BDNF Val66Met polymorphism, life stress and depression: A meta-analysis of gene-environment interaction

Highlights

• The interaction between life stress and BDNF Val66Met polymorphism in depression was found.

• Stressful life events and childhood adversity independently interacted with BDNF Val66Met polymorphism in depression.

• The interaction between life stress and BDNF Val66Met were strongest in Caucasian and cross-sectional studies.

Abstract

Background

Depression is thought to be multifactorial in etiology, including genetic and environmental components. While a number of gene-environment interaction studies have been carried out, meta-analyses are scarce. The present meta-analysis aimed to quantify evidence on the interaction between brain-derived neurotrophic factor (BDNF) Val66Met polymorphism and stress in depression.

Methods

Included were 31 peer-reviewed with a pooled total of 21060 participants published before October 2016 and literature searches were conducted using PubMed, Wolters Kluwer, Web of Science, EBSCO, Elsevier Science Direct and Baidu Scholar databases.

Results

The results indicated that the Met allele of BDNF Val66Met polymorphism significantly moderated the relationship between stress and depression (Z=2.666, p = 0.003). The results of subgroup analysis concluded that stressful life events and childhood adversity separately interacted with the Met allele of BDNF Val66Met polymorphism in depression (Z = 2.552, p = 0.005; Z = 1.775, p = 0.03).

Limitations

The results could be affected by errors or bias in primary studies which had small sample sizes with relatively lower statistic power. We could not estimate how strong the interaction effect between gene and environment was.

Conclusions

We found evidence that supported the hypothesis that BDNF Val66Met polymorphism moderated the relationship between stress and depression, despite the fact that many included individual studies did not show this effect.

Introduction

Depression is the leading cause of disability in individuals between 15 and 44 years of age and affects millions of individuals worldwide (Murray and Lopez, 1996; Reddy, 2010; Shenal et al., 2003). The morbidity is estimated to be approximately 10–15%. The World Health Organization predicted depression to be the leading cause of the total global burden of disease by the year 2030, which will account for 13% of the total burden and will replace cardiovascular disorders (Levav and Rutz, 2002). The etiology of depression is known to be complex (Sullivan et al., 2000). Genetic factors and nongenetic factors alone exert a relatively small contribution to depression (Ebmeier et al., 2006), the genetic contribution to depression is estimated to be approximately 35% (Otte et al., 2016), meaning that multiple partially overlapping sets of susceptibility genes interact with the environment, predisposing individuals to depression (Lohoff, 2010).

Gene-environment interactions refer to one or more genetic variant and one or more environmental factor contributing to the occurrence of a condition in the same individual, with genetic factors influencing the sensitivity to exposure of environmental factors (Uher, 2014). Environmental factors such as stressful life events and childhood adversity have been suggested to have an important impact on mental health outcomes. Stressful life events are a kind of stress closing to the onset or relapse of the illness which may have a negative effect on individuals (Ezquiaga et al., 1987). There is strong evidence indicating that stressful life events are connected with depression and stressful life events cluster before depressive episodes or aggravation of symptoms (Caspi et al., 2003). Childhood adversity is described as stressful life experiences occurring early in life. Previous studies suggested that childhood adversity was linked to psychiatric illnesses (Kessler et al., 2010, Morgan and Fisher, 2007, Scott et al., 2012), emotional problems (Agnafors et al., 2013) and physical disorders (Hosang et al., 2013, Scott et al., 2011). It is thus very important to explore the specific mechanisms that underlie stressful life events and childhood adversity/depression associations.

So far, gene-environment interaction studies in the context of depression almost solely focused on candidate genes or candidate polymorphisms (Uher, 2014). According to the neurotrophic model for stress-related mood disorders (Duman and Monteggia, 2006), stress decreases Brain-derived neurotrophic factor (BDNF) activity, which results in reduced function in limbic brain regions involved in emotion processing and cognition. BDNF is widely expressed in the central nervous system, including the hippocampus, amygdala, hypothalamus, neocortex and cerebellum, all of which are important regions in the regulation of mood. BDNF plays an important role in structural brain abnormalities observed in individuals with depression, such as reduced hippocampal volume or cognitive deficits (Calabrese et al., 2009; Martinowich et al., 2007; Nestler et al., 2002; Oral et al., 2012). There is some debate about the association between BDNF and neurogenesis (Hajek et al., 2012, Jessen et al., 2009, Terracciano et al., 2010). Given the evidence presented generally supports this association. If a functional single nucleotide polymorphism (rs6265) within the BDNF gene promoter with a substitution from valine to methionine at codon 66 in the BDNF gene (Val66Met) occurred, it could influence the activity of the BDNF protein (Baj et al., 2013; Rybakowski, 2008). It is hypothesized that life stress could cause a reduction in BDNF levels in the brain (especially in regions connected with emotion), which may affect mood and cause depression (Elzinga et al., 2011, Stein et al., 2008).

The first Gene × environment (G×E) study focused on the role of a common functional 43-bp insertion/ deletion polymorphism (5-HTTLPR) in the promoter region of the serotonin transporter gene (SLC6A4 or SERT) and found that this polymorphism involved either short (s) or long (l) alleles. The low expressing s allele predisposed to increased depressive symptoms, suicide behaviors and rates of diagnosed depression after exposure to stressful events and childhood maltreatment (Caspi et al., 2003). Based on this, a large body of studies has identified that the interaction of 5-HTTLPR variations with life stress can predict depression, but the results were inconsistent (Eley et al., 2004, Surtees et al., 2006) and three meta-analyses (Karg et al., 2011, Munafo et al., 2009, Risch et al., 2009) have recently been conducted. The interaction between BDNF Val66Met polymorphism and stress in depression has also been widely researched, however, it remains unclear if BDNF belongs to depression susceptibility genes Because of the conflicting conclusions in this studies. A possible explanation for inconsistencies is that there are methodological differences among studies, including differences in study population, study design and depression diagnostic criteria, all of which may influence the comparability of studies. Another possible speculation concerns small sample sizes in many studies that could decrease statistical power to claim statistical significance of low-risk susceptibility genes (Lohmueller et al., 2003). The aforementioned problems, especially the latter problem, can partly be solved by conducting meta-analyses (Levinson, 2005). A preliminary meta-analysis was published in 2014, which concluded that there was a statistical significance in the overall analysis. In the stratified analysis, however, the evidence was not supportive for interactions of BDNF with childhood adversity and there was no statistically significant difference in the stratified analysis of ancestry (Hosang et al., 2014). Since then, nine additional studies, one positive G×E effect in depression (Harkness et al., 2015)and eight negative G×E effect in depression (Cruz-Fuentes et al., 2014; Gutiérrez et al., 2015; Kudinova et al., 2015; Rawson et al., 2015; Rimay et al., 2015a; Shimasaki et al., 2014; Webb et al., 2016; Zhang et al., 2016a) replication, have been published and a reevaluation of this G × E hypothesis is warranted.

In this meta-analysis update, we focused on the entire body of evidence assessing the relationship between BDNF Val66Met polymorphism, life stress and depression. Firstly, we conducted a overall meta-analysis in all included studies. Secondly, three subgroup Meta-analyses were conducted stratified by the type of stressors, ancestry and the type of study designs. Usual meta-analytic methods for the combination of study findings rely on single-effect measures, such as mean differences and odds ratios (Sterne et al., 2001). However, studies addressing the interaction of gene-environment factors to predict depression often encounter the problem of heterogeneity. For that purpose, one should use appropriate methods for the meta-analysis of interaction studies which use values of the level of significance instead of raw data (Hedges and Ingram, 1985). The Lipták–Stouffer z-score method is a good example of a method that combines p-values of different studies and has been used in several relevant meta-analyses (Byrd and Manuck, 2014, Hosang et al., 2014, Karg et al., 2011) and in genomics and biostatistics (Hwang et al., 2005; Keith et al., 2012; Leung et al., 2011; Majeti et al., 2009; Richards et al., 2009). In the present study, we used the Lipták–Stouffer z-score method to combine the outcomes across the studies to investigate whether BDNF variants moderate the relationship between life stress and depression.