III-IV, EchoCG criteria for LV dilatation (EDDlv), sST2 level, occurred when patients were divided into groups with and without carbohydrate metabolism disorders according to the HbA 1c level. Therefore, when analyzing the CHF etiological factors and the therapy conducted before hospitalization, we also took as a basis the division by the HbA 1c level (Table 4). In patients with carbohydrate metabolism disorders, the ischemic etiology of CHF was statistically significantly more frequent, and there was a tendency to increase the Table 3 Comparative Note: * -significance of differences between groups 5 and 6. frequency of stage 2-3a chronic kidney disease in the dysglycemia group. The relationship between glycemic status and CHF therapy was not revealed.

Discussion
The high prevalence of dysglycemia (75,89% of patients) in the examined cohort of patients with CHF is comparable both with the results of registers in which the T2DM prevalence is on average 27% compared to 31,25% in our study [6], and with the results of clinical trials in which the prevalence of dysglycemia reached 80% [7][8][9][10].
In our study, patients with dysglycemia were statistically significantly younger than patients without carbohydrate metabolism disorders. This does not align with the data on number of clinical trials in which patients with dysglycemia were older [7][8][9][10]. Nevertheless, the average age of our patients corresponds to data of international and national epidemiological studies, including EPOCHA-CHF [1,6].
In our study, patients with impaired carbohydrate metabolism were statistically significantly more likely to have an ischemic etiology of CHF. According to the literature, the ischemic etiology of CHF and the presence of dysglycemia are interrelated, although this may not represent a clear causeeffect mechanism, but rather reflect the general pathogenesis components. Thus, in the Swedish Heart Failure Registry, T2DM was more common in patients with CHD than in patients without it (30% vs 19%) [11]. Note: * -significance of differences between groups 1 and 2. Abbreviations: HD -hypertensive disease, DBP -diastolic blood pressure, DCM -dilated cardiomyopathy, CAD -coronary artery disease, MI -myocardial infarction, ACA -acute cerebrovascular accident, SBD -systolic blood pressure, DM -diabetes mellitus, GFR -glomerular filtration rate, HF -heart failure, TIA -transient ischemic attack, AF -atrial fibrillation, CKD -chronic kidney disease, COPD -chronic obstructive pulmonary disease, HR -heart rate, TG -thyroid gland.
The severity of carbohydrate metabolism di sorders in our study was statistically significantly correlated with the severity of CHF, which does not contradict the data of both heart failure registers and clinical trials that demonstrated that dysglycemia compared to normoglycemia is associated with an increased risk of general and cardiovascular mortality, and in a number of studies, the highest risk of death was observed in patients with newly diagnosed T2DM [7,8,10].
When analyzing our data, attention was drawn to a statistically significantly higher level of sST2 in patients with dysglycemia compared to patients without lipid metabolism disorders when divided by the HbA 1c level. At the same time, patients with and without dysglycemia did not significantly differ in the NT-proBNP level.
ST2 and NT-proBNP reflect the course of two different but overlapping biological processes, so the markers provide independent and complementary information. As markers of hemodynamic instability or cardiomyocyte stretching, NT-proBNP/BNP are more suitable for the identification of CHF, but are less important for prognosis. ST2 is the most powerful and clinically significant prognostic marker of cumulative cardiovascular events and mortality rate, the degree of sST2 increase does not depend on CHF etiology, as well as on age, gender, heart rate, body mass index, hemoglobin level, and the presence of atrial fibrillation [12].
In our study, among the criteria used for carbohydrate metabolism disorders, the HbA 1c level was most closely associated with the criteria for CHF severity.
Some recommendations emphasize that the use of fasting plasma glucose determination, a 2-hour glucose tolerance test, or HbA 1c level is equally appropriate [13]. A number of studies substantiate the predominant value of HbA 1c as more associated with cardiovascular risk [14].
The stronger association found between HbA 1c and emerging cardiovascular diseases may be explained by the ability of HbA 1c to reflect average gly cemia. Fructosamine values reflect shorter-term glycemic levels than HbA 1c -2-3 weeks. Fructosamine may be a tool of choice when it is necessary to assess glycemic control in patients with severe chronic kidney disease (stages 4 and 5) [15], anemia or hemoglobinopathy [16].
Study limitations. The results of this study should be interpreted in the context of several constraints. This follow-up includes only hospitalized patients with CHF. The patients did not undergo a glucose tolerance test, which could be the reason for underestimating the true prevalence of carbohydrate metabolism disorders.

Conclusion
Thus, according to the local registry, dysglycemia was observed in almost 3/4 of patients with CHF. The severity of carbohydrate metabolism disorders was statistically significantly correlated with CHF severity according to such criteria as 6MWT, CHF FC, sST2 level, and some parameters of heart remodeling. Among the criteria used for carbohydrate metabolism disorders, the HbA 1c level was most closely associated with the criteria for CHF severity. Patients with any CHF etiology need to clarify the carbohydrate metabolism status.
Relationships and Activities: none.