Death rates from cardiovascular disease have fallen in recent years in New Zealand and most other Western countries.1,2 However, coronary heart disease and stroke continue to be the leading causes of death in this country and in most relatively affluent societies throughout the world.1,2 Thus, the identification of new, readily modifiable, risk factors invariably arouse considerable interest. When prospective epidemiological and animal studies as well as small randomised trials suggested that high intakes of antioxidant nutrients might be protective, there were high hopes that supplements containing such compounds might radically reduce cardiovascular risk.3–6 However, such hopes have been dashed by a series of randomised controlled trials (RCTs) that have convincingly shown that the most widely studied such nutrients – β-carotene, vitamin C, and vitamin E – at least in the amounts usually recommended, were ineffective.7–10 It is against this background that the role of homocysteine (Hcy), a sulphur-containing amino acid formed during the metabolism of methionine, must be considered (Figure 1). (Homocysteine as measured in most clinical studies refers to total homocysteine, which is the sum of free homocysteine, protein-bound homocysteine, homocysteine thiolactone, and mixed disulfides involving homocysteine measured in plasma (or serum).)

Evidence-based medicine demands an impressive array of consistent data before a recommendation is made regarding the measurement and treatment of a putative risk factor.

More than 30 years ago, McCully reported extensive arterial atherosclerosis and thrombosis at post mortem examination in two children with homocystinuria.11 He raised then the possibility that lesser degrees of homocystinaemia than those seen in homocystinuria might be more widely involved in the pathogenesis of atherosclerosis. These observations appear not to have been followed up until Wilcken and Wilcken published their less than convincing, at least by modern standards, case control series in 1976, which suggested that coronary heart disease patients frequently have abnormal Hcy metabolism.12 Indeed, it has only been in the past 10 years or so that more convincing data have started to appear.

The first impressive case control study, undertaken by the European Concerted Action Group and published in 1997, involved 730 cases of atherosclerotic vascular disease (cardiac, cerebral and peripheral) and 800 controls from 19 centres in nine European countries.13 The data showed, given the limitations imposed by the methodology, that Hcy confers an independent risk of vascular disease, similar to that of smoking or hyperlipidaemia, and that risk was more pronounced in smokers and in those with hypertension. There was also some evidence of a gradient (dose response) effect with increasing levels of Hcy. While the majority of case control and cross sectional studies have supported an association between vascular disease and Hcy, findings from prospective cohort studies have been less consistent.14 While prospective cohort studies have the advantage that blood samples are collected before the clinical event, it is also necessary to remember that measurements are usually made long after initial data gathering on stored blood samples, using methods that may not have been adequately validated under such circumstances.

A formal meta-analysis of all case control and cohort studies was published last year.14 For coronary heart disease, summary odds ratios for a 5 μmol/L increase in Hcy concentration were 1.70 for 26 publications of case control studies, 1.23 for 10 publications of nested case controls (based on cohort studies), and 1.06 for two publications of cohort studies. For cerebrovascular disease, the corresponding figures were 2.16 for 17 publications of case control studies, 1.58 for five publications of nested case control, and 1.10 for two publications of cohort studies.

Biological plausibility is an important attribute in assessing whether a putative risk factor is causally related to a disease. There is certainly ample experimental evidence to suggest that Hcy may be causally linked to vascular disease. For example, Hcy can undergo auto-oxidation in the plasma, during which potent reactive oxygen species (ROS), including superoxide and hydrogen peroxide, are generated.15 ROS can cause injury to the endothelial cells, exposing the underlying matrix and smooth muscle cells that proliferate and promote the activation of platelets and leukocytes.16 ROS are also responsible for oxidation of low-density lipoprotein that may be directly involved in the atherothrombotic process but additionally promotes the formation of foam cells, which in turn yield another source of ROS.15

Perhaps the most intriguing feature of Hcy as a potential cardiovascular risk factor is the relative ease in which it may be lowered. Folate, vitamins B6 and B12 are all coenzymes for critical steps in the metabolism of homocysteine, and blood concentrations of these vitamins are inversely associated with Hcy concentrations.17 Innumerable studies have shown that supplementing the diet with folic acid (the synthetic form of folate), and to a lesser extent vitamin B12, or substantially increasing bioavailable dietary sources of folate (eg citrus fruits and legumes), can appreciably reduce blood levels of total Hcy.18,19 A meta-analysis that included data from over 1000 individuals reported that supplementation with folic acid reduced Hcy concentrations by 25% based on a standardised pre-treatment Hcy of 12 µmol/L.18 The addition of vitamin B12 lowered Hcy by an additional 8%, whereas vitamin B6 produced no effect. The dose of these vitamins required to produce a maximum lowering of Hcy remains to be conclusively established. We have found in healthy individuals that 100–200 μg/day of folic acid (ie the amount present in 1–2 servings of folic acid-fortified breakfast cereals) produces an effect not appreciably different from that achieved by much larger amounts.20 However, in patients with renal disease and possibly in those with established cardiovascular disease, much larger amounts may be required and even then it may not be possible to normalise Hcy concentrations.21,22

Several randomised controlled trials with clinical endpoints are underway. One fairly small such study was published this year.23 Schnyder et al reported on 535 patients who, following successful coronary angioplasty, were randomised to placebo or treatment with a combination of folic acid (1 mg), vitamin B12 (400 μg), and pyridoxine (10 mg). Not only were levels of Hcy appreciably reduced (from an average of 11.4 to 7.4 μmol/L at six months) in the treatment group, but after 12 months the need for revascularisation was appreciably and significantly reduced. Further, a non-significant trend was reported towards fewer deaths and non-fatal myocardial infarctions in the group receiving the vitamins.23

Given that an RCT is usually regarded as both the ultimate proof of causality as well as good evidence for action, is there now a case to be made for measurement of Hcy in some or all those at risk of cardiovascular disease? In our opinion, the answer at present is no. The only RCT published to date is small and of short duration.23 The CHAOS study, which showed the benefit of vitamin E supplementation in terms of morbidity if not mortality, was the stimulation for the widespread use of such treatment.5 As is now well known, the results could not be replicated in several much larger subsequent studies and the confusion caused by CHAOS is only now, several years later, being resolved.7–10 Surely, it is more prudent to wait for the results of the much larger trials (including those in high risk patients with chronic renal failure) at present underway. Such trials, if they are positive, should indicate whether it will be appropriate to treat all, or at least most, high risk individuals as has recently been established with regard to cholesterol lowering by means of statin therapy, or whether only those with Hcy above a certain cut-off point should be treated. Further, although the availability of a reliable, simple and automated immunoassay based on Abbott immuno-diagnostic platforms makes Hcy determination possible at most clinical laboratories, there is at present no recognised international or national standardisation of Hcy measurement as there is for cholesterol.24,25 Efforts are underway to standardise Hcy measurement and several external quality programmes are available.26,27

The potential implications of elevated Hcy extend beyond cardiovascular disease. Of considerable interest at present is the potential role for elevated Hcy in deterioration of cognitive function in the elderly.28 Should a trial presently underway in Dunedin demonstrate an improvement in cognitive function as a consequence of treatment with homocysteine-lowering vitamins, and should the several trials being undertaken to determine cardiovascular risk reduction generate positive results,29 there would be a very strong case for mandatory fortification of a staple food such as flour with folic acid. This is especially so, given the potential for this nutrient to reduce the risk of neural tube defect when consumed in appropriate quantities prior to conception and during the early stages of pregnancy.30 However, before this alternative to supplementation is considered, it is first necessary to re-examine the risk posed by excessive intakes of folic acid. A major concern, especially in the elderly, is that folic acid could mask the haematological signs of vitamin B12 deficiency, delaying diagnosis, and allowing progression of neurological damage.31,32 The use of a synthetic alternative to folic acid that more closely resembles natural folate may offer a means of avoiding this risk.33 All these aspects are the subject of intense research in Dunedin and worldwide. Vitamin treatment to lower Hcy may not prove to be the magic bullet to further reduce cardiovascular risk, but homocysteine, the previously considered rather innocuous amino acid, may yet prove to be pivotal in our understanding of several disease processes and its manipulation have the potential to profoundly influence several health outcomes. Watch this spot!

Author information: Jim I Mann, Professor in Human Nutrition and Medicine; Timothy J Green, Lecturer in Human Nutrition, Department of Human Nutrition, University of Otago

Correspondence: Professor J Mann, Department of Human Nutrition, University of Otago, P O Box 56, Dunedin. Fax: (03) 479 7959; email:
jim.mann@stonebow.otago.ac.nz