Significant Coronary Stenosis in Asymptomatic Chinese With Different Glycemic Status

The current study was nested in an ongoing community-based cohort study that investigated associations between glucose dysregulation and cardiovascular complications. The study design and protocol of baseline data collection of the cohort study were described previously (15). Briefly, all the permanent residents aged 40 years or older in Songnan community in Shanghai were invited to participate in a screening examination for cardiometabolic diseases. Among 10,185 participants, we randomly selected 5,250 subjects ‪using a ratio of 1.0 [diabetes diagnosed previously or fasting plasma glucose (FPG) ≥7.0 mmol/L] to 1.2 (no previous diabetes and 5.6 ≤ FPG < 7.0 mmol/L) to 1.44 (no previous diabetes and FPG <5.6 mmol/L) and oversampling people with lower glucose levels because they might have a lower participation rate than those with higher glucose levels‬, to undergo a much detailed and comprehensive evaluation including a standard 75-g oral glucose tolerance test. We then reclassified the participating 4,012 subjects (attendance rate, 76.4%) into NGR, prediabetes, and diabetes groups based on their diabetes history and FPG and 2-h postload plasma glucose (PPG) levels according to the 1999 World Health Organization criteria. There was no significant difference in age and sex distribution between those included and those not included in the cohort.

For the current study, we randomly selected 150 individuals from the diabetes group and 150 NGR group and 150 prediabetes group individuals matched for age and sex of diabetic participants, respectively, after excluding subjects with the following characteristics: 1) age older than 60 years; 2) having symptoms of CAD (chest pain or shortness of breath); 3) having a history of cardiovascular diseases (myocardial infarction, unstable angina, percutaneous coronary intervention, or stroke); 4) having abnormal Q waves on resting electrocardiogram (ECG); 5) having a previous diagnosis of diabetes for >5 years; 6) having impaired liver or renal function [alanine aminotransferease more than twice the upper limit of the normal range, serum creatinine level >133 μmol/L (1.5 mg/dL), or glomerular filtration rate <60 mL/min]; 7) being pregnant or having significant medical comorbidities; 8) having X-ray examination or CT scan within 1 year; 9) having tachycardia (a heart rate >90 bpm) or arrhythmia such as atrial fibrillation on ECG that causes coronary artifacts during CTA examinations; and 10) having a history of allergic reaction to iodine-containing contrast agent. Finally, a total of 420 individuals (attendance rate, 93%) participated in the current study.

The study protocol was approved by the Institutional Review Board of Rui-Jin Hospital and written informed consent was obtained from each participant after providing a full explanation of the protocol and procedure.

During July 2009 and August 2010, all participants underwent a comprehensive examination including a detailed questionnaire, anthropometric measurements, biochemical evaluation, and CTA examination. Family history of CAD in first-degree relatives was recorded and history of chronic diseases and current use of medication were acquired. Smoking status was defined as current if a subject smoked cigarettes regularly in the past 6 months. The BMI was calculated as body weight in kilograms divided by body height in meters squared (kg/m²). Blood pressure was measured in the nondominant arm in a seated position three times consecutively at 1-min intervals after at least 5 min of rest using an automated electronic device (OMRON Model HEM-752; Omron Company, Dalian, China). The last two readings were averaged for analysis. Participants without a previous diagnosis of diabetes underwent an oral glucose tolerance test. All participants were told to fast for at least 10 h before blood samples were collected.

Plasma glucose, serum triglycerides, total cholesterol, HDL cholesterol, and LDL cholesterol were measured using an autoanalyser (Beckman CX-7 Biochemical Autoanalyser, Beckman Coulter, Brea, CA). Hemoglobin A_1c (HbA_1c) was determined by high-performance liquid chromatography using the VARIANT II Hemoglobin Testing System (Bio-Rad Laboratories, Berkeley, CA) in a National Glycohemoglobin Standardization Program–certified laboratory of Shanghai Institute of Endocrine and Metabolic Diseases. Fasting serum insulin was measured by an electrochemiluminescence assay (Roche Diagnostics). The single-void first morning urine samples were collected to assess urinary albumin-to-creatinine ratio. Urinary albumin concentrations were determined by immunoturbidimetry and urinary creatinine concentrations were measured by a modified Jaffe method on an automatic analyzer (Beckman LX-20).

Diabetes was defined either by a previous diagnosis and contemporary antidiabetic medication or by levels of plasma glucose during oral glucose tolerance test according to the 1999 World Health Organization criteria. Prediabetes was defined as FPG ≥6.1 mmol/L and <7.0 mmol/L plus PPG <7.8 mmol/L or FPG <7.0 mmol/L plus PPG ≥7.8 mmol/L and <11.1 mmol/L in participants without a previous diagnosis of diabetes. The indexes of homeostasis model assessment (HOMA) of insulin resistance and HOMA β-cell function were calculated according to the formulas: HOMA of insulin resistance = fasting insulin concentration (mIU/L) × FPG (mmol/L) / 22.5; HOMA β-cell function = 20 × fasting insulin concentration (mIU/L)/[FPG (mmol/L) – 3.5]. Urinary albumin-to-creatinine ratio was calculated by dividing the urinary albumin concentrations by the urinary creatinine concentrations and expressed in milligrams per gram. Microalbuminuria was defined as an albumin-to-creatinine ratio between 30 and 300 mg/g.

All examinations were performed on a dual-source CT scanner (SOMATOM Definition; Siemens Medical Solutions, Forchheim, Germany). None of the participants in the current study received β-blockers.

A standard retrospectively ECG-gated scanning protocol was applied, with 0.6-mm slice collimation, 330-ms gantry rotation time, 120-kV tube voltage, and a maximum tube current of 400 mAs/tube. All scans were performed using ECG-controlled tube current modulation. A bolus of 70 mL Iohexol injection (350 mg/mL iodine; Omnipaque; GE Healthcare Shanghai, Shanghai, China) was intravenously injected (4 mL/s) via an 18-gauge catheter placed in the antecubital vein, followed by a 40-mL saline chaser.

The CTA images were interpreted independently by an experienced senior radiologist, who was unaware of the clinical information of study participants, with an offline three-dimensional workstation (ADW 4.4; GE Healthcare, Waukesha, WI). Coronary arteries were divided into 15 segments according to the American Heart Association classification (16). Only segments with a diameter >1.5 mm were included for analysis. Furthermore, patients with uninterpretable segments attributable to motion artifacts were excluded. The presence of atherosclerotic plaques and luminal narrowing were evaluated using axial images and curved multiplanar reconstructions. Coronary plaques were considered when structures >1 mm² were detected within or adjacent to the coronary artery lumen, which could be clearly distinguished from vessel lumen and the surrounding pericardial tissue. Significant coronary stenosis was defined as >50% narrowing of vessel lumen.

Given an estimated response rate of 75%, we calculated that a sample size of 140 participants for each of the 3 groups was needed to provide 90% power at a 2-tailed significance level of 0.05 to detect a difference from 5 to 20% of estimated prevalence of significant coronary stenosis assessed by CTA among the groups. All data were analyzed using SAS 9.2 (SAS Institute, Cary, NC). Continuous variables were presented as means ± SD or medians (interquartile ranges) for skewed variables. Categorical variables were shown in absolute numbers and percentages. Demographic and metabolic features and characteristics of coronary arteries in NGR, prediabetes, and diabetes were described and compared using ANOVA for continuous variables and logistic regression analysis for categorical variables, adjusted for age and sex. The study population was reclassified into groups of increasing coronary stenosis (participants without coronary stenosis, with <50% coronary stenosis, and with significant coronary stenosis). FPG, PPG, and HbA_1c levels and diabetes prevalence were then compared among these three groups.

To investigate the associations of various cardiometabolic factors with significant coronary stenosis, logistic regression models were used to assess the crude and multivariate-adjusted odds ratios (OR) of advanced age (age 53 years or older, representing the median cut-off value), female sex, family history of CAD, ACE inhibitor/angiotensin receptor blocker medication, current smoking, high school education or more, overweight/obesity, hypertension, high LDL, low HDL, prediabetes, and diabetes for risks of significant coronary stenosis. Variables for multivariate regression analysis were chosen as confounding factors depending on their clinical plausibility and external evidence such as previous research and previous beliefs, i.e., their well-recognized clinical relevance to both dysglycemia (the independent variable of interest) and cardiovascular diseases (significant coronary stenosis, the dependent variable) in this study, rather than internal statistical evidence from the data.

To further elucidate the differences between diabetic participants with or without significant coronary stenosis, general characteristics were compared. Crude and multivariate-adjusted ORs for significant coronary stenosis in diabetes by each 1-SD increase in levels of different glucose evaluations were calculated using logistic regression procedures.

Significance tests were two-tailed and P < 0.05 was considered statistically significant.

Nineteen (4.5%) patients with uninterpretable segments attributable to motion artifacts were excluded from analysis, which resulted in 135 with NGR, 132 with prediabetes, and 134 with diabetes participating for data analysis. Plaques were analyzed on a per-patient level.

Characteristics of the study population according to glycemic status are shown in Table 1. Generally, percentages of individuals with a family history of CAD, educated beyond high school level, or currently smoking were not significantly different among groups, whereas BMI, blood pressure, and triglycerides were elevated and HDL was decreased substantially in parallel with deteriorations in glucose levels, insulin sensitivity, and β-cell function. Medications using ACE inhibitors or angiotensin receptor blockers were basically similar among the three groups. Daily use of aspirin or statins was extremely rare in study participants. Only two individuals were using aspirin. One was in the NGR group and the other was in the prediabetes group, and both were without coronary plaques. Statins were used by only one individual who was diabetic and had nonsignificant coronary stenosis.

As shown in Table 1, coronary plaques were found in 58 (43.0%) participants with NGR, in 77 (58.3%) with prediabetes, and in 74 (55.2%) with diabetes (P for trend = 0.081), whereas plaques causing significant coronary stenosis were detected in 10 (7.4%), 10 (7.6%), and 22 (16.4%) individuals with NGR, prediabetes, and diabetes, respectively (P for trend = 0.029). Therefore, a similar prevalence of CAD detected by CTA was found in prediabetes and diabetes, and it was relatively higher than that in NGR; however, prevalence of significant coronary stenosis was doubled in diabetes and was substantially increased compared with that in NGR or prediabetes.

When participants were reclassified according to existence or extent of coronary stenosis, individuals with significant stenosis had dramatically elevated levels of FPG, PPG, and HbA_1c and had a significantly increased prevalence of diabetes compared with both groups of participants with <50% coronary stenosis and without stenosis after controlling for a variety of confounding factors (all P < 0.05; Fig. 1).

Multivariate logistic regression analysis revealed few independent risk factors for significant coronary stenosis. Women were strongly protected from significant coronary stenosis [OR, 0.33 (95% CI, 0.14–0.81); P = 0.016; Table 2]. Prediabetes was not associated with an increased risk of significant coronary stenosis, whereas diabetes was associated with a significant 2.34-fold elevated risk [2.34 (1.01–5.43); P = 0.047] compared with NGR.

Among 134 patients with diabetes for <5 years, 22 (16.4%) had significant coronary stenosis. Diabetic individuals with significant coronary stenosis were more likely to be males and had more deteriorated glucose metabolism and β-cell function compared with those without significant coronary stenosis (Table 3). The presence of microalbuminuria tended to be increased in those with significant stenosis, but it failed to reach statistical significance. Among all cases of diabetes, 104 (77.6%) were newly diagnosed, but 14 (13.5%) already had significant coronary stenosis. Although newly diagnosed diabetes tended to occur more often in the group without significant coronary stenosis, the difference was not statistically significant.

In univariate and multivariate logistic analyses, glucose evaluation levels were independently and significantly associated with risks of significant coronary stenosis in diabetes. Each 1-SD increase in FPG, PPG, or HbA_1c conveyed 2.11-fold, 1.73-fold, or 1.81-fold higher risks of significant coronary stenosis, respectively, after controlling for other conventional cardiovascular risk factors (all P < 0.05; Table 4).

We used CTA, which is a noninvasive diagnostic modality, to characterize subclinical CAD in a community-dwelling adult population with NGR, prediabetes, or diabetes matched for age and sex. We found that although coronary stenosis was more prevalent in both prediabetes and diabetes, the risk of significant stenosis was only elevated in diabetes, despite a short duration of <5 years. Additionally, glucose evaluation levels were significantly and independently associated with risks of significant coronary stenosis in diabetic participants.

There has been a well-documented association between diabetes and cardiovascular diseases. Haffner et al. (17) reported that the presence of diabetes alone confers risk of cardiovascular mortality similar to that in nondiabetic individuals with a previous myocardial infarction. Heart disease is the most frequent cause of death in those with diabetes, accounting for 70% of all deaths (18). Coronary artery lesions in those with diabetes were often characterized as diffuse and multivessel, yet asymptomatic (19). In fact, silent myocardial ischemia is highly prevalent in the diabetic population. In the current study, 74 out of 134 diabetic patients (55.2%) were found to have a certain degree of coronary stenosis and 22 (16.4%) had significant stenosis, which was relatively low compared with most studies using CTA in asymptomatic patients with diabetes (11–14). The variation in prevalence may well reflect the important differences regarding population characteristics between studies. Those with higher prevalence rates of significant stenosis, ranging between ∼26 and ∼41%, generally included a higher proportion of male participants (for example, 66% in the study by Rivera et al. (11) compared with 50% in the current study) and more patients with hypertension (69 vs. 59%), dyslipidemia (78 vs. 32%), or family history of CAD (30 vs. 19%). One study included patients with abnormal stress test results (13). Another important and more relevant fact is that participants in those studies had much longer duration of diabetes (mean duration, 7–10 years). Diabetes duration is thought to contribute significantly to CAD risks, and the Framingham Heart Study reported a 1.38-fold increased risk for CAD and a 1.86-fold higher risk for CAD death for each 10-year increase in diabetes duration (20). In previous studies, much attention has been given to diabetes of an advanced stage and CAD evaluation. There is a paucity of data on coronary artery stenosis in asymptomatic diabetes that has been diagnosed for a relatively short period of time. In the current study, a majority of individuals in the diabetes group were newly diagnosed (104/134; 77.6%) and the rest of diabetes cases were detected within 5 years. These diabetic participants were found to have a 2.34-fold increased risk for significant coronary stenosis compared with NGR. It is well-known that diabetes exists for several years before diagnosis, and thus an elevated risk of significant coronary stenosis in recently diagnosed diabetes may reflect a cumulative effect of glucose dysregulation that has been going on for years and for much longer than it had been noticed.

Moreover, among diabetic patients in this study, those with significant coronary stenosis had much higher glucose and HbA_1c levels than those without. Glucose evaluations, i.e., FPG, PPG, and HbA_1c, were found to be significant and independent risk markers for significant coronary stenosis in diabetes. Although deterioration of glucose homeostasis is reflective of a longer diabetes duration, elevated levels of glucose evaluations still were directly associated with increased risks of significant coronary stenosis in diabetes in that each 1-SD increment of glucose or HbA_1c level independently conveyed 1.73-fold to 2.11-fold elevated risks, adequately controlled for newly diagnosed cases and antidiabetic medications.

Participants with prediabetes were found to have a significantly higher prevalence of any coronary plaques but a similar prevalence of plaques causing >50% stenosis compared with NGR in this study. The close relationship between prediabetes and future CAD has been demonstrated by several studies (21–23). The enhanced atherogenic risk profile in prediabetes, such as higher BMI values and greater prevalence of hypertension, together with a prediabetes glucose level, may contribute to the elevated CAD prevalence in this study. Nevertheless, significant coronary stenosis as measured by CTA was not increased in prediabetes and was statistically lower than that in diabetes (7.6 vs. 16.4%; P = 0.027). Studies investigating biomarker trajectories leading to diabetes diagnosis reported modest changes in insulin sensitivity and secretion from normal to impaired glucose tolerance but substantial decreases during further progression to diabetes, leading to abrupt and steep increases in FPG and PPG levels during the few years (most likely 3 years) immediately before diagnosis (24,25). This is consistent with our findings on HOMA of insulin resistance and HOMA β-cell function in three groups with different glycemic status, and it might be one of the potential mechanisms of a markedly elevated risk for significant coronary stenosis in early diabetes compared with prediabetes. Noticeably, coronary plaques were found in 43.0% of NGR participants. In fact, despite glucose levels being within a normal range, 42% of these individuals were overweight or obese, 31% had hypertension, and 30% had dyslipidemia. Nearly 80% of these subjects had at least one cardiovascular risk factor, such as cigarette smoking, family history of CAD, overweight/obesity, hypertension, or dyslipidemia, indicating a noteworthy overall CAD risk profile in middle-aged Chinese adults residing in metropolitan cities such as Shanghai.

Multislice CT coronary angiography recently has emerged as a powerful imaging modality for noninvasive assessment of CAD. Except for high diagnostic accuracy for detecting CAD compared with the gold standard invasive coronary angiography (26–28), with a sensitivity of 89%, a specificity of 96%, and positive and negative predictive values of 78 and 98%, respectively (29), CTA also has been found useful in predicting future cardiac events in patients with known or suspected CAD (30–35). However, consistency is lacking regarding the use of CTA in asymptomatic subjects, particularly in high-risk asymptomatic patients, such as those with diabetes. The major concerns are unnecessary CT scan radiation, increased health care spending, and subsequent invasive cardiac procedures (36). However, recent advances in CTA technology, such as dual-source CT coronary artery angiography, have dramatically improved spatial and temporal resolution, leading to substantial improvement in image quality and significant reduction in radiation dose and use of contrast material (37).

Our findings may have important clinical implications. Currently, the American Diabetes Association consensus guidelines recommend CAD screening in diabetic individuals with cardiovascular symptoms (38). However, our study showed a markedly increased risk for the presence of significant coronary stenosis in asymptomatic patients with diabetes, even soon after its diagnosis. Furthermore, evidence has indicated similar frequencies of CAD in diabetic patients with and without angina symptoms (39,40). Therefore, cardiac testing for asymptomatic patients with diabetes should not be considered unwarranted, especially with the latest development in CT technology, which provides satisfactory cardiac images while reducing radiation dose. Meanwhile, although it could be somewhat premature to suggest that CTA examination should be routinely used for patients with recently diagnosed diabetes, our findings may imply that CTA screening should be used more aggressively in patients with diabetes, especially in those with additional risk factors such as dyslipidemia and hypertension. Nevertheless, whereas clinical effectiveness proves promising, cost-effectiveness is another important issue of concern before integrating CTA into patient care algorithms. It is possible that clinical outcomes may be sufficiently improved by use of CTA, justifying the substantially higher spending associated with its use. Therefore, longitudinal studies comparing costs of downstream testing and treatment of cardiovascular complications in diabetes with and without use of CTA are urgently needed.

Despite the findings and the clinical implications for CAD management in asymptomatic early diabetes, this study suffers from several important limitations. First, the sample size was relatively small and thus did not allow separate analyses by sex or stratification by cardiovascular risks and did not take menopause status into account. Second, although participants with NGR, prediabetes, or diabetes were well-matched for age and sex, other important CAD risk factors such as hypertension were not considered when selecting participants. Therefore, a better-matched study population should have been used. Third, coronary stenosis was classified as <50% or ≥50% in this study, rather than a much detailed quantification to include severe stenosis, which is ≥75% narrowing of coronary artery lumen and is highly urgent for clinical intervention. Moreover, although many other cardiovascular risk factors such as smoking, hypertension, and dyslipidemia were adjusted for correlation of glucose dysregulation with significant coronary stenosis in the multivariate analyses, some residual or undetected confounding could not be ruled out.

In conclusion, we used noninvasive diagnostic CTA to anatomically assess subclinical CAD in middle-aged community-dwelling Chinese adults with different glycemic status. We detected a markedly increased risk of significant coronary stenosis in early diabetes without clinical manifestation of myocardial ischemia. Although prediabetes is thought to be predictive of future cardiac events, it was not found to be associated with a higher risk of significant coronary stenosis. Follow-up studies are needed to investigate the long-term cardiovascular outcomes of CTA-confirmed subclinical atherosclerosis in patients with diabetes.

This study was supported by the grants from the Key Laboratory for Endocrine and Metabolic Diseases of Ministry of Health (1994DP131044), the Sector Funds of Ministry of Health (201002002), the National Key New Drug Creation and Manufacturing Program of Ministry of Science and Technology (2012ZX09303006-001), the National Natural Science Foundation of China (81100564, 30911120493, 81130016, and 81222008), 863 Project (2012AA02A509), and Shanghai Municipal Health Bureau (2012).

No potential conflicts of interest relevant to this article were reported.

Y.X. conceived and designed the study, analyzed and interpreted data, drafted the article, performed critical revision for important intellectual content, provided final approval of the article, provided statistical expertise, and collected and assembled data. Y.B. and W.W. conceived and designed the study, analyzed and interpreted data, drafted the article, performed critical revision for important intellectual content, provided final approval of the article, and collected and assembled data. M.L. and T.W. analyzed and interpreted data, drafted the article, provided final approval of the article, provided statistical expertise, and collected and assembled data. J.Z. analyzed and interpreted data, provided final approval of the article, and collected and assembled data. M.X. drafted the article, performed critical revision for important intellectual content, provided final approval of the article, and collected and assembled data. J.L. drafted the article, performed critical revision for important intellectual content, and provided final approval of the article. Y.C. analyzed and interpreted data, performed critical revision for important intellectual content, provided final approval of the article, and collected and assembled data. X.L. analyzed and interpreted data, performed critical revision for important intellectual content, and provided final approval of the article. S.L. conceived and designed the study, analyzed and interpreted data, performed critical revision for important intellectual content, provided final approval of the article, and provided statistical expertise. G.N. conceived and designed the study, analyzed and interpreted data, performed critical revision for important intellectual content, and provided final approval of the article. G.N. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

The authors express gratitude to all the participants in this study.