Methylenetetrahydrofolate reductase A1298C genetic variant & risk of schizophrenia: A meta-analysis

Schizophrenia is a psychiatric disorder with the prevalence rate of 1 per cent12. According to the Diagnostic and Statistical Manual of Mental Disorders IV criteria (DSM-IV)3, the diagnosis is based on symptoms such as delusions, hallucinations and affective flattening and alogia. The most widely considered hypothesis of schizophrenia is neurodevelopment hypothesis which integrates influences of environment and genes. The neurodevelopment hypothesis of schizophrenia postulates that altered biochemical pathways and defective neural circuitry during foetal brain development lead to cognitive and emotional defects in later part of life4. In the brain of schizophrenics, several neurotransmitter network systems have been found to be defective. Numerous schizophrenia candidate genes have been reported4567.

Folate is an essential nutrient required for de novo nucleotide synthesis and DNA methylation and it has been associated with DNA hypomethylation8910 and DNA strand breaks11. Methylenetetrahydrofolate (MTHFR) is an important enzyme of folate/homocysteine metabolism which catalyzes the synthesis of 5-methyltetrahydrofolate. 5-methyltetrahydrofolate donates methyl group for remethylation of homocysteine to methionine, which is a precursor of S-adenosylmethionine (SAM). SAM is the main methyl donor of several cellular methylation reactions. MTHFR polymorphism raises the dietary requirement for folic acid to maintain normal re-methylation of homocysteine to methionine12; consequently, low folate status in individuals with the MTHFR polymorphism results in an increase in homocysteine and a decrease in methionine levels. A deficiency in cellular folates and methyl-donors may be associated with abnormal methylation of DNA, proteins and neurotransmitters and DNA strand breaks1314151617. If an adequate amount of folate is not available concentration of intracellular homocysteine is increased, and consequently, methionine synthesis is reduced18. SAM is the only source of methyl group for DNA methylation reactions in the brain, and this forms a plausible biologic explanation for potential associations between MTHFR gene polymorphism and schizophrenia risk19.

In MTHFR gene, more than 40 polymorphisms have been reported, of which two mutations C677T20 and A1298C2122 have been extensively studied. It has been reported that the C677T mutation decreases MTHFR activity and increases the plasma homocysteine level while the second polymorphism A1298C results in reduced enzymatic activity but to a lesser extent than the C677T mutation22. The MTHFR A1298C polymorphism is associated with a reduction in the bioavailability of cellular folate23. The mutant genotype (CC) prevalence ranges from 7 to 12 per cent in Caucasian, 4-5 per cent in Hispanic and 1-4 per cent in Asian populations2425.

Several case-control studies investigating the association between MTHFR gene A1298C polymorphism and schizophrenia have been published, but the results were conflicting. Some reported positive association262728293031 but others did not find any association between A1298C polymorphism and schizophrenia32333435. To shed some light on these controversial results, we performed a meta-analysis of published case-control studies, and a pooled estimate of the A1298C polymorphism as risk for schizophrenia was obtained for the allele contrast, homozygote, dominant, co-dominant and recessive models.

Material & Methods

Selection of studies: PubMed, Google Scholar, Science Direct and Springer Link databases were searched for the studies which investigated A1298C polymorphism as risk factor for schizophrenia up to June 2014. The articles were searched using the following terms - ’methylenetetrahydrofolate reductase,’ ’MTHFR’, ’A1298C’ and ’Schizophrenia’. Studies published between 2003 and 2012 were included in the present meta-analysis. Eligible studies met the following criteria: (i) the study should be a case-control association study, (ii) patients were diagnosed by psychiatrists according to the DSM-IV, (iii) the study should report the sample size, distribution of MTHFR A1298C genotypes, or other information necessary for estimation of the odds ratio (OR) and 95 per cent confidence interval (CI), and (iv) the studies should be published. The following exclusion criteria were used: (i) studies containing duplicate data, (ii) no usable data reported, and (iii) case reports, letter to editor, book chapters or reviews.

Data extraction: The following information was extracted from each study - first author's family name, journal name, year of publication, country name, number of cases and controls and the distribution of genotypes. If any author reported the results on different populations, each population was included as an independent study. The allele numbers were calculated from the corresponding genotype distributions.

Statistical analysis: Pooled OR with its corresponding 95 per cent CI were calculated to investigate the association between MTHFR A1298C polymorphism and risk of schizophrenia for the allele contrast (C vs. A), homozygote (CC vs. AA), heterozygote (AC vs. AA), recessive (CC vs. AC+AA) and dominant (CC+AC vs. AA) models. The degree of heterogeneity between the studies was assessed by the Q-test based on Chi-squared statistic36. We also measured the effect of heterogeneity by another measure, I²=100%×(Q−df)/Q37. I² ranges between 0 and 100 per cent, and I² values of 25, 50 and 75 per cent were defined as low, moderate and high estimates of heterogeneity, respectively. The pooled OR was estimated using both fixed effects (FE) and random effects models3839. Whether the genotype distributions of controls were in Hardy–Weinberg equilibrium (HWE) was evaluated in control genotypes of each study.

Publication bias: Publication bias was estimated by funnel plot and Egger's linear regression test40. All statistical analyses were done by software MIX version 1.741.

Results

Characteristic analysis of eligible studies: A total of 49 studies were retrieved after PubMed, Google Scholars, Science Direct and Springer Link databases. After reviewing abstracts of all retrieved studies, 24 were excluded and full text of the remaining 25 articles was considered. Of the 25 studies, four were review articles and one was not relevant for present meta-analysis, so five studies were excluded. A total of 19 studies (Fig. 1) that investigated the association between A1298C polymorphism and schizophrenia were found suitable for the inclusion in the present meta-analysis26272829303132333435424344454647. Two authors2633 studied different populations; we included data from each population as an independent study. All 19 studies were performed in different countries - Korea294546, China263047, Romania3444, Turkey4243, Bulgaria35, Denmark33, Germany27, Norway33, Poland32, Scotland26, Spain28, Sweden33 and Syria31 Table I. Except two studies2729, genotype distribution of control population in all studies was in HWE.

Close

Fig. 1

Flow diagram showing selection of studies.

Export to PPT

Table I Details of included studied (n=19)

In all 19 studies, total cases were 4049 with AA (2127), AC (1518) and CC (404), and controls were 5488 with AA (2950), AC (2045) and CC (493) genotypes. In controls’ genotypes, the percentage of AA, AC and CC were 53.75, 37.26 and 8.98 per cent, respectively. In total cases, genotype percentages of AA, AC and CC were 52.53, 37.49 and 9.98 per cent, respectively (Table II).

Table II Methylenetetrahydrofolate reductase A1298C genotypes and allele number for cases and controls in all included studies

The results of the overall meta-analysis and the heterogeneity test are shown in Table III. A1298C polymorphism was significantly associated with schizophrenia under all five models [C vs. A (allele contrast model): OR=1.13, 95% CI=1.02-1.26, P=0.02 (Fig. 2); CC vs. AA (homozygote model): OR=1.20, 95% CI=1.03-1.39, P=0.02 (Fig. 3); CC vs. AA+AC (recessive model): OR=1.17, 95% CI=1.01-1.35, P=0.04; AC vs. AA (co-dominant model): OR=1.13, 95% CI=1.03-1.23, P=0.009; CC+AC vs. AA (dominant model): OR=1.14, 95% CI=1.05-1.24, P=0.002) (Fig. 4)].

Table III Pooled odds ratio of methylenetetrahydrofolate reductase A1298C in different allele and genotype contrasts, P value of Q-test, the I² metric and P value of Egger test in all included studies

Close

Fig. 2

Random effect forest plot showing a significant association between methylenetetrahydrofolate reductase (MTHFR) A1298C polymorphism and risk of schizophrenia using allele contrast model.

Export to PPT

Close

Fig. 3

Fixed effect forest plot showing a significant association between methylenetetrahydrofolate reductase (MTHFR) A1298C polymorphism and risk of schizophrenia using homozygote model. Only 18 studies that had homozygous mutants were included.

Export to PPT

Close

Fig. 4

Forest plots (random effect) showing no association between methylenetetrahydrofolate reductase A1298C polymorphism and risk of schizophrenia using dominant model (CC+AC vs. AA).

Export to PPT

Subgroup analysis: Subgroup analysis based on ethnicity was performed. Of the 19 studies included in the present meta-analysis, seven were from Asian, and 12 from Caucasian populations. In Asian population, allele contrast meta-analysis showed significant association adopting fixed (OR=1.15; 95% CI=1.03-1.29; P=0.01; I²=48.61%; P_{heterogeneity}=0.07; P_Pb=0.38) effect model. Combined mutant genotypes also showed significant association with fixed (OR=1.17; 95% CI=1.02-1.35; P=0.02) and random (OR=1.18; 95% CI=0.99-1.41; P=0.06) effect models. In this subgroup, heterogeneity between studies (I²=33.91%; P_{heterogeneity}=0.17) was moderate and publication bias (P_Pb=0.51) was absent (Table III).

In Caucasian population, allele contrast meta-analysis showed significant association with FE model (OR=1.10; 95% CI=1.01-1.19; P=0.02) with significant heterogeneity (I²=59.66%; P_{heterogeneity}=0.004). The combined mutant genotype showed significant association with FE model (OR=1.12; 95% CI=1.01-1.24; P=0.03) with moderate heterogeneity (I²=57.39%; P_{heterogeneity}=0.007) and no publication bias (P_Pb=0.23) (Table III).

Publication bias: Publication bias was not observed as funnel plots were symmetrical (Fig. 5A–F). In addition, results of Eggers also did not suggest any evidence of publication bias (P=0.77 for C vs. A with additive model; P = 0.93 for AC vs. AA with co-dominant model; P=0.41 for CC vs. AA with homozygote model; P=0.66 for CC+AC vs. AA with dominant model; P=0.26 for CC vs. AA+AC with recessive model).

Close

Fig. 5

(A) Funnel plot of precision by log odds ratio (allele contrast model); (B) Funnel plot of standard error by log odds ratio (allele contrast model); (C) Funnel plot of precision by log odds ratio (homozygote model); (D) Funnel plot of standard error by log odds ratio (homozygote model); (E) Funnel plot of precision by log odds ratio (dominant model); (F) Funnel plot of standard error by log odds ratio (dominant model).

Export to PPT

Discussion

Several obstetric complications such as foetal hypoxia, maternal stress, infection, bleeding during pregnancy and pre-eclampsia are the risk factors for psychiatric disorders in offspring484950515253. In addition, perinatal folate deficiency and higher plasma concentration of homocysteine in mother have also been reported as a risk factor for schizophrenia54. Folate supplies the substrate for MTHFR reaction. During reduced serum folate and dysfunctional MTHFR, there is a profound deficit in MTHFR-dependent cellular and physiological processes. MTHFR enzyme activity is reduced in heterozygous (AC) and homozygous (CC) individuals, which leads to hyperhomocysteinemia22.

Folate deficiency along with MTHFR polymorphism leads to DNA hypomethylation55565758, and abnormal embryogenesis, which results into developmental malformation in foetus. The neurons in the developing brain are vulnerable to the effect of folate deficiency59. Two main metabolic pathways of homocysteine elimination, i.e., betaine remethylation and transsulfuration are not found in brain6061. Hence, MTHFR hypofunction/folate deficiency increases the concentration of homocysteine in brain which is toxic to neurons and blood vessels and can induce DNA strand breakage, oxidative stress and apoptosis61626364656667.

Several meta-analyses have been published accessing folate pathway genes polymorphism as risk factor for various diseases/disorders such as neural tube defects68, cleft lip and palate69, Down syndrome70, hyperurecemia71, autism72, bipolar disorder73, depression74, Alzheimer's disease75, recurrent pregnancy loss76 and cancer7778. Twelve meta-analyses have been published regarding MTHFR polymorphisms and schizophrenia334672798081828384858687. Of these, in only four meta-analyses MTHFR A1298C polymorphism has been investigated as a risk factor for schizophrenia72798185. Zintzaras81 pooled nine studies for A1298C meta-analysis and reported that the A1298C association with schizophrenia was only marginally significant with a fixed effect (FE) OR of 1.16 (95% CI=1.09-1.331). Gilbody et al72 performed a meta-analysis of only two studies examining the association between polymorphisms A1298C and schizophrenia and reported an association with fixed-effects [OR (CC vs. AA)=1.64 95% CI=1.04-2.54], with low heterogeneity (I²=0%). Peerbooms et al85 conducted a meta-analysis of 10 published case-control studies investigating associations between MTHFR A1298C and schizophrenia and reported that schizophrenia was not significantly associated with A1298C polymorphism. Hu et al86 pooled 12 studies and reported a significant association between A1298C polymorphism and schizophrenia (OR=1.13, 95% CI 1.03-1.24).

The present meta-analysis had several strengths along with limitations. The main strength of the present meta-analysis was that the publication bias was not detected and pooled number of cases and controls from different studies significantly increased the statistical power of the analysis. The limitations of the present meta-analysis were (i) use of crude OR, (ii) small sample size (9537 samples), (iii) inclusion of studies with small sample size, (iv) only one gene MTHFR and one polymorphism (A1298C) of folate pathway considered, and (v) gene-gene interactions not considered.

In conclusion, the present meta-analysis indicated that A1298C polymorphism was associated with schizophrenia risk. In future, several case-control studies with larger samples size are required to investigate gene-gene and gene-environment interactions involving the MTHFR A1298C polymorphism to determine the susceptibility for schizophrenia.