Several studies have reported an association between hyperhomocysteinemia, 5,10-methylenetetrahydrofolate reductase (MTHFR) polymorphisms and cleft lip with or without cleft palate (CLP), and congenital heart defects (CHDs). However, findings have been inconsistent. A meta-analysis was performed of published studies until September 2006 investigating these associations in both mothers and children. Homocysteine data were provided in two CLP and three CHD studies, and MTHFR polymorphisms were reported in ten CLP and eight CHD studies. Data were analyzed using the random effects model in the Cochrane Review Manager. The pooled odds ratio (OR) of maternal hyperhomocysteinemia was 2.3 (95% CI 0.4-11.9) for CLP, and 4.4 (2.6-7.3) for CHDs. The MTHFR C677T polymorphism and CLP showed pooled ORs of 1.2 (0.9-1.5) in mothers and 1.0 (0.9-1.2) in children, whereas these estimates for the A1298C polymorphism were 1.0 (0.7-1.2) in mothers and 0.9 (0.6-1.2) in children. The MTHFR C677T polymorphism in CHD studies demonstrated a pooled OR of 1.0 (0.8-1.3) for mothers and 1.1 (0.9-1.5) for children. Two studies investigating the maternal A1298C polymorphism in CHDs demonstrated a pooled OR of 1.2 (0.8-1.8). Only one CHD study reported an OR of 1.3 (0.8-2.1) for this polymorphism in children. In conclusion, this meta-analysis demonstrates that maternal hyperhomocysteinemia is a risk factor for CHDs. The MTHFR polymorphisms C677T and A1298C in both mothers and children are not independently associated with CLP or CHDs. Future studies should be performed to investigate the interactions between maternal hyperhomocysteinemia, B-vitamin intake, related polymorphisms and the risk of CLP and CHDs.

Periconceptional food folate and folic acid supplement intake, polymorphisms in the methylenetetrahydrofolatereductase-gene, and maternal hyperhomocysteinemia are associated with the risk of a child with cleft lip and/or palate (CLP) [1].Recently, it has been demonstrated that antiserum to folate receptors induces embryo resorption and malformation in rats [2]. Subsequently, folate receptors autoantibodies were identified in serum of women who had a pregnancy complicated by a neural tube defect [3]. Therefore, the hypothesis is that maternal autoantibodies against folate receptors play a role in the pathogenesis of CLP by affecting the folate and total homocysteine (tHcy) status.From the case-control family study on nonsyndromic CLP conducted in the Netherlands [1], 11 Caucasian case-mothers of a child with CLP and 10 Caucasian control-mothers of a healthy child without structural congenital malformations were selected. At around 14 months after the indexpregnancy standardized maternal blood samples were taken. After blinding of the samples, autoantibodies were determined as described previously [3]. The intra-and interassay coefficients of variations (CV) were 3% and 5%, respectively. Red blood cell folate (RBC) and serum folate were 0020-7292/$ -see front matter D