Abstract
Background
Type 2 diabetes is an independent risk factor for heart failure development, but the relationship between incident heart failure and antecedent glycaemia has not been evaluated.
Methods and Results
The Go-DARTS study holds data for 8683 individuals with type 2 diabetes. Dispensed prescribing, hospital admission data and echocardiography reports were linked to extract incident heart failure cases from December 1998 to August 2011. All available HbA1c measures until heart failure development or end of study were used to model HbA1c time-dependently. Individuals were observed from study enrolment until heart failure development or end of study. Proportional hazard regression calculated heart failure development risk associated with specific HbA1c ranges accounting for comorbidities associated with heart failure, including blood pressure, body mass index and coronary artery disease. Seven hundred and one individuals with type 2 diabetes (8%) developed heart failure during follow up (mean 5.5 years, ±2.8 years). Time-updated analysis with longitudinal HbA1c showed that both HbA1c <6% (HR = 1.60, 95% CI 1.38-1.86 P-value= <0.0001) as well as HbA1c> 10% (HR = 1.80, 95% CI 1.60-2.16 P-value= <0.0001) were independently associated with the risk of HF.
Conclusions
Both high and low HbA1c predicted heart failure development in our cohort, forming a U-shaped relationship.
Keywords: diabetes mellitus, glucose, echocardiography
Type 2 diabetes mellitus (T2DM) is a recognised risk factor for heart failure (HF) development, independent of coronary artery disease (CAD) and hypertension (1). HF development is associated with high mortality, making prevention of HF in T2DM a research and clinical priority (2).
Glycaemic control may influence HF development. Hyperglycaemia can exert detrimental effects on the myocardium (3) and has been shown to up-regulate the renin-angiotensin-aldosterone system, increase oxidative stress (4, 5), promote accumulation of advanced glycation end-products and cause interstitial fibrosis (6, 7). Hyperglycaemia has also been associated with myocardial lipotoxicity, mitochondrial dysfunction, abnormal substrate metabolism and impaired calcium handling (8, 9). Epidemiological evidence indicated glycosylated haemoglobin (HbA1c) and the risk of HF are related (10-13). A cohort study of ~50 000 patients with T2DM showed each 1% increase in HbA1c was associated with an 8% increased risk of HF (10).
However, large randomized controlled trials (RCTs), investigating intensive versus standard glycaemic control and vascular outcomes, did not show any benefit of intensive glycaemic control (14, 15) and one study found intensive control promoted increased mortality (16). These studies were not specific to HF and were underpowered to draw conclusions regarding intensive glycaemic control and HF development. Meta-analysis of these RCTs has not shown any reduction in HF risk associated with intensive glycaemic control (17).
In view of these findings, we investigated four principal questions. Firstly, was time-updated HbA1c, a longitudinal measure of glycaemic control, associated with HF development in a population with T2DM in Tayside, Scotland? Secondly, how was HbA1c variability related to HF development, since HbA1c variability has previously been linked to incident cardiovascular events (18)? Thirdly, since glycaemic control has been associated with CAD and CAD is an accepted risk factor for HF development (19), was this association dependent on the existence of CAD? Lastly, what attributable risk did specific HbA1c ranges convey and what range defined ‘optimum control’?
Methods
We exploited the established regional clinical informatics systems developed by the University of Dundee and NHS Tayside. These systems make use of a unique health record identifier - the Community Health Index (CHI) number to deterministically link clinical datasets through robust anonymisation protocols within the Health Informatics Centre (HIC) at an individual level with high accuracy. The clinical datasets include:
(1) Go-DARTS database
The Diabetes Audit and Research in Tayside Study (DARTS) is a well-validated, region-wide clinical information system containing data on effectively all patients with diabetes in the region (20). The DARTS clinical systems facilitated the creation of the Genetics of Diabetes Audit and Research Tayside Scotland (Go-DARTS) database. Population-based recruitment to Go-DARTS, using the data within DARTS, allowed essentially all patients with T2DM in the region to be recruited from 1998, reducing selection bias. All Go-DARTS participants provided written consent for their data to be linked anonymously for research purposes. Many of these individuals were recruited under the auspices of the Wellcome Trust UK Type 2 Diabetes Case Control Consortium.
We studied 8683 individuals with T2DM recruited to Go-DARTS between December 1998 and May 2009, observing them from recruitment to the end of the study period, August 2011. The Go-DARTS population was described in previous studies (20-22). Collection and analysis of data in DARTS and Go-DARTS was approved by the East of Scotland Research and Ethics Committee.
(2) The Tayside echocardiography database
The Tayside echocardiography database, maintained by the Department of Cardiology, contains data from all clinically requested echocardiograms performed in Ninewells Hospital, Dundee from September 1993. All echocardiograms were reported by British Society of Echocardiography accredited echocardiographers. This database was linked to the Go-DARTS data to identify echocardiographically defined cases of left ventricular systolic dysfunction (LVSD). The echocardiography database was previously used to identify structural heart disease for research (23-26).
(3) Other HIC datasets
Other datasets utilised included the dispensed prescribing dataset, containing information of drug prescriptions dispensed for all individuals in Tayside over the past 20 years. Hospital discharge data comprising the Scottish Morbidity Record (SMR01) containing ICD 9 and 10 classification codes, mortality data from the General Registrar’s Office (GRO) and data containing information on biochemical test results including HbA1c measurements.
Study Design
To investigate how HbA1c levels predicted incident HF we performed a retrospective cohort study between December 1998 and August 2011 to ascertain the size of risk of specific levels of HbA1c with respect to the rate of HF development in T2DM, taking into account all risk factors associated with HF.
HF definition
Incident HF was defined as the first inpatient or outpatient diagnosis of HF according to a previously validated method (23, 26-28). All individuals with T2DM and a hospital admission with HF according to ICD codes (ICD 9: 428, ICD10: I50) or an echocardiogram with evidence of LVSD and a prescription for a loop-diuretic (BNF code 2.2.2) were regarded as HF cases. A previous case validation exercise found 91% concordance between a clinical diagnosis of HF from case note review and definition of HF based on an echocardiographic evidence of LVSD requiring prescription of loop diuretics (23, 27-28). Only incident HF cases were included in the analysis. Individuals who had not been admitted to hospital with HF, had an echocardiogram showing LVSD nor received a prescription for a loop diuretic were classed as non-HF cases in subsequent analysis.
Time-updated analysis with longitudinal HbA1c
HbA1c was modelled as a time-dependent variable where the follow-up time was split each time the Hba1c category changed. Thus, if subject 1 was in category 6-7% HbA1c, the subject would have added person years to the 6-7% HbA1c group. If the subject switched to 7-8%, the subject would then have also added patient-years to that HbA1c category.
Calculation of HbA1c variability
To investigate the effects of intra-individual variations of HbA1c on incident HF we calculated intra-individual mean (HbA1c-MEAN) and standard deviation (HbA1c-SD), respectively. HbA1c values obtained preceding recruitment, at the enrolment and until the last recorded value of HbA1c were used. The inter-individual difference in the number of HbA1c assessments was adjusted according to the formula: adj-HbA1c-SD = SD/√[n/(n-1)] as previously described (29).
Statistical Analysis
The Chi-squared test of association was used to identify discrete variables and Student’s t test was used to identify continuous variables associated with HF. Variables showing significant association with HF development were used to model HF by proportional hazard regression.
Covariates
All factors associated with HF contributed to the model, including age, gender, diabetes duration, number of HbA1c measures, systolic blood pressure (SBP), diastolic blood pressure (DBP), history of CAD, body mass index (BMI), anti-hyperglycaemic medications and creatinine. Creatinine was categorised in 10 umol/L increments due to the broad range of values. Age was divided by 10 due to the broad range of values. Values for BMI, SBP, DBP, creatinine, age and diabetes duration were taken at baseline.
Diagnosis of CAD was based on hospital admission with corresponding discharge code (ICD 9: 410, 411, 412, 413 or 414 and ICD10: I20, I21, I22, I23, I24 or I25). Existence of CAD was modelled in a time-dependent manner.
Prescriptions for anti-hyperglycaemic medications at study enrolment were analysed using BNF codes: 6.1.2.1 for sulphonylureas, 6.1.2.2 for metformin, 6.1.1.2 for insulin and the approved names ‘pioglitazone’ and ‘rosiglitazone’ for TZDs. All participants were assigned to one of 5 treatment categories. All participants taking insulin at enrolment were placed in the ‘insulin’ group. Patients taking sulphonylureas at enrolment, who were not taking insulin, were placed in the ‘suphonylureas’ group. Patients receiving TZDs at enrolment, but not taking insulin or sulphonylureas, were placed in the ‘TZDs’ group. Individuals prescribed metformin only were assigned to the ‘metformin’ group and those who were not receiving anti-hyperglycaemic agents at enrolment were placed in the ‘diet’ group.
Treatment category at enrolment was included in the proportional hazard regression model with those receiving diet-control as the reference group. In this analysis, time-updated HbA1C was divided into only 3 categories:< 6.5, 6.5-9 and >9%, for simplicity, 6.5-9% was used as the reference category.
Statistical analyses were performed using SAS 9.2 for Windows and the R software (30).
Results
Out of 8 683 individuals, a total of 701 (8%) developed HF during the follow up period (mean 5.5 years, standard deviation 2.8 years, Figure 1). These comprised 196 cases of HF based on echocardiography data and 505 based on hospital admission with HF.
Figure 1.
Flow chart showing method of case selection
Clinical features associated with HF
Variations in baseline clinical characteristics with HbA1c are shown in Table 1. Simple tests of association showed associations between HF development and number of HbA1c measures, age, gender, diabetes duration, CAD, SBP, DBP, BMI, creatinine and HbA1c levels (p<0.01 for all above variables except gender, p=0.02).
Table 1.
Variation in baseline clinical characteristics according to HbA1c levels and their association with HF development, n=1112 each quintile
| HbA1c quintile (HbA1c range, %) | First (2.4-6.5) | Second (6.5-7.0) | Third (7.0-7.4) | Fourth (7.4-8.2) | Fifth (8.2-18.3) |
|---|---|---|---|---|---|
| Number HbA1c measures (SD) | 16.0 (10.4) | 20.3 (10.6) | 21.9 (11.3) | 24.0 (12.8) | 23.4 (12.8) |
| Age, yrs (SD) | 66.9 (11.0) | 66.6 (10.4) | 64.5 (11.5) | 63.3 (11.7) | 59.9 (11.9) |
| Number male (%) | 669 (60) | 688 (61) | 683 (61) | 682 (61) | 637 (57) |
| Duration T2DM, yrs (SD) | 3.8 (5.0) | 5.3 (5.1) | 5.8 (5.6) | 6.9 (5.8) | 7.9 (6.6) |
| Frequency CAD (%) | 266 (23) | 241 (21) | 233 (21) | 268 (24) | 277 (24) |
| SBP, mmHg (SD) | 138 (19) | 140 (17) | 140 (17) | 140 (18) | 141 (19) |
| DBP, mmHg (SD) | 76 (11) | 77 (10) | 77 (10) | 78 (10) | 79 (10) |
| BMI Kg/m2 (SD) | 29.9 (5.8) | 30.4 (5.0) | 30.8 (5.6) | 30.7 (5.4) | 31.6 (6.1) |
| Creatinine, umol/L (SD) | 92.0 (32.1) | 91.1 (22.4) | 91.4 (32.5) | 90.3 (23.5) | 90.1 (27.1) |
| Insulin at enrolment (%) | 26 (2) | 36 (3) | 80 (7) | 185 (16) | 469 (42) |
| SUs at enrolment (%) | 122 (11) | 280 (25) | 429 (38) | 511 (46) | 407 (37) |
| Metformin at enrolment (%) | 182 (16) | 350 (32) | 356 (32) | 264 (24) | 122 (1) |
| TZDs at enrolment (%) | 9 (0.8) | 30 (3) | 40 (4) | 42 (4) | 25 (2) |
| Diet control at enrolment (%) | 772 (70) | 416 (37) | 207 (19) | 110 (10) | 88 (8) |
| Frequency HF (%) | 134 (12) | 93 (8) | 121 (11) | 134 (12) | 218 (20) |
Time-updated analysis with longitudinal HbA1c
Time-varying HbA1c analysis (Figure 2 and Table 2) showed an increased risk of HF for HbA1c categories <6%, 7-8%, 8-9% and >9% compared to the reference category of 6-7%. Risk was significantly increased in the categories <6% (HR = 1.60, 95% CI 1.38-1.80 P-value= <0.0001) and >10% groups (HR 1.80, 95% CI 1.60-2.16 P-value= <0.0001). This relationship persisted when baseline anti-hyperglycaemic medications were accounted for (<6.5% HR 1.48, 95% CI 1.23-1.80, P value <0.0001, >9% HR 1.31, 95% CI 1.06-1.61, P value 0.01, Table 3). Use of insulin and sulphonylureas at baseline were also associated with increased risk of HF development in this cohort.
Figure 2.
Relationship between time-updated HbA1c and the development of heart failure in type 2 diabetes, including age, duration of diabetes, gender, history of coronary artery disease, body mass index, creatinine, systolic and diastolic blood pressure as covariate.
Table 2.
Risk of heart failure development associated with varying levels of glycaemic control
| Variable | HR | 95% CI | P value |
|---|---|---|---|
| Age | 1.07 | 1.07-1.07 | <0.0001 |
| Duration T2DM, yrs | 1.02 | 1.02-1.04 | <0.0001 |
| Female gender | 1.01 | 0.90-1.20 | 0.60 |
| CAD | 1.63 | 4.98-5.3 | <0.0001 |
| BMI, kg/m2 | 1.08 | 1.07-1.11 | <0.0001 |
| Creatinine category, umol/L | 1.10 | 1.09-1.14 | <0.0001 |
| SBP, mmHg | 0.96 | 0.94-0.98 | 0.0003 |
| DBP, mmHg | 0.91 | 0.88-0.95 | <0.0001 |
| HbA1c <6% | 1.60 | 1.38-1.86 | <0.0001 |
| HbA1c 7-8% | 1.04 | 0.84-1.25 | 0.68 |
| HbA1c 8-9% | 1.07 | 0.82-1.33 | 0.57 |
| HbA1c 9-10% | 1.18 | 0.88-1.50 | 0.27 |
| HbA1c >10% | 1.80 | 1.60-2.16 | <0.0001 |
Abbreviations: HR= hazard ratio, CI= confidence interval, P-value= probability value, HbA1c= glycosylated haemoglobin, yrs=years, T2DM=type 2 diabetes mellitus, CAD =coronary artery disease, SBP=systolic blood pressure, DBP=diastolic blood pressure, BMI=body mass index.
Table 3.
Risk of heart failure development associated with varying levels of glycaemic control accounting for baseline anti-hyperglycaemic medications
| Variable | HR | 95% CI | P value |
|---|---|---|---|
| Age | 1.08 | 1.07-1.09 | <0.0001 |
| Duration T2DM, yrs | 1.01 | 1.00-1.02 | 0.10 |
| CAD | 5.21 | 4.40-6.17 | <0.0001 |
| BMI, kg/m2 | 1.08 | 1.07-1.10 | <0.0001 |
| Creatinine category, umol/L | 1.10 | 1.07-1.14 | <0.0001 |
| SBP, mmHg | 0.97 | 0.95-0.99 | 0.0035 |
| DBP, mmHg | 0.93 | 0.90-0.97 | 0.0003 |
| HbA1c <6.5% | 1.48 | 1.23-1.80 | <0.0001 |
| HbA1c >9% | 1.31 | 1.06-1.61 | <0.0001 |
| Insulin | 1.81 | 1.43-2.29 | <0.0001 |
| Sulphonylureas | 1.17 | 1.00-1.42 | 0.11 |
| TZDs | 0.97 | 0.67-1.39 | 0.85 |
| Metformin | 1.10 | 0.93-1.30 | 0.28 |
Abbreviations: HR= hazard ratio, CI= confidence interval, P-value= probability value, HbA1c= glycosylated haemoglobin, yrs=years, T2DM=type 2 diabetes mellitus, CAD =coronary artery disease, SBP=systolic blood pressure, DBP=diastolic blood pressure, BMI=body mass index, TZDs=thiazolidinediones
HbA1c variability and incident HF
Intra-individual mean (HbA1c-MEAN) and its adjusted Standard deviation (HbA1c-SD) were included in the proportional hazard model as above. HbA1c variability (HbA1c-SD) was independently and significantly associated with incident HF (HR=0.80 CI 0.74-0.85, P-value<0.0001), with less variability in HbA1c having a protective effect on incident HF.
Assumptions of proportional hazard models
The assumption of proportional hazard with time dependent covariates were tested by creating interactions of the predictors as a function of survival time and including these values in the model using the function “Proportionality test” in Proc Phreg in SAS 9.2. Test of proportional hazards did not yield significant P-values for any of the covariates in the model or globally, suggesting the proportional hazard assumptions hold true.
Discussion
The U-shaped relationship between HbA1c and HF development
Our study generated 3 main findings; firstly, HbA1c >10% was associated with HF development in a progressive non-linear relationship even when CAD was accounted for. Secondly, HbA1c<6% was also associated with an increased HF risk. These findings suggest both high and low HbA1c predicted HF development in our cohort, forming the U-shaped relationship, shown in Figure 2. Thirdly, intra-individual variability in HbA1c was an independent risk factor for development of HF.
Currie et al (31) previously reported a U-shaped relationship between mean HbA1c and all-cause mortality and cardiac events in a retrospective study of patients with T2DM utilizing the UK General Practice Research Database (GPRD, 32). They showed HbA1c around 7.5% was associated with the lowest all-cause mortality and lowest progression to large vessel disease events (myocardial infarction, stroke, coronary revascularisation or angina), although they did not look at HF development. Our study extends these findings, showing the U-shaped relationship between mean HbA1c and cardiovascular risk applies to HF development.
Eight percent of our cohort of patients with T2DM developed HF over 5.5 years, which compares well with previous studies (11). This is 53 times higher than Scotland’s general population (approximately 0.15%). Echocardiographic and hospital admission data were used in our study whilst hospital admission data alone generated the Scotland-wide figure quoted (33).
We identified a number of risk factors for HF development in T2DM including increasing age, BMI, SBP and creatinine at baseline and a history of CAD. These findings are consistent with previous work (12).
Early observational studies showed raised HbA1c was associated with increased risk of HF development, where HF was identified through hospital admission data, regardless of CAD status (10, 13). These studies used a single measure of HbA1c, although calculation of the mean of serial HbA1c measures has been shown to better predict diabetic complications (34).
Lind and colleagues used the Swedish National Diabetes Registry to explore the relationship between mean HbA1c and the risk of HF and reported that individuals with mean HbA1c> 7% were at increased risk of developing HF. They reported ‘no further reduction in risk’ in patients with HbA1c<6% (42mmol/mol) compared with those with HbA1c 6-7%. However, they demonstrated an increased risk of HF hospitalisation in patients with mean HbA1c<6% compared to mean HbA1c 6-7% in their model accounting for the greatest number of confounding factors (11). These findings support a U-shaped relationship between HbA1c level and incident HF risk, in-keeping with the UK GPRD study (31). This fits with data on mortality and vascular events from recent trials investigating intensive versus standard glycaemic control and vascular outcomes (14-16). The relationship between HF development and HbA1c was not explained by an association between CAD and HbA1c, and HbA1c still predicts HF development independently of CAD. This is consistent with the findings described above (13).
Why was HbA1c<6% associated with HF development?
The U-shaped relationship between HbA1c and the risk of developing HF merits discussion. Multiple factors may promote HF development in patients with HbA1c<6%. Adverse ‘off-target’ drug effects from anti-hyperglycaemic medications may account for some of the incident HF cases in the group of individuals with HbA1c<6%.
Insulin and TZDs have been associated with HF development (35, 36). Only 0.9% of individuals with HbA1c<6% were taking TZDs so we were unable to assess whether their use was associated with increased risk of HF development in this group. Although sulphonylurea use was associated with HF development in our study, previous studies have provided conflicting evidence regarding the risk of HF development with sulphonylurea use. An early study using the GPRD database on 25,690 diabetics, showed the incidence of HF did not differ across oral anti-diabetic drug classes (37).
However, more recent analyses of observational data have shown an increased risk of HF with sulphonylurea use. McAlister and colleagues reported that users of high-dose sulphonylureas were more likely to develop incident HF than users of high-dose metformin monotherapy (HR 1.38, 95% CI 1.20–1.60, 38). A more recent analysis of the UK GPRD database of 91,521 patients with diabetes showed second-generation sulphonylureas were associated with an 18% higher risk of developing HF (HR 1.18, 95% CI 1.04-1.34, 39). It is impossible to draw firm conclusions regarding the role of anti-hyperglycaemic agents from this observational study. A randomized control trial would be required to resolve this issue, which may not be ethically feasible given the data from preceding RCTs (14-16). Hypoglycaemic events may also partially explain the U-shaped relationship as hypoglycaemia is an independent cardiovascular risk factor (40). We were unable to analyse the influence of hypoglycaemic events from our data so discussion regarding their importance is speculative.
Previous reports had linked HbA1c variability to the risk of microvascular complications of both type 1 (39, 40) and type 2 diabetes (41). Retrospective analyses of the DCCT (41, 42) showed variation in HbA1c was an independent risk factor for the development of both diabetic retinopathy and nephropathy in individuals with type 1 diabetes. In the FinnDiane study, HbA1c variability also predicted incident cardiovascular events (18) although the link with cardiovascular events was not a consistent finding.
We examined the association of intra-individual HbA1c variability and incident HF and showed that lower variability in HbA1c had was relatively protective against incident HF. The physiological mechanisms through which HbA1c variability contributes to HF development are not known. However, increased glucose variability has been reported to have promote cell apoptosis and oxidative stress (43-45). Both these processes can contribute to the pathogenesis of HF (46). The suggestion that HbA1c variability enhances cell apoptosis and oxidative stress in patients with diabetes remains speculative and cannot be concluded from this study.
Study limitations
Our study reflects a true population and a real-world scenario. We recognise the inherent limitations that come with a retrospective, non-randomised observational study. A potential limitation of our study was that we did not model use of anti-hyperglycaemic medication time-dependently and so cannot determine whether the association between anti-hyperglycaemic medications and HF development depended on medication category, medication-dose or the duration of drug exposure. However, in this observational cohort, modelling the impact of anti-hyperglycaemic treatment category as well as the inclusion of other cardiovascular protective medication would have been prohibitively complex, both to analyse and interpret.
Another limitation of our study was that we were unable to take into account hospitalizations occurring outside of Scotland or account for other potential confounding factors such as atrial fibrillation, smoking status, lipids and QT interval.
However, our study had multiple strengths. The primary strength of our analysis, relative to the observational studies described above, lies in the use of echocardiographic data to identify individuals with incident HF.
Conclusion
Both high and low HbA1c were associated with increased risk of HF development in T2DM, forming a U-shaped relationship. The association between HbA1c and HF development does not depend on the association between HbA1c and CAD. Sulphonylurea drugs and insulin also appear to be associated with HF development when compared to diet-control.
Supplementary Material
Acknowledgements
We acknowledge the support of the Health Informatics Centre, University of Dundee for managing and supplying the anonymised data and NHS Tayside, the original data owner. We are grateful to all the participants who took part in the Go-DARTS study, to the general practitioners, to the Scottish School of Primary Care for their help in recruiting the participants, and to the whole team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses.
Sources of Funding
The Wellcome Trust U.K. type 2 diabetes case control collection (Go-DARTS2) was funded by the Wellcome Trust (Award GR02960).
HMP’s salary is funded by the EC Seventh Framework Programme (FP7)-Co-operation-Health, reference number 242209.
Footnotes
Disclosures
Professor Chim Lang has consulted on saxagliptin (Astra Zeneca).
All other authors have no disclosures.
References
- 1.Gottdiener JS, Arnold AM, Aurigemma GP, Polak JF, Tracy RP, Kitzman DW, Gardin JM, Rutledge JE, Boineau RC. Predictors of congestive heart failure in the elderly: the Cardiovascular Health Study. J Am Coll Cardiol. 2000;35:1628–37. doi: 10.1016/s0735-1097(00)00582-9. [DOI] [PubMed] [Google Scholar]
- 2.Peterson L, McKenzie C, Schaffer J. Diabetic Cardiovascular Disease: Getting to the Heart of the Matter. J of Cardiovasc Trans Res. 2012;5:436–45. doi: 10.1007/s12265-012-9374-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Boudina S, Abel ED. Diabetic cardiomyopathy revisited. Circulation. 2007;115:3213–23. doi: 10.1161/CIRCULATIONAHA.106.679597. [DOI] [PubMed] [Google Scholar]
- 4.Cohn JN, Johnson G, Ziesche S, Cobb F, Francis G, Tristani F, Smith R, Dunkman WB, Loeb H, Wong M, Bhat G, Goldman S, Fletcher RD, Doherty J, Hughes CV, Carson P, Cintron G, Shabetai R, Haakenson C. A comparison of enalapril with hydralazine-isosorbide dinitrate in the treatment of chronic congestive heart failure. N Engl J Med. 1991;325:303–10. doi: 10.1056/NEJM199108013250502. [DOI] [PubMed] [Google Scholar]
- 5.Wold LE, Ceylan-Isik AF, Fang CX, Yang X, Li SY, Sreejayan N, Privratsky JR, Ren J. Metallothionein alleviates cardiac dysfunction in streptozotocin-induced diabetes: role of Ca2+ cycling proteins, NADPH oxidase, poly(ADP-Ribose) polymerase and myosin heavy chain isozyme. Free Radic Biol Med. 2006;40:1419–29. doi: 10.1016/j.freeradbiomed.2005.12.009. [DOI] [PubMed] [Google Scholar]
- 6.Ritchie RH, Quinn JM, Cao AH, Drummond GR, Kaye DM, Favaloro JM, Proietto J, Delbridge LM. The antioxidant tempol inhibits cardiac hypertrophy in the insulin-resistant GLUT4-deficient mouse in vivo. J Mol Cell Cardiol. 2007;42:1119–28. doi: 10.1016/j.yjmcc.2007.03.900. [DOI] [PubMed] [Google Scholar]
- 7.Miki T, Yuda S, Kouzu H, Miura T. Diabetic cardiomyopathy: pathophysiology and clinical features. Heart Fail Rev. 2012;18:149–66. doi: 10.1007/s10741-012-9313-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhou YT, Grayburn P, Karim A, Shimabukuro M, Higa M, Baetens D, Orci L, Unger RH. Lipotoxic heart disease in obese rats: implications for human obesity. Proc Natl Acad Sci U S A. 2000;97:1784–9. doi: 10.1073/pnas.97.4.1784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Trost SU, Belke DD, Bluhm WF, Meyer M, Swanson E, Dillmann WH. Overexpression of the sarcoplasmic reticulum Ca(2+)-ATPase improves myocardial contractility in diabetic cardiomyopathy. Diabetes. 2002;51:1166–71. doi: 10.2337/diabetes.51.4.1166. [DOI] [PubMed] [Google Scholar]
- 10.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;103:2668–73. doi: 10.1161/01.cir.103.22.2668. [DOI] [PubMed] [Google Scholar]
- 11.Lind M, Olsson M, Rosengren A, Svensson AM, Bounias I, Gudbjornsdottir S. The relationship between glycaemic control and heart failure in 83,021 patients with type 2 diabetes. Diabetologia. 2012;55:2946–53. doi: 10.1007/s00125-012-2681-3. [DOI] [PubMed] [Google Scholar]
- 12.Nichols GA, Joshua-Gotlib S, Parasuraman S. Independent contribution of A1C, systolic blood pressure, and LDL cholesterol control to risk of cardiovascular disease hospitalizations in type 2 diabetes: an observational cohort study. J Gen Intern Med. 2013;28:691–7. doi: 10.1007/s11606-012-2320-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pazin-Filho A, Kottgen A, Bertoni AG, Russell SD, Selvin E, Rosamond WD, Coresh J. HbA 1c as a risk factor for heart failure in persons with diabetes: the Atherosclerosis Risk in Communities (ARIC) study. Diabetologia. 2008;51:2197–204. doi: 10.1007/s00125-008-1164-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D, Hamet P, Harrap S, Heller S, Liu L, Mancia G, Mogensen CE, Pan C, Poulter N, Rodgers A, Williams B, Bompoint S, de Galan BE, Joshi R, Travert F. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358:2560–72. doi: 10.1056/NEJMoa0802987. [DOI] [PubMed] [Google Scholar]
- 15.Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren SR, Goldman S, McCarren M, Vitek ME, Henderson WG, Huang GD, VADT Investigators Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med. 2009;360:129–39. doi: 10.1056/NEJMoa0808431. [DOI] [PubMed] [Google Scholar]
- 16.Gerstein HC, Miller ME, Byington RP, Goff DC, Jr, Bigger JT, Buse JB, Cushman WC, Genuth S, Ismail-Beigi F, Grimm RH, Jr, Probstfield JL, Simons-Morton DG, Friedewald WT. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008;358:2545–59. doi: 10.1056/NEJMoa0802743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Castagno D, Baird-Gunning J, Jhund PS, Biondi-Zoccai G, MacDonald MR, Petrie MC, Gaita F, McMurray JJ. Intensive glycemic control has no impact on the risk of heart failure in type 2 diabetic patients: evidence from a 37,229 patient meta-analysis. Am Heart J. 2011;162:938–48 e2. doi: 10.1016/j.ahj.2011.07.030. [DOI] [PubMed] [Google Scholar]
- 18.Wadén J, Forsblom C, Thorn LM, Gordin D, Saraheimo M, Groop PH. Finnish Diabetic Nephropathy Study Group. A1C variability predicts incident cardiovascular events, microalbuminuria, and overt diabetic nephropathy in patients with type 1 diabetes. Diabetes. 2009;12:2649–2655. doi: 10.2337/db09-0693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Turner RC, Millns H, Neil HA, Stratton IM, Manley SE, Matthews DR, Holman RR. Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom Prospective Diabetes Study. BMJ. 1998;316:823–8. doi: 10.1136/bmj.316.7134.823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Morris AD, Boyle DI, MacAlpine R, Emslie-Smith A, Jung RT, Newton RW, MacDonald TM. The diabetes audit and research in Tayside Scotland (darts) study: electronic record linkage to create a diabetes register. BMJ. 1997;315:524–8. doi: 10.1136/bmj.315.7107.524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Doney ASF, Fischer B, Leese G, Morris AD, Palmer CNA. Cardiovascular Risk in Type 2 Diabetes Is Associated With Variation at the PPARG Locus. Arteriosclerosis, Thrombosis, and Vascular Biology. 2004;24:2403–7. doi: 10.1161/01.ATV.0000147897.57527.e4. [DOI] [PubMed] [Google Scholar]
- 22.Zhou K, Bellenguez C, Spencer CC, Bennett AJ, Coleman RL, Tavendale R, Hawley SA, Donnelly LA, Schofield C, Groves CJ, Burch L, Carr F, Strange A, Freeman C, Blackwell JM, Bramon E, Brown MA, Casas JP, Corvin A, Craddock N, Deloukas P, Dronov S, Duncanson A, Edkins S, Gray E, Hunt S, Jankowski J, Langford C, Markus HS, Mathew CG, Plomin R, Rautanen A, Sawcer SJ, Samani NJ, Trembath R, Viswanathan AC, Wood NW, MAGIC investigators. Harries LW, Hattersley AT, Doney AS, Colhoun H, Morris AD, Sutherland C, Hardie DG, Peltonen L, McCarthy MI, Holman RR, Palmer CN, Donnelly P, Pearson ER. Common variants near ATM are associated with glycemic response to metformin in type 2 diabetes. Nat Genet. 2011;43:117–20. doi: 10.1038/ng.735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Elder DHJ, Donnelly L. Wong A 011 HbA1c and mortality in diabetic individuals with heart failure: an observational cohort study. Heart. 2012;98:A9–A10. [Google Scholar]
- 24.Doney AS, Lee S, Leese GP, Morris AD, Palmer CN. Increased cardiovascular morbidity and mortality in type 2 diabetes is associated with the glutathione S transferase theta-null genotype: a Go-DARTS study. Circulation. 2005;111:2927–34. doi: 10.1161/CIRCULATIONAHA.104.509224. [DOI] [PubMed] [Google Scholar]
- 25.Nadir MA, Wei L, Elder DH, Libianto R, Lim TK, Pauriah M, Pringle SD, Doney AD, Choy AM, Struthers AD, Lang CC. Impact of renin-angiotensin system blockade therapy on outcome in aortic stenosis. J Am Coll Cardiol. 2011;58:570–6. doi: 10.1016/j.jacc.2011.01.063. [DOI] [PubMed] [Google Scholar]
- 26.Elder DH, Wei L, Szwejkowski BR, Libianto R, Nadir A, Pauriah M, Rekhraj S, Lim TK, George J, Doney A, Pringle SD, Choy AM, Struthers AD, Lang CC. The impact of renin-angiotensin-aldosterone system blockade on heart failure outcomes and mortality in patients identified to have aortic regurgitation: a large population cohort study. J Am Coll Cardiol. 2011;58:2084–91. doi: 10.1016/j.jacc.2011.07.043. [DOI] [PubMed] [Google Scholar]
- 27.Elder DHJ, Shearer F, Dawson A, Pradeep M, Parry H, Currie P, Pringle SD, George J, Choy A, Lang C. 012 Automated data capture from echocardiography reports to enhance heart failure population research. Heart. 2012;98(Supplement 1):A10. abstr. [Google Scholar]
- 28.Szwejkowski BR, Elder DH, Shearer F, Jack D, Choy AM, Pringle SD, Struthers AD, George J, Lang CC. Pulmonary hypertension predicts all-cause mortality in patients with heart failure: a retrospective cohort study. Eur J Heart Fail. 2012;14:162–7. doi: 10.1093/eurjhf/hfr159. [DOI] [PubMed] [Google Scholar]
- 29.Penno G, Solini A, Bonora E, Fondelli C, Orsi E, Zerbini G, Morano S, Cavalot F, Lamacchia O, Laviola L, Nicolucci A, Pugliese G. HbA1c Variability as an Independent Correlate of Nephropathy, but Not Retinopathy, in Patients With Type 2 Diabetes The Renal Insufficiency And Cardiovascular Events (RIACE) Italian Multicenter Study. Diabetes Care. 2013;36:2301–1. doi: 10.2337/dc12-2264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gentleman RIaR. R: A Language for Data Analysis and Graphics. Journal of Computational and Graphical Statistics. 1996;5:299–314. [Google Scholar]
- 31.Currie CJ, Peters JR, Tynan A, Evans M, Heine RJ, Bracco OL, Zagar T, Poole CD. Survival as a function of HbA(1c) in people with type 2 diabetes: a retrospective cohort study. Lancet. 2010;375:481–9. doi: 10.1016/S0140-6736(09)61969-3. [DOI] [PubMed] [Google Scholar]
- 32.Database GPR. General Practice Research Database. 2009. [Google Scholar]
- 33.Stewart S, MacIntyre K, Capewell S, McMurray JJV. Heart failure and the aging population: an increasing burden in the 21st century? Heart. 2003;89:49–53. doi: 10.1136/heart.89.1.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lind M, Oden A, Fahlen M, Eliasson B. The true value of HbA1c as a predictor of diabetic complications: simulations of HbA1c variables. PLoS ONE. 2009;4:e4412. doi: 10.1371/journal.pone.0004412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.McGuire DK, See R, Abdullah SM, Snell PG, McGavock JM, Ayers CR, Szczepaniak LS. The effect of rosiglitazone on integrated cardiovascular performance, cardiac structure, function and myocardial triglyceride: trial design and rationale. Diab Vasc Dis Res. 2009:43–50. doi: 10.3132/dvdr.2009.009. [DOI] [PubMed] [Google Scholar]
- 36.UK Prospective Diabetes Study (UKPDS) Group Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33) Lancet. 1998;352:837–53. [PubMed] [Google Scholar]
- 37.Maru S, Koch GG, Stender M, Clark D, Gibowski L, Petri H, White AD, Simpson RJ., Jr. Antidiabetic drugs and heart failure risk in patients with type 2 diabetes in the UK primary care setting. Diabetes Care. 2005;28:20–26. doi: 10.2337/diacare.28.1.20. [DOI] [PubMed] [Google Scholar]
- 38.McAlister FA, Eurich DT, Majumdar SR, Johnson JA. The risk of heart failure in patients with type 2 diabetes treated with oral agent monotherapy. Eur J Heart Fail. 2008;10:703–708. doi: 10.1016/j.ejheart.2008.05.013. [DOI] [PubMed] [Google Scholar]
- 39.Tzoulaki I, Molokhia M, Curcin V, Little MP, Millett CJ, Ng A, Hughes RI, Khunti K, Wilkins MR, Majeed A, Elliott P. Risk of cardiovascular disease and all cause mortality among patients with type 2 diabetes prescribed oral antidiabetes drugs: retrospective cohort study using UK general practice research database. BMJ. 2009;339:b4731. doi: 10.1136/bmj.b4731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Frier BM, Schernthaner G, Heller SR. Hypoglycemia and cardiovascular risks. Diabetes Care. 2011;34:S132–7. doi: 10.2337/dc11-s220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kilpatrick ES, Rigby AS, Atkin SL. A1C variability and the risk of microvascular complications in type 1 diabetes: data from the Diabetes Control and Complications Trial. Diabetes Care. 2008;12:2198–202. doi: 10.2337/dc08-0864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hsu CC, Chang HY, Huang MC, Hwang SJ, Yang YC, Lee YS, Shin SJ, Tai TY. HbA(1c) variability is associated with microalbuminuria development in type 2 diabetes: a 7-year prospective cohort study. Diabetologia. 2012;12:3163–3172. doi: 10.1007/s00125-012-2700-4. [DOI] [PubMed] [Google Scholar]
- 43.Li W, Liu X, Yanoff M, Cohen S, Ye X. Cultured retinal capillary pericytes die by apoptosis after an abrupt fluctuation from high to low glucose levels: a comparative study with retinal capillary endothelial cells. Diabetologia. 1996;39:537–47. doi: 10.1007/BF00403300. [DOI] [PubMed] [Google Scholar]
- 44.Quagliaro L, Piconi L, Assaloni R, Martinelli L, Motz E, Ceriello A. Intermittent high glucose enhances apoptosis related to oxidative stress in human umbilical vein endothelial cells: the role of protein kinase C and NAD(P)H-oxidase activation. Diabetes. 2003;52:2795–804. doi: 10.2337/diabetes.52.11.2795. [DOI] [PubMed] [Google Scholar]
- 45.Monnier L, Mas E, Ginet C, Michel F, Villon L, Cristol JP, Colette C. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycaemia in patients with type 2 diabetes. JAMA. 2006;295:1681–7. doi: 10.1001/jama.295.14.1681. [DOI] [PubMed] [Google Scholar]
- 46.Tutsui H, Kinugawa S, Matsushima S. Oxidative stress and heart failure. Am J Physiol Heart Circ Physiol. 2011;301:H2181–90. doi: 10.1152/ajpheart.00554.2011. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.