Plasma Homocysteine and Severity of Atherosclerosis in Young Patients With Lower-Limb Atherosclerotic Disease
Abstract Elevated plasma homocysteine levels are recognized as an independent risk factor for atherosclerotic disease. It is not known (1) whether the severity of atherosclerotic disease is related to hyperhomocyst(e)inemia or (2) whether any such relation differs between fasting and post–methionine loading plasma homocysteine levels. Therefore, in 171 consecutive patients under 55 years of age with first symptoms of lower-limb disease, we examined the relation between severity of atherosclerosis and plasma homocysteine concentration. Severity of atherosclerotic disease was estimated from the prevalence of coronary artery disease and cerebrovascular disease and from the angiographic extent of lower-limb disease. Plasma homocysteine was measured after a period of fasting and in response to methionine loading (0.1 g/kg). In multivariate analysis, the prevalence of coronary artery disease plus cerebrovascular disease was related to both fasting and postmethionine homocysteine levels (odds ratio [OR] for the upper quartile versus the lower three quartiles, 2.8, 95% confidence interval [CI], 1.1 to 7.5; and OR 3.0, 95% CI, 1.1 to 7.8, respectively). The extent of lower-limb disease was weakly related to the fasting homocysteine level (partial correlation coefficient, .12; P=.17) and more strongly related to the postmethionine homocysteine level (partial correlation coefficient, .25; P=.003). These relations tended to be more pronounced in women than in men. They were independent of age, total serum cholesterol, blood pressure, and smoking habit. We concluded that the severity of atherosclerotic disease in young patients with lower-limb atherosclerotic disease is associated with high postmethionine and fasting homocysteine concentrations.
Reprint requests to Dr M. van den Berg, Institute for Cardiovascular Research, Department of Vascular Surgery, Free University Hospital, PO Box 7057, 1007 MB Amsterdam, Netherlands.
- Received September 11, 1995.
- Accepted October 17, 1995.
Lower-limb atherosclerotic disease is a common finding among the elderly. The prevalence of lower-limb atherosclerotic disease among men and women over 60 years of age varies between 7% and 14.4% or 3% and 14.1%, respectively, when the Rose questionnaire for intermittent claudication is used and between 16% and 35% in men and 13% and 28% in women when the ankle-brachial blood pressure index is used.1 Even among young people, however, the prevalence of lower-limb atherosclerotic disease is not negligible. Studies in persons 60 years old or younger indicate a prevalence of intermittent claudication of 0.4% to 5.8% in men and 0.7% to 1.8% in women.1 Use of the ankle-brachial blood pressure index again yields higher estimates, both in men (4.2% to 16%) and in women (5.4% to 13%).1
Follow-up of patients with lower-limb atherosclerotic disease2 3 shows a fourfold to sixfold increased risk of death as a result of cardiovascular disease, which cannot be fully explained by traditional risk factors such as smoking, hypertension, diabetes, and male sex.2 3 4 This suggests that in patients with lower-limb atherosclerotic disease, other risk factors affect the severity of atherosclerosis, which, moreover, is known to vary considerably both in the legs and in the coronary and cerebral vasculature.5
Hyperhomocyst(e)inemia has been shown to be an independent risk factor for atherosclerotic disease. This association is not limited to severe hyperhomocyst(e)inemia6 ; it is also observed when homocysteine levels are moderately elevated.7 8 9 10 11 12 13 Studies in vitro and in animals show that excess homocysteine may promote atherogenesis by inducing endothelial cell injury14 15 and vascular smooth muscle cell proliferation.16
It is not known whether hyperhomocyst(e)inemia modulates the severity of atherosclerosis in patients with lower-limb atherosclerotic disease. We chose to investigate this question in young patients with lower-limb atherosclerotic disease because hyperhomocyst(e)inemia after oral methionine loading is known to be prevalent among such patients.7 9 Severity of atherosclerotic disease was estimated from angiography of the legs and history of coronary and cerebrovascular events. Plasma homocysteine concentration was measured both in the fasting state and after a methionine load (which stimulates homocysteine synthesis) because it is unclear which is the better predictor of vascular disease.
We studied 185 consecutive white patients under the age of 55 years who had been referred to the Department of Vascular Surgery at the Free University Hospital in Amsterdam between April 1991 and May 1994 because of first symptoms of lower-limb atherosclerotic disease. Patients with renal (serum creatinine >120 μmol/L) or liver dysfunction (abnormal serum aminotransferase concentrations and/or presence of physical signs) were excluded from the study (n=14). Thus, 171 patients were included. Lower-limb atherosclerotic disease was defined as an ankle-brachial blood pressure index <0.9.1 4 17 In view of their age, all patients underwent a methionine-loading test and angiography of the legs as part of their clinical workup. All patients gave informed consent for these studies, which were approved by the local ethics committee. All studies were performed on referral except in 2 patients in whom methionine loading was performed at the age of 56 and 57 years, respectively.
Severity of Vascular Disease
Extent of Lower-Limb Atherosclerotic Disease
Angiography was performed by use of the transfemoral Seldinger technique; results were examined by vascular surgeons or radiologists. The percent diameter reduction (stenosis) was determined by comparing the narrowest segment of the appropriate artery with the closest adjacent apparently normal segment of the same artery. The degree of anatomical obstruction was graded as 0% to 19%, 20% to 49%, 50% to 99%, or total occlusion and was assigned a score of 0, 1, 2, or 3, respectively. Estimates of disease extent were then obtained (1) by classifying patients as having either single-level (≥20% obstruction confined to the aortoiliac, femoropopliteal, or infrapopliteal level on one or both sides) or multilevel disease5 and (2) by calculating the “total run-off resistance,”18 a composite score that takes into account both the degree of anatomical obstruction (“a,” graded 0 to 3 as above) and the relative contribution to outflow (“b,” graded 1 to 3 with increasing contribution to outflow) of each of the 29 angiographically defined vessel segments (aorta and both legs).18 The score per segment is calculated as a×b. The total score was calculated as the sum score of all segments. In contrast to estimates of disease extent such as the ankle-brachial blood pressure index or the presence of multilevel disease, total runoff resistance considers all relevant vessels in both legs and therefore is thought to give a better estimate of the overall extent of disease.18
Additional Coronary and Cerebrovascular Disease
Coronary artery disease was defined as a history of myocardial infarction or angina pectoris; cerebrovascular disease was defined as a history of ischemic stroke or transient ischemic attack (World Health Organization clinical definitions). These diagnoses were accepted only when corroborated by a written report by a physician; in addition, diagnoses of myocardial infarction and ischemic stroke required confirmation by new Q waves on electrocardiography and/or diagnostic enzyme changes or by computed tomography, respectively.
The total (free plus protein-bound) homocysteine plasma concentration was measured in the fasting state and 6 hours after an oral methionine load (0.1 g/kg body weight) as previously described in detail.19 20 Samples were stored at −30°C and assayed within 1 week of blood sampling. The intra-assay and interassay coefficients of variation for total homocysteine were 2.1% and 5.1%, respectively.
Other Clinical and Laboratory Data
On referral, we recorded age, body weight, menopausal status (postmenopause was defined as absence of menstrual bleeds for >1 year), current smoking habit (yes or no), the presence of diabetes mellitus (WHO criteria), and blood pressure (measured after 15 minutes of supine rest without altering antihypertensive regimens). Hypertension was defined as diastolic blood pressure >90 mm Hg and/or systolic blood pressure >140 mm Hg and/or antihypertensive drugs having been prescribed for the subject. In addition, we measured fasting serum total cholesterol (enzymatically), glucose (glucose oxidase method), and creatinine (modified Jaffé reaction). In case of an abnormal postmethionine homocysteine concentration (>97.5% of previously proposed normal values20 ), we also assessed in fasting serum samples the concentrations of vitamin B6 (by fluorescent high-pressure liquid chromatography; normal, >17 nmol/L), vitamin B12 (radioassay, Becton Dickinson; normal, >80 pmol/L), and folic acid (radioassay, Becton Dickinson; normal, >3.4 nmol/L). Data collection and assessment of the severity of vascular disease were “blinded” to results of plasma homocysteine measurements.
Descriptive data are given as mean±SD or median (range) and, when appropriate, compared with use of standard parametric or nonparametric tests. Univariate logistic and linear regression were used to examine whether estimates of the severity of atherosclerosis (prevalence of coronary artery disease, cerebrovascular disease, and multilevel peripheral disease and the total runoff resistance score as dependent variables) were related to the fasting and postmethionine plasma homocysteine concentration or to other cardiovascular risk factors (age, sex, smoking habit, serum cholesterol, presence of diabetes mellitus, blood pressure [as a continuous variable or as presence versus absence of hypertension], and serum creatinine concentration as independent variables). We then repeated the analysis, using multivariate regression, to examine whether relations between the various estimates of severity of atherosclerosis and plasma homocysteine concentrations were modulated by the other cardiovascular risk factors.
There were five patients who had had both a myocardial infarction and a stroke. In the analyses that considered all cardiovascular events, we included only the first event; in the separate analyses of cardiac and cerebral disease, we included all events in these five patients (ie, 10 events).
All odds ratios (OR) are given with their 95% confidence interval (CI) in parentheses and contrast the odds in the upper quartile of the distribution (Q4) of the homocysteine concentration with the lower three quartiles (Q1 through Q3) or the odds per unit change of the other cardiovascular risk factors (eg, per 1 year of age and per 1 mmol/L of serum cholesterol).
In addition, the influence of homocysteine concentrations on severity of vascular disease was studied (1) as a continuous variable and (2) by comparing patients with elevated concentrations (>97.5% of previously proposed sex-specific normal values20 ) to those with normal homocysteine concentrations. A value of P<.05 was considered significant. All testing was two-sided.
Table 1⇓ shows clinical and laboratory characteristics of the patients. The combined prevalence of cardiac plus cerebrovascular disease was 37 (21.6%) of the 171 patients. Prevalence of coronary artery disease and cerebrovascular disease, respectively, was 18 (19.8%) and 5 (5.5%) of 91 men and 12 (15.0%) and 7 (8.8%) of 80 women (P=.39 and P=.41, respectively).
Median fasting homocysteine levels were higher in men than in women (difference, 1.5 μmol/L; P=.05). Postmethionine levels were higher in women than in men (difference, 2.8 μmol/L; P=.06). We therefore report sex-specific strata of homocysteine levels. Tables 2⇓ and 3⇓ show clinical and laboratory characteristics according to quartiles of fasting and postmethionine homocysteine levels, respectively. In both men and women, higher homocysteine levels (men, fasting; women, both fasting and postmethionine) were associated with the presence of hypertension and with higher serum creatinine concentrations. Twenty-six (63%) of 41 patients in the upper quartile of fasting homocysteine levels were also in the upper quartile of postmethionine homocysteine levels. Conversely, 26 (62%) of 42 patients in the upper quartile of postmethionine homocysteine levels were also in the upper quartile of fasting homocysteine levels.
Coronary Artery and Cerebrovascular Disease and Plasma Homocysteine Concentration
Fig 1⇓ shows that the combined prevalence of coronary artery disease plus cerebrovascular disease was significantly increased in the fourth quartile of both fasting (univariate OR, 2.8; 95% CI, 1.3 to 6.2), and postmethionine homocysteine concentrations (OR, 3.2; 95% CI, 1.5 to 6.9; Table 4⇓). Adjustments for other cardiovascular risk factors did not alter these results significantly (OR, 2.8 and 3.0; Table 4⇓), nor did the exclusion of angina and transient ischemic attack from the analysis (ie, limiting the analysis to myocardial infarction and stroke) (univariate OR 2.8; 95% CI, 1.2 to 6.8 for fasting homocysteine; OR 3.3 and 95% CI, 1.4 to 7.8 for postmethionine homocysteine).
Table 4⇑ shows that these associations were strongest for postmethionine homocysteine levels in women (adjusted OR, 25.2 versus 1.4 in men; P<.05). Further analysis indicated that the associations were most pronounced in postmenopausal women (adjusted OR, 14.8; 95% CI, 1.1 to 131 for fasting homocysteine; OR 36.0 and 95% CI, 2.8 to 471 for postmethionine homocysteine).
Analyses of coronary artery and cerebrovascular disease separately (Table 4⇑ and Fig 2⇓) showed similar results in that the prevalence of coronary artery disease and cerebrovascular disease was increased in the fourth quartile of both fasting and postmethionine concentrations. These associations tended to be stronger for women than for men, and in women, stronger for postmethionine than for fasting homocysteine levels. The presence of coronary artery disease was also associated with age (univariate OR, 1.1; 95% CI, 1.05 to 1.2) and serum cholesterol (OR, 1.5; 95% CI, 1.1 to 1.9). In multivariate analysis, only the association with age remained (OR, 1.2; 95% CI, 1.1 to 1.3). In contrast, prevalence of cerebrovascular disease was associated with the presence of hypertension (univariate OR, 6.8; 95% CI, 1.9 to 24.0), diastolic blood pressure (OR, 1.1; 95% CI, 1.01 to 1.2), and serum creatinine (OR, 1.03; 95% CI, 1.0 to 1.05). In multivariate analysis, only the association with diastolic blood pressure remained (OR, 1.1; 95% CI, 1.03 to 1.2).
It must be emphasized that these subgroup analyses of coronary artery and cerebrovascular disease concern small groups and therefore yield wide confidence intervals of the risk estimates. Thus, the differences between men and women or between fasting and postmethionine homocysteine were not statistically significant.
Lower-Limb Atherosclerotic Disease and Plasma Homocysteine Concentration
On angiography, 57 (63%) of 91 men and 41 (51%) of 80 women had abnormalities in both legs. Multilevel disease was present in 36 men (40%) and in 25 women (31%) and was present in both legs in 9 (25%) of these men and in 7 (28%) of these women.
The prevalence of multilevel disease was significantly increased in the upper quartile of fasting homocysteine levels (adjusted OR, 4.0). The total runoff resistance score was significantly associated with postmethionine homocysteine levels (partial correlation coefficient, .25; P=.003; Table 4⇑).
The association with multilevel disease was most pronounced for fasting homocysteine levels in men (adjusted OR, 7.2; P<.05 versus fasting homocysteine levels in women and versus postmethionine homocysteine levels in men; Table 4⇑), whereas the total runoff resistance score was more strongly associated with postmethionine homocysteine levels in women (partial correlation coefficient, .52; P=.0001). It is likely that the OR for multilevel disease and postmethionine homocysteine level in women is underestimated, because 5 (56%) of 9 women in the upper quartile of postload homocysteine had multilevel disease in both legs. Additional analysis showed that the presence of multilevel disease was associated with the presence of hypertension (univariate OR, 2.9; 95% CI, 1.4 to 5.9) and with systolic blood pressure (OR, 1.02; 95% CI, 1.004 to 1.04). In multivariate analysis, association with the presence of hypertension remained significant (OR, 2.6; 95% CI, 1.3 to 5.2). Similarly, the total runoff resistance score was associated with the presence of hypertension (r=.26; P=.001) and with systolic blood pressure (r=.19; P=.02). In multivariate analysis, the association with systolic blood pressure remained significant (P=.008).
Other Estimates of Hyperhomocyst(e)inemia
Analyses that considered homocysteine concentration as a continuous variable or as “elevated” versus “normal” yielded similar results (data not shown). By the latter criterion, 42 patients had an abnormal postmethionine concentration and 29 patients had an abnormal fasting level.
Serum Vitamin Concentrations and Plasma Homocysteine
The serum concentrations of folic acid and of vitamins B6 and B12 were measured in all patients with postmethionine homocysteine levels >97.5% of previously proposed normal values20 (n=42); these serum concentrations were within their respective normal ranges. In this group, plasma vitamin levels were not significantly related to postmethionine or fasting homocysteine concentrations.
The main finding in the present study is that the severity of atherosclerosis in young patients with lower-limb atherosclerotic disease is related to homocysteine concentrations. Our results were robust in that the association was present for each of the estimates of disease severity (coronary artery, cerebrovascular, and lower-limb disease), was not influenced by the various possible definitions of hyperhomocyst(e)inemia, and was independent of classic risk factors for atherosclerosis. The relatively high prevalence of coronary artery disease in the second quartile of homocysteine levels (Fig 2⇑) is probably a chance finding.
Several findings support the hypothesis that hyperhomocyst(e)inemia is causally related to atherosclerotic disease. First, the prevalence of coronary artery, cerebrovascular, and peripheral arterial disease is associated with hyperhomocyst(e)inemia.6 7 8 9 10 11 In men, higher homocysteine levels are associated with an increased incidence of myocardial infarction12 13 and possibly of stroke.21 In addition, a “dose-response” relationship is supported by the fact that the severity of atherosclerotic disease is associated with hyperhomocyst(e)inemia, as shown by the present study in patients with lower-extremity atherosclerotic disease and by a previous study in men with coronary heart disease.22 Second, data from in vitro studies14 suggesting that excess homocysteine may cause endothelial injury, a key event in atherogenesis, were recently supported by a small study23 in patients with lower-limb atherosclerotic disease that showed that endothelial dysfunction, as estimated by increased plasma von Willebrand factor concentrations, was ameliorated by treatment of hyperhomocyst(e)inemia with folic acid and vitamin B6.
It is currently debated whether the association between atherosclerotic disease and hyperhomocyst(e)inemia differs between men and women, by location of disease, and between fasting or random (ie, either fasting or nonfasting but without methionine loading) and postmethionine homocysteine levels. Our results suggest that among young patients with lower-limb atherosclerotic disease, the association is most pronounced in women (especially postmenopausal women), is about equally strong for coronary artery and cerebrovascular disease, and is not clearly different for postmethionine than for fasting homocysteine levels.
Malinow et al24 observed in asymptomatic subjects a graded increase in prevalence of carotid artery intimal-medial thickening with increasing random homocysteine levels that was more pronounced in women. In contrast, Selhub et al25 reported no significant sex difference in the relation between the prevalence of carotid artery stenosis and random homocysteine levels. A recent prospective study26 did not observe a significant relation between incidence of stroke and (random) homocysteine levels in healthy men or women, although the OR was in fact somewhat higher in women. Thus, notwithstanding the highly selected nature of our patient group, a sex difference in the risk associated with hyperhomocyst(e)inemia is a distinct possibility and needs further investigation.
Our study directly compared the risks associated with fasting and postmethionine homocysteine levels. The data raise the possibility that severity of atherosclerosis is more strongly related to postmethionine homocysteine than to fasting homocysteine levels, especially in women (Table 4⇑). This finding is supported by indirect comparisons of risks, ie, the risk associated with the fasting level in one population to that associated with the postmethionine level in another.10 12 21 24 25 If corroborated by additional studies, this conclusion has important clinical and theoretical implications. First, postmethionine homocysteine concentration would provide the most accurate risk estimate. Identification of patients at risk made only on the basis of their fasting homocysteine level would not be possible because, depending on the exact cutoff, about 38% to 54% of patients with high postmethionine homocysteine concentrations have a normal fasting level (Reference 20 and the present study). The latter finding, moreover, suggests the existence of multiple underlying metabolic deficits, because fasting and postmethionine homocysteine levels are thought to be determined by different pathways, ie, remethylation and transsulfuration of homocysteine, respectively.27 Second, because most models of homocysteine-induced vascular injury assume that the average extracellular homocysteine concentration is relevant,14 15 it is interesting to note that the fasting (or random) homocysteine level reflects the average daily level much more closely than does the postmethionine level.28 29 Thus, the high risk associated with postmethionine homocysteine levels indicates either that peak (not average) extracellular levels are important or that postmethionine levels reflect some crucial intracellular metabolic disturbance.
Homocysteine metabolism is modulated by genetic and environmental factors. The most important examples of the latter are renal failure and deficiencies of folic acid, vitamin B12, and vitamin B6.10 It is less clear whether variability in these vitamins within their conventionally defined normal ranges also affects homocysteine levels. Our results suggest that any such influence is not strong in young patients with high homocysteine levels. The explanation for this is not clear but may be related to a relatively high prevalence of genetic abnormalities of homocysteine metabolism in young patients with hyperhomocyst(e)inemia,10 which might alter the relations between homocysteine and vitamin levels.25
Our study has several limitations. Its sample size was sufficient to investigate the relationship between severity of atherosclerotic disease and homocysteine levels but insufficient to definitively answer the questions of whether these relationships differ between fasting and postmethionine homocysteine and/or between men and women. Because there are no published studies on these issues, our sample size considerations were based on indirect comparisons, which suggested that the risk of vascular disease might be much higher with postmethionine than with fasting hyperhomocyst(e)inemia (eg, relative risk, 23.99 versus 3.112 for coronary artery disease). Our results indicate that such large differences are unlikely in young patients with lower-limb disease.
We did not measure other known risk factors for atherosclerotic disease that may be relevant in patients with lower-limb atherosclerotic disease, such as lipid profile (other than total cholesterol), fibrinogen,30 and cross-linked fibrin degradation products.31 The possible relationship between fibrinogen and homocysteine concentrations22 needs further study. In addition, our results were obtained in young patients with lower-extremity atherosclerotic disease, almost all of whom were smokers. We cannot exclude a specific effect of hyperhomocyst(e)inemia in smokers, and extrapolation of our conclusions to the general population requires caution.
We conclude that the severity of atherosclerosis in young patients with lower-limb atherosclerotic disease is associated with high homocysteine levels. Lowering high homocysteine levels is possible by treatment with folic acid and vitamin B6,20 which stimulate homocysteine remethylation and transsulfuration, respectively. Whether such treatment slows the progression of atherosclerotic disease in patients with lower-limb arterial disease is not known and requires further investigation.
Dr van den Berg is supported by the Praeventiefonds (Prevention Fund). Dr Stehouwer is the recipient of a fellowship from the Netherlands Organisation for Scientific Research (NWO) and the Diabetes Fonds Nederland (Diabetes Research Fund).
Balkau B, Vray M, Eschwege E. Epidemiology of peripheral arterial disease. J Cardiovasc Pharmacol. 1994;23(suppl 3):S8-S16.
Smith GD, Shipley MJ, Rose G. Intermittent claudication, heart disease risk factors, and mortality. Circulation. 1990;82:1925-1931.
Malinow MR, Kang SS, Taylor LM, Wong PKW, Coull B, Inahara T, Mukerjee D, Sexton G, Upson B. Prevalence of hyperhomocyst(e)inemia in patients with peripheral arterial occlusive disease. Circulation. 1989;79:1180-1188.
Ueland PM, Refsum H, Brattström L. Plasma homocysteine and vascular disease. In: Francis RB Jr, ed. Atherosclerotic Vascular Disease, Hemostasis and Endothelial Function. New York, NY: Marcel Dekker Inc; 1992:183-236.
Arnesen E, Refsum H, Bonaa KH, Ueland PM, Forde OH, Nordrehaug JE. Serum total homocysteine and coronary heart disease. Int J Epidemiol. 1995;24:704-709.
Starkebaum G, Harlan JM. Endothelial cell injury due to copper-catalyzed hydrogen peroxide generation from homocysteine. J Clin Invest. 1986;77:1370-1376.
Harker LA, Ross R, Slichter SJ, Scott CR. Homocysteine-induced atherosclerosis: the role of endothelial cell injury and platelet response in its genesis. J Clin Invest. 1976;58:731-741.
Tsai J-C, Perrella MA, Yoshizumi M, Hsieh C-M, Haber E, Schlegel R, Lee M-E. Promotion of vascular smooth muscle cell growth by homocysteine: a link to atherosclerosis. Proc Natl Acad Sci U S A. 1994;91:6369-6373.
Fowkes FRG. The measurement of atherosclerotic peripheral arterial disease in epidemiological surveys. Int J Epidemiol. 1988;17:248-254.
te Poele-Pothoff MTWB, van den Berg M, Franken DG, Boers GHJ, Jakobs C, de Kroon IFI, Eskes TABK, Trijbels JMF, Blom HJ. Three different methods for the determination of total homocysteine in plasma. Ann Clin Biochem. 1995;32:218-220.
Verhoef P, Hennekens CH, Malinow R, Kok FJ, Willet WC, Stampfer MJ. A prospective study of plasma homocyst(e)ine and risk of ischemic stroke. Stroke. 1994;25:1924-1930.
von Eckardstein A, Malinow MR, Upson B, Heinrich J, Schulte H, Schönfeld R, Kohler E, Assmann G. Effects of age, lipoproteins, and hemostatic parameters on the role of homocyst(e)inemia as a cardiovascular risk factor in men. Arterioscler Thromb. 1994;14:460-464.
Malinow MR, Nieto FJ, Szklo M, Chambless LE, Bond G. Carotid artery intimal-medial wall thickening and plasma homocyst(e)ine in asymptomatic adults: the Atherosclerosis Risk in Communities Study. Circulation. 1993;87:1107-1113.
Selhub J, Miller JW. The pathogenesis of homocysteinemia: interruption of the coordinate regulation by S-adenosylmethionine of the remethylation and transsulfuration of homocysteine. Am J Clin Nutr. 1992;55:131-138.
Guttormsen AB, Schneede J, Fiskerstrand T, Ueland PM, Refsum HM. Plasma concentrations of homocysteine and other aminothiol compounds are related to food intake in healthy human subjects. J Nutr. 1994;124:1934-1941.