Abstract
Previous studies have investigated the association between common variants in FKBP5 and MDD; however, the results remain inconsistent. In order to conduct a comprehensive meta-analysis of the association between FKBP5 variants and MDD risk, seven studies involving 26582 subjects, including 12491 cases with MDD and 14091 controls, were enrolled totally. Four common SNPs (rs1360780, rs4713916, rs3800373 and rs755658) with complete data from two or more studies were analyzed. In the total sample, there was no evidence of a significant association between MDD and any of the four SNPs using a random-effects model. However, after removing one heterogeneous German study, as indicated by sensitivity analysis, both the rs1360780 T-allele (Z = 2.95, P = 0.003, OR = 1.06, 95% CI = 1.02–1.11) and the rs3800373 C-allele (Z = 3.05, P = 0.002, OR = 1.07, 95% CI 1.02–1.12) were significantly associated with MDD in a fixed-effect model. Our study thus provides support for an association between specific FKBP5 genetic variants and MDD risk. Rs4713916 was not significantly associated with MDD; However, this analysis had limited statistical power and larger sample sizes are required to further validate this result. Future research should also investigate possible gender- and ethnicity-specific differences in the association between FKBP5 and MDD.
Similar content being viewed by others
Introduction
Major depressive disorder (MDD) is one of the most common psychiatric disorders, affecting up to 20% of the world population across their lifetime1. Some of the core symptoms of MDD include persistent depressed mood, loss of interest or pleasure, and psychomotor retardation. The World Health Organization has predicted that MDD will be the second leading cause of disability worldwide by the year 20202.
Though the pathogenesis of MDD remains largely unknown, genetic factors have been shown to play an important role in conferring vulnerability to MDD and the heritability of MDD, estimated by family, twin and adoption studies, is 35–40%3,4. Uncovering the genetic basis of susceptibility to MDD has become an important task. Though risk variants range widely in frequency and penetrance5, together common variants identified by candidate gene or genome-wide association studies (GWAS) constitute a considerable portion of the genetic architecture of MDD6,7.
As opposed to GWASs, candidate gene studies investigate MDD genetic risk factors based on hypothesis driven approaches. Candidate genes of MDD have logically focused on the stress hormone-regulating hypothalamic-pituitary-adrenal (HPA) axis8 and neurotransmitter signaling pathways9, given their involvement in the development of MDD. Up till now, several dozen MDD susceptibility genes have been identified through candidate gene studies, for example, ACSM110 and BDNF11.
FKBP5 is located on chromosome 6p21.31 (GRCh38), a region associated with psychosis12, and encodes the heat shock protein 90 co-chaperone. On one hand, the expression of FKBP5 is regulated by corticosteroids at the transcriptional level through the hormone response element (HRE). On the other hand, FKBP5 can decrease the affinity of the glucocorticoid receptor (GR) to corticosteroids by a complex interaction with the mature GR, thus resulting in impaired GR signaling13,14. Given the regulatory role of FKBP5 in HPA axis signaling, Binder et al. first investigated the association between common variants in FKBP5 and MDD risk. They found that single nucleotide polymorphisms (SNPs) were significantly associated with the recurrence of depressive episodes and response to antidepressants, although no significant association was observed between SNPs and MDD15. Following this study, several other groups have reported an association between FKBP5 variants and MDD using American, German and Polish cohorts respectively16,17,18. However, other studies involving Spanish, Swedish and Italian populations have failed to replicate these finding19,20,21. Whether common variants in FKBP5 are associated with MDD susceptibility thus remains inconclusive.
Possible explanations for these inconsistent results might be genetic heterogeneity of MDD and allelic differences between ethnicities. Furthermore, the sample sizes in previous studies were relatively small and might be underpowered to detect the likely small effect sizes. Meta-analysis allows one to combine data from different studies, thus maximizing the statistical power to detect an association between common variants and MDD risk, if indeed one exists22. In this study, we report a comprehensive meta-analysis of the association between common FKBP5 variants and MDD risk.
Results
Eligible studies
According to the literature search strategy, a total of 113 references were identified. We excluded 107 studies for reasons specified in Fig. 1, leaving 6 independent studies for the meta-analysis16,17,18,19,20,21.
Although Binder et al. genotyped 5 SNPs across the FKBP5 locus in a German cohort (294 MDD cases and 338 controls), and reported no evidence of an association with MDD, none of the exact genotype frequency data in the case and control groups was available, either from the manuscript15 or by contacting the corresponding author. Besides, the samples used in the Binder et al. study were also part of the European GWAS sample. Therefore, we excluded this sample from this meta-analysis. One European GWAS data set (9240 cases with MDD and 9519 controls), which reported the P-values, ORs and S.Es., was included in our meta-analysis23. Overall, 7 studies totaling 12491 cases with MDD and 14091 controls were combined in the present study. Careful examination of the details of each sample, ensured there was no overlap between the population included in the six association studies and the GWAS. Table 1 lists the information on the recruited studies, including sample size, ethnicity, mean age, gender ratio, diagnosis of MDD, and genotyping method.
To date, 11 SNPs in FKBP5 have been investigated in candidate gene association studies of MDD. Seven of these SNPs were reported in only one study and thus excluded from the meta-analysis (Table S1).The remaining 4 SNPs were rs1360780, rs4713916, rs3800373 and rs755658 (Table 2).
Assessment of publication bias
Stata12.0 software was used to detect potential publication bias in the meta-analysis. No evidence of publication bias was observed for any of the SNPs: rs1360780 (Begg’s P = 0.652, Egger’s P = 0.862); rs4713916 (Begg’s P = 0.117, Egger’s P = 0.645); rs3800373 (Begg’s P = 0.174, Egger’s P = 0.581) (funnel plots not shown), and rs755668 (Begg’s P = 0.117, Egger’s P = 0.669).
Power analysis
Before the pooling procedure, statistical power was assessed with the following assumptions: P = 0.05, OR = 1.20 (corresponding to a “weak to moderate” gene effect) for the four SNPs, and minor allele frequencies (MAF) were estimated from the 1000 Genomes project (www.1000genomes.org/). The present sample size indicated 100% power for rs1360780 (MAF = 0.327) and rs3800373 (MAF = 0.325), 88% power for rs755658 (MAF = 0.059), and 57.9% power for rs4713916 (MAF = 0.222, 2013 cases and 1274 controls available) to detect a significant association with MDD, if one indeed existed.
FKBP5 variants and MDD susceptibility
There was significant evidence of between-study heterogeneity for the rs1360780 T-allele (I2 = 56.7%, P = 0.031), rs4713916 A-allele (I2 = 75.5%, P = 0.017) and rs3800373 C-allele (I2 = 68.4%, P = 0.023), therefore the random-effects model was used to combine samples. Meta-analysis of all available samples showed that none of the three SNPs, rs1360780 (OR = 1.05, 95% CI 0.95–1.15, Z = 0.97, P = 0.333), rs4713916 (OR = 0.97, 95% CI 0.76–1.24, Z = 0.24, P = 0.812) and rs3800373 (OR = 1.02, 95% CI 0.89–1.17, Z = 0.28, P = 0.779), was associated with MDD (Fig. S1). Furthermore, there was no association between the rs755658 T-allele and MDD with the fixed-effect model (OR = 1.06, 95% CI 0.99–1.13, Z = 1.65, P = 0.100), since no significant heterogeneity was observed among the different studies for this SNP (I2 = 43.8%, P = 0.169) (Fig. 2).
Sensitivity analyses
Considering the significant heterogeneity observed for rs1360780 T-allele, rs4713916 A-allele and rs3800373 C-allele, we conducted sensitivity analysis in order to determine the source of heterogeneity which might influence the findings. The sensitivity analysis indicated a large difference in the heterogeneity of the population for rs1360780 (from I2 = 56.7%, P = 0.031, to I2 = 33.6%, P = 0.184), rs4713916 (from I2 = 75.5%, P = 0.017, to I2 = 0%, P = 0.687), as well as rs3800373 (from I2 = 68.4%, P = 0.023, to I2 = 0.0%, P = 0.687), after removing the Zobel et al. study17.
A careful examination of the Zobel et al. study showed that 1) both the rs1360780 T-allele and the rs4713916 A-allele from the German population had the highest MAF and protective association with MDD, whereas the other studies indicated that both the other two SNPs contributed to MDD risk (Table 2); and 2) MAF of rs1360780 (0.339) was considerably different from the HapMap project (MAF = 0.268 in European samples).
After removing the Zobel et al. study, we pooled ORs again with the fixed-effect model. Meta-analysis showed that the rs1360780 T-allele was significantly associated with MDD (Z = 2.95, P = 0.003) with OR 1.062 (95% CI 1.02–1.11). For rs3800373, the C-allele conferred significantly increased MDD risk to MDD as well (Z = 3.05, P = 0.002) with an OR 1.07 (95% CI 1.02–1.12). However, there remained no significant association between MDD and rs4713916 (Z = 1.37, P = 0.170, OR = 1.09, 95% CI 0.96–1.23) (Fig. 2).
Discussion
It has been recognized as one of the greatest challenges to decipher the genetic architecture underlying MDD. Most of the inconsistent association results of MDD might be due to limited sample sizes, likely lacking statistical power, and heterogeneous patient populations, i.e. ethnicity and subtypes of MDD. Meta-analysis is a widely accepted technique used to produce solid conclusions by combining data from independent studies together24, and more importantly, serve as a powerful method to dissect potential source of heterogeneity. In the present study, we conducted a comprehensive meta-analysis of common variants across the FKBP5 locus and MDD risk. The meta-analysis involved a total sample size of 26582, including 12491 patients and 14091 healthy controls, which was sufficiently powered to detect MDD risk associated with genetic factors that had a low to moderate effect size (OR = 1.20). Although we found no significant evidence of an association between any of the SNPs in FKBP5 and MDD susceptibility when all samples were pooled, when one heterogeneous study was removed17, both the rs1360780 T-allele (Z = 2.95, P = 0.003, OR = 1.062, 95% CI 1.02–1.11) and the rs3800373 C-allele (Z = 3.05, P = 0.002, OR = 1.07, 95% CI 1.02–1.12) were significantly associated with MDD (Fig. 2).
No prior individual study has reported a significant association of MDD with either rs1360780 or rs3800373. For rs1360780, the majority of studies found that the T-allele frequency was higher among cases than in controls, with a moderate effect (OR ranging from 1.049 to1.210)16,18,19,20,23 but failed to reach statistical significance. Two studies have reported small effect sizes in the reverse direction (OR ranging from 0.762 to 0.912)17,21. For rs3800373, Lekman et al. reported a marginal association with MDD (P = 0.052)17. As we know, the statistical power of a case-control genetic association study is determined by a number of factors including the type I error probability, OR, the MAF of the genetic variants and the sample size. Simplified, the required sample size increases exponentially as the ORs declines. To detect a SNP with an OR of 1.1 or less, tens of thousands of samples will be required (based on 10% prevalence of MDD)25. Given the observed small effect size of the rs1360780 T-allele and the rs3800373 C-allele shown in this meta-analysis, this suggests that previous studies of this variant were under-powered, thus providing a possible explanation for their null findings26.
Differences in the allelic frequencies of a given SNP across ethnic groups, can influence the findings27. Given that the studies included in the meta-analysis involve predominantly individuals of European ancestry, it’s difficult to generalize these findings to other ethnic groups. For example, SNP rs1344707 in ZNF804A is significantly associated with schizophrenia risk in European populations, however is not in a Chinese population28. As is shown in the HapMap project (http://hapmap.ncbi.nlm.nih.gov/), the T-allele frequencies of rs1360780 in Japanese and Nigerian were 0.210 and 0.449 respectively, being quite different from that in European samples (0.268). Whether the rs1360780 T-allele or the rs3800373 C-allele are associated with MDD in non-European populations (i.e., Asian and African) remains to be determined. Furthermore, a number of genetic variants confer risk to MDD in a gender-specific way. For example, a meta-analysis of BDNF Val66Met polymorphism found that this variant was not significantly associated with MDD in the overall population or among females, but was associated with MDD in males (OR = 1.27, 95% CI 1.10–1.47)11. The rs619002 and rs644926 SNPs of the EHD3 locus were exclusively associated with MDD in females but not in males29. Given that each of the studies included in the meta-analysis did not provide genotypic information separately for each gender, the potential gender-specific association between rs1360780 T-allele and MDD was not analyzed. Whether or not the other FKBP5 SNPs, especially rs4713916 and rs755658, were associated with MDD in a gender-specific way also remains unknown. In addition, although our meta-analysis have the distinct advantage of combining data from multiple studies, strengthening the overall findings, the estimated statistical power to investigate the association between rs4713916 and MDD was indeed limited (57.9%), which might help explain the null finding. It remains possible that this SNP is associated with MDD, but this can only be association may be identified in even larger sample sizes.
FKBP5 is a heat shock protein 90 co-chaperone, which is essential for maturation and activation of GR13. In the absence of corticosteroids, FKBP5 can decrease the affinity of GR to corticosteroids by a complex interaction with the mature GR, thus leading to weakened GR signaling30. There is evidence to suggest that rs1360780 TT homozygous carriers have higher FKBP5 levels, and these are twice as high as those with C+ genotypes (P = 0.024)15. These different expression patterns of FKBP5 might be due to different transcriptional activity of FKBP5, which is regulated by a conserved HRE located less than 200 bp away from rs1360780 in intron 231. This FKBP5 variant has also been associated with hippocampal volume, an endophenotype of MDD17. Moreover, rs1360780 could regulate the sensitivity of GR to its ligand15. Furthermore work should now focus on exploring the functional differences of rs1360780 alleles, which would provide further insights into the involvement of this gene in MDD.
Methods
Search strategy
Meta-analysis was performed in agreement with previously described methods24,32. To identify eligible studies for the meta-analysis, PubMed (http://www.ncbi.nlm.nih.gov), SCOPUS (http://www.scopus.com), EMBASE (http://www.elsevier.com/online-tools/embase) and ISI Web of Knowledge (http://apps.webofknowledge.com/) were retrieved with the following searching terms “(FKBP5 or FKBP51 or FKBP54) and (depression or mood disorder or affective disorder)”. Since FKBP5 is also known as FKBP51 or FKBP54, the three gene symbols were retrieved simultaneously. Studies published in English before the 1st of November 2015 were considered. The references of retrieved articles were also reviewed to identify other eligible studies that were not indexed by the above-mentioned databases.
Inclusion and exclusion criteria
Only those studies investigating at least one FKBP5 SNP were included in the current meta-analyses. Eligible studies were also required to meet the following criteria: 1) be published in a peer-reviewed journal; 2) a case-control studies; 3) provide genotype and/or allele frequencies in both case and control samples, or statistics, including odds ratio (OR) with 95% confidence interval (95% CI); 4) samples were independent of other studies (if samples from different studies overlapped, only the first published data was included in this meta-analysis); 5) MDD was assessed in accordance with DSM-III, DSM-IIIR, DSM-IV, DSM-V (Diagnostic and Statistical Manual of Mental Disorders) or ICD-10 (International Classification of Diseases) diagnostic criteria; and 6) genotype frequencies in the controls were in Hardy–Weinberg equilibrium (HWE) (P > 0.05).
Studies with any of the following criteria were excluded from this meta-analysis: 1) case-only studies, family-based designs and population studies with healthy subjects which were designed to explore the effects of FKBP5 variants on depression-related phenotypes, i.e., anxiety or personality traits; 2) studies with insufficient data to calculate an effect size, even after we contacted the corresponding authors.
Eligible studies included a European GWAS data set, which was freely accessible from Ricopili (http://www.broadinstitute.org/mpg/ricopili/)23. We extracted information on FKBP5 SNPs, including P-value, odds ratio (OR) and 95% confidence intervals (95% CI) and standard error (S.E.). This GWAS enrolled a total of 9240 cases with MDD and 9519 controls, from nine distinct groups, including GAIN (Genetic Association Information Network studies, 1696 cases and 1765 controls), GenRED (Genetics of Recurrent Early-onset Depression, 1030 cases and 1253 controls), GSK (Glaxo-Smith-Kline, 887 cases and 864 controls), MDD2000-QIMR_610 (433 cases and 751 controls), MDD2000-QIMR_317 (1017 cases and 960 controls), MPIP (376 cases and 537 controls), RADIANT + Bonn/Mannhein (935 cases and 1290 controls), RADIANT (1625 cases and 1588 controls), and STAR*D (Sequenced Treatment Alternatives to Relieve Depression, 1241 cases and 511 controls)6,33,34,35,36,37,38,39.
Data extraction
For each eligible study, two independent investigators (Rao and Yao) extracted the following data using a standardized data extraction form: 1) first author and publication year; 2) study design; 3) sample origin; 4) sample size, gender ratio and age distribution if available; 5) MDD diagnosis criteria; 6) OR and 95% CI; and 7) genotypic and allelic distribution of cases and controls. If essential data were not available directly from the manuscripts, we calculated these values using the existing data, or contacted the authors for additional data.
Statistical analysis
Chi-square (X2) goodness-of-fit test was used to calculate the Hardy Weinberg Equilibrium (HWE) of genotype frequencies in controls, when this was not reported. P < 0.05 was used as the threshold of statistical significance. We applied the Power and Sample Size Program software to perform power analysis40, and Stata12.0 statistical software package (http://www.stata.com/) to conduct publication bias analysis, meta-analysis and sensitivity analysis. Potential publication bias was checked using the Egger regression test for a funnel plot41 and the Begg–Mazumdar test, which is based on Kendall’s-τ42.
Cochran’s x2-based Q-statistic was performed to assess the heterogeneity between individual OR estimates. The extent of inconsistency across studies was quantified with the I2 metric (I2 = Q-d.f./Q) that takes values between 0 and 100%, with 0–25% representing no heterogeneity, 25–50% moderate heterogeneity, 50–75% large heterogeneity and 75–100% extreme heterogeneity43. When heterogeneity was present, a random-effects model was used to combine the odds ratio and the corresponding 95% CI; otherwise, a fixed-effect model was used. The significance of the pooled OR was determined by the Z test. Sensitivity analysis was performed to test the potential effect of individual studies on the pooled OR by sequentially removing each study and recalculating the pooled OR and 95% CI. For all analyses, P < 0.05 was considered statistically significant.
Additional Information
How to cite this article: Rao, S. et al. Common variants in FKBP5 gene and major depressive disorder (MDD) susceptibility: a comprehensive meta-analysis. Sci. Rep. 6, 32687; doi: 10.1038/srep32687 (2016).
References
Alonso, J. et al. Prevalence of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta psychiatrica Scandinavica. Supplementum. 21–27 (2004).
Murray, C. J. & Lopez, A. D. Evidence-based health policy–lessons from the Global Burden of Disease Study. Science 274, 740–743 (1996).
Sullivan, P. F., Neale, M. C. & Kendler, K. S. Genetic epidemiology of major depression: review and meta-analysis. The American journal of psychiatry 157, 1552–1562 (2000).
Shih, R. A., Belmonte, P. L. & Zandi, P. P. A review of the evidence from family, twin and adoption studies for a genetic contribution to adult psychiatric disorders. International review of psychiatry 16, 260–283 (2004).
Sullivan, P. F., Daly, M. J. & O’Donovan, M. Genetic architectures of psychiatric disorders: the emerging picture and its implications. Nature reviews. Genetics 13, 537–551 (2012).
Wray, N. R. et al. Genome-wide association study of major depressive disorder: new results, meta-analysis, and lessons learned. Molecular psychiatry 17, 36–48 (2012).
consortium, C. Sparse whole-genome sequencing identifies two loci for major depressive disorder. Nature 523, 588–591 (2015).
de Kloet, E. R., Joels, M. & Holsboer, F. Stress and the brain: from adaptation to disease. Nat Rev Neurosci 6, 463–475 (2005).
Levinson, D. F. The genetics of depression: a review. Biol Psychiatry 60, 84–92 (2006).
Li, W. et al. Genetic association of ACSM1 variation with schizophrenia and major depressive disorder in the Han Chinese population. Am J Med Genet B Neuropsychiatr Genet 168B, 144–149 (2015).
Verhagen, M. et al. Meta-analysis of the BDNF Val66Met polymorphism in major depressive disorder: effects of gender and ethnicity. Molecular psychiatry 15, 260–271 (2010).
Cheng, R. et al. Genome-wide linkage scan in a large bipolar disorder sample from the National Institute of Mental Health genetics initiative suggests putative loci for bipolar disorder, psychosis, suicide, and panic disorder. Molecular psychiatry 11, 252–260 (2006).
Grad, I. & Picard, D. The glucocorticoid responses are shaped by molecular chaperones. Molecular and cellular endocrinology 275, 2–12 (2007).
Wochnik, G. M. et al. FK506-binding proteins 51 and 52 differentially regulate dynein interaction and nuclear translocation of the glucocorticoid receptor in mammalian cells. The Journal of biological chemistry 280, 4609–4616 (2005).
Binder, E. B. et al. Polymorphisms in FKBP5 are associated with increased recurrence of depressive episodes and rapid response to antidepressant treatment. Nat Genet 36, 1319–1325 (2004).
Lekman, M. et al. The FKBP5-gene in depression and treatment response–an association study in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) Cohort. Biol Psychiatry 63, 1103–1110 (2008).
Zobel, A. et al. DNA sequence variants of the FKBP5 gene are associated with unipolar depression. The international journal of neuropsychopharmacology/official scientific journal of the Collegium Internationale Neuropsychopharmacologicum 13, 649–660 (2010).
Szczepankiewicz, A. et al. FKBP5 polymorphism is associated with major depression but not with bipolar disorder. J Affect Disord 164, 33–37 (2014).
Papiol, S. et al. Genetic variability at HPA axis in major depression and clinical response to antidepressant treatment. J Affect Disord 104, 83–90 (2007).
Lavebratt, C., Aberg, E., Sjoholm, L. K. & Forsell, Y. Variations in FKBP5 and BDNF genes are suggestively associated with depression in a Swedish population-based cohort. J Affect Disord 125, 249–255 (2010).
Minelli, A. et al. Role of allelic variants of FK506-binding protein 51 (FKBP5) gene in the development of anxiety disorders. Depress Anxiety 30, 1170–1176 (2013).
Chapman, K. et al. A meta-analysis of European and Asian cohorts reveals a global role of a functional SNP in the 5′ UTR of GDF5 with osteoarthritis susceptibility. Human molecular genetics 17, 1497–1504 (2008).
Major Depressive Disorder Working Group of the Psychiatric, G. C. et al. A mega-analysis of genome-wide association studies for major depressive disorder. Mol Psychiatry 18, 497–511 (2013).
Rao, S. et al. Common variants in CACNA1C and MDD susceptibility: A comprehensive meta-analysis. Am J Med Genet B Neuropsychiatr Genet (2016).
Flint, J. & Kendler, K. S. The genetics of major depression. Neuron 81, 484–503 (2014).
Spencer, C. C., Su, Z., Donnelly, P. & Marchini, J. Designing genome-wide association studies: sample size, power, imputation, and the choice of genotyping chip. PLoS genetics 5, e1000477 (2009).
Li, M. et al. Allelic differences between Europeans and Chinese for CREB1 SNPs and their implications in gene expression regulation, hippocampal structure and function, and bipolar disorder susceptibility. Molecular psychiatry 19, 452–461 (2014).
Li, M. et al. Allelic differences between Han Chinese and Europeans for functional variants in ZNF804A and their association with schizophrenia. The American journal of psychiatry 168, 1318–1325 (2011).
Wang, L., Shi, C., Zhang, K. & Xu, Q. The gender-specific association of EHD3 polymorphisms with major depressive disorder. Neuroscience letters 567, 11–14 (2014).
Zhang, X., Clark, A. F. & Yorio, T. FK506-binding protein 51 regulates nuclear transport of the glucocorticoid receptor beta and glucocorticoid responsiveness. Investigative ophthalmology & visual science 49, 1037–1047 (2008).
Kawamura, Y. et al. DNA polymorphism in the FKBP5 gene affects impulsivity in intertemporal choice. Asia Pac Psychiatry 5, 31–38 (2013).
Yao, Y. et al. Meta-analysis indicates that SNP rs9939609 within FTO is not associated with major depressive disorder (MDD) in Asian population. J Affect Disord 193, 27–30 (2016).
Rietschel, M. et al. Genome-wide association-, replication-, and neuroimaging study implicates HOMER1 in the etiology of major depression. Biol Psychiatry 68, 578–585 (2010).
Sullivan, P. F. et al. Genome-wide association for major depressive disorder: a possible role for the presynaptic protein piccolo. Molecular psychiatry 14, 359–375 (2009).
Shi, J. et al. Genome-wide association study of recurrent early-onset major depressive disorder. Molecular psychiatry 16, 193–201 (2011).
Muglia, P. et al. Genome-wide association study of recurrent major depressive disorder in two European case-control cohorts. Mol Psychiatry 15, 589–601 (2010).
Lewis, C. M. et al. Genome-wide association study of major recurrent depression in the U.K. population. Am J Psychiatry 167, 949–957 (2010).
Shyn, S. I. et al. Novel loci for major depression identified by genome-wide association study of Sequenced Treatment Alternatives to Relieve Depression and meta-analysis of three studies. Molecular psychiatry 16, 202–215 (2011).
Ising, M. et al. A genomewide association study points to multiple loci that predict antidepressant drug treatment outcome in depression. Archives of general psychiatry 66, 966–975 (2009).
Dupont, W. D. & Plummer, W. D. Jr. Power and sample size calculations. A review and computer program. Control Clin Trials 11, 116–128 (1990).
Egger, M., Davey Smith, G., Schneider, M. & Minder, C. Bias in meta-analysis detected by a simple, graphical test. BMJ 315, 629–634 (1997).
Begg, C. B. & Mazumdar, M. Operating characteristics of a rank correlation test for publication bias. Biometrics 50, 1088–1101 (1994).
Lau, J., Ioannidis, J. P. & Schmid, C. H. Quantitative synthesis in systematic reviews. Ann Intern Med 127, 820–826 (1997).
Acknowledgements
We thank all the subjects who provided their DNA for this research. This work was supported by the Fundamental Research Funds for the Central Universities (To S.R.) and the National Natural Science Foundation of China (No. 81571319, to Y. X.).
Author information
Authors and Affiliations
Contributions
S.R., Q.X. and Y.X. designed the study and wrote the protocol. Y.Y. and S.R. managed systematic literature research and extracted data from eligible studies. S.R., J.R., T.L., D.W. and C.Z. undertook the statistical analysis. S.R. wrote the article and J.R. revised it.
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Electronic supplementary material
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Rao, S., Yao, Y., Ryan, J. et al. Common variants in FKBP5 gene and major depressive disorder (MDD) susceptibility: a comprehensive meta-analysis. Sci Rep 6, 32687 (2016). https://doi.org/10.1038/srep32687
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep32687
- Springer Nature Limited
This article is cited by
-
A large-scale genome-wide gene expression analysis in peripheral blood identifies very few differentially expressed genes related to antidepressant treatment and response in patients with major depressive disorder
Neuropsychopharmacology (2021)
-
No association between FKBP5 gene methylation and acute and long-term cortisol output
Translational Psychiatry (2020)
-
Association of FKBP5 genotype with depressive symptoms in patients with coronary heart disease: a prospective study
Journal of Neural Transmission (2020)
-
FKBP5 methylation predicts functional network architecture of the rostral anterior cingulate cortex
Brain Structure and Function (2020)
-
The effects of childhood maltreatment on epigenetic regulation of stress-response associated genes: an intergenerational approach
Scientific Reports (2019)