Psychiatry Research 245 (2016) 17

Contents lists available at ScienceDirect

Psychiatry Research
journal homepage: www.elsevier.com/locate/psychres

Review article

Lower folate levels in schizophrenia: A meta-analysis

Bing Cao a,1, Dong-Fang Wang a,1, Mei-Yan Xu b, Ya-Qiong Liu a, Lai-Lai Yan a,
Jing-Yu Wang a,n, Qing-Bin Lu a,n
a
b

School of Public Health, Peking University, Beijing 100191, PR China

Department of Nutrition, Aerospace Center Hospital, Beijing 100049, PR China

art ic l e i nf o

a b s t r a c t

Article history:
Received 20 August 2015
Received in revised form
1 March 2016
Accepted 2 March 2016
Available online 22 July 2016

This meta-analysis aimed to estimate the association between folate level and schizophrenia in order to provide the evidence for the treatment of schizophrenia. Data were extracted from all the studies meeting our
inclusion and exclusion criteria. The association between the folate level and schizophrenia was evaluated by
the standardized mean difference (SMD) and 95% condence interval (CI). The 20 published articles of our
meta-analysis included 1463 (53.4%) cases and 1276 (46.6%) controls. The folate level was signicantly lower in
schizophrenia cases than in healthy controls. Subgroup analysis showed the folate level was lower in cases from
Asia subgroup than in healthy controls. Sensitivity analysis showed that the current results were credible and
reliable and the funnel plots indicated no publication bias in our meta-analysis. Our study indicates that
schizophrenia patients may have lower folate levels. More epidemiological and laboratory studies are still
needed to conrm whether it is necessary to supplement folate in schizophrenia patients.
& 2016 Elsevier Ireland Ltd. All rights reserved.

Keywords:
Folate
Case-control study
Subgroup analysis
Psychotic disorders

Contents
1.
2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Literature searches. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Inclusion and exclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.
Data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.
Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Basic statistics of the studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
Homogeneity analysis and effect estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.
Subgroup analysis and meta-regression analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.
Sensitivity analysis and publication bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conict of interest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix A.
Supporting information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1
2
2
2
2
3
3
3
3
3
5
5
6
6
6
6
6
6

1. Introduction
n

Corresponding authors.
E-mail addresses: wjy@bjmu.edu.cn ( J.-Y. Wang),
qingbinlu@bjmu.edu.cn (Q.-B. Lu).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.psychres.2016.03.003
0165-1781/& 2016 Elsevier Ireland Ltd. All rights reserved.

The psychopathological manifestations of schizophrenia consist of

separation from reality with delusion formation, disorganized behavior, hallucinations, and emotional dysregulation (Millier et al., 2014).

B. Cao et al. / Psychiatry Research 245 (2016) 17

Fig. 1. Flow chart depicting exclusion/inclusion of individual studies for meta-analysis.

It is a debilitating disorder that affects 1% of the population worldwide (Arroll et al., 2014). Schizophrenia evolves in cycles of remissions and relapses (Millier et al., 2014) with nancial burden and
humanistic burden on society (Abouzaid et al., 2010).
Folate is one of the B vitamins which is required for cell division and cell maintenance and folate is the re-methylation of
plasma homocysteine to methionine (Mitchell et al., 2014). Studies
have been reported that folate is essential for neuronal function
(Czeizel et al., 2013). Severe folate deciencies may be linked to
the increased risk of neurodevelopmental disorders, psychiatric
diseases and dementia (Mitchell et al., 2014). The association between folate level and schizophrenia is still controversial (AyesaArriola et al., 2012; Bouaziz et al., 2010; Chen, 2014; Eren et al.,
2010; Feng et al., 2009; Garcia-Miss Mdel et al., 2010; Haidemenos
et al., 2007; Hei et al., 2014; Kim and Moon, 2011; Mabrouk et al.,
2011; Misiak et al., 2014; Muntjewerff et al., 2003; Ozcan et al.,
2008; Petronijevic et al., 2008; Qian et al., 2009; Saedisomeolia
et al., 2011; Song et al., 2014; Zhang et al., 2010; Zhang et al., 2007;
Zhu et al., 2010). Several studies revealed a positive relationship
between folate deciency and schizophrenia (Chen, 2014; Kim and
Moon, 2011; Mabrouk et al., 2011). While other studies did not
nd that folate deciency increased the risk of schizophrenia
(Ayesa-Arriola et al., 2012; Bouaziz et al., 2010; Garcia-Miss Mdel
et al., 2010; Ozcan et al., 2008). Therefore, the relationship needs
to be further evaluated.
Taking all ndings into consideration, we performed this metaanalysis to evaluate the association of them and then to provide
the evidence for the treatment of schizophrenia.

2. Methods
2.1. Literature searches
We performed a systematic literature search to identify eligible
relevant studies that reported the folate status of schizophrenia

and were published in English or Chinese in an electronic database. Search terms included schizophrenia OR psychotic disorders
OR psychosis, folate OR folic acid OR vitamin 9. PubMed, EMBASE literature database were used to search the studies published
in English. And China Biology Medical (CBM), China National
Knowledge Infrastructure (CNKI), VIP information (VIP) were used
to search the studies in Chinese. Searches were limited to articles
about human studies before the end of April 2015.
2.2. Inclusion and exclusion criteria
Inclusion criteria of the studies were met if they (1) used a
cohort or casecontrol design; (2) assessed a group of unaffected
controls (control subjects were selected from healthy volunteers
who were recruited from students, company employees and so on
documented to be free from psychiatric problems and histories of
mental illness); (3) diagnosed schizophrenia according to Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV criteria
by at least 2 psychiatrists on the basis of extensive clinical interviews and a review of medical records; (4) had available folate
levels of participants .
Exclusion criteria of the studies were met if they (1) did not
focus on evaluating the folate levels on schizophrenia; (2) did not
report the exact values of folate level; (3) included unhealthy
control groups; (4) were repetitive publications from the same
datasets by the same or different authors.
2.3. Data extraction
Data were extracted from all the studies that met our inclusion
and exclusion criteria. We designed a data form in Excel to extract
the relevant data for our review. Two investigators abstracted information independently for each eligible article, and reached a
consensus on all the items through discussion of disagreements.
Detailed information about the index test was extracted from the
studies which included rst author, publication year, country,

RIDA
ECLIA
ECLIA
ECLIA
MPEIA
ECLIA
ECLIA
MPEIA
ECLIA
ECLIA
ECLIA
ECLIA
RIDA
RIDA
FPIA
ECLIA
ECLIA
ECLIA
ECLIA
ECLIA

Detection method

B. Cao et al. / Psychiatry Research 245 (2016) 17

27/8
67/30
37/20
44/0
10/8
72/50
28/14
33/0
74/31
24/18
40/26
66/0
99/135
39/21
54/7
75/64
30/26
26/19
71/59
28/18
39.4 79.0
21.076.0
27.0 7 4.8
34.7 78.7
31.0 7 7.0
31.0 7 10.0
29.0 77.7
36.6 79.0
38.0 7 9.0
22.7 73.9
32.4 711.7
33.6 7 11
40.4 710.5
15.055.0
33.3 7 9.2
32.17 10.8
27.2 7 7.3
22.17 5.0
27.0 7 12.0
22.5 7 5.8

No
No
Yes
No
No
No
Yes
No
No
Yes
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes

First episode
M/F
Age (range or mean 7 SD)

The association between the folate level and schizophrenia was

evaluated by the standardized mean difference (SMD) and 95%
condence interval (CI). Heterogeneity of effect estimates within
each group of studies was assessed by chi-square statistics and I2
statistics. P o0.10 or I2 450% was considered that the heterogeneity had statistically signicant differences. If the data were
homogeneous, the xed effect model would be applied to evaluate
the folate status of schizophrenia. Otherwise, the random effect
model would be adopted. Sensitivity analysis was performed to
strengthen the results of the meta-analysis. Publishing bias was
tested using the funnel plot with Begg's test and Egger's test. The
meta-analysis data were analyzed by Stata 12.0 (Stata Corp LP,
College Station, TX, USA).

6.2
3.4
10.6
8.2
8.8
5.8
5.6
4.8
16.9
9.4
7.7
8.8
5.7
6.7
8.2
7.11
11.1
7.5
9
7.7
2.8
1.4
4.6
1.8
3.7
1.86
1.7
2.8
10.7
2.9
2.2
2.9
2.3
3.7
2.2
3.9
2.9
1.9
4
2.6
2003
2007
2007
2008
2008
2009
2009
2010
2010
2010
2010
2010
2011
2011
2011
2012
2014
2014
2014
2014

35
97
57
44
18
122
42
33
70
42
66
66
234
60
61
139
56
45
130
46

4.6
2.8
8.8
6.9
9.4
5.1
5.2
4.2
15.5
8.5
4.6
7.4
3.8
5.5
4.2
7.8
6.3
4.1
6
5.6

104
103
30
20
20
122
30
35
70
28
50
34
234
60
46
99
53
28
80
30

SD
Mean
SD
n

Mean

3.8
1.7
5
1.7
1.9
1.41
1.2
3.8
6.5
2.7
2.2
3.1
3.5
2.1
3.2
3.9
15.2
1.9
3
1.7

Netherlands
Greece
China
Serbia
Turkey
China
China
Tunisia
Mexico
China
China
Turkey
Korea
Iran
Tunisia
Spain
Poland
China
China
China

Europe
Europe
Asia
Asia
Europe
Asia
Asia
Africa
America
Asia
Asia
Europe
Asia
Asia
Africa
Europe
Europe
Asia
Asia
Asia

Geographic area

Plasma
Plasma
Serum
Serum
Serum
Serum
Serum
Plasma
Serum
Serum
Plasma
Serum
Plasma
Serum
Plasma
Serum
Plasma
Serum
Serum
Serum

Sample type

3. Results

Country
Healthy controls
Schizophrenic patients
Year

geographic area, gender ratio and age of patients, specimen needed for test, sample detection method, whether or not rst-episode of patients, sample size and mean folate levels with standard
deviation (SD) of participants. The Newcastle-Ottawa Scale (NOS)
criteria were used to assess the quality of each study.
2.4. Statistical analysis

3.1. Basic statistics of the studies

Our initial search identied 326 potentially relevant studies.
After screening of titles and abstracts, and removing the duplicates
and other unrelated references, 89 articles were remained for
further evaluation. Following detailed assessment, 20 references
met our inclusion criteria and were recruited into the meta-analysis (Ayesa-Arriola et al., 2012; Bouaziz et al., 2010; Chen, 2014;
Eren et al., 2010; Feng et al., 2009; Garcia-Miss Mdel et al., 2010;
Haidemenos et al., 2007; Hei et al., 2014; Kim and Moon, 2011;
Mabrouk et al., 2011; Misiak et al., 2014; Muntjewerff et al., 2003;
Ozcan et al., 2008; Petronijevic et al., 2008; Qian et al., 2009;
Saedisomeolia et al., 2011; Song et al., 2014; Zhang et al., 2010;
Zhang et al., 2007; Zhu et al., 2010). Fig. 1 provided the literature
review process.
The 20 studies included 1463 (53.4%) cases and 1276 (46.6%)
controls. They were all case-control studies. Most studies were
from Asia (11/20, 55.0%). There were 13 studies on serum and
seven studies on plasma. Detection methods used in the studies
were electro-chemo-luminescence immunoassay (ECLIA), micro
particle enzyme immunoassay (MPEIA), radio isotope dilution
assay (RIDA), and uorescence-polarization immunoassay (FPIA).
We evaluated all articles by NOS. All studies indicated good qualities with a score of more than or equal to 6. Table 1 presented the
characteristics of 20 included studies.
3.2. Homogeneity analysis and effect estimation
The random-effects model was used due to the high heterogeneity (I2 82.8%, P o0.001) (Fig. 2). Our meta-analysis identied
a signicant association between folate level and schizophrenia
(SMD  0.57; 95% CI:  0.76,  0.37; Po 0.001) (Fig. 3).

Author

Muntjewerff
Haidemenos
Zhang
Petronijevic
Ozcan
Feng
Qian
Bouaziz
Garca-Miss
Zhu
Zhang
Eren
Kim
Saedisomeolia
Mabrouk
Ayesa-Arriola
Misiak
Chen
Hei
Song
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

3.3. Subgroup analysis and meta-regression analysis

No.

Table1
Characteristics of the included studies in the meta-analysis.

Subgroup analysis was conducted by publish year (before/after

2010), age of patients ( o30/ Z30 years), gender ratio (F/M0;
0 oF/M r1; F/M 41), geographic area (Asia/Europe/Africa/America), sample type (Serum/Plasma), detection method (ECLIA/
MPEIA/RIDA/FPIA) and whether rst-episode of patients (Yes/No).
All the results of subgroup analysis showed the folate level of each

B. Cao et al. / Psychiatry Research 245 (2016) 17

Fig. 2. Meta-analysis for the difference of the folate levels between schizophrenia cases and healthy controls by random effect analysis. The forest plot illustrates the SMD
and its 95% CI for each study, and the results of the homogeneity test. SMD, standardized mean difference. The graph was produced using Stata 12.0.

Fig. 3. Sensitivity analysis through deletion of one study at a time to reect the inuence of the individual dataset to the pooled SMDs of the folate levels in schizophrenia
and healthy controls using Stata 12.0.

B. Cao et al. / Psychiatry Research 245 (2016) 17

Table 2
Meta-regression analysis on the heterogeneity for the association between folate
level and schizophrenia.
Factor

Coefent

95% CI

Age
First-episode
Sample type
Gender ratio
Geographic area
Publish year
Detection method
Constant

0.770
 0.604
 0.307
0.199
0.190
 0.385
 0.101
 0.358

0.538
0.563
0.302
0.340
0.170
0.255
0.169
0.908

1.43
 1.07
 1.02
0.59
1.12
 1.51
 0.59
 0.39

0.178
0.305
0.330
0.569
0.286
0.157
0.564
0.700

 0.40, 1.94
 1.83, 0.62
 0.96, 0.35
 0.54, 0.94
 0.18, 0.56
 0.94, 0.17
 0.47, 0.27
 2.34, 1.62

SE, standard error.

subgroup was lower in patients than in healthy controls (all

P o0.05) except for geographic area and detection method. The
subgroup analysis of geographic area showed the folate level was
lower in patients from Asia than in healthy controls (SMD 0.72;
95% CI:  0.94,  0.49; P o0.001), while there were no signicant
differences in Europe (SMD 0.27; 95% CI: 0.56, 0.02;
P 0.078), Africa (SMD  0.84; 95% CI:  2.13,  0.45; P 0.201),
and America (SMD  0.16; 95% CI:  0.49, 0.17; P 0.350) subgroups. No factors explain the existed heterogeneity in our metaanalysis. Meta-regression analysis was also used to explore the
sources of the heterogeneity. We performed a meta-regression
with the seven factors mentioned above, but still failed to nd the
sources of heterogeneity (all P 40.05) (Table 2).
3.4. Sensitivity analysis and publication bias
Each study included in this meta-analysis was omitted one by
one to determine the effect of an individual dataset to the pooled
SMDs. The results showed that the overall statistical signicance
did not change when each single study was omitted (Fig. 4).
Therefore, the current meta-analysis results were robustly credible
and reliable. The possibility of publication bias in the study was
evaluated with the Egger's test and Begg's test. The graphical
funnel plots of those 20 studies indicated no publication bias in
our meta-analysis (Fig. 5), and the same results were found in
Egger's test (Z  0.80, P 0.435) and Begg's test (Z  0.65,
P 0.516).

4. Discussion
The ndings in our meta-analysis have indicated that folate
level was signicantly lower in schizophrenia cases than in healthy
controls. Nevertheless, the exact mechanism that how folate level
is related to the development and progression of schizophrenia
has been not fully understood at present.
There are numerous evidence supporting our results. Mitchell
et al. have shown that people who had a low folate intake and genetic homocysteine metabolism disorder affected the occurrence of
schizophrenia (Mitchell et al., 2014). A review of Moustafa et al. indicated that low folate levels and high homocysteine levels might
inuence cognition, since homocysteine plays a role in working
memory processes, supplementation of folate has been proved to be
effective in lowering homocysteine levels (Moustafa et al., 2014). A
Korean study reported a high folate level reduced the possibility of
schizophrenia risk (Kim and Moon, 2011). Mudd et al. also asserted
the folate helped to prevent the incidence of schizophrenia (Mudd
and Freeman, 1974). Data from various original articles and metaanalyses suggested that high homocysteine level was associated with
schizophrenia (Haidemenos et al., 2007; Levine et al., 2005; Nishi
et al., 2014). The major cause of schizophrenia was a defect in the 5,
10-methyleneterahydrofolate reductase (MTHFR), which re-methylate homocysteine into methionine. The administration of folate
could treat the genetic defect of MTHFR (Levine et al., 2006; Muntjewerff et al., 2011). One possible pathway was that high homocysteine concentration affected schizophrenia by virtue of a neurotoxic mechanism, and schizophrenia patients with high homocysteine levels beneted from high folate ingestion, because the folate serves as a cofactor for methionine synthesis by transferring
methyl group to homocysteine (Akanji et al., 2007). The increase of
plasma homocysteine in the general population was caused by low
levels of serum folate, vitamin B12 and betaine (Brown and Roffman,
2014; Misiak et al., 2014; Roffman et al., 2013) which shared functions in one-carbon metabolism (Geller et al., 2013). Vitamin B12 and
folate deciency are contributors to this pathway (Fenech, 2012). A
16 weeks' parallel group trial of 140 schizophrenia patients indicated
folate plus vitamin B12 supplementation improved negative symptoms of schizophrenia, but treatment response was inuenced by
genetic variation in folate absorption (Roffman et al., 2013). Whether
it is necessary to supplement folate in schizophrenia patients still
needs more exploration.

Fig. 4. Begg's Funnel plot of the estimation of publication bias. Each point represents a separate study for the indicated association. The graph was produced using Stata 12.0.

B. Cao et al. / Psychiatry Research 245 (2016) 17

Fig. 5. Egger's Funnel plot of the estimation of publication bias. Each point represents a separate study for the indicated association. The graph was produced using Stata 12.0.

To investigate the exact relationship between folate level and

schizophrenia, we also carefully carried out subgroup analysis on
the basis of publish year, age of patients, gender ratio, geographic
area, sample type, detection method and whether or not rstepisode of patients. Our ndings of geographic area have revealed
that the folate level signicantly decreased in schizophrenia patients from Asian populations, but no signicantly statistical differences in European, African and American populations, which
may be due to differences of the racial, alimentary, treatment,
living standard, lifestyle or economic level. Owing to limited
number of studies, our results concerning subgroup analysis
should be interpreted with caution. More cohort or experimental
studies are needed to conrm the causality (Mitchell et al., 2014).
There are several limitations existing in our meta-analysis. Firstly,
the sample size was relatively small for some stratied analyses. The
literatures included in our study were not comprehensive, as we failed
to obtain the literature except for English and Chinese. Secondly, we
did not nd any cohort study to conrm the causality between schizophrenia and low folate level. Thirdly, we failed to nd the factors to
explain the existed heterogeneity in our meta-analysis.
In conclusion, our study indicates that schizophrenia patients
have lower folate levels. More epidemiological and laboratory
studies are still needed to conrm whether it is necessary to
supplement folate in schizophrenia patients.

Conict of interest
Regarding this report, the authors do not have any commercial
or other association that would be considered a conict of interest.

Contributors
Conceived and designed the experiments: Qing-Bin Lu and
Jing-Yu Wang. Searched the references and collected data: Bing
Cao, Dong-Fang Wang and Mei-Yan Xu. Performed the statistical
analysis: Bing Cao and Dong-Fang Wang. Drafted the manuscript:
Dong-Fang Wang, Bing Cao and Qing-Bin Lu. Contributed to the
discussion: Jing-Yu Wang, Rong Liu, Lai-Lai Yan and Ya-Qiong Liu.
All authors have read and approved the nal version of this article.

Funding
This work was supported by the Youth Talent Support Program by
School of Public Health, Peking University and The Medicine Interdisciplinary Seed Fund (BMU20140435) by Health Science Center,
Peking University. The funding agents had no role in the design and
conduct of the study; collection, management, analysis, interpretation
of the data; preparation, review, or approval of the manuscript.

Acknowledgments
Thanks to the researchers of the original studies included in our
meta-analysis. The authors alone are responsible for the content
and writing of the paper. We thank team members for their support and contributions to this study.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the
online version at http://dx.doi.org/10.1016/j.psychres.2016.03.003.

References
Abouzaid, S., Jutkowitz, E., Foley, K.A., Pizzi, L.T., Kim, E., Bates, J., 2010. Economic
impact of prior authorization policies for atypical antipsychotics in the treatment of schizophrenia. Popul. Health Manag. 13, 247254.
Akanji, A.O., Ohaeri, J.U., Al-Shammri, S.A., Fatania, H.R., 2007. Associations of blood
homocysteine concentrations in Arab schizophrenic patients. Clin. Biochem. 40,
10261031.
Arroll, M.A., Wilder, L., Neil, J., 2014. Nutritional interventions for the adjunctive
treatment of schizophrenia: a brief review. Nutr. J. 13, 91.
Ayesa-Arriola, R., Perez-Iglesias, R., Rodriguez-Sanchez, J.M., Mata, I., Gomez-Ruiz,
E., Garcia-Unzueta, M., Martinez-Garcia, O., Tabares-Seisdedos, R., VazquezBarquero, J.L., Crespo-Facorro, B., 2012. Homocysteine and cognition in rstepisode psychosis patients. Eur. Arch. Psychiatry Clin. Neurosci. 262, 557564.
Bouaziz, N., Ayedi, I., Sidhom, O., Kallel, A., Rafra, R., Jomaa, R., Melki, W., Feki, M.,
Kaabechi, N., El Hechmi, Z., 2010. Plasma homocysteine in schizophrenia: determinants and clinical correlations in Tunisian patients free from antipsychotics. Psychiatr. Res 179, 2429.
Brown, H.E., Roffman, J.L., 2014. Vitamin supplementation in the treatment of
schizophrenia. CNS Drugs 28, 611622.
Chen, X.-M., 2014. Changes of cortical thickness in drug naive rst episode schizophrenia and their correlations with serum folate, homocysteine, BDNF levels.
Chin. Med. J. 94, 990993.

B. Cao et al. / Psychiatry Research 245 (2016) 17

Czeizel, A.E., Dudas, I., Vereczkey, A., Banhidy, F., 2013. Folate deciency and folic
acid supplementation: the prevention of neural-tube defects and congenital
heart defects. Nutrients 5, 47604775.
Eren, E., Yegin, A., Yilmaz, N., Herken, H., 2010. Serum total homocystein, folate and
vitamin B12 levels and their correlation with antipsychotic drug doses in adult
male patients with chronic schizophrenia. Clin. Lab. 56, 513518.
Fenech, M., 2012. Folate (vitamin B9) and vitamin B12 and their function in the
maintenance of nuclear and mitochondrial genome integrity. Mutat. Res 733,
2133.
Feng, L.-G., Shao, C.-Q., Qi, P.-P., Tao, Y.-H., Hu, J., 2009. Homocysteine, folate and
vitamin B12 and the association with schizophrenia. Chin. J. Nerv. Ment. Dis. 35,
4041.
Garcia-Miss Mdel, R., Perez-Mutul, J., Lopez-Canul, B., Solis-Rodriguez, F., PugaMachado, L., Oxte-Cabrera, A., Gurubel-Maldonado, J., Arankowsky-Sandoval,
G., 2010. Folate, homocysteine, interleukin-6, and tumor necrosis factor alfa
levels, but not the methylenetetrahydrofolate reductase C677T polymorphism,
are risk factors for schizophrenia. J. Psychiatr. Res 44, 441446.
Geller, V., Friger, M., Sela, B.A., Levine, J., 2013. Elevated homocysteine level in
siblings of patients with schizophrenia. Psychiatry Res 210, 769772.
Haidemenos, A., Kontis, D., Gazi, A., Kallai, E., Allin, M., Lucia, B., 2007. Plasma
homocysteine, folate and B12 in chronic schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 31, 12891296.
Hei, G.-R., Pang, L.-J., Chen, X.-M., Zhang, W., Zhu, Q.-Y., Lv, L.-X., Song, X.-Q., 2014.
Association of serum folic acid and homocysteine levels and 5,10-methylenetetrahydrofolate reductase gene polymorphism with schizophrenia. Natl. Med.
J. China 94, 28972901.
Kim, T.H., Moon, S.W., 2011. Serum homocysteine and folate levels in korean
schizophrenic patients. Psychiatry Invest. 8, 134140.
Levine, J., Sela, B.A., Osher, Y., Belmaker, R.H., 2005. High homocysteine serum levels in young male schizophrenia and bipolar patients and in an animal model.
Prog. Neuropsychopharmacol. Biol. Psychiatry 29, 11811191.
Levine, J., Stahl, Z., Sela, B.A., Ruderman, V., Shumaico, O., Babushkin, I., Osher, Y.,
Bersudsky, Y., Belmaker, R.H., 2006. Homocysteine-reducing strategies improve
symptoms in chronic schizophrenic patients with hyperhomocysteinemia. Biol.
Psychiatry 60, 265269.
Mabrouk, H., Douki, W., Mechri, A., Younes, M.K., Omezzine, A., Bouslama, A., Gaha,
L., Najjar, M.F., 2011. Hyperhomocysteinemia and schizophrenia: case control
study. Encephale 37, 308313.
Millier, A., Schmidt, U., Angermeyer, M.C., Chauhan, D., Murthy, V., Toumi, M., CadiSoussi, N., 2014. Humanistic burden in schizophrenia: a literature review. J.
Psychiatr. Res. 54, 8593.
Misiak, B., Frydecka, D., Slezak, R., Piotrowski, P., Kiejna, A., 2014. Elevated homocysteine level in rst-episode schizophrenia patientsthe relevance of family
history of schizophrenia and lifetime diagnosis of cannabis abuse. Metab. Brain
Dis. 29, 661670.
Mitchell, E.S., Conus, N., Kaput, J., 2014. B vitamin polymorphisms and behavior:
evidence of associations with neurodevelopment, depression, schizophrenia,
bipolar disorder and cognitive decline. Neurosci. Biobehav. Rev. 47, 307320.

Moustafa, A.A., Hewedi, D.H., Eissa, A.M., Frydecka, D., Misiak, B., 2014. Homocysteine levels in schizophrenia and affective disorders-focus on cognition.
Front Behav. Neurosci. 8, 343.
Mudd, S.H., Freeman, J.M., 1974. N-5,10-methylenetetrahydrofolate reductase deciency and schizophrenia: a working hypothesis. J. Psychiatr. Res 11, 259262.
Muntjewerff, J.-W., van der Put, N., Eskes, T., Ellenbroek, B., Steegers, E., Blom, H.,
Zitman, F., 2003. Homocysteine metabolism and B-vitamins in schizophrenic
patients: low plasma folate as a possible independent risk factor for schizophrenia. Psychiatry Res. 121, 19.
Muntjewerff, J.W., Ophoff, R.A., Buizer-Voskamp, J.E., Strengman, E., den Heijer, M.,
2011. Effects of season of birth and a common MTHFR gene variant on the risk
of schizophrenia. Eur. Neuropsychopharmacol. 21, 300305.
Nishi, A., Numata, S., Tajima, A., Kinoshita, M., Kikuchi, K., Shimodera, S., Tomotake,
M., Ohi, K., Hashimoto, R., Imoto, I., Takeda, M., Ohmori, T., 2014. Meta-analyses
of blood homocysteine levels for gender and genetic association studies of the
MTHFR C677T polymorphism in schizophrenia. Schizophr. Bull. 40, 11541163.
Ozcan, O., Ipcioglu, O.M., Gultepe, M., Basogglu, C., 2008. Altered red cell membrane
compositions related to functional vitamin B(12) deciency manifested by
elevated urine methylmalonic acid concentrations in patients with schizophrenia. Ann. Clin. Biochem 45, 4449.
Petronijevic, N.D., Radonjic, N.V., Ivkovic, M.D., Marinkovic, D., Piperski, V.D., Duricic, B.M., Paunovic, V.R., 2008. Plasma homocysteine levels in young male
patients in the exacerbation and remission phase of schizophrenia. Prog.
Neuropsychopharmacol. Biol. Psychiatry 32, 19211926.
Qian, Z.-K., Jiang, X.-X., Du, X.-D., 2009. Relashionship between plasma homocysteine, serum folate levels and rst-episode schizophrenia patients. HangZhou Univ. J. Med. Sci. 29, 11401142.
Roffman, J.L., Lamberti, J.S., Achtyes, E., Macklin, E.A., Galendez, G.C., Raeke, L.H.,
Silverstein, N.J., Smoller, J.W., Hill, M., Goff, D.C., 2013. Randomized multicenter
investigation of folate plus vitamin B12 supplementation in schizophrenia.
JAMA Psychiatry 70, 481489.
Saedisomeolia, A., Djalali, M., Moghadam, A.M., Ramezankhani, O., Najmi, L., 2011.
Folate and vitamin B12 status in schizophrenic patients. J. Res Med. Sci. 16
(Suppl 1), S437S441.
Song, X., Fan, X., Li, X., Kennedy, D., Pang, L., Quan, M., Chen, X., Gao, J., Zhang, W.,
Zhang, J., Lv, L., 2014. Serum levels of BDNF, folate and homocysteine: in relation to hippocampal volume and psychopathology in drug naive, rst episode
schizophrenia. Schizophr. Res 159, 5155.
Zhang, G.-F., Tang, H., Ren, C.-L., Qi, S.-G., Zhang, Y.-B., Lu, Z.-X., Lv, Z.-L., 2010.
Changes of levels of plasma hydrogensulde, homocysteine, folic acid and vitamin B6 in schizophrenia patients. J. Xinxiang Med. Coll. 27, 603605.
Zhang, W.-Y., Lin, H., Zhao, H.-Q., Huang, H., Xuan, C.-M., Ma, J.-F., 2007. Cognitive
function in the rst-episode schizophrenics and its association w ith serum
level of homocysteine. Chin. J. Nerv. Ment. Dis. 33, 652655.
Zhu, Z.-X., Xiong, P., Jia, M., Wan, J., Zeng, Y., Xu, M.-J., 2010. Serum level of
homocysteine in the rst-episode schizophrenics and its correlation w ith
P300. Chin. J. Nerv. Ment. Dis. 36, 355357.