Background: Data are conflicting concerning risk for ischemic stroke associated with hyperhomocyst(e)inemia (hyper-Hcy) and a common polymorphism in the gene encoding 5,10-methylenetetrahydrofolate reductase (MTHFR 677C-->T), which predisposes to hyper-Hcy in vivo.
Methods: Search of MEDLINE, Science Citation Index, and abstracts of conference proceedings revealed relevant articles. Exposure was defined as follows: 1) prevalence of hyper-Hcy; 2) absolute difference in the mean Hcy concentration between subjects with and without ischemic stroke; and 3) the MTHFR TT genotype frequency. Outcome was defined as ischemic stroke with or without neuroimaging. Inclusion criteria were retrospective and prospective studies with reported odds ratios (OR) or hazard ratios (HR) or arithmetic mean Hcy levels. Exclusion criteria were absence of OR or HR, outcome defined as carotid atherosclerosis or intima-media thickening, stroke in patients younger than 18 years old, and studies in languages other than English. Statistical analyses for between-study heterogeneity and pooled risk estimates were performed using Stata software (Stata Corporation, College Station, TX).
Results: Among 16 studies (1,487 stroke and 2,554 nonstroke cases), the pooled mean Hcy level in patients with ischemic stoke was 2.32 micromol/L (95% CI, 1.6 to 3.04; p < 0.001) greater than that in those without ischemic stroke. Among 14 included studies (1,769 stroke and 7,400 nonstroke cases), the pooled OR estimate of ischemic stroke associated with hyper-Hcy was 1.79 (95% CI, 1.61 to 2.0; p < 0.001). Among 19 included studies (2,788 stroke and 3,962 nonstroke cases), the OR associated with the TT genotype was 1.23 (95% CI, 0.96 to 1.58; p = 0.1).
Conclusion: These data support an association between mild-to-moderate hyper-Hcy and ischemic stoke. The MTHFR TT genotype may have a small influence in determining susceptibility to ischemic stoke.