Professional Documents
Culture Documents
A B S T R A C T
Evidence shows that there is a rapid increase in the production of markers of oxidative damage
immediately following acute stroke and that endogenous antioxidant defences are rapidly depleted,
thus permitting further tissue damage. Several studies point to an antioxidant effect of B-group
vitamins and a pro-oxidant effect of elevated plasma tHcy (total homocysteine). In the present
study, we assessed whether supplementary B-group vitamins during this critical period will enhance
antioxidant capacity and mitigate oxidative damage. Forty-eight patients with acute ischaemic
stroke within 12 h of symptom onset were assigned to receive daily oral supplements of B-group
vitamins comprising 5 mg of folate, 5 mg of vitamin B2 , 50 mg of vitamin B6 and 0.4 mg of vitamin
B12 (n = 24) or no supplements (n = 24) for 14 days. The treatment group and controls were
matched for stroke subtype and age. Blood samples were obtained before intervention and also at 7
and 14 days post-recruitment for measurement of the following biomarkers: red cell folate (whole
blood folate corrected with haematocrit), erythrocyte glutathione reductase activity coefficient
(EGRAC; measure of vitamin B2 status), plasma pyridoxal phosphate (vitamin B6 status), plasma
vitamin B12 , plasma α-tocopherol, plasma ascorbic acid, plasma TAOC (total antioxidant capacity),
plasma MDA (malondialdehyde), plasma tHcy and CRP (C-reactive protein). Supplementation
for 14 days with B-group vitamins significantly increased the plasma concentrations of pyridoxal
phosphate and red blood cell folate and improved a measure of B2 status compared with the
control group (P < 0.05). Plasma tHcy decreased in both groups albeit less in the control group, but
differences in cumulative changes were not significant. There was, however, a decrease in plasma
MDA concentration in the treatment group, in contrast with the increase seen in the control
group and these differences were significant (P = 0.05). CRP concentration, a marker of tissue
inflammation, was significantly lower in the treatment group compared with controls (P < 0.05). In
conclusion, B-group vitamin supplementation immediately post-infarct may have antioxidant and
anti-inflammatory effects in stroke disease independent of a homocysteine-lowering effect.
Key words: B-group vitamin, C-reactive protein, homocysteine, ischaemia, stroke, malondialdehyde, total antioxidant capacity.
Abbreviations: CRP, C-reactive protein; CV, coefficient of variation; EGRAC, erythrocyte glutathione reductase activity coefficient;
LACI, lacunar infarct; LDL, low-density lipoprotein; MDA, malondialdehyde; PACI, partial anterior circulation infarct; POCI,
posterior circulation infarct; TACI, total anterior circulation infarct; TAOC, total antioxidant capacity; tHcy, total homocysteine.
Correspondence: Dr Salah E. Gariballa (email s.e.gariballa@sheffield.ac.uk).
C 2004 The Biochemical Society
478 R. Ullegaddi, H. J. Powers and S. E. Gariballa
There is strong indirect evidence that free radical pro- subject received thrombolytic therapy. Stroke patients
duction is an important mechanism of brain injury after with active gastrointestinal disease, severe medical
exposure to ischaemia and reperfusion [4]. Lipid perox- or psychiatric illness, serum creatinine concentration
idation with accumulation of both conjugated dienes and > 150 µmol/l, history of gout or acute renal failure,
thiobarbiturate-reactive material is consistently found supplemental vitamins or inability or refusal to give con-
when cerebral ischaemia is followed by reperfusion [4]. sent were excluded. The treatment group and controls
Recent studies in humans have reported an association were matched for age and stroke subtype. The class-
between plasma concentration of antioxidants, markers ification of stroke subtypes was based on clinically
of oxidative damage and early clinical outcome after acute identifiable subtypes of cerebral infarction as described
ischaemic stroke [5,6]. by Bamford et al. [13]. This classification has important
Elevated levels of plasma tHcy (total homocysteine) prognostic implications and is useful for planning stroke
and deficiencies in folate or vitamin B6 or B12 cofactors treatment trials. The classification is as follows: large
have been found in atherosclerotic patients, including anterior circulation infarcts with both cortical and sub-
strokes [7]. Several studies suggest that tHcy may have a cortical involvement [TACI (total anterior circulation
pro-oxidant effect and a role in the production of reactive infarcts)]; more restricted and predominantly cortical in-
oxygen species that results in oxidative damage to arterial farcts [PACI (partial anterior circulation infarcts)];
endothelial cells [8,9]. Through cofactor roles in homo- infarcts associated with the vertebro-basilar arterial
cysteine metabolism, increased intakes of B-vitamins territory [POCI (posterior circulation infarcts)]; and
have homocysteine-lowering effects [9]. The effect is infarcts confined to the territory of the deep perforating
strongest for folate, but vitamins B6 , B12 and B2 have arteries [LACI (lacunar infarcts)].
all been shown to be independently predictive of plasma
homocysteine. There is also recent evidence suggesting an Protocol
acute antioxidant effect of folate independent of its effect Non-fasting venous blood was obtained before treatment
on homocysteine [10,11]. Riboflavin may also have pro- (within 12 h of stroke onset) and at day 7 and day 14.
tective effects independent of homocysteine-lowering; Blood was collected in appropriate tubes, the haematocrit
this vitamin may protect tissues from ischaemia/reper- measured and an aliquot of whole blood stored at − 70 ◦ C
fusion injury, probably through an antioxidant effect [12]. with ascorbic acid for the measurement of red blood
We hypothesize that there is a rapid increase in the pro- cell folate. Following separation from red blood cells,
duction of markers of oxidative damage following acute plasma was stored, with stabilizer where appropriate, at
stroke due to the reperfusion event following ischaemia. − 70 ◦ C for analysis within 12 weeks. Red blood cells
Supplementary B-group vitamins during this critical were washed, haemolysed and stored at − 70 ◦ C for the
period will enhance antioxidant capacity and mitigate later measurement of vitamin B2 status. All analyses were
oxidative damage. The aim of the present study was, performed blind to the identity of the sample. Plasma B12
therefore, to test the effect of supplementary B-group was measured by competitive immunoassay using direct
vitamins during this critical period on plasma markers of chemiluminescence on an ADVIA Centaur system (kit
antioxidant capacity, oxidative damage and plasma tHcy reference 09544818; Bayer, Newbury, Berkshire, U.K.).
concentration. The inter-batch CV (coefficient of variation) was
7.0 %. The automated Ion Capture Assay system (IMX
folate assay; Abbott Laboratories, Abbott Park, IL,
METHODS U.S.A.) was utilized to determine the total folate con-
centration of whole blood, and had an inter-batch CV
Subjects of 6.1 %. Simultaneously measured haematocrit values
For a 12-month period, all patients admitted to a Uni- were then used to calculate red cell folate concentra-
versity Teaching Hospital with a diagnosis of ischaemic tions. Riboflavin status was assessed as the EGRAC
stroke according to the World Health Organization (erythrocyte glutathione reductase activity coefficient),
criteria were identified prospectively. The study was ap- using the Cobas BioAutoanalyser (Roche Diagnostics,
proved by the Local Health Research Ethics Committee. Indianapolis, IN, U.S.A.) [14], with an inter-batch CV
After providing informed written consent, 48 patients of 3.4 %. Ratios above 1.4 were considered to reflect
with acute ischaemic stroke admitted within 12 h of biochemical deficiency of riboflavin. Pyridoxine (vitamin
symptom onset (excluding those with cerebral or sub- B6 ) was measured as plasma pyridoxal phosphate by
arachnoid haemorrhage on CT head scan) were randomly HPLC [15]; this assay had an inter-batch CV of 8.0 %.
assigned to receive oral supplements of B-group vitamins Plasma tHcy was determined by an automated fluor-
comprising 5 mg of folate, 5 mg of vitamin B2 , 50 mg of escence polarization immunoassay (IMX Hcy assay;
vitamin B6 and 0.4 mg of vitamin B12 (n = 24) daily, or no Abbot Laboratories) with an inter-batch CV of 5.4 %.
supplement (n = 24), for 14 days. All patients, however, Plasma α-tocopherol was measured by HPLC [16],
received standard medical treatment for acute stroke. No and plasma total ascorbic acid was by a fluorescence assay
C 2004 The Biochemical Society
Effect of B-group vitamin supplementation following acute ischaemic stroke 479
automated for the Cobas BioAutoanalyser [17], giving tests (median and inter-quartile range) were used to
inter-batch CVs of 4.3 and 9.0 % respectively. TAOC describe the baseline characteristics of the subjects.
(total antioxidant capacity) was measured as the ability Kruskal–Wallis and Mann–Whitney U tests were used to
of plasma to inhibit free-radical-induced oxidation in an test between-group differences where appropriate. The
assay automated for the Cobas BioAutoanalyser, with Friedman test was used to test within-group differences.
an inter-batch CV of 5.0 % [18]. Serum total cholesterol Forward stepwise multiple regression analysis was per-
concentrations, measured by a cholesterol oxidase assay formed to determine the predictive importance of baseline
using the Synchron LX System (kit reference 467825; variables (age, stroke subtype, drugs, chronic disease and
Beckman, Fullerton, CA, U.S.A.), were used to adjust CRP) for plasma TAOC, MDA and tHcy. Adjusted
α-tocopherol values. The inter-batch CV of this assay and change in R2 values were then used to determine
was 6.0 %. Plasma MDA (malondialdehyde), a marker of the extent to which the TAOC, MDA and tHcy concen-
lipid peroxidation, was also measured by HPLC using trations could be explained by individual antioxidants
fluorescence detection [19], and this had an inter-batch (vitamin C, vitamin E and uric acid) and B-group vita-
CV of 9.0 %. CRP (C-reactive protein) concentration, a mins, but also other clinical variables included in the
marker of acute inflammation in stroke [20] and other model.
illnesses (normal range, < 6 mg/l), was measured by
a modified latex-enhanced immuno-turbidimetric assay
(Synchron LX System; kit reference 465131; Beckman). RESULTS
Plasma uric acid concentration was measured by an
enzymic uricase fluorescence method (normal range, 160– Table 1 shows baseline characteristics of the treatment and
400 µmol/l) on the Synchron LX System (kit reference control groups on entry into the study. The two groups
442785; Beckman). The inter-batch CVs for these assays were comparable with respect to clinical stroke subtype,
were 6.5 % and 5.0 % respectively. chronic disease, drug and alcohol intake and time between
All patients had demographic and medical data col- stroke onset and randomization. Although there were
lected at baseline, including history of hypertension, some differences between the two groups in age, rate
smoking, alcohol and drug intake, diabetes mellitus of atrial fibrillation, disability, blood glucose and serum
and cardiovascular diseases. Disability at baseline was creatinine, these differences did not reach statistical
assessed using the Barthel score on a 100-point scale significance, except for blood glucose (P < 0.05). Vitamin
[21]. The Barthel scores 10 functions on a scale 0 (fully supplements were prescribed in the drug cards of the
dependent) to 100 (independent). patients, and compliance for surviving patients was 100 %
with no treatment failure or withdrawal.
Allocation of treatment The concentrations of plasma vitamins B6 and B12 , red-
The method of minimization [22] is an accepted form of cell folate and EGRAC (vitamin B2 ) values are shown in
randomization and was chosen because it minimizes any Table 2. Supplementation of B-group vitamin for 14 days
differences and provides treatment groups very similar for significantly increased concentrations of plasma vitamin
major prognostic variables. The effect of this procedure B6 and red cell folate and reduced EGRAC (inversely
was that the groups were similar with regard to the associated with riboflavin status) in the intervention
chosen variables of interest. In this trial, clinical stroke group compared with controls. Concentrations of plasma
subtype (TACI, PACI, LACI and POCI) and age (< 75 vitamin B12 increased significantly in both groups over
and 75 years) were used for minimization. 2 weeks (P < 0.05 for within-group differences), but there
was no significant difference between groups.
Sample size calculation Table 3 shows plasma concentrations of ascorbic acid,
A previous study [23] has shown that daily supplement- α-tocopherol, TAOC and MDA during the study period.
ation with B-group vitamins (5 mg of folate, 50 mg of Within-group and between-group differences in plasma
vitamin B6 and 0.4 mg of vitamin B12 ) to patients with ascorbic acid, α-tocopherol and TAOC were not signi-
recurrent venous thrombosis (19 on supplements com- ficant. Although plasma TAOC concentrations decreased
pared with 15 on placebo) reduced plasma tHcy concen- in both groups, albeit less so in the treatment group
trations by up to 33 %. A sample size of 48 ischaemic over the study period, cumulative differences between the
stroke patients (24 treatment group + 24 controls) would two groups were not statistically significant. However,
therefore allow the detection of a 30 % difference between the reduction in plasma MDA concentration seen in the
groups in plasma tHcy with 80 % power and type 1 error treatment group was in contrast with the increase seen
probability of 0.05. in the control group and these different responses were
significant. Plasma tHcy concentrations decreased more
Data analysis in the treatment group than the controls, but difference in
Statistical analyses were performed with SPSS software, cumulative changes between the two groups was not
version 11.0 (SPSS Inc., Chicago, IL, U.S.A.). Descriptive significant (Table 4).
C 2004 The Biochemical Society
480 R. Ullegaddi, H. J. Powers and S. E. Gariballa
Table 1 Baseline characteristics of the treatment and in the multivariate model, only stroke subtype showed
control stroke groups significant correlation with TAOC at day 7 (Table 5).
Values are medians (inter-quartile range), or numbers (%). ∗ P < 0.05 compared Most of the variance in MDA was explained by age
with control. AF, atrial fibrillation; TIA, transient ischaemic attack. at baseline and α-tocopherol at day 7, but tHcy showed
significant positive correlation with MDA at baseline and
Treatment group Control group
day 14 (Table 6).
n 24 24 We also analysed the effects of B-group vitamins, age,
Age (years) 77 (68–81) 79 (73–84) stroke subtype, tissue inflammation (CRP), drugs and
Sex (males) 15 (63) 15 (63) alcohol on plasma tHcy during the study period using a
Stroke sub-type (n) multivariate model. Folate and age were significant pre-
TACI 6 (25) 6 (25) dictors of tHcy during the study period (Table 7).
PACI 8 (33) 8 (33) Figure 1 shows the CRP profile over a period of
LACI 6 (25) 6 (25) 3 months following the stroke. CRP concentrations were
POCI 4 (17) 4 (17) significantly lower in the treatment group compared with
Smoking history (n) controls (P < 0.05). There were no statistically significant
Never smoked 10 (42) 12 (50) differences in the rate of infective complications (pneu-
Ex-smoker 11 (46) 8 (33) monia, urinary tract infection and septicaemia) or re-
Current smoker 3 (12) 4 (17) infarction between the treatment group and controls
Alcohol intake > 21 units/week (n) 1 (4) 1 (4) during the first 2 weeks post-randomization [two (8 %)
Drugs/patient† (n) 1.9 1.5 and no patients in the treatment group had infective
Previous stroke/TIA (n) 8 (33) 10 (42) complications during the first and second weeks post-
AF on ECG (n) 3 (12) 8 (33) randomization respectively, compared with three (13 %)
Hypertension (n) 8 (33) 9 (38) and one (4 %) in the control group; two [8 %] stroke
Ischaemic heart disease (n) 8 (33) 6 (25) patients in the treatment group had recurrent cerebral
Diabetes mellitus (n) 1 (4) 1 (4) infarction compared with two (8 %) controls].
CRP (mg/l) 6 (6–12.5) 6 (6–9)
Blood glucose (mmol/l) 5.7 (5.3–7.0)∗ 6.7 (5.8–7.6)
Serum creatinine (µmol/l) 88 (71–117) 99 (69–142)
DISCUSSION
Barthel score 40 (16–59) 28 (20–65) Supplementation with B-group vitamins within 12 h
Time from onset of stroke to 8.0 (5.0–12.8) 9.5 (5.0–12.0) post-infarct after ischaemic stroke reduced lipid perox-
randomization (h) idation products (MDA) after 2 weeks. This occurred
† Diuretics, angiotensin-converting-enzyme (ACE) inhibitors, β-blockers, calcium channel in association with a predicted increase in the plasma
blockers, statins, aspirin, warfarin and clopidogrel. concentrations of certain B vitamins. Although tHcy de-
creased more in the treatment group compared with con-
trols, overall there was no significant difference between
In a multivariate analysis, plasma uric acid and red the two groups. Plasma CRP, a marker of tissue
cell folate explained most of the variance in TAOC du- inflammation, was also lower over 90 days in those
ring the study period. Of the other variables included subjects who received B-group vitamin supplements
Table 2 Changes in concentrations of B-group vitamins over the study period in the treatment and control stroke groups
Values are medians (inter-quartile range), or median percentage change over 2 weeks. ∗ P values are for the differences in cumulative changes between groups over
2 weeks.
Group Day Vitamin B12 (pmol/l) Vitamin B6 (nmol/l) Red cell folate (nmol/l) Vitamin B2 EGRAC
Treatment (n = 24) 0 302 (207–417) 23.1 (15.1–37.1) 481 (327–720) 1.33 (1.23–1.61)
7 482 (307–706) 212.4 (143.2–295.7) 603 (413–910) 1.15 (1.11–1.24)
14 404 (293–697) 222.4 (121.1–330.9) 616 (388–902) 1.17 (1.12–1.27)
Percentage change over 2 weeks 46 778 21 9
Control (n = 24) 0 345 (268–481) 31.7 (16.8–45.2) 606 (426–720) 1.28 (1.19–1.37)
7 417 (305–737) 35.5 (22.9–53.1) 590 (403–757) 1.24 (1.18–1.35)
14 452 (270–733) 32.6 (19.5–48.9) 546 (410–676) 1.27 (1.18–1.49)
Percentage change over 2 weeks 16 14 3 2
∗
P value 0.64 < 0.01 < 0.01 < 0.01
C 2004 The Biochemical Society
Effect of B-group vitamin supplementation following acute ischaemic stroke 481
Table 3 Change in plasma ascorbic acid, α-tocopherol, TAOC and MDA over the study period in the treatment and control
stroke groups
Values are medians (inter-quartile range), or median percentage change over 2 weeks. ∗ P values are for the differences in cumulative changes between groups over
2 weeks.
Group Day Ascorbic acid (µmol/l) Adjusted α-tocopherol (µmol/l) TAOC (mmol/l) MDA (µmol/l)
Treatment (n = 24) 0 39.7 (18.2–55.4) 4.68 (3.51–5.62) 1.30 (1.16–1.38) 0.60 (0.49–0.76)
7 28.6 (14.8–49.8) 4.3 (3.76–5.19) 1.25 (1.14–1.38) 0.57 (0.48–0.71)
14 28.0 (22.4–48.1) 4.31 (3.39–5.72) 1.26 (1.17–1.40) 0.53 (0.37–0.77)
Percentage change over 2 weeks 20 6 2 5
Control (n = 24) 0 33.6 (21.3–63.8) 4.40 (3.77–4.98) 1.29 (1.19–1.38) 0.58 (0.49–0.71)
7 30.2 (19.3–46.2) 4.42 (3.72–5.89) 1.24 (1.17–1.41) 0.68 (0.49–0.88)
14 33.6 (19.0–50.4) 4.98 (3.57–5.90) 1.21 (1.07–1.43) 0.70 (0.52–0.95)
Percentage change over 2 weeks 12 5 7 8
∗
P value 0.76 0.22 0.42 0.05
Table 4 Change in plasma tHcy over the study period in the variance in MDA was explained by age, vitamin E
the treatment and control stroke groups and tHcy. Folate status and age were the most significant
Values are medians (inter-quartile range), median change or median percentage independent predictors of tHcy.
change over 2 weeks. Many recent studies have reported an association
between plasma tHcy concentrations and risk of stroke
Group Day tHcy (µmol/l)
independent of other prognostic indicators [7]. Mech-
Treatment (n = 24) 0 16.73 (14.44–19.67) anisms whereby elevated plasma tHcy increases risk of
7 15.28 (12.02–22.19) stroke are not fully understood, but several studies point
14 14.46 (11.90–23.67) to a pro-oxidant effect. High homocysteine levels may
Change in tHcy lead to endothelial damage, affect platelet function and
Day 0–7 − 1.63 (− 4.59–+ 0.05) coagulation factors, and promote LDL (low-density lipo-
Day 0–14 − 2.30 (− 3.52–+ 0.92) protein) oxidation [8,9,24]. Hyperhomocysteinaemia has
Percentage change over 2 weeks 12 been suggested to play a role in the production of re-
active oxygen species, resulting in oxidative damage to ar-
Control (n = 24) 0 16.48 (11.97–24.06)
terial endothelial cells [9,25]. The homocysteine-lowering
7 13.21 (11.26–19.42)
activity of B-group vitamins may therefore be viewed
14 12.82 (10.82–15.56)
as potentially antioxidant in nature. However, folate can
Change in tHcy
evidently behave as a free-radical scavenger, which might
Day 0–7 − 0.91 (− 4.10–+ 1.43)
explain reported beneficial effects of high-dose folate
Day 0–14 − 0.70 (− 3.95–+ 2.66)
on the vasculature [26]. Nakano et al. [10] have shown
Percentage change over 2 weeks 4
recently that folate has a direct protective antioxidant
effect on LDL oxidation in vitro, and others have shown
for 14 days compared with controls. Plasma TAOC protective effects on the endothelium that appear to be
concentrations decreased in both groups and cumulative independent of homocysteine-lowering [11].
differences between the two groups were not statistically Elevated plasma CRP concentration during the acute
significant. Plasma TAOC was significantly accounted phase or recovery in stroke is independently predictive of
for by serum uric acid and red cell folate status. Most of poor long-term outcome [27,28]. We found a significant
Table 5 Multiple regression analysis for treatment and control stroke groups with TAOC as the dependent variable
P < 0.05 is significant.
C 2004 The Biochemical Society
482 R. Ullegaddi, H. J. Powers and S. E. Gariballa
Table 6 Multiple regression analysis for treatment and control stroke groups with MDA as the dependent variable
P < 0.05 is significant.
Table 7 Multiple regression analysis for treatment and control stroke groups with tHcy as the dependent variable
P < 0.05 is significant.
C 2004 The Biochemical Society
Effect of B-group vitamin supplementation following acute ischaemic stroke 483
C 2004 The Biochemical Society
484 R. Ullegaddi, H. J. Powers and S. E. Gariballa
20 Di Napoli, M., Papa, F. and Bocola, V. (2001) Prognostic 27 Winbeck, K., Poppert, H., Etgen, T., Conrad, B. and
influence of increased C-reactive protein and fibrinogen Sander, D. (2002) Prognostic relevance of early serial
levels in ischemic stroke. Stroke 32, 133–138 C-reactive protein measurements after first ischemic
21 (1992) Standardised assessment scales for older people. A stroke. Stroke 33, 2459–2464
report of joint workshop, pp. 1–27, The Royal College of 28 Muir, K. W., Weir, C. J., Alwan, W., Squire, I. B. and Lees,
Physicians of London and The British Geriatrics Society, K. R. (1999) C-reactive protein and outcome after ischemic
London stroke. Stroke 30, 981–985
22 Pocock, S. J. (1988) Clinical Trials: A practical Approach, 29 Friso, S., Jacques, P. F., Wilson, P. W., Rosenberg, I. H. and
John Wiley and Sons, Chichester Selhub, J. (2001) Low circulating vitamin B6 is associated
23 den Heijer, M., Brouwer, I. A., Bos, G. M. et al. (1998) with elevation of the inflammation marker C-reactive
Vitamin supplementation reduces blood homocysteine protein independently of plasma homocysteine levels.
levels: a controlled trial in patients with venous thrombosis Circulation 103, 2788–2791
and healthy volunteers. Arterioscler., Thromb., Vasc. Biol. 30 Gariballa, S. E., Hutchin, T. P. and Sinclair, A. J. (2002)
18, 356–361 Antioxidant capacity after acute ischaemic stroke.
24 Harker, L. A., Harlan, J. M. and Ross, R. (1983) Effect of Q. J. Med. 95, 685–690
sulfinpyrazone on homocysteine-induced endothelial 31 Patt, A., Harken, A. H., Burton, L. K. et al. (1988)
injury and arteriosclerosis in baboons. Circ. Res. 53, Xanthine oxidase-derived hydrogen peroxide contributes
731–739 to ischemia reperfusion-induced edema in gerbil brains.
25 El Kossi, M. M. and Zakhary, M. M. (2000) Oxidative J. Clin. Invest. 81, 1556–1562
stress in the context of acute cerebrovascular stroke. 32 Woodside, J. V., Yarnell, J. W., McMaster, D. et al. (1998)
Stroke 31, 1889–1892 Effect of B-group vitamins and antioxidant vitamins on
26 Wilmink, H. W., Stroes, E. S., Erkelens, W. D. et al. (2000) hyperhomocysteinemia: a double-blind, randomized,
Influence of folic acid on postprandial endothelial factorial-design, controlled trial. Am. J. Clin. Nutr. 67,
dysfunction. Arterioscler., Thromb., Vasc. Biol. 20, 185–188 858–866
C 2004 The Biochemical Society