Variability in Hemoglobin A1c and risk of incident heart failure

AHA sessions 2019

1. Variability in Hemoglobin A1c and Risk of Incident Heart Failure among Patients with Type 2 Diabetes Mellitus Matthew W. Segar, MD, MS PGY-2, University of Texas Southwestern Medical Center

2. DM: A Common Complication of HF Shah, Lancet Diabetes Endocrin , 2015 Bertoni , Diabetes Care , 2004

3. Refractory to Strict Glucose Control A single measurement of glycemia may not reflect chronic glucose exposure and expectant CV risk.

4. Fluctuations in Glycemic Control Mortality Diabetic retinopathy CV events Renal disease Gorst , Diabetes Care , 2015

5. Central Hypothesis Is long-term change and variability in glycemia associated with increased risk of heart failure among patients with T2DM?

6. Study Methodology ACCORD Trial Long-standing T2DM Patients [Intensive (<6%) vs. Standard (7-7.9%) HbA1c control] 3+ HbA1c recordings in first 3 years N = 8,576 Outcome Time to first hospitalization event for HF or death due to HF Percent change in HbA1c  HbA1c Glycemic variability of HbA1c ASV ( average absolute difference between successive values) CV ( standard deviation / mean) SD ( standard deviation) Covariates of Interest

7. Study Methodology Variability Outcome  HbA1c

8. Association between  HbA1c and risk of incident HF   Model 1 Model 2   HR (95% CI) P value HR (95% CI) P value <10% loss to <10% gain Referent Group ≥10% loss 1.84 (1.14, 2.95) 0.01 1.77 (1.09, 2.87) 0.02 ≥10% gain 1.98 (1.25, 3.12) 0.003 1.89 (1.18, 3.03) 0.008 Events = 161 (1.9%) Model 1 was adjusted for age, sex, race, level of education, treatment group, systolic blood pressure, history of cardiovascular disease, alcohol use, antihypertensive medication, insulin use, body mass index, serum creatinine, total cholesterol, and baseline HbA1c level. Model 2 was adjusted for the same variables as model 1 and for change in systolic blood pressure, change in body mass index, change in creatinine, and change in HbA1c level from enrollment baseline to year 3.

9. Association between  HbA1c and risk of incident HF

10. Long-term Variability in HbA1c and risk of HF

11. Long-term Variability in HbA1c and risk of HF Model 1 was adjusted for age, sex, race, level of education, treatment group, systolic blood pressure, history of cardiovascular disease, alcohol use, antihypertensive medication, insulin use, body mass index, serum creatinine, total cholesterol, and baseline HbA1c level. Model 2 was adjusted for the same variables as model 1 and for change in systolic blood pressure, change in body mass index, change in creatinine, and change in HbA1c level from enrollment baseline to year 3.

12. Sensitivity Analysis: Long-term Variability in FPG and risk of HF Model 1 was adjusted for age, sex, race, level of education, treatment group, systolic blood pressure, history of cardiovascular disease, alcohol use, antihypertensive medication, insulin use, body mass index, serum creatinine, total cholesterol, and baseline FPG level. Model 2 was adjusted for the same variables as model 1 and for change in systolic blood pressure, change in body mass index, change in creatinine, and change in FPG level from enrollment baseline to year 3.

13. Limitations Study population was derived from the ACCORD trial which included patients with T2DM who had prevalent cardiovascular disease or risk factors. Left ventricular ejection fraction and HF subtype data were not available for the present analysis.

14. Conclusions Observed a non-linear relationship between changes in HbA1c during the trial and risk of HF. Higher long-term variability in HbA1c and FPG were also significantly associated with higher risk of HF. Findings highlight the prognostic significance of substantial fluctuations in glycemia among patients with T2DM in forecasting future HF risk

15. Acknowledgements Primary Mentor Ambarish Pandey, MD Collaborators Kershaw Patel, MD Muthiah Vaduganathan , MD, MPH Melissa Caughey , PhD Javed Butler, MD, MBA Gregg Fonarow , MD Justin Grodin , MD, MPH Darren McGuire, MD

16. Thank you ! ScientificSessions.org #AHA19

17.   Quintile 1 (n=1748) Quintile 2 (n=1700) Quintile 3 (n=1716) Quintile 4 (n=1702) Quintile 5 (n=1710) P value Age, years 63.4 (6.4) 63.1 (6.6) 62.7 (6.5) 62.5 (6.4) 61.3 (6.4) <0.001 Female 641 (36.7) 659 (38.8) 645 (37.6) 661 (38.8) 683 (39.9) 0.33 BMI, kg/m 2 31.9 (5.2) 31.7 (5.3) 32.0 (5.4) 32.3 (5.4) 32.8 (5.5) <0.001 Intensive glycemic control arm 1369 (78.3) 1077 (63.4) 852 (49.7) 553 (32.5) 415 (24.3) <0.001 Race           <0.001 Black 222 (12.7) 272 (16.0) 294 (17.1) 350 (20.6) 430 (25.1)   Hispanic 79 (4.5) 115 (6.8) 97 (5.7) 117 (6.9) 181 (10.6)   Other 223 (12.8) 232 (13.6) 197 (11.5) 203 (11.9) 169 (9.9)   White 1224 (70.0) 1081 (63.6) 1128 (65.7) 1032 (60.6) 930 (54.4)   Established CVD 499 (28.5) 522 (30.7) 572 (33.3) 576 (33.8) 585 (34.2) 0.001 Systolic blood pressure, mmHg 135.7 (16.3) 136.5 (16.6) 136.0 (16.2) 136.3 (17.0) 136.6 (17.4) 0.46 Serum creatinine, mg/dL 0.90 (0.2) 0.89 (0.2) 0.91 (0.2) 0.91 (0.2) 0.91 (0.2) 0.04 HgbA1c, mg/dL 8.0 (0.9) 8.1 (0.9) 8.3 (1.0) 8.4 (1.1) 8.6 (1.2) <0.001 Anithypertensive medication 1406 (80.5) 1407 (83.0) 1451 (84.7) 1415 (83.3) 1424 (83.6) 0.02 Insulin use 245 (15.2) 368 (23.7) 432 (28.3) 448 (29.5) 448 (29.4) <0.001 ASV, % 0.2 (0.05) 0.3 (0.03) 0.4 (0.04) 0.6 (0.06) 1.1 (0.44) <0.001

18.   Model 1 Model 2   HR (95% CI) P value HR (95% CI) P value <10% loss to <10% gain Referent Group ≥10% loss 1.84 (1.14, 2.95) 0.01 0.01 1.77 (1.09, 2.87) 0.02 ≥10% gain 1.98 (1.25, 3.12) 0.003 0.003 1.89 (1.18, 3.03) 0.008   <20% loss to <20% gain Referent Group ≥20% loss 2.44 (1.19, 4.99)   0.003 2.32 (1.12, 4.81)   0.008 ≥20% gain 2.08 (1.31, 3.60)   <0.001 1.97 (1.12, 3.49)   0.002 Association of baseline and percent change in HbA1c level with risk of incident HF. Percent change is from enrollment baseline to year 3 of follow-up.

19. HbA1c variability and risk of incident HF across different levels of baseline HbA1c (Panel A) and percent change in HbA1c (Panel B) and FPG variability and risk of incident HF across different levels of baseline FPG (Panel C) and percent change in FPG (Panel D).