Homocysteine enhances LDL fatty acid peroxidation,promoting microalbuminuria in type 2 diabetes

Abstract

Background

We aimed to establish the relationship between glycated haemoglobin (HbA_1c), hypertension and microalbuminuria onset in type 2 diabetes. We also intended to ascertain the metabolic action of homocysteine on LDL fatty acids and on renal function.

Methods

The study was carried out on 200 patients with type 2 diabetes and 200 healthy subjects. HbA_1c, apolipoprotein B (apo B) and microalbuminuria were measured using immunoturbidimetric methods. Cholesterol, peroxide, urea and uric acid were assayed using colorimetric methods. Creatinine clearance was calculated using the Cockroft-Gault equation. Homocysteine was measured by immunological fluorescence polarization. LDL fatty acids were quantified by gas chromatography.

Results

Creatinine and microalbuminuria significantly increased in type 2 diabetes when compared with controls. Microalbuminuria was significantly correlated with HbA_1c and with the presence of high blood pressure. Homocysteinaemia significantly correlated with creatinine clearance in diabetes. Linoleic acid (C18:2ω6) did not differ between groups. C18:2ω6/C18:3ω3 ratio was three times higher in diabetics than in controls. Total saturated fatty acids, homocysteine, H₂O₂ and LDL-thiobarbituric reactive substances significantly increased in microalbuminuric when compared with normoalbuminuric diabetes. Total polyunsaturated fatty acids, arachidonic acid (C20:4ω6), LDL-cholesterol, apo B and creatinine clearance significantly decreased in microalbuminuric when compared with normoalbuminuric diabetes.

Conclusion

Microalbuminuria onset is associated with renal protein oxidation that is preceded by LDL fatty acid oxidation. The latter is initiated by H₂O₂ produced from an auto-oxidation of homocysteine and increased metabolism of arachidonic acid towards its pro-inflammatory eicosanoids. An oxidative stress state is the common ground of diffused vasculopathy.

Introduction

Diabetes mellitus is a well-known risk factor for micro and macroangiopathy. Increasing evidence is compelling for fatty acids having a major role in diabetes type 2 complications. Further to our previous studies,^1,2 we decided to quantify LDL fatty acids and LDL-cholesterol rather than plasma fatty acids. Consistent findings indicate that LDL fatty acid oxidation is the key symptom of atherosclerosis.³ Indeed, lipid dysregulation, hypercholesterolaemia and hyperhomocysteinaemia are associated with increased risk of coronary artery disease.⁴ Increased risk for cardiovascular disease is also associated with microalbuminuria.⁵ The latter reflects a diffused vasculopathy and an endothelial dysfunction.⁶ Endothelial dysfunction and impaired vasodilatory capacity were reported to occur in association with increased concentrations of homocysteine.⁷

The hypothesis that endothelial dysfunction is the common result of lipid dysregulation, hyperhomocysteinaemia and impaired vasodilatory capacity in type 2 diabetes provides a plausible mechanism linking microalbuminuria and LDL composition.

In this context, we aimed to establish the relationship between glycated haemoglobin (HbA_1c), blood pressure and microalbuminuria onset in diabetes. We also aimed to underline the metabolic action of homocysteine on LDL fatty acids and on renal function. This study is one of the large number of studies examining the microalbuminuria–homocysteine relationship in patients with diabetes and one of the few relating microalbuminuria to LDL fatty acid peroxidation. Elucidating these mechanisms is of great importance and may lead to knowledge about the pathophysiological mechanisms necessary for therapeutic innovations.

Materials and methods

Patient population

Two hundred type 2 diabetes patients (48% men and 52% women) with mean age of 55 years were admitted to the Endocrinology Unit of University Hospital Center of Sousse (Tunisia). Type 2 diabetes diagnosis was based on the World Health Organization criteria. The admission was restricted to patients having body mass index (BMI) <30 kg/m², HbA_1c <12%, absence of thyroid disease, renal deficiency, or pregnancy. Patients were recruited, with informed consent, for this study and were under observation.

Control population

Two hundred healthy controls (50% men and 50% women), whose mean age was 50 years and who were blood bank donors.

Clinical samples

Clinical samples were collected from venous blood into four different tubes (ethylenediaminetetraacetic acid, lithium heparin, fluoride oxalate and without anticoagulant) after fasting for 12 h.

HbA_1c, apolipoprotein B (apo B) and microalbuminuria were assayed using an immunoturbidimetric technique (Randox, Antrim, UK). Total cholesterol (T-chol), LDL-cholesterol (LDL-chol), urea and uric acid were evaluated using commercial kits (Randox). Creatinine was determined with a kinetic design using alkaline picrate (Jaffé method). Creatinine clearance was calculated using the Cockroft-Gault equation.⁸ Homocysteine was measured using an automated analyzer system (AxSYM Abbott, Wiesbaden, Germany). The measure was based on fluorescence polarization immunology assay. Peroxide (H₂O₂) determination was performed by PerOx kits (Immundiagnostik, Wiesenstr., Bensheim) based on colorimetric method. Absorbance was measured at 450 nm in a microtiter plate reader Σ960 Metertech (Metertech Inc., Taipei, Taiwan).

LDL-thiobarbituric reactive substances

LDL-thiobarbituric reactive substances (TBARS) were determined by Oxitek kits (Zeptometrix, Buffalo, NY, USA). LDL fractions were selectively precipitated by heparin and manganese chloride solution (60:1, v/v). The obtained pellet was suspended in saline solution and divided into two aliquots. The first one was designed for LDL-TBARS analysis and the second one for fatty acid quantification. Oxidized substances were evaluated by a colorimetric reaction with thiobarbituric acid at 90°C for 15 min. Fluorescence supernatants were read with excitation set at 530 nm and emission at 550 nm. TBARS were quantified by reference to a calibration curve of tetraethoxypropane, submitted to the same thiobarbituric acid colorimetric procedure.⁹

The second aliquot of LDL fraction was used to extract fatty acids. Fatty acid extraction was performed by chloroform/methanol (2:1, v/v) in the presence of 1 mL of butylated hydroxytoluene (BHT) in 5% methanol as an antioxidant. Heptadecanoic acid (C17) was the internal standard. Chloroformic phase was evaporated under nitrogen stream. Fatty acid methylation was achieved with toluene/methanol (2:3, v/v) and acetyl chloride (5% in methanol) at 95°C over 4 h. Methylated fatty acids were then extracted by hexane and evaporated. Fatty acid quantification was performed by gas chromatography HP5890 equipped with a flame ionization detector. Chromatography (HP5890 model) conditions were as follows: the column type was HP INNOWAX fused silica capillary (30 m, Ø: 0.32 nm with 0.5 μm film thickness). The gas vector was N₂ with flow rate of 1 mL/min. The injector temperature was 220°C, the detector temperature was 275°C and the column temperature was 180°C.

Statistical analyses

All statistical analyses were performed by SPSS for windows (version 10.0). Data of continuous variables are expressed as medians and interquartile ranges. Differences between groups were assessed by independent samples t-test. The criterion for significance was P < 0.05.

Results

Clinical data and cardiovascular risk factors of the study population are presented in Table 1. The diabetes and control populations were gender and age-matched. Fifty-one percent of diabetics presented with a family history of diabetes. Diabetes covered certain cardiovascular risk factors such as hypertension, dyslipidaemia and high BMI. Fourteen percent of diabetics were under insulin.

Table 1

Clinical characteristics of type 2 diabetes and healthy controls

	Controls (n = 200)	Diabetes (n = 200)
Age (years)	50 (7)	56 (11)
Diabetes duration (years)		11 (6)
Sex: men/women (%)	50/50	48/52
Smokers (%)		19
Body mass index (kg/m²)	24 (2)	27 (5)
Dyslipidaemia (%)		70
Hypertension (%)		18
Blood pressure (mmHg)
Systolic		140 (40)
Diastolic		80 (20)
Family history
Diabetes type 2 (%)		51
Blood hypertension (%)
Nephropathy (%)		22
Personal history
Macroangiopathy (%)		18
Retinopathy (%)		25
Neuropathy (%)		50
Hypoglycaemic agents
Metformin (%)		26
Sulphonylureas (%)		20
Metformin + Sulphonylureas (%)		40
Insulin (%)		14

Renal function results are shown in Table 2. Creatinine and microalbuminuria significantly increased in diabetes when compared with controls. Creatinine clearance decrease did not reach a significant difference between diabetics and controls. Forty percent of diabetics had microalbuminuria >30 mg/L, among whom 50% were hypertensive. Microalbuminuria was significantly correlated to HbA_1c (r = 0.6; P < 0.001) and to blood pressure (r = 0.7; P < 0.001). We found that creatinine clearance depended on homocysteine concentration (r = −0.5; P < 0.001). In fact, when homocysteinaemia varied between 4 and 10 μmol/L, creatinine clearance was normal and varied between 87 and 93 mL/min. When homocysteinaemia was between 10 and 14 μmol/L, creatinine clearance increased, varying between 95 and 111 mL/min. However, when homocysteinaemia exceeded 14 μmol/L, creatinine clearance decreased, varying between 65 and 70 mL/min.

Table 2

Renal function and homocysteine in type 2 diabetes and controls

	Controls (n = 200)	Diabetes (n = 200)
Urea (mmol/L)	6.0 (5.5–6.6)	6.7 (5.3–7.4)
Uric acid (μmol/L)	213 (199–221)	218 (166–298)
Creatinine (mmol/L)	76 (73–78)	81 (67–95)*
Creatinine clearance (mL/min)	91 (68–102)	86 (70–91)**
Microalbuminuria (mg/L)	20 (18–23)	37 (32–59)**
Homocysteine (μmol/L)	7.5 (7–7.8)	14 (10–14.5)**

*P < 0.05; **P < 0.001

We noted a significant increase in myristic acid (C14:0), palmitic acid (C16:0), stearic acid (C18:0) and total saturated fatty acids (ΣSFA) in diabetes when compared with controls (Table 3). Oleic acid (C18:1; ω9) was significantly lower in diabetics than in controls. Total polyunsaturated fatty acids (ΣPUFA) significantly decreased in diabetes when compared with controls. Eicosapentanoic acid (C20:5; ω3) and docosahexanoic acid (C22:6; ω3) also significantly decreased in diabetics when compared with controls. Linoleic acid (C18:2; ω6), the ω6 family precursor, did not differ between groups. But arachidonic acid (C20:4; ω6) significantly decreased in diabetes when compared with controls. The ratio of linoleic acid (C18:2; ω6)/linolenic acid (C18:3; ω3) was three times higher in diabetics than in controls.

Table 3

LDL fatty acid percentages in type 2 diabetes and controls

	Controls (n = 200)	Diabetes (n = 200)
C14:0	1.9 (0.9–2.7)	2.6 (1–5)**
C16:0	7.6 (7–9.3)	9.8 (5–12)**
C18:0	5.7 (5–6)	6 (5–7)**
C20:0	0.1 (0.1–0.2)	0.4 (0.3–1)**
C22:0	0.4 (0.3–0.5)	0.7 (0.4–2)**
C24:0	0.2 (0.1–0.2)	0.3 (0.2–0.6)**
ΣSFA	16 (13.7–19)	19.2 (12–27)***
C14:1	1.1 (0.5–2)	1.7 (0.6–2.7)**
C16:1 ω7	18.7 (16–20)	17.5 (14–20)**
C18:1 ω7	0.1 (0–0.2)	0.5 (0.3–2.7)**
C18:1 ω9	22.6 (20–26)	19.2 (15–23)**
C22:1	0.1 (0–0.2)	4.5 (3–10)**
C24:1	0.7 (0–2)	0.5 (0.4–1.5)
C20:1 ω9	0.1 (0–2)	0.1 (0–0.3)**
ΣMUFA	43.5 (36.5–49)	44.2 (33.5–60)**
C18:2 ω6	25.5 (23–27)	25.2 (22–29)
C20:2 ω6	0.1 (0–0.3)	0.2 (0.1–0.4)**
C20:3 ω6	0.3 (0.2–0.4)	0.3 (0.1–0.6)*
C20:4 ω6 AAr	0.6 (0.3–0.7)	0.3 (0.2–0.4)***
C22:4 ω6	4.5 (5–7)	3 (2–4)**
ΣPUFA ω6	31.5 (28.5–35.4)	29 (22.4–35)***
C18:3 ω3	1 (0.1–2.3)	0.3 (0.2–0.6)***
C20:5 ω3 EPA	4.5 (4–5)	3.5 (3–4.6)***
C22:5 ω3	0.2 (0.1–0.3)	0.3 (0.2–0.9)**
C22:6 ω3 DHA	3.3 (2.5–5.5)	2.9 (2.1–3.9)***
ΣPUFA ω3	9.1 (6–13.3)	7 (6–8.5)***
ΣPUFA	41 (35.5–47)	36 (28–46)***
C18:2 ω6/C18:3 ω3	25 (20–30)	84 (50–90)***

ΣSFA, total saturated fatty acids; ΣMUFA, total monounsaturated fatty acids; ΣPUFA, total polyunsaturated fatty acids; AAr, arachidonic acid; EPA, eicosapentanoic acid; DHA, docosahexanoic acid

*P < 0.5; **P < 0.01; ***P < 0.001

ΣSFA significantly increased in microalbuminuric diabetes when compared with normoalbuminuric diabetes (Table 4), whereas ΣPUFA and arachidonic acid (C20:4; ω6) were significantly decreased in microalbuminuric diabetes when compared with normoalbuminuric diabetes. LDL-chol and apo B significantly increased in microalbuminuric diabetes when compared with normoalbuminuric diabetes. Homocysteine, H₂O₂ and LDL-TBARS significantly increased in microalbuminuric diabetes when compared with normoalbuminuric diabetes. Creatinine clearance significantly decreased in microalbuminuric diabetes when compared with normoalbuminuric diabetes.

Table 4

LDL fatty acids, lipid profile, oxidant state and renal function in normoalbuminuric and microalbuminuric type 2 diabetes

	Controls (n = 200)	Normoalbuminuric (n = 120)^†	Microalbuminuric (n = 80)^‡
LDL fatty acids
ΣSFA	16 (13.7–19)	17.5 (15–21)**	20 (16–25)***
ΣMUFA	43.5 (36.5–49)	43 (37–45)**	44 (40–50)***
ΣPUFA ω6	31.5 (28–35.4)	35 (31–39)**	29 (20–30)
ΣPUFA ω3	9.1 (6–13.3)	8 (5–11)	8 (6–10)
ΣPUFA	41 (35.5–47)	41 (38–44)	37 (34–40)***
C20:4 ω6 AAr	0.6 (0.3–0.7)	0.4 (0.5–0.6)***	0.3 (0.2–0.4)***
Lipid profile
T-chol (mmol/L)	4.2 (3.9–4.5)	4.8 (4.5–5)***	5 (5–6)**
LDL-chol (mmol/L)	2.8 (2.7–2.9)	2.9 (2.5–3.7)**	3.2 (3–4)**
apo B (g/L)	0.8 (0.7–0.9)	0.9 (0.8–1)**	1 (1–1.3)**
Oxidant status
Homocysteine (μmol/L)	7.6 (7.5–7.7)	12 (11–12)***	14.5 (12.5–16.5)***
H₂O₂ (μmol/L)	131 (110–150)	250 (200–300)***	320 (280–400)***
LDL-TBARS	1.4 (1.3–1.5)	2.8 (2.2–3.4)*	3.3 (2.6–4)**
Renal function
Creatinine clearance (mL/min)	91 (62–102)	83 (70–95)	75 (60–86)**

Microalbuminuria >30 mg/L; ΣSFA, total saturated fatty acids; MUFA, monounsaturated fatty acids; ΣPUFA, total polyunsaturated fatty acids; AAr, arachidonic acid; chol, cholesterol; LDL-TBARS, low-density lipoprotein-thiobarbituric reactive substances; H₂O₂, hydrogen peroxide

*P < 0.5; **P < 0.01; ***P < 0.001

^†Statistical difference between normoalbuminuric and microalbuminuric diabetes; ^‡Statistical difference between controls and normoalbuminuric diabetes

Discussion

Microalbuminuria was detected in 40% of diabetics, whereas creatinine clearance was still normal. This translated a predisposition to develop renal dysfunction in microalbuminuric diabetes.¹⁰ This finding may also reflect a generalized cardiovascular disease settling down¹⁰, since 20% of our microalbuminuric diabetics presented with macroangiopathy. We found two complementary correlations in diabetes: the first was between microalbuminuria and HbA_1c; the second was between microalbuminuria and blood pressure. We concluded that chronic hyperglycaemia contributes to microalbuminuria drain onset via an endothelial tone alteration. Chronic hyperglycaemia leads to basement membrane protein glycation and contributes to capillary permeability increase. This disturbs the autoregulation of glomerular capillary pressure and permits a more important fraction of systemic pressure to penetrate the glomerular.¹¹

The relationship between hyperhomocysteinaemia and creatinine clearance indicates a renal function adaptation to a vasodilation caused by homocysteine increase up to 14 μmol/L. At high concentrations, homocysteine alters the balance between reactive oxygen species (ROS) production and their neutralization.¹² Such events decrease nitric oxide bioavailability and thus encourage the development of vasodilation.¹² However, when homocysteine was >14 μmol/L, renal function dropped. This decrease may be due to glomerular basal membrane damage caused by either ROS released from homocysteine auto-oxidation or protein homocysteine bound. In fact, at high concentrations, homocysteine may undergo complicated rearrangements to form homocysteine thiolactone. The latter is chemically reactive and acylates free amino groups of the basal membrane proteins,¹³ leading to progressive glomerular drainage.¹⁴ Indeed, uric acid increase observed in our diabetics may result in crystal formation leading to tubular obstruction.¹⁵

The significant increase of SFA in diabetes reveals that either they were consuming fatty diets despite the nutritionists' advices or they displayed lipid pathophysiology.¹⁶ In metabolic studies, different classes of SFA have different effects on plasma lipid and lipoprotein concentrations. Specifically, SFA with 12–16 carbon atoms tend to increase plasma total and LDL-chol. Myristic acid (C14:0), which was higher in diabetics than controls, increases thrombosis risk by inhibiting cholesterol metabolism¹⁷ or by stimulating secretion of factor VII.¹⁸ Indeed, palmitic acid (C16:0) was higher in diabetes than in controls. This may increase muscle cell insulin resistance.¹⁹ Stearic acid (C18:0) which was more important in diabetes than in controls does not have a cholesterol-raising effect, but may decrease HDL-chol and increase Lp(a) concentration.¹⁸

ΣPUFA decrease in diabetes can be attributed to LDL peroxidation due to an oxidant stress evidenced by an increased LDL-TBARS.^20,21 In both diabetics and controls, linoleic acid (C18:2; ω6)/alpha-linolenic acid (C18:3; ω3) ratio exceeded 5, the recommended value in the literature.²² The emphasis is on the ratio value, rather than their absolute amounts. The ratio value is critical for disease prevention. This is due to the competitive metabolism between these two essential PUFA for their entry into the elongation and desaturation pathways leading to the synthesis of their respective eicosanoids.

Eicosapentanoic acid (C20:5; ω3) and docosahexanoic acid (C22:6; ω3) significantly decreased in diabetics when compared with controls. This may be due to a synthesis inhibition owing to ω6 pathway accentuation in our diabetics. In fact, eicosapentanoic acid (C20:5; ω3) and docosahexanoic acid (C22:6; ω3) have an anti-inflammatory effect, since they competitively inhibit eicosanoid products delivered by arachidonic acid (C20:4; ω6).²³

We found that diabetes and controls have the same percentage of linoleic acid (C18:2; ω6), whereas arachidonic acid (C20:4; ω6) significantly decreased in diabetics when compared with controls. This decrease does not reflect any inflammatory protection. On the contrary, it reveals an increase in arachidonic acid (C20:4; ω6) metabolism to pro-inflammatory eicosanoids such as leucotrienes and prostaglandin synthesis. This metabolism includes two essential enzymes, lipo-oxygenase and cyclo-oxygenase, which are considered as a source of superoxide anion .^24,25 This finding constitutes a supplementary argument for the H₂O₂ increase found in our diabetics. Arachidonic acid (C20:4; ω6) decrease may also reflect its peroxidation, since it is the favoured site of ROS attack. The resulting products are endowed with powerful inflammatory properties.

PUFA ω6 decrease in microalbuminuric diabetes when compared with normoalbuminuric diabetes is due to arachidonic acid (C20:4; ω6) mobilization. Arachidonic acid (C20:4; ω6) decrease in micoalbuminuric diabetes may be due to its peroxidation by H₂O₂, whose concentration was nearly two-fold greater in microalbuminuric than in normoalbumunuric diabetes. LDL-TBARS rise in microalbuminuric diabetes reflected that lipid peroxidation reactions were undergone.

The high concentration of H₂O₂ in microalbuminuric diabetes may arise from homocysteine auto-oxidation. In fact, the oxidation of two homocysteine molecules yields the oxidized disulphide (homocysteine), two protons (H⁺) and two electrons (e⁻) while promoting the formation of ROS such as H₂O₂.¹³ Thus, the important concentration of H₂O₂ found in microalbuminuric diabetes may attack arachidonic acid (C20:4; ω6) and produce further peroxides. The resulting outcome may attack glomerular basal membrane proteins and subsequently charge–size selectivity alteration. First, these facts lead to an albumin reabsorption decrease, thus albumin would be drained in urine; and second, a decrease in glomerular creatinine clearance.

Homocysteine was significantly greater in microalbuminuric than normal albuminuric diabetes. This result may demonstrate that homocysteine has a causal role in albuminuria onset. In fact in high concentrations, homocysteine may combine with LDL apolipoproteins. Once structure of LDL particles are modified, they are sustained within the intima and subsequent inflammatory foam cell formation associated with atherogenesis.

SFA increase explains the hypercholesterolaemia found in microalbuminuric diabetes. Hypercholesterolaemia and hyperapolipoproteinaemia B reflect an increased LDL concentration in microalbuminuric diabetes. LDL accumulation in the presence of high concentrations of H₂O₂ may initiate PUFA-LDL peroxidation followed by protein oxidation of renal endothelium and mesangial cells. This results in glomerular charge–size selectivity alteration promoting microalbuminuria leakage. This finding partially explains why microalbuminuria is considered as a powerful predictor for generalized vascular disease. It is well documented that increasing glycosylation impairs homocysteine–albumin binding, leading to vascular damage in diabetes.²⁶ Wotherspoon et al. ²⁷ have reported the presence of an increased homocysteine concentration associated to a reduced antioxidant defence in microalbuminuric type 1 diabetes.

Our data suggest that chronic hyperglycaemia contributes to microalbuminuria drain via endothelial tone alteration. Renal function can withstand homocysteine increase to a certain threshold but it declines in the presence of high homocysteine concentration. An important novel finding of our study is that increased H₂O₂ produced by homocysteine and arachidonic acid oxidation may be the key role of glomerular basal membrane damage promoting albumin leakage.

We strongly thought that microalbuminuria is an ‘alarm alert’ of oxidant stress. The favoured target is LDL fatty acids, followed by further production of ROS. If these results are validated in future studies, early diagnosis and treatment of oxidant stress, especially H₂O₂ neutralization, may provide a novel way of preventing microalbumunuria.

References

Kassab

, Ferchichi , Amira

, Hammami

, Laradi

, Miled

. Plasma phospholipids fatty acids composition in Tunisian diabetes type 2. Arab J Clin Lab Invest 2004;1:69–78

Kassab-Chekir

, Ferchichi

, Kerkeni

, [Cholesterol ester fatty acid composition in Tunisian type 2 diabetics with and without cardiovascular complications]. Ann Biol Clin (Paris) 2004;62:555–62

Spiteller

. The relation of lipid peroxidation processes with atherogenesis: a new theory on atherogenesis. Mol Nutr Food Res 2005;49:999–1013

Kerkeni

, Addad

, Chauffert

, Hyperhomocysteinaemia, methylenetetrahydrofolate reductase polymorphism and risk of coronary artery disease. Ann Clin Biochem 2006;43(Pt 3):200–206

Tedesco

, Natale

, Mocerino

, Tassinario

, Calabro

. Renal resistive index and cardiovascular organ damage in a large population of hypertensive patients. J Hum Hypertens 2007;21:291–6

Helaly

, Sheashaa

, Hatata

, Youssef

, Hegazi

, Abdel-Aal

. Endothelial dysfunction in geriatric diabetic patients: the role of microalbuminuria in elderly type 2 diabetic patients? A randomized controlled study. Int Urol Nephrol 2006;39:333–8

Dayal

, Lentz

. ADMA and hyperhomocysteinemia. Vasc Med 2005;10(Suppl. 1):S27–33

Cockcroft

, Gault

. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41

Lefèvre

, Beljean-Leymarie

, Beyerle

. Evaluation of lipid peroxidation by measuring thiobarbaturic acid reactive substances. Ann Biol Clin (Paris) 1998;56:305–19

10.

Ochodnicky

, Vettoretti

, Henning

, Buikema

, Van Dokkum

, de Zeeuw

. Endothelial dysfunction in chronic kidney disease: determinant of susceptibility to end-organ damage and therapeutic response. J Nephrol 2006;19:246–58

11.

Bonnardel-Phu

, Wautier

, Vicaut

. Advanced glycation end products are involved in microvascular permeability changes observed in microcirculation of diabetic rats in vivo . J Mal Vasc 2000;25:122–7

12.

Garalienė

. The main determinants of endothelial dysfunction. Medicina (Kaunas) 2006;42:362–9

13.

Hayden

, Tyagi

. Homocysteine and reactive oxygen species in metabolic syndrome, type 2 diabetes mellitus, and atheroscleropathy: the pleiotropic effects of folate supplementation. Nutr J 2004;3:4–27

14.

Nakajima

, Hasegawa

, Kamiuchi

, Differential regulation of intracellular redox state by extracellular matrix proteins in glomerular mesangial cells: potential role in diabetic nephropathy. Redox Rep 2006;11:223–30

15.

Hess

. Pathophysiology, diagnosis and conservative therapy in calcium kidney calculi. Ther Umsch 2003;60:79–87

16.

Pankov

IuA

. Adiponectin, a new hormone: its role in the pathogenesis of diabetes mellitus. Vestn Ross Akad Med Nauk 2006;9:99–104

17.

Nagaya

, Nakaya

, Takahashi

, Yoshida

, Okamoto

. Relationships between serum saturated fatty acids and serum total cholesterol and HDL-cholesterol in humans. Ann Clin Biochem 1994;31(Pt 3):240–4

18.

Khor

. Dietary fat quality: a nutritional epidemiologist's view. Asia Pac J Clin Nutr 2004;13(suppl.):pS22

19.

Dimopoulos

, Watson

, Sakamoto

, Hundal

. Differential effects of palmitate and palmitoleate on insulin action and glucose utilization in rat L6 skeletal muscle cells. Biochem J 2006;399:473–81

20.

Ozdemir

, Ozden

, Maral

, Kuskay

, Cetinalp

, Tarkun

. Malondialdehyde, glutathione, glutathione peroxidase and homocysteine levels in type 2 diabetic patients with and without microalbuminuria. Ann Clin Biochem 2005;42(Pt 2):99–104

21.

Kassab

, Laradi

, Ferchichi

, Paramètres du stress oxydant dans le diabète de type 2. Immunoanal Biol Spéc 2003;18:79–85

22.

Crawford

. Commentary on the workshop statement. Essentiality of and recommended dietary intakes for Omega-6 and Omega-3 fatty acids. Prostaglandins Leukot Essent Fatty Acids 2000;63:131–4

23.

Delarue

. Acides gras polyinsaturés n-3: anti-catabolisme et prévention de l'insulino-résistance. Nutr Clin Métabol 2000;14:221–34

24.

Bonizzi

, Piette

, Merville

, Bours

. Cell type-specific role for reactive oxygen species in nuclear factor-kappaB activation by interleukin-1. Biochem Pharmacol 2000;59:7–11

25.

Van der Donk

, Tsai

, Kulmacz

. The cyclooxygenase reaction mechanism. Biochemistry 2000;52:15451–8

26.

Mathai

, Radford

, Holland

. Progressive glycosylation of albumin and its effect on the binding of homocysteine may be a key step in the pathogenesis of vascular damage in diabetes mellitus. Med Hypotheses 2007;69:166–72

27.

Wotherspoon

, Laight

, Browne

, Plasma homocysteine, oxidative stress and endothelial function in patients with type 1 diabetes mellitus and microalbuminuria. Diabet Med 2006;23:1350–56