Purpose: Inconsistency of reported associations between the C1420T polymorphism in the cytosolic serine hydroxymethyltransferase (SHMT1) gene and colorectal cancer (CRC) prompted us to undertake a meta-analysis.
Methods: We conducted searches of published literature in MEDLINE through PubMed up to April 2012. Individual data on 5,043 cases and 6,311 controls from 15 published case-control studies were evaluated. Meta-analyses were performed on the compiled dataset.
Results: In the overall analysis, association was lacking between the C1420T polymorphism and CRC risk (odds ratio [OR] 0.96-1.04, p = 0.47-0.77), materially unchanged when reanalyzed without the Hardy-Weinberg equilibrium-deviating studies (OR 1.03-1.09, p = 0.22-0.55) or subjected to outlier treatment (OR 0.89-0.99, p = 0.10-0.8). In the ethnic subgroups, Europeans were susceptible (OR 1.11-1.17, p = 0.13-0.48) and Americans, slightly protected (OR 0.86-0.87, p = 0.49-0.61). The increased risk effects, however, became null following outlier treatment (OR 0.95-1.06). Test for interaction between decreased risk associations in the low-folate subgroup (OR 0.60-0.85, p = 0.009-0.03) with the susceptible effects in the high-folate category (OR 1.14-1.22, p = 0.19-0.32) was significant (p interaction = 0.004).
Conclusions: Overall summary estimates imply no associations but suggest geography-specific effects of the SHMT1 polymorphism that render Europeans susceptible, but not Americans. Folate status appears to show an inverse association of this polymorphism with CRC.