Plasma homocysteine levels in patients with type 2 diabetes in a Mediterranean population: relation with nutritional and other factors

Nutrition, Metabolism & Cardiovascular Diseases (2005) 15, 109e117

www.elsevier.com/locate/nmcd

Plasma homocysteine levels in patients with type 2 diabetes in a Mediterranean population: relation with nutritional and other factors E. Diakoumopoulou a, N. Tentolouris a, E. Kirlaki a, D. Perrea b, E. Kitsou a, M. Psallas a, D. Doulgerakis a, N. Katsilambros a,* a

First Department of Propaedeutic Medicine, Athens University Medical School, Laiko Hospital, 5 Doryleou street, 115 21 Athens, Greece b Experimental Surgery Department, Athens University Medical School, 5 Doryleou street, 115 21 Athens, Greece Received 21 October 2003; accepted 16 January 2004

KEYWORDS Homocysteine; Diet; Vitamins; Lifestyle; Diabetes

Summary Background and Aim: Hyperhomocysteinemia is a major and independent risk factor for atherothrombotic vascular disease. It may be promoted by genetic factors, nutritional deficiencies of the vitamin cofactors required for homocysteine metabolism, and other modifiable factors. This cross-sectional study investigated the effect of dietary habits and lifestyle on plasma total homocysteine (tHcy) levels in patients with type 2 diabetes in a Mediterranean population. Methods and results: A total of 126 diabetic and 76 healthy subjects were interviewed using a food-frequency questionnaire. Information consisted of dietary and smoking habits, coffee and alcohol consumption and physical activity recording, during the month prior to enrolment. Measurements included blood pressure, body mass index (BMI), waist-to-hip ratio (WHR), plasma tHcy, folate, vitamin B12, lipids, HbA1c, creatinine, uric acid, and glomerular filtration rate (GFR). Plasma tHcy levels were not different between diabetic and control subjects (11.49G3.68 vs 12.67G3.79 mmol/l respectively, PZ0.40). Diabetic subjects had significantly higher plasma folate levels and consumed more fish, fruit and vegetables, in comparison with controls. Controls consumed more red meat, coffee, and alcohol. Multivariate analysis in diabetic subjects, after controlling for age, sex, systolic blood pressure, duration of diabetes, GFR, plasma uric acid levels, and the amount of the weekly consumption of fruit and vegetables, demonstrated that age,

* Corresponding author. Tel.: C30 210 745 6261; fax: C30 210 779 1839. E-mail address: [email protected] (N. Katsilambros). 0939-4753/$ - see front matter Ó 2005 Published by Elsevier Ltd. doi:10.1016/j.numecd.2004.01.001

110

E. Diakoumopoulou et al. GFR and the weekly amount of fruit and vegetable consumption were independently associated with plasma tHcy concentrations [regression coefficient (B)Z0.11, SE (B)Z0.03, PZ0.001, BZÿ0.07, SE (B)Z0.01, P!0.0001, and BZÿ0.05, SE (B)Z0.02, PZ0.04, respectively]. The weekly amount of coffee, alcohol and red meat consumption, and physical activity level were not related with plasma tHcy levels in either study group. Conclusions: 1) Plasma tHcy levels were not different in the diabetic group as compared to the control group. 2) In patients with type 2 diabetes age, GFR and the consumption of fruit and vegetables were strong and independent determinants of plasma tHcy levels. Ó 2005 Published by Elsevier Ltd.

Introduction Hyperhomocysteinaemia is positively associated with the risk of arteriosclerotic vascular disease [1]. According to recent meta-analyses there is evidence for causality between plasma total homocysteine (tHcy) and ischaemic heart disease, deep vein thrombosis, pulmonary embolism, and stroke [2,3]. Genetic factors [4e6] and nutritional deficiencies of folate or of the vitamin cofactors (vitamins B12, B6 and B2) involved in Hcy metabolism may also promote hyperhomocysteinemia. Among the main determinants of tHcy levels in non-diabetic subjects are age, sex, renal function, several diseases, drugs, coffee and chronic alcohol consumption, smoking and physical inactivity [7]. An altered vitamin status could play an important role in deranging plasma tHcy equilibrium. Furthermore, dietary habits and lifestyle factors have not been studied in diabetic subjects, so as to evaluate their influence on tHcy levels. In addition, literature data concerning plasma tHcy concentrations in diabetic and non-diabetic subjects are not unanimous [7]. The primary aim of this cross-sectional study was to evaluate the influence of dietary habits and lifestyle on plasma tHcy levels in subjects with type 2 diabetes in a Mediterranean population. A secondary aim was to examine the potential differences in tHcy levels between diabetic and control subjects and the relationships between plasma tHcy levels and sex, age, anthropometric parameters, duration and degree of hyperglycemia, plasma vitamin B12 and folate levels as well as insulin resistance, in these subjects.

Methods Subjects A total of 126 type 2 diabetic patients, consecutively attending the outpatient diabetes clinic of

our hospital, were included in the study. In addition, 76 healthy controls, matched for age, sex and smoking habits with the diabetic subjects participated in the study. Control subjects were hospital staff and their relatives or friends. The control and diabetic subjects who participated, met the following criteria: their age was 40 to 75 years, they had no history of nephropathy or microalbuminuria, hypothyroidism, liver disease, active infection, autoimmune diseases, stage III or IV heart failure according to NYHA criteria, uncontrolled hypertension, pregnancy, known vitamin B12 of folate deficiency, cancer or leukaemia, psoriasis, and alcohol abuse. In addition, patients on medications known to affect plasma tHcy levels (methotrexate, phenytoin, fibrates, oral contraceptives, hormonal replacement therapy, sulfonamides, L-dopa, theophylline, and vitamin supplementation) were also excluded [7]. An ad hoc committee of our hospital approved the study. The purpose of the study was clearly explained to all the study subjects, who then volunteered to participate. Informed consent was obtained from all participants. The demographic and clinical characteristics of the study subjects are shown in Table 1.

Data collection The dietary habits of the subjects during the previous month were documented using a food-frequency questionnaire, which focused on special foods (fruits and vegetables, red meat, fish), coffee and alcohol consumption [8]. The food-frequency questionnaire was based on the one used in the multicenter study of the Mediterranean group for the study of diabetes and has been validated as described previously [9]. The portion sizes were subsequently quantified using published data [10]. The doctors involved in the present study recorded all information provided by the diabetic and control subjects concerning their weekly consumption of the above foods (expressed as portions consumed per week), coffee, and alcohol.

Diet, lifestyle and homocysteine in diabetes

111

Table 1 Demographic and clinical characteristics of the study subjects n Males n (%) Females n (%) Age (years) BMI (Kg/m2) WHR Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Duration of diabetes (years)

Controls

Diabetics

P

76 39 (51.3) 37 (48.7) 55.8G13.6 26.7G3.8 0.88G0.09 123.5G14.7

126 67 (53.2) 59 (46.8) 58.4G9.1 29.4G4.8 0.92G0.08 135.4G18.9

0.79 0.42 0.12 0.14 0.01

77.6G7.5

82.7G9.5

0.03

e

Type of antidiabetic treatment Diet only e Tablets e Insulin e

9.1G8.0

27 (21.4) 68 (54.0) 31 (24.6)

Data are shown as meanGSD or as n (%). BMI: Body mass index; WHR: Waist-to-hip ratio.

Information about coffee consumption included the amount (number of cups per day) and type of coffee and was expressed as g of caffeine consumed per week. Alcohol consumption was classified according to the type and amount of alcohol (number of glasses consumed per day) and was expressed as g of alcohol consumed per week. Subjects were classified according to smoking habits as non-smokers, ex-smokers, and current smokers. Physical activity records were used to obtain an account of all physical activities performed by the subjects in the month prior to enrolment. Subjects were asked to describe the average type and duration of all physical activities performed in that period. Recording of physical activity was based on a 5-digit code obtained from the Compendium of Physical Activities [11]. The obtained data were expressed as metabolic equivalents (METs)-min per day (the product of the minutes for each activity times the MET intensity level). One MET-min is roughly equivalent to 1 Kcal/min for a 60-kg person [12].

Assays Blood was drawn early in the morning after a 8e10 fast for measurement of biochemical and other study parameters. Plasma lipids (total cholesterol, triglycerides, high-density lipoprotein cholesterol), creatinine and glucose were measured enzymatically on a Technicon RA-XT analyzer

(Technicon Ltd, Dublin, Ireland). Low-density lipoprotein cholesterol levels were estimated using the Friedewald equation [13]. Plasma tHcy was measured on plasma using a fluorescence polarization immunoassay (FPIA method) on a IMx analyzer (Abbott Diagnostics, c.v.Z4.3G0.8%). Plasma levels of vitamin B12 and folate were measured by radioimmunoassay (RIA) methods [SimulTRAC-SNB Radiassay Kit Vitamin B12 [57Co]/ Folate [125I], ICN Pharmaceuticals, Diagnostic Division, Orangeburg, New York, c.v.Z4.8G1.3% and 6.3G1.9%, respectively]. HbA1c levels were determined by ion-exchange HPLC with a non-diabetic reference range of 4.1e6.2%. Plasma insulin was measured by RIA (Biosure, Brussels, Belgium, c.v.Z3.3G1.2%). Insulin resistance was assessed by the homeostasis model assessment (HOMA) insulin resistance index [14]. GFR values were calculated by the Cockroft-Gault formula. Urine albumin and creatinine were measured in a first morning void urine sample. Urine albumin was measured by the RIA method (Albumin RIA; Pharmacia Diagnostics AB, Upsala, Sweden, c.v. !10%), while urine creatinine on Technicon RA-XT analyzer. Microalbuminuria was defined as albumin to creatinine ratio (O3.5 mg/mmol for women and O2.5 mg/ mmol for men). Hypertension was considered in all patients treated with antihypertensive drugs and/ or in subjects previously diagnosed with high blood pressure (BP) levels (systolic BP O140 and diastolic BP O90 mmHg).

Statistical analysis Statistical analysis was performed using the SPSS program (SPSS, Chicago, IL, USA). All variables were tested for normal distribution of the data. Because the plasma values of folate and vitamin B12 were skewed, they were log-transformed to improve normality for statistical testing and back-transformed for presentation in the tables. A two-sample t-test or a Wilcoxon test was used to compare differences between non-diabetic and diabetic subjects. A chi-square test was used for categorical variables. Univariate linear regression analysis was applied to examine the relationship between plasma tHcy levels and the variables of interest in the control and the diabetic group. Then, multivariate linear regression analyses were performed to look for independent associations between plasma tHcy levels and the variables that were found to have a statistical association (P!0.05) or a trend for an association (P!0.20) with the plasma tHcy concentrations in univariate analysis and included: in the control group, age,

112

E. Diakoumopoulou et al.

sex, GFR, amount of the weekly consumption of red meat, fish and coffee as well as physical activity; in the diabetic group, age, sex, systolic blood pressure, GFR, duration of diabetes, plasma uric acid levels, and the amount of the weekly consumption of fruit and vegetables. All independent variables in the multivariate analyses models were tested for multicollinearity. P values !0.05 were considered as statistically significant.

Results Blood pressure, plasma glucose, HbA1c, triglyceride and folate levels were all significantly higher in diabetic subjects as compared to controls, while plasma HDL cholesterol concentrations were lower in the diabetic group (Tables 1 and 2). Plasma levels of tHcy were not different between diabetic and non-diabetic subjects (PZ 0.40) (Table 2). Non-diabetic subjects consumed more red meat (PZ0.02) and coffee (PZ0.02) per week in comparison to diabetic subjects. On the contrary, diabetic subjects consumed more fish (PZ0.01), fruit and vegetables (P!0.0001) per week. Physical activity level (PZ0.96) and alcohol consumption (PZ0.35) was not different between the two groups (Table 3). In addition, plasma tHcy concentrations were higher in men than in women (13.4G4.0 vs 10.6G3.7 mmol/l, respectively, PZ0.03). Similarly, men with diabetes had higher plasma tHcy levels in comparison with the women with diabetes (12.3G4.2 vs 10.7G3.1 mmol/l, respectively, PZ0.02).

Table 2

Univariate linear regression analysis in non-diabetic subjects showed that plasma tHcy levels were significantly associated with male sex (PZ0.04), portions of red meat consumed per week (PZ0.005) and GFR (PZ0.01). In addition, there was a trend for association with age (PZ0.09), portions of fish and coffee consumed per week (PZ0.10 and PZ0.20, respectively), and with physical activity (PZ0.10) (Table 4). Blood pressure, BMI, WHR, smoking status, insulin resistance index, serum albumin, and plasma folate as well as vitamin B12 levels were not associated significantly (PO0.05) with the outcome variable. The same analysis in diabetic subjects showed significant relationships between plasma tHcy levels and age (PZ0.02), male sex (PZ0.02), systolic blood pressure (PZ0.03), plasma uric acid (PZ0.001), GFR (PZ0.02) and the amount of the weekly consumption of fruit and vegetables (PZ0.02). Furthermore, there was a trend for an association with duration of diabetes (PZ0.10) (Table 4). Diastolic blood pressure, BMI, WHR, duration of diabetes, HBA1c, smoking status, insulin resistance, serum albumin, and plasma folate as well as vitamin B12 levels were not associated significantly (PO0.05) with plasma tHcy concentrations. Multivariate linear regression analysis in the nondiabetic group, after controlling for age, sex, GFR, the amount of the weekly consumption of red meat, fish and coffee, and physical activity, showed that only GFR was independently and significantly associated with plasma tHcy levels [regression coefficient (B)Zÿ0.002, SE(B)Z0.001, PZ0.03]. The same analysis in the diabetic group, after controlling for age, sex, systolic blood pressure, duration of diabetes, plasma uric acid levels, GFR and the

Profile of biochemical parameters

Fasting glucose (mmol/l) HBA1c (%) Total cholesterol (mmol/l) HDL cholesterol (mmol/l) LDL cholesterol (mmol/l) Triglyceride (mmol/l) Uric acid (mmol/l) Creatinine (mmol/l) GFR (ml/min) Folate (nmol/l) Vitamin B12 (pmol/l) Albumin (g/l) Homocysteine (mmol/l)

Controls

Diabetics

5.06G0.92 5.4G0.3 5.64G1.05 1.18G0.28 3.88G1.09 1.35G0.80 0.24G0.06 82.2G12.4 107.9G33.5 14.5G6.0 320.6G135.9 52.3G2.1 12.67G3.79

9.53G3.22 7.6G1.8 5.52G1.03 1.12G0.55 3.72G0.92 1.61G0.92 0.24G0.07 75.1G15.0 102.6G30.4 19.4G14.6 311.9G145.3 53.5G2.9 11.49G3.68

!0.0001 !0.0001 0.81 0.04 0.27 0.008 0.99 0.60 0.30 0.01 0.39 0.59 0.40

Data are shown as meanGSD. BMI: Body mass index; WHR: Waist-to-hip ratio; HDL: High density lipoprotein; LDL: Low density lipoprotein. GFR: glomerular filtration rate.

Diet, lifestyle and homocysteine in diabetes Table 3

113

Nutritional and lifestyle factors Controls

Consumption of: Red meat (portions per week) Fish (portions per week) Fruit and vegetables (portions per week) Coffee (g of caffeine per week) Alcohol (g per week) Smoking status n (%) Current smokers Ex-smokers Non-smokers MET-min/day

3.5 1.0 14.0 1.34 45.0 19 26 31 291.4

Diabetics

(2.0, 5.0) (1.0, 1.1) (8.3, 28.1) (0.64, 1.23) (0.0, 94.3)

3.0 2.0 28.2 1.17 30.0

(25.0) (34.2) (40.8) (117.9, 711.4)

32 34 60 323.6

P

(2.0, 4.0) (1.0, 2.0) (21.1, 35.4) (0.64, 1.29) (0.0, 105.0)

0.02 0.01 !0.0001 0.02 0.35

(25.4) (27.0) (47.6) (106.1, 739.3)

0.51 0.96

Data are shown as median values (interquartile range, IQR) or as n (%). MET: metabolic equivalent.

amount of the weekly consumption of fruit and vegetables, demonstrated that age, GFR and the weekly consumption of fruits and vegetables were independently and significantly associated with plasma tHcy concentrations [BZ0.11, SE(B)Z0.03, PZ0.001, BZÿ0.07, SE(B)Z0.01, P!0.0001, and BZÿ0.05, SE(B)Z0.02, PZ0.04, respectively]. Because treatment with metformin may reduce plasma vitamin B12 levels, and consequently may rise plasma tHcy levels, a sub-analysis was performed between diabetic patients treated (nZ37) and not-treated (nZ69) with metformin. This analysis showed that patients treated with metformin had lower plasma levels of vitamin B12 and folate than patients not treated with metformin (293.5G152.7 vs 396.2G367.5 pmol/l and 14.5 G6.5 vs 19.6G9.7 nmol/l, PZ0.04 and PZ0.03, respectively). However, plasma concentrations of tHcy were not different between the two groups (11.3G3.5 vs 11.6G3.8 mmol/l, PZ0.87).

Discussion The consecutive selection of the diabetic patients from the outpatient diabetes clinic ensures that no selection bias has been introduced. However, a selection bias cannot be excluded in the nondiabetic subjects, although they were matched for age, sex and smoking habits with their diabetic counterparts. Diabetic subjects constitute a patient population, at high risk for cardiovascular disease, due to the influence of a clustering of risk factors. Plasma tHcy is considered as an emerging independent nontraditional risk factor for atherosclerotic vascular disease, which may enhance the effect of the traditional risk factors [15,16]. It is also a strong predictor of cardiovascular and all-cause mortality [17]. Therefore, it is important to know if dietary habits and lifestyle can affect plasma tHcy levels in this population, and eventually select the patients

Table 4 Univariate linear regression analysis: the potential effect of various confounding factors on plasma homocysteine concentrations in the control and diabetic subjects Controls Age (1year) Male vs female sex SBP (mmHg) GFR (1 ml/min) Duration of diabetes (1 year) Uric acid (mmol/l) Weekly consumption of Red meat (1 portion) Fish (1 portion) Coffee (100 mg) Alcohol (25 g) Fruit and vegetables (1 portion)

Subjects with diabetes

B

SE (B)

P

B

SE (B)

P

0.002 0.08 0.01 ÿ0.002 e 14.59

0.001 0.04 0.75 0.001 e 19.50

0.09 0.04 0.74 0.01 e 0.46

0.08 1.48 0.04 ÿ0.02 0.06 20.47

0.03 0.66 0.01 0.01 0.04 4.66

0.02 0.02 0.03 0.02 0.13 !0.0001

0.03 ÿ0.04 0.16 0.19 ÿ0.001

0.01 0.03 0.12 0.17 0.002

0.005 0.10 0.20 0.27 0.37

0.12 0.09 0.007 0.08 ÿ0.02

0.24 0.05 0.04 0.09 0.01

0.62 0.53 0.88 0.39 0.02

B: regression coefficient; SBP: Systolic blood pressure; DBP: GFR: Glomerular filtration rate; PA: Physical activity. Data in parenthesis shows the incremental unit of the independent variable.

114 who would be at a higher risk for developing such complications.

Dietary habits A novel finding of this study was the significant inverse relationship between plasma tHcy levels and the weekly consumption of fruits and vegetables in the diabetic group, which was independent of the plasma concentrations of folate and vitamin B12. This relationship might be due to the presence of other factors or antioxidants, which are important components in a diet rich in fruit and vegetables such as a-tocopherol, beta-carotene, vitamin C and vitamin B6, which, according to recent reports [18], may exert a lowering effect on plasma tHcy concentrations. Furthermore, this inverse relationship between plasma tHcy and the high intake of fruits and vegetables was also reported by Fung et al. and Lasheras et al. in nondiabetic individuals. Both studies reported that a prudent dietary pattern, characterized by higher intakes of fruit and vegetables, was associated with higher plasma folate levels and inversely correlated with lower homocysteine plasma concentrations [19,20].

Lifestyle Literature data concerning the effect of physical activity on plasma tHcy levels are limited. The Hordaland Homocysteine Study [21] showed an inverse relationship between plasma tHcy levels and physical activity in middle-aged non-diabetic men and women who regularly engaged in moderate physical activity. However, in the present study no significant association was found between physical activity and plasma tHcy levels in either study group. A significant positive relationship between plasma tHcy and smoking has been described [21e24], which was attributed to the lower plasma levels of folate, vitamin B6 and B12 in smokers. Smoking probably increases plasma tHcy mainly by impairing vitamin status, both because of a lower intake of fruit and vegetables [25] and by increased vitamin requirement. In addition, consumption of large, but not of moderate amounts of coffee, has been associated with high plasma tHcy levels in healthy individuals [25e27]. Several mechanisms have been proposed: caffeine, which is partly responsible for the Hcy raising effect of coffee [23], together with chlorgenic acid which raises Hcy several hours after consumption [28]. Grubben et al. (2000), speculated that this effect is due to the cholesterol raising

E. Diakoumopoulou et al. diterpene alcohols, cafestol and kahweol present exclusively in unfiltered coffee [26]. However, we did not find any significant effect of smoking or coffee consumption on plasma tHcy levels in either study group. It is speculated that the relatively high intake of fruits and vegetables in our Mediterranean population might overcome the possible increase of plasma tHcy induced by smoking. In addition, the diabetic patients in our study consumed rather small amounts of coffee. Alcohol consumption is probably associated with tHcy concentration in a J-shaped fashion [29]; moderate alcohol consumers have a lower tHcy concentration compared with nondrinkers, whereas alcoholics have elevated tHcy concentrations [7]. Some investigators have reported a weak positive association between plasma tHcy with moderate alcohol intake [23]. Chronic alcohol consumption decreases the intestinal absorption of folate and may increase the folate requirement [30]. Nevertheless, we did not find any association between the alcohol consumption and plasma tHcy levels in the groups studied.

Biochemical factors Previous studies have shown an inverse association between either higher plasma folate or vitamin B12 and tHcy plasma concentrations in diabetic [31,32] as well as in non-diabetic subjects [22,23,33]. In this study we did not find a significant relationship between plasma folate and tHcy levels in either study group nor any sex differences in plasma folate levels in either diabetic and nondiabetic. This may be due to the fact that plasma folate primarily reflects the recent dietary intake, and measurement of the red cell folate content is a better index of the body stores in folate than plasma folate [34]. Homocysteine is metabolized in the renal tubular ephithelium, and this explains why plasma tHcy levels rise even with the modest deterioration of the renal function [35]. GFR is an independent determinant of plasma tHcy concentrations; its decline explains the age-related increase in plasma tHcy and hyperfiltration explains the lower than normal mean plasma tHcy in some populations of diabetic patients [24,31]. Other investigators also report that GFR is an independent predictor of plasma tHcy and inversely correlated with tHcy levels in diabetic subjects [36,37]. High plasma tHcy concentrations are associated with increased albumin excretion, especially in patients with type 2 diabetes [35,38]. Most data from these studies suggest that microalbuminuric type 2 diabetic

Diet, lifestyle and homocysteine in diabetes subjects have higher tHcy levels than normoalbuminurics, but the relationship between tHcy and albumin excretion rate is mainly due to changes in renal function. In our study, patients with microalbuminuria were excluded in order to avoid its confounding effect on tHcy levels. In accordance with the above findings, we found that plasma tHcy levels were independently and significantly associated with GFR, in diabetic as well as in nondiabetic subjects.

Demographic factors Plasma tHcy levels have been studied extensively in diabetic as well as in non-diabetic subjects. Some investigators demonstrated a high prevalence of hyperhomocysteinaemia in subsets of diabetic patients compared to control subjects [39,40], while others report lower levels of tHcy in diabetics versus controls [24,31,32]. However, our findings are in accordance with a number of studies which did not find any differences in plasma tHcy values between diabetic and control subjects [32,38,41]. In this study we found a positive association between plasma tHcy levels and male sex only in univariate analysis in both study groups. Other studies have also described a positive association between plasma tHcy concentrations and male sex [21e24,32,33]. It is noteworthy that we did not find any sex differences in either plasma folate or vitamin B12 levels. Furthermore, no significant difference was found in the weekly consumption of fruit and vegetables between diabetic men and women. Salardi et al. [32], found that diabetic patients had lower plasma tHcy values than nondiabetic controls, but this difference was present only in females. Furthermore, the females showed higher folate levels than males only in the diabetic group. So, in contrast with the findings of Salardi et al, the gender difference in plasma tHcy levels in our study cannot be attributed to differences in plasma folate or in vitamin B12 levels. However, this difference may be due to the higher fat-free mass in men and to the high estradiol concentrations in women, factors which are known to influence plasma tHcy concentrations [33]. The findings of this study confirm the known positive association between plasma tHcy levels and age in the diabetic subjects [21,31,32]. Possible mechanisms include vitamin malabsorption due to atrophic gastritis [42] or inadequate nutrition with a low content of folate and vitamins B6 and B12, which are common in the elderly, as well as the decline of renal function, which is a natural phenomenon for this age population [31].

115

Medications Regarding the effect of treatment with metformin on plasma tHcy concentrations, we found that although this treatment was associated with lower plasma vitamin B12 and folate levels, it had no significant effect on plasma tHcy levels. Other authors have reported an increase in plasma tHcy in type 2 diabetic patients treated with metformin [43,44]. Hoogeveen et al. (1997) however, reported that this effect is minimal [44] and others found e as we did e no significant difference between subjects receiving and not receiving metformin [36]. In conclusion, our data confirm the findings of other studies which do not show any difference in plasma tHcy levels between type 2 diabetic and non-diabetic subjects. In addition, in patients with type 2 diabetes, age, renal function, as defined by GFR, and the weekly consumption of fruit and vegetables are important determinants of plasma tHcy levels. This last finding has practical implications since it emphasizes the need for a relatively high intake of fruit and vegetables in the diabetic population. Furthermore, it was shown that metformin treatment is not associated with any ‘‘harmful’’ effect on plasma tHcy levels.

Acknowledgments This work was supported by a grant from the University of Athens.

References [1] Boushey CJ, Beresford SAA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease: probable benefits of increasing folic acid intakes. J Am Med Assoc 1995;274:1049e57. [2] Wald DS, Law M, Morris JK. Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. Br Med J 2002;325:1202. [3] Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke. J Am Med Assoc 2002;288:2015e22. [4] Mudd SH, Skovby F, Levy HL, Pettigrew KD, Wilcken B, Pyeritz RE, et al. The natural history of homocystinouria due to cystathionine b-synthase deficiency. Am J Hum Genet 1985;37:1e31. [5] Frosst P, Blom HJ, Milos R, Goyette P, Sheppard CA, Matthews RG, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 1995;10:111e3. [6] Klerk M, Verhoef P, Clarke R, Blom HJ, Schouten EG; MTHFR Studies Collaboration Group. MTHFR 677C/T polymorphism and risk of coronary heart disease: a metaanalysis. J Am Med Assoc 2002;288:2023e31

116 [7] De Bree A, Verschuren WM, Kromhout D, Kluijtmans LA, Blom HJ. Homocysteine determinants and the evidence to what extent homocysteine determines the risk of coronary heart disease. Pharmacol Rev 2002;54:599e618. [8] Levine J, Morgan M. Assessment of dietary intake in man: a review of available methods. J Nutr Med 1991;2:65e81. [9] Thanopoulou AC, Karamanos BG, Angelico FV, AssaadKhalil SH, Barbato AF, Del Ben MP, et al. Dietary fat intake as risk factor for the development of diabetes: multinational, multicenter study of the Mediterranean Group for the Study of Diabetes (MGSD). Diabetes Care 2003;26: 302e7. [10] Crawley H. Food Portion Sizes. 2nd ed. London: HMSO; 1993. [11] Ainsworth BE, Haskell WL, Leon AS, Jacobs DR, Montoye HJ, Sallis JF, et al. Compendium of Physical Activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 1993;25:71e80. [12] Taylor H, Jacobs D, Schucker B. A questionnaire for the assessment of leisure time physical activities. J Chronic Dis 1978;31:741e55. [13] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499e502. [14] Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and b-cell function from fasting plasma glucose and insulin concentration in man. Diabetologia 1985;28:412e9. [15] Graham IM, Daly LE, Refsum HM, Robinson K, Brattstro ¨m LE, Ueland PM, et al. Plasma homocysteine as a risk factor for vascular disease. The European Concerted Action Project. J Am Med Assoc 1997;277(22):1775e81. [16] Hackam DG, Anand SS. Emerging risk factors for atherosclerotic vascular disease. J Am Med Assoc 2003;290: 932e40. [17] Bostom AG, Silberhatz H, Rosenberg IH, Selhub J, D’Agostino RB, Wolf PA, et al. Non-fasting plasma total homocysteine levels and all-cause and cardiovascular disease mortality in elderly Framingham men and women. Arch Intern Med 1999;159:1077e80. [18] Steptoe A, Perkins-Porras L, McKay C, Rink E, Hilton S, Cappucio FP. Behavioral counselling to increase consumption of fruit and vegetables in low income adults: randomized trial. Br Med J 2003;326:855e60. [19] Fung TT, Rimm EB, Spiegelman D, Rifai N, Tofler GH, Willett WC, et al. Associations between dietary patterns and plasma biomarkers of obesity and cardiovascular disease risk. Am J Clin Nutr 2001;73:61e7. [20] Lasheras C, Huerta JM, Gonza ´lez S, Prada M, Braga S, Ferna ´ndez S, Patterson AM. A diet score is associated with plasma homocysteine in a healthy institutionalized elderlypopulation. Nutr Metab Cardiovasc Dis 2003;16:384e90. [21] Nyga ˚rd O, Vollset SE, Refsum H, Stensvold I, Tverdal A, Nordehaug JE, et al. Total plasma homocysteine and cardiovascular risk profile: the Hordaland Homocysteine Study. J Am Med Assoc 1995;274:1526e33. [22] Nyga ˚rd O, Refsum H, Ueland PM, Vollset SE. Major lifestyle determinants of plasma total homocysteine distribution: the Hordaland Homocysteine Study. Am J Clin Nutr 1998;67: 263e70. [23] Jacques PF, Bostom AG, Wilson PW, Rich S, Rosenberg IH, Selhub J. Determinants of plasma total homocysteine concentration in the Framingham Offspring cohort. Am J Clin Nutr 2001;73:613e21. [24] Matteucci E, Rossi L, Mariani S, Fagnani F, Quilici V, Cinapri V, et al. Blood levels of total homocysteine in

E. Diakoumopoulou et al.

[25]

[26]

[27]

[28]

[29] [30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

patients with type 1 diabetes (with no complications, diabetic nephropathy and/or retinopathy) and in their non-diabetic relatives. Nutr Metab Cardiovasc Dis 2002;12: 184e9. Nyga ˚rd O, Refsum H, Ueland PM, Stensvold I, Nordehaug JE, Kva ˚le G, et al. Coffee consumption and plasma total homocysteine: the Hordaland Homocysteine Study. Am J Clin Nutr 1997;65:136e43. Grubben MJ, Boers GH, Blom HJ, Broekhuizen R, de Jong R, van Rijt L, et al. Unfiltered coffee increases plasma homocysteine concentrations in healthy volunteers: a randomized trial. Am J Clin Nutr 2000;71:480e4. Urgert R, van Vliet T, Zock PL, Katan MB. Heavy coffee consumption and plasma homocysteine: a randomized controlled trail in healthy volunteers. Am J Clin Nutr 2000;72:1107e10. Verhoef P, Pasman WJ, van Vliet T, Urgert R, Katan MB. Contribution of caffeine to the homocysteine-raising effect of coffee: a randomized controlled trial in humans. Am J Clin Nutr 2002;76:1244e8. Halsted CH. Lifestyle effects on homocysteine and an alcohol paradox. Am J Clin Nutr 2001;73:501e2. Cravo ML, Camilo ME. Hyperhomocysteinemia in chronic alcoholism: relations to folic acid and vitamins B(6) and B(12) status. Nutrition 2000;16:296e302. Wollesen F, Brattsro ¨m L, Refsum H, Ueland PM, Berglund L, Berne C. Plasma total homocysteine and cysteine in relation to glomerular filtration rate in diabetes mellitus. Kidney Int 1999;55:1028e35. Salardi S, Cacciari E, Sassi S, Grossi G, Mainetti B, Dalla Casa C, et al. Homocysteinemia, serum folate and vitamin B12 in very young patients with diabetes mellitus type 1. J Pediatr Endocrinol Metab 2000;13:1621e7. Dierkes J, Jeckel A, Ambrosch A, Westphal S, Luley C, Boeing H. Factors explaining the difference of total homocysteine between men and women in the European Investigation into Cancer and Nutrition Potsdam Study. Metabolism 2001;50:640e5. Dunphy DM. Evaluation of the red cell folate activity as an assessment of body stores of folic acid. Can J Med Technol 1972;34:142e54. Smulders YM, Raki M, Slaats EH, Treskes M, Sijbrands EJ, Oderkerken DA, et al. Fasting and post-methionine homocysteine levels in NIDDM. Determinants and correlations with retinopathy, albuminuria, and cardiovascular disease. Diabetes Care 1999;22:125e32. Buysschaert M, Dramais AS, Wallemacq PE, Hermans MP. Hyperhomocysteinemia in type 2 diabetes: relationship to macroangiopathy, nephropathy, and insulin resistance. Diabetes Care 2000;23:1816e22. Davies L, Wilmshurst EG, McElduff A, Gunton J, Cliffton-Blich P, Fulcher GR. The relationship among homocysteine, creatinine clearance, and albuminuria in patients with type 2 diabetes. Diabetes Care 2001;24: 1805e9. Lanfrendini M, Fiorina P, Peca MG, Veronelli A, Mello A, Astorri E, et al. Fasting and post-methionine load homocyst(e)ine values are correlated with microalbuminuria and could contribute to worsening vascular damage in non-insulin-dependent diabetes mellitus patients. Metabolism 1998;47:915e21. Passaro A, D’Ellia K, Pareschi PL, Calzoni F, Carantoni M, Fellin R, et al. Factors influencing plasma homocysteine levels in type 2 diabetes. Diabetes Care 2000;23:420e1. Yeromenko Y, Lavie L, Levy Y. Homocysteine and cardiovascular risk in patients with diabetes mellitus. Nutr Metab Cardiovasc Dis 2000;11:108e16.

Diet, lifestyle and homocysteine in diabetes [41] Pavia C, Ferrer I, Valls C, Artuch R, Colome ´ C, Vilaseca MA. Total homocysteine in patients with type 1 diabetes. Diabetes Care 2000;23:84e7. [42] van Asselt DZ, de Groot LC, van Staveren WA, Blom HJ, Wevers RA, Biemond I, et al. Role of cobalamin intake and atrophic gastritis in mild cobalamin deficiency in older Dutch subjects. Am J Clin Nutr 1998;68: 328e34.

117 [43] Carlsen SM, Folling I, Grill V, Bjerve KS, Schneede J, Refsum H. Metformin increases total serum homocysteine levels in non-diabetic male patients with coronary heart disease. Scand J Lab Invest 1997;57:521e7. [44] Hoogeveen EK, Kostense PJ, Jakobs C, Bouter LM, Heine RJ, Stehouwer CD. Does metformin increase the serum total homocysteine in non-insulin-dependent diabetes mellitus? J Intern Med 1997;242:389e94.