Association of the BDNF C270T polymorphism with schizophrenia: Updated meta-analysis
Correspondence: Yuichiro Watanabe, MD, PhD, Department of Psychiatry, Niigata University Graduate School of Medical and Dental Sciences, 757 Asahimachidori-ichibancho, Niigata 951-8510, Japan. Email: firstname.lastname@example.org
Brain-derived neurotrophic factor (BDNF) has been suggested to play a role in the pathophysiology of schizophrenia. The C270T polymorphism (rs56164415) in the BDNF 5′-non-coding region has been extensively investigated for an association with schizophrenia, but with conflicting results. An updated meta-analysis was therefore performed of 13 case–control association studies (3505 patients and 3992 controls). An association was found between the T allele and schizophrenia. The association was significant in the East Asian population, but not in the Caucasian population. It is suggested that the BDNF C270T polymorphism contributes to schizophrenia susceptibility, especially in East Asian subjects.
BRAIN-DERIVED NEUROTROPHIC FACTOR (BDNF) regulates neuronal growth and synaptic plasticity. Several studies have reported altered BDNF levels in the brain and blood of schizophrenia patients. These findings suggest that BDNF plays a role in the pathophysiology of schizophrenia.
The C270T polymorphism (rs56164415) in the BDNF 5′-non-coding region has been extensively investigated for an association with schizophrenia, but with conflicting results.[3-14] In the most recent meta-analysis, there was a significant association between the T allele and schizophrenia in a fixed-effect model, but not in a random-effect model. In addition, the association in the fixed-effect model did not remain significant after omission of the study by Szekeres et al. In the SZGene meta-analysis (http://www.szgene.org/meta.asp?geneID=2), the association was not significant. Thus, the true nature of the association between C270T and schizophrenia remains to be determined. To assess whether the BDNF C270T polymorphism confers increased susceptibility to schizophrenia, an updated meta-analysis was performed.
We included 13 case–control association studies of the BDNF C270T polymorphism with schizophrenia in the meta-analysis. We asked the correspondence authors for the genotype counts of C270T if those were not available in the publications,[11, 15] or if there was a typographical error in how these were reported. We assessed the Hardy–Weinberg equilibrium (HWE) in the patients and controls of each study using Haploview v4.2 (http://www.broadinstitute.org/scientific-community/science/programs/medical-and-population-genetics/haploview/haploview). We performed fixed- and random-effect model meta-analyses using Catmap (http://cran.r-project.org/src/contrib/Archive/catmap/). Publication bias was assessed using a linear regression analysis to measure funnel plot asymmetry. P < 0.05 was considered to indicate statistical significance.
The genotype distributions did not deviate significantly from HWE in either group in all studies (data not shown), except for the control group of Sun et al. (P = 0.048). In both fixed- and random-effect models, the T allele was significantly associated with schizophrenia (Table 1), and the combined odds ratios remained significant after the sequential omission of individual studies (data not shown). In both models, the association was significant in the East Asian population, but not in the Caucasian population. Linear regression analysis showed no significant funnel plot asymmetry.
Table 1. BDNF C270T polymorphism and schizophrenia: Meta-analysis of case–control studies
|Nanko et al. (2003)||Japanese||178||339||17||332||649||15||2.17||1.07–4.40||0.028|
|Szekeres et al. (2003)||Hungarian||101||174||28||68||132||4||5.31||1.82–15.51||0.001|
|Anttila et al. (2005)||Finnish||94||171||17||98||182||14||1.29||0.62–2.70||0.495|
|Galderisi et al. (2005)||Italian||106||199||13||111||211||11||1.25||0.55–2.86||0.592|
|Szczepankiewicz et al. (2005)||Polish||397||755||39||380||725||35||1.07||0.67–1.71||0.777|
|Jönsson et al. (2006)||Swedish||187||354||20||275||521||29||1.02||0.57–1.82||0.960|
|Watanabe et al. (2006)||Japanese||349||672||26||423||827||19||1.68||0.92–3.07||0.086|
|Zhang et al. (2006)||European American||84||158||10||250||470||30||0.99||0.47–2.07||0.982|
|Li et al. (2007)||Han Chinese||130||246||14||144||279||9||1.76||0.75–4.15||0.188|
|Xu et al. (2007)||Han Chinese||275||534||16||297||574||20||0.86||0.44–1.68||0.658|
|Kawashima et al. (2009)||Japanese||1115||2159||71||1102||2138||66||1.07||0.76–1.50||0.716|
|Zai et al. (2010)||European Canadian||166||314||18||166||316||16||1.13||0.57–2.26||0.725|
|Sun et al. (2011)||Han Chinese||323||606||40||346||661||31||1.41||0.87–2.28||0.163|
|All (fixed)†|| ||3505|| || ||3992|| || ||1.25||1.06–1.48||0.007|
|All (random)†|| || || || || || || ||1.28||1.06–1.54||0.011|
|Asian (fixed)‡|| ||2370|| || ||2644|| || ||1.29||1.04–1.60||0.021|
|Asian (random)‡|| || || || || || || ||1.32||1.03–1.70||0.029|
|Caucasian (fixed)§|| ||1135|| || ||1348|| || ||1.20||0.93–1.55||0.157|
|Caucasian (random)§|| || || || || || || ||1.24||0.91–1.68||0.176|
The conflicting findings of earlier association studies between the BDNF C270T polymorphism and schizophrenia may have been a result of insufficient statistical power in individual studies. The most recent meta-analysis did not yield clear evidence for the association. The present updated meta-analysis provides the most comprehensive assessment to date of the association. In the present updated meta-analysis, the T allele was found to be associated with schizophrenia. The significance of this association was not influenced by any single study. Stratification of the studies by ethnicity indicated that the T allele was associated with schizophrenia in the East Asian population, but not in the Caucasian population. This finding may be explained by genetic heterogeneity between these two populations. Indeed, there was a difference in the mean frequency of T allele between the East Asian (0.031) and Caucasian (0.051) populations. Notably, a recent exploratory analysis reported that the T allele was associated with a higher prevalence of neuropsychiatric symptoms and specifically with the presence of hallucinations in patients with Alzheimer's disease. Taken together, the T allele may predispose to schizophrenia, as well as psychotic symptoms of psychiatric disorders. The functional implications of the BDNF C270T polymorphism are still unclear, however, and should be clarified in future studies. In conclusion, the present updated meta-analysis provides evidence for the contribution of the BDNF C270T polymorphism to schizophrenia susceptibility, especially in East Asian subjects.
We thank the authors who kindly provided the data necessary for our meta-analysis. There is no conflict of interest.