Introduction
Prediabetes is a pathological condition caused by initial disruption of glucose metabolism before development of diabetes. Prediabetes is accepted as a risk factor of diabetes and cardiovascular diseases [1]. Vascular inflammation is triggered by the effects of hyperglycemia in prediabetic patients and subclinical atherosclerotic changes detected in vascular system [2]. Furthermore, subclinical atherosclerotic changes occur with similar incidence in cerebral, carotid and coronary arteries [3]. For that reason, carotid intima-media thickness (CIMT) has been evaluated for demonstration of subclinical atherosclerosis in many studies due to its availability for ultrasound examination (US) [4, 5]. Increased CIMT indicates subclinical atherosclerosis and possible cardiovascular problems in the future [6].
Another measurement used for the diagnosis of vascular pathology is the ankle-brachial pressure index (ABPI). Lower ABPI scores are considered to be associated with peripheral artery disease [7], while higher ABPI is suspicious for calcified vessels and arterial stiffness. ABPI is also accepted as an indicator of atherosclerosis in diabetic patients [8]. Furthermore, the blood pressure index (BPI) is a novel marker for demonstration of increased artery stiffness and atherosclerosis among subjects with the risk of cognitive impairment [9].
It is well known that prediabetic patients also have an increased cardiovascular risk. Demonstration of subclinical atherosclerosis in prediabetic patients is important for early prevention of cardiovascular diseases. Associations between BPI, CIMT and ABPI have not been well clarified in prediabetic patients. We believe that BPI, CIMT and ABPI may be good indicators of subclinical atherosclerosis in prediabetic patients. We evaluated BPI, CIMT and ABPI in prediabetic patients and compared these values in diabetic and control groups.
Material and methods
The present research was designed as an age- and gender-adjusted case-control study. Ninety persons referred to the internal medicine clinic were enrolled in the study. Exclusion criteria were familial hyperlipidemia, infection diseases, hematological diseases, coronary artery disease, cerebrovascular diseases, asthma, chronic obstructive pulmonary disease, hypertension, hypo/hyperthyroidism and smoking.
The present study was conducted in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee. All participants signed an informed consent prior to the study.
Demographic characteristics, blood count and biochemical parameters were obtained from their files.
Definitions
Diagnoses of prediabetes and diabetes were made in accordance with the criteria of the American Diabetes Association (ADA) [10].
The criteria for prediabetes were fasting serum glucose 100-125 mg/dL or postprandial serum glucose 140-200 mg/dL after oral glucose tolerance test (OGTT) or HbA1c level within 5.7-6.4% according to the ADA guidelines. The criteria for diabetes were fasting serum glucose ≥126 mg/dL, postprandial serum glucose 200 mg/dL and HbA1c >6.5% in accordance with the ADA guidelines [10]. The control group consisted of individuals without any chronic diseases or previous medication.
Anthropometric measurements
Body weight (kg) was measured using an electronic scale (Seca 770, Seca United Kingdom, Birmingham, UK) after sufficient calibration (sensitivity ± 0.1 kg). Measurements were rounded up to 100 g. Height (m) was measured by stadiometer (Seca 222).
Body mass index (BMI) was determined as body weight (kg) divided into body surface area (m2). All measurements were carried out in a standing position in light clothing without shoes.
Ankle-brachial pressure index and blood pressure index
Systolic blood pressure was measured using Medisana MTM U80C sphygmomanometer (Medisana AG, Neuss, Germany) and portable Doppler device on the upper arm and the ankle. ABPI was calculated by dividing the ankle pressure (dorsalis pedis) into systolic pressure of the brachial artery for each lower extremity. BPI was calculated as systolic blood pressure divided into diastolic blood pressure [11, 12]. Measurements were performed in a silent room at a fixed temperature after keeping the supine position for 10 minutes.
Carotid intima-media thickness
CIMT was measured using ACUSON Sequoia US device (high-resolution US machine, Image Point HX, Agilent) equipped with a 5-10 MHz linear probe. CIMT was measured in supine position and the neck was slightly turned to the opposite side for analysis of carotid arteries. Maximum value after bilateral measurements was accepted as CIMT in a certain patient. Presence or absence of plaques was noted. All scans were performed by an experienced radiologist who was not previously informed about clinical characteristics of patients.
Statistical analysis
Statistical analysis was conducted using SPSS 20 software for Windows (IBM Corp., Armonk, NY, USA). Normal distribution of data was evaluated by Kolmogorov-Smirnov test. Numeric variables with normal distribution were presented as mean ± standard deviation, whereas those without normal distribution were presented as median (min-max). Differences of numeric variables with normal distribution were evaluated by ANOVA test and numeric variables without normal distribution were analyzed using the Kruskal-Wallis H test. The chi-square test was used to compare categorical data. The association between numeric variables was analyzed using the Spearman or Pearson correlation analysis. ROC analysis and the Youden index were used to find out cut-off values for markers in predicting subclinical atherosclerosis in prediabetes patients. The cut-off value with the highest Youden index was accepted as the best one. Differences were significant at p-value <0.05.
Results
Mean age and BMI were 49.6±9.9 years and 32.4±7.5 kg/m2, respectively. There were 52.4% of females. Comparison of socio-demographic parameters, anthropometric measurements and laboratory parameters is shown in Table 1. Comparison of lipid panels (total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides) did not reveal any significant difference between prediabetic, diabetic patients and the control group (p>0.05). Hemoglobin level was 13.5±1.0 g/dL in the prediabetic group, 13.3±1.1 g/dL in the control group and 14.5±1.7 g/dL in the diabetic group. Difference in hemoglobin level between diabetic patients and other groups was significant (p=0.02). Leukocyte count was 6.4·103 µL–1 in prediabetic patients, 7.6·103 µL–1 in the control group and 8·103 µL–1 in the diabetic group. Significant difference was detected between the control group, diabetic and prediabetic groups (p<0.001). Fasting plasma glucose was 99 (81-121) mg/dL in the prediabetic group, 96 (76-99) and 131 (93-232) mg/dL in the control and diabetic groups, respectively. HbA1c was 5.9% (5.7-6.2%) in the prediabetic group, 5.5% (4.7-5.6%) in the control group and 7.0% (6.0-13.0%) in the diabetic group. Differences between fasting plasma glucose and HbA1c were significant between prediabetes, diabetes and the control group (p<0.001). The ratio of monocytes to HDL cholesterol (MHR) was 12.3, 13.2 and 10.8 in prediabetic, diabetic patients and control subjects, respectively. Between-group differences were significant (p=0.005).
Table 1. Demographic, anthropometric and laboratory parameters.
|
Variables |
Control (n=30) |
Prediabetes (n=30) |
Diabetes (n=30) |
p-value |
|
Age, years |
51.4±8.6 |
51.8±7.7 |
51.7±9.7 |
0.123 |
|
Height, cm |
159.6±7.2 |
157.8±7.4 |
163.9±11.3 |
0.291 |
|
Weight, kg |
92.3±14.1 |
100.5±20.4 |
89.8±22.4 |
0.071 |
|
BMI, kg/m2 |
38.6±6.1 |
39.5±7.4 |
34.7±7.8 |
0.506 |
|
Laboratory parameters |
||||
|
Leucocytes, µl–1 |
7.6 (4.8-11.3) |
6.4 (4.6-9.5)b |
8.0 (5.4-15.2)a |
<0.001* |
|
Monocytes, µl–1 |
0.54 (0.3-1.2) |
0.45 (0.3-1.7)b |
0.66 (0.4-1.4)a |
0.001* |
|
Hemoglobin, g/dl |
13.3±1.1b |
13.5±1.0b |
14.5±1.7a,c |
0.002* |
|
Platelets, *103µl–1 |
266 (172-449) |
275 (178-425) |
268 (196-459) |
0.977 |
|
Glucose, mg/dl |
96 (76-99)a,b |
99 (81-121)b,c |
131 (93-302)a,c |
<0.001* |
|
HbA1c, % |
5.5 (4.7-5.6)a,b |
5.9 (5.7-6.2)b,c |
7.0 (6.0-13.0)a,c |
<0.001* |
|
AST, U/l |
24 (8-50) |
20 (10-43) |
23 (16-46) |
0.062 |
|
Creatinine, mg/dl |
0.8 (0.6-1.0) |
0.8 (0.6-1.0) |
0.86 (0.4-1.1) |
0.244 |
|
Total cholesterol, mg/dl |
218±34 |
227±43 |
214±47 |
0.446 |
|
LDL cholesterol, mg/dl |
138±29 |
145±32 |
126±36 |
0.069 |
|
HDL cholesterol, mg/dl |
48 (33-84) |
53 (35-72) |
48 (30-78) |
0.390 |
|
Triglyceride, mg/dl |
138 (76-341) |
140 (55-318) |
195 (54-422) |
0.080 |
|
MHR |
10.8 (6.1-28.5) a,b |
12.3 (4.8-33.5) b,c |
13.2 (5.3-31.3) a,c |
0.005* |
|
Other measurements |
||||
|
BPI |
1.76±0.22a,b |
1.82±0.26b,c |
2.08±0.50a,c |
0.034* |
|
ABPI |
1.08±0.12a,b |
1.10±0.12b,c |
1.18±0.11a,c |
0.004* |
|
CIMT, mm |
0.71±0.18a,b |
0.84±0.20b,c |
0.91±0.22a,c |
0.001* |
|
Plaque, % |
1 (3.33) |
7 (23.3) |
5 (16.7) |
0.081 |
CIMT was 0.71±0.18 mm in the control group, 0.84±0.20 mm in prediabetic patients and 0.91±0.22 mm in diabetic patients (p<0.001) (Fig. 1). There was no significant between-group difference in terms of plaque (p=0.08).
Fig. 1. CIMT in study groups.
ABPI was 1.18±0.11 in diabetic patients, 1.10±0.12 in prediabetic patients and 1.08±0.12 in the control group. Difference was significant for diabetic patients compared to prediabetic patients and the control group. However, differences between prediabetes group and control group were not significant (p=0.004 and p=0.65, respectively) (Fig. 2). BPI was 2.1±0.5 in diabetic patients, 1.8±0.3 in prediabetic patients and 1.8±0.2 in the control group (p<0.001) (Fig. 3).
Fig. 2. ABPI in study groups.
Fig. 3. BPI in study groups.
Correlation analysis revealed significant positive relationships between CIMT and age (r=0.377, p<0.001), HbA1c (r=0.416, p<0.001), MHR (r=0.361, p<0.01), BPI (r=0.284, p<0.01), fasting plasma glucose (r=0.296, p=0.005) and between ABPI and HbA1c (r=0.229, p=0.03). Significant positive correlations between BPI and age (r=0.221, p=0.037), HBA1c (r=0.342, p<0.01), fasting plasma glucose (r=0.232, p=0.028) were also detected (Table 2).
Table 2. Correlation analysis.
|
Variable |
CIMT |
ABPI |
BPI |
|||
|
r |
p-value |
r |
p-value |
r |
p-value |
|
|
Age |
0.377 |
<0.001* |
–0.132 |
0.214 |
0.221 |
<0.037* |
|
CIMT |
1 |
— |
–0.004 |
0.972 |
0.342 |
<0.001* |
|
BPI |
0.342 |
<0.001* |
0.256 |
0.345 |
1 |
|
|
ABPI |
–0.004 |
0.972 |
1 |
— |
–0.178 |
0.093 |
|
Total cholesterol |
–0.021 |
0.846 |
–0.019 |
0.061 |
0.061 |
0.565 |
|
LDL cholesterol |
–0.076 |
0.477 |
–0.058 |
0.062 |
0.069 |
0.516 |
|
Glucose |
0.296 |
0.005** |
0.159 |
0.135 |
0.232 |
0.028** |
|
HDL cholesterol |
0.117 |
0.274 |
–0.077 |
0.469 |
–0.055 |
0.608 |
|
Triglyceride |
0.061 |
0.568 |
–0.033 |
0.759 |
0.145 |
0.173 |
|
HbA1c |
0.416 |
<0.001** |
0.229 |
0.030** |
0.256 |
0.015** |
|
MHR |
0.361 |
0.008** |
0.180 |
0.784 |
0.124 |
0.243 |
ROC analysis was applied to estimate predictive ability of measurements and optimal cut-off values for prediabetes. The optimal cut-off values, as well as sensitivity and specificity of each index are summarized in Table 3. MHR with largest AUC (0.684) was followed by BPI (0.644) and CIMT (0.678) (Fig. 4).
Table 3. Diagnostic accuracy of markers for subclinical atherosclerosis in prediabetes.
|
Variable |
AUC |
Optimal cut-off value |
Sensitivity (%) |
Specificity (%) |
Youden index |
|
CIMT |
0.678 |
0.73 |
76.7 |
58.3 |
0.35 |
|
BPI |
0.664 |
1.81 |
70.0 |
60.0 |
0.30 |
|
MHR |
0.684 |
12.64 |
60.0 |
75 |
0.35 |
Fig. 4. ROC curves for MHR (a), CIMT (b) and BPI (c).
Discussion
Over the last few years, subclinical atherosclerosis has been examined widely in prediabetic patients. However, to our knowledge, this is the first study examining the role of BPI as a new simple and non-invasive marker specifically in prediabetic patients. This parameter also correlates with CIMT. Furthermore, we have found that BPI, CIMT and ABPI were higher in diabetic patients than in other groups. Moreover, prediabetic patients had higher BPI and CIMT compared to the control group. Thus, we can suppose that atherosclerotic process may start before diabetic period.
Prediabetes is accepted as a risk factor of diabetes mellitus, cardiovascular diseases, myocardial infarction and stroke [13, 14]. The increased risk of cardiovascular diseases in prediabetic patients has been reviewed in many previous studies [15, 16]. Inflammation and oxidative stress induced by hyperglycemia lead to development of atherosclerosis [17]. Demonstration of subclinical atherosclerosis is also important for prediction of cardiovascular disease [18]. Therefore, many markers have been evaluated as indicators of atherosclerosis [19-22]. High-resolution B-mode US enables non-invasive assessment of arterial wall thickening and progression of subclinical atherosclerosis. For that reason, CIMT was also accepted as an indicator of subclinical atherosclerosis [23]. Furthermore, increase of CIMT is considerably associated with atherosclerosis in other segments of arterial system [24]. Increased CIMT was shown as a strong predictive factor of cardiovascular morbidity and mortality [3]. In fact, it is known that diabetes mellitus is associated with cardiovascular diseases. Significant increase of CIMT was also detected in diabetic patients [25]. Moreover, a few studies found thickened carotid intima in prediabetic patients compared to healthy population [19].
CIMT was also found to be significantly higher in prediabetic and diabetic patients in the present study. Considering literature data, we can state that atherosclerosis process starts with prediabetic progression and there is an increased risk of cardiovascular diseases throughout this period. It may be assumed that increased CIMT is detected in parallel with subclinical atherosclerosis induced by hyperglycemia. Furthermore, we detected positive correlations between age, HbA1c, MHR, BPI, fasting plasma glucose and CIMT.
ABPI as essential aspect in diagnosis of peripheral artery disease was investigated in diabetic patients in many recent clinical studies [26-28]. It is accepted as an important screening method for diagnosis of peripheral artery disease in diabetic patients [29]. Previous study detected lower ABPI in prediabetic and diabetic elderly smoking patients [30]. ABPI was also detected to be significantly lower when diabetic and prediabetic patients were scanned for peripheral artery disease. Negative correlation between HbA1c and ABPI was also observed. Moreover, another study has shown that ABPI was significantly lower in prediabetic and hypertensive patients [31].
ABPI increment is also important for calcification and stiffness of vascular system. More importantly, previous studies showed that ABPI may be an indicator of arterial stiffness in patients without peripheral artery disease. ABPI 1.0-1.5 particularly correlated with arterial stiffness [32]. Another study comprising 167 patients detected significantly higher CIMT and ABPI in recently diagnosed diabetic patients compared to prediabetic patients and the control group [33].
Other blood pressure indices were also well studied for demonstration of subclinical atherosclerosis in previous trials. Recently, it was demonstrated that higher BPI as a novel non-invasive simple tool could be associated with arterial stiffness and atherosclerosis [9]. Although the role of atherosclerosis has been postulated in cardiovascular adverse outcomes, there are no data on association between blood pressure and subclinical atherosclerosis in prediabetic patients. Our study ensures additional data on the role of blood pressure for identifying subclinical atherosclerosis in prediabetic patients. Furthermore, we detected positive correlations between age, HbA1c, CIMT and BPI.
In the present study, we detected that BPI and ABPI were significantly higher in diabetic patients compared to the control group and prediabetic patients. None of the patients enrolled in this study were diagnosed with peripheral artery disease. Therefore, ABPI >0.9 in such patients should evoke a possible association with arterial stiffness. Similar to literature data on positive association between arterial stiffness and HbA1c, we detected a positive correlation between ABPI and HbA1c. In our study, all participants had ABPI >1. This finding may be related to arterial stiffness. All three groups of patients had mean BMI >30 kg/m2. Obesity in most participants might explain an increased arterial stiffness. On the other hand, higher ABPI in prediabetic patients compared to the control group may be related to increased risk of arterial stiffness in this early stage.
Inflammation and oxidative stress are important mechanisms of development and progression of atherosclerosis [34]. Monocytes are essential in this process. Activated monocytes interact with endothelium and cause overexpression of proinflammatory cytokines and adhesion molecules. Monocytes transform into macrophages on arterial vascular wall and create foamy cells as a result of oxidized low-density lipoprotein (LDL) cholesterol digestion [35]. High-density lipoprotein (HDL) cholesterol molecules prevent migration of macrophages and cause removal of oxidized cholesterol from these cells. Recent studies also indicated the role of HDL cholesterol in control of activation, adhesion and inflammation of monocytes [36]. In addition to anti-inflammatory and antioxidant effects of HDL cholesterol molecules, the last ones increase expression of endothelial nitric oxide synthase and enables vascular relaxation [37]. Therefore, monocytes present a pro-inflammatory and pro-oxidant effect. However, HDL cholesterol acts as a reverse factor during these reactions. MHR may be considered as an indicator of inflammation and oxidative stress [38, 39].
We also showed that MHR was significantly increased in prediabetic and diabetic patients. We think that oxidative stress and inflammation may be important indicators of pathophysiology of diabetes.
Conclusion
The present study has reviewed the indicators of subclinical atherosclerosis in prediabetic patients. Although increased BPI and CIMT as indicators of subclinical atherosclerosis were observed in both diabetic and prediabetic patients, we detected a significant increase of ABPI in diabetic patients only. We also showed that MHR as an indicator of oxidative stress and inflammation was significantly higher in prediabetic and diabetic patients.
Limitations
The present study has some limitations. Firstly, the study was conducted in one center with small sample size. Secondly, we could not check possible daily changes in blood pressure measurements. Further prospective studies on this issue are needed.
This study did not receive any specific grant from any funding agency in the public, commercial or non-profit sector.
The authors declare no conflicts of interest.