Editorials

Homocysteine, the kidney, and vascular disease

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e3925 (Published 13 June 2012) Cite this as: BMJ 2012;344:e3925
  1. Richard Haynes, research fellow ,
  2. Robert Clarke, reader in epidemiology and public health medicine
  1. 1Clinical Trial Service Unit and Epidemiological Studies Unit, Oxford OX3 7LF, UK
  1. richard.haynes{at}ctsu.ox.ac.uk

Moderate differences in homocysteine concentrations do not cause vascular disease

Chronic kidney disease is common and those who have it are at substantially higher risk of cardiovascular disease.1 The association between these two diseases is partly explained by their shared causes (such as diabetes) and by disturbances in known vascular risk factors caused by chronic kidney disease (such as higher blood pressure and altered lipid metabolism).2 However, these risk factors do not seem to fully explain the excess risk, and other risk factors, including homocysteine, have been implicated. The linked meta-analysis by Jardine and colleagues (doi:10.1136/bmj.e3533) examines the relevance for cardiovascular disease risk of lowering blood homocysteine concentrations in people with chronic kidney disease.3 If low cost and effective interventions to reduce homocysteine (such as folic acid supplements) could also reduce risk of cardiovascular disease in these patients then the benefits for public health could be substantial (in addition to explaining the link between these diseases).

The homocysteine hypothesis of coronary heart disease was initially prompted by observations of occlusive vascular disease in children with extreme increases in plasma homocysteine (>100 µmol/L; about 10 times the normal value).4 A meta-analysis of retrospective observational studies in 1995 reported that a 5 µmol/L higher concentration of homocysteine was associated with a 60-80% higher risk of coronary heart disease.5 However, a meta-analysis of prospective observational studies in 2002 reported that after adjustment for known cardiovascular disease risk factors, a homocysteine concentration 25% lower than usual (about 3 µmol/L, a difference typically expected with folic acid supplementation) was associated with a more modest 11% lower risk of coronary heart disease.6

The enzyme methylenetetrahydrofolate reductase, encoded by the MTHFR gene, uses folate to metabolise and remove homocysteine. Genetic variants in MTHFR result in lifelong differences in homocysteine concentrations, and “Mendelian randomisation” studies, which rely on the random assortment of alleles during meiosis, can provide an unbiased assessment of causality. The initial meta-analysis of MTHFR and coronary heart disease in 2002 reported that people who were TT rather than CC homozygotes for the MTHFR C677T polymorphism had a 16% (95% confidence interval 5% to 28%) higher risk of coronary heart disease.7 The apparently concordant results of the observational and genetic studies increased interest in the results of randomised trials of homocysteine lowering B vitamins in people at risk of vascular disease (including those with chronic kidney disease).

However, the meta-analysis of eight large trials involving 37 485 people carried out by the B vitamin Treatment Trialists’ (BVTT) collaboration reported in 2010 that, on average, lowering homocysteine concentrations by 25% for five years had no effect on risk of major vascular events (relative risk 1.01, 0.97 to 1.05).8 Although some of the trials (and previous meta-analyses) had reported significant effects for some vascular outcomes or in particular subgroups, this meta-analysis robustly refuted any beneficial (or hazardous) effects of lowering homocysteine concentrations with folic acid treatment on cardiovascular disease, cancer, and all cause mortality.

The current meta-analysis examined the effect on cardiovascular disease in people with chronic kidney disease.3 The authors identified 11 trials of 10 951 people with chronic kidney disease and extracted data from the published reports. Given the greater exposure to homocysteine and higher risks of cardiovascular disease in people with chronic kidney disease, it was expected that this population might benefit from folic acid supplementation, but the results were again resoundingly null (risk ratio 0.97, 0.92 to 1.03). Indeed, the new results are consistent with the lack of heterogeneity in the effects of vitamin B treatment on cardiovascular disease by baseline kidney function in the BVTT meta-analysis. Many patients with end stage renal disease still receive folic acid supplements, and this meta-analysis shows the futility of such treatment. It also illustrates the requirement for large scale randomised trials to identify effective and safe strategies to reduce vascular risk.

An updated meta-analysis of MTHFR and coronary heart disease in 2012, involving 19 unpublished datasets, reported no association of MTHFR genotype with risk of coronary heart disease (odds ratio 1.02, 0.98 to 1.07).9 By contrast, a meta-analysis of published studies (28 617 cases of coronary heart disease) had suggested an odds ratio of 1.15 (1.09 to 1.21). The discrepant results of MTHFR studies reflect the effects of publication bias and other methodological problems.9 After two decades of research, meta-analyses of randomised trials of folic acid and the unbiased genetic studies have convincingly shown that moderate differences in homocysteine concentrations are not causally relevant to vascular disease.

Notes

Cite this as: BMJ 2012;344:e3925

Footnotes

  • Research, doi:10.1136/bmj.e3533
  • Competing interests: Both authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.

References