Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The Interaction between Childhood Bullying and the FKBP5 Gene on Psychotic-Like Experiences and Stress Reactivity in Real Life

  • Paula Cristóbal-Narváez ,

    Contributed equally to this work with: Paula Cristóbal-Narváez, Tamara Sheinbaum

    ‡ These authors are joint co-first authors on this work.

    Affiliation Departament de Psicologia Clínica i de la Salut, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain

  • Tamara Sheinbaum ,

    Contributed equally to this work with: Paula Cristóbal-Narváez, Tamara Sheinbaum

    ‡ These authors are joint co-first authors on this work.

    Affiliation Departament de Psicologia Clínica i de la Salut, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain

  • Araceli Rosa,

    Affiliations Secció de Zoologia i Antropologia, Departament de Biologia Evolutiva, Ecologia i Ciències Ambientals, Facultat de Biologia, Universitat de Barcelona (UB), Barcelona, Spain, Centre for Biomedical Research Network on Mental Health (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain, Institut de Biomedicina de la Universitat de Barcelona (IBUB), Barcelona, Spain

  • Sergi Ballespí,

    Affiliation Departament de Psicologia Clínica i de la Salut, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain

  • Marta de Castro-Catala,

    Affiliations Secció de Zoologia i Antropologia, Departament de Biologia Evolutiva, Ecologia i Ciències Ambientals, Facultat de Biologia, Universitat de Barcelona (UB), Barcelona, Spain, Institut de Biomedicina de la Universitat de Barcelona (IBUB), Barcelona, Spain

  • Elionora Peña,

    Affiliations Secció de Zoologia i Antropologia, Departament de Biologia Evolutiva, Ecologia i Ciències Ambientals, Facultat de Biologia, Universitat de Barcelona (UB), Barcelona, Spain, Institut de Biomedicina de la Universitat de Barcelona (IBUB), Barcelona, Spain

  • Thomas R. Kwapil,

    Affiliation Department of Psychology, University of North Carolina at Greensboro (UNCG), Greensboro, NC, United States of America

  • Neus Barrantes-Vidal

    neus.barrantes@uab.cat

    Affiliations Departament de Psicologia Clínica i de la Salut, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain, Centre for Biomedical Research Network on Mental Health (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain, Department of Psychology, University of North Carolina at Greensboro (UNCG), Greensboro, NC, United States of America, Sant Pere Claver–Fundació Sanitària, Barcelona, Spain

The Interaction between Childhood Bullying and the FKBP5 Gene on Psychotic-Like Experiences and Stress Reactivity in Real Life

  • Paula Cristóbal-Narváez, 
  • Tamara Sheinbaum, 
  • Araceli Rosa, 
  • Sergi Ballespí, 
  • Marta de Castro-Catala, 
  • Elionora Peña, 
  • Thomas R. Kwapil, 
  • Neus Barrantes-Vidal
PLOS
x

Abstract

Aim

The present study employed Experience Sampling Methodology to examine whether the interaction between childhood bullying and FKBP5 variability (i) is associated with the expression of psychotic-like experiences, paranoia, and negative affect, and (ii) moderates psychotic-like, paranoid, and affective reactivity to different forms of momentary stress (situational and social) in daily life.

Methods

A total of 206 nonclinical young adults were interviewed for bullying with the Childhood Experience of Care and Abuse and were prompted randomly eight times daily for one week to complete assessments of their current experiences, affect, and stress appraisals. Participants were genotyped for three FKBP5 single nucleotide polymorphisms (SNPs) (rs3800373, rs9296158, and rs1360780) that have been linked to hypothalamus-pituitary-adrenal axis reactivity. Multilevel analyses were conducted to examine the effect of the interaction between childhood bullying and the FKBP5 haplotype derived from these three SNPs.

Results

The interaction between bullying and the FKBP5 haplotype was associated with positive, but not negative, psychotic-like experiences, paranoia, and negative affect. The bullying x FKBP5 interaction also moderated the association of a social stress appraisal (specifically, being alone because people do not want to be with you) with psychotic-like experiences and negative affect in daily life. Simple slopes analyses indicated that, in all cases, the associations were significantly increased by exposure to bullying in participants with the risk haplotype, but not for those with the non-risk haplotype.

Discussion

The present study provides the first evidence of the interplay between childhood bullying and FKBP5 variability in the real-world expression of psychosis proneness and social stress reactivity. The findings underscore the importance of investigating how gene-environment interactions are involved in mechanistic pathways to the extended psychosis phenotype and lend further support to the increasing relevance given to socially defeating appraisals in the experience of reality distortion.

Introduction

Mounting evidence indicates that childhood adversity is associated with an increased risk for psychosis phenotypes [1]. The association has been repeatedly observed with psychotic disorders, subclinical psychotic symptoms, and schizotypy traits [24]—consistent with converging research supporting the notion of etiological continuity between the clinical and nonclinical manifestations of psychosis [5, 6]. Among different types of interpersonal childhood adversity, experiences within the family milieu have been more extensively studied. However, there is increasing recognition that peer relations are key to children’s and adolescent’s developmental outcomes [7] and that exposure to bullying by peers can have a host of long-term detrimental effects [8]. Bullying involves intentional aggressive/hostile behavior in the context of actual or perceived imbalance of power [9, 10] and recent findings across different study designs (including prospective studies) and populations have demonstrated its association with psychotic phenomena [1114].

It has been proposed that exposure to childhood interpersonal adversities may increase the risk for psychosis through a process of behavioral and biological sensitization involving hypothalamus-pituitary-adrenal (HPA) axis dysregulation and contributing to a final common pathway of dopamine sensitization in mesolimbic regions [15]. Dysregulation of the HPA axis involves dysregulation of the hypothalamic peptides of the corticotropin-releasing and arginine vasopressin hormones, resulting in enhanced release of plasma adrenocorticotropic hormone and glucocorticoid cortisol. Glucocorticoids promote the physiological stress response of fight-or-flight and are crucial for terminating the response through a negative feedback loop [16]. Such impaired negative feedback regulation via the glucocorticoid receptor (GR) has been proposed as a potential risk factor for stress-related psychopathology [16]. A critical regulator of GR activity is the FK506 binding protein 5 (FKBP5), a 51-kDa protein encoded by the FKBP5 gene (located on chromosome 6p21.31 in humans) [17]. Notably, a functional haplotype that comprises up to 18 single nucleotide polymorphisms (SNPs) in the FKBP5 gene has been related to increased expression of FKBP5 in response to GR activation and variation in GR sensitivity [1618]. This haplotype is tagged by three SNPs (rs3800373, rs9296158, and rs1360780) that, for this reason, have been the most studied and characterized FKBP5 polymorphisms [19]. Several studies pointed out that the risk alleles of these polymorphisms are the C, A, and T alleles, respectively [17]. Research has also suggested that the rs1360780 is the variant most likely conferring the haplotype functionality [17].

An increasing number of gene-environment interaction (GxE) studies have investigated the interaction between psychosocial stressors and FKBP5 variability, with findings suggesting that adverse childhood experiences in interaction with the above mentioned haplotype are associated with risk for a range of psychopathological phenotypes (for review, see [17]). In this regard, recent studies showed that genetic variation in the FKBP5 gene interacted with childhood trauma in the expression of psychotic phenomena in clinical and nonclinical samples [2022]. Nevertheless, to the best of our knowledge, the role of the interaction between FKBP5 variability and childhood bullying in particular has not been previously investigated.

There are also no GxE studies examining whether the interaction between childhood adverse experiences and FKBP5 variability plays a role in heightening affective, paranoid, and psychotic-like responses to stress in daily life. Researchers have increasingly employed momentary assessment strategies to examine with ecological validity the experience and expression of psychological constructs in daily life as well as their environmental triggers (e.g., [23, 24]). The Experience Sampling Method (ESM) is a within-day self-assessment technique that prompts participants at random intervals to complete brief questionnaires about their current experiences, including stress, cognition, affect, and symptoms. By assessing participants in real time and in their real-life settings, ESM offers several advantages in comparison to traditional assessment techniques. These include the minimization of retrospective bias, enhanced ecological validity, and the ability to capture the context in which experiences occur [23, 25]. ESM measures exhibit good psychometric properties and have proven useful for examining the phenomenology and stress reactivity dynamics of the clinical and subclinical psychosis phenotypes [2630]. In addition, it has been highlighted that utilizing prospective and repeated assessments of environmental exposures increases precision and reliability in the realm of GxE research [31]. Thus, the features of ESM data should enhance the power and quality of GxE studies and increase mechanistic insights that complement findings from large-scale epidemiological investigations [3235].

In a previous ESM study we found that bullying was associated with psychotic-like experiences (PLEs) in daily life as well as with increased affective and paranoid reactivity to daily life stressors [36]. However, that study did not examine what factors may interact with bullying in shaping the expression of psychotic phenomena and stress reactivity in daily life. Therefore, the present study sought to examine in a non-clinically ascertained sample of young adults whether the interaction between bullying and FKBP5 (i) is associated with PLEs, paranoia, and negative affect, and (ii) moderates psychotic-like, paranoid, and affective reactivity to different forms of momentary stress (i.e., situational and social) in daily life. We predicted that the interaction between bullying and the CAT risk haplotype of the FKBP5 gene would be associated with higher levels of PLEs, paranoia, and negative affect, but not negative-like symptoms. We also expected that the previously reported association of stress with PLEs, paranoia, and affective experiences in daily life [26], and particularly the association with social stress, would be moderated by the bullying x FKBP5 interaction, such that the association would be stronger for risk haplotype participants with childhood bullying exposure.

Methods

Ethics Statement

The present study was approved by the Universitat Autònoma de Barcelona Ethics Committee and conformed to the Helsinki Declaration. Participants had full capacity to consent to participation in research and gave written informed consent before taking part in the study.

Participants

The sample forms part of PSYRIS-Barcelona, a longitudinal study examining psychosis risk and resilience. The sample was comprised of 206 nonclinically ascertained young adults from whom usable interview, ESM, and FKBP5 genotype data were obtained. The participants were drawn from a screening sample of 589 undergraduate students (547 had complete usable data) at the Universitat Autònoma de Barcelona (Spain). Participants with high schizotypy scores were oversampled in order to ensure adequate representation of schizotypy in the current sample. A detailed description of the sample selection procedure has been provided elsewhere [26, 37]. The mean age of the sample was 21.3 years (SD = 2.4) and 78.6% were women. Ninety two percent were of European origin (subjects and both parents born in a European country).

Materials and Procedure

Bullying.

Questions from the Childhood Experience of Care and Abuse (CECA) [38] were used to assess bullying by peers. The CECA is a retrospective investigator-based interview that measures childhood experiences prior to the age of 17 years. Bullying is scored on a 4-point scale ranging from “marked” to “little/none”, based on specific rating rules and benchmarked thresholds. The interviews were conducted by psychologists and advanced graduate students in clinical psychology. Consensus meetings to discuss ratings were held regularly throughout the data collection period. The continuous severity ratings of bullying victimization were used for analyses.

ESM assessments.

The ESM data collection was conducted with personal digital assistants (PDAs) that signaled participants randomly 8 times a day (between 10 a.m. and 10 p.m.) for one week to complete short questionnaires. When participants were signaled by the PDA, they had 5 minutes to initiate responding. After this time interval or the completion of the questionnaire, the PDA shut down until the next signal. The full list of ESM items can be found in Barrantes-Vidal et al. [26]. The social contact item was answered dichotmously (alone/with others), whereas the remaining items employed in the present study were answered on 7-point scales from “not at all” to “very much”.

ESM measures of symptoms, negative affect, social contact, and stress were used for analyses. Following Barrantes-Vidal et al. [26], we created indices of PLEs (8 items: unusual senses, unusual thoughts, feeling weird, losing control, difficulty controlling thoughts, familiar things seeming strange, hearing/seeing things others could not, and passivity; alpha index = .74) and paranoia (2 items: feeling suspicious and mistreated; alpha index = .70), and used the experience of diminished thoughts/emotions (“Right now I have no thoughts or emotions”) as a measure of negative-like symptoms. Negative affect was measured by an index composed of 4 items (feeling anxious, sad, angry, and guilty; alpha index = .80). With regard to stress, note that consistent with previous ESM research (e.g., [27, 30, 39, 40]), the present study did not focus on objective environmental stressors but rather on subjective appraisals of stress in daily life. The item “My current situation is stressful” was used to assess situational stress. Regarding social stress appraisals, we distinguished between social stress when participants were alone (assessed by the item “I am alone because people do not want to be with me”) and social stress when participants were with others (assessed by 2 items: “I feel close to this person (people)” and “Right now I would prefer to be alone”). Additionally, the social contact item was included in the analyses in order to distinguish the effects of social stress appraisals from the effects of simply being alone or with others at the time of the signal.

Genotyping.

Genomic DNA was extracted using the Real Extraction DNA kit (Durviz S.L.U., Valencia, Spain). The three FKBP5 SNPs rs3800373, rs9296158, and rs1360780 were genotyped using TaqMan 5’-exonuclease allelic discrimination assay (Applied Biosystems, custom assays: C_27489960_10, C_1256775_10, and C_8852038_10, respectively). Minor allele frequencies were 0.34 for rs3800373 (allele C), 0.33 for rs9296158 (allele A), and 0.33 for rs1360780 (allele T). No differences were observed between the allele frequencies in our sample and those reported in European (EUR from 1000 genomes) and Spanish (IBS from HapMap) reference populations (p > 0.05). Genotypic frequencies for each polymorphism are reported in Table 1. All SNPs were in accordance with Hardy-Weinberg Equilibrium (all p > 0.5).

thumbnail
Table 1. Description of the FKBP5 Single Nucleotide Polymorphisms (SNPs) and Haplotype Groups (n = 206).

https://doi.org/10.1371/journal.pone.0158809.t001

Linkage disequilibrium, which is the tendency of SNPs to be inherited together, was examined by pair-wise comparisons of r2 and D’using Haploview version 4.2 [41]. The three studied SNPs were observed to be in high linkage disequilibrium (D’ = 0.89). Haplotypes considering these three polymorphisms were estimated using a bayesian approach implemented with PHASE software [42]. The frequencies of the CAT (risk haplotype) and the AGC (non-risk haplotype) were 0.29 and 0.62, respectively. For analyses, participants were classified in two groups: i) risk carriers, which included carriers of at least one risk haplotype (i.e., CAT/CAT, CAT/XXX, or CAT/AGC), and ii) non-risk carriers, which included non-carriers of the risk haplotype (i.e., AGC/AGC, AGC/XXX, or XXX/XXX) (Table 1).

Statistical Method

ESM data have a hierarchical structure in which ESM ratings (level 1 data) are nested within participants (level 2 data). Multilevel or hierarchical linear modeling provides a more appropriate method than conventional unilevel analyses for analyzing nested data and is standard for the analysis of ESM data [43, 44].

Two types of multilevel analyses were conducted in the present study. First, in order to examine the impact of the interaction of bullying and FKBP5 on PLEs, paranoia, and negative affect in daily life, we assessed the independent effect of level 2 predictors (bullying, FKBP5, and the bullying x FKBP5 interaction) on level 1 dependent measures (ESM ratings in daily life). Note that as described above we have already reported the association of bullying with ESM ratings in a study examining a wide variety of adversity exposures on daily-life experiences [36]; therefore, the main effects of bullying are not the object of the current study and will be solely described as a necessary step required to yield the GxE interaction. Second, to analyze whether the bullying x FKBP5 interaction moderates the association of momentary stress with experiences in daily life, cross-level interactions (or slopes-as-outcomes) were computed. Cross-level interactions were used to examine whether level 1 relationships (e.g., the association between feeling unwanted and PLEs in daily life) vary as a function of level 2 variables (e.g., the bullying x FKBP5 interaction). In both the analyses of direct effects and cross-level interactions, the effect of bullying and the FKBP5 haplotype were examined separately (i.e., two separate models were used, one in which bullying was the predictor and another in which the FKBP5 haplotype was the predictor).

In order to examine the effect of the bullying x FKBP5 interaction, bullying, the FKBP5 haplotype, and the interaction term were entered simultaneously in the same model. When a significant bullying x FKBP5 interaction was found, the effect of the interaction was examined within each haplotype group using simple slopes. The multilevel analyses were computed with MPlus 6 [45]. Graphics and simple slopes were computed with HLM 7.01 program [46]. Level 1 predictors were group mean centered and level 2 predictors were grand mean centered. The data departed from normality in some cases, so parameter estimates were calculated using robust standard errors. Furthermore, level 1 criteria exhibiting significant skew were treated as categorical.

Results

Participants completed an average of 40.8 usable ESM questionnaires (SD = 9.1). The FKBP5 risk haplotype and bullying were not correlated (r = 0.08) and neither was associated with the number of usable records (r = 0.03 and -0.04, respectively). Additionally, there were no sex differences in either variable (bullying: t = 0.849, p = 0.397; FKBP5 haplotype: χ² = 1.658, p = 0.198).

As shown in Table 2, bullying was associated with PLEs and negative affect but not with paranoia or negative-like symptoms. The FKBP5 haplotype was not associated with daily life symptoms or negative affect. The interaction of bullying and the FKBP5 haplotype was significantly associated with PLEs, paranoia, and negative affect. Simple slopes analyses indicated that, as expected, bullying increased PLEs and paranoia for participants with the risk haplotype (0.059, SE = 0.021, t = 2.78, p < 0.01; 0.085, SE = 0.035, t = 2.41, p < 0.05, respectively), but not for those with the non-risk haplotype (0.002, SE = 0.019, t = 0.08, ns; -0.018, SE = 0.031, t = -0.58, ns, respectively). Similarly, bullying was associated with increased negative affect for participants with the risk haplotype (0.188, SE = 0.051, t = 3.72, p < 0.001), but not for those with the non-risk haplotype (0.037, SE = 0.058, t = 0.64, ns).

thumbnail
Table 2. Main Effects of Bullying, the FKBP5 Haplotype, and their Interaction on Psychosis Spectrum Experiences and Negative Affect (n = 206).

https://doi.org/10.1371/journal.pone.0158809.t002

Cross-level interaction analyses examined whether bullying, the FKBP5 haplotype, and their interaction moderated the association of social contact and stress appraisals with PLEs, paranoia, and negative affect in daily life (Table 3). Bullying moderated the association of situational stress and preference to be alone with paranoia. It also moderated the association of social contact with PLEs and that of decreased social closeness with negative affect. The FKBP5 haplotype did not moderate the associations of situational stress or social stress with experiences in daily life. The bullying x FKBP5 interaction moderated the association of feeling unwanted when alone and PLEs in daily life. Simple slopes analyses indicated that the association between feeling unwanted and PLEs was significantly increased by exposure to bullying in participants with the risk haplotype (0.056, SE = 0.027, t = 2.07, p < 0.05; see Fig 1), but not for those with the non-risk haplotype (-0.034, SE = 0.020, t = -1.74, ns).

thumbnail
Fig 1. Association between feeling unwanted and PLEs across levels of bullying in FKBP5 risk-haplotype participants.

https://doi.org/10.1371/journal.pone.0158809.g001

thumbnail
Table 3. Cross-Level Interactions with Bullying, the FKBP5 Haplotype, and the Bullying x FKBP5 Interaction (n = 206).

https://doi.org/10.1371/journal.pone.0158809.t003

The bullying x FKBP5 interaction also moderated the association of feeling unwanted with negative affect. Simple slopes analyses showed that the association between this appraisal and negative affect was significantly increased by exposure to bullying in risk haplotype participants (0.144, SE = 0.051, t = 2.80, p < 0.01; see Fig 2), but not in non-risk haplotype participants (-0.092, SE = 0.075, t = -1.23, ns).

thumbnail
Fig 2. Association between feeling unwanted and negative affect across levels of bullying in FKBP5 risk-haplotype participants.

https://doi.org/10.1371/journal.pone.0158809.g002

Additional Analyses

Following the recommendation of a reviewer, we performed exploratory analyses partialing out the CECA parental antipathy ratings to examine whether the effects of the bullying x FKBP5 interaction on daily life outcomes were found over-and-above the effects of another form of childhood maltreatment. Specifically, we partialed the parental antipathy score out of the main effects of bullying and FKBP5, and the antipathy rating and the antipathy x bullying and antipathy x FKBP5 interaction out of the analysis of the bullying x FKBP5 interaction. This followed the reviewer's suggestion that we select a single CECA exposure to use as a confounding measure. For this sample CECA parental antipathy and role reversal exposures were available (peak rating taking into account behavior of mother and father figure). We selected parental antipathy because (i) it shares similarities with peer bullying (e.g., involving rejection, coldness, hostility) and (ii) it is carried out by different figures than peer bullying. Note that parental antipathy had a modest (although not significant) correlation with bullying in our sample (r = .13)–making it an ideal covariate. This method follows recommendations by Keller [47], who suggested that in order to properly control for potential confounders, all the covariate-by-environment and the covariate-by-gene interaction terms should be entered in the same model that tests the gene-by-environment interaction term. However, it should be interpreted cautiously in the current study because of the post hoc nature of the analyses and the lack of a priori selection of parental antipathy. We reran all of the analyses in Tables 1 and 2 using this strategy. The results were largely unchanged (see S1 and S2 Tables). All the significant bullying x FKBP5 interactions remained, and the association between situational stress and paranoia became significantly moderated by the bullying x FKBP5 interaction. Simple slopes indicated that the association was significantly increased by exposure to bullying in participants with the risk haplotype (0.044, SE = 0.018, t = 2.46, p < 0.05), but not for those with the non-risk haplotype (0.007, SE = 0.014, t = 0.48, ns). These additional analyses add support to the interpretation that the bullying x FKBP5 effects are robustly significant and attributable to this particular type of adversity.

Discussion

To the best of our knowledge, the current study is the first to examine the interplay between bullying and FKBP5 variability in the expression of psychotic phenomena and stress reactivity in the realm of daily life. The results indicated that the interaction between bullying and the risk FKBP5 haplotype was associated with PLEs, paranoia, and negative affect, and that it moderated psychotic-like and affective reactivity to a social stress appraisal (i.e., feeling unwanted by others) in a nonclinical sample. This work expands on previous GxE research supporting that the interaction between FKBP5 variability and childhood adversity exposure increases the risk for psychosis phenotypes. The findings contribute to our understanding of how the complex interplay between genetic and environmental factors is involved in the real-world expression of psychosis proneness.

The results regarding the interaction between bullying and FKBP5 variability on psychotic-like, paranoid, and affective experiences were in line with our hypotheses and provide evidence of a GxE interaction on subclinical psychotic phenomena in real life. Furthermore, the finding that the interaction was not associated with negative-like symptoms is consistent with the contention that positive and negative psychotic features may involve different etiological pathways [48], with environmental adversity exposures and biological mechanisms involved in regulating the stress response thought to be particularly relevant for the positive symptom dimension (e.g., [15]).

Prior research has consistently shown that exposure to interpersonal childhood adversities increases the risk for several psychopathological phenotypes in carriers of the functional haplotype associated with higher FKBP5 induction and prolonged cortisol responses [17]. Although next-generation sequencing projects have enabled to catalogue the broad range of variants in the FKBP5 gene (e.g., [49]), the majority of these studies have investigated FKBP5 variability using a tagging approach and focusing on some of the most common tag SNPs (rs3800373, rs9296158, or rs1360780) of this haplotype [18]. However, the investigated SNPs have not been the same in all studies and, to our knowledge, there are no studies examining specifically the role of the haplotype comprised by these three tag SNPs on psychosis proneness. Nevertheless, the finding that the bullying and FKBP5 interaction was associated with positive psychotic phenomena is consistent with a recent study showing that childhood abuse was associated with increased PLEs in carriers of the risk alleles of rs1360780 in a nonclinical sample [22]. Our results are also in agreement with the first study examining the role of FKBP5 in psychosis [20], which found that carriers of the rs1360780 and rs9296158 risk alleles (as well as rs1043805, which was not investigated here) were more vulnerable to the effect of childhood trauma on PLEs in a general population sample. In the same study, they also found that rs9296158 moderated the effect of trauma on psychotic symptoms in patients with a psychotic disorder. Of note, neither our study nor previous ones in nonclinical and clinical samples [2022, 50] found that FKBP5 variability by itself was associated with positive psychotic phenomena or presence of a psychotic disorder. Therefore, taken together, findings are in line with the notion that the contribution of FKBP5 variability to psychosis risk may be dependent upon the presence of specific environmental exposures [51].

The finding that the association of the risk alleles of the haplotype with psychotic phenomena is commonly triggered by exposure to childhood adversity is interesting in light of recent molecular studies suggesting that childhood trauma exposure could induce allele-specific epigenetic modifications that may increase the risk for stress-related phenotypes [52]. Specifically, Klengel et al. [52] found that childhood abuse exposure was associated with preferential demethylation of DNA (near a glucocorticoid response element in FKBP5) in risk allele carriers, which enhances differences in glucocorticoid receptor sensitivity and entails a dysregulation of the stress system that may eventually increase vulnerability for certain psychopathological phenotypes. Importantly, this reduced methylation seemed to be dependent specifically on childhood abuse exposure, but not adult trauma exposure or current levels of cortisol, indicating that there may be a critical developmental stage for such epigenetic effects [52].

Regarding stress reactivity, we found that the GxE interaction moderated the association of appraisals of being unwanted when alone with PLEs and negative affect. In particular, our results indicated that these associations were significantly increased by exposure to bullying in risk haplotype participants, but not in non-risk haplotype participants. By contrast, the interaction did not moderate affective and symptomatic reactivity to situational stress and other forms of social stress (i.e., appraisals of diminished closeness and increased preference for being alone). In light of these results, it is attractive to speculate that social defeat is a mechanism involved in increasing reactivity in individuals with the risk haplotype.

More specifically, it has been suggested that childhood adversity may increase psychosis vulnerability by inducing a state of social defeat, characterized by feelings of outsider status and decreased self-value [53, 54]. Of note, recent research has indicated that social defeat plays a mediating role in the association between childhood trauma and psychotic phenotypes at the population level [55], and that a history of social defeat increases the likelihood of psychotic responses during social interactions in an experimental social environment generated by Virtual Reality in clinically at-risk individuals [56]. Bullying has been conceptualized as a socially defeating experience and its parallels with animal models of social defeat have been highlighted [57]. Likewise, the appraisal of being alone because others do not want to be with you could be considered a proximal micro-level experience of social defeat. Previous work indicated that mice lacking the FKBP5 gene showed decreased neuroendocrine/physiological responses to chronic social defeat stress (as compared with wild-type animals), pointing to an increased glucocorticoid negative feedback of the HPA axis that may be modulated by heightened GR sensitivity [58]. Such findings support human studies suggesting that FKBP5 risk alleles may increase sensitivity to psychosocial adversities through an enhanced FKBP5 expression and thereby diminished GR sensitivity [16, 17]. In this context, our results may therefore suggest that the FKBP5 risk haplotype amplifies the likelihood that distal experiences of social defeat will increase psychotic-like reactivity to proximal socially defeating appraisals.

Strengths of the present study include the use of an interview measure to assess bullying, which allowed to obtain in-depth information and minimize biases related to subjective responding [59]. The estimation of the risk haplotype is also a strength given that it reports the full variability of a DNA fragment and increases the power to find genetic associations [60]. In addition, we employed ecologically valid measures of experiences obtained prospectively and repeatedly during a one-week period, increasing the power and reliability of GxE research [31, 32]. Finally, although we computed multiple analyses, they were limited to a priori goals and hypotheses of the study to avoid exploratory analyses that would increase the risk of Type I error. Also, following the suggestion of a reviewer, we confirmed that the findings reported remain largely unchanged after partialing out the effect of another type of adversity (parental antipathy), which strengthens the role of bullying as a relevant exposure. Limitations of the study include the cross-sectional nature of the data, which limits interpretations about the causal effects of GxE interactions. Similarly, given that predictor and criterion ESM variables were measured concurrently, causal inferences regarding the effects of stress appraisals cannot be definitively made. Furthermore, the generalizability of the present results is limited by the use of a predominantly female university student sample. Future studies should investigate at-risk and clinical samples to identify whether the interaction between bullying and FKBP5 variability is relevant across the psychosis continuum. Likewise, further research may consider assessing whether the direct and cross-level effects reported in the current study are found over-and-above the effects of other childhood adversity exposures.

To conclude, the present study provides a novel contribution by showing that bullying and the FKBP5 risk haplotype interact in shaping the expression of reality distortion and social stress reactivity in real life. The current study concurs with and expands previous work by providing evidence supporting the 3-hit [51] and sensitization [15] hypotheses, that is, the relevance of the interaction of 1) genetic risk, 2) distal environmental factors and 3) proximal environmental re-exposures on the expression of psychosis proneness. Our findings highlight that examining the interplay between genetic and environmental factors should increase our understanding of the mechanistic pathways leading to the extended psychosis phenotype and further support the increasing relevance given to socially defeating appraisals in the experience of reality distortion.

Supporting Information

S1 Table. Main Effects of Bullying, the FKBP5 Haplotype, and their Interaction on Psychosis Spectrum Experiences and Negative Affect Partialing out the Effects of Parental Antipathy (n = 206).

https://doi.org/10.1371/journal.pone.0158809.s001

(DOCX)

S2 Table. Cross-Level Interactions with Bullying, the FKBP5 Haplotype, and the Bullying x FKBP5 Interaction Partialing out the Effects of Parental Antipathy (n = 206).

https://doi.org/10.1371/journal.pone.0158809.s002

(DOCX)

Author Contributions

Conceived and designed the experiments: NBV TRK AR. Performed the experiments: TS SB NBV EP MCC AR. Analyzed the data: PCN TRK. Contributed reagents/materials/analysis tools: TRK. Wrote the paper: PCN TS NBV. Critically revised the manuscript: TRK AR MCC SB EP.

References

  1. 1. Varese F, Smeets F, Drukker M, Lieverse R, Lataster T, Viechtbauer W, et al. Childhood adversities increase the risk of psychosis: a meta-analysis of patient-control, prospective- and cross-sectional cohort studies. Schizophr Bull. 2012;38: 661–671. pmid:22461484
  2. 2. Matheson SL, Shepherd AM, Pinchbeck RM, Laurens KR, Carr VJ. Childhood adversity in schizophrenia: a systematic meta-analysis. Psychol Med. 2013;43: 225–238. pmid:22716913
  3. 3. van Winkel R, van Nierop M, Myin-Germeys I, van Os J. Childhood trauma as a cause of psychosis: linking genes, psychology, and biology. Can J Psychiatry. 2013;58: 44–51. pmid:23327756
  4. 4. Velikonja T, Fisher HL, Mason O, Johnson S. Childhood trauma and schizotypy: a systematic literature review. Psychol Med. 2015;45: 947–963. pmid:25273151
  5. 5. Barrantes-Vidal N, Grant P, Kwapil TR. The role of schizotypy in the study of the etiology of schizophrenia-spectrum disorders. Schizophr Bull. 2015;41 Suppl 2: S408–S416. pmid:25810055
  6. 6. van Os J, Linscott RJ, Myin-Germeys I, Delespaul P, Krabbendam L. A systematic review and meta-analysis of the psychosis continuum: evidence for a psychosis proneness-persistence-impairment model of psychotic disorder. Psychol Med. 2009;39: 179–195. pmid:18606047
  7. 7. Parker JG, Rubin KH, Erath SA, Wojslawowicz JC, Buskirk AA. Peer relationships, child development, and adjustment: a developmental psychopathology perspective. In: Cicchetti D D., Cohen DJ, editors. Developmental psychopathology. Vol. 1: Theory and method. 2nd ed. Hoboken, NJ: John Wiley & Sons. 2006. pp. 419–493.
  8. 8. Wolke D, Lereya ST. Long-term effects of bullying. Arch Dis Child. 2015;100: 879–885. pmid:25670406
  9. 9. Olweus D. Bullying at school: basic facts and effects of a school based intervention program. J Child Psychol Psychiatry. 1994;35: 1171–1190. pmid:7806605
  10. 10. Ostrov JM, Kamper KE. Future directions for research on the development of relational and physical peer victimization. J Clin Child Adolesc Psychol. 2015;44: 509–519. pmid:25751392
  11. 11. Arseneault L, Cannon M, Fisher HL, Polanczyk G, Moffitt TE, Caspi A. Childhood trauma and children's emerging psychotic symptoms: a genetically sensitive longitudinal cohort study. Am J Psychiatry. 2011;168: 65–72. pmid:20952460
  12. 12. Kelleher I, Keeley H, Corcoran P, Ramsay H, Wasserman C, Carli V, et al. Childhood trauma and psychosis in a prospective cohort study: cause, effect, and directionality. Am J Psychiatry. 2013;170: 734–741. pmid:23599019
  13. 13. van Dam DS, van der Ven E, Velthorst E, Selten JP, Morgan C, de Haan L. Childhood bullying and the association with psychosis in non-clinical and clinical samples: a review and meta-analysis. Psychol Med. 2012;42: 2463–2474. pmid:22400714
  14. 14. Wolke D, Lereya ST, Fisher HL, Lewis G, Zammit S. Bullying in elementary school and psychotic experiences at 18 years: a longitudinal, population-based cohort study. Psychol Med. 2014;44: 2199–2211. pmid:24342773
  15. 15. van Winkel R, Stefanis NC, Myin-Germeys I. Psychosocial stress and psychosis. A review of the neurobiological mechanisms and the evidence for gene-stress interaction. Schizophr Bull. 2008;34: 1095–1105. pmid:18718885
  16. 16. Binder EB. The role of FKBP5, a co-chaperone of the glucocorticoid receptor in the pathogenesis and therapy of affective and anxiety disorders. Psychoneuroendocrinology. 2009;34 Suppl 1: S186–S195. pmid:19560279
  17. 17. Zannas AS, Binder EB. Gene-environment interactions at the FKBP5 locus: sensitive periods, mechanisms and pleiotropism. Genes Brain Behav. 2014;13: 25–37. pmid:24219237
  18. 18. Binder EB, Salyakina D, Lichtner P, Wochnik GM, Ising M, Pütz B, et al. Polymorphisms in FKBP5 are associated with increased recurrence of depressive episodes and rapid response to antidepressant treatment. Nat Genet. 2004;36: 1319–1325. pmid:15565110
  19. 19. Zannas AS, Wiechmann T, Gassen NC, Binder EB. Gene-stress-epigenetic regulation of FKBP5: clinical and translational implications. Neuropsychopharmacology. 2016;41: 261–274. pmid:26250598
  20. 20. Collip D, Myin-Germeys I, Wichers M, Jacobs N, Derom C, Thiery E, et al. FKBP5 as a possible moderator of the psychosis-inducing effects of childhood trauma. Br J Psychiatry. 2013;202: 261–268. pmid:23429203
  21. 21. Ajnakina O, Borges S, Di Forti M, Patel Y, Xu X, Green P, et al. Role of environmental confounding in the association between FKBP5 and first-episode psychosis. Front Psychiatry. 2014;5.
  22. 22. Alemany S, Moya J, Ibáñez MI, Villa H, Mezquita L, Ortet G, et al. Childhood trauma and the rs1360780 SNP of FKBP5 gene in psychosis: a replication in two general population samples. Psychol Med. 2016;46: 221–223. pmid:26399750
  23. 23. Conner TS, Tennen H, Fleeson W, Barrett LF. Experience sampling methods: a modern idiographic approach to personality research. Soc Personal Psychol Compass. 2009;3: 292–313. pmid:19898679
  24. 24. Oorschot M, Kwapil T, Delespaul P, Myin-Germeys I. Momentary assessment research in psychosis. Psychol Assess. 2009;21: 498–505. pmid:19947784
  25. 25. Hektner JM, Schmidt JA, Csikszentmihalyi M. Experience sampling method: Measuring the quality of everyday life. Thousand Oaks: Sage Publications; 2007.
  26. 26. Barrantes-Vidal N, Chun CA, Myin-Germeys I, Kwapil TR. Psychometric schizotypy predicts psychotic-like, paranoid, and negative symptoms in daily life. J Abnorm Psychol. 2013;122: 1077–1087. pmid:24364610
  27. 27. Chun CA, Barrantes-Vidal N, Sheinbaum T, Kwapil TR. Expression of schizophrenia-spectrum personality traits in daily life. Personal Disord. 2015.
  28. 28. Kwapil , Kwapil TR, Brown LH, Silvia PJ, Myin-Germeys I, Barrantes-Vidal N. The expression of positive and negative schizotypy in daily life: an experience sampling study. Psychol Med. 2012;42: 2555–2566. pmid:22716971
  29. 29. Myin-Germeys I, Delespaul P, van Os J. The Experience Sampling Method in psychosis research. Curr Opin Psychiatry. 2003;16 Suppl 2: S33–S38.
  30. 30. Reininghaus U, Kempton MJ, Valmaggia L, Craig TK, Garety P, Onyejiaka A, et al. Stress sensitivity, aberrant salience, and threat anticipation in early psychosis: an experience sampling study. Schizophr Bull. In press.
  31. 31. Moffitt TE, Caspi A, Rutter M. Strategy for investigating interactions between measured genes and measured environments. Arch Gen Psychiatry. 2005;62: 473–481. pmid:15867100
  32. 32. Myin-Germeys I, Oorschot M, Collip D, Lataster J, Delespaul P, van Os J. Experience sampling research in psychopathology: opening the black box of daily life. Psychol Med. 2009;39: 1533–1547. pmid:19215626
  33. 33. Sheinbaum T, Barrantes-Vidal N. Mechanisms mediating the pathway from environmental adversity to psychosis proneness. In: Mason O, Claridge G, editors. Schizotypy: New dimensions. New York: Routledge; 2015. pp. 116–131.
  34. 34. van Winkel M, Peeters F, van Winkel R, Kenis G, Collip D, Geschwind N, et al. Impact of variation in the BDNF gene on social stress sensitivity and the buffering impact of positive emotions: replication and extension of a gene-environment interaction. Eur Neuropsychopharmacol. 2014;24: 930–938. pmid:24613654
  35. 35. Wichers M. The dynamic nature of depression: a new micro-level perspective of mental disorder that meets current challenges. Psychol Med. 2014;44: 1349–1360. pmid:23942140
  36. 36. Cristóbal-Narváez P, Sheinbaum T, Ballespí S, Mitjavila M, Myin-Germeys I, Kwapil TR, et al. Impact of adverse childhood experiences on psychotic-like symptoms and stress reactivity in daily life in nonclinical young adults. Manuscript submitted for publication.
  37. 37. Barrantes-Vidal N, Gross GM, Sheinbaum T, Mitjavila M, Ballespí S, Kwapil TR. Positive and negative schizotypy are associated with prodromal and schizophrenia-spectrum symptoms. Schizophr Res. 2013;145: 50–55. pmid:23402694
  38. 38. Bifulco A, Brown GW, Harris TO. Childhood Experience of Care and Abuse (CECA): a retrospective interview measure. J Child Psychol Psychiatry. 1994;35: 1419–1435. pmid:7868637
  39. 39. Palmier-Claus JE, Dunn G, Lewis SW. Emotional and symptomatic reactivity to stress in individuals at ultra-high risk of developing psychosis. Psychol Med. 2012;42: 1003–1012. pmid:22067414
  40. 40. Hernaus D, Collip D, Lataster J, Viechtbauer W, Myin E, Ceccarini J, et al. Psychotic reactivity to daily life stress and the dopamine system: a study combining experience sampling and [18F]fallypride positron emission tomography. J Abnorm Psychol. 2015;124: 27–37. pmid:25688430
  41. 41. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21: 263–265. pmid:15297300
  42. 42. Stephens M, Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet. 2003;73: 1162–1169. pmid:14574645
  43. 43. Bolger N, Laurenceau JP. Intensive longitudinal methods: An introduction to diary and experience sampling research. New York: Guilford Press; 2013.
  44. 44. Nezlek J. Multilevel modeling for social and personality psychology. London: Sage Publications; 2011.
  45. 45. Muthén LK, Muthén BO. Mplus user’s guide. 6th ed. Los Angeles: Muthén & Muthén; 1998–2010.
  46. 46. Raudenbush SW, Bryk AS, Congdon R. HLM 7.01 for Windows [Computer software]. Skokie, IL: Scientific Software International, Inc; 2013.
  47. 47. Keller MC. Gene × environment interaction studies have not properly controlled for potential confounders: the problem and the (simple) solution. Biol Psychiatry. 2014;75: 18–24. pmid:24135711
  48. 48. Kwapil TR, Barrantes-Vidal N. Schizotypy: looking back and moving forward. Schizophr Bull. 2015;41 Suppl 2: S366–S373. pmid:25548387
  49. 49. Ellsworth KA, Moon I, Eckloff BW, Fridley BL, Jenkins GD, Batzler A, et al. FKBP5 genetic variation: association with selective serotonin reuptake inhibitor treatment outcomes in major depressive disorder. Pharmacogenet Genomics. 2013;23: 156–166. pmid:23324805
  50. 50. Gawlik M, Moller-Ehrlich K, Mende M, Jovnerovski M, Jung S, Jabs B, et al. Is FKBP5 a genetic marker of affective psychosis? A case control study and analysis of disease related traits. BMC Psychiatry. 2006;6.
  51. 51. Daskalakis NP, Binder EB. Schizophrenia in the spectrum of gene-stress interactions: the FKBP5 example. Schizophr Bull. 2015;41: 323–329. pmid:25592294
  52. 52. Klengel T, Mehta D, Anacker C, Rex-Haffner M, Pruessner JC, Pariante CM, et al. Allele-specific FKBP5 DNA demethylation mediates gene-childhood trauma interactions. Nat Neurosci. 2013;16: 33–41. pmid:23201972
  53. 53. Selten JP, van der Ven E, Rutten BP, Cantor-Graae E. The social defeat hypothesis of schizophrenia: an update. Schizophr Bull. 2013;39: 1180–1186. pmid:24062592
  54. 54. Hammels C, Pishva E, De Vry J, van den Hove DL, Prickaerts J, van Winkel R, et al. Defeat stress in rodents: from behavior to molecules. Neurosci Biobehav Rev. 2015;59: 111–140. pmid:26475995
  55. 55. van Nierop M, van Os J, Gunther N, van Zelst C, de Graaf R, ten Have M, et al. Does social defeat mediate the association between childhood trauma and psychosis? Evidence from the NEMESIS-2 Study. Acta Psychiatr Scand. 2014;129: 467–476. pmid:24571736
  56. 56. Valmaggia LR, Day F, Garety P, Freeman D, Antley A, Slater M, et al. Social defeat predicts paranoid appraisals in people at high risk for psychosis. Schizophr Res. 2015;168: 16–22. pmid:26276306
  57. 57. Björkqvist K. Social defeat as a stressor in humans. Physiol Behav. 2001;73: 435–442. pmid:11438372
  58. 58. Hartmann J, Wagner KV, Dedic N, Marinescu D, Scharf SH, Wang XD, et al. Fkbp52 heterozygosity alters behavioral, endocrine and neurogenetic parameters under basal and chronic stress conditions in mice. Psychoneuroendocrinology. 2012;37: 2009–2021. pmid:22641006
  59. 59. Bifulco A, Thomas G. Understanding adult attachment in family relationships: Research, assessment, and intervention. Abingdon, UK: Routledge; 2013.
  60. 60. Crawford DC, Nickerson DA. Definition and clinical importance of haplotypes. Annu Rev Med. 2005;56: 303–220. pmid:15660514