Can protein biomarkers provide an index of coronary artery calcification in patients with Type 2 diabetes?

OBJECTIVES: By exploring differences between patients with high and low coronary artery calcification score (CACS), a plasma protein biomarker might be identified as an alternative to CACS screening. METHODS: We selected stored samples (12 per group) from a cohort study of patients with Type 2 diabetes and CACS >1000 or <100 Agatston units, with matching for age, BMI, blood pressure, lipids and lipoproteins and fibrinogen. Multiplex, immunobead-based assay or ELISA measured 18 cardiovascular-related protein biomarkers. SELDI-TOF mass spectrometry (MS) screened for proteins differing significantly between high and low CACS. RESULTS: Only monocyte chemotactic protein-1 was higher in the high compared with the low CACS group but concentrations overlapped appreciably. On SELDI-TOF MS, several mass/charge ratio peak intensities significantly discriminated high and low CACS but these differences were not confirmed in larger samples from the cohort. CONCLUSIONS: Plasma protein biomarkers are unlikely to provide an effective alternative to measurement of CACS.

Coronary artery calcium and cardiovascular risk in diabetes.

Measurement of coronary artery calcium score (CACS) by electron beam tomography has been shown to a powerful predictor of coronary heart disease events in asymptomatic non-diabetic subjects. In type 2 diabetes, measurement of CACS was found to be a powerful predictor of cardiovascular events which could enhance prediction provided by established risk models. 23% of type 2 diabetic subjects with low CACS were found to be at low risk for cardiovascular events. Moreover mortality was similar for type 2 diabetic and non-diabetic subjects with undetectable coronary artery calcification. Conversely type 2 diabetic subjects with high CACS were identified who were at high cardiovascular risk. Thus not all those with type 2 diabetes are at similar cardiovascular risk. Measurement of CACS enables cardiovascular risk in type 2 diabetes to be stratified so that the level of preventive therapy could be reduced in some and intensified in others. Although prospective data for the power of CACS to predict CHD events in type 1 diabetes are lacking, measurement of CACS could help in deciding on preventive therapy in type 1 diabetes.

Coronary calcification, homocysteine, C-reactive protein and the metabolic syndrome in Type 2 diabetes: the Prospective Evaluation of Diabetic Ischaemic Heart Disease by Coronary Tomography (PREDICT) Study.

AIMS: The PREDICT Study aims to determine: (i) the association between cardiovascular risk factors and coronary artery calcification score (CACS) obtained by electron beam tomography and (ii) the predictive value of CACS for coronary heart disease (CHD) events in Type 2 diabetes. METHODS: Having previously reported relationships between CACS and conventional risk factors, we have now studied the novel risk factors, plasma high-sensitivity C-reactive protein (CRP) and homocysteine, insulin resistance, serum apoprotein A1 and B concentrations, the serum triglyceride/high-density lipoprotein cholesterol ratio and metabolic syndrome (International Diabetes Federation definition) in 573 subjects of the PREDICT Type 2 diabetes cohort. RESULTS: In univariate analyses, the only significant positive novel correlate of CACS was homocysteine (P = 0.0004). CRP was increased in those with detectable calcification, but decreased with increasing calcification score (P = 0.006). In a multivariate model that included all significant univariate correlates, CACS was independently associated with age (P < 0.0001), waist-hip ratio (P < 0.02), male gender (P < 0.05) and duration of diabetes (P < 0.05), but the association with homocysteine was no longer significant. The negative association between CACS and CRP remained in multivariate analysis, and was independent of statin use. CONCLUSIONS: Age was the major factor influencing CACS in Type 2 diabetes, with weaker contributions from waist hip-ratio and duration of diabetes. Other novel cardiovascular risk factors appear to have little positive effect.

The association of coronary calcium score and conventional cardiovascular risk factors in Type 2 diabetic subjects asymptomatic for coronary heart disease (The PREDICT Study).

AIM: To determine the association between coronary calcification score (CACS) obtained by electron beam computed tomography (EBCT) and cardiovascular risk factors in Type 2 diabetic subjects entered into a prospective cohort study. METHODS: Type 2 diabetic subjects attending routine hospital diabetic clinics without known coronary heart disease (CHD) underwent EBCT to measure CACS. Demographic data were obtained and conventional cardiovascular risk factors were measured at baseline. RESULTS: Four hundred and ninety-five subjects were assessed of whom 67.7% were male. They had a mean (SD) age of 62.9 (7.1) years, with median (inter-quartile range) duration of diabetes of 8 (4-13) years. None had a history of coronary artery disease. Forty-five per cent were receiving lipid-lowering agents (including 36% statins). In a univariate analysis, there were significant associations between increased CACS and age, duration of diabetes, male gender, waist-hip ratio (WHR), systolic blood pressure, and the use of statins. In a multivariate model adjusting for the possible interaction of these and other factors, the significant association between CACS and WHR, systolic blood pressure, male gender and statin use remained. CONCLUSIONS: The close association between CACS and WHR and the association with systolic blood pressure suggest that coronary calcification may be particularly linked to the metabolic syndrome in Type 2 diabetes.

Type 2 diabetes presenting as diabetic ketoacidosis in adolescence.

We report two black adolescent subjects who presented with diabetic ketoacidosis, but who lacked autoimmune markers and demonstrated clinical and biochemical characteristics more typical of Type 2 diabetes, including obesity, acanthosis nigricans, positive family history for Type 2 diabetes, and Type 2 diabetic dyslipidaemia. Subsequent to acute presentation, insulin was discontinued in both subjects and excellent glycaemic control was achieved with metformin therapy alone. Four months following acute presentation, both had adequate C-peptide responses to intravenous glucagon. Type 2 diabetes can present as diabetic ketoacidosis in obese adolescent subjects.

Rates of cholesterol esterification and esterified cholesterol net mass transfer between high-density lipoproteins and apolipoprotein B-containing lipoproteins in Type 1 diabetes.

AIMS: Type 1 diabetes is associated with a high incidence of coronary heart disease (CHD) despite paradoxically normal or high high-density lipoprotein (HDL) cholesterol concentrations. Triglyceride (TG) concentrations have been shown to be important determinants of two aspects of HDL metabolism: cholesterol esterification rate and esterified cholesterol (EC) net mass transfer rate between HDL and the apolipoprotein B-containing lipoproteins. In order to try to explain the paradox, we aimed to assess the relationships between plasma TG and these two processes in Type 1 diabetic compared with non-diabetic subjects. METHODS: Rates of cholesterol esterification and EC net mass transfer between HDL and the apolipoprotein B-containing lipoproteins were assessed by incubating whole plasma at 37 degrees C; intra-assay coefficients of variation were 6% and 30%, respectively. RESULTS: Ten Type 1 diabetic and 10 non-diabetic subjects, with similar ages, sex distributions, body mass indices and total cholesterol and TG concentrations, were assessed. Apolipoprotein A1, HDL unesterified cholesterol, and HDL phospholipid concentrations were greater in the Type 1 diabetic subjects. There were no significant differences in the rates of cholesterol esterification or EC net mass transfer between the groups. There were strong associations between plasma TG and the rate of cholesterol esterification and between plasma TG and the rate of EC net mass transfer in Type 1 diabetic subjects (r = 0.83, P = 0.0027 and r = 0.88, P = 0.0009, respectively) and in non-diabetic subjects (r = 0.91, P = 0.0002 and r = 0.79, P = 0.0070, respectively). However, the slopes of the associations with plasma TG were significantly steeper in the Type 1 diabetic subjects (analyses of covariance P = 0.0053 and P = 0.0146, respectively). CONCLUSIONS: Increases in TG may therefore promote more EC enrichment of atherogenic apolipoprotein B-containing lipoproteins in Type 1 diabetes while also promoting more cholesterol esterification, thereby maintaining HDL cholesterol concentrations. This could contribute to the paradox of high CHD incidence despite normal or high HDL cholesterol concentrations in Type 1 diabetes.

High-density lipoprotein composition and paraoxonase activity in Type I diabetes.

Type I diabetes is associated with a high incidence of coronary heart disease (CHD), despite a normal or even increased concentration of high-density lipoprotein (HDL) cholesterol. This paradox may be explained by changes in the antioxidant capacity of HDL, related to paraoxonase (PON1) activity. HDL compositional changes in subjects with Type I diabetes may result in changes in PON1 activity that are associated with a higher incidence of CHD. Single-vertical-spin density-gradient ultracentrifugation was used to isolate seven HDL fractions from serum according to density. PON1 activity was measured in serum and in the HDL fractions using phenyl acetate as substrate. The mean recovery of PON1 activity in the HDL fractions was 87% (S.D. 12%). CHD risk was assessed using B-mode ultrasound to measure carotid artery intima-media thickness (IMT). Groups of 35 subjects with Type I diabetes [duration of diabetes 18 years (12-32 years) [median (interquartile range)]; glycated haemoglobin 7.67% (1.17%)] and 24 non-diabetic control subjects were studied. Carotid IMT was greater in the diabetic subjects [0.60 (0.55-0.70) compared with 0.55 (0.45-0.64) mm; P=0.042] and HDL cholesterol concentration was higher [1.53 (0.36) compared with 1.32 (0.34) mmol/l; P=0.031]. There were qualitative differences in HDL in subjects with Type I diabetes: HDL particles were triacylglycerol-deplete, and there were greater numbers of the larger, more buoyant HDL particles. These properties were not those found to determine PON1 activity. PON1 activity increased as HDL particle density increased and particle size decreased; the increase in PON1 activity was associated with an increase in the ratio of the two HDL surface lipid components, phospholipid and unesterified cholesterol, as particle density increased. PON1 activity was similar in diabetic and non-diabetic subjects [121 (28) and 120 (36) micromol x min(-1) x ml(-1) respectively; P=0.887]. PON1 activity was not associated with carotid IMT in either group. Our results suggest that the PON1 activities of HDL particles relate to the density, size and composition of the particles. However, PON1 activity does not appear to contribute to the greater risk of CHD in subjects with Type I diabetes.

Total antioxidant status and coronary artery calcification in type 1 diabetes.

OBJECTIVE: Type 1 diabetes is associated with a high risk of coronary heart disease (CHD), despite the absence of dyslipidemia. Oxidative modification may render LDLs more atherogenic. We aimed to assess antioxidant status in type 1 diabetes and its association with coronary artery calcification (CAC). RESEARCH DESIGN AND METHODS: Total antioxidant status (TAS) of serum was measured using the Trolox equivalent antioxidant capacity assay in 48 type 1 diabetic and 25 nondiabetic subjects. The presence of CAC was assessed in the diabetic subjects using electron beam computed tomography. RESULTS: TAS was reduced in type 1 diabetic subjects compared with nondiabetic subjects (Mann-Whitney U test, P < 0.0001). There were associations between TAS and HbA(1c) (r = -0.43; P = 0.0026) and duration of diabetes (r = -0.35; P = 0.0157). Significant CAC was considered present if the Agatston score was >10. The diabetic subjects with significant CAC were older (P < 0.0001); had longer duration of diabetes (P = 0.0002); were more likely to have high blood pressure (P = 0.040); had higher total cholesterol concentration (P = 0.039), serum creatinine concentration (P = 0.003), and urinary albumin-to-creatinine ratio (P = 0.022); and had lower serum TAS (P = 0.018) compared with those without significant calcification. In logistic regression with CAC as the dependent variable, TAS was entered as a predictor, and the effects on its predictive value of adding other explanatory variables in bivariate analyses were assessed. The power of TAS to predict CAC was independent of many of the traditional CHD risk factors. Whereas TAS as a predictor was no longer statistically significant when age or duration of diabetes were entered into the model, the odds ratio for a TAS concentration above the median value predicting significant CAC only increased from 0.19 to 0.26 and 0.32, respectively. CONCLUSIONS: TAS is reduced in type 1 diabetes and is associated with the presence of CAC.

Correlation between carotid artery distensibility and serum vascular endothelial growth factor concentrations in type 1 diabetic subjects and nondiabetic subjects.

The relationships between serum vascular endothelial growth factor (VEGF) concentrations and vessel wall ultrasonic characteristics in type 1 diabetic and nondiabetic subjects were assessed. Serum VEGF concentration was measured, and ultrasound imaging and blood pressure recordings were performed in 41 type 1 diabetic subjects (hemoglobin A(1c) [HbA(1c)], 7.63 [1.17%]; duration of diabetes, 12 (0 to 23) years), and 50 nondiabetic subjects. Change in carotid artery luminal diameter during the cardiac cycle was measured using M-mode ultrasound, from which percentage increase in carotid artery luminal diameter was calculated; the carotid artery distensibility index was calculated as the ratio of percentage increase in carotid artery luminal diameter and pulse pressure. Serum VEGF concentration was higher in the diabetic subjects (217 [135 to 336] v 137 [80 to 237] pg/mL; P =.009). The percentage increase in carotid luminal diameter during the cardiac cycle was not significantly different between the 2 groups (12.9 [10.2 to 15.7] v 13.0 [10.6 to 15.0%]; P =.270) despite significantly greater pulse pressure in the type 1 diabetic group (55 [45 to 71] v 46 [41 to 51] mm Hg; P =.0003). The distensibility index was therefore lower in the diabetic subjects (0.24 [0.10] v 0.28 [0.08%]/mm Hg; P =.031). There was a significant negative correlation between serum VEGF concentrations and mean percentage increase in carotid luminal diameter during the cardiac cycle in the diabetic group (r = -.36, P =.021) and in the nondiabetic group (r = -.28, P =.047). This negative correlation could be strengthened by relating mean percentage increase in luminal diameter to pulse pressure to give the distensibility index. Therefore, serum VEGF concentrations correlated strongly and inversely with the distensibility index in the diabetic group (r = -.49, P =.001), in the nondiabetic group (r = -.29, P =.041), and in both groups analyzed together (r = -.42, P <.0001). Vessel wall distensibility may be an important determinant of serum VEGF concentrations in both diabetic and nondiabetic populations and may underlie the previously observed association between blood pressure and serum VEGF concentrations. The pathophysiologic relevance of these findings remains to be elucidated.

The relationship between active renin concentration and plasma renin activity in Type 1 diabetes.

AIMS: Circulating activity of the renin-angiotensin-aldosterone system (RAAS) can be assessed by measuring plasma active renin concentration (ARE), as well as by measuring plasma renin activity (PRA). We aimed to assess the relationships between ARE and PRA in Type 1 diabetic compared with non-diabetic control subjects. We also assessed concentrations of the active renin precursor, prorenin. PATIENTS AND METHODS: Thirty-five Type 1 diabetic subjects and 34 non-diabetic control subjects were assessed. Groups had similar ages, sex distributions, body mass indices, systolic and diastolic blood pressures. PRA was measured by radioimmunoassay of angiotensin I generation from endogenous substrate. ARE and total renin concentration (TRE) were measured by immunoradiometric assay (Nichols Institute Diagnostics, USA). Prorenin concentration was calculated as the difference between ARE and TRE. RESULTS: PRA was significantly lower in Type 1 diabetic than in control subjects (0.8 (0.4-1.1) vs. 1.1 (0.9-1.9) pmol/ml per h; P < 0.005), while ARE was similar (17 (9-33) vs. 18 (15-25) mU/l; P = 0.548). PRA (loge transformed) correlated strongly with ARE in diabetic (r = 0.49; P = 0.003) and control subjects (r = 0.59; P = 0.0002), but there was significant vertical separation of the regression lines for the two groups (P < 0.0001). Prorenin concentrations were significantly higher in Type 1 diabetic subjects (249 (170-339) vs. 171 (153-219) mU/l; P = 0.005). Diabetic subjects with high prorenin concentrations (> 400 mU/l (control mean + 3 SD)) were more likely to have microalbuminuria (P = 0.027) and peripheral neuropathy (P = 0.049). CONCLUSIONS: Type 1 diabetes is associated with an altered relationship between ARE and PRA, such that ARE is higher for a given PRA compared with non-diabetic control subjects. Both ARE and PRA are used to assess circulating RAAS activity. The altered relationship between the two in Type 1 diabetic subjects suggests that neither parameter alone is necessarily an adequate and reliable index of such activity. Higher prorenin concentrations, particularly in association with microvascular complications, were confirmed in the Type 1 diabetic subjects. Diabet. Med. 18, 451-458 (2001)

Smoking is associated with increased hepatic lipase activity, insulin resistance, dyslipidaemia and early atherosclerosis in Type 2 diabetes.

We have studied the relationships between hepatic lipase activity, smoking, dyslipidaemia insulin resistance, and early atherosclerosis in 67 Type 2 diabetic subjects, 47 non-smokers and 20 smokers. Insulin resistance was measured using an insulin modified frequently sampled intravenous glucose tolerance test. Early atherosclerosis was assessed using high-resolution ultrasound to measure carotid intima media thickness (IMT) and an arterial ultrasonic score (AUS). Smokers had higher serum cholesterol and triglyceride, lower HDL and HDL2 cholesterol as well as increased hepatic lipase activity. They were also more insulin resistant than non-smokers. Smokers also had higher patient AUS scores. On multiple regression analysis, hepatic lipase activity emerged as the most significant variable affecting patient AUS. We suggest that smoking accentuates the dyslipidaemia of Type 2 diabetic subjects and this is associated with increased hepatic lipase activity. This may be one mechanism whereby smoking further increases the risk of cardiovascular disease in Type 2 diabetes.

Insulin resistance, cardiovascular risk factors and ultrasonically measured early arterial disease in normotensive Type 2 diabetic subjects.

BACKGROUND: The objective of the study was to examine the relationship between serum fasting insulin, insulin sensitivity S(i), cardiovascular risk factors, and asymptomatic early atherosclerosis in normotensive Type 2 diabetic subjects. METHODS: Specific insulin was measured using an enzyme-linked immunosorbent assay (ELISA) and insulin sensitivity was assessed with an insulin-modified frequently sampled intravenous glucose tolerance test (FSIVGTT). Early atherosclerotic change was assessed using carotid intima media thickness (IMT) and an arterial ultrasound score (AUS) measured by high-resolution B-mode ultrasound. RESULTS: On bivariate analysis, there was a positive correlation between S(i) and high density lipoprotein (HDL) cholesterol (r(s)=0.27, p<0.05), and a negative correlation between S(i) and body mass index (BMI) (r(s)=-0.42, p<0.001), HbA(1c) (r(s)=-0.29, p<0.05) and serum triglyceride (r(s)=-0.30, p<0.05). There was a positive correlation between carotid IMT and age (r(s)=0.41, p<0.0005), and a positive association with male sex (p<0.0001) as well as with smoking (p<0.0001). However, we found no correlation between carotid IMT and fasting specific insulin (r(s)=-0.04) or S(i) (r(s)=-0.08). On multiple regression analyses, only age and serum triglycerides appeared to be significant independent variables with respect to carotid IMT whereas age, male sex and smoking emerged jointly significant with respect to AUS. There were no independent associations between carotid IMT or AUS with other variables including using either fasting specific insulin or S(i) as markers on insulin resistance separately. CONCLUSION: Carotid IMT and AUS in Type 2 diabetes are closely associated with age, male sex and smoking. The relationships between serum insulin and insulin resistance with ultrasonically measured early arterial disease in Type 2 diabetes remain unclear.

Prevalence of renal artery stenosis in subjects with type 2 diabetes and coexistent hypertension.

OBJECTIVE: To assess the prevalence of renal artery stenosis (RAS) in subjects with type 2 diabetes and coexistent hypertension by using magnetic resonance angiography (MRA) of the renal arteries, to assess clinical and biochemical predictors of RAS, and to assess the hemodynamic significance of RAS, by using the captopril test (a measure of the response of plasma renin activity to a single oral dose of captopril). RESEARCH DESIGN AND METHODS: A total of 117 subjects with type 2 diabetes and coexistent hypertension between 40 and 70 years of age and with creatinine concentrations < 150 micromol/l were recruited from two inner-city general diabetes clinics. All subjects underwent MRA of the renal arteries. In a subgroup of 85 subjects, data concerning possible clinical and biochemical predictors of RAS were collected, and the captopril test was performed. For comparison of a continuous variable between subjects with a positive MRA and those with a negative MRA, the Mann-Whitney test was used. For comparison of a discrete variable between subjects with a positive MRA and those with a negative MRA, Fisher's exact test was used. RESULTS: The prevalence of RAS detected by using MRA in 117 hypertensive type 2 diabetic subjects was 17%; 19 subjects had unilateral RAS, and only 1 subject had bilateral RAS. A femoral bruit was significantly more common in subjects with a positive MRA versus subjects with a negative MRA (21 vs. 0%; Fisher's exact test P < 0.005); however, other clinical features of atherosclerotic disease were not statistically associated. Greater duration of hypertension and treatment with statins were features of subjects with RAS (P < 0.05). The captopril test was negative in all subjects, although the antihypertensive response to oral captopril was significantly greater in subjects with RAS detected by MRA. CONCLUSIONS: RAS is common in hypertensive type 2 diabetic subjects. The presence of a femoral bruit is a useful predictive clinical marker. The captopril test is not useful in predicting the hemodynamic significance of RAS in this patient group.

Variation in the PPARalpha gene is associated with altered function in vitro and plasma lipid concentrations in Type II diabetic subjects.

AIMS/HYPOTHESIS: Peroxisome proliferator activated receptor alpha (PPARalpha) regulates genes involved in lipid metabolism, haemostasis and inflammation, in response to fatty acids and fibrates, making it a candidate gene for risk of dyslipidaemia, atherosclerosis and coronary artery disease. Plasma non-esterified fatty acids are increased in subjects with Type II (non-insulin-dependent) diabetes mellitus, suggesting that PPARalpha could link Type II diabetes and dyslipidaemia, and affect response to fibrates. This has been investigated in association studies in healthy and diabetic subjects and in vitro studies. METHODS: The human PPARalpha gene was isolated and screened for variation by single strand conformation polymorphism analysis. Genotypes were determined for 129 Type II diabetic subjects and 2508 healthy men. The association with plasma lipid concentrations was examined. The function of the V162 variant was examined in co-transfection assays. RESULTS: We identified two polymorphisms, one in intron 3 and a missense mutation, leucine 162 to valine, in the DNA binding domain. In Type II diabetic patients, V162 allele carriers had higher total cholesterol, HDL cholesterol and apoAI whereas intron 3 rare allele carriers had higher apoAI concentrations. By contrast, no effect was observed in healthy rare allele carriers. In vitro, the V162 variant showed greater transactivation of a reporter gene construct. CONCLUSION/INTERPRETATION: Naturally occurring variation alters PPARalpha function, influencing plasma lipid concentrations in Type II diabetic patients but not healthy people. This demonstrates that PPARalpha is a link between diabetes and dyslipidaemia, and so could influence the risk of coronary artery disease, the greatest cause of morbidity and mortality in Type II diabetes.