Hemoglobin A1C in non-diabetic patients: An independent predictor of coronary artery disease and its severity

diabetes research and clinical practice 102 (2013) 225–232

Contents available at ScienceDirect

Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres

Hemoglobin A1C in non-diabetic patients: An independent predictor of coronary artery disease and its severity Haleh Ashraf a, Mohammad Ali Boroumand b, Alireza Amirzadegan a, Shaghayegh Ashraf Talesh c, Gholamreza Davoodi a,* a

Department of Cardiology, Tehran Heart Center, Tehran University of Medical Sciences, North Kargar Street, Tehran, Iran b Tehran Heart Center, Tehran University of Medical Sciences, North Kargar Street, Tehran, Iran c General Practitioner, Free Researcher, Iran

article info

abstract

Article history:

Aims: To determine the association between glycated hemoglobin (HbA1c) and angiograph-

Received 17 May 2013

ically proven coronary artery disease (CAD) and its severity in nondiabetic individuals.

Received in revised form

Methods: We enrolled 299 consecutive individuals undergoing coronary angiography for

27 July 2013

suspected ischemia. Patients were included if they had no history of prior revascularization

Accepted 2 October 2013

or diabetes mellitus and had fasting blood glucose < 126 mg/dl (7.0 mmol/l) and

Available online 9 October 2013

HbA1c < 6.5% (47 mmol/mol). The severity of the CAD was also evaluated using the Gensini score. Serum HbA1c (NGSP certified Method), highly sensitive C-reactive protein (hsCRP),

Keywords:

lipid profile, insulin and APO lipoprotein A1 and B100 levels were measured.

Diabetes mellitus

Results: Mean age was 58.8  10.4 year; 60.9% men. One hundred forty seven patients had

Haemoglobin A1c

significant CAD (50% stenosis in any major vessel). With increasing HbA1c levels, there

Coronary artery disease

was a significant increase in the prevalence of CAD and number of vessels involved. In

Preventive cardiology

multivariate analysis, HbA1c emerged as an independent predictor of significant CAD (OR: 2.8, 95% CI: 1.3–6.2, p = 0.009). Adjusted ORs for the occurrence of CAD were highest in subjects with both hsCRP and HbA1c in the upper 2 quartiles (OR: 4.183; 95% CI: 1.883–9.290, p < 0.0001). There was a significant association between Gensini score and increasing HbA1c tertiles ( p = 0.038). The ideal cut-off value of HbA1c for prediction of the occurrence of CAD was 5.6% 38 mmol/mol) (sensitivity: 60.5%, specificity: 52%). Conclusions: In non-diabetic subjects, HbA1c could be utilized for risk stratification of CAD and its severity, independent of traditional cardiovascular risk factors, insulin resistance and inflammatory markers. # 2013 Elsevier Ireland Ltd. All rights reserved.

1.

Introduction

Although most studies have demonstrated significant reductions in the risk of developing microvascular complications with reduction in Hemoglobin A1c (HbA1c) values [1,2] in

people with diabetes, the association of hyperglycemia with cardiovascular disease (CVD) is less obvious [3,4]. Moreover, a growing literature indicates that the cardiovascular risk associated with hyperglycemia seems to be a continuum without a threshold effect and extends well beyond the threshold currently defined as diabetes [1,5]. Although, fasting

* Corresponding author. Tel.: +98 2188029600; fax: +98 2188029724. E-mail address: [email protected] (G. Davoodi). 0168-8227/$ – see front matter # 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.diabres.2013.10.011

226

diabetes research and clinical practice 102 (2013) 225–232

hyperglycemia below the diabetic threshold has been associated with adverse cardiovascular outcome in some investigations [6,7], data relating fasting hyperglycemia to cardiovascular risk are inconclusive in general [8,9]. However, other studies have denoted that postprandial hyperglycemia may be superior in determining cardiovascular risk [10], but diagnostic testing to detect such an abnormality is cumbersome to carry out. HbA1c is indicative of ambient plasma glucose concentrations, reflecting both fasting and postprandial spikes in the blood glucose levels over the preceding 2–3 months. Determination of HbA1c provides a convenient, singlesample determination of glycemia that does not require a fasting state or glucose load and has higher reproducibility than fasting glucose; all of these factors make it a more feasible measure for large population studies. With the introduction of reference method standardization, issues concerning high inter-laboratory and inter-assay analytic variability have been greatly resolved [11]. Given these characteristics, recently studies have attempted to extend the role of HbA1c as an indicator of cardiovascular risk assessment in people without diabetes. Some have shown HbA1c to be a predictor of future CVD events in the general population [12], while others have found no association [13] or an association only in women [14]. In addition, few studies have examined the association between HbAlc and coronary artery lesions in non-diabetes. Most previous studies have diagnosed diabetes by fasting glucose levels or medical records (conventional diabetes), and oral glucose tolerance test or HbA1c were not used to exclude diabetes. Hyperglycemia below the threshold required for the diagnosis of diabetes, is frequently accompanied by obesity, insulin resistance, hyperlipidemia, and increased levels of pro-inflammatory markers which are also risk factors for poor cardiovascular outcomes [15]. The aim of the present study was to explore whether nondiabetic hyperglycaemia, assessed by HbA1c, is associated with angiographic CAD prevalence or severity, independent of traditional cardiovascular risk factors, insulin resistance and inflammatory markers. In addition, we sought to determine the ideal cut-off value of HbA1c for prediction of the occurrence of CAD, which would contribute to better risk stratification in the nondiabetic population and potentially improved strategies for prevention of CVD.

2.

Methods

2.1.

Study population

We enrolled 382 consecutive individuals without history of known diabetes who underwent their first coronary angiography for suspected ischemia at Tehran Heart center hospital between November 2011 and December 2012. These patients had either angina or angina-like chest pain and evidence of ischemia (either ischemic electrocardiographic changes, positive stress or other noninvasive tests). None of these patients had history of revascularization procedures (percutaneous transluminal coronary angioplasty or coronary artery bypass grafting). Exclusion criteria included newly detected

diabetes (defined as fasting blood sugar  126 mg/dl [7.0 mmol/l], HbA1c  6.5% [53 mmol/l] or 2-h postload glucose  200 mg/dl [11.1 mmol/l] during an oral glucose tolerance test, OGTT), hemoglobin < 11 mg/dl, concomitant systemic diseases such as autoimmune disease, cancer, or active infection, splenectomy or acute blood loss in the last month. As a result, a total of 299 individuals were enrolled and included in the final analyses; 182 (60.9%) men; mean age of 58.8 (median 58.0, range 24–88) years. The study was approved by the Institutional Review Board of the University and the ethics committee of the hospital, complied with the Declaration of Helsinki, and written informed consent was obtained from all of the patients.

2.2.

Assessment of risk factors and comorbidities

Traditional cardiovascular risk factors and demographic data were assessed using a standardized questionnaire. Anthropometric indices (weight and height) were measured and Body mass index (BMI) calculated by dividing weight (kg) by the square of height (m2). Normal fasting glucose (NFG), impaired fasting glucose (IFG) and diabetes were defined as fasting glucose levels < 100 mg/dl (5.5 mmol/l), 100–125 mg/dl (5.5– 6.9 mmol/l) and 126 mg/dl (7.0 mmol/l), respectively [16]. In patients with fasting plasma glucose 100 mg/dl (5.5 mmol/l), plasma glucose concentrations at 120 min following ingestion of 75 g glucose were also analyzed. Individuals with a documented history of diabetes, receiving antidiabetic treatment, HbA1c  6.5 (53 mmol/mol) and/or impaired OGTT were also considered to have diabetes, irrespective of their fasting glucose levels [16]. Blood pressure was measured after 5 min of rest, with a mercury sphygmomanometer in the right arm with the seated, and the average of three recordings was used for analysis. Hypertension was defined as a self-reported history of hypertension and/or use of antihypertensive medication, or a blood pressure of at least 140/90 mmHg [17]. Hyperlipidemia was identified in patients already on lipid lowering treatment, or having fasting triglyceride levels  150 mg/dl (1.7 mmol/l) or total cholesterol 200 mg/dl (5.2 mmol/l) [18].

2.3.

Laboratory measurements

Following overnight fasting, blood samples were drawn from the antecubital vein into EDTA-treated and plain tubes on the day of coronary angiography for biochemical assay and stored at 70 8C prior to analysis. Serum glucose, lipid, creatinine, HbA1c and high sensitive C-reactive protein (hsCRP) concentrations were assessed immediately after admission. Serum triglycerides, total cholesterol, high density lipoprotein-cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and glucose were determined by standard enzymatic procedures. HbAlc was assessed using turbidimetric inhibition immunoassay (Roche Tina-quant Gen.2 HbA1c on Integra 800) with inter- and intra-assay coefficient of variation (CV) of 1.3% and 0.8% at a mean level of 5.3% (34 mmol/mol) and 1.0% and 0.9% at a mean level of 10.2% (88 mmol/mol), respectively, and lower detection limit of 2.9%. HbAlc was calculated as the National Glycohemoglobin Standardization Program (NGSP) equivalent value.

diabetes research and clinical practice 102 (2013) 225–232

The creatinine was measured by Jaffe kinetic reaction buffer without deproteinization. The glomerular filtration rate was calculated using the level-modified Modification of Diet in Renal Disease formula: estimated glomerular filtration rate (eGFR) = 0.741  175  age 0.203  serum creatinine 1.I54 [19]. The female sex adjustment (eGFR female = eGFR  0.742) was applied when appropriate. Apolipoprotein A1, B100 concentrations (Apo A1 and Apo B) were determined by immunoturbidimetry (Cobas Integra 700, Roche) with interand intra-assay CV of 2.4% and 1.0% at a mean level of 88 mg/dl and 1.7% and 0.8% at a mean level of 164 mg/dl, for Apo A1, and inter- and intra-assay CV of 2.9% and 1.2% at a mean level of 80 mg/dl and 3.2% and 1.1% at a mean level of 150 mg/dl, for Apo B, respectively. The sensitivity was 20.0 mg/dl for both Apo A1 and Apo B. Serum CRP concentration was measured by immunoturbidimetry with latex particles sensitized with specific antibodies (Cobas Integra 700, Roche) with interand intra-assay CV of 2.9% and 1.8% at a mean level of 0.62 mg/ dl and 2.7% and 1.5% at a mean level of 14.2 mg/dl, respectively, and lower detection limit of 0.1 mg/dl. Insulin was measured by ELISA (Demeditic diagnostic GmbH, lisemeitner-strabe2, D-24145-kiel, Germany) with inter- and intra-assay CV of 2.9% and 2.6% at a mean level of 17.5 mIU/l and 6.0% and 1.8% at a mean level of 66.5 mIU/l, respectively, and sensitivity of 1.76 mIU/l. Insulin resistance was estimated by homeostasis model assessment basal insulin resistance (HOMA-IR = (fasting glucose (mg/dl) fasting insulin (U/l)/405) [20].

2.4. Assessment of coronary atherosclerosis by coronary angiography Coronary angiography was performed using standard techniques and all coronary angiograms were reported and reviewed by two physicians who were not aware of the status of the patients’ HbA1c level. CAD was defined as 50% luminal narrowing of at least one major epicardial vessel. We attempted to quantify the ‘‘severity of CAD’’ by ascertaining the prevalence of multivessel disease, extent of CAD [one, two, or three vessel disease or left main stem stenosis (50%)] and Gensini scoring system [21]. According to the number of diseased arteries, patients were categorized as having no disease, or one-, two- or three-vessel disease. Given the small number of patients with left main disease, three-vessel and left main disease were grouped together into one, as a threevessel disease.

2.5.

Statistical analysis

Data were analyzed using SPSS software (version16.0). Binary and ordinal data were summarized using percentages and compared using x2 test and Fisher’s exact test when required. Continuous (Scale) data are presented as mean  standard deviation (SD) and compared using the Mann–Whitney U test or independent-samples T test, as appropriate. Pearson’s and spearman correlation analysis were utilized to determine unadjusted bivariate correlation between two variables. Binary and multivariate logistic regression analysis was performed to identify the relative risk of the serum HbA1c levels for the presence and severity of CAD (extent of CAD), as

227

odds ratios (OR) with 95% confidence intervals (CI). A multivariate linear regression analysis was also used to test the independent contribution of HbA1c levels in predicting severity of CAD according to Gensini scoring system. In the multivariate analyses, all covariates with p < 0.2 in univariate analysis were entered. Adjustments were made for age and sex in Model 1. Model 2 was adjusted for the variables in model 1 plus cigarette smoking, hypertension, APOA1, HDL-C, nonHDL-C and BMI, and finally, Model 3 was adjusted for the variables in model 2 plus HOMA-IR, hsCRP and hemoglobin. The predictive values of HbA1c and hsCRP for the presence of CAD were calculated by constructing receiveroperating characteristic (ROC) curves and the area under the curve (AUC) was calculated. To determine the ideal thresholds, the point on the ROC curve with maximum Youden index [sensitivity (1 specificity)]), and the point with shortest distance value form the point (0,1) [(1 sensitivity)2 + (1 specificity)2] were calculated [22]. These are the two most commonly used methods for establishing the ideal cut-off [23]. All tests were 2-tailed, and probability values were considered significant at the 0.05 level.

3.

Result

3.1.

Clinical characteristics

A total of 299 individuals were included in the final analyses; 182 (60.9%) men, mean age of 58.8  10.4, range 24–88 years. The average fasting blood sugar was 99.3  9.0 mg/dl (5.5 mmol/l) and HbA1c was 5.6%  0.4% (38 mmol/mol); 150 (51.5%) individuals had NFG, and 141 (48.5%) had IFG. One hundred and forty one patients (47.2%) had typical chest pain and 158 (52.8%) did not but were candidates because of atypical chest pain or shortness of breath and evidence of ischemia, which persuaded us to perform coronary angiography. The levels HbA1c and hsCRP were comparable in two groups (HbA1c, 5.60 vs. 5.58% p = 0.671 and hsCRP, 0.45 vs. 0.35 p = 0.245, respectively). Coronary angiography revealed that 148 (50%) had a significant stenosis (22% single-vessel, 16% two-vessel and 11% three-vessel disease), and 72 (24%) individuals had nonobstructive disease, while no plaque was observed in 79 (26%.) participants. Tables 1 and 2 outline the demographic, clinical and biochemical characteristics of the patients according to CAD category. Patients with CAD were older, more were cigarette smokers, opium addicts and of male gender compared with the non CAD patients (Table 1). They were also more likely to have lower HDL-C and ApoA1 levels compared with non CAD patients (Table 2).

3.2. Relationship between HbA1c and other demographic and biochemical parameters Univariate analysis showed that the glycated hemoglobin levels were closely related to the serum admission glucose (r = 0.169, p = 0.004), insulin (r = 0.118, p = 0.044), total cholesterol (r = 0.137, p = 0.021), and non-HDL-C levels (r = 0.130, p = 0.03), but not with the hsCRP ( p = 0.097).

228

diabetes research and clinical practice 102 (2013) 225–232

Table 1 – Clinical characteristics of the study population according to coronary artery disease (CAD) category. All N = 299

CAD ( ) N = 152

CAD (+) N = 147

p value

Male (%) Age (years) BMI (kg/m2) Abdominal length Smoking (%) Never Former Current

182 (60.9%) 58.8  10.4 28.4  4.3 99.7  10.03

70 (46.1%) 56.6  10.6 28.9  4.4 100.9  9.1

112 (76.2%) 61.0  9.7 27.8  4.0 98.4  10.8

<0.0001 <0.0001 0.027 0.027

185 (61.9%) 39 (13.0%) 75 (25.1%)

115 (75.7%) 16 (10.5%) 21 (13.8%)

70 (47.6%) 23 (15.6%) 54 (36.7%)

<0.0001

Cigarette PACK.YEAR Opium (%) Never Former Current

20 (8.8–35)

10.3 (8.0–23.8)

25.0 (10.5–40.0)

255 (85.3%) 8 (2.7%) 36 (12.0%)

142 (93.4%) 2 (1.3%) 8 (5.3%)

113 (76.9%) 6 (4.1%) 28 (19.0%)

<0.0001

Hypertension (%) Hyperlipidemia (%) Family history of CAD (%) Left ventricular Ejection fraction (%)

160 (53.5%) 208 (69.6%) 45 (15.1%) 52.2  9.0

77 (50.7%) 105 (69.1%) 21 (13.8%) 53.9  8.2

83 (56.5%) 103 (70.1%) 24 (16.3%) 50.6  9.5

0.314 0.853 0.544 0.005

0.014

Data are mean  standard deviation or frequency (prevalence rates).

3.3.

HbA1c and hsCRP levels and CAD presence

HbA1c (5.6  0.4% vs. 5.5  0.4% p = 0.017) and hsCRP (0.53  0.92 vs. 0.27  0.31 mg/dl, p = 0.003) levels were significantly higher in patients with CAD than in the group without CAD, but the serum fasting glucose and insulin levels did not significantly differ (Table 2). Logistic-regression analysis revealed that HbA1c levels were still predictive of CAD prevalence even after considering the effects of age, sex, smoking, hypertension, ApoA1, HDL-C, nonHDL-C and BMI, HOMA-IR, hsCRP and hemoglobin levels. For every unit increase in HbA1c, there was a 2.8-fold increase in CAD prevalence (adjusted OR: 2.845; 95% CI: 1.296–6.244, p = 0.009) (Table 3). Male gender (OR: 2.183; 95% CI: 1.009–4.726, p = 0.047) and older age (OR: 1.047; 95% CI: 1.015–1.079, p = 0.004) were also found to be independent risk factors for the presence of CAD. To analyze the combined effects of hsCRP and HbA1c, the subject population was then further subdivided into 3 groups: group I (n = 68, 22.7%) consisted of subjects with both hsCRP and HbA1c in the lower 2 quartiles; group II (n = 134, 44.8%) consisted of subjects with 1 of the parameters in the lower 2 quartiles and the other parameter in the upper 2 quartiles; and group III (n = 97, 32.4%) consisted of subjects with both parameters in the upper 2 quartiles. Adjusted ORs for the occurrence of CAD were highest in subjects with both parameters in the upper 2 quartiles (OR: 4.183; 95% CI: 1.883–9.290, p < 0.0001, as model 3).

3.4.

Subgroup analysis

When patients were divided into two groups of NFG and IFG, the association between higher HbA1c levels and CAD was observed only in NFG group (OR: 5.276; 95% CI: 1.440–19.336, p = 0.012), but HbA1c ceased to be associated with the CAD in IFG group (OR: 1.727; 95% CI: 0.544–5.484, p = 0.354).

3.5.

Glycaemic status and CAD severity

Pearson’s test (controlled for age and sex) revealed that increasing tertiles of HbA1c correlated positively with both the

number of significantly diseased coronary vessels (r = 0.155, p = 0.008) and multivessel disease (r = 0.137, p = 0.019). Upon further review of angiograms and after assigning Gensini scores, in bivariate correlation analysis HbA1c level as a continuous variable was not significantly related to the severity of CAD (Spearman r = 0.105, p = 0.059). But linearregression analysis after adjustment for age, sex and other cardiovascular risk factors revealed a trend of increasing Gensini score with increasing HbA1c. When HbA1c was employed in regression models as categorical scale using HbA1c tertile (the lowest tertile was the reference category) a significant association between Gensini score and increasing levels of HbA1c was observed. Table 4 illustrates the multivariate linear regression associations between HbA1c tertiles. The regression coefficient between HbA1c and the Gensini score was 2.031 ( p < 0.0001).

3.6.

Ideal cut-off level of HbA1c for the diagnosis of CAD

ROC curve for HbA1c and hsCRP in predicting CAD demonstrated that the HbA1c level was a significant predictor for the presence of CAD (AUC: 0.620, 95% CI 0.548–0.691, p = 0.001), whereas hsCRP was not (AUC: 0.543, 95% CI 0.473–0.613, p = 0.243). The ideal threshold of HbA1c for CAD diagnosis was chosen according to the point on the ROC curve with maximum Youden index and the minimized distance on the ROC curve, as 5.6% (38 mmol/mol) (sensitivity: 60.5% specificity: 52%) (Fig. 1).

4.

Discussion

The principal findings of our study indicate that HbAlc values, even in the normal range, are associated with the presence and severity of coronary lesions in people without diabetes, even after multivariable analysis models which adjusted for traditional cardiovascular risk factors, insulin resistance and inflammatory markers. An increase in HbA1c of one percent was associated with a 2.8-fold increase in CAD prevalence in

229

diabetes research and clinical practice 102 (2013) 225–232

Table 2 – Biochemical characteristics of the study population according to coronary artery disease (CAD) category.

HbA1c Fasting blood sugar (mg/dl) Total cholesterol (mg/dl) Triglyceride (mg/dl) LDL-cholesterol (mg/dl) HDL-cholesterol (mg/dl) NonHDL-cholesterol (mg/dl) Cholesterol/HDL APOA1 (mg/dl) APOB100 (mg/dl) APOB100/APOA1 Fasting insulin (IU/l) HOMA-IR units) hsCRP (mg/l) Creatinin (mg/dl) eGFR (cc) Hemoglobin (mg/dl)

All N = 299

CAD ( ) N = 152

CAD (+) N = 147

p value

5.59  0.36 99.3  9.0 181.1  49.7 124.0 (95.0–169.0) 115.4  40.3 44.0  10.8 136.8  46.5 4.3  1.2 134.4  22.8 95.1  28.1 0.7  0.2 11.0 (7.8–16.2) 2.8 (1.9–4.1) 0.39  0.69 0.9  0.2 63.5  14.9 52.2  9.0

5.54  0.36 98.9  8.5 182.8  43.2 122.0 (97.0–159.0) 115.9  35.8 46.0  10.1 135.6  40.3 4.1  1.0 138.5  21.7 96.0  27.5 0.7  0.2 11.0 (7.9–17.1) 2.8 (1.9–4.2) 0.27  0.31 0.8  0.2 64.7  15.1 14.4  1.6

5.64  0.36 99.8  9.5 179.4  55.3 127.5 (91.3–188.3) 114.9  44.2 42.1  11.2 138.0  51.8 4.4  1.4 130.2  23.2 94.2  28.7 0.7  0.2 11.0 (7.7–16.0) 2.8 (1.9–4.1) 0.53  0.921 0.9  0.2 62.3  14.5 14.8  1.6

0.017 0.442 0.566 0.241 0.837 0.002 0.669 0.009 0.002 0.587 0.189 0.712 0.999 0.003 0.003 0.168 0.032

Data are mean  standard deviation or median (inter-quartile range). HbA1c, hemoglobin A1c; HDL, high density lipoprotein; LDL, low density lipoprotein; APO A1, apolipoprotein A1; APO B100, apoipoprotein B100, HOMA-IR, homeostasis model assessment basal insulin resistance; hsCRP, high sensitive C-reactive protein; eGFR, estimated glomerular filtration rate.

the multivariate analysis and the best threshold of HbA1c for prediction of the presence of CAD was 5.6% (38 mmol/mol). The current study also showed that glycemia in nondiabetic individuals, as a contributing factor for angiographic CAD, may be more important in individuals with normal fasting glucose. In addition, hyperglycemia might enhance the proatherogenic effects of hsCRP. Previous studies reported the relationship between HbAlc values and cardiovascular events in nondiabetic individuals and its prognostic value was also observed in patients with acute coronary syndrome. In the Rancho Bernardo cohort [24] of 1239 older non-diabetic adults, baseline HbA1c but not fasting or post-challenge glucose predicted cardiovascular mortality in women but not in men. A threshold effect was noted, such that women in the highest (6.7% [50 mmol/mol]) vs. the lower four quintiles had a near 3-times elevation in adjusted risk. In a report from the European Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) [25], a 1% increment in HbA1c was associated with a 21% increase in cardiovascular risk after multivariable adjustment in both men and women. However, when subjects with prior diabetes and CVD were excluded this association was attenuated and no longer statistically significant. In the Hoorn Study [26], which also presented categorical analyses, the age-adjusted risk in the highest vs. lowest category (6.5% [48 mmol/mol]

vs. <5.2% [33 mmol/mol]), was 3.8 (95% CI: 1.6–8.0). However, after additional adjustment for gender, hypertension, dyslipidemia, and smoking, this effect was diminished and not statistically significant. The womens health study (WHS) cohort [13], suggested that HbA1c levels were elevated well in advance of the clinical development of type 2 diabetes supporting recent recommendations for lowering of diagnostic thresholds for glucose metabolic disorders. In contrast, the association of HbA1c with incident cardiovascular events appeared weak and did not persist after accounting for established cardiovascular risk factors. Pai et al. [27], conducted parallel nested case-control studies in 2 cohorts of US health professionals, in non diabetic women (Nurses’ Health Study) and men (Health Professionals Follow-up Study) and found that compared with HbA1c of 5.0% to <5.5%, those with an HbA1c of 6.0% to <6.5%, the pooled relative risk of CAD was 1.29 (95% CI 1.11–1.50) for every 0.5%-increment increase in HbA1c levels and 1.67 (95% CI 1.23–2.25) for every 1%increment increase, with the risk plateauing around 5.0%. Furthermore, participants with HbA1c levels between 6.0% and <6.5% and CRP levels >3.0 mg/l had a 2.5-fold higher risk of CAD compared with participants in the lowest categories of both biomarkers. A recent meta-analysis also examined the association between HbA1c and risk of CAD in people without diabetes,

Table 3 – Multivariate regression models for presence of coronary artery disease according to HbA1C.

p value OR (95% CI) per 1–percentage point increase of HbA1c OR (95% CI) per 0.5–percentage point increase of HbA1c

Model 1

Model 2

Model 3

0.003 2.907 (1.430–5.909) 1.70 (1.20–2.43)

0.005 3.000 (1.398–6.439) 1.73 (1.18–2.54)

0.009 2.845 (1.296–6.244) 1.69 (1.14–2.50)

Model 1, adjusted for age and sex; Model 2, adjusted for age, sex, smoking, hypertension, apolipoprotein A1, high density lipoproteincholesterol, non-HDL-cholesterol and body mass index; Model 3, adjusted for the variables in model 2 plus, homeostasis model assessment basal insulin resistance, high sensitive C-reactive protein and hemoglobin. OR (95% CI): Odds Ratio (95% confidence interval).

230

diabetes research and clinical practice 102 (2013) 225–232

Table 4 – Multivariate linear regression models for Gensini score according to HbA1C. Model 1

Model 2

Model 3

HbA1c (scale) b R2 p value

0.099 0.342 0.076

0.108 0.369 0.067

0.116 0.367 0.057

HbA1c tertiles b R2 p value

0.109 0.345 0.049

0.124 0.374 0.035

0.125 0.370 0.038

Models as Table 3.

although most studies included participants with HbA1c well beyond the established clinical threshold of 6.5%. Using data from 9 cohorts (1639 cases), the meta-analysis reported a relative risk of 1.2 (95% CI 1.10–1.31) for each 1%-increment increase in levels of HbA1c (not adjusted for established CAD risk factors) [28]. With regard to the association between HbAlc and angiographic coronary artery lesion extension in people without diabetes, limited prior studies of smaller size have demonstrated variable results. Kowalska et al. [29] in a cohort of nondiabetic men referred for coronary angiography, demonstrated that the number of diseased vessels significantly correlated with increasing levels of HbAlc. In a small angiographic study by Pajunen et al. [30] that recruited asymptomatic members of high-risk families, fasting postload glycaemia and HbA1c levels, although significantly correlated with the number of diseased vessels in the univariate analysis, failed to predict angiographic severity when incorporated into a multivariate model. Konstantinou et al. [31] also demonstrates that deterioration of the glycaemic profile even in the non-diabetic range is an independent determinant of angiographic CAD prevalence

Fig. 1 – Optimal cut-off of hemoglobin A1c for the diagnosis of coronary artery disease.

and severity. Recently, Ikeda et al. [32] showed that HbAlc is significantly associated with the complexity of coronary lesions. This association was even observed in non-diabetic adults. In all of these studies, the diagnosis of diabetes was based solely on fasting glucose levels, and neither oral glucose tolerance test nor HbA1c (except the study of Pai et al. [27]) were considered for exclusion of diabetes. Thus, it is possible that some of the individuals categorized in these studies as not having diabetes indeed have the disease based on these test results. In addition, they cannot rule out residual confounding with inflammatory markers or insulin resistance, and in some of these reports, some known conventional cardiovascular risk markers were not included in the adjustment which might affect the odds ratios (e.g., smoking data in Ikeda et al. [32] study and family predisposition in WHS cohort [13]). The pathophysiological mechanisms responsible for glycemia-induced endothelial dysfunction have not been fully elucidated. According to current understanding, hyperglycemia induces oxidative stress which, in combination with soluble advanced glycation end products (AGEs) and lipid peroxidation products leads to endothelial dysfunction and expression of inflammatory genes [33], thereby promoting atherogenesis. HbA1c, rather than reflecting ambient glucose levels, might indicate more widespread protein glycation [34] and associated inflammation which may precede the development of CAD. Although the current study was not designed to examine mediating mechanisms, the observation that the association of HbA1c with CAD was independent of basal insulin resistance, hsCRP (as a marker for low grade inflammation) and other metabolic disorders, implies a direct effect of elevated glucose on the endothelium. The observation that HbA1c ceased to be associated with the CAD in IFG group might be because of heavily clustered metabolic abnormalities in this group. The current study also showed that hyperglycemia might strengthen the proatherogenic effects of hsCRP. Several findings suggest that hsCRP may be related to atherogenesis by diminishing nitric oxide [35]. Interestingly, acute hyperglycemia also reduces nitric oxide bioavailability [36], pointing to the possibility that the combined effect of raised hsCRP concentrations and increased HbA1c may jointly reduce nitric oxide bioavailability. This study has some limitations. First, because the patients enrolled in our study were candidates for coronary angiography, our findings cannot be extrapolated to the general population. Diagnostic thresholds derived from high-risk populations may not be generalizable to lower risk groups as screening characteristics vary with underlying glucose frequency distributions. Second, given the cross-sectional design of our study, the directionality of the association cannot be determined. Third, we used a single baseline measurement of HbA1c. We therefore cannot evaluate the effects of changes in this parameter over time. Fourth, physical activity or sedentary lifestyle may also have an independent effect on CAD; however, because we have incomplete data on these variables, they were not included as potential confounders. Finally, on account of the observational design we cannot report or make conclusions with regard to event outcomes. Prospective studies are needed to evaluate whether individuals without diabetes with increasing HbA1c levels have worse cardiovascular outcomes.

diabetes research and clinical practice 102 (2013) 225–232

Our study suggest that an elevated HbA1c even at levels generally considered within the normal range is a significant determinant of angiographically verified CAD prevalence and severity, independent of traditional CAD risk factors. These findings demonstrate the potential role of this biomarker at levels, even in the non-diabetic range, in usual clinical care in risk stratification and early identification of CAD patients who may warrant a greater emphasis in primary prevention. Future prospective studies are needed to validate this potential mechanistic explanation, and to determine if aggressive interventions in early-stage glycometabolic disorders would improve the risks of cardiovascular disease.

[10]

[11]

[12]

[13]

Conflict of interest [14]

Authors declare no conflicts of interest. [15]

Acknowledgements This work was supported by a grant from Tehran University of Medical Sciences (900430-15947). We gratefully acknowledge the help of Mr. Mehrdad Bagheri, Mr. Mahmoud Ruhzendeh, and Mr. Hasan Yaghoubzadeh, laboratory technicians.

[16]

[17]

references

[1] Diabetes Control and Complications Trial Research Group: The absence of a glycemic threshold for the development of long-term complications: the perspective of the Diabetes Control and Complications Trial. Diabetes 1996;45:12891298. [2] 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;352:837-853. [3] Hoff JA, Quinn L, Sevrukov A, Lipton RB, Davigllus M, Garside DBea. The prevalence of coronary artery calcium among diabetic individuals without known coronary artery disease. J Am Coll Cardiol 2003;41:1008–12. [4] Rosamond W, Flegal K, Furie Kea. Heart disease and stroke statistics–2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2008;117:e25–146. [5] Standards of medical care in diabetes–2013. American Diabetes Association. Diabetes Care 2013;36(Suppl. 1): S11–66. [6] Simons L, Friedlander Y, McCallum J. Fasting plasma glucose in non-diabetic elderly women predicts increased all-cause mortality and coronary heart disease risk. Aust N Z J Med 2000;30:41–7. [7] Fisman EZ, Motro M, Tenenbaum A, Boyko V, Mandelzweig L, Behar S. Impaired fasting glucose concentrations in nondiabetic patients with ischemic heart diseasea marker for a worse prognosis. Am Heart J 2001;141:485–90. [8] Hu G, Qiao Q, Tuomilehto J. Glucose tolerance and cardiovascular mortality. Cardiovasc Rev Rep 2001;22:649–54. [9] Yarnell JW, Pickering JE, Elwood PC, Baker IA, Bainton D, Dawkins C, et al. Does non-diabetic hyperglycemia predict

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

231

future IHD? Evidence from the Caerphilly and Speedwell studies. J Clin Epidemiol 1994;383–8. Shah J, Hodge A. Isolated post-challenge hyperglycemia confirmed as a risk factor for mortality. Diabetologia 1999;42:1050–4. Nathan DM, Singer DE, Hurxthal K, Goodson JD. The clinical information value of the glycosylated hemoglobin assay. N Engl J Med 1984;310:341–6. Khaw K, Wareham N, Luben R, Bingham S, Oakes S, Welch A, et al. Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of European Prospective Investigation of Cancer and Nutrition (EPICNorfolk). BMJ 2001;322. Blake GJ, Pradhan AD, Manson JE, Williams GR, Buring J, Ridker PM, et al. Hemoglobin A1c level and future cardiovascular events among women. Arch Intern Med 2004;164:757–61. Singer DE, Nathan DM, Anderson KM, Wilson PW, Evans JC. Association of HbA1c with prevalent cardiovascular disease in the original cohort of the Framingham Heart Study. Diabetes 1992;41:202–8. Stentz FB, Umpierrez GE, Cuervo R, Kitabchi AE. Proinflammatory cytokines, markers of cardiovascular risks, oxidative stress, and lipid peroxidation in patients with hyperglycemic crises. Diabetes 2004;53:2079–86. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2012;13:S64–71. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo Jr JL, et al. Seventh report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Hypertension 2003;42:1206–52. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285:2486–97. Imai E, Horio M, Nina K, Yamagata K, Iseki K, Hara S, et al. Estimation of glomerular filtration rate by the MDRD study equation modified for Japanese patients with chronic kidney disease. Clin Exp Nephrol 2007;11:41–50. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–9. Gensini GG. A more meaningful scoring system for determining the severity of coronary heart disease. Am J Cardiol 1983;51:606–7. Pepe M. The statistical evaluation of medical tests for classification and prediction. New York: Oxford University Press; 2003. Perkins NJ, Schisterman EF. The inconsistency of optimal cutpoints obtained using two criteria based on the receiver operating characteristic curve. Am J Epidemiol 2006;163:670–5. Park S, BarrettConnor E, Wingard DL, Shan J, Edelstein S. Ghb is a better predictor of cardiovascular disease than fasting or postchallenge plasma glucose in women without diabetes. The Rancho Bernardo study. Diabetes Care 1996;19:4506. Khaw K-T, Wareham N, Bingham S, Luben R, Welch A, Day N. Association of Hemoglobin A1c with Cardiovascular Disease and Mortality in Adults: The European Prospective Investigation into Cancer in Norfolk. Ann Intern Med 2004;141:413–20.

232

diabetes research and clinical practice 102 (2013) 225–232

[26] Henry RM, Ferreira I, Kostense PJ, Dekker JM, Nijpels G, Heine RJ, et al. Type 2 diabetes is associated with impaired endothelium-dependent, flow-mediated dilation, but impaired glucose metabolism is not; The Hoorn Study. Atherosclerosis 2004;174:49–56. [27] Pai JK, Cahill LE, Hu FB, Rexrode KM, Manson JE, Rimm EB. Hemoglobin a1c is associated with increased risk of incident coronary heart disease among apparently healthy, nondiabetic men and women. J Am Heart Assoc 2013;2:e000077. [28] Sarwar N, Aspelund T, Eiriksdottir G, Gobin R, Seshasai SR, Forouhi NG, et al. Markers of dysglycaemia and risk of coronary heart disease in people without diabetes: Reykjavik prospective study and systematic review. PLoS Med 2010;e1000278. [29] Kowalska I, Prokop J, Bacho´rzewska-Gajewska H, Telejko B, Kinalskal I, Kochman W, et al. Disturbances of glucose metabolism in men referred for coronary arteriography. Postload glycemia as predictor for coronary atherosclerosis. Diabetes Care 2001;24:897–901. [30] Pajunen P, Syva¨nne M, Nieminen MS, Kareinen A, Viitanen L, Lehto S, et al. Serum homocysteine, creatinine, and glucose as predictors of the severity and extent of coronary

[31]

[32]

[33] [34] [35]

[36]

artery disease in asymptomatic members of high-risk families. Eur J Clin Invest 2002;32:472–8. Konstantinou DM, Chatzizisisa YS, Louridasa GE, Parcharidisa GE, Giannoglou GD. Non-diabetic hyperglycaemia correlates with angiographic coronary artery disease prevalence and severity. Diabetes Metab 2010;36:402–8. Ikeda N, Iijima R, Hara H, Moroi M, Nakamura M, Sugi K. Glycated hemoglobin is associated with the complexity of oronary artery disease, even in non-diabetic adults. J Atheroscler Thromb 2012;19:1066–72. Gillery P. Oxidative stress and protein glycation in diabetes mellitus. Ann Biol Clin 2006;64:309–14. Brownlee M. Lilly Lecture 1993. Glycation and diabetic complications. Diabetes Care 1994;43:836–41. Verma S, Wang CH, Li SH, Dumont AS, Fedak PW, Badiwala MV, et al. A self-fulfilling prophecy: C-reactive protein attenuates nitric oxide production and inhibits angiogenesis. Circulation 2002;106:913–9. Giugliano D, Marfella R, Coppola L, Verrazzo G, Acampora R, Giunta R, et al. Vascular effects of acute hyperglycemia in humans are reversed by L-arginine: evidence for reduced availability of nitric oxide during hyperglycemia. Circulation 1997;95:1783–90.