Homocysteine metabolism. SAM indicates S-adenosyl methionine; SAH, S-adenosyl homocysteine; CBS, cystathionine β-synthase; MS, methionine synthase; 5 MTHF, 5-methyltetrahydrofolate; MTHFR, 5-methyltetrahydrofolate reductase; NADPH, nictotinamide adenine dinucleotide phosphate; and TH Folate, tetrahydrofolate.
Treatment of hyperhomocysteinemia. tHcy indicates total homocysteine; QD, every day.
Stein JH, McBride PE. Hyperhomocysteinemia and Atherosclerotic Vascular DiseasePathophysiology, Screening, and Treatment. Arch Intern Med. 1998;158(12):1301-1306. doi:10.1001/archinte.158.12.1301
Hyperhomocysteinemia has recently been identified as an important risk factor for atherosclerotic vascular disease. This article reviews homocysteine metabolism, causes of hyperhomocysteinemia, the pathophysiological findings of this disorder, and epidemiological studies of homocysteine and vascular disease. Screening for hyperhomocysteinemia should be considered for patients at high risk for vascular disease or abnormalities of homocysteine metabolism. For primary prevention of vascular disease, treatment of patients with homocysteine levels of 14 µmol/L or higher should be considered. For secondary prevention, treatment of patients with homocysteine levels of 11 µmol/L or higher should be considered. Treatment is most conveniently administered as a folic acid supplement (400-1000 µg) and a high-potency multivitamin that contains at least 400 µg of folate. Higher doses of folic acid and cyanocobalamin supplements may be required in some patients. Until prospective clinical trial data become available, these conservative recommendations provide a safe, effective, and evidence-based approach to the diagnosis, evaluation, and management of patients with hyperhomocysteinemia.
The observation that up to 19% of patients in large clinical trials of low-density lipoprotein cholesterol–lowering therapy experienced adverse cardiovascular events, despite this powerful intervention, has intensified the search for new, nonlipid risk factors for atherosclerotic vascular disease (ASVD).1- 4 Hyperhomocysteinemia, a metabolic abnormality that can be detected in up to 30% of patients with coronary artery disease (CAD) and 42% of patients with cerebrovascular disease, recently has been identified as an important risk factor for ASVD.4- 6 The purpose of this article is to provide an evidence-based approach to the diagnosis and management of patients with hyperhomocysteinemia. In this context, homocysteine metabolism, the causes of hyperhomocysteinemia, and the pathophysiological findings of this disorder are reviewed, as are the epidemiological studies that have implicated hyperhomocysteinemia as a predictor of increased risk of ASVD. On the basis of these discussions, practical recommendations for the screening and treatment of patients with this disorder are provided.
Homocysteine is an amino acid intermediate in the metabolism of methionine, an essential amino acid found in both animal and plant proteins (Figure 1). The recommended daily allowance of methionine is 0.9 g; however, the average American diet contains approximately 2 g/d of methionine, the excess of which is converted via enzymatic transmethylation to homocysteine.6 Homocysteine is converted to cystathionine via a transsulfuration pathway that is dependent on the vitamin B6–dependent enzyme cystathionine β-synthase.6- 8 Cystathionine is then converted into cysteine, which is eventually degraded and excreted in the urine. Homocysteine also may be recycled back into methionine by either of 2 remethylation pathways, the most important of which involves the vitamin B12–dependent enzyme methionine synthase and its cosubstrate, 5-methyltetrahydrofolate.6- 8 The other remethylation pathway is independent of vitamin B12 and folate but uses betaine as a cofactor.6- 8
Hyperhomocysteinemia may result from abnormalities in the function of any of the enzymes involved in homocysteine metabolism or from deficiencies of enzyme cofactors or cosubstrates (ie, folate, vitamin B6, or vitamin B12) (Table 1).
Diminished activity of methionine synthase or 5-methyltetrahydrofolate reductase because of genetic abnormalities, vitamin deficiencies, or medication use may cause hyperhomocysteinemia.6,8- 11 Indeed, the most common form of genetic hyperhomocysteinemia results from production of a thermolabile variant of 5-methyltetrahydrofolate reductase with decreased activity.12 Homozygosity for this mutant enzyme is present in 9% to 17% of the population, and heterozygosity can be detected in 30% to 41% of the general population.12- 16 Homozygous cystathionine β-synthase deficiency causes homocystinuria, a rare disorder characterized by mental retardation, arterial and venous thrombosis, and premature atherosclerosis. Affected individuals may have plasma total homocysteine (tHcy) levels as high as 300 to 500 µmol/L.6,8,9 Although homocystinuria is a rare disorder, heterozygous deficiency of this enzyme is prevalent in the United States at a frequency of 1 per 300 and may cause moderately elevated tHcy levels.9
Because these vitamins function as cofactors or cosubstrates in the enzymatic reactions involved in homocysteine metabolism, severe hyperhomocysteinemia may be detected in patients with low levels of folate or vitamin B12, even when serum levels of these vitamins are in the low-normal range.17- 23 The association between vitamin B6 deficiency and hyperhomocysteinemia is less clear; however, vitamin B6 and homocysteine levels seem to be inversely related.17,22,24
Clinically, tHcy levels tend to increase with older age and tobacco use.25- 27 Men tend to have higher tHcy levels than women, and tHcy levels tend to be elevated in individuals with renal dysfunction, unexplained deep venous thrombosis, systemic lupus erythematosus, malignant neoplasms, psoriasis, and solid organ transplantation.6,25- 36 After a myocardial infarction or a cerebrovascular accident, tHcy levels are subject to an acute-phase response, characterized by an initial reduction of approximately 25%, followed by a convalescent increase of approximately 20% to 22%, that can interfere with interpretation of laboratory values obtained up to 3 months after these events.7,37,38 The effect of surgery and other systemic illnesses on tHcy levels has not been well characterized. Finally, several commonly used medications, including methotrexate, nitrous oxide, phenytoin, carbamazepine, nicotinic acid, colestipol, and thiazide diuretics, increase tHcy levels.6,28,34,39
The vascular and hematologic abnormalities associated with hyperhomocysteinemia lead to a proatherogenic and prothrombotic metabolic milieu (Table 2).6,8 These abnormalities include (1) endothelial cell injury, the initial event in the development of atherosclerosis, manifested as impaired endothelium-dependent vasodilation and impaired endogenous tissue-type plasminogen activator activity; (2) increased platelet aggregation, related to increased synthesis of thromboxane A2 and decreased synthesis of prostacyclin; and (3) abnormalities of the clotting cascade, such as activation of factors V, X, and XII, and inhibition of natural anticoagulants, such as antithrombin III and factor C.6,8,40 Homocysteine promotes the binding of lipoprotein(a) to fibrin and the growth of smooth muscle cells, and tHcy levels correlate with levels of fibrinogen, an independent risk factor for ASVD.41- 44
Several studies have demonstrated strong associations between hyperhomocysteinemia and CAD (especially premature CAD),5,7,17,23,45- 50 cerebrovascular disease,5,45,50- 53 and peripheral arterial vascular disease.5,45,50,53
One of the earliest studies that related hyperhomocysteinemia and ASVD was conducted in Dublin, Ireland.5 Hyperhomocysteinemia, defined as a tHcy level greater than the high threshold level of 24 mmol/L, was more predictive of ASVD than any other risk factor.5 After adjustment for hypercholesterolemia, hypertension, and tobacco abuse, hyperhomocysteinemia was associated with an overall odds ratio (OR; lower limits of 95% confidence limits) for ASVD of 1.39.
In the Physicians' Health Study, 14916 male physicians without known ASVD were followed up prospectively for 5 years.46 Plasma tHcy levels were higher in the 271 patients with myocardial infarction than in healthy controls, and the adjusted risk for the highest fifth percentile vs the bottom 90th percentile of tHcy levels was 3.1.46 This finding was verified by a study of 21826 patients in Tromso, Norway, in which the relative risk of CAD increased by 1.32 for each 4-µmol/L increase in tHcy levels.47 The increased risk associated with increasing tHcy levels was observed in both males (10983 patients) and females (10863 patients) and across a wide range of ages (12-61 years) and tHcy values (3.5-35.0 µmol/L).47 Furthermore, no threshold effect was observed.47
In the British Regional Heart Study,52 5661 middle-aged men from the United Kingdom were followed up prospectively for up to 13 years. Plasma tHcy levels were higher in the 109 patients with stroke than in controls, and the adjusted relative risk of cerebrovascular disease for the highest quartile of tHcy levels was 2.8.52 These data were verified by the Framingham Heart Study of 1041 patients, in which the OR of having significant carotid artery stenosis was 2.0 for patients with tHcy levels greater than 14.4 µmol/L.51 A critical observation in this study was that tHcy elevations within the "normal" range (11.4-14.3 µmol/L) also were associated with increased risk of cerebrovascular disease (OR=1.6).51 Furthermore, the risk of significant carotid artery stenosis was associated with decreasing folate levels.51
Recently, the European Concerted Action Project50 addressed the interaction between hyperhomocysteinemia and conventional risk factors for ASVD in a study of 750 case and 800 control patients younger than 60 years.50 The risk for ASVD in patients in the highest quintile of tHcy levels (≥12 µmol/L) relative to the lower quintiles was 2.2. The risk of increasing tHcy levels was continuous, and a 5-µmol/L increment in the tHcy level was associated with an OR for ASVD of 1.3 for men and 1.4 for women. The magnitude of increased risk of ASVD associated with hyperhomocysteinemia was greater than that associated with hypercholesterolemia (OR=1.4), less than that for hypertension (OR=3.9), and similar to that for tobacco use (2.2). For both sexes, the ASVD risk associated with hyperhomocysteinemia was multiplicative in the presence of tobacco use or hypertension.50
The association between CAD and decreasing folate levels also has been verified in several studies.7,17,22,23,31 Significant but less consistent and less powerful associations between CAD and low levels of vitamins B6 and B12 also have been described.7,17,22- 24,31,50
Regarding the CAD risk attributed to homocysteine levels, a recent meta-analysis established that a 5-µmol/L increase in the tHcy level was equivalent to an approximately 0.52-mmol/L (20-mg/dL) increase in the total cholesterol level (ie, 20% increased risk).8 The relative risk for CAD associated with increasing tHcy levels in this meta-analysis8 was lower than reported in the European Concerted Action Project in which a 5-µmol/L change was equivalent to a 2.22 mmol/L (86 mg/dL) change in cholesterol.50,54
Most clinical laboratories use high-performance liquid chromatography to measure plasma tHcy levels, although some laboratories still use older immunoassays. Although both techniques are sensitive, the newer techniques are more specific. Charges for these assays are typically $45 to $100 per measurement. Blood samples should be collected in tubes containing an anticoagulant such as EDTA and centrifuged within 30 minutes to avoid a false elevation of homocysteine levels due to its release from red blood cells. Accordingly, serum homocysteine measurements have a higher normal range. Samples may then be refrigerated or frozen for several weeks. Free homocysteine levels are not useful clinically.
The "normal" range for fasting tHcy values is approximately 5 to 15 µmol/L. The upper limit of this range, however, should be revised downward because the increased risk of atherosclerosis associated with tHcy levels in this range has been well documented.6,8,17,47,50,51 Certain patients with abnormal homocysteine metabolism may have normal fasting plasma tHcy levels and may require provocative testing (ie, methionine loading) to expose this abnormality.6,8,50 The methionine loading test involves oral administration of 100 mg/kg of methionine and measurement of the plasma tHcy level 6 to 8 hours later.6,50 The European Concerted Action Project50 reported that methionine loading identified an additional 27% of cases, despite a strong correlation between baseline and postload tHcy values. The role of methionine loading in clinical practice is controversial, however, because of its excess cost and time commitment and interindividual variation in the time-to-peak tHcy response to methionine loading. Because most of the recent epidemiological studies that associate hyperhomocysteinemia with ASVD only measured fasting plasma tHcy levels, maintaining this practice seems reasonable, especially until the clinical utility and cost-effectiveness of methionine loading is established. The sensitivity of fasting plasma tHcy levels for identification of patients at risk for ASVD may be improved by using a lower criterion than that used in the European Concerted Action Project (12 µmol/L),50 especially for high-risk patients, such as those with established ASVD.
Screening for hyperhomocysteinemia (Table 3) should be considered for patients at risk for this disorder and for patients at high risk for ASVD who may benefit from homocysteine-lowering therapy (discussed below). Such patients include those with (1) ASVD without conventional risk factors; (2) premature ASVD (clinical ASVD event before age 60 years); (3) premature ASVD in a first-degree relative or ASVD risk factors associated with hyperhomocysteinemia, such as tobacco abuse and hypertension; (4) chronic renal failure; (5) unexplained deep venous thrombosis; (6) systemic lupus erythematosus; (7) severe psoriasis; (8) solid organ transplantation; and (9) prolonged use of medications associated with hyperhomocysteinemia (Table 1). Screening for hyperhomocysteinemia should be deferred until 8 to 12 weeks after any serious systemic illness, including myocardial infarction or stoke, because of the potential for spurious underestimations of baseline tHcy levels.7,37,38
Clinical trial data regarding the potential benefits of homocysteine-lowering therapy for the prevention of ASVD are not yet available. In the European Concerted Action Project,50 however, users of multivitamins containing folic acid, pyridoxine hydrochloride, and cyanocobalamin had a significantly lower risk of all forms of ASVD than nonusers, even after adjustment for conventional risk factors (relative risk, 0.38), and at least some of this effect was attributable to lower tHcy levels. A similar observation was made in the Nurse's Health Study, where participants who used moderate doses of vitamin supplements containing folate and pyridoxine had fewer coronary events than nonusers.55
Therapy with folic acid at doses of 400 µg or more daily has been associated with a 30% to 42% decrease in tHcy levels.56,57 Doses lower than this do not result in a sustained reduction in tHcy levels.6,56 Cyanocobalamin supplementation has been associated with a less dramatic reduction in tHcy levels of approximately 15%.56 In the absence of vitamin B6 deficiency, pyridoxine hydrochloride supplementation does not lower tHcy levels significantly.53,56 Combination therapy with all 3 of these vitamins, administered orally or as a monthly intramuscular injection, has been associated with 15% to 72% reductions in tHcy levels.6,22,56,58 It is doubtful that pyridoxine hydrochloride or cyanocobalamin supplementation has an additional homocysteine-lowering effect in addition to folic acid supplementation in patients without deficiencies of these vitamins.56 The US Food and Drug Administration recently mandated fortification of flours and cereal products with 140 µg of folic acid per 100 g. This intervention is expected to have a population effect of lowering tHcy levels by an average of approximately 3 µmol/L and may potentially prevent 17000 deaths due to ASVD each year.45,59
The safety of folic acid supplementation has been well documented.59,60 Exacerbation of vitamin B12 deficiency has been reported rarely and only occurs with higher doses of folic acid (≥10 mg/d).60 Although folic acid supplementation may mask the red blood cell macrocytosis associated with vitamin B12 deficiency, the red blood cell mean corpuscular volume is an insensitive indicator of vitamin B12 deficiency, and more reliable indicators of vitamin B12 status are available, such as direct measurement of vitamin B12 and methylmalonic acid levels.60 Furthermore, folic acid supplementation does not mask the neurologic or cutaneous manifestations of vitamin B12 deficiency, exacerbate seizure disorders, interfere with effectiveness of antifolate medications, or increase the risk of cancer.60
In the absence of clinical trial evidence that treatment of hyperhomocysteinemia decreases the risk of ASVD or its clinical manifestations, therapy for this disorder is based on the strong epidemiological association between increasing tHcy levels and ASVD and on the safety and low cost of folic acid supplementation. There is also preliminary evidence that vitamin therapy may prevent progression of cerebral atherosclerosis.61 The optimal dose, combination, and route of administration of homocysteine-reducing therapies have not yet been clarified. Because of these limitations, the following recommendations (Figure 2) were formulated to provide a safe and effective strategy for addressing hyperhomocysteinemia in high-risk patients until clinical trials addressing these issues are completed.
For the primary prevention of ASVD, screening for hyperhomocysteinemia should be considered in patients at risk for this disorder and for patients at high risk for ASVD who may benefit from homocysteine-lowering therapy (Table 3). Special attention should be given to patients who smoke tobacco or have hypertension because of the association between tobacco use, hypertension, hyperhomocysteinemia, and premature ASVD.26,50 Treatment of patients with tHcy levels of 14 µmol/L or higher, a level near the 95th percentile in the Framingham Heart Study and clearly associated with an increased risk of ASVD, should be considered.7,17,45,50,51 For the secondary prevention of ASVD, a more aggressive approach to screening and therapy may be warranted, as manifest ASVD dramatically increases the risk of vascular death.1,2 In patients with established ASVD, consideration should be given to treating patients with tHcy levels of 11 µmol/L or higher, a level clearly associated with increased vascular events in the Framingham Heart Study and other studies but previously considered normal.7,17,45,50,51
Because of the small risk that folic acid supplements may delay the diagnosis or potentiate the neurologic manifestations of vitamin B12 deficiency, which is relatively common in older patients, we conservatively recommend measurement of serum vitamin B12 levels in all patients with hyperhomocysteinemia, as defined by the clinical settings described above.58- 60 If vitamin B12 deficiency is identified it should be evaluated, and cyanocobalamin supplementation may lower tHcy levels to an acceptable level.
If hyperhomocysteinemia persists after vitamin B12 deficiency has been excluded or treated, and offending medications have been discontinued (if possible), folic acid supplementation may be initiated as primary therapy. This is most conveniently administered as a folic acid supplement (400-1000 µg) and a high-potency multivitamin (without iron for males and postmenopausal females) that contains at least 400 µg of folic acid and the US recommended daily allowance of pyridoxine hydrochloride and cyanocobalamin. This ensures adequate intake of pyridoxine hydrochloride and cyanocobalamin during folic acid supplementation.
Levels of tHcy should be remeasured 6 to 8 weeks after initiation of homocysteine-lowering therapy. If elevated levels persist despite therapy as described above, the dose of the folic acid supplement may be increased to 2000 µg/d, with another measurement of the tHcy level after 6 to 8 additional weeks. Higher doses of folic acid (up to 5000 µg/d) may be needed for patients with end-stage renal disease or continuing sources of folic acid loss. If hyperhomocysteinemia persists, cyanocobalamin supplementation (400 µg/d), even in the absence of overt deficiency, may be beneficial. Although it is an uncommon disorder, vitamin B6 deficiency should be considered in patients with marked hyperhomocysteinemia (tHcy level ≥24 µmol/L) or who are refractory to folic acid therapy.
Although estrogen replacement and penicillamine therapies have reduced tHcy levels in small studies, therapy with these medications for the sole purpose of reducing homocysteine levels cannot be recommended.28,62 Administration of betaine, a choline derivative that functions as a cofactor in the non–vitamin B12–dependent remethylation pathway of homocysteine metabolism, effectively lowers tHcy levels in patients with homocystinuria who are unresponsive to pyridoxine hydrochloride supplementation and may prevent arterial and venous thrombotic events in these patients.63,64 Use of this orphan drug in patients with less severe hyperhomocysteinemia is not recommended because of its expense and the availability and effectiveness of the therapies described above.
These recommendations for screening and treating patients with hyperhomocysteinemia, for the purposes of preventing ASVD and its complications, may be modified when more epidemiological data and prospective clinical trial data, such as from the Heart Protection Study II, become available. Until then, these conservative recommendations provide a safe, effective, and evidence-based approach to the diagnosis, evaluation, and management of patients with hyperhomocysteinemia.
Accepted for publication November 4, 1997.
Reprints: James H. Stein, MD, University of Wisconsin Medical School, 600 Highland Ave, H6/315 CSC, Madison, WI 53792 (e-mail: firstname.lastname@example.org).