Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jan 1.
Published in final edited form as: J Diabetes Complications. 2015 Sep 21;30(1):150–154. doi: 10.1016/j.jdiacomp.2015.09.009

Glycemic Variability During Algorithmic Titration of Insulin Among Hospitalized Patients with Type 2 Diabetes and Heart Failure

Kathleen Dungan a, Philip Binkley b,c, Kwame Osei a
PMCID: PMC4698074  NIHMSID: NIHMS730597  PMID: 26475503

Abstract

AIMS

The objective of this study is to assess hypoglycemia and glycemic variability (GV) in hospitalized patients with and without heart failure (HF) exacerbation.

METHODS

Hospitalized patients with type 2 diabetes (T2D) with (N=35) or without (N=16) HF who had hyperglycemia or significant insulin use were included. Subjects underwent continuous glucose monitoring during algorithmic titration of basal bolus insulin.

RESULTS

HF subjects had lower glucose coefficient of variation ([CV], 31+/−12 vs. 22+/−8.2, p=0.02), lower Low Blood Glucose Index (LBGI) and less hypoglycemia (25 vs. 2.6%, p=0.02), but similar mean glucose and glycemic lability index as non-HF subjects on day 1, but not on day 2. Sensor CV was correlated with hypoglycemia (ρ 0.32, p=0.02), HF status (ρ −0.35, p=0.013), T2D duration (ρ 0.29, p=0.04), insulin use prior to admission (ρ 0.42, p=0.002) and catecholamine levels. After controlling for differences in age, HbA1c, hypoglycemia, catecholamine levels, QT interval, and beta blocker use, only HF and diabetes duration or insulin use prior to admission were independent predictors of CV. HF had less robust associations with LBGI in multivariable models.

CONCLUSIONS

HF is not associated increased GV or hypoglycemia risk during initial titration of insulin. Further research is needed to determine prognostic implications.

Keywords: heart failure, hypoglycemia, glycemic variability, insulin, hospital

1.0 INTRODUCTION

Heart failure (HF) is a frequent comorbidity of diabetes that poses an enormous medical, societal and financial burden, affecting 5 million Americans, and leading to $27.9 billion in costs annually.[American Heart Association 2005] Diabetes is an independent predictor of mortality in patients with HF. [De Groote et al. 2004. Gustafsson et al. 2004] Experts recommend relaxed glucose targets among patients with significant comorbidities and an individualized approach based upon the perceived risk of hypoglycemia as well as the potential for adverse sequelae related to hypoglycemia.[Inzucchi et al. 2012, Ismail-Beigi et al. 2011] Hypoglycemia may be particularly concerning in HF patients, due to the predisposition for arrhythmias and ischemic events.[Kennel et al. 1998, Uretsky et al. 2000] However, glucose control has not been well characterized in patients with HF.

In patients with heart failure (HF), higher HbA1c has been associated with increased mortality in some studies, [Gerstein et al. 2008, Romero et al. 2013]. However, other data support a paradoxical [Auilar et al. 2009, Tomova et al. 2012] or J-shaped relationship [Eshaghian et al. 2006] between HbA1c and outcomes, indicating that hypoglycemia may mitigate possible benefits of lower HbA1c. Unfortunately, it can not be determined from these studies whether the low HbA1c per se is harmful, or even whether hypoglycemia plays a role. Furthermore, observations are confounded by non-glycemic factors, which may disproportionately affect the measurement of HbA1c in sicker patients.

Continuous glucose monitoring has the potential to uncover patterns in glucose control which are not captured by HbA1c. Measures of glycemic variability (GV) have garnered interest since numerous studies have demonstrated that increasing measures of GV are associated with higher mortality during critical illness, [Eslami et al. 2011] and possibly HF exacerbation.[Dungan et al. 2011] Patients with long-standing diabetes have increasing GV with both beta cell and counterregulatory hormone failure, and GV is a predictor of counterregulatory failure in response to hypoglycemia.[Murata et al. 2004, Alghothani and Dungan 2011] However, HF itself is characterized by profound neuroendocrine disturbances, and thus it is unclear if this may play a role in the development of hypoglycemia or GV.[Braunwald 2008, Jankowska et al. 2006, Niskanen et al. 2009] Assessing measures of GV may be useful for assessing hypoglycemia risk[Niskanen et al. 2009, Monnier et al 2011] which may limit the titration of therapies.

The objective of this study is to assess whether hypoglycemia or GV differs among hospitalized patients with or without HF exacerbation.

2.0 MATERIAL AND METHODS

2.1 Patients

Study subjects were enrolled as part of separate studies of hospitalized patients with type 2 diabetes, one in patients admitted with HF as the primary diagnosis and the other in subjects without a history of HF.[Dungan et al. 2014, Dungan et al. 2013a] Inclusion criteria for both studies included significant insulin use (>20 units/day) or hyperglycemia (BG >180 mg/dl [10 mmol/l] on at least 2 occasions separated by at least 4 hours apart). Exclusion criteria for both studies included type 1 diabetes, hyperglycemic emergency, critical illness (such as the need for mechanical ventilation and hypotension requiring vasopressors), corticosteroid use, end stage renal or liver disease, hospital stay expected to be less than 48 hr, inability to consent, prisoners and pregnancy. The HF study also excluded patients with acute myocardial infarction within the previous 3 months or predominantly right-sided heart failure. The non-HF group also excluded patients with arrhythmia or autonomic neuropathy. All studies were approved by the Institutional Review Board at the study institution and all patients signed informed consent.

2.2 Intervention

Patients were randomly assigned to intravenous (IV) or subcutaneous (SQ) insulin. However, due to differences in the IV insulin protocols, only subjects receiving SQ insulin could be analyzed for the current study. The SQ insulin algorithm was identical for both studies. All oral or non-insulin agents were discontinued. In insulin naïve patients the total daily dose of SQ insulin was 0.4 or 0.5 times the body weight in kg for an enrollment glucose of <180 mg/dl or >180 mg/dl respectively. In patients admitted on insulin, the total daily dose of SQ insulin was estimated as 100 or 120% of the total home dose of insulin in patients with an enrollment glucose of <180 mg/dl or >180 mg/dl respectively. Basal insulin was administered as approximately half of the estimated total daily dose of insulin. Prandial insulin was delivered according to carbohydrate intake as described previously.[Dungan et al. 2014, Dungan et al. 2013a] The target glucose range was 100–150 mg/dl and adjustments were made in the total daily insulin dose of +/−10–20% per day.

A continuous glucose monitor (CGMS Ipro®, Medtronic) was used in accordance with manufacturer instructions. The sensor was inserted on the abdomen and downloaded after at least 48 hours using CGMS solutions software. Capillary glucose values (Accu-Chek Inform®, Roche) were measured every 4–6 hours (before meals and bedtime when eating) in the SQ group. Calibrations were performed at 4 pre-determined time points each day (closest to 7AM, 11AM, 4PM and 9PM) within the allowable glucose limits (40–400 mg/dl) of the software. CGM data had a correlation of 0.88 (p-value <0.0001) and a mean absolute difference of 9.6% compared to capillary blood glucose assessments.[Dungan et al. 2013b]

2.3 Analysis

Glycemic variability was measured with the coefficient of variation (CV, standard deviation/mean glucose) and glycemic lability index (GLI, which is calculated by first finding the square of the difference between successive glucose measurements, dividing this value by the difference in time between measurements, and then calculating the sum of the quotients).[Ryan et al. 2004] Hypoglycemia was defined as a blood glucose 70 mg/dl (<3.9 mmol/l) due to the concern for low accuracy of continuous glucose monitoring in the hypoglycemic range.[Zijlstra et al. 2013] Due to the relatively low number of hypoglycemic events, a hypoglycemic risk score, the Low Blood Glucose Index (LBGI), was also calculated as reported previously using a transformed scale to correct the skewness of the glucose range.[Kovatchev et al. 1998]

The QT interval was obtained from patients who had a 12-lead electrocardiogram within 24 hours of enrollment. QT interval was corrected for heart rate, gender, and QRS interval as previously reported.[Rautaharju et al. 2004, Rautaharju et al. 2009] Change in plasma volume was calculated with the hemoglobin and hematocrit from successive days as published previously.[Kalra et al. 2002]

Continuous variables were reported as mean (standard deviation) or median (interquartile range) for normal and non-normal distributions respectively. Unpaired t-tests or Wilcoxon rank-sum tests were used to compare groups as appropriate. Dichotomous variables were reported as number (percentage) and between group comparisons were made using Fisher’s exact test. Statistical significance was determined at a p-value <0.05. Spearman’s correlation coefficients were calculated. Multiple linear regression analyses were performed for CV using least squares linear regression and backward stepwise methodology. Variables were chosen for entry into the model based upon univariable effect estimates (cut-off p-value of 0.1). Age and beta blockade were added to models due to baseline differences between groups. Statistical analyses were performed using JMP 10.0 software.

3.0 RESULTS

A total of 35 patients with HF and 16 patients without HF met the inclusion and exclusion criteria. Baseline characteristics, stratified by HF status, are presented in Table 1. Patients with HF were older (63 +/−12 vs. 55 +/−10.4 years, p=0.02), more likely to be on a beta blocker (90 vs. 44%, p=0.0007), and had lower HbA1c (7.7 +/−1.4 vs. 9.2 +/−2.5, p=0.04) than patients without HF. Patients with HF had higher norepinephrine (1167 +/−698 vs. 389 +/−264 pg/ml, p<0.0001), epinephrine (69 +−51 vs. 20 +/−14 pg/ml, p<0.0001), and corrected QT interval (342 +/− 10.2 vs. 360 +/− 27.5, p=0.008) compared to those without HF. Otherwise, baseline characteristics were similar.

Table 1.

Baseline Characteristics by Heart Failure Status

No Heart Failure Heart Failure P-value
Age 55 (10.4) 63 (12.4) 0.02
Male 12 (75%) 26 (70%) 0.75
Caucasian 14 (88%) 27 (74%) 0.47
Diabetes duration (years) 12 (7–17) 11 (9–20) 0.66
Coronary artery disease 6 (38%) 25 (64%) 0.08
Beta Blockade 7 (44%) 35 (90%) 0.0007
CHF treatment
 Dobutamine - N (%) - 6 (15%) -
 Milrinone - N (%) - 5 (13%) -
 Nesiratide - N (%) - 6 (15%) -
 Intravenous diuretic - days (IQR) - 4 (2–7) -
BNP^
 day 1 - 431 (224–976) -
 day 4 - 299 (135–713) -
Insulin prior to admission 13 (81%) 33 (85%) 0.71
 Insulin total dose day 1 (units) 58 (19–85) 35 (18–58) 0.19
 Insulin total dose day 2 (units) 43 (15–74) 38 (24–72) 0.49
Body mass index (kg/m2) 35 (7.4) 39 (8.8) 0.08
HbA1c (%) 9.2 (2.5) 7.7 (1.4) 0.04
HbA1c (mmol/mol) 77 (27.3) 61 (15.3)
Admit serum glucose (mg/dl) 212 (64) 175 (61) 0.06
Estimated GFR (mg/dl)* 57 (27) 50 (20) 0.35
Hemoglobin (gm/dl)
 day 1 11.6 (2.55) 11.3 (1.85) 0.65
 day 2 10.9 (2.17) 11.3 (1.78) 0.62
 %change plasma volume 2.45 (−3.73– 14.8) 0.59 (−0.61– 7.97) 0.31
Norepinephrine (pg/ml) 389 (264) 1167 (698) <0.0001
Epinephrine (pg/ml) 20 (14) 69 (51) <0.0001
QT interval† 342 (10.2) 360 (28) 0.008
Hospital length of stay (days) 7.5 (5.3–12.3) 8 (5–12) 0.55
Open in a new tab
^

P-value within group over time =0.0003.

*

Estimated GFR is assessed using the Modified Diet in Renal Disorders equation.

†

QT interval corrected for heart rate, gender, and QRS; N=10 without HF and N=24 with HF.

HF=heart failure, GFR=glomerular filtration rate.

HF subjects had lower glucose CV, (31+/−12 vs. 22+/−8.2, p=0.02), less hypoglycemia (25 vs. 2.6%, p=0.02), and tended to have lower LBGI compared to non-HF subjects overall (Table 1). Mean glucose and GLI were similar between CH and non-HF subjects. Daily differences were evident for CV, LBGI, and hypoglycemia on day 1 but not day 2.

Sensor CV was correlated with hypoglycemia (ρ 0.32, p=0.02), HF status (ρ −0.35, p=0.013), duration of diabetes (ρ 0.29, p=0.04), corrected QT interval (ρ −0.38, p=0.03) and catecholamine levels, but was not correlated with age, beta blocker use, body mass index, renal function or other variables (Table 2). GLI was not correlated with any of the variables analyzed (Table 2).

Table 2.

Glucose Measures by Heart Failure Status

Sensor glucose
 Mean (mg/dl) 156 (36) 166 (40) 0.37
  Mean day 1 155 (36) 169 (49) 0.27
  Mean day 2 156 (70) 164 (41) 0.68
 GLI ([mg/dl]2/hr*day−1) 0.94 (0.45–2.02) 0.75 (0.37–2.62) 0.37
  GLI day 1 0.88 (0.74–2.74) 0.66 (0.29–2.01) 0.15
  GLI day 2 0.32 (0.19–0.92) 0.51 (0.30–1.51) 0.30
 Coefficient of Variation (%) 31 (12) 22 (8.2) 0.019
  CV day 1 29 (13) 19 (8.5) 0.008
  CV day 2 16 (6.6) 19 (8.0) 0.13
 Low Blood Glucose Index (LBGI) 7.73 (1.2–12.2) 2.44 (1.05–5.93) 0.08
  LBGI day 1 8.42 (1.09–18.2) 2.29 (0.52–6.89) 0.04
  LBGI day 2 1.35 (0.51–11.4) 0.99 (0.58–3.30) 0.60
Hypoglycemia day 1–2 4 (25%) 1 (2.6%) 0.02
Day 1 hypoglycemia 4 (25%) 1 (2.6%) 0.02
Day 2 hypoglycemia 0 (0%) 1 (2.6%) >0.99
Open in a new tab

GLI=glycemic lability index. Hypoglycemia is defined as blood or interstitial fluid glucose <70 mg/dl.

LBGI was correlated with T2D duration (ρ 0.35, p=0.01), admission on insulin (ρ 0.28, p=0.042), epinephrine (ρ −0.35, p=0.01), renal function (ρ −0.33, p=0.02)), and QTc (ρ −0.40, p=0.02) (Table 2). LBGI showed only a trend for correlation with HF status (ρ −0.24, p=0.08).

Multivariable models were created using CV as the dependent variable and HF status as well as potential confounders or mediators as independent variables (Table 3). In model 1, HF status was the only significant independent variable (p=0.008) after controlling for diabetes duration, beta blockade, epinephrine level, and hypoglycemia. Following backward linear regression, both HF status and diabetes duration were independent predictors in the final model, such that HF was associated with 5.18% lower CV compared to non-HF status (p=0.001), and each year of diabetes duration was associated with a 0.35% increase in CV (p=0.01). After log-transformation of diabetes duration, only HF was a significant predictor (data not shown). In model 2, HF status (p=0.004) and insulin use prior to hospitalization (p=0.04) were independent predictors of CV, independent of age, beta blockade, and HbA1c. Following backward linear regression, both HF and insulin use prior to admission were independent predictors in the final model, such that HF was associated with a 4.36% lower CV compared to no HF (p=0.003), and insulin use was associated with 4.46% increase in CV compared to non-use (p=0.01). Finally, in model 3, corrected QT interval was not a significant predictor of CV after controlling for HF, beta blockade, and insulin use prior to admission.

Table 3.

Correlations with Glucose Variability

Coefficient of Variation Glycemic Lability Index Low Blood Glucose Index
Variable Spearman ρ P-value Spearman ρ P-value Spearman ρ P-value
Sensor Mean 0.18 0.19 0.36 0.0006 −0.16 0.28
Hypoglycemia day1–2 0.32 0.024 0.03 0.83 0.32 0.020
Heart failure −0.35 0.013 −0.13 0.37 −0.24 0.08
Age −0.03 0.85 −0.15 0.28 −0.08 0.59
Male −0.17 0.23 −0.20 0.15 −0.16 0.25
Caucasian 0.13 0.36 −0.05 0.74 −0.07 0.62
Diabetes duration 0.29 0.04 0.06 0.69 0.35 0.013
Beta blocker 0.02 0.90 −0.04 0.77 −0.10 0.47
Insulin prior to admission 0.42 0.002 −0.11 0.45 0.28 0.04
Body mass index −0.14 0.32 0.05 0.70 −0.04 0.78
HbA1c 0.19 0.18 −0.02 0.90 −0.07 0.63
Admission glucose 0.26 0.067 0.21 0.14 0.15 0.27
Norepinephrine −0.29 0.046 0.08 0.61 −0.11 0.44
Epinephrine −0.52 0.0002 −0.21 0.16 −0.35 0.01
Estimated GFR −0.09 0.52 0.04 0.81 −0.33 0.02
QTc-RR,QRS* −0.38 0.03 0.018 0.92 −0.40 0.02
Open in a new tab
*

QT interval was corrected for gender, heart rate, QRS interval.

GFR=glomerular filtration rate

Multivariable models were also conducted using LBGI as the dependent variable and HF status and other variables as independent variables (Table 4). In initial models adjusting for HF status, age, diabetes duration, insulin use prior to admission, epinephrine level, beta blockade, renal function, and QTc interval, HF was not a significant predictor. However, HF was a significant predictor of LBGI in final models following backward linear regression.

Table 4.

Multivariable Models for Coefficient of Variation

Final Model 1
Term Estimate SE P-value
Heart failure −5.18 1.44 0.001
Diabetes duration 0.35 0.14 0.01
Final Model 2
Term Estimate SE P-value
Heart failure −4.36 1.38 0.003
Insulin prior to admission 4.46 1.67 0.01
Final Model 3
Term Estimate SE P-value
Heart failure −4.36 1.38 0.003
Insulin prior to admission 4.46 1.67 0.01
Open in a new tab

Model 1 initially included heart failure, diabetes duration, beta blockade, epinephrine and hypoglycemia. Model 2 initally included age, heart failure, beta blockade, insulin prior to admission, and HbA1c. Model 3 initially included heart failure, insulin prior to admission, QTc (corrected for gender, heart rate, QRS interval) and beta blockade.

4.0 DISCUSSION

In this study, hospitalized patients with HF had lower GV, assessed as CV, on day 1 of titration of insulin compared to those without HF, even after controlling for other differences between groups. In addition, the crude frequency of hypoglycemia and LBGI, a measure of hypoglycemia risk, was lower in HF patients compared to non-HF patients. HF was a predictor of LBGI in final models but was not a significant predictor of LBGI in fully adjusted models.

Patients with HF have more risk factors for hypoglycemia, including renal failure; thus the results are somewhat unexpected. However, HF like many illness states, is marked by “decomplexification”, or increased regularity of physiologic rhythms such as heart rate and respiratory rate”.[Seely and Macklem 2004] Thus, there could be important interactions between physiologic parameters in various disease states that are not very well understood. For example, it is known that changes in glucose in controlled settings may alter autonomic tone, measured with heart rate variability and QT interval.[Takei et al 2007, Santini et al. 2007, Koivikko et al. 2005] In HF, these autonomic measures are abnormal at baseline and are blunted to outside stimuli.[Guzzetti et al 2001] In previous studies, IV insulin resulted in higher GV (measured with GLI) and greater sympathetic tone (measured with pre-ejection period) compared to SQ insulin in non-HF patients[Dungan et al 2013a], but the effect was very weak in a study of HF patients.[Dungan et al 2014] Thus, susceptibility of effects of autonomic perturbations on glucose fluctuations could be dampened in HF. In this study, increasing catecholamine levels were inversely correlated with CV, but there was no association after controlling for other variables, suggesting that other factors are more important for explaining differences in glycemic variability between HF and non-HF patients. However, tissue specific differences in autonomic reactivity may still play a role and are not completely assessed by circulating catecholamine levels.

It is possible that HF patients may have had less awareness of hypoglycemia, but the use of CGM should have identified asymptomatic hypoglycemia. CGM does have limitations in accuracy, particularly at low glucose values, [Choudhary et al. 2001] but the LBGI is reported to be a strong predictor of future severe hypoglycemia.[Kovatchev et al. 1998] Other studies have shown that GV is a predictor of hypoglycemia and may be a useful tool for hypoglycemia risk stratification.[Niskanen et al 2009] While cardiac autonomic neuropathy may increase the risk of severe hypoglycemia through impaired sympathoadrenal responses to hypoglycemia, [Cryer 2001] the autonomic responses to hypoglycemia in the setting of HF are unknown.

Hyperglycemia related treatment factors are expected to play a minimal role in explaining differences between HF and non-HF patients, as all patients were managed using a consistent insulin dosing algorithm and titration scale. While the use of beta blockers differed between groups, it was not significantly associated with GV or hypoglycemia. These findings are in agreement with literature supporting the general safety of beta blockade in patients with diabetes, [Murad et al 2009, Kerr et al. 1990] and additionally, provide reassurance of the safety of their use in insulin-requiring patients with T2D and HF. In fact, beta blockade has been shown to prevent the impaired autonomic response to hypoglycemia known to follow an antecedent hypoglycemic event.[Ramanathan and Cryer 2011] Several factors, including age, HbA1c, catecholamine levels, and QT interval differed by HF status, but none were significant predictors of CV or LBGI after controlling for HF and diabetes duration or insulin use. Unfortunately, QT interval was only available in a subset of patients. Thus, a larger sample size is needed to further evaluate this issue. Of note, the differences in hypoglycemia and GV dissipated after study day 1, suggesting that they are modified by other factors, such as changes in acuity of illness or active insulin titration. Thus, these findings warrant confirmation with other treatment regimens and in other settings.

Although these results are reassuring, it does not automatically follow that lower GV or hypoglycemia risk translates to improvement in outcomes in HF patients. In fact, it could be a sign of more advanced HF and portend worse outcomes. Nor can we conclude from this data that hypoglycemia, when it occurs, is more or less harmful in HF patients compared to those without HF. This will require future investigation, though our preliminary studies suggest possibly blunted autonomic responses in HF patients. In the meantime, providers should continue to individualize treatment goals based upon known factors that increase the risk related to hypoglycemia and the potential benefits of glucose control in a given individual.

Several limitations are worthy of mention. The sample size was small, the observation period was short, and titration was performed in an unblinded manner. Residual confounding by unmeasured factors is therefore still possible and these findings should be confirmed in larger studies of greater duration. Strengths include a novel patient population at high risk for potential adverse effects of hypoglycemia, the use of CGM measures for a more complete picture of glucose patterns, and consistent hyperglycemia therapy.

In conclusion, HF was not associated with increased GV or hypoglycemia risk in hospitalized patients with type 2 diabetes during algorithmic titration of basal bolus insulin. Further study is needed to verify these findings, identify mechanisms, and assess glucose targets.

Table 5.

Multivariable Models for Low Blood Glucose Index

Final Model 1
Term Estimate SE P-value
Heart failure −2.88 0.88 0.002
Insulin prior to admission 2.346 1.05 0.031
Final Model 2
Term Estimate SE P-value
Heart failure −2.88 0.88 0.002
Insulin prior to admission 2.34 1.05 0.031
Final Model 3
Term Estimate SE P-value
Heart failure −2.79 0.90 0.003
Open in a new tab

Model 1 initially included heart failure, insulin prior to admission, estimated glomerular filtration rate, epinephrine, and QT interval corrected for gender, heart rate, and QRS interval. Model 2 initially included age, heart failure, beta blocker, insulin prior to admission, and epinephrine. Model 3 initially included heart failure, diabetes duration, estimated glomerular filtration rate, epinephrine, and QT interval corrected for gender, heart rate, and QRS interval.

Acknowledgments

The project was supported by the OSU Clinical and Translational Research Center, Award Number UL1RR025755 from the National Center for Research Resources, by NIH grant numbers K23DK080891 and R21DK081877.

Footnotes

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health or other affiliated parties.

CONTRIBUTOR STATEMENTS

K.D. performed study design, researched the data, and wrote the manuscript, P.B. and K.O. assisted with study design, reviewed/edited the manuscript and contributed to the discussion.

AUTHOR DISCLOSURE STATEMENT

K.D. reports research funding from Merck, and Astra Zeneca, and advisory board activities with Eli Lilly and consulting activities with Glaxo Smith Kline. K.O. and P.B have nothing to report.

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. Aguilar D, Bozkurt B, Ramasubbu K, Deswal A. Relationship of hemoglobin A1C and mortality in heart failure patients with diabetes. J Am Coll Cardiol. 2009;54(5):422–8. doi: 10.1016/j.jacc.2009.04.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alghothani N, Dungan KM. The effect of glycemic variability on counterregulatory hormone responses to hypoglycemia in young children and adolescents with type 1 diabetes. Diabetes Technol Ther. 2011;13:1085–9. doi: 10.1089/dia.2011.0026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. American Heart Association. Heart Disease and Stroke Statistics: 2005 Update. Dallas, Tex: American Heart Association; 2005. [Google Scholar]
  4. Braunwald E. Biomarkers in Heart Failure. N Engl J Med. 2008;358:2148–2159. doi: 10.1056/NEJMra0800239. [DOI] [PubMed] [Google Scholar]
  5. Burger AJ, Aronson D. Blunted sympathetic response in diabetic patients with decompensated congestive heart failure. Int J Cardiol. 2001;81:243–249. doi: 10.1016/s0167-5273(01)00573-3. [DOI] [PubMed] [Google Scholar]
  6. Choudhary P, Lonnen K, Emery CJ, Freeman JV, McLeod KM, Heller SR. Relationship between interstitial and blood glucose during hypoglycemia in subjects with type 2 diabetes. Diabetes Technol Ther. 2011;13:1121–7. doi: 10.1089/dia.2011.0059. [DOI] [PubMed] [Google Scholar]
  7. Cryer PE. Hypoglycemia associated autonomic failure in diabetes. Am J Physiol Endocrinol metab. 2001;281:E1115–E1121. doi: 10.1152/ajpendo.2001.281.6.E1115. [DOI] [PubMed] [Google Scholar]
  8. De Groote P, Lamblin N, Mouquet F, et al. Impact of diabetes mellitus on long-term survival in patients with congestive heart failure. Eur Heart J. 2004;25:656–662. doi: 10.1016/j.ehj.2004.01.010. [DOI] [PubMed] [Google Scholar]
  9. Dungan K, Graessle K, Sagrilla C. The effect of congestive heart failure on sensor accuracy among hospitalized patients with type 2 diabetes. Diabetes Technol Ther. 2013;15:817–24. doi: 10.1089/dia.2013.0094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dungan KM, Binkley P, Nagaraja HN, Schuster DS, Osei K. Assessment of glycemic control and variability in patients hospitalized with congestive heart failure during the implementation of hospital-wide initiatives and effects on mortality. Diabetes Metabolism Research & Reviews. 2011;27(1):85–93. doi: 10.1002/dmrr.1155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dungan KM, Osei K, Gaillard T, Moore J, Binkley P. A comparison of continuous intravenous insulin and subcutaneous insulin among patients with type 2 diabetes and congestive heart failure exacerbation. Diabetes Metab Res Rev. 2014;31:93–101. doi: 10.1002/dmrr.2569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dungan KM, Osei K, Sagrilla C, Binkley P. Effect of the approach to insulin therapy on glycaemic fluctuations and autonomic tone in hospitalized patients with diabetes. Diabetes Obes Metab. 2013;15:558–63. doi: 10.1111/dom.12069. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Eshaghian S, Horwich TB, Fonarow GC. An unexpected inverse relationship between HbA1c levels and mortality in patients with diabetes and advanced systolic heart failure. Am Heart J. 2006;151:91. doi: 10.1016/j.ahj.2005.10.008. [DOI] [PubMed] [Google Scholar]
  14. Eslami S, Taherzadeh Z, Schultz MJ, Abu-Hanna A. Glucose variability measures and their effect on mortality: a systematic review. Intensive Care Med. 2011;37:583–93. doi: 10.1007/s00134-010-2129-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gerstein HC, Swedberg K, Carlsson J, et al. 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]
  16. Goldberger AL. Fractal variability versus pathologic periodicity: complexity loss and stereotypy in disease. Perspect Biol Med. 1997;40:543–61. doi: 10.1353/pbm.1997.0063. [DOI] [PubMed] [Google Scholar]
  17. Gustafsson I, Brendorp B, Seibaek M, et al. Danish Investigatord of Arrhythmia and Mortality on Dofetilde 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(5):771–7. doi: 10.1016/j.jacc.2003.11.024. [DOI] [PubMed] [Google Scholar]
  18. Guzzetti S, Magatelli R, Borroni E, Mezzetti S. Heart rate variability in chronic heart failure. Auton Neurosci. 2001;90:102–5. doi: 10.1016/S1566-0702(01)00274-0. [DOI] [PubMed] [Google Scholar]
  19. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycaemia in type 2 diabetes: a patient-centered approach. Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) Diabetologia. 2012;55(6):1577–96. doi: 10.1007/s00125-012-2534-0. [DOI] [PubMed] [Google Scholar]
  20. Ismail-Beigi F, Moghissi E, Tiktin M, Hirsch IB, Inzucchi SE, Genuth S. Individualizing glycemic targets in type 2 diabetes mellitus: implications of recent clinical trials. Ann Intern Med. 2011;154(8):554–9. doi: 10.7326/0003-4819-154-8-201104190-00007. [DOI] [PubMed] [Google Scholar]
  21. Jankowska EA, Ponikowski P, Piepoli MF, Banasiak W, Anker SD, Poole-Wilson PA. Autonomic imbalance and immune activation in chronic heart failure - pathophysiological links. Cardiovasc Res. 2006;70(3):434–45. doi: 10.1016/j.cardiores.2006.01.013. [DOI] [PubMed] [Google Scholar]
  22. Kalra PR, Anagnostopoulos C, Bolger AP, Coats AJ, Anker SD. The regulation and measurement of plasma volume in heart failure. J Am Coll Cardiol. 2002;39:1901–8. doi: 10.1016/s0735-1097(02)01903-4. [DOI] [PubMed] [Google Scholar]
  23. Kannel WB, Wilson PW, D’Agostino RB, Cobb J. Sudden coronary death in women. Am Heart J. 1998 Aug;136(2):205–12. doi: 10.1053/hj.1998.v136.90226. [DOI] [PubMed] [Google Scholar]
  24. Kerr D, MacDonald IA, Heller SR, Tattersall RB. Beta-adrenoceptor blockade and hypoglycaemia. A randomised, double-blind, placebo controlled comparison of metoprolol CR, atenolol and propranolol LA in normal subjects. Br J Clin Pharmacol. 1990;29:685–93. doi: 10.1111/j.1365-2125.1990.tb03689.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Koivikko ML, Salmela PI, Airaksinen KE, et al. Effects of sustained insulin-induced hypoglycemia on cardiovascular autonomic regulation in type 1 diabetes. Diabetes. 2005;54:744–750. doi: 10.2337/diabetes.54.3.744. [DOI] [PubMed] [Google Scholar]
  26. Kovatchev BP, Cox DJ, Gonder-Frederick LA, Young-Hyman D, Schlundt D, Clarke W. Assessment of risk for severe hypoglycemia among adults with IDDM: validation of the low blood glucose index. Diabetes Care. 1998;21:1870–5. doi: 10.2337/diacare.21.11.1870. [DOI] [PubMed] [Google Scholar]
  27. Monnier L, Wojtusciszyn A, Colette C, Owens D. The contribution of glucose variability to asymptomatic hypoglycemia in persons with type 2 diabetes. Diabetes Technol Ther. 2011;13:813–8. doi: 10.1089/dia.2011.0049. [DOI] [PubMed] [Google Scholar]
  28. Murad MH, Coto-Yglesias F, Wang AT, et al. Clinical review: Drug-induced hypoglycemia: a systematic review. J Clin Endocrinol Metab. 2009;94:741–5. doi: 10.1210/jc.2008-1416. [DOI] [PubMed] [Google Scholar]
  29. Murata GH, Duckworth WC, Shah JH, Wendel CS, Hoffman RM. Sources of glucose variability in insulin-treated type 2 diabetes: the Diabetes Outcomes in Veterans Study (DOVES) Clin Endocrinol (Oxf) 2004;60:451–6. doi: 10.1111/j.1365-2265.2004.02001.x. [DOI] [PubMed] [Google Scholar]
  30. Niskanen L, Virkamäki A, Hansen JB, Saukkonen T. Fasting plasma glucose variability as a marker of nocturnal hypoglycemia in diabetes: evidence from the PREDICTIVE study. Diabetes Res Clin Pract. 2009;86(2):e15–8. doi: 10.1016/j.diabres.2009.08.005. [DOI] [PubMed] [Google Scholar]
  31. Ramanathan R, Cryer PE. Adrenergic mediation of hypoglycemia-associated autonomic failure. Diabetes. 2011;60:602–6. doi: 10.2337/db10-1374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Rautaharju PM, Surawicz B, Gettes LS, et al. American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; American College of Cardiology Foundation; Heart Rhythm Society, AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol. 2009;53:982–91. doi: 10.1016/j.jacc.2008.12.014. [DOI] [PubMed] [Google Scholar]
  33. Rautaharju PM, Zhang ZM, Prineas R, Heiss G. Assessment of prolonged QT and JT intervals in ventricular conduction defects. Am J Cardiol. 2004;93:1017–21. doi: 10.1016/j.amjcard.2003.12.055. [DOI] [PubMed] [Google Scholar]
  34. Romero SP, Garcia-Egido A, Escobar MA, et al. Impact of new-onset diabetes mellitus and glycemic control on the prognosis of heart failure patients: a propensity-matched study in the community. Int J Cardiol. 2013;167(4):1206–16. doi: 10.1016/j.ijcard.2012.03.134. [DOI] [PubMed] [Google Scholar]
  35. Ryan EA, Shandro T, Green K, et al. Assessment of the severity of hypoglycemia and glycemic lability in type 1 diabetic subjects undergoing islet transplantation. Diabetes. 2004;53:955–962. doi: 10.2337/diabetes.53.4.955. [DOI] [PubMed] [Google Scholar]
  36. Santini V, Ciampittiello G, Gigli F, et al. QTc and autonomic neuropathy in diabetes: effects of acute hyperglycaemia and n-3 PUFA. Nutr Metab Cardiovasc Dis. 2007;17:712–718. doi: 10.1016/j.numecd.2006.09.006. [DOI] [PubMed] [Google Scholar]
  37. Seely AJ, Macklem PT. Complex systems and the technology of variability analysis. Crit Care. 2004;8:R367–84. doi: 10.1186/cc2948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Takei Y, Tomiyama H, Tanaka N, Yamashina A. Close relationship between sympathetic activation and coronary microvascular dysfunction during acute hyperglycemia in subjects with atherosclerotic risk factors. Circ J. 2007;71:202–6. doi: 10.1253/circj.71.202. [DOI] [PubMed] [Google Scholar]
  39. Tomova GS, Nimbal V, Horwich TB. Relation Between Hemoglobin A(1c) and Outcomes in Heart Failure Patients With and Without Diabetes Mellitus. Am J Cardiol. 2012;109(12):1767–73. doi: 10.1016/j.amjcard.2012.02.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Uretsky BF, Thygesen K, Armstrong PW, Cleland JG, Horowitz JD, Massie BM, Packer M, Poole-Wilson PA, Ryden L. Acute coronary findings at autopsy in heart failure patients with sudden death: results from the assessment of treatment with lisinopril and survival (ATLAS) trial. Circulation. 2000 Aug 8;102(6):611–6. doi: 10.1161/01.cir.102.6.611. [DOI] [PubMed] [Google Scholar]
  41. Zijlstra E, Heise T, Nosek L, Heinemann L, Heckermann S. Continuous glucose monitoring: quality of hypoglycaemia detection. Diabetes Obes Metab. 2013;15:130–135. doi: 10.1111/dom.12001. [DOI] [PubMed] [Google Scholar]

RESOURCES