Apolipoprotein(a) in insulin-dependent diabetic patients with and without diabetic nephropathy

Lipoprotein(a) serum levels in diabetic patients with retinopathy

BACKGROUND

Atherogenic lipoproteins, such as total cholesterol, LDL cholesterol, oxidized low density lipoprotein, and triglycerides, are associated with progression of retinopathy. Aim. To evaluate the relationship between lipoprotein(a) and retinopathy in patients with type 2 diabetes mellitus.

MATERIALS AND METHODS

We enrolled 145 diabetic consecutive patients (82 females, 63 males; mean age 66.8 ± 12 years, mean duration of diabetes 9.4 ± 6.8 years). Presence and severity of retinopathy were evaluated. Serum lipid profile, including Lp(a) level, was assessed.

RESULTS

High Lp(a) levels have been observed in 54 (78.3%) subjects and normal levels in 13 (18.85%) subjects as regards diabetic patients with retinopathy. Lp(a) levels were high in 15 subjects (21.75%) and normal in 63 subjects (91.35%) as regards patients without retinopathy.

CONCLUSIONS

Lp(a) levels are increased in a significant percentage of patients with retinopathy compared to diabetic patients without retinopathy. The impact of Lp(a) levels on diabetic retinopathy needs to be further investigated.

The role of lipids in the development of diabetic microvascular complications: implications for therapy

Dyslipidemia is a major factor responsible for coronary heart disease and its reduction decreases coronary risk in patients with diabetes mellitus. However, the association of dyslipidemia with microvascular complications and the effect of intervention with lipid-lowering therapy in diabetes have been less investigated. We present the systematic review of association and intervention studies pertaining to dyslipidemia and microvascular disease in diabetes and also review possible mechanisms. Dyslipidemia may cause or exacerbate diabetic retinopathy and nephropathy by alterations in the coagulation-fibrinolytic system, changes in membrane permeability, damage to endothelial cells and increased atherosclerosis. Hyperlipidemia is associated with faster decline in glomerular filtration rate and progression of albuminuria and nephropathy. Recent evidence also suggests a role of lipoprotein(a) in progression of retinopathy and nephropathy in patients with diabetes mellitus. Lipid-lowering therapy, using single agents or a combination of drugs may significantly benefit diabetic retinopathy and diabetic nephropathy. In particular, hydroxymethyl glutaryl coenzyme A reductase inhibitors may be effective in preventing or retarding the progression of microvascular complications because of their powerful lipid-lowering effects and other additional mechanisms. However, most of the data are based on short-term studies, and need to be ascertained in long-term studies. Until more specific guidelines are available, aggressive management of diabetic dyslipidemia, according to currently accepted guidelines, should be continued for the prevention of macrovascular disease which would also benefit microvascular complications.

Plasma lipids and urinary albumin excretion rate in Type 1 diabetes mellitus: the EURODIAB IDDM Complications Study

AIMS

To examine the relationship between increased urinary albumin excretion rate and fasting plasma lipids among male and female respondents to the EURODIAB IDDM Complications Study, and attempt to explain inconsistencies in previous reports.

METHODS

A cross-sectional study of 3250 randomly selected Type 1 diabetic patients from 31 diabetes clinics in 16 European countries was carried out between 1989 and 1990. Plasma lipids and urinary albumin were measured centrally. The present analysis was confined to the subgroup of 2205 patients attending after a 10-12 h overnight fast. Mean age was 33 years (SD 10) and mean duration of Type 1 diabetes mellitus was 15 years (SD 9).

RESULTS

The prevalence of microalbuminuria (24-h urinary albumin excretion rate 20-200 microg/min) was 21.7% (95% confidence interval 19.9-23.5) and macroalbuminuria (24-h urinary albumin excretion rate > 200 microg/min) 7.8% (6.6-9.0). In comparison to patients with normal urinary albumin excretion rate (< 20 microg/min), and after controlling for age, sex, glycaemic control, duration of diabetes and current smoking, macroalbuminuria was associated with significantly (P<0.01) increased fasting plasma triglycerides, cholesterol, LDL-cholesterol, cholesterol:HDL-cholesterol ratio and, in women, reduced HDL-cholesterol. In men and women with microalbuminuria, the only significant association was with increased plasma triglycerides.

CONCLUSIONS

These data confirm that there is an association between fasting plasma lipids and increasing urinary albumin excretion rate in European Type 1 diabetic patients. In microalbuminuric patients, however, the association was weaker than previously reported and partly explained by confounding factors.

Glycation amplifies lipoprotein(a)-induced alterations in the generation of fibrinolytic regulators from human vascular endothelial cells

Increased lipoprotein(a) [Lp(a)] in plasma is an independent risk factor for premature cardiovascular diseases. The levels of glycated Lp(a) are elevated in diabetic patients. The present study demonstrated that glycation enhanced Lp(a)-induced production of plasminogen activator inhibitor-1 (PAI-1), and further decreased the generation of tissue-type plasminogen activator (t-PA) from human umbilical vein endothelial cells (HUVEC) and human coronary artery EC. The levels of PAI-1 mRNA and its antigen in the media of HUVEC were significantly increased following treatments with 5 microgram/ml of glycated Lp(a) compared to equal amounts of native Lp(a). The secretion and de novo synthesis of t-PA, but not its mRNA level, in EC were reduced by glycated Lp(a) compared to native Lp(a). Treatment with aminoguanidine, an inhibitor for the formation of advanced glycation end products (AGEs), during glycation normalized the generation of PAI-1 and t-PA induced by glycated Lp(a). Butylated hydroxytoluene, a potent antioxidant, inhibited native and glycated Lp(a)-induced changes in PAI-1 and t-PA generation in EC. The results indicate that glycation amplifies Lp(a)-induced changes in the generation of PAI-1 and t-PA from venous and arterial EC. This may attenuate fibrinolytic activity in blood circulation and potentially contributes to the increased incidence of cardiovascular complications in diabetic patients with hyperlipoprotein(a). EC-mediated oxidative modification and the formation of AGEs may be implicated in glycated Lp(a)-induced alterations in the generation of fibrinolytic regulators from vascular EC.

The relationship of prorenin values to microvascular complications in patients with insulin-dependent diabetes mellitus

We have performed a cross-sectional analysis of the relationship between prorenin values and the microvascular complications of diabetes in a well controlled population of insulin-dependent diabetes mellitus (IDDM) subjects. One hundred and thirty-nine subjects (75 men, 64 women, age 44 +/- 17 years; duration of diabetes 19 +/- 15 years), formed the study group. Sixty-seven subjects (48.2%) had no complications, 55 (39.6%) had retinopathy alone, and 17 (12.2%) had retinopathy and albuminuria. Patients with no complications had lower prorenin values than those with microvascular complications (p < 0.001), whilst patients with both albuminuria and retinopathy had higher values than those with retinopathy alone (p < 0.05). Retinopathy was associated with duration of diabetes (p < 0.0001), diastolic blood pressure (p < 0.02) and albuminuria (p < 0.0001) while albuminuria was associated with prorenin (p < 0.02), serum triglyceride (p < 0.01) and retinopathy (p < 0.001). Patients with albuminuria were 5.5 times more likely to have raised prorenin values (>80 ng/mL/h) than those with normal albumin excretion [95% confidence interval (CI): 1.48-20.12] and those with retinopathy alone were 2.5 times as likely (95% CI: 1.19-5.15). Eighty patients with IDDM (40 males, 40 females; age: 47 +/- 17 years; duration of diabetes: 20 +/- 15 years), had retinal photography performed to determine the association between the severity of retinopathy and prorenin values. Retinopathy was more severe in patients with retinopathy and albuminuria than in those with retinopathy alone (p < 0.002). When the prorenin values of patients with more marked retinopathy (eye grade greater than 3) were compared, prorenin values of those with retinopathy and albuminuria were greater than those of patients with retinopathy alone [269 (139-1406) versus 91 (41-273) ng/mL/h: geometric mean (range); p < 0.05]. Furthermore, when patients without albuminuria were considered, there was no significant difference between the prorenin levels of patients with more severe retinopathy (eye grade >3) when compared to patients with lesser degrees of retinopathy [91 (41-273) versus 69 (23-375). In patients with microvascular complications, prorenin values were independently predicted by albuminuria (p < 0.0001) and diastolic blood pressure (p < 0.02) but not the severity of retinopathy. In conclusion, prorenin values are significantly associated with the presence of microvascular complications in patients with IDDM. The association with albuminuria may be stronger than the association with retinopathy.

Lipoprotein(a) in health and disease

Lipoprotein(a) [Lp(a)] represents an LDL-like particle to which the Lp(a)-specific apolipoprotein(a) is linked via a disulfide bridge. It has gained considerable interest as a genetically determined risk factor for atherosclerotic vascular disease. Several studies have described a correlation between elevated Lp(a) plasma levels and coronary heart disease, stroke, and peripheral atherosclerosis. In healthy individuals, Lp(a) plasma concentrations are almost exclusively controlled by the apo(a) gene locus on chromosome 6q2.6-q2.7. More than 30 alleles at this highly polymorphic gene locus determine a size polymorphism of apo(a). There exists an inverse correlation between the size (molecular weight) of apo(a) isoforms and Lp(a) plasma concentrations. The standardization of Lp(a) quantification is still an unresolved task due to the large particle size of Lp(a), the presence of two different apoproteins [apoB and apo(a)], and the large size polymorphism of apo(a) and its homology with plasminogen. A working group sponsored by the IFCC is currently establishing a stable reference standard for Lp(a) as well as a reference method for quantitative analysis. Aside from genetic reasons, abnormal Lp(a) plasma concentrations are observed as secondary to various diseases. Lp(a) plasma levels are elevated over controls in patients with nephrotic syndrome and patients with end-stage renal disease. Following renal transplantation, Lp(a) concentrations decrease to values observed in controls matched for apo(a) type. Controversial data on Lp(a) in diabetes mellitus result mainly from insufficient sample sizes of numerous studies. Large studies and those including apo(a) phenotype analysis came to the conclusion that Lp(a) levels are not or only moderately elevated in insulin-dependent patients. In noninsulin-dependent diabetics, Lp(a) is not elevated. Conflicting data also exist from studies in patients with familial hypercholesterolemia. Several case-control studies reported elevated Lp(a) levels in those patients, suggesting a role of the LDL-receptor pathway for degradation of Lp(a). However, recent turnover studies rejected that concept. Moreover, family studies also revealed data arguing against an influence of the LDL receptor for Lp(a) concentrations. Several rare diseases or disorders, such as LCAT- and LPL-deficiency as well as liver diseases, are associated with low plasma levels or lack of Lp(a).

Levels of lipoprotein(a), apolipoprotein B, and lipoprotein cholesterol distribution in IDDM. Results from follow-up in the Diabetes Control and Complications Trial

Levels of lipoprotein(a) [Lp(a)], apolipoprotein (apo) B, and lipoprotein cholesterol distribution using density-gradient ultracentrifugation were measured as part of a cross-sectional study at the final follow-up examination (mean 6.2 years) in the Diabetes Control and Complications Trial. Compared with the subjects in the conventionally treated group (n = 680), those subjects receiving intensive diabetes therapy (n = 667) had a lower level of Lp(a) (Caucasian subjects only, median 10.7 vs 12.5 mg/dl, respectively; P = 0.03), lower apo B (mean 83 vs. 86 mg/dl, respectively; P = 0.01), and a more favorable distribution of cholesterol in the lipoprotein fractions as measured by density-gradient ultracentrifugation with less cholesterol in the very-low-density lipoprotein and the dense low-density lipoprotein fractions and greater cholesterol content of the more buoyant low-density lipoprotein. Compared with a nondiabetic Caucasian control group (n = 2,158), Lp(a) levels were not different in the intensive treatment group (median 9.6 vs. 10.7 mg/dl, respectively; NS) and higher in the conventional treatment group (9.6 vs. 12.5 mg/dl, respectively; P < 0.01). No effect of renal dysfunction as measured by increasing albuminuria or reduced creatinine clearance on Lp(a) levels could be demonstrated in the diabetic subjects. Prospective follow-up of these subjects will determine whether these favorable lipoprotein differences in the intensive treatment group persist and whether they influence the onset of atherosclerosis in insulin-dependent diabetes.

Lipoprotein (a) and vascular disease in diabetic patients

In order to assess the potential role of lipoprotein (a) as a risk factor for cardiovascular disease in diabetes mellitus, plasma concentrations were measured in a large group (n = 500) of non-insulin-dependent (NIDDM, n = 355) and insulin-dependent (IDDM, n = 145) patients. Concentrations of lipoprotein (a) were compared in diabetic patients with (n = 153) or without (347) documented vascular disease (ischaemic heart disease, peripheral vascular disease or macroangiopathy). They were significantly higher (p < 0.05) in patients with ischaemic heart disease (mean [interquartile range] 15.5 (5.0-38.0) vs 9.0 (4.5-26.0) mg/dl) or macroangiopathy (13.0 (5.0-38.0) vs 9.0 (4.0-25.0) mg/dl) compared to patients without manifestations of vascular disease. In addition, stepwise logistic regression analysis identified lipoprotein (a) levels > or = 30 mg/dl as being independently associated with the presence of cardiovascular disease. Lipoprotein (a) was an independent risk factor for ischaemic heart disease and macroangiopathy in this group of IDDM and NIDDM patients.

Longitudinal study of lipoprotein(a) in peripubertal children with insulin-dependent diabetes

We aimed to examine the longitudinal relationship between lipoprotein(a) and haemoglobin A1c, albumin excretion rate, and puberty in peripubertal children with insulin-dependent diabetes. A total of 114 patients aged 11.5 +/- 3.6 years (mean (SD)) were followed prospectively for 15.2 +/- 2.8 months. Lipoprotein(a), apolipoproteinB-100, haemoglobin A1c, mean overnight albumin excretion rate and Tanner stage were determined at the beginning and end of the study period. Lipoprotein(a) and apolipoproteinB-100 were measured using nephelometry. This method was correlated with radioimmunoassay and there was no significant change in mean bias during the study. Lipoprotein(a) fell significantly over time (214, (152, 276); 160 (84, 236) mg l-1 geometric mean (0.95 confidence intervals), p < 0.001); apolipoproteinB-100 did not change. Lipoprotein(a) and apolipoproteinB-100 did not differ in 233 cross-sectional controls of similar age. The change in lipoprotein(a) did not correlate with a small fall in haemoglobin A1c or with overnight albumin excretion rate, Tanner stage or insulin dose. Separate analysis of male and female patients and prepubertal and pubertal patients continued to show a significant fall in lipoprotein(a) independent of change in haemoglobin A1c or albumin excretion rate. Likewise, 53 patients with a change in haemoglobin A1c of greater than 1%, and 20 patients who progressed from normal albumin excretion rate to albumin excretion rate above the 95th centile, showed no relationship between lipoprotein(a) and haemoglobin A1c or albumin excretion rate. In conclusion, longitudinal changes in lipoprotein(a) do not relate to metabolic control or early changes in albuminuria in young patients with insulin-dependent diabetes.

The role of lipoprotein(a) in the vascular complications of diabetes mellitus

Lipoprotein(a) has been identified as an independent risk factor for atherosclerotic vascular disease in non-diabetic populations. Because of its potential role in the pathogenesis of both microvascular and macrovascular complications in diabetes, there have recently been many reports on lipoprotein(a) in diabetic populations. Some studies indicate an association between elevated lipoprotein(a) and macrovascular disease in non-insulin-dependent diabetes mellitus (NIDDM), but this link has not been found with insulin-dependent diabetes mellitus (IDDM). In IDDM, elevated lipoprotein(a) has been found in groups with diabetic nephropathy and retinopathy, raising the possibility that it plays a causative role. The relationship between glycaemic control and the lipoprotein(a) level has not been fully resolved. Most studies have not found any connection in NIDDM, but some found higher lipoprotein(a) levels in hyperglycaemic IDDM patients. Potentially, lipoprotein(a) is an important factor linking the microvascular and macrovascular complications of diabetes.

Lipoprotein(a) in type 1 diabetic patients with renal disease

Lp(a) was measured in 64 normoalbuminuric, 52 microalbuminuric, and 37 proteinuric Type 1 diabetic patients and 54 healthy subjects. Microalbuminuric and proteinuric Type 1 diabetic patients had higher median Lp(a) values (133 (16-1932) and 169 (17-1149) mg l-1) than patients with normal AER (73 (15-1078) mg l-1; p = 0.048 and p = 0.027). Lp(a) in healthy subjects (110 (15-1630)mg l-1) did not differ from the diabetic subgroups. The frequency of Lp(a) values in the upper quarter of the normal distribution was similar in the diabetic groups and did not differ between diabetic and control subjects. The cumulative distribution of Lp(a) was similar in all groups. Lp(a) concentrations were not related to AER, age, gender, duration of diabetes, body mass index, glycaemic control, serum creatinine, free insulin or systolic blood pressure. Cholesterol, LDL-cholesterol, triglycerides, and apo B were higher in microalbuminuric and proteinuric than in normoalbuminuric Type 1 diabetic patients. Lp(a) was independently related to diastolic blood pressure, fibrinogen, and macroangiopathy. In conclusion, median Lp(a) concentrations tend to be higher in Type 1 diabetic patients with early and established renal disease, although the differences are small and the overlap between groups large. Lp(a) is related to diastolic blood pressure and fibrinogen, and this association of powerful risk factors suggests that Lp(a) may play a role in the pathogenesis of cardiovascular disease in Type 1 diabetic patients with proteinuria. Whether Lp(a) is an independent determinant of increased cardiovascular risk in these patients needs to be elucidated by prospective studies.

The behaviour of lipoprotein(a) in patients with various diseases

The plasma Lp(a) concentrations were evaluated in several groups of patients. Groups with liver cirrhosis (n = 20), type-1 diabetes mellitus (n = 148), type-2 diabetes mellitus (n = 65), hypertension (n = 51), lung cancer (n = 48) and deep venous thrombosis (n = 31) were compared with a group of healthy volunteers (n = 69). Significantly higher median values were found in the hypertension (142 mgl-1 vs. 43 mgl-1, p < 0.001) and lung cancer groups (241 mgl-1 vs. 43 mgl-1; p < 0.0001). Significantly lower values were recorded in the group with liver cirrhosis (11 mgl-1 vs. 43 mgl-1; p = 0.02). But in this last group there were significant differences between patients in the Child-Turcotte severity stages A to C.

Microalbuminuria: prognostic and therapeutic implications in diabetes mellitus

Thirty years following the development of the first radioimmunoassay for albumin, microalbuminuria is widely acknowledged as an important predictor of overt nephropathy in patients with Type 1 diabetes and of cardiovascular mortality in Type 2 diabetes. In addition, there is accumulating evidence to suggest that diabetic patients with microalbuminuria may have more advanced retinopathy, higher blood pressure, and worse dyslipidaemia than patients with normal albumin excretion rates. Recent studies have focused on the role of intervention, principally with antihypertensive therapy and intensive glycaemic control, in reducing microalbuminuria. While successful in reducing urinary albumin excretion it remains to be established whether such therapies will be translated into a reduction in renal failure and decreased cardiovascular morbidity and mortality.

Lp(a) concentrations and phenotypes in children with insulin-dependent diabetes mellitus

Subjects with insulin-dependent diabetes mellitus (IDDM) have an increased incidence of coronary heart disease. Several studies have suggested that Lp(a) levels may be increased in IDDM subjects, although these studies have been limited by the lack of information on apo(a) phenotype and urinary albumin excretion. We compared Lp(a) concentrations in 66 children with IDDM and 18 non-diabetic children; all were non-Hispanic whites and none had detectable albuminuria. Lp(a) concentrations (mg/dl) were lower in subjects with IDDM than in non-diabetic subjects (12.0 +/- 2.2 vs. 20.0 +/- 6.1, respectively), although these means were not significantly different (P = 0.276). Postpubertal subjects, particularly males, had increased Lp(a) concentrations relative to prepubertal subjects (P = 0.041). Higher apo(a) molecular weight was associated with decreased Lp(a) concentrations in both diabetic and non-diabetic subjects. However, apo(a) size was not different in diabetic and non-diabetic subjects. Lp(a) concentrations were not significantly correlated with glycosylated hemoglobin levels in diabetic subjects (r = 0.11, P = NS). We also found similar Lp(a) concentrations in postpubertal IDDM subjects compared with adult non-Hispanic white non-diabetic subjects (n = 208) from the San Antonio Heart Study, a population-based study. These observations do not support increased Lp(a) concentrations in young normoalbuminuric IDDM subjects.

Lipoprotein(a) levels in relation to diabetic complications in patients with non-insulin-dependent diabetes

The relationship between serum levels of lipoprotein(a) Lp(a)) and the presence of chronic diabetic complications was studied in 194 patients with non-insulin-dependent diabetes mellitus (NIDDM; 75 males, 119 females; age 66 +/- 11 years; duration of diabetes, 11 (range 1-35) years). They were taking various treatments (diet alone, oral hypoglycaemic agents and/or insulin). Metabolic status and prevalence of diabetic complications were assessed by detailed history, physical examination, laboratory analysis and ECG. Average metabolic control was moderate (HbA1c 8.2 +/- 1.7%). Median serum Lp(a) level was 183 U l-1 (range 8-2600 U l-1), which was significantly higher than in control subjects of comparable age (median 101; range 8-1747 U l-1; P < 0.05), while HDL-cholesterol levels were lower (1.14 +/- 0.38 vs. 1.35 +/- 0.35 mmol l-1; P = 0.001), and total cholesterol levels were comparable. No significant relationships between diabetes treatment or metabolic control and Lp(a) levels were observed. In the quartile of patients with the highest Lp(a) levels, total cholesterol and triglycerides were slightly higher (P < 0.05), whereas HDL-cholesterol was not different. With increasing Lp(a) levels, higher prevalences of preproliferative retinopathy and of coronary artery disease (CAD) were observed, but not of the other complications. No relationship was found between the degree of albuminuria and Lp(a) levels. We conclude that in NIDDM patients, Lp(a) levels are elevated compared with non-diabetic subjects, and that higher Lp(a) levels are associated with higher prevalences of CAD and of retinopathy.

Improved blood glucose control by insulin therapy in type 2 diabetic patients has no effect on lipoprotein(a) levels

The effects of improved blood glucose control by insulin therapy on lipoprotein(a) and other lipoproteins were studied in 54 patients with Type 2 diabetes (mean +/- SD: age 67 +/- 9 years, body mass index 26.1 +/- 4.4 kg m-2, median duration of diabetes 10 (range 1-37) years, 23 males, 31 females), who were poorly controlled despite diet and maximal doses of oral hypoglycaemic agents. After 6 months of insulin treatment, mean fasting blood glucose concentrations had decreased from 14.1 +/- 2.2 mmol l-1 to 8.4 +/- 1.8 mmol l-1 (p < 0.001), and HbA1c had fallen from 11.1 +/- 1.4% to 8.2 +/- 1.1% (p < 0.001). Significant decreases of total and LDL cholesterol, triglycerides, apolipoprotein B, and free fatty acids were observed, while HDL-cholesterol and apoA1 increased by 10%. Baseline serum Lp(a) levels were elevated compared to non-diabetic subjects of similar age (median 283, range 8-3050 mg l-1, vs 101, range 8-1747 mg l-1, p < 0.05), but did not change with insulin, and there was no correlation with the degree of metabolic improvement and changes in Lp(a) levels. It is concluded that improved blood glucose control by insulin therapy does not alter elevated Lp(a) levels in Type 2 diabetic patients, but has favourable effects on the other lipoproteins.

Apolipoprotein(a) and cardiovascular disease in type 2 (non-insulin-dependent) diabetic patients with and without diabetic nephropathy

The relative mortality from cardiovascular disease is on average increased five-fold in Type 2 (non-insulin-dependent) diabetic patients with diabetic nephropathy compared to non-diabetic subjects. We assessed the possible contribution of dyslipidaemia in general and elevated serum apolipoprotein(a) (apo(a)) in particular. Type 2 diabetic patients with normo-, micro- and macroalbuminuria were compared with healthy subjects. Each group consisted of 37 subjects matched for age, sex and diabetes duration. Serum creatinine in the nephropathy group was 105 (54-740) mumol/l. The prevalence of ischaemic heart disease (resting ECG, Minnesota, Rating Scale) was 57, 35, 19 and 2% in macro-, micro- and normoalbuminuric diabetic patients and healthy subjects, respectively. The prevalence of ischaemic heart disease was higher in all diabetic groups as compared to healthy subjects (p < 0.05), and higher in macroalbuminuric as compared to normoalbuminuric diabetic patients (p < 0.01). There was no significant difference between apo(a) in the four groups: 161 (10-1370), 191 (10-2080), 147 (10-942), 102 (10-1440) U/l (median (range)) in macro-, micro- and normoalbuminuric groups and healthy subjects. Serum total-cholesterol, HDL-cholesterol and LDL-cholesterol were not significantly different when comparing healthy subjects and each diabetic group. Apolipoprotein A-I was lower (p < 0.05) in all diabetic groups as compared to healthy subjects (nephropathy vs healthy subjects): 1.50 +/- 0.25 vs 1.69 +/- 0.32 g/l (mean +/- SD). Triglyceride was higher (p < 0.05) in patients with nephropathy and microalbuminuria as compared to healthy subjects (nephropathy vs healthy subjects): 2.01 (0.66-14.7) vs 1.09 (0.41-2.75) mmol/l (median (range)).(ABSTRACT TRUNCATED AT 250 WORDS)