Abstract
Among patients with diabetes mellitus (DM) in the general population, elevated glycosylated hemoglobin (HbA1c) increases the risk of developing heart failure (HF). However, in patients with established HF, the association of HbA1c level with outcomes is not well established. This study investigated the relation between HbA1c and outcomes in HF patients with and without diabetes. We studied 845 advanced HF patients followed at the Ahmanson-UCLA Cardiomyopathy Center, stratified by presence (n=358) or absence (n=487) of DM and by DM-specific HbA1c quartiles (Q) (Diabetics: Q1≤6.4%, Q2 6.5–7.2%, Q3 7.3–8.5%, Q4≥8.6%; Non-diabetics: Q1≤5.6%, Q2 5.7–6.0%, Q3 6.1–6.5%, Q4≥6.6%). The primary outcomes analyzed were death and death/urgent heart transplantation (Utx). In the cohort with DM, 2-year event-free survival was 61% and 65% for Q3 and Q4 compared to 48% and 42% in Q1 and Q2 (p=0.005). In the cohort without DM, there was no difference in outcomes by HbA1c Q. Risk-adjusted analysis in the diabetic cohort showed significantly increased hazard ratio (HR) of death/Utx in Q1 and Q2 compared to Q4. For every unit HbA1c increase, there was a 15% decreased HR of death/Utx and all-cause mortality (p=0.006 and p=0.036, respectively). In the cohort without DM, multivariable models revealed similar HR among HbA1c Q. In this cohort of patients with advanced HF, higher HbA1c levels were associated with improved outcomes in patients with diabetes. This relationship was not observed in patients without DM. Further investigations into mechanisms underlying the relationship between HbA1c, DM, and survival in advanced HF are warranted.
Keywords: diabetes, heart failure, glycosylated hemoglobin, outcomes
Introduction
Diabetes mellitus (DM) is a common co-morbidity in patients with HF.1 The prevalence of DM is approximately 25% in stable HF, compared to only 7% in the general population.2 This relation may in part be due to pathophysiological processes that underlie both HF and DM, including similar patterns of neurohormonal activation, endothelial dysfunction, and oxidative stress.3 A useful marker of glycemic control in diabetic subjects is glycosylated hemoglobin, or HbA1c. Elevated HbA1c is a risk factor for the development of HF in patients with DM, with a 10–15% increased HF risk for every one unit increase in HbA1c.4 Although DM and elevated HbA1c levels are associated with an increased risk of new-onset HF2–4, the data regarding HbA1c and outcomes in patients with pre-existing HF has been conflicting.5–7 This study aimed to re-evaluate the relationship between HbA1c levels and outcomes in a large, well-characterized cohort of advanced HF patients, including patients both with DM and without DM.
Methods
The study population consisted of 845 patients with advanced HF referred to a single university center between January 1, 1999 and July 1, 2010 who had HbA1c measurements recorded within 6 months of referral. The presence of DM was determined by patient self-report and documentation by physician or documentation of treatment with one or more hypoglycemic agents.
Baseline patient characteristics were collected at time of referral; all testing including measurement of HbA1c and other laboratory testing, echocardiography, cardiopulmonary exercise testing, and right-heart catheterization in selected cases were performed within 6 months of initial referral. Patients were further classified into subgroup-specific quartiles (Q) based on HbA1c levels (Table 1). The cohort was analyzed as a whole and by subgroups based on presence or absence of DM.
Table 1.
HbA1c Quartiles by Subgroup
| Diabetes Mellitus | |||
|---|---|---|---|
| Quartile | Whole Cohort | Yes | No |
| 1 | ≤5.8 (n=221) | ≤6.4 (n=90) | ≤5.6 (n=130) |
| 2 | 5.9–6.3 (n=205) | 6.5–.2 (n=94) | 5.7–6.0 (n=120) |
| 3 | 6.4–7.1 (n=213) | 7.3–8.5 (n=86) | 6.1–6.5 (n=143) |
| 4 | ≥7.2 (n=206) | ≥8.6 (n=88) | ≥6.6 (n=94) |
The end-points evaluated in this study included 1) death or need for urgent heart transplant (Utx, Status 1A) and 2) all-cause mortality, in which all heart transplants (Status IA, IB, and II) were coded as non-fatal events. Ventricular assist device (VAD) placement was not analyzed as a final endpoint, but rather the subsequent heart transplant or death after VAD was considered as the final outcome. Of 42 subjects who received VADs, 24 subsequently received Utx, 5 received non-urgent transplant, and 13 died.
Baseline characteristics were presented as mean ± SD for normally distributed continuous variables, median (interquartile range) for non-normally distributed variables, and percent of total for categorical variables. Statistical methods employed included independent samples t-test, ANOVA, chi-square test, and non-parametric tests for comparison of variables as appropriate. Kaplan-Meier survival curves were calculated at one- and two-year follow-up. Cox proportional hazards regression analyses were used to estimate the hazard ratio (HR) of the endpoints of death/Utx and all-cause mortality.Variables included in the multivariate were patient age, gender, body mass index (BMI), and left ventricular ejection fraction (LVEF). A p value ≤ 0.05 was considered significant. PASW v.19 (IBM corp. Sommers, NY) was used for analyses.
Results
The cohort was 72% men with a mean age of 55 ± 12 years. Patients with NYHA III or IV comprised 78% of the patient population. The mean HbA1c was 6.7 ± 1.4% (range 3.6–14.3%), and mean LVEF was 25 ± 12%. Of the patients in the cohort, 42% (n=358) had DM and 58% (n=487) did not have DM. In the patients with DM, the mean HbA1c was 7.6% (range 4.5 – 14.3%). In those without DM, the mean HbA1c was 6.0% (range 4.4–8.0%). Table 2 shows characteristics of the cohort stratified by presence or absence of DM. Diabetic subjects were older, with higher BMI, but similar LVEF.
Table 2.
Baseline Characteristics by Subgroup
| Diabetes Mellitus | |||
|---|---|---|---|
| Variable | Yes | NO | P-value |
| Age(years) | 57.3 ± 10.4 | 53.0 ± 13.5 | <0.001 |
| Female | 28.5% | 27.1% | 0.671 |
| Myocardial ischemic etiology | 59.2% | 33.3% | <0.001 |
| Hemoglobin A1c (%) | 7.6 ± 1.7 | 6.0 ± 0.6 | <0.001 |
| Body mass index (kg/m2) | 28.6 ± 5.4 | 26.5 ± 5.7 | <0.001 |
| New York Heart Association I–II / III–IV | 25.3% / 74.7% | 17.9% / 82.1% | 0.030 |
| Left ventricle ejection fraction(%) | 25.3 ± 9.6 | 25.3 ± 13.0 | 0.946 |
| Left ventricular end-diastolic diameter index (mm/m2) | 33.6 ± 9.2 | 35.6 ± 11.5 | 0.020 |
| Peak oxygen consumption(L/kg/min) | 11.6 ± 3.5 | 13.5 ± 4.8 | <0.001 |
| Heart rate (bpm) | 80.9 ± 15.5 | 81.6 ± 17.2 | 0.566 |
| Mean blood pressure (mmHg) | 80.7 ± 13.5 | 77.9 ± 12.4 | 0.004 |
| Systolic blood pressure (mmHg) | 111.5 ± 21.0 | 104.0 ± 18.1 | <0.001 |
| Pulmonary capillary wedge (mmHg) (n=560) | 22.3 ± 8.1 | 21.6 ± 8.5 | 0.305 |
| Hemoglobin (mL/g) | 12.7 ± 2.1 | 13.2 ± 2.5 | 0.006 |
| Creatinine(mg/dL) | 1.6 ± 1.7 | 1.4 ± 1.4 | 0.092 |
| Blood urea nitrogen(mg/dL) | 34.8 ± 24.4 | 27.8 ± 18.1 | <0.001 |
| B-type natriuretic peptide (pg/dL) | 597 (253 – 1300) | 570 (216 – 1410) | 0.876 |
| Total cholesterol(mg/dL) | 154.0 ± 58.2 | 147.6 ± 50.0 | 0.105 |
| Cardiac resynchronization therapy | 31.2% | 40.6% | 0.005 |
| Implantable cardioverter defibrillator placement | 63.3% | 67.1% | 0.253 |
| Heart Failure Medications | |||
| Beta blocker | 83.8% | 81.3% | 0.349 |
| Aldosterone antagonist | 58.1% | 59.5% | 0.673 |
| Angiotensin converting enzyme inhibitor or Angiotensin receptor blocker | 83.8% | 76.8% | 0.012 |
| Outcomes | |||
| Death / Urgent transplantation 2 year | 42.7% | 42.7% | 0.994 |
| All-cause mortality 2 year | 26.5% | 17.4% | 0.001 |
Baseline patient characteristics of the total cohort by HbA1c Q are demonstrated in Table 3 and baseline patient characteristics of cohort stratified by HbA1c in the presence or absence of DM are presented in Tables 4 and 5, respectively. In the diabetic cohort, patients with higher HbA1c were younger and had higher BMI, cholesterol, and blood pressure; these trends were not observed in the non-diabetic cohort.
Table 3.
Baseline Characteristics of the Cohort Stratified by Hemoglobin A1c Quartile
| Hemoglobin A1e Level (%) | |||||
|---|---|---|---|---|---|
| Variable | ≤ 5.8 | 5.9 – 6.3 | 6.4 – 7.1 | ≥ 7.2 | P-value |
| Age (years) | 51.9 ± 12.8 | 55.1 ± 13.3 | 57.0 ± 12.2 | 55.6 ± 10.9 | <0.001 |
| Female | 32.6% | 25.4% | 26.9% | 25.4% | 0.283 |
| Myocardial ischemic etiology | 38.9% | 40.5% | 41.1% | 57.3% | <0.001 |
| Body mass index (kg/m2) | 26.6± 6.2 | 26.6± 5.3 | 27.1± 5.2 | 29.4± 5.6 | <0.001 |
| New York Heart Association I / II / III / IV | 6.4% / 20.4% / 38.9% / 34.3% | 2.5% / 20.8% / 50.9% / 25.8% | 5.2% / 14.2% / 46.4% / 34.2% | 2.7% / 16.3% / 44.9% / 36.1% | 0.198 |
| History of Diabetes Mellitus | 17.6% | 20.5% | 43.0% | 90.3% | <0.001 |
| Left ventricle ejection fraction (%) | 27.5 ± 14.1 | 24.9 ± 11.9 | 23.9 ± 11.4 | 25.0 ± 8.6 | 0.028 |
| Left ventricular end-diastolic diameter index (mm/m2) | 36.1 ± 11.6 | 36.5 ± 12.0 | 34.4 ± 9.5 | 31.9 ± 8.4 | <0.001 |
| Peak oxygen consumption (L/kg/min) | 13.5 ±4.2 | 13.1 ± 5.0 | 12.4 ± 4.3 | 11.6 ± 3.8 | 0.014 |
| Heart rate (bpm) | 83.5 ± 18.8 | 81.2 ± 17.6 | 83.6 ± 15.6 | 81.5 ± 15.8 | 0.391 |
| Mean blood pressure (mmHG) | 77.0 ± 12.6 | 75.8 ± 11.5 | 74.9 ± 11.6 | 81.3 ± 14.4 | <0.001 |
| Systolic blood pressure (mmHG) | 104.7 ± 17.4 | 105.7 ± 17.1 | 103.9 ± 19.2 | 114.6 ± 22.7 | <0.001 |
| Pulmonary capillary wedge (mmHG) (n=560) | 20.8 ± 8.6 | 21.5 ± 8.2 | 22.6 ± 8.3 | 22.6 ± 8.1 | 0.173 |
| Hemoglobin (mL/g) | 12.7 ± 3.1 | 13.2 ± 2.0 | 13.1 ± 2.1 | 13.0 ± 2.0 | 0.127 |
| Creatinine (mg/dL) | 1.5 ± 1.2 | 1.4 ± 0.8 | 1.6 ± 1.7 | 1.7 ± 2.1 | 0.296 |
| Blood urea nitrogen (mg/dL) | 26.1 ± 16.1 | 28.1 ± 17.1 | 33.6 ± 22.9 | 35.7 ± 26.2 | <0.001 |
| B-type natriuretic peptide (pg/dL) | 447 (179.7–240.0) | 520.5 (240.5–1277.5) | 706.0 (316.5–625.0) | 588.0 (251.0–1300.0) | 0.012 |
| Total cholesterol (mg/dL) | 152.3 ± 54.8 | 147.4 ± 47.4 | 144.9 ± 50.4 | 156.7 ± 60.1 | 0.110 |
| Low-density lipoprotein cholesterol(mg/dL) | 88.3 ± 40.0 | 83.5 ± 35.2 | 83.5 ± 35.0 | 85.4 ± 41.8 | 0.577 |
| High-density lipoprotein cholesterol(mg/dL) | 38.5 ± 16.5 | 36.2 ± 12.5 | 37.7 ± 16.2 | 35.7 ± 13.8 | 0.237 |
| Triglycerides (mg/dL) | 104.0 (69.5–164.5) | 111.0 (77.0–180.0) | 102.0 (71.0–146.0) | 127.0 (76.0–257.0) | 0.001 |
| Heart Failure Medications | |||||
| Beta blocker | 76.9% | 83.9% | 82.7% | 86.4% | 0.066 |
| Aldosterone antagonist | 52.5% | 62.4% | 59.3% | 62.1% | 0.126 |
| Angiotensin converting enzyme inhibitor or Angiotensin receptor blocker | 73.3% | 79.5% | 80.8% | 85.9% | 0.013 |
Table 4.
Baseline Characteristics of the Diabetic Subgroup Stratified by Hemoglobin A1c Quartile
| Hemoglobin A1c Level (%) | |||||
|---|---|---|---|---|---|
| Variable | ≤ 6.4 | 6.5–7.2 | 7.3–8.5 | ≥ 8.6 | P-value |
| Age (years) | 58.6 ± 9.6 | 59.0 ± 10.1 | 58.0 ± 10.9 | 53.7 ± 10.0 | 0.001 |
| Female | 28.9% | 29.3% | 24.4% | 31.0% | 0.797 |
| Myocardial ischemic etiology | 61.1% | 58.5% | 67.4% | 50.0% | 0.130 |
| Body mass index (kg/m2) | 27.5 ± 5.1 | 27.8 ± 5.2 | 28.6 ± 5.1 | 30.7 ± 5.9 | <0.001 |
| New York Heart Association I / II / III / IV | 2.9% / 17.4% / 43.4% / 36.2% | 4.3% / 10.1% / 47.8% / 37.7% | 3.3% / 11.5% / 54.1% / 31.1% | 1.6% / 20.6% / 38.1% / 39.7% | 0.678 |
| Hemoglobin A1c (%) | 5.8 ± 0.4 | 6.8 ± 0.2 | 7.9 ± 0.4 | 10.0± 1.3 | <0.001 |
| Left ventricle ejection fraction (%) | 26.4 ± 10.5 | 24.5 ± 10.7 | 24.5 ± 8.4 | 25.5 ± 8.5 | 0.501 |
| Left ventricular end-diastolic diameter index (mm/m2) | 35.5 ± 11.0 | 35.1 ± 8.8 | 32.7 ± 8.3 | 30.8 ± 8.0 | 0.014 |
| Peak oxygen consumption (L/kg/min) | 11.9 ± 2.7 | 11.1 ± 3.6 | 11.6 ± 3.3 | 11.8 ± 4.1 | 0.692 |
| Heart rate (bpm) | 79.7 ± 16.0 | 78.4 ± 15.1 | 82.5 ± 15.9 | 82.9 ± 14.7 | 0.213 |
| Mean blood pressure (mmHg) | 76.4 ± 12.1 | 77.5 ± 12.1 | 83.9± 13.1 | 85.2 ± 14.7 | <0.001 |
| Systolic blood pressure (mmHg) | 107.2 ± 16.6 | 107.0 ± 20.1 | 115.4 ± 21.1 | 116.6 ± 24.2 | 0.003 |
| Pulmonary capillary wedge (mmHg) (n=560) | 21.4 ± 7.5 | 22.5 ± 8.6 | 22.6 ± 7.6 | 22.7 ± 8.8 | 0.801 |
| Hemoglobin (mL/g) | 12.2 ± 2.3 | 12.8 ± 2.1 | 12.9 ± 1.9 | 13.0 ± 1.9 | 0.096 |
| Creatinine (mg/dL) | 1.5 ± 0.8 | 1.5 ± 0.8 | 1.8 ± 2.6 | 1.6 ± 1.8 | 0.540 |
| Blood urea nitrogen (mg/dL) | 30.1 ± 18.0 | 35.9 ± 23.8 | 41.7 ± 32.8 | 31.8 ± 19.1 | 0.008 |
| B-type natriuretic peptide (pg/dL) | 654.0 (255.0–1300.0) | 516.0 (252.0–1220.0) | 743.0 (340.0–1301.0) | 503.0 (202.5–1055.0) | 0.557 |
| Total cholesterol (mg/dL) | 151.2 ± 56.1 | 148.2 ± 52.8 | 155.9 ± 64.1 | 161.1 ± 60.0 | 0.507 |
| Low-density lipoprotein cholesterol(mg/dL) | 83.9 ± 44.0 | 83.8 ± 33.1 | 81.5 ± 38.8 | 90.4 ± 45.3 | 0.575 |
| High-density lipoprotein cholesterol(mg/dL) | 35.3 ± 11.7 | 37.6 ± 16.3 | 33.8 ± 9.6 | 37.4 ± 15.5 | 0.266 |
| Triglycerides (mg/dL) | 116.0 (84.0–188.0) | 102.0 (65.0–140.0) | 126.0 (75.0–273.0) | 139.0 (81.0–248.0) | 0.029 |
| Heart Failure Medications | |||||
| Beta blocker | 80.0% | 81.9% | 84.9% | 88.6% | 0.425 |
| Aldosterone antagonist | 53.3% | 59.6% | 60.5% | 59.1% | 0.763 |
| Angiotensin converting enzyme inhibitor or Angiotensin receptor blocker | 83.3% | 81.9% | 79.1% | 90.9% | 0.175 |
| Anti-Diabetic Medications | |||||
| Insulin | 33.3% | 21.4% | 35.2% | 48.9% | 0.003 |
| Sulfonylureas | 25.0% | 23.8% | 30.7% | 42.2% | 0.033 |
| Glitazones | 8.3% | 5.1% | 14.8% | 15.4% | 0.276 |
| Metformin | 8.3% | 37.5% | 20.5% | 26.6% | 0.019 |
Table 5.
Baseline Characteristics of the Non-Diabetic Subgroup Stratified by Hemoglobin A1c Quartile
| Hemoglobin A1c Level (%) | P-value | ||||
|---|---|---|---|---|---|
| ≤ 5.6 | 5.7–6.0 | 6.1–6.5 | ≥ 6.6 | ||
| Age (years) | 50.1 ± 14.0 | 53.9 ± 13.4 | 54.1 ± 13.7 | 54.2 ± 12.4 | 0.038 |
| Female | 30.8% | 28.3% | 26.2% | 21.7% | 0.500 |
| Myocardial ischemic etiology | 35.4% | 35.0% | 34.3% | 26.6% | 0.499 |
| Body mass index (kg/m2) | 26.2 ± 6.6 | 26.9 ± 5.3 | 26.4 ± 5.9 | 26.6 ± 4.7 | 0.807 |
| New York Heart Association I / II / III / IV | 6.7% / 21.3% / 38.2% / 33.7% | 5.9% / 22.4% / 42.4% / 29.4% | 3.5% / 19.1% / 52.2% / 25.2% | 4.5% / 17.9% / 44.8% / 32.8% | 0.783 |
| Hemoglobin A1c (%) | 5.3 ± 0.3 | 5.9 ± 0.1 | 6.3 ± 0.1 | 6.9 ± 0.3 | <0.001 |
| Left ventricle ejection fraction (%) | 27.8 ± 14.5 | 26.8 ± 14.3 | 23.1 ± 11.1 | 23.4± 11.1 | 0.009 |
| Left ventricular end-diastolic diameter index (mm/m2) | 36.7 ± 12.8 | 34.7 ± 9.0 | 36.4 ± 12.9 | 34.3 ± 10.3 | 0.380 |
| Peak oxygen consumption (L/kg/min) | 13.8 ± 4.1 | 13.5 ± 5.1 | 13.4 ± 5.5 | 13.4 ± 4.4 | 0.968 |
| Heart rate (bpm) | 84.3 ± 20.2 | 78.3 ± 15.5 | 82.3 ± 16.5 | 81.3 ± 15.7 | 0.107 |
| Mean blood pressure (mmHG) | 76.3 ± 11.7 | 79.1 ± 12.9 | 79.1 ± 12.2 | 76.8 ± 12.7 | 0.208 |
| Systolic blood pressure (mmHg) | 103.2 ± 16.5 | 105.8 ± 17.3 | 104.8 ± 18.0 | 101.9 ± 20.7 | 0.476 |
| Pulmonary capillary wedge (mmHg) (n=560) | 20.1 ± 9.3 | 21.6 ± 7.9 | 21.8 ± 8.3 | 22.9 ± 8.3 | 0.257 |
| Hemoglobin (mL/g) | 12.6 ± 2.2 | 13.4 ± 3.6 | 13.4 ± 1.9 | 13.4 ± 2.1 | 0.040 |
| Creatinine (mg/dL) | 1.5 ± 1.4 | 1.3 ± 0.9 | 1.5 ± 0.8 | 1.6 ± 2.4 | 0.656 |
| Blood urea nitrogen (mg/dL) | 24.7 ± 15.3 | 25.9 ± 17.7 | 30.4 ± 19.5 | 30.8 ± 19.2 | 0.017 |
| B-type natriuretic peptide (pg/dL) | 386.0 (135.0–1093.0) | 402.0 (185.0–838.5) | 687.0 (371.0–1610.0) | 1020.0 (477.0–2150.0) | <0.001 |
| Total cholesterol (mg/dL) | 148.8 ± 49.4 | 152.1 ± 52.3 | 143.2 ± 47.0 | 146.5 ± 50.0 | 0.557 |
| Low-density lipoprotein cholesterol (mg/dL) | 87.4 ± 36.3 | 87.1 ± 36.4 | 81.7 ± 35.1 | 85.8 ± 37.8 | 0.606 |
| High-density lipoprotein cholesterol (mg/dL) | 38.3 ± 15.8 | 39.0 ± 16.0 | 36.8 ± 14.9 | 36.5 ± 16.2 | 0.584 |
| Triglycerides (mg/dL) | 92 (69 – 141) | 108 (75 – 174) | 96 (69–142) | 105 (79 – 170) | 0.236 |
| Heart Failure Medications | |||||
| Beta blocker | 76.9% | 80.0% | 82.5% | 87.2% | 0.252 |
| Aldosterone antagonist | 53.8% | 56.7% | 64.3% | 63.8% | 0.236 |
| Angiotensin converting enzyme inhibitor or Angiotensin receptor blocker | 70.0% | 74.2% | 79.0% | 86.3% | 0.031 |
For the total population, there were 123 deaths during the first year of follow-up and 180 total deaths by two years. Of the deaths by two years, congestive HF accounted for 62.2%, whereas 15.6% of deaths were sudden, and 22.2% occurred from unknown or other causes. At two years, 181 patients (21.4%) had received Utx.
In analyzing the total cohort, Q4 (HbA1c ≥7.2%) had the best survival free from death/Utx (Figure 1a), although this relationship was not significant for the endpoint of all-cause mortality (Figure 1b). After adjusting for age, gender, BMI, and LVEF on multivariable analysis, Q1–Q3 compared to Q4 were at higher risk of death/Utx[HR Q1: 1.52 (1.12–2.05) p=0.007, Q2: 1.27 (0.93 – 1.74) p=0.132; Q3 1.54 (1.14–2.07) p=0.005]. Analyzing HbA1c as a continuous variable in a multivariable analysis, for each unit increase in HbA1c there was an 8% decreased risk of death or need for Utx (HR 0.92; 95% confidence interval [CI] 0.84–1.00, p=0.050).
Figure 1.
Survival outcomes of advanced heart failure patients by HbA1c Quartiles. 1a. Whole Cohort: 2 Year Survival Free from Death/UTx by HbA1c Quartile (Q1 49.6%, Q2 55.3%, Q3 47.5%, Q4 60.4%) 1b. Whole Cohort: 2 Year Survival Free from All-Cause Mortality by HbA1c Quartile (Q1 70.2%, Q2 77.3%, Q3 65.7%, Q4 71.2%) 1c. DM+: 2 Year Survival Free from Death/UTx by HbA1c Quartile (Q1 47.9%, Q2 41.5%, Q3 60.7%, Q4 65.3%) 1d. DM+: 2 Year Survival Free from All-Cause Mortality by HbA1c Quartile (Q1 62.5%, Q2 53.5%, Q3 72.9%, Q4 74.4%) 1e. DM−: 2 Year Survival Free from Death/UTx by HbA1c Quartile (Q1 50.4%, Q2 60.6%, Q3 51.1%, Q4 49.9%) 1f. DM−: 2 Year Survival Free from All-Cause Mortality by HbA1c (Q1 69.8%, Q2 84.2%, Q3 77.1%, Q4 71.1%)
In the cohort with DM, both survival free from death/Utx and overall survival were significantly better in the higher HbA1c Q (Figure 1c, 1d). After multivariate analysis, Q1and Q2 compared to Q4 were at higher risk of death/Utx [HR Q1: 1.74 (1.04–2.92) p=0.036, Q2: 1.86 (1.13–3.07) p=0.015]. Furthermore, on analyzing HbA1c as a continuous variable, for each unit increase in HbA1c in patients with DM, there was an associated 15% decreased risk of death/Utx and a 15% decreased risk of all-cause mortality [HR 0.85 (0.76–0.95) p=0.006 and HR 0.85 (0.74–0.99) p=0.036, respectively]. For patients with DM, we also performed a separate multivariate analysis in which we additionally adjusted for use of the diabetic medications - insulin, metformin, sulfonylureas, and glitazones – in addition to age, gender, BMI, and LVEF. After adjustment for anti-diabetic medicines, Q1 and Q2 remained at a significantly higher risk compared to Q4 [HR Q1: 1.9 (1.1–3.2) p=0.026, Q2: 2.0 (1.2–3.4) p=0.008, Q3: 1.1 (0.6–1.9) p=0.775].
In patients without DM, both primary outcomes tended to be better in Q3 (HbA1c 5.7% – 6.0%); however, this relationship was not statistically significant (Figure 1e, 1f). HbA1c was also not an independent predictor of outcomes on multivariable analysis.
In order to further investigate the relationship between HbA1c levels and outcomes in HF patients with DM, outcomes by deciles of HbA1c were analyzed. Higher HbA1c deciles were, in general, associated with improved outcomes, with both the best survival free from Utx and best overall survival seen in Decile 8 (HbA1c 8.3 – 8.9%). Comparing Decile 8 to Decile 1, survival free from death/Utx was 79.7% and 34.1%, respectively, and overall survival was 83.2% and 53.8%, respectively. Since gender may modify the risk attributable to DM8,9 we analyzed survival by HbA1c levels in strata of men vs. women; elevated HbA1c was associated with better survival free from Utx in both men and women (data not shown).
Discussion
HF remains a leading cause of death and disability and affects approximately 6 million people in the U.S. DM is common in HF and associated with higher mortality.9,10 Despite the abundance of evidence linking DM, insulin resistance, and hyperglycemia to impaired functional status and worse outcomes in patients with HF, there is lack of data and guidelines on optimal strategies to manage DM in patients with chronic HF.HbA1c reflects the prior 2–3 months of glycemic control and is a useful marker for determining long-term glycemia without reference to prandial state.11 In this cohort of patients with advanced HF, higher HbA1c levels were associated with improved outcomes in patients with DM. These findings may have important implications for patients with HF and DM.
Recent studies in patients with DM at high cardiovascular risk have failed to show a positive effect of intensive glucose control on macrovascular outcomes or mortality.12In the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Study Group, the intensive-therapyarm (goal HbA1c<7%) showed an increase in the overall mortality by 21% which prompted early termination.13 Potential explanations for the adverse outcomes include increased rates of hypoglycemia, increased use of insulin, and weight gain in the intensive control group.14 Thus, for patients with cardiovascular disease and DM, glycemic control to HbA1c ≤ 7% is recommendedfor reduction of cardiovascular risk, rather than more intensive glucose control.14
Previous studies have investigated the relationship between HbA1c and outcomes in patients with HF.5–7In an analysis of the Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) Program, a graded positive association between HbA1c and cardiovascular risk was demonstrated.7Aguilar et al. found a U-shaped relationship with the best survival in the group with modest glycemic control (HbA1c 7.1–7.8% range). Our cohort significantly differed from the Aguilar cohort, which was comprised mainly of older men at a less advanced stage of HF.6 In our advanced HF population, and specifically in the diabetic cohort, we did not find a U-shaped or positive correlation between HbA1c and outcomes in diabetic HF patients but rather found an increase in risk above a HbA1c level of approximately 7.2%, similar to our previous findings in a smaller cohort.5
The mechanisms of the relationship between higher HbA1c and improved outcomes in this advanced HF population with DM require further study. HbA1c is a measure that reflects the metabolic state of the patient as a whole, and includes factors like hypoglycemia, cachexia, and catabolic stress.15These factors could be indicative of protein energy malnutrition and an increased inflammatory response associated with decreasing HbA1c levels and worse clinical outcomes. A recent study of HbA1c levels in a general non-diabetic US population (n=14,099) also found that very low HbA1c levels were associated with increased all-cause mortality.16 HF is also associated with insulin resistance, neurohormonal activation, and deranged myocardial glucose and fatty acid metabolism,17 perturbations which may render HbA1c an insufficient marker for glycemia in this patient population. Anti-diabetic medication usage may play a role; in our cohort patients in higher HbA1c Q were more often on metformin and glitazones, known to be projective in HF,18,19but also were more often on insulin therapy, which may have adverse effects in HF.3,10Lastly, in HF patients traditional risk factors such as obesity, high cholesterol, and high blood pressure – all associated with higher HbA1c – are associated with improved outcomes in patients with HF, a relation often termed “reverse epidemiology”.20Thus, one or all of these factors may be contributing to the improved outcomes seen in patients with higher HbA1c in this advanced HF cohort.
We acknowledge potential limitations of this study. This was an observational study and there were baseline differences in the patient population quartiles. Inherent to the study design, there is the possibility of residual, unmeasured confounding variables that may be contributing to the observed results. For example, blood pressure may be contributing to the observed results and its interaction with HbA1c and outcomes in diabetic patients with advanced HF requires further study. In addition, since severely ill patients are more likely to have their HbA1c levels measured and recorded as part of a heart transplant evaluation, the cohort pool may be skewed towards more chronically or acutely ill patients. Furthermore, in light of recent changes in the American Diabetes Association's guidelines on the diagnosis of DM to include patients with HbA1c levels >6.5%,21there are likely to be patients in the subgroup “without DM” who have subclinical or undiagnosed DM, as evidenced by some above-normal HbA1c in this cohort. Lastly, as this study is retrospective in nature, we cannot infer recommendations regarding optimal glycemic control management strategies in patients with advanced HF.
Acknowledgments
Funding Sources: Dr. Horwich was supported by NHLBI / NIH 1K23HL085097.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Preiss D, Zetterstrand S, McMurray JJV, Östergren J, Michelson EL, Granger CB, Yusuf S, Swedberg K, Pfeffer MA, Gerstein HC, Sattar N. Predictors of development of diabetes in patients with chronic heart failure in the candesartan in heart failure assessment of reduction in mortality and morbidity (CHARM) program. Diabetes Care. 2009;32:915–920. doi: 10.2337/dc08-1709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nichols GA, Gullion CM, Koro CE, Ephross SA, Brown JB. The incidence of congestive heart failure in type 2 diabetes. Diabetes Care. 2004;27:1879–1884. doi: 10.2337/diacare.27.8.1879. [DOI] [PubMed] [Google Scholar]
- 3.Kostis JB, Sanders M. The association of heart failure with insulin resistance and the development of type 2 diabetes. Am J Hypertens. 2005;18:731–737. doi: 10.1016/j.amjhyper.2004.11.038. [DOI] [PubMed] [Google Scholar]
- 4.Iribarren C, Karter AJ, Go AS, Ferrara A, Liu JY, Sidney S, Selby JV. Glycemic control and heart failure among adult patients with diabetes. Circulation. 2001;103:2668–2673. doi: 10.1161/01.cir.103.22.2668. [DOI] [PubMed] [Google Scholar]
- 5.Eshaghian S, Horwich TB, Fonarow GC. An unexpected inverse relationship between HbA1c levels and mortality in patients with diabetes and advanced systolic heart failure. American heart journal. 2006;151:91. doi: 10.1016/j.ahj.2005.10.008. [DOI] [PubMed] [Google Scholar]
- 6.Aguilar D, Bozkurt B, Ramasubbu K, Deswal A. Relationship of hemoglobin A1C and mortality in heart failure patients with diabetes. Journal of the American College of Cardiology. 2009;54:422–428. doi: 10.1016/j.jacc.2009.04.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gerstein HC, Swedberg K, Carlsson J, McMurray JJV, Michelson EL, Olofsson B, Pfeffer MA, Yusuf S. The hemoglobin A1c level as a progressive risk factor for cardiovascular death, hospitalization for heart failure, or death in patients with chronic heart failure: an analysis of the candesartan in heart failure: assessment of reduction in mortality and morbidity (CHARM) program. Arch Intern Med. 2008;168:1699–1704. doi: 10.1001/archinte.168.15.1699. [DOI] [PubMed] [Google Scholar]
- 8.MacDonald MR, Jhund PS, Petrie MC, Lewsey JD, Hawkins NM, Bhagra S, Munoz N, Varyani F, Redpath A, Chalmers J, MacIntyre K, McMurray JJV. Discordant short- and long-term outcomes associated with diabetes in patients with heart failure: importance of age and sex. Circulation: Heart Failure. 2008;1:234–241. doi: 10.1161/CIRCHEARTFAILURE.108.794008. [DOI] [PubMed] [Google Scholar]
- 9.Gustafsson I, Brendorp B, Seibaek M, Burchardt H, Hildebrandt P, Kober L, Torp-Pedersen C, DIAMOND Study Group Influence of diabetes and diabetes-gender interaction on the risk of death in patients hospitalized with congestive heart failure. J Am Coll Cardiol. 2004;43:771–777. doi: 10.1016/j.jacc.2003.11.024. [DOI] [PubMed] [Google Scholar]
- 10.Smooke S, Horwich TB, Fonarow GC. Insulin-treated diabetes is associated with a marked increase in mortality in patients with advanced heart failure. American heart journal. 2005;149:168–174. doi: 10.1016/j.ahj.2004.07.005. [DOI] [PubMed] [Google Scholar]
- 11.Tests of glycemia in diabetes. Diabetes Care. 2003;26:s106–s108. doi: 10.2337/diacare.26.2007.s106. [DOI] [PubMed] [Google Scholar]
- 12.Montori VM, Fernández-Balsells M. Glycemic control in type 2 diabetes: time for an evidence-based about-face? Annals of Internal Medicine. 2009;150:803–808. doi: 10.7326/0003-4819-150-11-200906020-00008. [DOI] [PubMed] [Google Scholar]
- 13.ACCORD Study Group. Gerstein HC, Miller ME, Genuth S, Ismail-Beigi F, Buse JB, Goff DC, Jr, Probstfield JL, Cushman WC, Ginsberg HN, Bigger JT, Grimm RH, Jr, Byington RP, Rosenberg YD, Friedewald WT. Long-term effects of intensive glucose lowering on cardiovascular outcomes. New England Journal of Medicine. 2011;364:818–828. doi: 10.1056/NEJMoa1006524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Skyler JS, Bergenstal R, Bonow RO, Buse J, Deedwania P, Gale EAM, Howard BV, Kirkman MS, Kosiborod M, Reaven P, Sherwin RS. Intensive glycemic control and the prevention of cardiovascular events: implications of the ACCORD, ADVANCE, and VA diabetes trials: a position statement of the American Diabetes Association and a scientific statement of the American College of Cardiology Foundation and the American Heart Association. Journal of the American College of Cardiology. 2009;53:298–304. doi: 10.1016/j.jacc.2008.10.008. [DOI] [PubMed] [Google Scholar]
- 15.Kalantar-Zadeh K, Block G, Horwich T, Fonarow GC. Reverse epidemiology of conventional cardiovascular risk factors in patients with chronic heart failure. Journal of the American College of Cardiology. 2004;43:1439–1444. doi: 10.1016/j.jacc.2003.11.039. [DOI] [PubMed] [Google Scholar]
- 16.Carson AP, Fox CS, McGuire DK, Levitan EB, Laclaustra M, Mann DM, Muntner P. Low hemoglobin A1c and risk of all-cause mortality among US adults without diabetes. Circulation. Cardiovascular quality and outcomes. 2010;3:661–667. doi: 10.1161/CIRCOUTCOMES.110.957936. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ashrafian H, Frenneaux MP, Opie LH. Metabolic mechanisms in heart failure. Circulation. 2007;116:434–448. doi: 10.1161/CIRCULATIONAHA.107.702795. [DOI] [PubMed] [Google Scholar]
- 18.Shah DD, Fonarow GC, Horwich TB. Metformin therapy and outcomes in patients with advanced systolic heart failure and diabetes. Journal of Cardiac Failure. 2010;16:200–206. doi: 10.1016/j.cardfail.2009.10.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Masoudi FA, Inzucchi SE, Wang Y, Havranek EP, Foody JM, Krumholz HM. Thiazolidinediones, metformin, and outcomes in older patients with diabetes and heart failure. Circulation. 2005;111:583–590. doi: 10.1161/01.CIR.0000154542.13412.B1. [DOI] [PubMed] [Google Scholar]
- 20.Guder G, Frantz S, Bauersachs J, Allolio B, Wanner C, Koller MT, Ertl G, Angermann CE, Stork S. Reverse epidemiology in systolic and nonsystolic heart failure: cumulative prognostic benefit of classical cardiovascular risk factors. Circulation. Heart failure. 2009;2:563–71. doi: 10.1161/CIRCHEARTFAILURE.108.825059. [DOI] [PubMed] [Google Scholar]
- 21.Association AD. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2010;33:S62–S69. doi: 10.2337/dc10-S062. [DOI] [PMC free article] [PubMed] [Google Scholar]