Pro-protein convertase subtilisin/kexin 9 concentrations correlate with coronary artery disease atheroma burden in a Pakistani cohort with chronic chest pain

OBJECTIVE

To determine the relationship between proprotein convertase subtilisin kexin 9 (PCSK9) levels and atheroma burden in Pakistanis presenting to an ambulatory centre with chest pain.

METHODS

A prospective matched case-control study of 400 patients selected for presence/absence of angiographic disease referred between 2001 and 2003. A comprehensive cardiovascular disease risk factor profile was assessed including demographics, environmental and biochemical risk factors including insulin resistance and PCSK-9 levels. Coronary atheroma burden was quantified by Gensini score.

RESULTS

In this population, PCSK-9 levels were weakly correlated (r = 0.23) with male gender (p = 0.06) and number of diabetes years (p = 0.09), and inversely with log10 of lipoprotein (a) concentration (p = 0.07) but not LDL-C. In multiple regression analysis, Gensini score was associated with age (p = 0.002), established angina (p = 0.001), duration of diabetes (p = 0.05), low HDL-C (p < 0.001), lipoprotein (a) (p = 0.01), creatinine (p < 0.001), C-Reactive Protein (p = 0.02) and PSCK-9 (p = 0.05) concentrations. PCSK9 added to the regression model. Neither total cholesterol nor LDL-C were significant risk factors in this study.

CONCLUSIONS

Proprotein convertase subtilisin kexin 9 concentrations are correlated with atheroma burden in Indian Asian populations from the sub-continent, not taking statin therapy, independent of LDL-C or other CVD risk factors.

Angiographic coronary artery disease and high-sensitivity troponin-T in a native Pakistani cohort presenting with chronic chest pain

OBJECTIVE

To determine the relationship between troponin-T levels and atheroma burden in Pakistanis presenting to an ambulatory centre with chest pain.

METHODS

A prospective case-control study of 400 patients selected for presence/absence of angiographic disease referred between 2001 and 2003. A comprehensive cardiovascular disease (CVD) risk factor profile was assessed including demographics, environmental and biochemical risk factors including insulin resistance and troponin-T levels. Coronary atheroma burden was quantified by Gensini score.

RESULTS

Clinically significant elevated troponin-T levels (> 30 pmol/l) were found in 40 patients (10%) with equal numbers in groups selected with or without angiographic disease. Troponin-T elevation (> 13 pmol/l) was present in 59 vs. 47 patients (30% vs. 24%; p = 0.04). Troponin-T levels did not correlate with any measured demographical, environmental, drug therapy or biochemical risk factor. No difference was found in concentrations of lipids, apolipoproteins, insulin resistance, C-reactive protein or sialic acid in cohorts stratified by troponin-T concentrations. In univariate analysis comparing patients with high (> 30 pmol/l) and low troponin-T levels (< 13 pmol/l) higher plasma total protein (91 g/l vs. 85 g/l; p = 0.01), increased immunoglobulin levels (41 g/l vs. 36 g/l; p = 0.02) and prevalence of hyperparathyroidism (40% vs. 21%; p = 0.04) were associated with higher troponin-T concentrations.

CONCLUSIONS

This study shows that measurement of troponin-T is not an alternative to imaging in an Indian asian population, but that it does identify a separate potentially high-risk population that would not be identified by the use of imaging alone which is potentially at higher risk of CVD events.

Insulin resistance phenotypes and coronary artery disease in a native Pakistani cohort

OBJECTIVE

To determine the relationship between insulin resistance (IR) and atheroma burden in Pakistanis.

METHODS

A prospective case-control study of 400 patients selected for the presence/absence of angiographic disease. Coronary atheroma burden was quantified and IR and cardiovascular risk factors were measured.

RESULTS

The patients were divided into two groups by QuickI score. Waist circumference (90 +/- 10 vs. 90 +/- 9 cm; p = 0.7) was similar but the groups differed in body mass index (26.5 +/- 3.7 vs. 24.2 +/- 3.5 kg/m(2); p < 0.001) and waist:hip ratio (0.94 +/- 0.09 vs. 0.90 +/- 0.06; p < 0.001). Lipid parameters showed similar high-density lipoprotein cholesterol (HDL-C) (0.77 +/- 0.23 vs. 0.82 +/- 0.22 mmol/l; p = 0.1) differences in triglycerides [1.32 (0.08-3.98) vs. 1.12 (0.37-3.61) mmol/l; p = 0.01], but no difference in low-density lipoprotein cholesterol (LDL-C) (2.75 +/- 1.00 vs. 2.90 +/- 0.94 mmol/l; p = 0.14). In insulin-resistant patients C-reactive protein (CRP) [6.8 (0.3-175.1) vs. 3.9 (0.2-57.9) mg/l: p < 0.001], sialic acid (82 +/- 14 vs. 77 +/- 15 mg/l; p < 0.001) aspartate transaminase [24 (7-171) vs. 21 (7-83) IU/l; p < 0.001] and gamma-glutamyl transferase [27 (8-482) vs. 21 (7-168) IU/l; p = 0.005] levels were increased. In insulin-resistant patients (n = 187), coronary artery disease (CAD) burden correlated (r = 0.55) with age (beta = 1.62; p < 0.001), HDL-C (beta = -53.2; p < 0.001), lipoprotein (a) (beta = 11.4; p = 0.007), smoking (beta = 7.98; p = 0.004), CRP (beta = 6.06; p = 0.03) and QuickI index (beta = -146; p = 0.04). In contrast in insulin-sensitive patients (n = 178) CAD burden (r = 0.46) correlated with LDL-C (beta = 10.0; p = 0.02), CRP (beta = 7.13; p = 0.03), HDL-C (beta = -38.1; p = 0.03), and weakly with age (beta = 0.73; p = 0.07) and smoking (beta = 5.52; p = 0.09).

CONCLUSIONS

Indian Asians show a dichotomous insulin-resistance phenotype. Atheroma is associated with low HDL-C and inflammation associated in all but LDL-C is a factor in the insulin sensitive in contrast to age and extent of IR in the insulin resistant.

Erectile dysfunction and statin treatment in high cardiovascular risk patients

Erectile dysfunction (ED) has been associated with risk factors for atherosclerosis. Medications used for atherosclerosis have also been implicated in ED. The aim of this study is to investigate the relationship of erectile function to cardiovascular risk factors and specific drug therapies before and after 6 months of statin therapy. In this prospective observational study, International Index of Erectile Function (IIEF) scores were measured in 93 men attending cardiovascular risk clinics. Cardiovascular risk factors and drug therapies were assessed prior to initiation and after 6 months of statin therapy. Prior to statin therapy, the median IIEF score was 21 (range 0-25), and 57% had impairment of erectile function. After statin therapy, IIEF scores were reduced to 6.5 (range 0-25) (p < 0.001), and 22% experienced new onset ED. Before statin therapy no correlation was observed between IIEF score and any individual cardiovascular risk factor. After 6 months of statin therapy, correlations were observed between lower IIEF scores (r = 0.62; p < 0.001) and age and diabetes and weakly with smoking. Differences in dose, relative efficacy or relative lipophilicity of statin prescribed showed no correlation with change in IIEF score. This study suggests ED following statin therapy is more likely in patients with severe endothelial dysfunction due to established cardiovascular risk factors including age, smoking and diabetes.

Metabolic syndrome and risk of coronary heart disease in a Pakistani cohort

OBJECTIVE

To assess the relation of the metabolic insulin resistance syndrome (M-IRS) with coronary heart disease (CHD) in Pakistani patients.

SUBJECTS

200 patients with angiographic disease (CHD(+)) matched with 200 patients with chest pain without occlusive disease (CHD(-)).

DESIGN

Prospective case-control study.

SETTING

Tertiary referral cardiology unit in Pakistan.

RESULTS

M-IRS was present in 37% of CHD(+) versus 27% of CHD(-) patients by criteria for white patients or 47% versus 42%, respectively, by Asian criteria (p < 0.001). After adjustment for other risk factors, M-IRS was not a significant predictor for CHD or angiographic disease. Age (p = 0.03), smoking (p < 0.001), diabetes-years (p = 0.003), sialic acid (p = 0.01), and creatinine (p = 0.008) accounted for the excess risk of CHD. Similarly, age (p = 0.005), creatinine (p < 0.001), cigarette pack-years (p = 0.02), diabetes-years (p = 0.003), and sialic acid (p = 0.08) were predictors of greater angiographic disease. M-IRS differed between Pakistani and white patients, as waist circumference correlated weakly (r = -0.03-0.08, p = 0.45-0.52) with triglycerides, high density lipoprotein cholesterol, systolic blood pressure, or glucose. Sialic acid was the only inflammatory marker associated with M-IRS.

CONCLUSIONS

Despite strong associations between individual risk factors associated with M-IRS and a univariate association between M-IRS and CHD in native Pakistanis, the principal discriminant risk factors in this group are age, smoking, inflammation, diabetes-years, and impaired renal function. The poor sensitivity of M-IRS for CHD reflects the high underlying prevalence of M-IRS, thus reducing sensitivity, confounding by other urban lifestyle traits, or a lack of association of waist circumference with M-IRS risk factors. The definition of M-IRS may have to be revised to increase its power as a discriminant risk factor for CHD in Pakistani populations.

Molecular genetics and phenotypic characteristics of MODY caused by hepatocyte nuclear factor 4alpha mutations in a large European collection

AIMS/HYPOTHESIS

Heterozygous mutations in the gene of the transcription factor hepatocyte nuclear factor 4alpha (HNF-4alpha) are considered a rare cause of MODY with only 14 mutations reported to date. The description of the phenotype is limited to single families. We investigated the genetics and phenotype of HNF-4alpha mutations in a large European Caucasian collection.

METHODS

HNF-4alpha was sequenced in 48 MODY probands, selected for a phenotype of HNF-1alpha MODY but negative for HNF-1alpha mutations. Clinical characteristics and biochemistry were compared between 54 HNF-4alpha mutation carriers and 32 familial controls from ten newly detected or previously described families.

RESULTS

Mutations in HNF-4alpha were found in 14/48 (29%) probands negative for HNF-1alpha mutations. The mutations found included seven novel mutations: S34X, D206Y, E276D, L332P, I314F, L332insCTG and IVS5nt+1G>A. I314F is the first reported de novo HNF-4alpha mutation. The average age of diagnosis was 22.9 years with frequent clinical evidence of sensitivity to sulphonylureas. Beta cell function, but not insulin sensitivity, was reduced in diabetic mutation carriers compared to control subjects (homeostasis model assessment of beta cell function 29% p<0.001 vs controls). HNF-4alpha mutations were associated with lower apolipoprotein A2 (p=0.001), A1 (p=0.04) and total HDL-cholesterol (p=0.02) than in control subjects. However, in contrast to some previous reports, levels of triglycerides and apolipoprotein C3 were normal.

CONCLUSIONS/INTERPRETATION

HNF-4alpha mutations are common when no HNF-1alpha mutation is found in strictly defined MODY families. The HNF-4alpha clinical phenotype and beta cell dysfunction are similar to HNF-1alpha MODY and are associated with reduced apolipoprotein A2 levels. We suggest that sequencing of HNF-4alpha should be performed in patients with clinical characteristics of HNF-1alpha MODY in whom mutations in HNF-1alpha are not found.

Ezetimibe-induced hyperlipidaemia

Ezetimibe is intestinally active cholesterol absorption inhibitor used to reduce low-density lipoprotein-cholesterol levels. This case report describes a novel side effect with this agent: ezetimibe-induced hyperlipidaemia in a patient with statin intolerance and familial combined hyperlipidaemia. Ezetimibe therapy induced an asymptomatic 770% increase in triglycerides (TGs) (3.51-27.1 mmol/l) and a 190% increase in total cholesterol (9.8-18.5 mmol/ 1) secondary to an increase (4.6-25.9 micromol/l; 560%) in hepatic cholesterol (lathosterol) synthesis. This lipid profile resolved 9 months after cessation of ezetimibe therapy. This report shows that ezetimibe may have long-lasting effects in man far exceeding its plasma half-life and that ezetimibe monotherapy can induce a large increase in hepatocyte very-low-density lipoprotein synthesis in rare individuals with a consequent mixed hyperlipidaemia or possibly hypercholesterolaemia depending on the metabolism and clearance of TG-rich lipoproteins.

Effect of moxonidine on lipid subfractions in patients with hypertension

Moxonidine is centrally acting imidazoline type-1 receptor agonist that significantly lowers blood pressure and has some insulin-sensitising actions. Its effects on plasma lipid profiles are uncertain. This study examined the effects of moxonidine on detailed lipid and lipoprotein profiles in 12 patients with hypertension and type 2b Fredricksen hyperlipidaemia. Treatment with moxonidine in six patients who completed the study resulted in a 10/5 mmHg reduction in 24-h ambulatory blood pressure (p = 0.01). A significant reduction in total and low-density-lipoprotein cholesterol (LDL-C) of 10% (p = 0.04) and 18% (p = 0.03), respectively, was seen. Triglycerides were reduced non-significantly by 23%, and high-density-lipoprotein cholesterol (HDL-C) was increased by 16%. There were no significant changes in apolipoprotein (apo) A-1 and B concentrations. No significant shifts were seen in HDL-C, LDL-C, very-low-density-lipoprotein cholesterol (VLDL-C) or apolipoprotein peak positions with therapy. Analysis of area under curve for each subfraction showed that moxonidine therapy resulted in a redistribution within the apoB profile. A slight non-significant reduction in VLDL apoB was seen. There was a reduction in the dense LDL apoB peak (p = 0.02) but less in the buoyant LDL apoB peak (p = 0.17) with a countervailing increase in LDL-C in the buoyant fraction (p = 0.01). The HDL-C and apoA-1 profile showed a shift from dense HDL apoA-1 (p = 0.01) to a buoyant HDL apoA-1sub-species (p = 0.01). These changes are consistent with a tendency for moxonidine to improve atherogenic lipid and lipoprotein profiles by actions on insulin-sensitisation and possibly through a direct cholesterol-reducing effect as seen with other imidazoles.

Effects of growth hormone (GH) replacement therapy on low-density lipoprotein apolipoprotein B100 kinetics in adult patients with GH deficiency: a stable isotope study

GH replacement therapy has been shown to improve the dyslipidemic condition in a substantial proportion of patients with adult GH deficiency. The mechanisms are not yet fully elucidated. Low-density lipoprotein (LDL) apolipoprotein B100 (apoB) formation and catabolism are important determinants of plasma cholesterol concentrations. This study examined the effect of GH replacement therapy on LDL apoB metabolism using a stable isotope turnover technique. LDL apoB kinetics was determined in 13 adult patients with GH deficiency before and after 3 months GH/placebo treatment in a randomized, double-blind, placebo-controlled study. LDL apoB (13)C-leucine enrichment was determined by isotope-ratio mass spectrometry. Plasma volume was assessed by standardized radionuclide dilution technique. GH replacement therapy significantly decreased LDL cholesterol, LDL apoB concentrations, and LDL apoB pool size compared with placebo. Compared with baseline, GH replacement therapy resulted in a significant increase in plasma volume and fractional catabolic rate, whereas LDL formation rate remained unchanged. LDL lipid content did not significantly change after GH and placebo. This study suggests that short-term GH replacement therapy decreases the LDL apoB pool by increasing removal of LDL particles without changing LDL composition or LDL apoB production rate. In addition, it is possible that the beneficial effects of GH on the cardiovascular system contribute to these findings.

Comparison of therapy with simvastatin 80 mg and 120 mg in patients with familial hypercholesterolaemia

High-dose preparations of simvastatin and atorvastatin have recently become available to treat resistant hypercholesterolaemia aggressively, but few studies have compared these two agents. This study compared the efficacy of simvastatin 80 mg and 120 mg in 22 patients with severe familial hypercholesterolaemia over a three-month period using an open label format. Simvastatin 120 mg was reasonably well tolerated and delivered a further 8% reduction in LDL over 80 mg, giving a total reduction of 55 +/- 13%, while further decreasing triglycerides (18%) and continuing to raise HDL (13%) further than the 80 mg dose. However, transient adverse changes were noted in both lipoprotein (a) and fibrinogen and 20% of patients were unable to tolerate the higher dose. One late case of rhabdomyolysis was observed, suggesting patients on the 120 mg dose require continued regular review.

Normal VLDL metabolism despite altered lipoprotein composition in type 1 diabetes mellitus

OBJECTIVES

Patients with type 1 diabetes are at increased risk of cardiovascular disease, which may be related to abnormal lipid metabolism. Secretion and clearance of VLDL apolipoprotein B100 (apoB) are important determinants of plasma lipid concentrations and are known to be influenced by hormones, including insulin and growth hormone.

PATIENTS

This study examined overnight VLDL apoB metabolism and VLDL composition in six lean patients with type 1 diabetes during euglycaemia (controlled by a varying insulin infusion) and in six age-, sex- and BMI-matched control subjects.

METHODS

VLDL apoB kinetics were determined using a primed constant 1-13C leucine infusion, and VLDL apoB enrichment was measured by gas-chromatography mass-spectrometry. Fasting lipid profile, IGF-I, IGFBP-3, overnight GH profiles and free insulin concentrations were also assessed.

RESULTS

Fasting concentrations of triglycerides (TG), total cholesterol (TC), HDL-cholesterol (HDL-C) and LDL-cholesterol (LDL-C) were similar in both groups. The VLDL apoB secretion and metabolic clearance rates were not significantly different between the two groups, but the VLDL-TGNLDL apoB and the VLDL-CNLDL apoB ratios were significantly increased in those with diabetes (P < 0.02 and P < 0.03, respectively). Total IGF-I concentrations were similar between the two groups; however, the GH area under the curve and free insulin concentrations were increased in patients with type 1 diabetes (GH: diabetes: 94.8 +/- 15.1 vs. controls: 45.6 +/- 10-6, mU/L/h, P < 0.04; free insulin: diabetes: 78.4 +/- 5.0 vs. controls: 28.3 +/- 3.26, pmol/l, P < 0.001). IGFBP-3 concentrations were lower in diabetic patients (diabetes: 2,454.2 +/- 68.7 vs. controls: 3,219.4 +/- 76.4, ng/ml, P < 0.001). In the control group overnight GH secretion correlated negatively with fasting TC (P < 0.01) and LDL-C (P < 0.03) concentrations, whereas free insulin concentrations correlated positively with fasting TG concentrations (P < 0.009). No significant correlations were found in the patients with diabetes.

CONCLUSION

This study suggests that in euglycaemic conditions patients with type 1 diabetes mellitus have normal VLDL apoB kinetics but altered VLDL composition. The altered VLDL composition may be associated with accelerated atherogenesis. We speculate that the disrupted hormonal balance and, in particular, the increased GH secretion might be responsible for the compositional changes of VLDL particles in type 1 diabetes mellitus.

Influence of sialic acid content of low-density lipoprotein particles on endothelium-dependent vasorelaxation in rabbit aorta

Low-density lipoprotein (LDL) concentration in plasma is an important predictor for atherosclerosis, and desialylated LDL has been proposed to be particularly atherogenic. Atherosclerosis is also associated with vascular endothelial dysfunction. We therefore wished to test the hypothesis that removal of sialic acid residues from LDL increases its ability to inhibit endothelium-dependent vasorelaxation. We studied vasorelaxant responses to acetylcholine (ACh) in isolated rabbit aortic rings as a model of endothelium-dependent relaxation, in the presence or absence of LDL treated either with saline or with neuraminidase, to cleave sialic acid residues. Vasorelaxant responses to ACh were inhibited by 300 microg protein per ml saline-treated LDL (E(max) 77.5+/-4.5 vs. 89.7+/-2.2% in the absence of LDL, P<0.05). This inhibitory effect was not altered by neuraminidase treatment of LDL. In contrast, 300 microg protein per ml LDL, either saline- or neuraminidase-treated, did not affect vasorelaxant responses to the endothelium-independent dilator sodium nitroprusside. We found no correlation between sialic acid content of saline-treated LDL and its ability to inhibit endothelium-dependent vasorelaxation, in rabbit aortic rings, at a concentration of 300 microg protein per ml. Our results therefore suggest that sialic acid content is not an important determinant of the effect of LDL on vascular endothelium-dependent relaxation.

Relationship between plasma sialic acid concentration and microvascular and macrovascular complications in type 1 diabetes: the EURODIAB Complications Study

OBJECTIVE

To test the hypothesis that an increased plasma concentration of sialic acid, a marker of the acute-phase response, is related to the presence of diabetic micro- and macrovascular complications in type 1 diabetes.

RESEARCH DESIGN AND METHODS

We investigated the relationship between plasma sialic acid concentration and nephropathy, retinopathy, neuropathy, and coronary heart disease (CHD) in a cross-sectional survey of 1,369 people with type 1 diabetes. Subjects were participants in the EURODIAB IDDM Complications Study, which involved 31 centers in 16 European countries.

RESULTS

There was a significantly increasing trend of plasma sialic acid with severity of retinopathy (P < 0.001 in men) and with degree of urinary albumin excretion (P < 0.001 men, P < 0.01 women). Plasma sialic acid correlated with increasing plasma creatinine concentration (P < 0.009 men, P < 0.0002 women), and men with neuropathy had a higher plasma sialic acid concentration than those without (P < 0.006). There was no significant correlation between plasma sialic acid and CHD in either sex. Elevated plasma sialic acid concentrations were also associated with several risk factors for diabetic vascular disease: diabetes duration, HbA1c, plasma triglyceride and cholesterol concentrations, waist-to-hip ratio, hypertension and smoking (in men), and low physical exercise (in women). In multiple logistic regression analysis, plasma sialic acid was independently related to proliferative retinopathy and urinary albumin excretion rate in men.

CONCLUSIONS

We conclude that an elevated plasma sialic concentration is strongly related to the presence of microvascular complications in type 1 diabetes, especially retinopathy and nephropathy. Further study of acute-phase response markers and mediators as indicators or predictors of diabetic microvascular complications is therefore justified.

Fibrinogen response with simvastatin versus atorvastatin in familial hypercholesterolemia

The clinical and biochemical determinants of the fibrinogen response to simvastatin or atorvastatin therapy were assessed in 130 patients with severe polygenic or familial hypercholesterolemia treated in a randomized open-trial format design. Hyperfibrinogenemia was associated with atorvastatin, baseline fibrinogen, and initial concentration and change in concentration of apolipoprotein B or low-density lipoprotein cholesterol.

Relation between sodium-lithium countertransport and hypertriglyceridemia in type V hyperlipidemia

Sodium-lithium countertransport (SLC) kinetics were measured in 30 patients with type V hyperlipidemia, 30 patients with type IIB hyperlipidemia on similar treatment, and 30 age- and sex-matched healthy controls. Clinical and laboratory data including basic anthropometry and blood pressure were obtained and blood was taken for detailed lipid biochemistry, glucose, insulin, and leptin measurements. Patients with type V hyperlipidemia were normotensive but more obese than controls, had elevated triglycerides, very low-density lipoprotein, glucose, and insulin; and reduced HDL cholesterol compared with type IIb controls. The median SLC activity (0.23 v 0.21 mmol Li+/L RBC/h) and median maximal velocity (0.33 v 0.31 mmol Li+/L RBC/h) were increased, but not significantly, compared to controls. In patients with type V hyperlipidemia SLC maximal velocity correlated with log triglycerides (r2 = 0.853; P < .001) and log very low-density lipoprotein (VLDL) triglycerides (r2 = 0.947; P < .001). Sodium-lithium countertransport maximal velocity correlated weakly with the homeostasis model assessment index of insulin resistance (r2 = 0.224; P = .06). The sodium affinity of the transporter did not differ between the groups and was independent of any of clinical or biochemical parameter studied. We conclude that VLDL triglyceride is strongly correlated with SLC maximal velocity and activity in patients with type V hyperlipidemia.

Differing relationships of methylene tetrahydrofolate reductase genotypes with cardiovascular risk in familial and polygenic hypercholesterolaemia

BACKGROUND

Plasma homocysteine and the methylene tetrahydrofolate reductase C677T polymorphism have been suggested as being risk factors for cardiovascular disease.

OBJECTIVE

To determine whether plasma homocysteine and the methylene tetrahydrofolate reductase C677T polymorphism are risk factors for coronary heart disease in patients with heterozygous familial hypercholesterolaemia as compared with those with polygenic hyperlipidaemia.

METHODS

Plasma homocysteine and the methylene tetrahydrofolate reductase polymorphism were assessed with other risk factors in 112 patients with familial hypercholesterolaemia and 72 patients with polygenic hyperlipidaemia, of whom 29 (25.8%) and 30 (41%) respectively had established cardiovascular disease and in 100 healthy normal subjects.

RESULTS

Plasma homocysteine was not significantly elevated in patients with and without coronary heart disease with familial hypercholesterolaemia or polygenic hyperlipidaemia compared with controls. The allele frequencies for C677T were significantly different in patients with coronary heart disease and with polygenic hyperlipidaemia (0.35 versus 0.29) (P = 0.02) as opposed to those with coronary heart disease and familial hypercholesterolaemia (0.25 versus 0.30) (P = 0.63). Methylene tetrahydrofolate reductase genotype but not homocysteine had a weak association with coronary heart disease in logistic regression analysis in patients with polygenic hyperlipidaemia (P = 0.05) but neither methylene tetrahydrofolate reductase genotype or plasma homocysteine was a risk factor in patients with familial hypercholesterolaemia.

CONCLUSION

Whilst methylene tetrahydrofolate reductase genotype may be a weak risk factor for coronary heart disease in polygenic hyperlipidaemia as opposed to familial hypercholesterolaemia, homocysteine does not seem to be an important risk factor for coronary heart disease in patients in southern UK.

Renin-angiotensin system polymorphisms and coronary events in familial hypercholesterolemia

The role of renin-angiotensin system polymorphisms as risk factors for coronary heart disease (CHD) is controversial. This study investigated their role in patients with heterozygous familial hypercholesterolemia (FH). Polymorphism frequencies for angiotensin-I-converting enzyme insertion/deletion (ACE I/D), angiotensinogen M235T, and angiotensin-II type I receptor (AG2R) A1166C were determined in 112 patients with FH and 72 patients with polygenic hypercholesterolemia, of whom 26.7% and 41.6%, respectively, had established CHD. None of the polymorphisms were associated with risk of CHD in patients with polygenic hypercholesterolemia in this study. Logistic regression analysis of risk factors for CHD in patients with FH identified male sex (odds ratio [OR]=3.03; 95% CI, 3.07 to 3.72; P=0.05), smoking (OR=2.91; 95% CI, 2.16 to 4.24; P=0.05), diastolic blood pressure (OR=3.70; 95% CI, 3.43 to 3.97; P=0.02), plasma glucose (OR=3.31; 95% CI, 3. 10 to 3.52; P=0.04), and the AG2R A1166C polymorphism as risk factors. The OR for the AG2R A1166C polymorphism was 2.26 (95% CI, 1.26 to 3.72; P=0.06) and increased to 3.10 (95% CI, 1.20 to 7.52; P=0.04) after adjustment for other risk factors. The AG2R A1166C polymorphism may interact with severe hypercholesterolemia and other risk factors to increase risk of CHD in FH patients.

Effects of lipids in patients with familial hypercholesterolaemia on the kinetics of the sodium-lithium countertransporter

Sodium-lithium countertransport kinetics were measured in 87 patients (50 male; 37 female) with heterozygous familial hypercholesterolaemia (FH) and a group of 38 age range and sex-distribution matched controls. Basic clinical data including basic anthropometry, blood pressure were obtained and blood was taken for detailed lipid biochemistry, glucose and insulin measurement. Patients with FH had elevated total cholesterol, low-density lipoprotein (LDL)-cholesterol and apolipoprotein B concentrations compared to controls. The activity and log transformed maximal velocity (Vmax) of the sodium-lithium countertransporter unlike the affinity (Km) were reduced in patients with FH compared to controls (geometric means 0.172 vs 0.217 mmol Li+/L.RBC.hr; P = 0.02; 0.237 vs. 0.317 mmol Li+/L.RBC.hr; P = 0.009 respectively). In multiple regression analysis, log normalised SLC activity correlated weakly with log triglyceride (beta = 0.225; P = 0.06) and cholesterol (beta = -0.112 P = 0.06). Log Vmax correlated with log triglyceride (beta = 0.307; P = 0.02), and high-density lipoprotein (HDL) (beta = 0.74; P = 0.03) whilst Km correlated with HDL (beta = 1.73; P<0.001) and apoAI (beta = -1.76; P = 0.0048), LDL (beta = -0.14; P = 0.05), and creatine kinase (beta = 0.003; P = 0.01). Cholesterol and triglyceride concentrations rather than insulin resistance seem to be the key features affecting the environmental alteration of sodium lithium countertransporter Vmax in patients with familial hypercholesterolaemia.

High-density lipoprotein cholesterol and triglyceride response with simvastatin versus atorvastatin in familial hypercholesterolemia

The clinical and biochemical determinants of high-density lipoprotein (HDL) and triglyceride response to simvastatin and atorvastatin were assessed in 150 patients with severe hyperlipidemia treated in a randomized open-trial format design. Triglyceride reduction was only dependent on HDL:apolipoprotein A1, change in apolipoprotein B, and dose response, whereas an increase in HDL was dependent on initial LDL, change in LDL or dose response, and therapy with simvastatin.

A randomised placebo controlled trial of the effects of tibolone on blood pressure and lipids in hypertensive women

The effects of hormone replacement therapy in hypertensive women are controversial. This randomised placebo controlled trial assessed the effect of tibolone 2.5 mg on blood pressure and fasting plasma lipids in 29 hypertensive postmenopausal women over 6 months using a 2:1 randomisation to tibolone. The primary clinical end-point was mean office blood pressure. At 6 months systolic blood pressure declined by 5.30 +/- 2.87% vs 4.94 +/- 3.37% whilst diastolic blood pressure declined 5.38 +/- 2.65% vs 0.85 +/- 3.69% on tibolone and placebo respectively. These differences were not statistically significant. Triglycerides decreased by 33.3 +/- 6.1% vs 7.6 +/- 7.9% (P < 0.01) and high-density lipoprotein (HDL)-cholesterol by 21.7 +/- 3.8% vs 2.4 +/- 2.6% (P < 0.01) with tibolone as opposed to placebo. No significant differences were observed in total cholesterol, low-density lipoprotein (LDL)-cholesterol and lipoprotein (a). Fibrinogen levels were reduced by 13.6 +/- 6.8% on tibolone compared to a 19.3 +/- 15.4% rise (P < 0.05) on placebo. This study suggests that tibolone has no deleterious effect on blood pressure in women with hypertension but has contrasting effects on biochemical risk factors. Large-scale studies are required to determine the overall effect of tibolone on cardiovascular morbidity and mortality.

Comparison of therapy with simvastatin 80 mg and atorvastatin 80 mg in patients with familial hypercholesterolaemia

This study compared the efficacy of simvastatin 80 mg and atorvastatin 80 mg in the treatment of 26 patients with familial hypercholesterolaemia over 12 weeks using an open crossover trial format. Both, similarly, reduced LDL by 47 +/- 13% and 43 +/- 16% and median triglycerides by 22% and 27% respectively. However, atorvastatin reduced HDL by 2 +/- 24% compared with 8 +/- 30% increase with simvastatin (p = 0.05) affecting the LDL:HDL ratio achieved (4.478 +/- 1.56 vs 3.74 +/- 0.93, p = 0.001). Atorvastatin raised median fibrinogen by 15% compared with a non-significant 5% increase with simvastatin (p = 0.05). Simvastatin reduced lipoprotein (a) by a median 20% compared with baseline (p = 0.05) compared with 5% for atorvastatin. Side-effects, mostly gastrointestinal, were seen in four patients (16%) with atorvastatin compared with one case of myalgia with simvastatin (4%). We conclude both drugs are equally effective in LDL reduction but that simvastatin is superior in raising HDL and causes fewer side-effects. These results require confirmation in larger studies.

Growth hormone (GH) replacement therapy reduces serum sialic acid concentrations in adults with GH-deficiency: a double-blind placebo-controlled study

OBJECTIVES

Patients with adult GH-deficiency are thought to have an increased risk of cardiovascular disease. Sialic acid (SA) concentrations have been proposed as a marker of atherosclerotic disease probably related to an inflammatory response of the arterial wall. SA as a marker of cardiovascular disease in adult GH-deficiency and its relation to changes in fasting lipid profile and hormone concentrations have not yet been investigated.

PATIENTS

We performed a randomised, double-blind placebo-controlled study in 18 patients with adult GH-deficiency before and after 3 months GH replacement therapy (0.036 U/kg/d; GH-treated group: 6 females, 3 males; age: 47.3 +/- 5.4 years., mean +/- SEM; placebo-group: 5 females, 4 males; mean age 50.2 +/- 4.7). In addition, SA concentrations were measured in 18 sex and age matched healthy control subjects.

METHODS

Blood samples were obtained after an overnight fast. Serum SA, triglycerides and cholesterol were measured using enzymatic methods. Lipoprotein classes were separated by ultracentrifugation. Insulin and IGF-I were determined by radioimmunoassay, HbA1C was measured by anion exchange liquid chromatography.

RESULTS

SA concentrations of the patients with adult GH-deficiency were not significantly different compared to the control group (GH-deficient group: 2.29 +/- 0.02 mmol/l, mean +/- SEM vs. control group: 2.09 +/- 0.13 mmol/l, P = 0.25). Before GH replacement therapy SA concentrations correlated positively with the patients age (r = 0.45; P < 0.04) and fasting insulin concentrations (r = 0.5; P < 0.03) but not with fasting lipid profile. GH replacement therapy significantly increased IGF-I (GH: + 27 +/- 2.6 vs. placebo: + 1.0 +/- 0.8 nmol/l, P < 0.001) and fasting insulin concentrations (GH: + 71.9 +/- 8.0 vs. placebo: + 19.6 +/- 22.6 pmol/l, P < 0.04) compared to placebo therapy. SA concentrations (GH: - 0.41 +/- 0.15 vs. placebo: - 0.01 +/- 0.12 mmol/l, P < 0.05), total cholesterol (GH: - 0.71 +/- 0.16 vs. placebo: 0.23 +/- 0.21 mmol/l, P < 0.003) and LDL-cholesterol (- 0.71 +/- 0.14 vs. placebo: - 0.12 +/- 0.21 mmol/l P < 0.04) significantly decreased after GH replacement therapy compared to placebo therapy. No significant correlation between changes in SA concentrations and changes in lipid profile were observed following GH replacement therapy.

CONCLUSION

These results suggest that, firstly, GH replacement therapy may have a beneficial effect on the pathogenesis of atherosclerosis despite the increase in insulin concentrations, a surrogate marker of insulin resistance, secondly, the proposed beneficial effect of GH on the atherosclerotic process is likely to be multifactorial and cannot only be explained by changes in lipid profile and finally, SA might be a useful marker for the process of atherosclerotic disease in interventional studies.

Atorvastatin compared with simvastatin-based therapies in the management of severe familial hyperlipidaemias

We compared atorvastatin with simvastatin-based therapies in a prospective observational study of 201 patients with severe hyperlipidaemia. Atorvastatin 10 mg therapy was substituted for simvastatin 20 mg, 20 mg for 40 mg, 40 mg for simvastatin 40 mg plus resin, and 80 mg for simvastatin-fibrate-resin therapy. Lipid and safety profiles were assessed. Atorvastatin reduced total cholesterol by 31 +/- 11-40 +/- 14% vs. 25 +/- 12-31 +/- 11%; LDL by 38 +/- 16-45 +/- 18% vs. 31 +/- 18-39 +/- 18% and geometric mean triglycerides by 29.3-37.3% vs. 16.6-24.8%, but reduced HDL 11% +/- 47% at 80 mg compared with a 16% +/- 34% increase with simvastatin-based therapy. Target LDL < 3.5 mmol/l was achieved more often with atorvastatin (63% vs. 50%; p < 0.001). Atorvastatin increased geometric mean fibrinogen by 12-20% vs. a 0-6% fall with simvastatin (p << 0.001). Side effects were noted in 10-36% of patients, including one case of rhabdomyolysis, and 36% discontinued therapy. These data suggest that atorvastatin is more effective than current simvastatin-based therapies in achieving treatment targets in patients with familial hypercholesterolaemia but at the expense of a possible increase in side-effects. This issue needs further study in randomized controlled trials.

Transport of phytanic acid on lipoproteins in Refsum disease

Patients with Refsum disease accumulate significant quantities of phytanic acid in adipose and neural tissue. The accumulation can be reversed by following a diet low in phytanic acid, yet the mechanism of transport of this fatty acid is obscure. We investigated the distribution of phytanic acid in different lipoprotein subfractions in 11 patients with Refsum disease and 9 unaffected siblings. Plasma phytanic acid was distributed on VLDL (16.2% +/- 12.2%), IDL (1.77% +/- 1.64%), LDL (34.8% +/- 12.6%) and HDL (14.3% +/- 7.87%). No correlations with any parameter were seen with total phytanic acid content. Weak nonsignificant correlations were found with the fractional distribution of phytanic acid and VLDL triglyceride (r = 0.35; p = 0.12) and plasma HDL-cholesterol (r = 0.32; p = 0.16) and with LDL:HDL cholesterol ratio (r = 0.33; p = 0.14). Significant correlation of the fractional distribution of phytanic acid on lipoprotein particles was noted with the ratio of apolipoprotein B: apolipoprotein A1-containing particles (r = 0.46; p = 0.03) and apolipoprotein B: apolipoprotein A1 in HDL2 (r = 0.53; p = 0.01). This suggests that the import-export balance for phytanic acid in plasma is related to forward and reverse cholesterol transport on lipoprotein particles, and only weakly to plasma cholesterol and triglycerides. These ratios of apolipoprotein particles may play a significant role in determining the rate of phytanic acid elimination in patients with Refsum disease.

Effects of growth hormone (GH) replacement therapy on very low density lipoprotein apolipoprotein B100 kinetics in patients with adult GH deficiency: a stable isotope study

Patients with adult GH deficiency are often dyslipidemic and may have an increased risk of cardiovascular disease. The secretion and clearance of very low density lipoprotein apolipoprotein B 100 (VLDL apoB) are important determinants of plasma lipid concentrations. This study examined the effect of GH replacement therapy on VLDL apoB metabolism using a stable isotope turnover technique. VLDL apoB kinetics were determined in 14 adult patients with GH deficiency before and after 3 months GH or placebo treatment in a randomized double blind, placebo-controlled study using a primed constant [1-(13)C]leucine infusion. VLDL apoB enrichment was determined by gas chromatography-mass spectrometry. GH replacement therapy increased plasma insulin-like growth factor I concentrations 2.9 +/- 0.5-fold (P < 0.001), fasting insulin concentrations 1.8 +/- 0.6-fold (P < 0.04), and hemoglobin A1C from 5.0 +/- 0.2% to 5.3 +/- 0.2% (mean +/- SEM; P < 0.001). It decreased fat mass by 3.4 +/- 1.3 kg (P < 0.05) and increased lean body mass by 3.5 +/- 0.8 kg (P < 0.01). The total cholesterol concentration (P < 0.02), the low density lipoprotein cholesterol concentration (P < 0.02), and the VLDL cholesterol/VLDL apoB ratio (P < 0.005) decreased. GH therapy did not significantly change the VLDL apoB pool size, but increased the VLDL apoB secretion rate from 9.2 +/- 2.0 to 25.9 +/- 10.3 mg/kg x day (P < 0.01) and the MCR from 11.5 +/- 2.7 to 20.3 +/- 3.2 mL/min (P < 0.03). No significant changes were observed in the placebo group. This study suggests that GH replacement therapy improves lipid profile by increasing the removal of VLDL apoB. Although GH therapy stimulates VLDL apoB secretion, this is offset by the increase in the VLDL apoB clearance rate, which we postulate is due to its effects in up-regulating low density lipoprotein receptors and modifying VLDL composition.

High-dose atorvastatin therapy in severe heterozygous familial hypercholesterolaemia

Lipid targets can be difficult to attain in familial hypercholesterolaemia. To compare atorvastatin with simvastatin-fenofibrate and simvastatin-cholestyramine therapy, we studied 54 patients with familial hypercholesterolaemia over periods of 2-6 months on each therapeutic regimen. The atorvastatin regimen reduced total cholesterol by 41.2 +/- 11.2%, LDL by 45.6 +/- 15.5%, triglycerides by 33.8 +/- 24.8%, and increased HDL by 2.3 +/- 37.0%. Simvastatin-fenofibrate therapy achieved reductions of 33.9 +/- 8.5% in cholesterol, 42.0 +/- 12.2% in LDL, 34.7 +/- 38.3% for triglycerides, and a 25.4 +/- 55.1% increase in HDL. Simvastatin-cholestyramine gave a reduction of 31.3 +/- 11.8% in cholesterol, 36.0 +/- 14.4% in LDL, 13.7 +/- 36.3% in triglycerides, and a 1.1 +/- 30.3% rise in HDL. The atorvastatin regimen was marginally but not significantly better than simvastatin-fenofibrate in improving the LDL:HDL ratio, LDL:apoB and and apolipoprotein B:A1 ratios. Eleven patients (20.4%) had side-effects: two discontinued atorvastatin due to side-effects; two patients had rashes; six had myalgia and two had diarrhoea. Gastrointestinal side-effects were described in 16 (30.1%) patients on simvastatin-cholestyramine therapy and four cases of myalgia (11.2%) were seen with simvastatin-fenofibrate. In nine patients on atorvastatin (20.4%) a 30% or greater fall in HDL was observed, compared to five patients with resin therapy (9.2%) and two with fibrate therapy (5.5%). There were no significant differences in liver or muscle biochemistry between the regimens, but atorvastatin did raise transaminase and creatine kinase concentrations significantly compared to pre-treatment values (p = 0.001). Atorvastatin significantly improves the lipid profile in most patients compared with other regimens. It has a comparable incidence of side-effects to combination therapy regimens.

Fenofibrate plus simvastatin therapy versus simvastatin plus cholestyramine therapy for familial hypercholesterolaemia

Combination therapy is routinely used to achieve improved cholesterol reduction in familial hypercholesterolaemia. We compared the standard simvastatin plus bile-acid sequestrant (cholestyramine) therapy with simvastatin plus fenofibrate in 29 patients with severe familial hypercholesterolaemia. The fibrate regimen resulted in an 35.1 +/- 10.7% reduction in total cholesterol, a 40.6 +/- 20.5% in LDL cholesterol, 17.2 +/- 56.5% reduction in triglycerides and a 20.3 +/- 52.0% increase in HDL cholesterol. The cholestyramine regimen produced reductions of 29.3 +/- 13.2% in cholesterol, 37.1 +/- 21.9% in LDL cholesterol, and 12.5 +/- 48.9% in triglycerides, and a 5.0 +/- 25.4% rise in HDL cholesterol. The fibrate regimen was significantly more effective in reducing total cholesterol (p < 0.001) and LDL-cholesterol (p = 0.004), and also reduced triglycerides significantly (p = 0.05), compared to the cholestyramine regimen. There were significant improvements in the LDL:HDL cholesterol ratio (3.62 +/- 1.54 vs. 4.00 +/- 1.36; p = 0.05) and in the apolipoprotein B:A1 ratio (1.13 +/- 0.036 vs. 1.20 +/- 0.34; p = 0.05). Gastrointestinal side-effects occurred in 10 patients on cholestyramine therapy, and four patients on fibrate therapy had myalgia. There were no cases of rhabdomyolysis with either regime. No significant differences in liver biochemistry or creatine kinase were seen with either regimen.

Abnormalities of very low density lipoprotein apolipoprotein B-100 metabolism contribute to the dyslipidaemia of adult growth hormone deficiency

Increased cardiovascular mortality in adult growth hormone deficiency (GHD) may be, in part, explained by the dyslipidaemia associated with this condition. It is possible that abnormalities of very low density lipoprotein apolipoprotein B-100 (VLDL apoB) metabolism contribute to this dyslipidaemia. To test this hypothesis, we measured VLDL apoB kinetics in adult GH deficient patients (4 females, 3 males; age 50.1 +/- 4.7 yr (mean +/- SEM); BMI 28.2 +/- 1.1 kg/m2; total cholesterol (TC) 6.6 +/- 0.3 mmol/l; triglyceride (TG) 2.8 +/- 0.6 mmol/l; HDL cholesterol 1.1 +/- 0.1 mmol/l) and in control subjects (4 females, 3 male; age 47.0 +/- 4.7 yr; BMI 27.0 +/- 2.6 kg/m2; TC 5.0 +/- 0.4 mmol/l; TG 0.9 +/- 0.2 mmol/l; HDL cholesterol 1.4 +/- 0.1 mmol/l). [1-(13)C] leucine was administered by a primed (1 mg/kg), constant intravenous infusion (1 mg/kg/hr) and VLDL apoB enrichment with 13C leucine was determined using gas-chromatography mass-spectrometry. The GHD patients had a significantly higher hepatic secretion rate of VLDL apoB (15.5 +/- 1.8 mg/kg/day vs 9.4 +/- 0.6 mg/kg/day p = 0.007) and reduced catabolism ofVLDL apoB (metabolic clearance rate; 12.3 +/- 1.7 ml/min vs 24.3 +/- 4.8 ml/min p < 0.05) compared with control subjects. These findings suggest that GH is integrally involved in the regulation of VLDL apoB metabolism.

Comparison of immunoturbidimetric and Lowry methods for measuring concentration of very low density lipoprotein apolipoprotein B-100 in plasma

To assess whether the Lowry-tetramethylurea method for measuring apolipoprotein B-100 (apo-B) in very low density lipoprotein (VLDL) could be replaced by direct assay of VLDL apo-B using a highly practicable immunological method. Seventy five fasting blood samples were collected from patients attending the lipid clinic at this hospital. Plasma was separated immediately and VLDL isolated by preparative ultracentrifugation at solution density 0.93-1.006 kg/l. Apo-B was precipitated from an aliquot of the VLDL fraction using the tetramethylurea (TMU) technique and protein mass determined by the Lowry method (LM); mean apo-B 83.02 micrograms/ml (SD 74.85). Apo-B was also measured in VLDL using direct immunoturbidimetry on the Cobas-Fara analyser; mean apo-B 82.32 micrograms/ml (SD 72.88). There was a very close correlation between methods (immunoturbidimetry = 0.94.LM + 3.95, r = 0.97, p < 0.001). The mean difference between methods (constant error) was small (0.70 microgram/ml) and not significant (p = 0.742). Random error was 13.01 micrograms/ml by analysis of variance. It is concluded that immunoturbidimetry, a more rapid and convenient test, may replace the LM and TMU techniques for measuring VLDL apo-B concentration and that this method could be applied to research studies requiring analysis of large numbers of samples.

Automated spectrophotometric method for determining oxidized and reduced glutathione in liver

An enzymatic recycling method has been applied to the measurement of total and oxidized glutathione with a centrifugal analyzer. When the reduced form of glutathione (GSH) was masked with 2-vinylpyridine to measure the oxidized glutathione (GSSG), the time to ensure full derivatization was three times longer than has been reported. The method is quick, simple, accurate, and precise (1.27% for GSH, 3.3% for GSSG intraassay CV; 2.15% for GSH, 5% for GSSG interassay CV), and the automation allows large numbers of samples to be conveniently assayed.

Automated spectrophotometric method for determining oxidized and reduced glutathione in liver

An enzymatic recycling method has been applied to the measurement of total and oxidized glutathione with a centrifugal analyzer. When the reduced form of glutathione (GSH) was masked with 2-vinylpyridine to measure the oxidized glutathione (GSSG), the time to ensure full derivatization was three times longer than has been reported. The method is quick, simple, accurate, and precise (1.27% for GSH, 3.3% for GSSG intraassay CV; 2.15% for GSH, 5% for GSSG interassay CV), and the automation allows large numbers of samples to be conveniently assayed.

Determination of serum cholesterol concentration in the presence of ascorbate

Pretreatment of a serum or plasma sample with ascorbate oxidase removed interfering ascorbate and allowed the determination of cholesterol to be carried out by a current enzymatic cholesterol method available in kit form. The Cobas-Fara was programmed to carry out pretreatment of the sample with ascorbate oxidase before addition of the cholesterol colour reagent.

Insulin therapy in type 2 diabetic subjects suppresses plasminogen activator inhibitor (PAI-1) activity and proinsulin-like molecules independently of glycaemic control

Eleven Type 2 diabetic subjects (10 male 1 female: age 56.2 +/- 9.7 (SD) yr) were treated in random order either with insulin or with sulphonylureas for 8 weeks each, without attempting to alter glycaemic control between the two treatment periods. Insulin treatment was associated with suppression of endogenous insulin secretion (fasting C-peptide levels -35.0 +/- 24.2%; p = 0.006), and of intact proinsulin (-43.1 +/- 36.8%; p = 0.03) and 32,33 split proinsulin -20.1 +/- 27.0%; p = 0.03). Activity of plasminogen activator inhibitor (PAI-1), a fast acting inhibitor of fibrinolysis, decreased significantly (-14.3% +/- 27.5%; p = 0.02) but no changes occurred in concentration of lipoproteins or apoproteins between therapies. Changes in concentrations of 32,33 split and intact proinsulin were closely and significantly related (rs = 0.83; p < 0.001) to each other but not with changes in concentrations of C-peptide (intact proinsulin rs = -0.41; p = 0.11) and 32,33 split proinsulin rs = -0.27; (p = 0.21). Percentage changes in intact proinsulin concentrations were positively correlated with those in PAI-1 (rs = 0.51; p = 0.05). There was, however a paradoxical negative relationship between changes in C-peptide concentrations and those of PAI-1 (rs = -0.73; p = 0.006). These preliminary observations suggest that insulin treatment in Type 2 diabetic subjects without any changes in glycaemic control is associated with a reduced activity of PAI-1, but is without effect on any other cardiovascular risk factors. Concentrations of insulin precursor molecules may play a role in determining fibrinolytic activity.