Abstract
To investigate the associations between glucose metabolism, left ventricular (LV) contractile reserve, and exercise capacity in patients with chronic systolic heart failure (HF).
From an outpatient HF clinic, 161 patients with systolic HF were included (mean age 70 ± 10 years, 69% male, 59% had ischaemic heart disease, mean LV ejection fraction (LVEF) 37 ± 9%). Thirty-four (21%) patients had known diabetes mellitus (DM). Oral glucose tolerance testing (OGTT) classified patients without a prior DM diagnosis as normal glucose tolerance (NGT), impaired glucose tolerance or new DM. All patients completed low-dose dobutamine echocardiography (LDDE) and 154 patients a 6-min walking distance test (6MWD). Compared with patients with NGT, patients with known DM had lower resting LVEF (33.4 vs. 39.1%, P < 0.05) and higher E/e′ (13.9 vs. 11.4, P < 0.05). During LDDE, an increase in LVEF could be observed in all glycemic groups (mean 8.2% absolute increase), but the contractile reserve was lower in patients with known DM (−5.4%, P = 0.001) and new DM (−3.5%, P = 0.035) compared to patients with NGT. 6MWD was lower in known DM (349 m) and new DM (379 m) compared with NGT (467 m) (P < 0.001). Differences in clinical variables, resting echocardiographic parameters or contractile reserve, did not explain the exercise intolerance related to diabetes.
Diabetes, known or newly detected by OGTT, is independently associated with reduced LV contractile reserve and exercise intolerance in outpatients with systolic HF. These findings may offer one explanation for the excess mortality related to diabetes in HF.
Introduction
Heart failure (HF) with left ventricular (LV) systolic dysfunction is a common disease with a poor prognosis. This is particularly true for patients with co-existent diabetes mellitus (DM), in whom the mortality rate is elevated 1.5–2 times.1–3 Additionally, diabetic HF patients appear more susceptible to exercise intolerance.4,5 We have previously shown that screening with an oral glucose tolerance test (OGTT) in an outpatient HF clinic reveals a substantial proportion with unrecognized DM associated with poor prognosis and a more severe HF phenotype.6 Also, abnormal glucose metabolism (AGM) below the diabetic threshold has been associated with increased mortality and exercise intolerance.7,8
The reason for increased mortality and exercise intolerance in HF patients with AGM is not clear and is apparently unrelated to the resting LV ejection fraction (LVEF). Diastolic dysfunction,9–11 reduced longitudinal myocardial tissue velocities,12,13 and reduced contractile reserve during dobutamine stress echocardiography14–16 have been demonstrated in AGM patients without HF. Since these markers of LV function are associated with mortality and exercise incapacity in HF patients,17–19 they may provide a pathophysiological link between AGM and HF that has not previously been thoroughly investigated.
Accordingly, we hypothesized that the severity of AGM is associated with LV diastolic dysfunction, reduced contractile reserve, and exercise intolerance in patients with chronic systolic HF.
Methods
Patient population
From May 2008 through June 2010, 268 patients referred to the outpatient HF clinic at Frederiksberg University Hospital were screened. Eligible patients had signs and symptoms of chronic systolic HF according to the definition by the European Society of Cardiology20 and an LVEF of 45% or less as established by echocardiography prior to referral. In total, 107 patients were excluded for the following reasons: unwillingness to participate (n = 39), HF diagnosis not new (n = 23), referral back to the general practitioner before inclusion (n = 11), compliance unlikely (n = 5), significant valvular heart disease (n = 9), paced heart rhythm (n = 5), mental or language barrier preventing informed consent (n = 9), or death before echocardiography (n = 2). Thus, 161 patients (60%) were enrolled.
The study was conducted in accordance with the Declaration of Helsinki, and approvals from the Regional Scientific Ethics Committee and the Danish National Board of Registration were obtained prior to the study start. All patients provided written informed consent.
Definitions and assays
Patients were categorized as known DM if the diagnosis was made prior to the referral to the HF clinic. Within 2 weeks from the baseline visit, an OGTT was conducted in patients without a prior DM diagnosis. A standardized procedure was performed using 75 g of glucose dissolved in 250 mL of water. Plasma glucose was measured in the fasting state (fasting plasma glucose, FPG), and 2 h after glucose ingestion (2hPG). Glycemic categories were defined according to WHO definitions21: normal glucose tolerance (NGT): FPG < 7.0 mmol/L and 2hPG < 7.8 mmol/L, impaired glucose tolerance (IGT): FPG < 7.0 mmoL/L and 2hPG > 7.8 mmol/L and < 11.1 mmol/L, newly detected DM: FPG ≥ 7.0 and/or 2hPG ≥ 11.1 mmol/L. Patients with IGT, new DM, or known DM are collectively termed AGM throughout this paper.
Atrial fibrillation was diagnosed from the recorded rhythm at the time of echocardiography. Ischaemic heart disease (IHD) was diagnosed if any of the following characteristics were present: previous myocardial infarction, previous percutaneous coronary intervention or coronary artery bypass grafting, evidence of obstructive coronary artery disease on angiogram, or signs of ischaemia on myocardial perfusion imaging. All patients, where a coronary angiogram was not judged clinically indicated, were subjected to myocardial perfusion imaging to exclude IHD.
Medication including dosage of ACE inhibitors, angiotensin-receptor blockers, and beta blockers were recorded at baseline. The loop diuretic dose was calculated as daily furosemide dose or equivalent.
From the fasting blood samples, the following biochemical parameters were analysed: haemoglobin, creatinine, N-terminal pro Brain Natriuretic Peptide (NT-proBNP), haemoglobin A1c (HbA1c), and insulin. Insulin resistance was estimated by the Homeostasis Model of Assessment – Insulin Resistance (HOMA-IR).22
Echocardiography
Resting and low-dose dobutamine echocardiography (LDDE) were performed within 2 weeks after the baseline visit on a commercially available ultrasound machine (IE-33, Philips Inc.) and stored for later offline analysis blinded from the results of OGTT and 6-min walking distance (6MWD). For analysis of Doppler measurements, an average of 5 beats for patients in sinus rhythm and 10 beats in patients in atrial fibrillation were averaged. Echocardiography was performed on therapy, including the use of beta blockers.
LV volumes and LVEF were calculated by the biplane method of discs from apical four- and two-chamber views. An average of three consecutive beat measurements was used for patients with atrial fibrillation. Maximal left atrial volume was calculated by the biplane method of discs from the apical four- and two-chamber views adjusted to maximize left atrial size and divided by body surface area to obtain the left atrial volume index (LAVI).
Peak early (E) and late (A) diastolic transmitral flow velocities were recorded by pulsed-wave Doppler with a 2 mm sample volume placed between the tips of the mitral leaflets and aligned parallel with flow. Pulsed-wave tissue Doppler was used to measure systolic (s′), early diastolic (e′) and late diastolic (a′) longitudinal tissue velocities, and these were averaged from six corners of the mitral annulus in apical four-, two-chamber and long-axis views, and the E/e′ ratio calculated.
The grade of diastolic dysfunction was defined by LAVI, e′ (average of septal and lateral values), and E/e′ according to the most recent ASE/ESE recommendations.23
Cardiac output was estimated by multiplying the velocity–time integral (stroke length) from the pulsed-wave Doppler profile of the LV outflow tract in apical five-chamber or long-axis view with the LV outflow tract area and heart rate.
After acquisition of resting images, a standardized LDDE protocol was initialized. Dobutamine was infused through a peripheral vein at a rate of 5 µg/kg/min for 3min followed by a rate of 10 µg/kg/min for 3min after which the resting examination was repeated during continuous infusion. Blood pressure was measured prior to infusion, at 5 µg/kg/min, and at 10 µg/kg/min and the heart rate and rhythm were monitored by continuous ECG. Contractile reserve was defined as the absolute difference between LVEF at rest and during dobutamine stress, likewise regarding s′.
Six-minute walking test
Exercise capacity was measured by an indoor standardized 6-min walking test (6MWD). Seven (4%) patients were unable to perform 6MWD.
Pulmonary function
Forced expiratory volume during the first second was recorded by conventional spirometry on a commercially available spirometer (Microlab, Micro Medical Inc.). A variation <5% in FEV-1 from three measures was required for successful completion. Twenty-three (14%) patients did not complete a successful spirometry.
Statistical analysis
Continuous variables with a normal distribution are presented as mean (SD), and differences in these variables between the four glycemic categories (NGT, IGT, New DM, known DM) are compared using one-way ANOVA. A least significant difference test was used post hoc to compare individual AGM groups with the reference group (NGT). Skewed variables are presented as median (interquartile range) and compared by the independent Kruskal–Wallis test. Proportions are presented as percentages and compared by χ2 test. Within-group changes in echocardiographic parameters from rest to LDDE were assessed using the paired t-test.
Associations between glycemic categories and exercise capacity or echocardiographic parameters were evaluated by univariable and multivariable ANCOVA models. Correlations of continuous measures of HbA1c, FPG, 2hPG concentrations, and HOMA-IR with exercise capacity and LV function were evaluated by univariable and multivariable linear regression models. HOMA-IR was logarithmically transformed prior to entering models to satisfy the requirement for a normal distribution in ANOVA models and for equal variability in regression analyses. The most important confounders considered of potential relevance by the authors were corrected for in the multivariate models. Thus age, sex, IHD, atrial fibrillation, resting heart rate, and eGFR were included in all the multivariable models (forced entry). Echocardiographic outcome variables were additionally adjusted for systolic blood pressure, LV end-diastolic volume index, and resting LVEF (when LVEF contractile reserve, s′ contractile reserve, or LDDE E/e′ were the outcome variable). 6MWD was additionally adjusted for body mass index (BMI) and haemoglobin concentrations, and in a separate model echocardiographic variables (LDDE LVEF and E/e′) were added. A P-value of <0.05 was considered statistically significant. SPSS (IBM Corp.) version 20 was used for all analyses.
Results
Baseline demographics and clinical characteristics
The mean age was 70 ± 10 years, 69% were male, and 34 (21%) had a prior diagnosis of DM at the baseline visit. Based on the results of the OGTT performed in the 127 patients without known diabetes, 72 (45%) was classified as NGT, 27 (17%) as IGT and 28 (17%) as newly detected DM. Table 1 displays baseline characteristics according to glycemic categories. Patients with IGT and new DM were older than patients with NGT. No differences in the NYHA class distribution, presence of IHD, or atrial fibrillation were found. BMI was higher in patients with known diabetes compared with NGT. As expected, HbA1c and HOMA-IR increased with severity of AGM, but it is noteworthy that glycemic control was excellent in patients with known DM with a mean HbA1c value of 7.1%, which may explain that HOMA-IR was slightly lower in this group than among newly detected diabetic patients. No difference in the NT-proBNP level was found. Forced expiratory volumes decreased significantly with the severity of AGM. Diabetic patients required higher daily loop diuretic doses.
Baseline parameters according to glycemic classification
| Normal | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Patients, n (%) | 72 (45) | 27 (17) | 28 (17) | 34 (21) | |
| Age, years | 68.2 (10.8) | 74.7 (9.5)* | 72.7 (8.7)* | 69.2 (9.8) | 0.02 |
| Male, n (%) | 50 (69) | 18 (67) | 17 (61) | 26 (77) | 0.60 |
| NYHA III + IV, n (%) | 10 (14) | 6 (25) | 6 (21) | 9 (27) | 0.39 |
| IHD, n (%) | 44 (61) | 10 (37) | 19 (68) | 22 (65) | 0.08 |
| Previous myocardial infarction, n (%) | 33 (47) | 5 (21) | 13 (48) | 12 (38) | 0.12 |
| Atrial fibrillation, n (%) | 11 (16) | 10 (40) | 9 (33) | 9 (27) | 0.10 |
| Body mass index, kg/m2 | 26.3 (5.3) | 26.4 (3.9) | 27.4 (4.6) | 28.9 (5.9)* | 0.11 |
| 6MWD, m | 467 (104) | 437 (79) | 379 (119)** | 349 (110)** | <0.001 |
| FEV-1, % of predicted | 0.78 (0.16) | 0.78 (0.21) | 0.71 (0.15) | 0.68 (0.17)* | 0.02 |
| 2hPG, mmol/L | 5.7 (1.3) | 9.0 (0.9)** | 11.7 (2.9)** | – | <0.001 |
| FPG, mmol/L | 5.6 (0.5) | 5.7 (0.5) | 7.0 (1.3)** | 8.1 (2.1)** | <0.001 |
| HOMA-IR | 1.98 (1.45) | 2.45 (1.82) | 4.74 (3.70)** | 3.99 (2.80)** | <0.001 |
| eGFR, mmol/L/1.73 m2 | 76 (21) | 69 (18) | 64 (24)* | 70 (27) | 0.12 |
| Haemoglobin A1c, % | 5.7 (0.4) | 5.9 (0.4) | 6.2 (0.7)* | 7.1 (1.2)** | <0.001 |
| NT-proBNP, pg/mL*** | 736 (245–1547) | 1336 (372–2300) | 1387 (355–2351) | 989* (465–2207) | 0.19 |
| Haemoglobin, mmol/L | 8.7 (0.9) | 8.5 (1.1) | 8.7 (1.1) | 8.6 (1.0) | 0.81 |
| ACE-I or ARB, n (%) | 67 (96) | 20 (74) | 19 (68) | 29 (88) | 0.001 |
| Beta blocker, n (%) | 55 (79) | 19 (70) | 27 (96) | 25 (76) | 0.08 |
| Aldosterone receptor antagonist, n (%) | 12 (17) | 7 (26) | 9 (32) | 13 (39) | 0.09 |
| Loop diuretic dose, mg/24 h | 33 (68) | 44 (44) | 57 (60) | 85 (96)** | 0.005 |
| Normal | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Patients, n (%) | 72 (45) | 27 (17) | 28 (17) | 34 (21) | |
| Age, years | 68.2 (10.8) | 74.7 (9.5)* | 72.7 (8.7)* | 69.2 (9.8) | 0.02 |
| Male, n (%) | 50 (69) | 18 (67) | 17 (61) | 26 (77) | 0.60 |
| NYHA III + IV, n (%) | 10 (14) | 6 (25) | 6 (21) | 9 (27) | 0.39 |
| IHD, n (%) | 44 (61) | 10 (37) | 19 (68) | 22 (65) | 0.08 |
| Previous myocardial infarction, n (%) | 33 (47) | 5 (21) | 13 (48) | 12 (38) | 0.12 |
| Atrial fibrillation, n (%) | 11 (16) | 10 (40) | 9 (33) | 9 (27) | 0.10 |
| Body mass index, kg/m2 | 26.3 (5.3) | 26.4 (3.9) | 27.4 (4.6) | 28.9 (5.9)* | 0.11 |
| 6MWD, m | 467 (104) | 437 (79) | 379 (119)** | 349 (110)** | <0.001 |
| FEV-1, % of predicted | 0.78 (0.16) | 0.78 (0.21) | 0.71 (0.15) | 0.68 (0.17)* | 0.02 |
| 2hPG, mmol/L | 5.7 (1.3) | 9.0 (0.9)** | 11.7 (2.9)** | – | <0.001 |
| FPG, mmol/L | 5.6 (0.5) | 5.7 (0.5) | 7.0 (1.3)** | 8.1 (2.1)** | <0.001 |
| HOMA-IR | 1.98 (1.45) | 2.45 (1.82) | 4.74 (3.70)** | 3.99 (2.80)** | <0.001 |
| eGFR, mmol/L/1.73 m2 | 76 (21) | 69 (18) | 64 (24)* | 70 (27) | 0.12 |
| Haemoglobin A1c, % | 5.7 (0.4) | 5.9 (0.4) | 6.2 (0.7)* | 7.1 (1.2)** | <0.001 |
| NT-proBNP, pg/mL*** | 736 (245–1547) | 1336 (372–2300) | 1387 (355–2351) | 989* (465–2207) | 0.19 |
| Haemoglobin, mmol/L | 8.7 (0.9) | 8.5 (1.1) | 8.7 (1.1) | 8.6 (1.0) | 0.81 |
| ACE-I or ARB, n (%) | 67 (96) | 20 (74) | 19 (68) | 29 (88) | 0.001 |
| Beta blocker, n (%) | 55 (79) | 19 (70) | 27 (96) | 25 (76) | 0.08 |
| Aldosterone receptor antagonist, n (%) | 12 (17) | 7 (26) | 9 (32) | 13 (39) | 0.09 |
| Loop diuretic dose, mg/24 h | 33 (68) | 44 (44) | 57 (60) | 85 (96)** | 0.005 |
2hPG, 2-h plasma glucose; 6MWD, 6-min walking distance; ACE-I or ARB, angiotensin-converting enzyme inhibitor or angiotensin-receptor blocker; DM, diabetes mellitus; FEV-1, forced expiratory volume first second; FPG, fasting plasma glucose; HOMA-IR, homeostasis model of assessment-insulin resistance; IGT, impaired glucose tolerance; NT-proBNP, N-terminal pronatriuretic peptide; NYHA, New York Heart Association.
*P < 0.05 vs. NGT.
**P< 0.001 vs. NGT.
***Median (interquartile range).
Baseline parameters according to glycemic classification
| Normal | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Patients, n (%) | 72 (45) | 27 (17) | 28 (17) | 34 (21) | |
| Age, years | 68.2 (10.8) | 74.7 (9.5)* | 72.7 (8.7)* | 69.2 (9.8) | 0.02 |
| Male, n (%) | 50 (69) | 18 (67) | 17 (61) | 26 (77) | 0.60 |
| NYHA III + IV, n (%) | 10 (14) | 6 (25) | 6 (21) | 9 (27) | 0.39 |
| IHD, n (%) | 44 (61) | 10 (37) | 19 (68) | 22 (65) | 0.08 |
| Previous myocardial infarction, n (%) | 33 (47) | 5 (21) | 13 (48) | 12 (38) | 0.12 |
| Atrial fibrillation, n (%) | 11 (16) | 10 (40) | 9 (33) | 9 (27) | 0.10 |
| Body mass index, kg/m2 | 26.3 (5.3) | 26.4 (3.9) | 27.4 (4.6) | 28.9 (5.9)* | 0.11 |
| 6MWD, m | 467 (104) | 437 (79) | 379 (119)** | 349 (110)** | <0.001 |
| FEV-1, % of predicted | 0.78 (0.16) | 0.78 (0.21) | 0.71 (0.15) | 0.68 (0.17)* | 0.02 |
| 2hPG, mmol/L | 5.7 (1.3) | 9.0 (0.9)** | 11.7 (2.9)** | – | <0.001 |
| FPG, mmol/L | 5.6 (0.5) | 5.7 (0.5) | 7.0 (1.3)** | 8.1 (2.1)** | <0.001 |
| HOMA-IR | 1.98 (1.45) | 2.45 (1.82) | 4.74 (3.70)** | 3.99 (2.80)** | <0.001 |
| eGFR, mmol/L/1.73 m2 | 76 (21) | 69 (18) | 64 (24)* | 70 (27) | 0.12 |
| Haemoglobin A1c, % | 5.7 (0.4) | 5.9 (0.4) | 6.2 (0.7)* | 7.1 (1.2)** | <0.001 |
| NT-proBNP, pg/mL*** | 736 (245–1547) | 1336 (372–2300) | 1387 (355–2351) | 989* (465–2207) | 0.19 |
| Haemoglobin, mmol/L | 8.7 (0.9) | 8.5 (1.1) | 8.7 (1.1) | 8.6 (1.0) | 0.81 |
| ACE-I or ARB, n (%) | 67 (96) | 20 (74) | 19 (68) | 29 (88) | 0.001 |
| Beta blocker, n (%) | 55 (79) | 19 (70) | 27 (96) | 25 (76) | 0.08 |
| Aldosterone receptor antagonist, n (%) | 12 (17) | 7 (26) | 9 (32) | 13 (39) | 0.09 |
| Loop diuretic dose, mg/24 h | 33 (68) | 44 (44) | 57 (60) | 85 (96)** | 0.005 |
| Normal | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Patients, n (%) | 72 (45) | 27 (17) | 28 (17) | 34 (21) | |
| Age, years | 68.2 (10.8) | 74.7 (9.5)* | 72.7 (8.7)* | 69.2 (9.8) | 0.02 |
| Male, n (%) | 50 (69) | 18 (67) | 17 (61) | 26 (77) | 0.60 |
| NYHA III + IV, n (%) | 10 (14) | 6 (25) | 6 (21) | 9 (27) | 0.39 |
| IHD, n (%) | 44 (61) | 10 (37) | 19 (68) | 22 (65) | 0.08 |
| Previous myocardial infarction, n (%) | 33 (47) | 5 (21) | 13 (48) | 12 (38) | 0.12 |
| Atrial fibrillation, n (%) | 11 (16) | 10 (40) | 9 (33) | 9 (27) | 0.10 |
| Body mass index, kg/m2 | 26.3 (5.3) | 26.4 (3.9) | 27.4 (4.6) | 28.9 (5.9)* | 0.11 |
| 6MWD, m | 467 (104) | 437 (79) | 379 (119)** | 349 (110)** | <0.001 |
| FEV-1, % of predicted | 0.78 (0.16) | 0.78 (0.21) | 0.71 (0.15) | 0.68 (0.17)* | 0.02 |
| 2hPG, mmol/L | 5.7 (1.3) | 9.0 (0.9)** | 11.7 (2.9)** | – | <0.001 |
| FPG, mmol/L | 5.6 (0.5) | 5.7 (0.5) | 7.0 (1.3)** | 8.1 (2.1)** | <0.001 |
| HOMA-IR | 1.98 (1.45) | 2.45 (1.82) | 4.74 (3.70)** | 3.99 (2.80)** | <0.001 |
| eGFR, mmol/L/1.73 m2 | 76 (21) | 69 (18) | 64 (24)* | 70 (27) | 0.12 |
| Haemoglobin A1c, % | 5.7 (0.4) | 5.9 (0.4) | 6.2 (0.7)* | 7.1 (1.2)** | <0.001 |
| NT-proBNP, pg/mL*** | 736 (245–1547) | 1336 (372–2300) | 1387 (355–2351) | 989* (465–2207) | 0.19 |
| Haemoglobin, mmol/L | 8.7 (0.9) | 8.5 (1.1) | 8.7 (1.1) | 8.6 (1.0) | 0.81 |
| ACE-I or ARB, n (%) | 67 (96) | 20 (74) | 19 (68) | 29 (88) | 0.001 |
| Beta blocker, n (%) | 55 (79) | 19 (70) | 27 (96) | 25 (76) | 0.08 |
| Aldosterone receptor antagonist, n (%) | 12 (17) | 7 (26) | 9 (32) | 13 (39) | 0.09 |
| Loop diuretic dose, mg/24 h | 33 (68) | 44 (44) | 57 (60) | 85 (96)** | 0.005 |
2hPG, 2-h plasma glucose; 6MWD, 6-min walking distance; ACE-I or ARB, angiotensin-converting enzyme inhibitor or angiotensin-receptor blocker; DM, diabetes mellitus; FEV-1, forced expiratory volume first second; FPG, fasting plasma glucose; HOMA-IR, homeostasis model of assessment-insulin resistance; IGT, impaired glucose tolerance; NT-proBNP, N-terminal pronatriuretic peptide; NYHA, New York Heart Association.
*P < 0.05 vs. NGT.
**P< 0.001 vs. NGT.
***Median (interquartile range).
LV function and glucose metabolism
Parameters obtained during echocardiography are presented in Table 2. Blood pressure was similar among the glycemic groups during rest and LDDE, but AGM patients had higher average heart rates at both stages. However, there were no significant changes in blood pressure and heart rate from rest to LDDE in any group.
Resting and low-dose dobutamine echocardiographic parameters according to glycemic classification
| NGT | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Resting echocardiography | |||||
| Systolic blood pressure, mmHg | 139 (45) | 136 (16) | 140 (27) | 133 (28) | 0.84 |
| Diastolic blood pressure, mmHg | 82 (12) | 82 (10) | 84 (14) | 82 (14) | 0.96 |
| Heart rate, bpm | 70 (15) | 75 (14) | 73 (13) | 80 (15)* | 0.01 |
| End-diastolic volume, mL | 132 (48) | 115 (41) | 122 (44) | 134 (40) | 0.31 |
| End-systolic volume, mL | 82 (40) | 72 (29) | 81 (37) | 90 (35) | 0.32 |
| Cardiac output, L/min | 4.3 (1.8) | 4.0 (1.6) | 4.2 (1.7) | 4.2 (1.7) | 0.87 |
| LVEF, % | 39.1 (9.6) | 37.8 (8.5) | 35.3 (9.1) | 33.4 (9.5)* | 0.03 |
| s′, cm/s | 5.8 (1.3) | 5.5 (1.1) | 5.5 (1.3) | 5.3 (1.6)* | 0.21 |
| E, cm/s | 74.7 (23.9) | 82.4 (24.4) | 82.1 (20.0) | 83.1 (29.9 | 0.28 |
| e′, cm/s | 7.1 (2.3) | 7.4 (2.4) | 6.5 (2.3) | 6.3 (2.2) | 0.17 |
| E/e′ | 11.4 (5.0) | 11.9 (4.0) | 13.9 (5.9)* | 13.9 (4.8)* | 0.04 |
| LAVI, mL/m2 | 38.1 (12.0) | 47.3 (12.3)* | 44.2 (10.5) | 42.1 (13.1) | 0.005 |
| Grade 2 + 3 diastolic dysfunction, % | 73 | 83 | 94 | 95 | 0.11 |
| Low-dose dobutamine echocardiography | |||||
| Systolic blood pressure, mmHg | 143 (26) | 131 (18) | 142 (31) | 134 (31) | 0.21 |
| Diastolic blood pressure, mmHg | 80 (12) | 77 (11) | 82 (15) | 77 (14) | 0.34 |
| Heart rate, bpm | 71 (18) | 79 (18)* | 73 (16) | 82 (20)* | 0.01 |
| Cardiac output, L/min | 5.2 (2.3) | 4.8 (1.6) | 4.5 (2.1) | 4.7 (2.0) | 0.55 |
| LVEF, % | 49.1 (11.7) | 46.2 (11.2) | 41.8 (10.6)* | 38.9 (11.6)** | 0.001 |
| S′, cm/s | 8.6 (2.8) | 8.0 (2.2) | 7.2 (2.1)* | 7.5 (2.3)* | 0.046 |
| E, cm/s | 84.8 (23.1) | 88.8 (26.1) | 90.3 (20.7) | 88.2 (31.2) | 0.76 |
| e′, cm/s | 8.5 (2.6) | 8.6 (2.7) | 7.6 (2.5) | 7.8 (2.6) | 0.27 |
| E/e′ | 10.7 (4.2) | 10.6 (3.1) | 13.4 (5.9)* | 11.9 (4.5) | 0.05 |
| NGT | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Resting echocardiography | |||||
| Systolic blood pressure, mmHg | 139 (45) | 136 (16) | 140 (27) | 133 (28) | 0.84 |
| Diastolic blood pressure, mmHg | 82 (12) | 82 (10) | 84 (14) | 82 (14) | 0.96 |
| Heart rate, bpm | 70 (15) | 75 (14) | 73 (13) | 80 (15)* | 0.01 |
| End-diastolic volume, mL | 132 (48) | 115 (41) | 122 (44) | 134 (40) | 0.31 |
| End-systolic volume, mL | 82 (40) | 72 (29) | 81 (37) | 90 (35) | 0.32 |
| Cardiac output, L/min | 4.3 (1.8) | 4.0 (1.6) | 4.2 (1.7) | 4.2 (1.7) | 0.87 |
| LVEF, % | 39.1 (9.6) | 37.8 (8.5) | 35.3 (9.1) | 33.4 (9.5)* | 0.03 |
| s′, cm/s | 5.8 (1.3) | 5.5 (1.1) | 5.5 (1.3) | 5.3 (1.6)* | 0.21 |
| E, cm/s | 74.7 (23.9) | 82.4 (24.4) | 82.1 (20.0) | 83.1 (29.9 | 0.28 |
| e′, cm/s | 7.1 (2.3) | 7.4 (2.4) | 6.5 (2.3) | 6.3 (2.2) | 0.17 |
| E/e′ | 11.4 (5.0) | 11.9 (4.0) | 13.9 (5.9)* | 13.9 (4.8)* | 0.04 |
| LAVI, mL/m2 | 38.1 (12.0) | 47.3 (12.3)* | 44.2 (10.5) | 42.1 (13.1) | 0.005 |
| Grade 2 + 3 diastolic dysfunction, % | 73 | 83 | 94 | 95 | 0.11 |
| Low-dose dobutamine echocardiography | |||||
| Systolic blood pressure, mmHg | 143 (26) | 131 (18) | 142 (31) | 134 (31) | 0.21 |
| Diastolic blood pressure, mmHg | 80 (12) | 77 (11) | 82 (15) | 77 (14) | 0.34 |
| Heart rate, bpm | 71 (18) | 79 (18)* | 73 (16) | 82 (20)* | 0.01 |
| Cardiac output, L/min | 5.2 (2.3) | 4.8 (1.6) | 4.5 (2.1) | 4.7 (2.0) | 0.55 |
| LVEF, % | 49.1 (11.7) | 46.2 (11.2) | 41.8 (10.6)* | 38.9 (11.6)** | 0.001 |
| S′, cm/s | 8.6 (2.8) | 8.0 (2.2) | 7.2 (2.1)* | 7.5 (2.3)* | 0.046 |
| E, cm/s | 84.8 (23.1) | 88.8 (26.1) | 90.3 (20.7) | 88.2 (31.2) | 0.76 |
| e′, cm/s | 8.5 (2.6) | 8.6 (2.7) | 7.6 (2.5) | 7.8 (2.6) | 0.27 |
| E/e′ | 10.7 (4.2) | 10.6 (3.1) | 13.4 (5.9)* | 11.9 (4.5) | 0.05 |
DM, diabetes mellitus; E, early diastolic mitral flow velocity; e′, early diastolic longitudinal tissue velocity; IGT, impaired glucose tolerance; LV, left ventricle; s′, systolic longitudinal tissue velocity.
*P < 0.05 vs. NGT.
**P < 0.001 vs. NGT.
Bold values implies statistical significance of between group differences in ANOVA analysis.
Resting and low-dose dobutamine echocardiographic parameters according to glycemic classification
| NGT | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Resting echocardiography | |||||
| Systolic blood pressure, mmHg | 139 (45) | 136 (16) | 140 (27) | 133 (28) | 0.84 |
| Diastolic blood pressure, mmHg | 82 (12) | 82 (10) | 84 (14) | 82 (14) | 0.96 |
| Heart rate, bpm | 70 (15) | 75 (14) | 73 (13) | 80 (15)* | 0.01 |
| End-diastolic volume, mL | 132 (48) | 115 (41) | 122 (44) | 134 (40) | 0.31 |
| End-systolic volume, mL | 82 (40) | 72 (29) | 81 (37) | 90 (35) | 0.32 |
| Cardiac output, L/min | 4.3 (1.8) | 4.0 (1.6) | 4.2 (1.7) | 4.2 (1.7) | 0.87 |
| LVEF, % | 39.1 (9.6) | 37.8 (8.5) | 35.3 (9.1) | 33.4 (9.5)* | 0.03 |
| s′, cm/s | 5.8 (1.3) | 5.5 (1.1) | 5.5 (1.3) | 5.3 (1.6)* | 0.21 |
| E, cm/s | 74.7 (23.9) | 82.4 (24.4) | 82.1 (20.0) | 83.1 (29.9 | 0.28 |
| e′, cm/s | 7.1 (2.3) | 7.4 (2.4) | 6.5 (2.3) | 6.3 (2.2) | 0.17 |
| E/e′ | 11.4 (5.0) | 11.9 (4.0) | 13.9 (5.9)* | 13.9 (4.8)* | 0.04 |
| LAVI, mL/m2 | 38.1 (12.0) | 47.3 (12.3)* | 44.2 (10.5) | 42.1 (13.1) | 0.005 |
| Grade 2 + 3 diastolic dysfunction, % | 73 | 83 | 94 | 95 | 0.11 |
| Low-dose dobutamine echocardiography | |||||
| Systolic blood pressure, mmHg | 143 (26) | 131 (18) | 142 (31) | 134 (31) | 0.21 |
| Diastolic blood pressure, mmHg | 80 (12) | 77 (11) | 82 (15) | 77 (14) | 0.34 |
| Heart rate, bpm | 71 (18) | 79 (18)* | 73 (16) | 82 (20)* | 0.01 |
| Cardiac output, L/min | 5.2 (2.3) | 4.8 (1.6) | 4.5 (2.1) | 4.7 (2.0) | 0.55 |
| LVEF, % | 49.1 (11.7) | 46.2 (11.2) | 41.8 (10.6)* | 38.9 (11.6)** | 0.001 |
| S′, cm/s | 8.6 (2.8) | 8.0 (2.2) | 7.2 (2.1)* | 7.5 (2.3)* | 0.046 |
| E, cm/s | 84.8 (23.1) | 88.8 (26.1) | 90.3 (20.7) | 88.2 (31.2) | 0.76 |
| e′, cm/s | 8.5 (2.6) | 8.6 (2.7) | 7.6 (2.5) | 7.8 (2.6) | 0.27 |
| E/e′ | 10.7 (4.2) | 10.6 (3.1) | 13.4 (5.9)* | 11.9 (4.5) | 0.05 |
| NGT | IGT | New DM | Known DM | P | |
|---|---|---|---|---|---|
| Resting echocardiography | |||||
| Systolic blood pressure, mmHg | 139 (45) | 136 (16) | 140 (27) | 133 (28) | 0.84 |
| Diastolic blood pressure, mmHg | 82 (12) | 82 (10) | 84 (14) | 82 (14) | 0.96 |
| Heart rate, bpm | 70 (15) | 75 (14) | 73 (13) | 80 (15)* | 0.01 |
| End-diastolic volume, mL | 132 (48) | 115 (41) | 122 (44) | 134 (40) | 0.31 |
| End-systolic volume, mL | 82 (40) | 72 (29) | 81 (37) | 90 (35) | 0.32 |
| Cardiac output, L/min | 4.3 (1.8) | 4.0 (1.6) | 4.2 (1.7) | 4.2 (1.7) | 0.87 |
| LVEF, % | 39.1 (9.6) | 37.8 (8.5) | 35.3 (9.1) | 33.4 (9.5)* | 0.03 |
| s′, cm/s | 5.8 (1.3) | 5.5 (1.1) | 5.5 (1.3) | 5.3 (1.6)* | 0.21 |
| E, cm/s | 74.7 (23.9) | 82.4 (24.4) | 82.1 (20.0) | 83.1 (29.9 | 0.28 |
| e′, cm/s | 7.1 (2.3) | 7.4 (2.4) | 6.5 (2.3) | 6.3 (2.2) | 0.17 |
| E/e′ | 11.4 (5.0) | 11.9 (4.0) | 13.9 (5.9)* | 13.9 (4.8)* | 0.04 |
| LAVI, mL/m2 | 38.1 (12.0) | 47.3 (12.3)* | 44.2 (10.5) | 42.1 (13.1) | 0.005 |
| Grade 2 + 3 diastolic dysfunction, % | 73 | 83 | 94 | 95 | 0.11 |
| Low-dose dobutamine echocardiography | |||||
| Systolic blood pressure, mmHg | 143 (26) | 131 (18) | 142 (31) | 134 (31) | 0.21 |
| Diastolic blood pressure, mmHg | 80 (12) | 77 (11) | 82 (15) | 77 (14) | 0.34 |
| Heart rate, bpm | 71 (18) | 79 (18)* | 73 (16) | 82 (20)* | 0.01 |
| Cardiac output, L/min | 5.2 (2.3) | 4.8 (1.6) | 4.5 (2.1) | 4.7 (2.0) | 0.55 |
| LVEF, % | 49.1 (11.7) | 46.2 (11.2) | 41.8 (10.6)* | 38.9 (11.6)** | 0.001 |
| S′, cm/s | 8.6 (2.8) | 8.0 (2.2) | 7.2 (2.1)* | 7.5 (2.3)* | 0.046 |
| E, cm/s | 84.8 (23.1) | 88.8 (26.1) | 90.3 (20.7) | 88.2 (31.2) | 0.76 |
| e′, cm/s | 8.5 (2.6) | 8.6 (2.7) | 7.6 (2.5) | 7.8 (2.6) | 0.27 |
| E/e′ | 10.7 (4.2) | 10.6 (3.1) | 13.4 (5.9)* | 11.9 (4.5) | 0.05 |
DM, diabetes mellitus; E, early diastolic mitral flow velocity; e′, early diastolic longitudinal tissue velocity; IGT, impaired glucose tolerance; LV, left ventricle; s′, systolic longitudinal tissue velocity.
*P < 0.05 vs. NGT.
**P < 0.001 vs. NGT.
Bold values implies statistical significance of between group differences in ANOVA analysis.
Crude and adjusted resting and reserve values of LVEF and s′ are displayed in Figure 1. Mean resting LVEF was 36.8 ± 9.4% and a significant progressive decline in resting LVEF with more severe AGM was found, the absolute difference between NGT and known DM being 5.7%. The resting s′ was significantly lower in known DM only. Both measures of systolic contraction increased significantly during LDDE, LVEF by 8.2% (22% relative increase), and s′ by 2.4 cm/s (43% relative increase). Although an LVEF increase could be observed in all groups, it was significantly lower in the AGM groups, and this lower LVEF reserve exacerbated the differences in LVEF observed at rest. The increase in s′ during LDDE was significantly lower among new DM patients. After multivariable adjustments, the differences in the contractile parameters at rest were attenuated and became insignificant. However, the LVEF reserve remained significantly reduced in patients with new DM (−3.5%, P = 0.035) and known DM (−5.4%, P = 0.001) compared with NGT. Substituting the IHD status with previous myocardial infarction in the multivariable model did not substantially alter these estimates (new DM: −3.4%, P = 0.032 and known DM: −5.1%, P = 0.01).
Distribution of LV contractile parameters and E/e′ at rest and during low-dose dobutamine infusion according to glycemic classification. Left: unadjusted mean values. Right: mean values after adjustment for age, sex, estimated glomerular filtration rate, IHD, atrial fibrillation, systolic blood pressure, resting heart rate, and LV end-diastolic volume index. Resting LVEF was further adjusted for in case of LDDE outcomes.
Moderate-to-severe diastolic dysfunction (Grade 2 or 3) was highly prevalent in this population, affecting 83% of all patients (Table 2). There was a trend towards a higher prevalence of moderate-to-severe diastolic dysfunction with more severe AGM. Neither E nor e′ were significantly different among the groups, but the E/e′ ratio was significantly higher in patients with new and known DM (both +2.5 compared with NGT, P < 0.05), a difference that became insignificant after multivariable adjustment (Figure 1). E/e′ during LDDE lowered slightly and was not different in the glycemic categories (multivariable analysis).
In patients without known diabetes, 2hPG correlated with decreasing LVEF reserve (β = −0.45% per mmol/L, P = 0.026), s′ reserve (β = −0.16 cm/s per mmol/L, P = 0.004), and E/e′ (β = 0.3 per mmol/L, P = 0.048). These correlations became insignificant in multivariable analysis. LogHOMA-IR correlated with decreasing LVEF reserve in both uni- and multivariable models (β = −5.8% per unit, P = 0.004), but neither FPG nor HbA1c were associated with contractile reserve or E/e′.
Exercise capacity and glucose metabolism
Among the161 patients, 154 completed a 6MWD test. Clinical and echocardiographic predictors of 6MWD are shown in Table 3. There was a highly significant step-wise reduction in exercise capacity with increasing severity of AGM (Table 1). Known DM status conferred a 25% reduction in 6MWD compared with NGT whereas IGT and new DM were associated with milder exercise intolerance. Crude and adjusted differences in 6MWD are displayed in Figure 2. After adjusting for baseline clinical differences, both known and new DM were associated with significantly shorter 6MWD. Additional adjustments for LV contractile capacity (LDDE LVEF) and filling pressure (E/e′) parameters did not influence the impact of AGM on 6MWD significantly, but the estimated impact of new DM vs. NGT was slightly reduced and lost the statistical significance. If FEV-1 was added to the model (n = 131), new DM was no longer associated with reduced 6MWD compared with NGT (P = 0.36), but known DM remained associated with reduced 6MWD (−72 m, P = 0.001). Adding spironolactone treatment, ACE inhibitor and beta blocker doses to the model did not alter the estimates of 6MWD in the groups substantially (data not shown).
Clinical and echocardiographic significant predictors of 6MWD in patients with chronic HF
| Beta coefficient (SE) | R | P | |
|---|---|---|---|
| Clinical parameters | |||
| 2-h plasma glucose, per mmol/L | −13.7 (3.0) | −0.38 | <0.001 |
| Age, per year | −5.3 (0.8) | −0.48 | <0.001 |
| Haemoglobin, per mmol/L | 26.0 (9.2) | 0.22 | 0.006 |
| eGFR, per mmol/L | −0.6 (0.18) | 0.37 | <0.001 |
| FEV-1 in % of predicted, per % | 256 (54) | 0.38 | <0.001 |
| Body mass index, per kg/m2 | −4.1 (1.8) | −0.19 | 0.02 |
| Resting heart rate, per bpm | −1.6 (0.6) | −0.21 | 0.008 |
| Resting echocardiography | |||
| LVEF rest, per % | 3.3 (1.0) | 0.28 | 0.001 |
| s′ rest, per cm/s | 24.7 (6.7) | 0.29 | <0.001 |
| e′ rest, per cm/s | 14.9 (3.9) | 0.30 | <0.001 |
| E/e′ rest, per unit | −6.8 (1.8) | −0.30 | <0.001 |
| LAVI, per mL/m2 | −1.6 (0.8) | −0.17 | 0.04 |
| Low-dose dobutamine echocardiography | |||
| LVEF, per % | 2.9 (0.8) | 0.31 | <0.001 |
| s′, per cm/s | 13.9 (3.6) | 0.30 | <0.001 |
| e′, per cm/s | 13.3 (3.5) | 0.30 | <0.001 |
| E/e′, per unit | −6.4 (2.1) | −0.25 | 0.002 |
| Beta coefficient (SE) | R | P | |
|---|---|---|---|
| Clinical parameters | |||
| 2-h plasma glucose, per mmol/L | −13.7 (3.0) | −0.38 | <0.001 |
| Age, per year | −5.3 (0.8) | −0.48 | <0.001 |
| Haemoglobin, per mmol/L | 26.0 (9.2) | 0.22 | 0.006 |
| eGFR, per mmol/L | −0.6 (0.18) | 0.37 | <0.001 |
| FEV-1 in % of predicted, per % | 256 (54) | 0.38 | <0.001 |
| Body mass index, per kg/m2 | −4.1 (1.8) | −0.19 | 0.02 |
| Resting heart rate, per bpm | −1.6 (0.6) | −0.21 | 0.008 |
| Resting echocardiography | |||
| LVEF rest, per % | 3.3 (1.0) | 0.28 | 0.001 |
| s′ rest, per cm/s | 24.7 (6.7) | 0.29 | <0.001 |
| e′ rest, per cm/s | 14.9 (3.9) | 0.30 | <0.001 |
| E/e′ rest, per unit | −6.8 (1.8) | −0.30 | <0.001 |
| LAVI, per mL/m2 | −1.6 (0.8) | −0.17 | 0.04 |
| Low-dose dobutamine echocardiography | |||
| LVEF, per % | 2.9 (0.8) | 0.31 | <0.001 |
| s′, per cm/s | 13.9 (3.6) | 0.30 | <0.001 |
| e′, per cm/s | 13.3 (3.5) | 0.30 | <0.001 |
| E/e′, per unit | −6.4 (2.1) | −0.25 | 0.002 |
E, early diastolic mitral flow velocity; e′, early diastolic longitudinal tissue velocity; eGFR, estimated glomerular filtration rate; FEV-1, forced expiratory volume first second; LAVI, left atrial volume index; LV, left ventricular; s′, systolic longitudinal tissue velocity.
Clinical and echocardiographic significant predictors of 6MWD in patients with chronic HF
| Beta coefficient (SE) | R | P | |
|---|---|---|---|
| Clinical parameters | |||
| 2-h plasma glucose, per mmol/L | −13.7 (3.0) | −0.38 | <0.001 |
| Age, per year | −5.3 (0.8) | −0.48 | <0.001 |
| Haemoglobin, per mmol/L | 26.0 (9.2) | 0.22 | 0.006 |
| eGFR, per mmol/L | −0.6 (0.18) | 0.37 | <0.001 |
| FEV-1 in % of predicted, per % | 256 (54) | 0.38 | <0.001 |
| Body mass index, per kg/m2 | −4.1 (1.8) | −0.19 | 0.02 |
| Resting heart rate, per bpm | −1.6 (0.6) | −0.21 | 0.008 |
| Resting echocardiography | |||
| LVEF rest, per % | 3.3 (1.0) | 0.28 | 0.001 |
| s′ rest, per cm/s | 24.7 (6.7) | 0.29 | <0.001 |
| e′ rest, per cm/s | 14.9 (3.9) | 0.30 | <0.001 |
| E/e′ rest, per unit | −6.8 (1.8) | −0.30 | <0.001 |
| LAVI, per mL/m2 | −1.6 (0.8) | −0.17 | 0.04 |
| Low-dose dobutamine echocardiography | |||
| LVEF, per % | 2.9 (0.8) | 0.31 | <0.001 |
| s′, per cm/s | 13.9 (3.6) | 0.30 | <0.001 |
| e′, per cm/s | 13.3 (3.5) | 0.30 | <0.001 |
| E/e′, per unit | −6.4 (2.1) | −0.25 | 0.002 |
| Beta coefficient (SE) | R | P | |
|---|---|---|---|
| Clinical parameters | |||
| 2-h plasma glucose, per mmol/L | −13.7 (3.0) | −0.38 | <0.001 |
| Age, per year | −5.3 (0.8) | −0.48 | <0.001 |
| Haemoglobin, per mmol/L | 26.0 (9.2) | 0.22 | 0.006 |
| eGFR, per mmol/L | −0.6 (0.18) | 0.37 | <0.001 |
| FEV-1 in % of predicted, per % | 256 (54) | 0.38 | <0.001 |
| Body mass index, per kg/m2 | −4.1 (1.8) | −0.19 | 0.02 |
| Resting heart rate, per bpm | −1.6 (0.6) | −0.21 | 0.008 |
| Resting echocardiography | |||
| LVEF rest, per % | 3.3 (1.0) | 0.28 | 0.001 |
| s′ rest, per cm/s | 24.7 (6.7) | 0.29 | <0.001 |
| e′ rest, per cm/s | 14.9 (3.9) | 0.30 | <0.001 |
| E/e′ rest, per unit | −6.8 (1.8) | −0.30 | <0.001 |
| LAVI, per mL/m2 | −1.6 (0.8) | −0.17 | 0.04 |
| Low-dose dobutamine echocardiography | |||
| LVEF, per % | 2.9 (0.8) | 0.31 | <0.001 |
| s′, per cm/s | 13.9 (3.6) | 0.30 | <0.001 |
| e′, per cm/s | 13.3 (3.5) | 0.30 | <0.001 |
| E/e′, per unit | −6.4 (2.1) | −0.25 | 0.002 |
E, early diastolic mitral flow velocity; e′, early diastolic longitudinal tissue velocity; eGFR, estimated glomerular filtration rate; FEV-1, forced expiratory volume first second; LAVI, left atrial volume index; LV, left ventricular; s′, systolic longitudinal tissue velocity.
Estimates of 6MWD according to glycemic classification. Left: unadjusted mean values; Middle: mean values after multivariable adjustment for the clinical variables age, sex, body mass index, IHD, estimated glomerular filtration rate, resting heart rate, haemoglobin, and atrial fibrillation; Right: mean values after multivariable adjustment for the clinical variables, LVEF during low-dose dobutamine infusion, and E/e′ at rest.
In patients without known DM, 2hPG was the only glycemic parameter associated with exercise capacity, where a weak correlation was found in univariable (β = −13.7 m per mmol/L, P < 0.001) (Figure 3) and multivariable linear regression analysis adjusted for clinical parameters, LDDE LVEF and E/e′ (β = −6.5 m per mmol/L, P = 0.014).
Correlation of 2hPG and 6MWD.
Discussion
This study is the first to report an association between glucose metabolism and contractile reserve in a chronic HF population. We found that more severe AGM was associated with less improvement in LVEF in response to low-dose dobutamine, and this association was independent of important confounders such as IHD and previous myocardial infarction. Furthermore, AGM was associated with exercise intolerance as evidenced by a markedly reduced 6MWD. Some of the differences in 6MWD were explained by the differences in clinical baseline parameters, while contractile reserve and non-invasively assessed LV filling pressure did not explain the reduced exercise capacity in patients with new and known diabetes.
Known or newly diagnosed DM in patients with systolic HF is associated with a two-fold increase in long-term mortality compared with patients without DM.1–3,6 This has been unrelated to differences in the resting LVEF. In the current study, resting LVEF was lower in patients with known DM, but after adjustments for differences in baseline variables and treatment, LVEF was comparable in all groups. The improvement in LVEF with inotropic stimulation, however, was significantly reduced in both new and known DM compared with NGT. Since several studies have found reduced contractile reserve during LDDE to be a predictor of mortality in HF,24,25 this association could provide a hitherto missing pathophysiological link between DM, LV dysfunction, and mortality in chronic systolic HF. Several cardiac metabolic and structural deficits linked to AGM could explain our findings. Insulin resistance and postprandial hyperglycemia is associated with impaired coronary microvascular function and may limit coronary flow reserve and metabolic supply during stress.26 Autonomic neuropathy and changes in cardiac beta-receptor subtype populations in DM patients may cause a blunted LV contractility response to beta-adrenergic stimulation.27,28 Oxidative stress induced by hyperglycemia has been associated with deficits in intracellular calcium handling and alterations in contractile proteins.29,30 As a consequence, peak myocyte relaxation and tension development are delayed, which could exacerbate both systolic and diastolic dysfunction in HF patients particularly during states of increased chronotropy. Finally, DM is associated with increased myocardial fibrosis which has been shown to reduce contractile reserve in dilated cardiomyopathy.31
Exercise intolerance due to dyspnoea or fatigue on exertion is the most prominent symptom of HF. Quantification of exercise capacity by the 6MWD is associated with prognosis and may be viewed as an estimate of cardiovascular reserve. Reduced pulmonary function can also affect the exercise capacity and a recent systematic review concluded that DM is associated with reduced lung function, which was also documented in the current study.32 We found a highly significant and stepwise reduction in 6MWD according to the severity of AGM, and our data correlate well with the previous findings in DM and IGT patients.4,5,33 A range of clinical, resting, and LDDE echocardiographic parameters predicted the 6MWD response. In multivariable analysis, we sought to assess whether the association of AGM with decreased 6MWD primarily was due to an unfavourable distribution of clinical determinants, lower contractile reserve, or elevated LV filling pressure. After adjustments for clinical and echocardiographic parameters, the association of new and known DM groups as well as increasing concentrations of 2hPG with a shorter 6MWD persisted. However, higher BMI, older age, and reduced pulmonary function explained a significant part of the observed AGM-related exercise intolerance, whereas the estimates of contractile function during dobutamine or resting LV filling pressure did not. We conclude that the exercise intolerance in AGM patients is not entirely accounted for by clinical parameters, reduced contractile reserve, or elevated LV filling pressure. Similarly, vascular dysfunction, skeletal muscle ergo receptor dysfunction, or inability of multivariable modelling to account for comorbidities could potentially explain the poor 6MWD performance associated with DM and 2hPG.34 However, even though the present data suggest that reduced contractile reserve does not cause reduced exercise capacity in the diabetic patients, it might be a predictor for mortality and morbidity. Future studies should investigate this.
Some limitations of the study deserve emphasis. The sample size was small and therefore most likely underpowered to detect the moderate differences between the NGT, IGT, and new DM groups, particularly with respect to diastolic dysfunction and 6MWD. Information on the site of previous myocardial infarction, incomplete revascularization, or the prevalence of patients judged unfit for revascularization procedures was not available. The possible contribution of muscular, joint, and skeletal disorders to exercise intolerance was not recorded, and if the prevalence was different among the groups, it would be a source of bias.
Conclusions
In an outpatient population with chronic systolic HF, AGM detected by oral glucose tolerance testing, but not by HbA1c, is independently associated with reduced LVEF reserve during low-dose dobutamine infusion and with exercise intolerance. However, the reduced exercise capacity does not seem to be explained by the reduced contractile reserve.
Funding
M.E. was supported by the Danish Heart Foundation [09-04-R72-A2426-22546]. I.G. was supported by The Danish Agency for Science Technology and Innovation [09-066331].
Conflict of interest: none declared.
