Polymorphisms of the Plasminogen Activator Inhibitor- 1 Gene in Type 1 and Type 2 Diabetes, and in Patients with Diabetic Retinopathy

To identify whether genotype contributes to the difference in PAI-1 levels in type 1 and type 2 diabetic subjects and whether genotype relates to the development of retinopathy, a Hind III restriction fragment length polymorphism and two dinucleotide repeat polymorphisms were studied. In 519 Caucasian diabetic subjects (192 type 1, 327 type 2) and 123 Caucasian control subjects there were no differences in the frequency of the Hind III restriction alleles (type 1 vs type 2 vs control: allele 1 0.397 vs 0.420 vs 0.448; allele 2 0.603 vs 0.580 vs 0.552) nor in the allelic frequency at either dinucleotide repeat sequence. In 86 subjects with no retinopathy at 15 years or more from diagnosis of diabetes and 190 subjects with diabetic retinopathy there was no difference in the frequency of Hind III restriction alleles (retinopathy present vs retinopathy absent: allele 1 0.400 vs 0.467; allele 2 0.600 vs 0.533) nor in the allelic frequencies at either dinucleotide repeat sequence. The results indicate that there is no or minimal influence of the PAI-1 gene on either PAI-1 levels or the development of diabetic retinopathy in patients with diabetes mellitus.

Hormonal Control of Haemostasis During Hypoglycaemia in Diabetes Mellitus

Factor VIII (FVIII) and plasminogen activator activity (PAA) rise during hypoglycaemia, and this might contribute to the vascular complications of diabetes. Similar changes in haemostasis accompany raised plasma levels of vasopressin (aVP) and adrenaline. To investigate the effects of these hormones on haemostasis during hypoglycaemia and the role of plasma insulin concentrations, eight insulin-dependent diabetic patients underwent controlled hypoglycaemia for 20 min and 13 diabetic patients were investigated during hyperinsulinaemia with blood glucose maintained at 8.0 mmol/1. During hypoglycaemia, insulin levels increased to median values of 114 mU/1, aVP rose from 0.5 to 4.4 (p <0.005) pg/ml and adrenaline from 0.4 to 4.4 nmol/l (p <0.005). FVIII coagulant activity (FVIII :C) rose from 0.75 to 1.09 iU/ml (p <0.01) and the ristocetin co-factor (FVIIIR:Co) and von Willebrand factor antigen (vWF:Ag) showed similar responses. PAA increased from 156 to 745 units (p <0.005). During hyperinsulinaemia, insulin rose following infusion from 24 to 52 and 118 mU/l, maintained for an hour at each level. Despite this, plasma aVP, FVIII :C, FVIIIR:Co, vWF:Ag and PAA remained unchanged. This study indicates that the marked changes in FVIII, vWF and PAA concentrations which accompany hypoglycaemia depend on low blood glucose and not raised plasma insulin. The response in probably mediated by increases in adrenaline and aVP, which are part of the physiological response to hypoglycaemia.

RAGE and arthritis: the G82S polymorphism amplifies the inflammatory response

The receptor for advanced glycation end products (RAGE) and its proinflammatory S100/calgranulin ligands are enriched in joints of subjects with rheumatoid arthritis (RA) and amplify the immune/inflammatory response. In a model of inflammatory arthritis, blockade of RAGE in mice immunized and challenged with bovine type II collagen suppressed clinical and histologic evidence of arthritis, in parallel with diminished levels of TNF-alpha, IL-6, and matrix metalloproteinases (MMP) 3, 9 and 13 in affected tissues. Allelic variation within key domains of RAGE may influence these proinflammatory mechanisms, thereby predisposing individuals to heightened inflammatory responses. A polymorphism of the RAGE gene within the ligand-binding domain of the receptor has been identified, consisting of a glycine to serine change at position 82. Cells bearing the RAGE 82S allele displayed enhanced binding and cytokine/MMP generation following ligation by a prototypic S100/calgranulin compared with cells expressing the RAGE 82G allele. In human subjects, a case-control study demonstrated an increased prevalence of the 82S allele in patients with RA compared with control subjects. These data suggest that RAGE 82S upregulates the inflammatory response upon engagement of S100/calgranulins, and, thereby, may contribute to enhanced proinflammatory mechanisms in immune/inflammatory diseases.

Characterization of allelic and nucleotide variation between the RAGE gene on chromosome 6 and a homologous pseudogene sequence to its 5' regulatory region on chromosome 3: implications for polymorphic studies in diabetes

Activation of the receptor for advanced glycation end products (RAGE) appears to be a key mechanism in the pathogenesis of diabetic vascular disease, making RAGE a candidate gene for investigation. RAGE is located in the major histocompatibility complex locus on chromosome 6, which contains a multitude of overlapping and duplicated genes involved predominantly in inflammatory and immune responses. The RAGE 5' flanking region from -505 in a 5' direction overlaps with PBX2, a gene that has a pseudogene copy on chromosome 3, making any studies of polymorphisms in this duplicated region potentially fraught with error. In this study we have addressed these issues by confirming RAGE as a predominantly single-copy gene and PBX2 to have two gene copies in the haploid human genome. We have characterized the gene:pseudogene differences between RAGE/PBX2 on chromosome 6 and PsiPBX2 on chromosome 3, which include a change from C to A at position -1139 RAGE/+2298 PBX2, previously reported as a polymorphism. Single chromosome-specific DNA amplification of the duplicated region has clarified five polymorphisms to be on chromosome 3 and one (at -1202 RAGE/+2234 PBX2) to be on chromosome 6. In conclusion, this study provides essential data for the study of RAGE and its genetics.

Effects of novel polymorphisms in the RAGE gene on transcriptional regulation and their association with diabetic retinopathy

Interactions between advanced glycation end products (AGEs) and the receptor for AGE (RAGE) are implicated in the vascular complications in diabetes. We have identified eight novel polymorphisms, of which the -1420 (GGT)n, -1393 G/T, -1390 G/T, and -1202 G/A were in the overlapping PBX2 3' untranslated region (UTR), and the -429 T/C (66.5% TT, 33.5% TC/CC), -407 to -345 deletion (99% I, 1% I/D, 0% D), -374 T/A (66.4% TT, 33.6% TA/AA), and +20 T/A were in the RAGE promoter. To evaluate the effects on transcriptional activity, we measured chloramphenicol acetyl transferase (CAT) reporter gene expression, driven by variants of the -738 to +49 RAGE gene fragment containing the four polymorphisms identified close to the transcriptional start site. The -429 C, -374 A, and 63-bp deletion alleles resulted in a mean increase of CAT expression of twofold (P < 0.0001), threefold (P < 0.001), and fourfold (P < 0.05), respectively, with the -374 T and A alleles yielding highly differential binding of nuclear protein extract from both monocyte- and hepatocyte-derived cell lines. The prevalence of the functional polymorphisms were investigated in subjects with type 2 diabetes (106 with and 109 without retinopathy), with the -429 C allele showing an increase in the retinopathy group (P < 0.05). These data suggest that the polymorphisms involved in differences in RAGE gene regulation may influence the pathogenesis of diabetic vascular complications.

alpha-fibrinogen Thr312Ala polymorphism and venous thromboembolism

The Aalpha-fibrinogen Thr312Ala polymorphism, which occurs in a region involved in factor XIII (FXIII)-dependent cross-linking processes, is associated with poststroke mortality in subjects with atrial fibrillation, suggesting an influence either on intraatrial clot formation or embolization. We have determined the association of Thr312Ala with deep vein thrombosis (DVT) and pulmonary embolism (PE) and have assessed the interaction of Thr312Ala with the FXIII Val34Leu polymorphism in 122 patients with DVT, 99 patients with PE, and 254 healthy control subjects. The genotype distribution of patients with PE (TT = 49%, TA = 36%, AA = 15%), but not DVT (TT = 50%, TA = 42%, AA = 8%), differed significantly from healthy control subjects (TT = 60%, TA = 34%, AA = 6%, P =.02). A significant interaction of Thr312Ala and Val34Leu was also identified (P =.01), indicating an inverse association between Leu34 and Ala312. These results support the hypothesis that Thr312Ala alters FXIII-dependent cross-linking, making formed fibrin clot more susceptible to embolization.

Association of a common polymorphism in the factor XIII gene with venous thrombosis

We have shown an association between a common mutation in the factor XIII a-subunit gene, coding for an amino acid change, 3 amino acids from the thrombin activation site (factor XIII Val34Leu) that may protect against myocardial infarction and predisposes to intracranial hemorrhage. To investigate the possible role of factor XIII Val34Leu in the pathogenesis of venous thromboembolism (VTE) and potential interactions with factor V Leiden (FV:Q506) and prothrombin G --> A 20210, we studied 221 patients with a history of VTE and 254 healthy controls. Patients with VTE showed an increased frequency of the FXIII Val/Val genotype (63% v 49%) and a lower frequency of the Val/Leu genotype (31% v 42%) than controls (P =. 007). FV:Q506 heterozygotes were more frequent in VTE patients (11%) than controls (5%; P =.04). The prothrombin G --> A 20210 mutation was present in only 3 patients and no controls (P =.10). In a logistic regression model for a history of VTE, the odds ratio (95% confidence interval) for FXIII Val/Leu or Leu/Leu genotype was 0.63 (0.38 to 0.82) and for possession of FV:Q506 2.40 (1.17 to 4.90). There was no evidence for an interaction between factor XIII Val34Leu genotype and FV:Q506, prothrombin G --> A 20210, sex, or age. It is concluded that possession of the Leu allele at factor XIII Val34Leu is protective against deep venous thrombosis.

The effect of number of days in culture and plasminogen activator inhibitor-1 (PAI-1) 4G/5G genotype on PAI-1 antigen release by cultured human umbilical vein endothelial cells

An insertion/deletion (4G/5G) polymorphism in what has been shown to be an enhancer/repressor binding site in the promoter region of the PAI-1 gene has been related to plasma PAI-1 activity. Transfection studies demonstrated increased interleukin-1 stimulated PAI-1 synthesis in cells containing the 4G sequence. To study this response in endothelial cells, first passage HUVEC from 26 umbilical cords were stimulated with interleukin-1 and tumor necrosis factor-alpha. PAI-1 antigen was measured in 24-hour conditioned medium and allele-specific PCR utilized to determine genotype at the 4G/5G locus. Analysis of covariance was used to determine whether the effect of a variable time in culture was masking a difference between genotypes. A trend towards higher PAI-1 levels with increasing time in culture was observed. The geometric mean (95% confidence interval) of the basal rate of PAI-1 release was, 4G/4G 9.7 (7.0, 13.5) ng/24 hours (n=11), 4G/5G 9.5 (6.5, 13.9) ng/24 hours (n=9), and 5G/5G 10.9 (7.8, 15.1) ng/24 hours (n=6). In cells of the same cultures, the interleukin-1 stimulated levels were 25.9 (23.1, 29.1), 27.2 (23.6, 31.3), and 23.1 (19.5, 27.3) ng/24 hours, respectively, corresponding to ratios of stimulated to basal levels of 2.68, 2.87, and 2.12. After adjustment for time in culture the basal PAI-1 release was 4G/4G 10.7, 4G/5G 9.1, and 5G/5G 9.7 ng/24 hours. For interleukin-1 stimulated release the adjusted levels were 26.3, 27.0, and 22.7 ng/24 hours, respectively. Adjusted levels in 4G/4G genotype cells were non-significantly greater than those in cells of 5G/5G genotype by a factor of 1.16 (0.95, 4.08). This study did not demonstrate a significant difference in basal or cytokine stimulated PAI-1 release from cells of different PAI-1 promoter (4G/5G) genotypes but does not exclude increased interleukin-1 stimulated PAI-1 release in the 4G/4G compared with the 5G/5G genotype.

Association of a common polymorphism in the factor XIII gene with myocardial infarction

Factor XIII when activated by thrombin, crosslinks fibrin, however its role in thrombotic disorders is unknown. A common point mutation (G-->T) in exon 2 of the A-subunit gene which codes for an amino acid change three amino acids from the thrombin activation site (Factor XIIIVal34Leu) is a candidate for a role in the pathogenesis of acute myocardial infarction. Factor XIII genotype frequencies were determined in a case-control study of 398 caucasian patients and 196 healthy controls. Patients had undergone angiography for investigation of coronary artery disease and were evaluated for a history of myocardial infarction. The prevalence of the mutation was lower in patients with myocardial infarction than without (32% vs. 50%), p = 0.0009 and than in controls (32% vs. 48%), p = 0.005. Patients possessing the mutation with a history of myocardial infarction had higher PAI-1 concentrations (mean, 27.9 vs. 16.7 ng/ml, p = 0.004) and the PAI-1 4G/4G genotype was commoner (43% vs. 26%, p = 0.03). There was no difference in PAI-1 4G/4G genotype (33% vs. 32%) and PAI-1 levels (mean, 21.0 vs. 20.9 ng/ml) in patients possessing wild type with MI compared to those without MI. These results indicate that the G-->T mutation coding for factor XIIIVal34Leu is protective against myocardial infarction and suggest a mechanism whereby elevated levels of PAI-1 may contribute to vascular risk.

Levels of activated FXII in survivors of myocardial infarction--association with circulating risk factors and extent of coronary artery disease

Little data is available regarding the activated form of factor XIIa (FXIIa) in survivors of myocardial infarction. 292 Caucasian patients characterised for extent of coronary atheroma by angiography and for a past history of myocardial infarction and 77 healthy controls were included in the study. To investigate the relationship between coronary artery disease, activated factor XII and other circulating factors, we studied levels of FXIIa, cholesterol, triglycerides, fasting insulin, fibrinogen, FVII:C, t-PA antigen and PAI-1 antigen. Factor XIIa levels were higher in all patients [2.5 (2.3-2.6) ng/ml] and in patients with a history of MI [2.6 (2.4-2.9) ng/ml] than in controls [1.9 (1.7-2.1) ng/ml], p < 0.0001. In patients, FXIIa levels positively correlated with FVII:C, BMI, cholesterol, insulin, PAI-1 antigen, t-PA antigen and triglycerides. In controls FXIIa levels only correlated with PAI-1 antigen and triglycerides. FXIIa levels were strongly associated with extent of coronary stenosis: 2.8 (2.6-3.1) ng/ml and 2.6 (2.3-2.9 ng/ml) in those with 2 and 3 vessels stenosed compared to 2.1 (1.9-2.3) ng/ml in those with 0 vessel stenosed (p = 0.0004). Activated FXII relates to both extent of coronary atheroma and to a past history of myocardial infarction and clusters with features of insulin resistance.

The paraoxonase Gln-Arg 192 polymorphism in subjects with ischaemic heart disease

BACKGROUND

Human serum paraoxonase activity is related to the paraoxonase Gln-Arg 192 polymorphism genotype. The purpose of this study was to investigate the association between the Gln-Arg 192 polymorphism of paraoxonase and ischaemic heart disease (IHD).

METHODS

Four hundred and forty patients with a history suggestive of IHD, and characterized by coronary angiography, and 527 healthy controls were studied. Patients were grouped according to paraoxonase genotype, presence or absence of diseased coronary arteries (on the basis of 50% stenosis), and history of myocardial infarction as judged by World Health Organization criteria. Patients were genotyped for the paraoxonase Gln-Arg 192 polymorphism by polymerase chain reaction.

RESULTS

No significant relationship was found between paraoxonase genotype and age, sex, body mass index, smoking, triglycerides or hypertension. However, by oneway analysis of variance, cholesterol was found to be significantly associated with paraoxonase genotype in male patients [AA 5.9 (5.8-6.1), AB 6.2 (6.0-6.4), BB 5.7 (5.4-6.1); P = 0.04]. The Gln-Arg 192 polymorphism was found to have no significant effect on the number of patients having diseased coronary arteries, or having myocardial infarction (P = 0.97 for both). In logistic regression models, paraoxonase genotype did not remain a significant independent predictor of stenosis or myocardial infarction.

CONCLUSION

This study failed to show an association between the Gln-Arg 192 polymorphism of paraoxonase and the clinical phenotypes of coronary atheroma and acute myocardial infarction.

Insulin receptor substrate-1 gene polymorphism and cardiovascular risk in non-insulin dependent diabetes mellitus and patients undergoing coronary angiography

Clustering of risk factors for cardiovascular disease related to insulin resistance may account for the increased incidence of vascular disease in these conditions and in non-diabetic subjects. To investigate the relationship between a coding polymorphism in the insulin receptor substrate-1 gene and the presence of cardiovascular risk factors, 209 patients with NIDDM and 452 subjects investigated for coronary artery disease (CAD) were studied. In the NIDDM subjects 22 (10.5%) were heterozygous at codon 972 for a polymorphism which codes for a glycine to arginine substitution and 187 (89.5%) were homozygous for the wild type. Patients with the mutation had lower levels of cholesterol compared with wild type (mean, 95% confidence intervals), 5.3 (4.9-5.8) vs 6.0 (5.9-6.2) mmol/l, respectively (P = 0.002), triglyceride 1.7 (1.4-2.1) vs 2.2 (2.0-2.4) mmol/l (P = 0.051), factor VII:C activity 109.5 (85.5-133.5) vs 133.5 (127-140)% (P = 0.057) and PAI-1 antigen, 16.0 (10.5-24.3) vs 22.2 (20.0-24.6) ng/ml (P = 0.054). There were no differences in body mass index, indices of glycaemic control, fasting insulin or the prevalence of hypertension. In patients with CAD, 55 (12.7%) were carriers of the mutation (including three homozygotes) (NIDDM vs CAD, NS). Although similar trends in cholesterol, factor VII, PAI-1 antigen and triglyceride existed between carriers of the mutation and the wild type, none reached statistical significance. The results indicate that the IRS-1 gene is not implicated in the pathogenesis of NIDDM or CAD.

PAI-1 concentrations in first-degree relatives of patients with non-insulin-dependent diabetes: metabolic and genetic associations

AIM

To identify whether levels of PAI-1 are elevated in relatives of NIDDM patients and the extent to which they relate to features of insulin resistance and to genotype at a common PAI-1 promoter polymorphism.

RESULTS

In 132 first-degree relatives of NIDDM patients and 151 controls PAI-1 activity was higher in relatives 14.4 U/ml than controls 9.3 U/ml (p < 0.0005) with higher body mass index 27.3 v. 24.7 kg/m2 p < 0.0005, fasting insulin 9.3 v. 7.6 mU/l p < 0.005 and triglyceride 1.4 v. 1.2 mmol/l p < 0.05. PAI-1 activity levels were higher in men and in smokers and showed a trend to being higher with increasing number of 4G alleles at the PAI-1 promoter polymorphism. After adjustment for age, sex, smoking, BMI and triglyceride levels, PAI-1 levels remained 26% higher in the NIDDM relatives (p = 0.01).

CONCLUSION

In relatives of NIDDM patients PAI-1 levels are elevated, presenting an additional mechanism for their increased cardiovascular risk.

Plasminogen activator inhibitor-1 promoter 4G/5G genotype and plasma levels in relation to a history of myocardial infarction in patients characterized by coronary angiography

To investigate the relationship between an insertion/deletion (4G/5G) polymorphism in the promoter region of the plasminogen activator inhibitor-1 (PAI-1) gene and the phenotypes of PAI-1 levels, coronary atheroma, and a past history of coronary thrombosis, we studied 453 patients (320 men and 133 women) characterized by coronary angiography. Patients were classified as having normal vessels (n = 125) or single-vessel (n = 92) or multivessel (n = 232) coronary disease on the basis of > or = 50% stenosis. PAI-1 antigen levels were highest in patients with the 4G/4G genotype (22.5 ng/mL), with a stepwise decrease in levels as the number of 4G alleles decreased (21.5 ng/mL for 4G/5G and 15.8 ng/mL for 5G/5G, P = .02) after adjusting for age, sex, triglyceride levels, and body mass index (BMI). The association between triglyceride level and PAI-1 was genotype specific, with a steeper slope in subjects with the 4G/4G genotype (P = .004). A gene-environment interaction between BMI, PAI-1, and genotype was observed, with a steeper association in patients with the 5G/5G genotype (P = .02). The 4G/4G genotype was significantly associated with a history of myocardial infarction (P < .03; odds ratio, 2.0; 95% CI, 1.1 to 3.7). This relationship was stronger in subjects with diseased vessels (P = .006). There was no relationship between either PAI-1 genotype or levels and the presence of atheroma. Our data suggest that PAI-1 promoter polymorphism influences the development of myocardial infarction through its effect on thrombus formation in patients with preexisting coronary atheroma.

Promoter (4G/5G) plasminogen activator inhibitor-1 genotype in Pima Indians: relationship to plasminogen activator inhibitor-1 levels and features of the insulin resistance syndrome

Elevated plasminogen activator inhibitor-1 may contribute to vascular disease in diabetes mellitus. Pima Indians have a low incidence of cardiovascular disease despite having a high prevalence of non-insulin-dependent diabetes mellitus (NIDDM) which in this population is not associated with elevated plasminogen activator inhibitor-1 activity. In Caucasians an insertion/deletion (4G/5G) polymorphism in the promoter region of the plasminogen activator inhibitor-1 gene that has been related to activity levels of its protein in plasma differentially binds repressor and enhancer elements. In 265 Pima Indians (133 diabetic, 132 non-diabetic, 129 male, 136 female, male, mean age 46.6, range 34-68 years) the promoter genotype frequencies were 23.0% for 4G/4G, 49.8% for 4G/5G and 27.2% for 5G/5G compared to 35.4%, 50.8% and 13.8% respectively, (chi 2 = 15.3, 2 df, p < 0.0005) previously reported in Caucasians with NIDDM. The mean plasma activity levels in the three genotypes in the Pima Indians were 18.2, 19.1 and 18.1 U/ml, respectively. Plasminogen activator inhibitor-1 activities correlated with plasma insulin (r = 0.38, p < 0.0001), body mass index (r = 0.24, p < 0.0001), and with triglyceride level (r = 0.12, p = 0.054) but there was no relationship between promotor genotype and activity. A steeper regression slope between plasminogen activator inhibitor-1 activity and triglycerides has been observed in Caucasians with the 4G/4G genotype as compared to Caucasians with the other genotypes. This was not found in the Pima population which may indicate a functional difference in this gene associated with reduced cardiovascular risk and may be involved in the lack of association of plasminogen activator inhibitor-1 levels with NIDDM in Pima Indians.

Beta-fibrinogen gene-455 G/A polymorphism and fibrinogen levels. Risk factors for coronary artery disease in subjects with NIDDM

OBJECTIVE

To investigate the association of a G/A polymorphism at position -455 of the beta-fibrinogen gene and fibrinogen levels in the development of coronary artery disease (CAD) in subjects with NIDDM.

RESEARCH DESIGN AND METHODS

In 187 Caucasian subjects with NIDDM, presence of CAD was taken as a clinical history of angina, myocardial infarction, coronary angioplasty, or coronary artery bypass grafting, confirmed from medical case notes.

RESULTS

Fibrinogen levels were significantly higher in subjects with CAD (n = 38, 3.41 [3.12-3.73] g/l) than those without (n = 149, 2.99 [2.87-3.11] g/l, P = 0.004 [geometric mean, 95% CI]). Comparing the genotype frequencies by chi 2 testing, there was a significant difference between subjects with CAD (GG = 0.84, GA + AA = 0.16) and those without (GG = 0.64, GA + AA = 0.36, P = 0.02). In a logistic regression model, including age, BMI, cholesterol, sex, smoking, triglycerides, and plasminogen activator inhibitor antigen as covariates, genotype and fibrinogen levels were significantly associated with CAD. The odds ratio for having CAD in individuals homozygous for the G allele compared with those possessing the A allele was 1.77 (1.08-2.90), P = 0.03; and for an increase of 1 g/l in fibrinogen levels was 1.60 (1.00-2.60), P = 0.05.

CONCLUSIONS

These data suggest a relationship between the -455 G/A beta-fibrinogen gene polymorphism and the development of CAD in subjects with NIDDM. This relationship was not associated with variations in fibrinogen levels, suggesting that this polymorphism may be in linkage with a polymorphism within the coding region of the beta-fibrinogen gene, which results in an alteration in the stability of the fibrin clot.

Angiotensin converting enzyme (ACE) insertion/deletion (I/D) polymorphism, and diabetic retinopathy in subjects with IDDM and NIDDM

Angiotensin 1 converting enzyme (ACE) catalyses the step which generates angiotensin II, and also inactivates bradykinin, peptides which play a key role in modulating vascular tone. Plasma ACE levels are under genetic control and up to 50% of the variation is due to an insertion/deletion (I/D) polymorphism of ACE gene with highest levels found in DD homozygotes. Studies have shown an association of diabetic nephropathy and ischaemic heart disease with angiotensin converting enzyme gene polymorphism in subjects with diabetes. We examined the association between diabetic retinopathy and ACE gene insertion/deletion polymorphism in 363 subjects with NIDDM (aged 68.3 +/- 10.7 years; 201 male, 162 female), 186 subjects with IDDM (aged 42.4 +/- 15.0 years; 100 male, 86 female) and 98 controls. These subjects were characterized for ACE I/D polymorphism employing standard primers. Diabetic retinopathy was diagnosed by ophthalmoscopy through dilated pupils by an ophthalmologist and classified as non-proliferative or proliferative retinopathy. As expected, diabetic retinopathy was strongly associated with duration of diabetes (p < 0.001) in both IDDM and NIDDM. Any retinopathy was present in 51% subjects with IDDM and 49% of subjects with NIDDM, while 22% of IDDM subjects and 5% of subjects with NIDDM had proliferative retinopathy. The frequency of I allele was 0.477 vs 0.482 vs 0.510 and D allele was 0.523 vs 0.518 vs 0.490, among subjects with IDDM, NIDDM and controls, respectively. The frequency of ACE I/D genotype was similar in subjects with IDDM, NIDDM, and controls (chi 2 = 0.46, df = 4, p = ns). Presence or absence of retinopathy was not significantly associated with ACE genotype in subjects with IDDM (chi 2 = 3.42, df = 2, p = ns) or NIDDM (chi 2 = 0.51, df = 2, p = ns). Among subjects with retinopathy, there was no significant association between ACE genotype and type of retinopathy. Controlled for duration of diabetes, the frequency of I/D genotype was not significantly different in 271 subjects with retinopathy (IDDM and NIDDM combined) when compared with 86 subjects without retinopathy at 15 years or more after diagnosis of diabetes (chi 2 = 1.29, df = 2, p = ns). These findings indicate that I/D polymorphism of ACE gene is not a useful marker and is unlikely to play a major role in determining genetic susceptibility to diabetic retinopathy or the severity of diabetic retinopathy.

Plasminogen activator inhibitor-1 (PAI-1) promoter polymorphism and coronary artery disease in non-insulin-dependent diabetes

Elevated levels of PAI-1 are found in coronary artery disease (CAD) and non-insulin-dependent diabetes (NIDDM). PAI-1 may be involved in the pathogenesis of CAD through suppression of fibrinolysis, alternatively the high levels may result from vascular damage. There is evidence that PAI-1 levels are related to genotype at a PAI-1 promoter polymorphism. Genotype at this 4G/5G polymorphism was determined in 160 NIDDM (90 males and 70 females) patients with (n = 38) or without (n = 122) clinical evidence of CAD. Levels of cholesterol were higher (6.5 vs 5.9 mM, p < 0.01) and PAI-1 tended to be higher (PAI-1 activity 23.0 vs 20.4 U/ml) with CAD. The frequency of the 4G/4G genotype was increased and the 5G/5G genotype decreased, in the group CAD compared to those without (p < 0.05). These results suggest that possession of the 4G/4G PAI-1 promoter genotype is a risk factor for the development of CAD in subjects with NIDDM.

Environmental and genetic factors in relation to elevated circulating levels of plasminogen activator inhibitor-1 in Caucasian patients with non-insulin-dependent diabetes mellitus

To investigate the interaction of metabolic and genetic factors in relation to PAI-1, genotype was determined at a 4G/5G polymorphism in the PAI-1 gene promoter and a Hind III RFLP of the PAI-1 gene in 189 Caucasian NIDDM patients. PAI-1 levels were equivalent in each genotype group and PAI-1 activity correlated with fasting insulin (r=0.45), triglyceride (r=0.39) body mass index (r=0.44), cholesterol (r=0.17) and glucose (r=0.15). The regression slope (B) of PAI-1 activity on triglycerides was steeper in the 4G/4G group than the other two groups: 4G/4G B = 0.91, r = 0.62; 4G/5G B = 0.36, r=0.27; 5G/5G B=0.31, r=0.29 (difference between slopes p=0.02) and the association between PAI-1 activity and glucose remained only in the 4G/4G group (r=0.35). These results confirm the association of PAI-1 levels with the features of insulin resistance and indicate that the association between PAI-1 levels and both triglyceride and glucose is influenced by genotype in the region of PAI-1 gene promoter.

Polymorphisms of the plasminogen activator inhibitor-1 gene in type 1 and type 2 diabetes, and in patients with diabetic retinopathy

To identify whether genotype contributes to the difference in PAI-1 levels in type 1 and type 2 diabetic subjects and whether genotype relates to the development of retinopathy, a Hind III restriction fragment length polymorphism and two dinucleotide repeat polymorphisms were studied. In 519 Caucasian diabetic subjects (192 type 1, 327 type 2) and 123 Caucasian control subjects there were no differences in the frequency of the Hind III restriction alleles (type 1 vs type 2 vs control: allele 1 0.397 vs 0.420 vs 0.448; allele 2 0.603 vs 0.580 vs 0.552) nor in the allelic frequency at either dinucleotide repeat sequence. In 86 subjects with no retinopathy at 15 years or more from diagnosis of diabetes and 190 subjects with diabetic retinopathy there was no difference in the frequency of Hind III restriction alleles (retinopathy present vs retinopathy absent: allele 1 0.400 vs 0.467; allele 2 0.600 vs 0.533) nor in the allelic frequencies at either dinucleotide repeat sequence. The results indicate that there is no or minimal influence of the PAI-1 gene on either PAI-1 levels or the development of diabetic retinopathy in patients with diabetes mellitus.

The effects of insulin-like growth factor-1 on plasminogen activator inhibitor-1 synthesis and secretion: results from in vitro and in vivo studies

In vitro studies have shown that insulin and IGF-1 releases the fibrinolytic inhibitor plasminogen activator inhibitor-1 (PAI-1) from cells of hepatic origin. To investigate the effects of IGF-1 on fibrinolysis: 1) cultured hepatoma cells were grown in the presence of IGF-1 and media collected for secreted PAI-1 and cells probed for PAI-1 mRNA, 2) 8 hypopituitary patients were treated with recombinant human growth hormone (rhGH) and 3) 5 type 2 diabetic patients were treated with recombinant human IGF-1 (rhIGF-1). Treatment of Hep G2 cells with IGF-1 (1000 ng/ml) increased secretion of PAI-1 from a median value of 80 ng/10(6) cells (range 21-91) to 144 ng/10(6) cells (range 128-169) after 24 h (p < 0.01). Synthesis of PAI-1 mRNA increased in a similar fashion. Treatment of hypopituitary patients with rhGH led to an increase in circulating IGF-1 from a mean value of 166 (range 41-324) ng/ml at baseline to 322 (77-575) ng/ml at 4 weeks and 259 (104-533) ng/ml after 8 weeks (p < 0.02). Despite this, no changes in circulating PAI-1 or fibrinolysis occurred. Type II diabetic patients treated with rhIGF-1 showed an increase in circulating IGF-1 from a mean value of 120 ng/ml (range 109-196), at baseline to 823 ng/ml (585-894) after 5 days. This also was not associated with changes in circulating PAI-1 or in fibrinolysis. The results confirm that IGF-1 induces the synthesis of PAI-1 in Hep G2 cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Thrombin generation and factor VIII:C levels in patients with type 1 diabetes complicated by nephropathy

The association between plasma coagulant activity and the presence of diabetic nephropathy was investigated in 31 patients with Type 1 diabetes, 12 with and 19 without nephropathy, and in 11 healthy subjects. Thrombin generation was assessed by computer assisted chromogenic method and expressed as time (in seconds) to 50% maximal thrombin activity (T50). Factor VIII:C levels related to thrombin activity, glycaemic control, and renal function. Median (IQ) FVIII:C concentration was increased in patients with nephropathy (1590 (1130-1900) IU l-1) compared to those without renal disease and with controls (960 (750-1090); 1020 (810-1100) IU l-1, p < 0.01, respectively). There were no significant differences in T50 values between the groups. FVIII:C correlated with age in all subjects and in the diabetic group (r = 0.33, p = 0.036; r = 0.39, p = 0.031) and inversely with T50 in all subjects and in controls (r = -0.35, p = 0.02; r = -0.62, p = 0.04). In all subjects and in patients, FVIII:C was related to urinary albumin excretion and creatinine clearance. FVIII:C and T50 were not related to HbA1c. The results show that FVIII:C levels are increased in Type 1 diabetes complicated by nephropathy and are related to degree of renal impairment but not levels of glycaemia. No associated enhancement of plasma procoagulant activity was detected.

The Effects of Insulin-Like Growth Factor-1 on Plasminogen Activator Inhibitor-1 Synthesis and Secretion: Results from In Vitro and In Vivo Studies

In vitro studies have shown that insulin and IGF-1 releases the fibrinolytic inhibitor plasminogen activator inhibitor-1 (PAI-1) from cells of hepatic origin. To investigate the effects of IGF-1 on fibrinolysis: 1) cultured hepatoma cells were grown in the presence of IGF-1 and media collected for secreted PAI-1 and cells probed for PAI-1 mRNA, 2) 8 hypopituitary patients were treated with recombinant human growth hormone (rhGH) and 3) 5 type 2 diabetic patients were treated with recombinant human IGF-1 (rhIGF-1). Treatment of Hep G2 cells with IGF-1 (1000 ng/ml) increased secretion of PAI-1 from a median value of 80 ng/106 cells (range 21-91) to 144 ng/106 cells (range 128-169) after 24 h (p <0.01). Synthesis of PAI-1 mRNA increased in a similar fashion. Treatment of hypopituitary patients with rhGH led to an increase in circulating IGF-1 from a mean value of 166 (range 41-324) ng/ml at baseline to 322 (77-575) ng/ml at 4 weeks and 259 (104-533) ng/ml after 8 weeks (p <0.02). Despite this, no changes in circulating PAI-1 or fibrinolysis occurred. Type II diabetic patients treated with rhIGF-1 showed an increase in circulating IGF-1 from a mean value of 120 ng/ml (range 109-196), at baseline to 823 ng/ml (585-894) after 5 days. This also was not associated with changes in circulating PAI-1 or in fibrinolysis. The results confirm that IGF-1 induces the synthesis of PAI-1 in Hep G2 cells. However, marked increases in IGF-1 had no effect on circulating PAI-1 or fibrinolysis.

Comparison of measures of body composition in a trial of low dose growth hormone replacement therapy

OBJECTIVE

We assessed the effects of the administration of low dose growth hormone in growth hormone deficient adults on body composition and physical performance. We compared the validity of different measures of body composition in GH treated adults.

DESIGN

An uncontrolled longitudinal study of eight patients with GH deficiency who were treated with 4 units of biosynthetic growth hormone (Norditropin), three times a week. Subjects were studied for 8 weeks.

PATIENTS

Eight patients with acquired growth hormone deficiency as defined by < 5 mU/l GH following standard provocative investigations in whom other hormone replacement was optimized.

MEASUREMENTS

IGF-I was measured on day 1 and after 4 and 8 weeks of treatment with recombinant human growth hormone. Body composition was estimated by dual energy X-ray absorptiometry, bioelectrical impedance, skinfold anthropometry, total body potassium measurement and by computerized tomography of a representative cross-sectional area of thigh on day 1 and following 8 weeks of treatment. Exercise capacity was measured on an electromechanical bicycle and palmar grip strength was measured using a Jamar dynamometer on day 1 and after 8 weeks.

RESULTS

IGF-I increased significantly. Exercise capacity and maximum heart rate achieved on the electromechanical bicycle increased significantly. Grip strength did not change. There was a significant increase in fat-free mass and a decline in fat mass as shown by dual energy X-ray absorptiometry and bioelectrical impedance. However, this was not confirmed by other methods. No side-effects were noted.

CONCLUSIONS

This study shows that a low dose of biosynthetic growth hormone can elevate IGF-I levels and have a pronounced physical impact in the growth hormone deficient adult without the side-effects seen at higher dosage schedules. Over a 2 month period the increase in fat-free mass may be due to an increase in total body water; however, the decline in fat mass is a genuine effect.

Metformin causes a reduction in basal and post-venous occlusion plasminogen activator inhibitor-1 in type 2 diabetic patients

The effects of metformin on the fibrinolytic system were studied pre- and post-venous occlusion in 38 Type 2 diabetic patients in a double-blind, placebo-controlled trial. After a 3-week run-in period, 21 patients received metformin and 17 placebo, for 6 weeks. In the metformin-treated patients basal plasminogen activator inhibitor-1 antigen (PAI-1Ag) fell from 57.4 micrograms l-1 before treatment to 36.1 (p less than 0.05) and 41.0 micrograms l-1 (p less than 0.01) after 3 and 6 weeks therapy. In this group post-venous occlusion PAI-1Ag also fell after 3 weeks (p less than 0.002) and 6 weeks (p less than 0.05) treatment. There were no changes in either basal or post-venous occlusion concentrations of PAI-1Ag in the placebo treated group. The fall in PAI-1Ag was not associated with an increase in basal plasminogen activator activity (PAA) which remained unchanged in both groups. Post-venous occlusion values for PAA in the metformin treated patients were increased at 3 weeks (p less than 0.05) although there was no difference at 6 weeks.

Increased urinary excretion of transferrin in children with type 1 diabetes mellitus

Urinary transferrin excretion was measured by radioimmunoassay in 74 children with Type 1 diabetes mellitus and in 40 normal children, and compared with urinary excretion of albumin, alpha-1-microglobulin, and N-acetyl-beta-D-glucosaminidase. Urinary transferrin excretion was significantly elevated in diabetic (median (range) 186 (18-1671) mg mol-creatinine-1) compared with normal (85 (27-668) mg mol-creatinine-1) children (p less than 0.001). Seventeen diabetic children had transferrin excretion above the 95th centile for normal children. In contrast there was no significant increase in urinary albumin excretion in the diabetic children although 8 had urinary albumin excretion which exceeded the 95th centile for normal children (6 of these 8 patients having coexistent urinary hyperexcretion of transferrin). Urinary transferrin excretion correlated significantly with urinary albumin excretion in both normal (rs = 0.62, p less than 0.001) and diabetic (rs = 0.61, p less than 0.001) children. The indices of proximal renal tubular function (urinary excretion of alpha-1-microglobulin and N-acetyl-beta-D-glucosaminidase) correlated significantly with transferrin excretion in both diabetic (rs = 0.43 and rs = 0.41, p less than 0.001) and normal (rs = 0.40, p less than 0.02 and rs = 0.53, p less than 0.001) children, but not with albumin excretion (rs = 0.20, p greater than 0.05 and rs = 0.22, p greater than 0.05). In addition urinary transferrin excretion significantly correlated with urinary glucose concentration (rs = 0.34, p less than 0.007) in Type 1 diabetic children. The discrepancy in urinary excretion of transferrin and albumin may reflect impaired proximal renal tubular reabsorption of transferrin and/or altered glomerular basement membrane selectivity for the two proteins.

Insulin increases plasma somatomedin C (IGF-1) concentrations in adult type 1 diabetic patients

The role of insulin in the control of somatomedin release has been investigated in people with Type 1 diabetes. Six patients underwent insulin-induced hypoglycaemia of 20 min duration and 8 patients were studied using the hyperinsulinaemic euglycaemic clamp technique at insulin infusion rates of 0.25, 1.25, 2.5, and 0.25 mU kg-1 min-1 for 1 h at each rate. In the first study plasma insulin concentrations rose from (median, range) 23 (10-36) to 114(60-200) mU l-1 at the onset of hypoglycaemia, and fell to 53 (23-100) mU l-1 after 20 min hypoglycaemia and 30 (15-73) mU l-1 on return to normoglycaemia. Plasma IGF-1 rose from 140 (96-292) to 179 (127-352) micrograms l-1 (p less than 0.05) at the onset of hypoglycaemia and fell to 131 (125-173) micrograms l-1 after 20 min and to 154 (121-287) micrograms l-1 at the end of the study. During hyperinsulinaemia plasma insulin rose from 23 (19-40) mU l-1 at the lowest infusion rate to 61 (33-84) and 148 (68-200) mU l-1 at the two higher infusion rates. Over the same period, plasma IGF-1 increased from 91 (13-203) to 123 (98-182) micrograms l-1 (p less than 0.05) and 109 (84-160) micrograms l-1. There was no correlation between growth hormone levels and IGF-1 in either study. These results suggest that insulin produces short-term increases in IGF-1 levels in man in the absence of a growth hormone response.

Acute changes in blood glucose concentration do not promote thrombin generation or fibrin breakdown in type 1 diabetes

To investigate the effect of blood glucose concentration on thrombin generation and fibrinolytic activity, six Type 1 patients had the blood glucose concentration maintained for 1 h at 5, 15, and 25 mmol l-1, and 8 patients underwent hypoglycaemia of 20 min duration after the blood glucose had been kept at 8 mmol l-1 for 1 h. During hyperglycaemia plasminogen activator activity rose from 214 (11-625) (median, range) to 478 (18-772) units (p less than 0.05) at a blood glucose of 5 mmol l-1 and to 511 (89-816) (p less than 0.05) and 535 (33-976) (p less than 0.05) units at a blood glucose of 15 and 25 mmol l-1, respectively. Cross-linked fibrin degradation products (FDP) were 45 and 53 micrograms l-1 at a blood glucose of 5 mmol l-1 and remained unchanged at higher glucose levels. Fibrinopeptide A was 1.3 (0.6-2.8) nmol l-1 at a blood glucose of 5 mmol l-1, and remained unchanged with hyperglycaemia, being 1.3 (0.9-1.3) nmol l-1 after 1h at 25 mmol l-1. During hypoglycaemia, plasminogen activator activity rose from 155 to 745 units (p less than 0.05) while both fibrinopeptide A and cross-linked FDP remained unchanged. The results indicate that acute fluctuations in blood glucose concentration do not lead to thrombin generation. Additionally, increased fibrinolytic activity measured in vitro is not associated with an increase in cross-linked FDP. This suggests that short-term hyper- and hypoglycaemia do not affect the end-products of the coagulation and fibrinolytic pathways.

Laboratory, nurse and patient assessment of Glucostix blood glucose reagent strips

The new Glucostix (Ames) blood glucose reagent strips have been assessed for accuracy and acceptability by nurses using them on the ward and by patients at home. When used with a Glucometer II, Glucostix showed close correlation with blood glucose measured by Yellow Springs Analyser (r = 0.97, p less than 0.001). Using the strips without a meter patients achieved a good correlation with the Yellow Springs Analyser when measuring their own capillary blood glucose (r = 0.94, p less than 0.001). Ward nurses were less precise (r = 0.87, p less than 0.001) and this, perhaps, reinforces the importance of regular quality control to maintain the precision of blood glucose monitoring using reagent strips. Comparing Glucostix with BM Glycemie 1-44 strips, 14 of 15 nurses preferred Glucostix whereas patients showed no overall preference, 37% preferring BM Glycemie 1-44, 33% Glucostix, and 30% stating no preference. There were only minor differences between the two reagent strips. Both groups considered the Glucostix colour change to be easier to read, particularly when using the perforated bench readers, but some thought Glucostix needed more blood and that it was more difficult to remove from the pad. Thus, the Glucostix system provides an accurate measure of blood glucose and is an acceptable alternative to BM Glycemie 1-44 strips for use at home or on the ward.

Effect of controlled hyperglycaemia on factor VIII concentrations in insulin dependent diabetes mellitus

Patients with diabetes mellitus have higher levels of coagulation factor VIII than the non-diabetic population. This may be a result of poor metabolic control and could contribute to the development of microvascular complications. During ketoacidosis there are acute changes in plasma concentrations of coagulation factors, some of which may be mediated by the rise in vasopressin that occurs. We have investigated the effects of hyperglycaemia without ketosis on some aspects of haemostasis by manipulating blood glucose concentrations using a Biostator. After a 1h run-in period with the blood glucose at 5 mmol/l, the blood glucose was maintained at 5, 15 and 25 mmol/l and maintained for one hour at each level in six male patients with insulin-dependent diabetes. Insulin was infused at 0.25 mu/kg/min. Venous blood samples were taken at the beginning and end of each hour after the run-in period for assays of factor VIII coagulant activity (FVIII:C), von Willebrand factor antigen (vWF:Ag), ristocetin co-factor (FVIIIR:Co), activated partial thromboplastin time (APTT) and vasopressin (aVP). There was a slight, though statistically insignificant fall in median factor VIII:C concentration at each incremental level of increase in blood glucose. Values (at the beginning and end of each hour) were: 1.0 and 1.1 iu/ml at 5 mmol/l; 0.95 and 0.79 iu/ml at 15 mmol/l; and 0.74 and 0.84 iu/ml at 25 mmol. vWF:Ag and FVIIIR:Co were unchanged. Plasma aVP fell slightly from 1.1 to 0.5 pg/ml. The results indicate that high levels of FVIII seen in diabetes are not due to short-term increases in blood glucose and that acute hyperglycaemia does not promote pro-coagulant changes in blood.

Regional variation in skin blood flow response to hypoglycaemia in type 1 (insulin-dependent) diabetic patients without complications

Body temperature falls during hypoglycaemia, perhaps as a protective mechanism. To test the hypothesis that the skin blood flow response to hypoglycaemia is specifically designed to facilitate heat loss we studied both nutritional blood flow and arteriovenous shunting of blood in skin during prolonged, controlled hypoglycaemia in man. We studied eight otherwise healthy, male, Type 1 (insulin-dependent) diabetic patients. Under Biostator control blood glucose was clamped at 8.0 (7.9-8.9), mmol/l (median and range) for 30 min, reduced to symptomatic hypoglycaemia, 1.7 (1.0-2.6) mmol/l for 20 min then raised to 4.9 (3.3-6.7) mmol/l. Interdigital skin web blood flow (laser doppler flowmeter, nutritional flow) fell during hypoglycaemia from 3.1 (2.2-3.8) to 2.4 (1.2-2.8) volts and remained depressed. In contrast, finger blood flow (venous occlusion plethysmography, arteriovenous shunt flow) started high at 54.7 (17.4-85.6), remained high at 52.7 (38.1-81.4) during hypoglycaemia but fell sharply to 25.3 (4.2-66.2) ml.min-1.100 ml-1 when symptoms were relieved. Plasma adrenaline and vasopressin both rose during hypoglycaemia from 0.4 (0.05-0.8) to 4.5 (2.3-20.2) nmol/l and from 0.5 (0.5-3.5) to 4.4 (2.0-13.9) pg/ml, respectively, and both fell sharply thereafter.(ABSTRACT TRUNCATED AT 250 WORDS)

The relationship between peripheral nerve resistance to ischaemia and diabetic control

The cause of the abnormal resistance to ischaemia of peripheral nerve function in diabetes is unknown. Median nerve function was more resistant to ischaemia in diabetic patients than in control subjects. In diabetic patients the degree of resistance to ischaemia correlated closely with HbAlc but not with the coincident blood glucose level, the duration of diabetes, the vibration perception threshold at the thumb or the initial median nerve action potential amplitude. Thus in diabetes the resistance of peripheral nerve function to ischaemia is dependent on medium term metabolic control and is not directly related to the presence or absence of neuropathy.

Hormonal control of haemostasis during hypoglycaemia in diabetes mellitus

Factor VIII (FVIII) and plasminogen activator activity (PAA) rise during hypoglycaemia, and this might contribute to the vascular complications of diabetes. Similar changes in haemostasis accompany raised plasma levels of vasopressin (aVP) and adrenaline. To investigate the effects of these hormones on haemostasis during hypoglycaemia and the role of plasma insulin concentrations, eight insulin-dependent diabetic patients underwent controlled hypoglycaemia for 20 min and 13 diabetic patients were investigated during hyperinsulinaemia with blood glucose maintained at 8.0 mmol/l. During hypoglycaemia, insulin levels increased to median values of 114 mU/l, a VP rose from 0.5 to 4.4 (p less than 0.005) pg/ml and adrenaline from 0.4 to 4.4 nmol/l (p less than 0.005). FVIII coagulant activity (FVIII:C) rose from 0.75 to 1.09 IU/ml (p less than 0.01) and the ristocetin co-factor (FVIIIR:Co) and von Willebrand factor antigen (vWF:Ag) showed similar responses. PAA increased from 156 to 745 units (p less than 0.005). During hyperinsulinaemia, insulin rose following infusion from 24 to 52 and 118 mU/l, maintained for an hour at each level. Despite this, plasma aVP, FVIII:C, FVIIIR:Co, vWF:Ag and PAA remained unchanged. This study indicates that the marked changes in FVIII, vWF and PAA concentrations which accompany hypoglycaemia depend on low blood glucose and not raised plasma insulin. The response in probably mediated by increases in adrenaline and aVP, which are part of the physiological response to hypoglycaemia.

Red cell sorbitol levels during hyper and hypoglycaemic clamp studies in insulin-dependent diabetic patients

The red cell sorbitol concentration has been suggested as a measure of polyol pathway activity. Red cell sorbitol levels were higher in 53 patients having insulin-dependent diabetes mellitus (IDDM) than in 16 control subjects. Six patients having IDDM underwent hyperglycaemic 'clamp' studies; the red cell sorbitol level returned to the normal range when the blood glucose was clamped at 5 mmol/l for 1 h and rapidly increased when it was clamped at 15 and 25 mmol/l for a further hour at each level. Seven patients with IDDM were rendered hypoglycaemic; red cell sorbitol levels rapidly fell to a level less than, but not significantly different from normal. The results of these studies suggest that in IDDM red cell sorbitol levels are a reflection of prevailing blood glucose concentration and do not indicate long-term sorbitol accumulation in other tissues.

Haemoglobin A1c levels in normal and diabetic pregnancies

Serial measurements of the HbA1c levels were performed during pregnancy in 4 groups of patients attending Antenatal Clinics: 36 normal pregnancies; 16 pregnancies in established insulin-dependent diabetic patients; 9 patients with gestational diabetes diagnosed during that pregnancy; and 21 patients who had been diagnosed as having gestational diabetes in at least one previous pregnancy. In the normal pregnancy HbA1c levels showed a small but significant increase from the end of the first trimester to delivery despite blood glucose levels remaining constant throughout. In the insulin-dependent and gestational diabetic patients, blood glucose levels remained significantly higher than in the normal throughout pregnancy but only in insulin-dependent diabetic patients and the newly diagnosed untreated gestational diabetic patients were the HbA1c levels significantly higher than in the normal. In those patients who had previous pregnancies complicated by gestational diabetes, blood glucose levels were significantly higher than in the normal but HbA1c levels were not. This dissociation between blood glucose and HbA1c levels in gestational diabetic pregnancies in particular limits the value of HbA1c levels in monitoring antidiabetic treatment in such pregnancies.

The Measurement of Haemoglobin A1c by isoelectric focussing in diabetic patients

Incubation in vitro of human red cells in increasing glucose concentrations results in a rise in both haemoglobin A1c levels and an intermediate band when measured by an isoelectric focussing method. There are strong correlations between blood glucose levels, levels of haemoglobin A1c and the intermediate band both in vitro and in blood samples taken from diabetic patients. As the intermediate band is also included in the measurement of haemoglobin A1c by the usual analytical methods, this may lead to inaccurate results.