Risk and resilience to depression and anxiety disorders after
childhood maltreatment: a sibling study
The role of personality traits neuroticism, extraversion and locus of control
Author: F. A. J. Gigase ID: S2085097
Supervisor: Marie-Louise Kullberg, MSc
Master Thesis Clinical Psychology Institute of Psychology
Universiteit Leiden April 1, 2019
Contents Page: Abstract 3 Introduction 4 Methods 7 Sample 7 Procedure 8 Measures 9
Presence of depression and/or anxiety disorder(s) 9
Depressive and anxiety symptoms 9
Childhood maltreatment 9
Neuroticism & Extraversion 10
Locus of Control 10
Statistical Analysis 10
Results 11
Sample description and between group differences 11
Correlation 13
Multilevel modeling 15
Model depression scores (IDS) 15
Model anxiety scores (BAI) 18
Discussion 20
Appendix: table 5 26
Abstract
Introduction: Childhood maltreatment has been associated with adult psychopathology. Individuals differ in the extent to which they develop depression and/or anxiety disorders after childhood maltreatment. Personality traits influence the relation between childhood maltreatment and adult psychopathology in between-family studies. The current study investigates the role of neuroticism, extraversion and locus of control as risk factors to depression and/or anxiety after childhood maltreatment using a sibling design, to control for shared familial background. We assess whether lifetime affected subjects, their unaffected siblings and a control group differ in terms of neuroticism, extraversion and locus of control and whether these personality traits moderate the relationship between childhood maltreatment and depression and/or anxiety symptom severity. Methods: Data was collected at two timepoints using various self-report questionnaires (CTQ, NEO-FFI, PM, BAI, IDS) as part of an ongoing 8-year longitudinal cohort study (NESDA). For the first analyses participants were divided into three groups: 447 subjects with lifetime depression and/or anxiety; 189 unaffected siblings with no lifetime depression and/or anxiety and 389 control subjects. For the second analyses, to investigate the role of personality traits on the association between childhood maltreatment and depression and anxiety, multilevel model regression analyses were used to account for nested data. Results: Unaffected siblings report significantly higher depression scores and lower extraversion and locus of control scores compared to the control group. Neuroticism was found to moderate the relation between childhood maltreatment and anxiety, but not depression. Conclusion: We emphasize the importance of studying childhood maltreatment, depression and anxiety in the context of the family. In addition, we identified neuroticism as a risk factor of developing anxiety disorder after childhood maltreatment while controlling for shared family background. We point out that unaffected siblings seem to be at increased risk of developing depression and that among siblings especially those who are more neurotic are more likely to develop anxiety disorder after childhood maltreatment.
Introduction
Childhood maltreatment is a complex global problem and can cast a long shadow over one’s life (Gilbert et al., 2009). Childhood maltreatment has been associated with adverse adult implications including drug use, suicide attempts, risky sexual and criminal behavior and the development of psychopathology (Gilbert et al., 2009; Norman et al., 2012). The experience of childhood adversities is related to first onset of numerous psychiatric disorders in low to high income countries around the globe (Kessler et al., 2010). Individuals who were subjected to child maltreatment are considered to be at greater risk of developing depression and/or anxiety at some point in their lives (Collishaw et al., 2007; Nanni, Uher, & Danese, 2012). More specifically, it was found that maltreated children are twice as likely to develop both recurrent and persistent depressive episodes (Nanni, 2012). Emotional abuse and physical and emotional neglect in particular appear to be strongly related to adverse mental health outcomes after childhood maltreatment as compared to sexual abuse (Hovens, Giltay, Spinhoven, Van Hemert, & Penninx, 2015; Norman et al., 2012; Powers, Ressler, & Bradley, 2009). In addition, childhood maltreatment was found to be predictive of unfavorable course of illness and treatment in depression and has been associated with an increased persistence of comorbidity and chronicity in affected adults (Hovens et al., 2015; Nanni et al., 2012; P. Spinhoven et al., 2011).
Interestingly, individuals differ to the extent to which they develop mental health problems after childhood maltreatment, with some individuals experiencing psychiatric symptoms whereas others do not develop symptoms at all (Amstadter, Myers, & Kendler, 2014; Bonanno, 2004). Risk factors that possibly modify the association between childhood maltreatment and adult psychopathology include the experience of complex family circumstances, low socio-economic background and lack of social support (Brown, Cohen, Johnson, & Smailes, 1999; A. E. Farmer & McGuffin, 2003; Kendler et al., 2011). In addition, various resilience factors can be at play including perceived parental care and social support and inter-personal qualities such as adult peer and romantic relationships (Cheong, Sinnott, Dahly, & Kearney, 2017; Collishaw et al., 2007; Powers et al., 2009).
Moreover, personality may affect the likelihood of developing psychiatric illness after childhood maltreatment. Personality traits neuroticism, extraversion and locus of
control are characteristics that may differ between individuals, even between siblings from the same family, and alter their perception and/or coping mechanisms (Plomin, Asbury, & Dunn, 2001). A neurotic personality is characterized by persistent and excessive worry about the occurrence of an adverse event (Costa & McCrae, 1992; A. Farmer et al., 2002). Extraversion is associated with outgoing behavior and extraverts often feel a sense of reward in social situations, networking experiences and when receiving attention (Magnus, Diener, Fujita, & Pavot, 1993). Locus of control represents one’s perceived level of mastery (Rotter, 1966). An individual reporting an internal locus of control believes that events and their outcomes are under their own influence, whereas someone with an external locus of control tends to lay the blame for these events with outside forces (Dağ & Şen, 2018; Rotter, 1966).
Various studies have looked into the role of neuroticism, extraversion and locus of control in the development of psychopathology after child maltreatment. It was found that neuroticism and extraversion are related to prognosis and/or chronicity of various affective disorders after childhood maltreatment (A. Farmer et al., 2002; Plomin et al., 2001; P. Spinhoven et al., 2011; Philip Spinhoven, Roelofs, et al., 2011; Struijs, Lamers, Spinhoven, van der Does, & Penninx, 2018). More specifically, neuroticism appears to be predictive of a poor prognosis and extraversion of a more favorable prognosis measured in terms of diagnostic status after two years, time to remission and clinical course trajectory (Spinhoven, 2011). The level of mastery has been shown to be a protective buffer for one’s mental health when facing persistent life stresses (Pearlin & Schooler, 1978; Pudrovska, Schieman, Pearlin, & Nguyen, 2005). In addition, it has been shown that external locus of control is associated with chronicity of various affective disorders including depression, anxiety, borderline personality disorder and psychosis (Hope, Wakefield, Northey, & Chapman, 2018; Struijs et al., 2018; Sullivan, Thompson, Kounali, Lewis, & Zammit, 2017).
Previous studies on the role of personality traits in the development of adult psychopathology after childhood maltreatment are mostly based on samples of unrelated individuals and between-family designs, whilst few studies have looked at siblings. Studying associations between traits and outcomes in siblings has some benefits over using a between-family design (Dick, Johnson, Viken, & Rose, 2000). Namely, a within-family design largely eliminates the role of heritable and experiential influences that are common to children raised in the same family and controls for
variation in family structure and environment, socio-economic factors, neighborhood and education (Dick et. al., 2000; Laporte, Paris, Guttman, Russell, & Correa, 2012). Therefore, it reduces the likelihood that unmeasured confounding factors explain the observed associations (Laporte 2012; McGrath, 2010, Dunn 1991). Consequently, a within-family study allows for the comparison of siblings who have almost identical familial and environmental surroundings yet show differences regarding their susceptibility to depression and/or anxiety. To illustrate, a divorce of parents casts a shadow over the whole family and puts all children of that family at increased risk of developing adult depression (L. & T.S.S., 2011). Siblings may be affected differently due to varying experiences and interpretations of their shared environment (Pike & Plomin, 1996). One child may be equipped with sufficient psychological resilience and experiences no identifiable adverse consequences of the divorce, whereas another child may somehow be predisposed to the development of depression and the divorce may prompt its onset. Shared environmental factors, such as socioeconomic status, are unlikely to contribute to the differences in developmental outcomes for these children (Plomin, 2011; Laporte, 2012). Non-shared environmental factors, on the other hand, may account for differences in psychological development in siblings, as such that siblings in the same family can grow up to be very different from each other (Plomin, 2011, Dick, 2000). By assessing children from the same family, we can identify non-shared environmental risk and resilience factors that possibly explain why siblings are so different in terms of psychopathology outcome after childhood maltreatment.
It is important to gain a better understanding of what is involved in the negative adaptation of maltreated individuals. Although there is considerable uncertainty about the prevalence and severity of child maltreatment, the World Health Organization (WHO) estimates the frequency of physical abuse in childhood to be around 25-50% (Norman et al., 2012). The serious long-term consequences of child maltreatment should encourage better identification of those at increased risk of developing adult psychopathology (Norman et al., 2012). To our knowledge, neuroticism, extraversion and locus of control have not yet been investigated in relation to the development of psychopathology after childhood maltreatment in a sibling study. Through the assessment of siblings and their personality traits, this study aims to add to the ongoing investigation of risk and resilience factors affecting the development of adult depression and anxiety disorder after childhood maltreatment using a design that
inherently models for shared environmental influences. This may lead to promising insights for future treatments, preventions and intervention techniques.
The current study aims to assess the relation between childhood maltreatment and adult depression and/or anxiety disorder severity and the involvement of neuroticism, extraversion and locus of control using a within-family design. We intent to answer the following questions: 1) do individuals with a lifetime depression and/or anxiety disorder, their unaffected siblings and healthy controls differ from each other with respect to neuroticism, extraversion and locus of control scores; and 2) taking into account the shared family environment of the lifetime affected individuals and their unaffected siblings using a sibling design, do neuroticism, extraversion and locus of control moderate the relation between childhood maltreatment and adult depression and/or anxiety severity? It is hypothesized that: 1) individuals with a lifetime depression and/or anxiety disorder, their unaffected siblings and healthy controls differ from each other with respect to neuroticism, extraversion and locus of control scores. More specifically, we hypothesize that high neuroticism, low extraversion and external locus of control is more prevalent in lifetime affected individuals compared to their unaffected siblings and the healthy controls and that these scores are slightly more prevalent among unaffected siblings compared to the control group. In addition, it is hypothesized that 2) childhood maltreatment is associated with adult depressive and/or anxiety disorder severity and that this association is moderated by neuroticism, extraversion and locus of control, controlling for a shared family background using a sibling study.
Methods
Sample
This study draws data from the Netherlands Study of Depression and Anxiety (NESDA) database (Penninx et al., 2008). NESDA is an ongoing 8-year longitudinal cohort study designed to investigate the determinants, disease course and consequences of depression and anxiety in individuals from different social, biological and cultural backgrounds. The sample includes 2981 participants with a history of depressive and anxiety disorders, subjects currently affected with a depressive and/or anxiety disorder
and healthy controls. Subjects were recruited from community, primary care and specialized mental health care, as to represent a variety of health care settings and developmental stages of disorder. Subjects were excluded on the basis of a primary diagnosis of other psychiatric conditions.
For the current study, 244 subjects from the original NESDA cohort were included, based on the following additional inclusion criteria: 1) they reported a depressive or anxiety episode at two time points during NESDA measurements, meaning that they are lifetime affected; 2) they share 100% the same parents with their siblings; 3) they participated in at least three out of four NESDA meetings; 4) GWA data is available and 5) they approve of contacting siblings for research purposes. These subjects are referred to as ‘targets’. The targets’ siblings were addressed and 367 were recruited based on the following inclusion criteria: 1) the sibling is currently living in the Netherlands; 2) in line with the NESDA baseline inclusion criteria, the sibling is not familiar with serious psychiatric problems, but may have depression and or anxiety disorder; 3) the sibling is aged between 18 and 77 and 4) the sibling is willing to participate in a face to face investigation. In addition, sibling pairs who were already included in the NESDA database and who met the inclusion criteria were added to the sample and divided between the target and sibling group, resulting in 256 additional targets and 380 additional siblings (N=636). Note that the sibling participants have at least one sibling with lifetime depression and/or anxiety included in the study. In addition, note that siblings can be affected (lifetime depression and/or anxiety) or unaffected (no lifetime depression and/or anxiety). Lastly, 389 unrelated healthy controls who have no history or current episode of depression and/or anxiety disorder were included. For the analyses, the participants are divided into groups based on presence of depression and/or anxiety disorder: group 1 consists of the NESDA targets and affected siblings, that is all subjects with lifetime depression and/or anxiety (N=447); group 2 consists of unaffected siblings (N=189) and group 3 includes healthy control subjects (N=389).
Procedure
Data is collected on childhood maltreatment, severity of depression and/or anxiety symptoms and personality characteristics neuroticism, extraversion and locus of control through various self-report measures and interviews. Measurements were obtained at
wave 4 (W4), between 2010 and 2013, six years after baseline and at wave 6 (W6), between 2013 and 2016, nine years after baseline.
Measures
Presence of depression and/or anxiety disorder(s)
The presence of DSM-IV depressive disorder (Major depressive disorder, Dysthymia) and/or anxiety (Social Anxiety Disorder, Generalized Anxiety Disorder, Panic Disorder with/without Agoraphobia) was established using the Composite Interview Diagnostic Instrument (CIDI). The CIDI is a standardized and validated interview that can be used to classify psychiatric diagnoses according to the DSM-IV criteria (Wittchen, 1994). Data was obtained at W6.
Depressive and anxiety symptoms
The severity of anxiety was assessed using the self-report questionnaire Beck Anxiety Index (BAI), a 21-item questionnaire assessing the presence and frequency of symptoms related to generalized anxiety, including heart pounding, nervousness, and fear of dying (Beck, Epstein, Brown, & Steer, 1988). Cronbach’s alpha for the BAI was .980. The severity of depression was assessed using the self-report questionnaire Inventory of Depressive Symptoms (IDS), a 30-item questionnaire addressing one’s sleep, feelings of misery and other symptoms associated with depression (John Rush et al., 1986). Cronbach’s alpha for the IDS was .950. Both were measured at W6.
Childhood maltreatment
Childhood maltreatment was retrospectively assessed through the self-report measure Childhood Trauma Questionnaire (CTQ). The short form CTQ (CTQ-SF) is a 25-item test that sets out to measure five types of maltreatment including three domains of childhood abuse (sexual, physical and emotional) and two domains of childhood neglect (physical and emotional). Cronbach’s alpha for the CTQ was .880. The CTQ has shown good sensitivity and specificity and is a good way to classify cases of abuse and neglect (Bernstein, Ahluvalia, Pogge, & Handelsman, 1997). Data was obtained at W4 for targets and at W6 for siblings and healthy controls.
Neuroticism & Extraversion
Personality traits neuroticism and extraversion were assessed using the NEO Five-Factor Inventory (NEO-FFI). The NEO-FFI is a self-report questionnaire used to measure five personality domains, namely: neuroticism (N), extraversion (E), openness (O), agreeableness (A) and conscientiousness (C). For the current analyses, the measures of neuroticism (negative affect and self-reproach; 12 items) and extraversion (positive affect, sociability, activity; 12 items) were included (Costa & McCrae, 1992). Cronbach’s alpha for neuroticism was .752 and for extraversion .793. Data was obtained at W6.
Locus of control
The locus of control (LoC) is measured through the self-report measure Pearlin & Schooler Mastery Scale (PM). The PM measures a person’s level of mastery, namely the extent to which one regards that one’s chances and events are under one’s own control (internal locus of control) as opposed to being fatalistically ruled (external locus of control) (Pearlin & Schooler, 1978). A higher score indicates a higher internal locus of control as opposed to external. The PM version used in NESDA consists of five items. Cronbach’s alpha was .960. Data was obtained at W6.
Statistical Analysis
For the first analyses, the sample is divided into three groups: group 1 consists of the NESDA targets and affected siblings, that is all subjects with lifetime depression and/or anxiety (N=447); group 2 consists of unaffected siblings (N=189) and group 3 includes healthy control subjects (N=389).
To answer the first research question we compared N, E and LoC scores between groups 1, 2 and 3 with a one-way ANOVA to test whether groups differ significantly in terms of personality scores. Post-hoc Bonferroni tests are performed to control for multiple testing and check for significant differences in predictor variables between groups. Next, a correlation analysis is carried out of childhood maltreatment (CTQ), depressive and/or anxiety symptom severity (BAI; IDS, respectively) and personality characteristics (N, E, LoC) to confirm the relationship between childhood maltreatment and depressive and/or anxiety symptom severity. In addition, correlation coefficients are compared between the three groups to assess whether groups differ significantly from each other in terms of associations between childhood maltreatment,
clinical symptom severity and personality traits, by transforming r coefficients to z scores using the Fisher’s Z transformation formula: Zob se rv ed = (z1 – z2) / (square root of [ (1 / N1 – 3) + (1 / N2 – 3) ].
For the next analyses, the healthy control group is excluded and only affected individuals and unaffected siblings are included (groups 1 and 2). To assess the moderation effect of neuroticism, extraversion and locus of control on the relationship between childhood maltreatment and depression/anxiety symptom severity, taking into account that data is nested and that the subjects from the same family are related to each other, a multilevel model is constructed. Using a multilevel model analysis allows to assess whether being part of a family partly explains the variance in severity of depression and/or anxiety, in addition to the variance explained by the fixed variables childhood maltreatment (CTQ), N, E and LoC. A mixed effects random intercept model is constructed to assess whether and to what extent variance in IDS and BAI scores are explained by family, age, gender, educational level, CTQ, LoC, N, and E as fixed factors. The first model contains the control variables age, sex and education level. In model 2, the predictor CTQ is added. Model 3 also includes N, E and LoC. In model 4, the interaction variables of CTQ with N, E and LoC respectively are added. Separate multilevel analyses are carried out for dependent variables depression and anxiety symptom severity. The intra class correlation (ICC) is calculated for each model (between family variance / total variance) to assess the variance within families that is explained by the predictor(s) added to the model. Model improvement of the new model relative to previous models was assessed using the Akaike Information Criterion (AIC).
Missing values, due to different assessment waves, are replaced by the series mean. A significance level of p< .05 is used for all analyses. Analyses are run using SPSS version 25 (IBM Corp. Released 2015, 2014).
Results
Sample description and between group differences
Data was available for 447 NESDA targets and affected siblings, 189 unaffected siblings and 389 control subjects. Descriptive results show that groups are of a similar age (affected subjects M=49.3, SD=13.1; unaffected siblings M=50.7, SD=13.6; controls M=50.6, SD=14.5; F=1.27, p=.280). The unaffected sibling group contained
significantly more males than the affected subjects and control group (affected subjects 30% male; unaffected siblings 55% male; controls 42% male; χ2= 36.4, p<.001). Education level was highest among control subjects and differed significantly from the other groups (affected subjects 2.42; unaffected siblings 2.37; controls 2.56; χ2= 24.4,
p<.001). The large majority of each group has completed intermediate and/or high
education. Of affected subjects, most suffer from comorbid depression and anxiety diagnosis (60%), remarkably fewer were diagnosed with either depression (24%) or anxiety (16%) only. Descriptive statistics are shown in table 1.
Reports of childhood maltreatment, depression/anxiety symptoms and personality characteristics were compared between groups (see table 1). Results reveal that affected subjects report significantly higher rates of childhood maltreatment on the CTQ (affected subjects M=38.6, SD=12.7; unaffected siblings M=33.8, SD=8.32; controls M=32.7, SD=9.42; F=32.7, p<.001). As for anxiety symptoms, affected subjects report significantly higher scores than the other groups (affected subjects
M=8.81, SD=8.21; unaffected siblings M=2.95, SD=3.60; controls M=2.16, SD=3.26; F=145, p<.001). Depression symptom scores are highest in the affected subjects group
and are also significantly increased in the unaffected siblings group compared to the control group (affected subjects M=16.8, SD=10.8; unaffected siblings M=8.24,
SD=6.51; controls M=5.73, SD=5.24; F=196, p<.001). Neuroticism scores too were
significantly higher among affected subjects compared to the other groups (affected subjects M=36.8, SD=8.93; unaffected siblings M=25.7, SD 7.8; controls M=25.3, SD 6.96; F=233, p<.001). Results reveal that groups differ significantly on extraversion and locus of control scores, with controls reporting the highest extraversion scores (affected subjects M=36.8, SD=7.71; unaffected siblings M=40.6, SD=8.07; controls
M=42.5, SD=7.06; F=60.0, p<.001) and the sibling group reporting the lowest locus of
control scores, indicating a more internal locus of control than the other groups (affected subjects M=15.6, SD=6.15; unaffected siblings M=8.94, SD=4.74; controls
M=21.3, SD 5.47; F=312, p<.001). Post-hoc Bonferroni testing yielded a significant
difference between the groups for CTQ, IDS, BAI, N, E and LoC (see table 1; values that differ significantly from the other two groups are shown in bold).
Table 1.
Sociodemographic characteristics, clinical status and rate of personality traits of three experimental groups. Affected subjects (n=447) Unaffected siblings (n=189) Controls (n=389) Between-group comparisons Mean (SD) Mean (SD) Mean (SD) Test
statistic Significance Age (mean; SD) 49.3 (13.1) 50.7 (13.6) 50.6 (14.51) F =1.27 p =.280 Gender (male %) 135 (30%) 104 (55%) 164 (42%) χ2= 36.4 p <.001*** Lifetime status Depression Anxiety Comorbid depression/anxiety 109 (24%) 71 (16%) 267 (60%) Education level (1 to 3) Not clear
Basic (lower vocational education) Intermediate (higher vocational education) High (college/university education) 2.42 7 (1.6%) 8 (1.8%) 221 (49%) 211 (47%) 2.37 7 (3.7%) 5 (2.6%) 89 (47%) 88 (47%) 2.56 0 10 (2.6%) 153 (39%) 226 (58%) χ2= 24.4 p <.001*** Childhood maltreatment (CTQ. Mean; SD) 38.6 (12.7) 33.8 (8.32) 32.7 (9.42) F = 32.7 p <.001*** Anxiety symptoms (BAI. Mean; SD) 8.81 (8.21) 2.95 (3.60) 2.16 (3.26) F =145 p <.001*** Depression symptoms (IDS. Mean; SD) 16.8 (10.8) 8.24 (6.51) 5.73 (5.24) F =196 p <.001*** Neuroticism (mean; SD) 36.3 (8.9) 25.7 (7.80) 25.3 (6.96) F =233 p <.001*** Extraversion (mean; SD) 36.8 (7.7) 40.6 (8.07) 42.5 (7.06) F =60.9 p <.001***
Mastery (LoC. Mean; SD) 15.6 (6.15) 8.94 (4.74) 21.3 (5.47) F =312 p <.001***
Significance: p<.05 = *; p<.01= **; p<.001=***. In bold is shown the value that is significantly different for that group compared to the other groups.
Correlation
Pearson’s correlations of childhood maltreatment, neuroticism, extraversion, locus of control, depression and anxiety symptom scores were assessed for the sample as a whole and separately for each group, i.e. affected subjects, unaffected siblings and controls, in order to confirm a relation between childhood maltreatment and depression/anxiety symptoms and to evaluate whether correlations differ between groups. Results confirm that childhood maltreatment (CTQ) is significantly related to
depression symptoms (IDS, r =.353, p <.01) and anxiety symptoms (BAI; r =.293, p <.01). Interestingly, associations between childhood maltreatment and depression and anxiety symptom severity are strongest for unaffected siblings. Correlations are shown in supplementary table 5.
Comparing the correlation coefficients between groups reveals that the affected subjects, unaffected siblings and controls differ significantly from each other for the majority of correlations (see table 2). Interestingly, unaffected siblings differ from controls with respect to correlations of CTQ with IDS, BAI, N, E and LoC, as well as IDS score with N and LoC, BAI score with E and LoC, and lastly N with LoC. To illustrate: the correlation between CTQ and BAI scores is significantly different from the control group, not only for the affected subjects (β=.226, p<.010), but also for the unaffected siblings (β=.260, p<.010). In addition, the correlation between CTQ and IDS scores is significantly stronger for the unaffected siblings (β =.363, p<.010) than for the controls (β =.229, p<.010), again emphasizing the importance of this group.
Table 2.
Comparison of correlations between groups.
Correlation Affected Subjects vs. Unaffected Siblings Z, p-value Affected Subjects vs. Controls Z, p-value Unaffected Siblings vs. Controls Z, p-value CTQ-IDS -2.26, .023 1.46, .144 3.39, <.001 CTQ-BAI -.922, .360 2.99, <.001 3.36, <.001 CTQ-N -3.93, <.001 -.207, .834 3.65, <.001 CTQ-E 22.4, <.001 22.1, <.001 -5.02, <.001 CTQ-LoC -9.35, <.001 -5.41, <.001 4.34, <.001 IDS-BAI 4.01, <.001 2.59, <.001 -1.84, .064 IDS-N -2.74, <.001 6.50, <.001 7.93, <.001 IDS-E -1.19, .234 .101, .920 1.24, .214 IDS-LoC -4.15, <.001 6.25, <.001 9.41, <.001 BAI-N 2.78, <.001 5.64, <.001 1.73, .083 BAI-E 2.76, <.001 -1.19, .234 -3.52, <.001 BAI-Loc 1.71, .087 13.4, <.001 11.2, <.001 N-E 3.50, <.001 2.29, .022 -1.77, .077 N-LoC -6.94, <.001 2.57, .010 8.86, <.001 E-LoC -.173, .865 -1.97, .049 -1.33, .184
Correlations that differ significantly are shown in bold. Legend: BAI, Beck Anxiety Index; CTQ, Childhood Trauma Questionnaire; E, extraversion; IDS, Inventory of Depressive Symptoms; LoC, locus of control; N, neuroticism.
Multilevel Modeling
Most assumptions of linear regression were met. A linear relationship was found between independent and dependent variables. Homoscedasticity was confirmed using a standardized residual plot. There is no multicollinearity between the independent variables, namely the Variance Inflation Factor (VIF) does not exceed 1.3. Markedly, BAI and IDS data is skewed to the right. Separate multilevel models were constructed for depression symptoms (IDS) and anxiety symptoms (BAI). The sample contained missing values at random, which were replaced by the variable mean. Control subjects were excluded from the multilevel analyses.
Model depression scores (IDS)
Table 3 shows the multilevel model for depression symptom severity (IDS). The baseline model suggests that most of the variance in IDS score resides between siblings. Namely, the null model results in an ICC of .200, indicating that siblings from the same family have 20% of the variance in IDS score in common. This means that there is an effect of being part of that family and confirms our choice of performing a multilevel model analysis in which individual variance is taken into account. Adding age (β = .020, p<.050), gender (β = 2.45, p<.010) and education level (β = -1.06, p>.050) in model 1 was needed to control for any effect these factors might have on IDS scores. The AIC was reduced compared to the null model. In model 2, CTQ outcome was significantly predictive for IDS score (β =.328, p<.001), suggesting that higher levels of reported childhood maltreatment are related to higher depression symptom scores. Adding CTQ to the model resulted in a lower AIC, making the model a better fit to explain variance in depression scores as compared to model 1. In model 3, upon adding the predictor variables neuroticism, extraversion and locus of control, childhood maltreatment still contributed significantly to the model, albeit that the relative contribution of CTQ on IDS score decreased when these factors were added (β =.167,
p<.001). As for neuroticism, a strong contribution to IDS score was found (β =.545,
p<.001), suggesting that increased neuroticism scores are associated with higher IDS
scores. Extraversion (β =-.183, p<.001) and locus of control (β = .221, p<.001), too, were found to contribute significantly to IDS scores. Adding neuroticism, extraversion and locus of control to the model results in a lower AIC, indicating an improved model fit. Upon adding the interaction variables of childhood maltreatment and the personality traits in model 4, the AIC is further reduced. Note that the interaction variables do not
significantly contribute to the model, indicating that there is no moderation effect of neuroticism, extraversion and locus of control on the relation between childhood maltreatment and depression.
We conclude that model 4 best explains the variance in IDS scores between siblings, as it has the lowest AIC, indicating the best model fit. In addition, model 4 has the lowest ICC, indicating that in this model the highest proportion of variance in IDS scores is explained by within-family differences, or in other words by the uniqueness of each child in a family. Namely, siblings from the same family have 12% of the variance in IDS scores in common whereas 88% of the variance can be explained due to within-family differences, that is differences between siblings. Therefore, the optimal model to explain variation in IDS scores among siblings includes the covariates age, gender, educational level and the fixed predictor variables childhood maltreatment, neuroticism, extraversion and locus of control, and their interaction variables (see table 3).
Table 3.
Multilevel model of fixed effects and variance in the prediction of depression score (IDS).
Null model β (SE) Model 1 β (SE) Model 2 β (SE) Model 3 β (SE) Model 4 β (SE) Fixed effects Intercept 14.6 (.468)*** 12.0 (2.76)*** 2.00 (2.85) .089 (3.27) -2.17 (3.68) Age Sex Level of education .020 (.034) 2.45 (.819)** -1.06 (.635) -.017 (.033) 1.79 (.778)* -.810 (.601) 0.018 (.027) 0.339 (.683) -.605 (.505) .023 (.027) .388 (.685) -.591 (.505) Childhood Maltreatment .328 (.038)*** .167 (.033)*** .155 (.046)*** Neuroticism Extraversion Locus of Control .545 (.042)*** -.183 (.044)*** .221 (.058)*** .542 (.046)*** -.139 (.051)*** -.174 (.059)** Interaction ChmN Interaction ChME Interaction ChMLoC .304 (.287) -.081 (.289) .286 (.331) Variances Within Variance 84.4*** 83.1*** 74.6*** 56.2*** 55.6*** Between variance 21.1*** 20.3*** 17.7*** 8.22* 7.60* AIC 4754 4748 4678 4465 4419 ICC .200 (20%) .197 (20%) .192 (19%) .128 (13%) .120 (12%)
AIC = Akaike Information Criterion; ICC = Intra Class Correlation; Significance p<.05 = *; p<.01= **; p<.001=***
Model anxiety scores (BAI)
Table 4 shows the multilevel model for anxiety symptom severity (BAI). The baseline model suggests that most of the variance in BAI score resides between siblings. Namely, the null model results in an ICC of .140, suggesting that siblings from the same family have 14% of the variance in BAI score in common. This means that there is an effect of being part of that family and again confirms our choice of performing a multilevel model analysis in which individual variance is taken into account. In model 1, age (β =-.012, p>.050), sex (β =2.38, p<.001) and education level (β =-.356, p>.050) were added to control for any effect these factors might have on BAI scores. The AIC was reduced compared to the null model. In model 2, CTQ outcome was significantly predictive for BAI score (β =.207, p<.001), suggesting that higher levels of reported childhood maltreatment are related to higher anxiety symptom scores. Note that the association between childhood maltreatment is slightly higher with IDS scores (β=.328,
p<.001) than with BAI scores (β=.207, p<.001). The AIC was further reduced, making this model a better fit to explain variance in anxiety scores as compared to model 1. In model 3, upon adding the predictor variables neuroticism, extraversion and locus of control, childhood maltreatment still contributed significantly to the model, yet its contribution lowered when these factors were added (β =.112, p<.001). As for neuroticism, a moderate contribution to BAI score was found (β =.356, p<.001), suggesting that increased neuroticism scores are associated with higher BAI scores. A trend was observable for the prediction of BAI score by extraversion (β =.017, p>.100) and locus of control (β = -.020, p>.100), however these findings were not significant indicating that E and LoC do not significantly contribute to the variance in BAI scores among siblings. Adding neuroticism, extraversion and locus of control to the model results in a lower AIC, indicating an improved model fit. In model 4, the interaction variable of neuroticism with CTQ significantly adds to the model (β =.948, p<.001), suggesting that neuroticism moderates the relation between childhood maltreatment and anxiety symptoms and that it partly accounts for the variance in BAI scores between siblings. The AIC is slightly increased compared to the previous model.
Despite the increase in AIC, we conclude that model 4 is the least complex fit to the data. Model 4 shows an ICC of .120, indicating that siblings from the same family have 12% of the variance in BAI scores in common and that the largest proportion of variance is explained by within-family differences and can be ascribed to the
uniqueness of each child in a family. Thus, a model with covariates age, gender, education level and the fixed predictor variables childhood maltreatment, neuroticism and the interaction of neuroticism and ChM is best suited to explain the variation in BAI scores between siblings (see table 4).
Table 4.
Multilevel model of fixed effects and variance in the prediction of anxiety score (BAI).
Null model β (SE) Model 1 β (SE) Model 2 β (SE) Model 3 β (SE) Model 4 β (SE) Fixed effects Intercept 7.24 (.331)*** 4.81 (1.99) -2.10 (2.09) -10.4 (3.23)*** -7.59 (2.06)*** Age Sex Level of education -.012 (.025) 2.38 (.604)*** -.356 (.465) -.037 (.023) 1.95 (.579)*** .037 (.444) -.001 (.022) .784 (.554) -.085 (.404) .002 (.021) .896 (.531) -.137 (.398) Childhood Maltreatment .207(.028)*** .112 (.026)*** .066 (.029)* Neuroticism Extraversion Locus of Control .356 (.038)*** .017 (.046) -.020 (.048) .321 (.029)*** Interaction ChmN .948 (.268)*** Variances Within Variance 48.6*** 47.5*** 42.3*** 36.0*** 34.6*** Between variance 7.95** 7.60** 7.72** 4.11* 4.64* AIC 4368 4358 4161 4011 4027 ICC .141 (14%) .138 (14%) .157 (16%) .102 (10%) .120 (12%)
AIC = Akaike Information Criterion; ICC = Intra Class Correlation; Significance p<.05 = *; p<.01= **; p<.001=***
Discussion
Childhood maltreatment has been associated with increased risk of developing depression and anxiety disorders in later life (Norman et al., 2012). Individuals vary in the extent to which they experience psychiatric symptoms after childhood maltreatment (Amstadter et al., 2014). The current study aimed to investigate the role of personality traits neuroticism, extraversion and locus of control in the relationship between childhood maltreatment and adult psychopathology in patients with depression and/or anxiety and their unaffected siblings, by taking into account a shared family background.
The first aim of this study was to examine whether individuals with a lifetime depression and/or anxiety disorder, their unaffected siblings and controls differ from each other with respect to neuroticism, extraversion and locus of control. Results reveal that groups are indeed distinct in terms of personality traits. As was hypothesized, neuroticism scores are highest and extraversion scores are lowest among affected subjects, suggesting that there might be a correlation between increased neuroticism and low extraversion and adult psychopathology (see table 1). In addition, results indicate that control subjects report highest extraversion and mastery scores, suggesting that high extraversion and internal locus of control is to a lesser extent associated with adult psychopathology. These findings are in line with previous work showing that high neuroticism and external locus of control is associated with chronicity of various affective disorders, whereas high extraversion is related to a more favorable prognosis (A. Farmer et al., 2002; Hope et al., 2018; P. Spinhoven et al., 2011; Philip Spinhoven, Roelofs, et al., 2011; Struijs et al., 2018; Sullivan et al., 2017). The distinction between groups is further confirmed by the comparison of correlation analyses, which reveal that the majority of assessed relations differs significantly for the affected targets compared to the unaffected siblings and controls (see table 2).
Interestingly, not only the affected subjects are distinct from the other groups, but also the unaffected siblings report higher depression scores and lower extraversion scores when compared to controls and show lowest locus of control scores. In addition, correlation analyses show that associations between childhood maltreatment and psychopathology are strongest for unaffected siblings and comparison of correlations reveals that unaffected siblings show distinct correlations between CTQ with
IDS/BAI/N/E/LoC; IDS with N/LoC; BAI with LoC and N with LoC when compared to the control group. These findings imply that unaffected siblings experience more severe depression symptoms than the control group and in line with what would be expected based on previous studies, report to be less extravert and have a more external locus of control (Philip Spinhoven, Elzinga, et al., 2011; Struijs et al., 2018). Heightened depression scores among unaffected siblings suggests that having a depressed or anxious sibling may influence the other (unaffected) sibling and predispose him/her to depression (Sander & McCarty, 2005; van Dam, Korver-Nieberg, Velthorst, Meijer, & de Haan, 2014). Possibly, the extent to which siblings are affected by each other’s misery is partly based on the closeness of their relationship, and those who are less close to each other are also less affected by the sibling’s misery (Bowes, Lereya, Lewis, Joinson, & Wolke, 2014; McHale, Updegraff, & Whiteman, 2012). The current findings emphasize the importance of studying childhood maltreatment, depression and anxiety in the context of the family.
Our second aim was to assess whether neuroticism, extraversion and locus of control moderate the relation between childhood maltreatment and adult depression and anxiety symptom severity, taking into account the shared family environment of lifetime affected individuals and their unaffected siblings by using a sibling design. As hypothesized and in congruence with previous work, results confirm a moderate association between childhood maltreatment and depression & anxiety symptom severity (Collishaw et al., 2007; Nanni et al., 2012).
From the multilevel model it can be deduced that neuroticism, extraversion and locus of control play a prominent role in explaining the variance in reported depression symptom severity between children from the same family (see table 3). The decrease in ICC upon adding the personality traits to the model suggests that variance within families increases, implying that children within the same family show variation in their personality scores. Interestingly, we found no moderation of neuroticism, extraversion or locus of control in the relationship between childhood maltreatment and depression symptom severity. This is contradictory to previous studies, which found a moderating effect of maladaptive personality traits on the association between childhood maltreatment severity and psychological distress (Philip Spinhoven, Elzinga, Van Hemert, De Rooij, & Penninx, 2016). A possible explanation for the lack of significant moderation is that the sample used in the current study is shown to be more depressed
than samples used in previous NESDA studies. Namely, healthy controls were excluded from the multilevel analysis, so that participants were either lifetime affected or unaffected siblings of lifetime affected subjects who report increased depression symptoms. Based on previous studies and as confirmed in table 1, higher depression scores in these groups are related to higher neuroticism, lower extraversion and a more external locus of control. In that case, the sample may be more neurotic, introvert and have an external locus of control compared to previously studied populations. As a consequence, the constructs childhood maltreatment, neuroticism, extraversion and locus of control may be such a strong predictor that their main effect explains the maximum of variance in depression scores and leaves no additional variance to be explained by a possible moderation effect.
In addition, it should be taken into account that both neuroticism and depression are independent constructs that show large resemblance to each other, as reflected in their strong correlation. Personality and psychopathology can relate to each other in multiple ways, namely; they can influence the presentation of one another, share a common etiology and have a causal role in the development of the other (Widiger, 2011). Possibly, in the case of neuroticism and depression, it is difficult to differentiate between the constructs and appoint a moderator in their complex relationship. Mood may influence personality scores and vice versa and it may not be possible to predict one based on the other (Jylha, Melartin, Rytsala, & Isometsä, 2009; Ormel, Rosmalen, & Farmer, 2004).
Lastly, it should be noted that the sibling design of the current study takes into account variance within families in addition to variance between families. Based on the reduction of the ICC in model 3, we find that personality scores vary among siblings of the same family (see table 3). However, it is likely that personality traits neuroticism, extraversion and locus of control are to some extent a familial aspect and show some resemblance for members of the same family. Despite individual differences among the siblings, the general level of neuroticism may be increased for the whole family compared to other families, making it difficult to pinpoint a moderating relationship. Future studies should assess the within-family correlations of these personality traits.
Neuroticism, extraversion and locus of control explain a large part of the variance in anxiety scores between siblings (see table 4). Note that the ICC is greatly reduced upon adding these personality traits to the model, suggesting that differences in anxiety
scores are partly due to differences in personality traits among siblings from the same family. Neuroticism was found to be an important predictor for anxiety symptom severity. This finding is in line with previous studies stating an association between neuroticism and anxiety symptoms (P. & E., 2006). We find no significant effect of extraversion and locus of control on anxiety symptom scores, which concurs with conclusions from previous studies showing an adverse effect of extraversion and external locus of control on anxiety symptoms (Helvik et al., 2016; Hope et al., 2018; Struijs et al., 2018; Sullivan et al., 2017; Yu & Fan, 2016). Possibly, the lack of correlation between anxiety symptom severity and locus of control is due to the duality of this trait, with each individual being somewhere on the scale between internal and external locus of control. Previous research has linked external locus of control to psychopathology (Benassi, Sweeney, & Dufour, 1988; Helvik et al., 2016). Internal locus of control may, however, also be influential in the development of mood disorders such as depression and/or anxiety. It has been suggested that maximum happiness is achieved in those with a balanced locus of control expectancy (April, Dharani, & Peters, 2012). Internalizing the responsibility of a problem and blaming oneself for e.g. maltreatment or thinking that one called it upon themselves may result in feelings of stress, guilt, insecurity and loneliness and may lead to serious mood problems (April et al., 2012). It may be that there is a delicate balance between internal and external locus of control that is highly context-dependent and can thus not be summarized in a clear trend.
Interestingly, neuroticism was found to moderate the relation between childhood maltreatment and anxiety symptoms. This finding implies that neuroticism scores can to some extent predict the risk of developing anxiety disorder after childhood maltreatment, and that being more neurotic in addition to having experienced childhood maltreatment makes one more vulnerable to the development of anxiety symptoms than someone who is less neurotic. The reduced ICC in model 3 implies that siblings vary with respect to neuroticism scores, implying that those with increased neuroticism are more vulnerable to developing anxiety symptoms than other family members.
In addition, the finding suggests that neuroticism is more strongly related to the development of anxiety as compared to depression after childhood maltreatment. This association may partly be explained by content overlap of measures of neuroticism and anxiety (Uliaszek et al., 2009). Namely, the construct of neuroticism consists of various facets, some of which are linked more strongly to depression, whereas others are related
more to anxiety (Uliaszek et al., 2009). Possibly, the difference in moderation effect of neuroticism on depression versus anxiety in the current study is due to disorder-specific neuroticism facets that show more similarity to BAI items than to IDS items and as such inflate the moderation effect of neuroticism on anxiety (Claridge & Davis, 2001).
In interpreting our findings, several methodological limitations should be taken into consideration. As was previously mentioned, the sample in the current study is more depressed than the general population, posing a problem for generalizability. In addition, there was a preponderance of men among the unaffected siblings. It may be that the relationship between childhood maltreatment and psychopathology is more strongly moderated by personality traits in females than in males. Namely, it was found that females generally report higher neuroticism and extraversion than males (Weisberg, De Young, & Hirsh, 2011). Gender differences in personality factors may even play a role in the gender difference in depression, namely neuroticism has been shown to moderate the relation between female gender and adult depression (Goodwin & Gotlib, 2004). As a result, the moderating effect of personality traits on the relation between childhood maltreatment and adult psychopathology may have been slightly subdued by the preponderance of males among unaffected siblings. Furthermore, data was not normally distributed and future studies should aim to use non-parametric tests. Lastly, it has been suggested that the use of retrospective self-report questionnaires to measure childhood maltreatment introduces the problem of recall bias, however plentiful studies have rejected this idea and state that the little bias is not sufficient to invalidate retrospective self-reports (Brewin, Andrews, & Gotlib, 1993; Fergusson, Horwood, & Boden, 2011; Hardt & Rutter, 2004; Newbury et al., 2018). An important strength of the current study is the use of a sibling design, which allowed for the assessment of the role of various personality traits on the relation between childhood maltreatment and adult depression and/or anxiety symptom severity among siblings, taking into account their shared family background.
In conclusion, this study shows that lifetime affected subjects, unaffected siblings and healthy controls are different in terms of neuroticism, extraversion and locus of control scores. We emphasize the importance of studying the unaffected sibling population, as they report more severe depression symptoms than do healthy controls and in line with what would be expected, report to be less extravert and have a more external locus of
control. Furthermore, we found that neuroticism moderates the relationship between childhood maltreatment and anxiety, putting those with higher neuroticism at increased risk of developing anxiety symptoms after childhood maltreatment. Therefore, when a history of childhood maltreatment is present in a family, it is important that clinicians look out for dysfunctional personality styles such as high neuroticism. Despite the general belief that personality traits are not modifiable, knowledge of their predicting role in psychopathology after childhood maltreatment may be helpful in intervention and therapy. To illustrate, it has been shown that mindfulness practice can moderate the relation between neuroticism and depression (Barnhofer, Duggan, & Griffith, 2011; Feltman, Robinson, & Ode, 2009). Interventions in people with depressive and anxiety disorders should aim to create awareness for traits that could be a deal-changer in those at risk after childhood maltreatment. Therapies should focus on psycho-education, mindfulness and meditation, implementing coping strategies and providing practical tips on how to handle certain characteristics in an optimal way in everyday life.
Appendix
Table 5.
Pearson’s R correlations between childhood maltreatment (CTQ), depression - (IDS) and anxiety symptoms (BAI), and personality traits neuroticism (N), extraversion (E) and locus of control.(LoC).
IDS (β, p) BAI (β, p) N (β, p) E (β, p) LOC (β, p)
CTQ IDS BAI N E .353** .293** .762** .334** .582** .520** -.176** -.363** -.260** -.356** -.041 -.102** -.039 p=.211 .026 p=.404 .152** Affected Subjects (447) CTQ .273 ** .226** .224** -.950* .028 p =.552 IDS .716** .450** -.243** -.240 p =.612 BAI .417** -.144** .380 p =.425 N -.274** .230** E .076 p =.111 Unaffected Siblings (189) CTQ .363** .260** .375** .043 p =.559 .272** IDS .552** .569** -.288** .403** BAI .300** -.066 p =.364 .309** N -.145* .509** E .080 p =.273 Controls (389) CTQ .229** .145** .230** -.202** -.126* IDS .632** .235** -.240** -.079 p =.118 BAI .233** -.174** .005 p =.922 N -.206** -.159** E .116* Significance: p<.05 = *; p<.01= **; p<.001=***
References
Amstadter, A. B., Myers, J. M., & Kendler, K. S. (2014). Psychiatric resilience: longitudinal twin study. The British Journal of Psychiatry, 205(4), 275–280.
http://doi.org/10.1192/bjp.bp.113.130906
April, K. A., Dharani, B., & Peters, K. (2012). Impact of locus of control expectancy on level of well-being. Review of European Studies, 4(2), 124–137.
http://doi.org/10.5539/res.v4n2p124
Barnhofer, T., Duggan, D. S., & Griffith, J. W. (2011). Dispositional mindfulness moderates the relation between neuroticism and depressive symptoms. Personality and Individual Differences, 51(8), 958–962. http://doi.org/10.1016/j.paid.2011.07.032
Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An Inventory for Measuring Clinical Anxiety: Psychometric Properties. Journal of Consulting and Clinical Psychology, 56(6), 893–897. http://doi.org/10.1037/0022-006X.56.6.893
Benassi, V. A., Sweeney, P. D., & Dufour, C. L. (1988). Is There a Relation Between Locus of Control Orientation and Depression? Journal of Abnormal Psychology, 97(3), 357– 367. http://doi.org/10.1037/0021-843X.97.3.357
Bernstein, D. P., Ahluvalia, T., Pogge, D., & Handelsman, L. (1997). Validity of the
childhood trauma questionnaire in an adolescent psychiatric population. Journal of the American Academy of Child and Adolescent Psychiatry, 36(3), 340–348.
http://doi.org/10.1097/00004583-199703000-00012
Bonanno, G. A. (2004). Loss, Trauma, and Human Resilience: Have We Underestimated the Human Capacity to Thrive after Extremely Aversive Events? American Psychologist. http://doi.org/10.1037/0003-066X.59.1.20
Bowes, L., Lereya, S. T., Lewis, G., Joinson, C., & Wolke, D. (2014). Sibling Bullying and Risk of Depression, Anxiety, and Self-Harm: A Prospective Cohort Study.
PEDIATRICS, 134(4), e1032–e1039. http://doi.org/10.1542/peds.2014-0832
Brewin, C. R., Andrews, B., & Gotlib, I. H. (1993). Psychopathology and Early Experience: A Reappraisal of Retrospective Reports. Psychological Bulletin, 113(1), 82–98. http://doi.org/10.1037/0033-2909.113.1.82
Brown, J., Cohen, P., Johnson, J. G., & Smailes, E. M. (1999). Childhood abuse and neglect: Specificity of effects on adolescent and young adult depression and suicidality. Journal of the American Academy of Child and Adolescent Psychiatry, 38(12), 1490–1496. http://doi.org/10.1097/00004583-199912000-00009
Cheong, E. Von, Sinnott, C., Dahly, D., & Kearney, P. M. (2017). Adverse childhood experiences (ACEs) and later-life depression: Perceived social support as a potential protective factor. BMJ Open, 7(9). http://doi.org/10.1136/bmjopen-2016-013228 Claridge, G., & Davis, C. (2001). What’s the use of neuroticism? Personality and Individual
Differences, 31(3), 383–400. http://doi.org/10.1016/S0191-8869(00)00144-6 Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., & Maughan, B. (2007).
Resilience to adult psychopathology following childhood maltreatment: Evidence from a community sample. Child Abuse and Neglect, 31(3), 211–229.
http://doi.org/10.1016/j.chiabu.2007.02.004
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI). Odessa, Florida.
http://doi.org/10.1016/j.msea.2007.11.110
Dağ, İ., & Şen, G. (2018). The mediating role of perceived social support in the relationships between general causality orientations and locus of control with psychopathological symptoms. Europe’s Journal of Psychology, 14(3), 531–553.
http://doi.org/10.5964/ejop.v14i3.1563
Dick, D. M., Johnson, J. K., Viken, R. J., & Rose, R. J. (2000). Testing between-family associations in within-family comparisons. Psychological Science, 11(5), 409–413. http://doi.org/10.1111/1467-9280.00279
difficulties: A comparison of depressed subjects, healthy controls and their siblings. Psychological Medicine, 33(7), 1169–1175. http://doi.org/10.1017/S0033291703008419 Farmer, A., Redman, K., Harris, T., Mahmood, A., Sadler, S., Pickering, A., & McGuffin, P.
(2002). Neuroticism, extraversion, life events and depression: The Cardiff Depression Study. British Journal of Psychiatry, 181(AUG.), 118–122.
http://doi.org/10.1192/bjp.178.6.549
Feltman, R., Robinson, M. D., & Ode, S. (2009). Mindfulness as a moderator of neuroticism-outcome relations: A self-regulation perspective. Journal of Research in Personality, 43(6), 953–961. http://doi.org/10.1016/j.jrp.2009.08.009
Fergusson, D. M., Horwood, L. J., & Boden, J. M. (2011). Structural equation modeling of repeated retrospective reports of childhood maltreatment. International Journal of Methods in Psychiatric Research, 20(2), 93–104. http://doi.org/10.1002/mpr.337
Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burden and consequences of child maltreatment in high-income countries. The Lancet.
http://doi.org/10.1016/S0140-6736(08)61706-7
Goodwin, R. D., & Gotlib, I. H. (2004). Gender differences in depression: The role of personality factors. Psychiatry Research, 126(2), 135–142.
http://doi.org/10.1016/j.psychres.2003.12.024
Hardt, J., & Rutter, M. (2004). Validity of adult retrospective reports of adverse childhood experiences: Review of the evidence. Journal of Child Psychology and Psychiatry and Allied Disciplines. http://doi.org/10.1111/j.1469-7610.2004.00218.x
Helvik, A. S., Bjørkløf, G. H., Corazzini, K., Selbæk, G., Laks, J., Østbye, T., & Engedal, K. (2016). Are coping strategies and locus of control orientation associated with health-related quality of life in older adults with and without depression? Archives of
Gerontology and Geriatrics, 64, 130–137. http://doi.org/10.1016/j.archger.2016.01.014 Hope, N. H., Wakefield, M. A., Northey, L., & Chapman, A. L. (2018). The association
between locus of control, emotion regulation and borderline personality disorder features. Personality and Mental Health, 12(3), 241–251.
http://doi.org/10.1002/pmh.1419
Hovens, J. G. F. M., Giltay, E. J., Spinhoven, P., Van Hemert, A. M., & Penninx, B. W. J. H. (2015). Impact of childhood life events and childhood trauma on the onset and
recurrence of depressive and anxiety disorders. Journal of Clinical Psychiatry, 76(7), 931–938. http://doi.org/10.4088/JCP.14m09135
IBM Corp. Released 2015. (2014). IBM SPSS Statistics for Mac, Version 22.0. 2014. http://doi.org/10.1080/02331889108802322
John Rush, A., Giles, D. E., Schlesser, M. A., Fulton, C. L., Weissenburger, J., & Burns, C. (1986). The inventory for depressive symptomatology (IDS): Preliminary findings. Psychiatry Research, 18(1), 65–87. http://doi.org/10.1016/0165-1781(86)90060-0 Jylha, P., Melartin, T., Rytsala, H., & Isometsä, E. (2009). Neuroticism, introversion, and
major depressive disorder - Traits, states, or scars? Depression and Anxiety, 26(4), 325– 334. http://doi.org/10.1002/da.20385
Kendler, K. S., Eaves, L. J., Loken, E. K., Pedersen, N. L., Middeldorp, C. M., Reynolds, C., … Gardner, C. O. (2011). The impact of environmental experiences on symptoms of anxiety and depression across the life span. Psychological Science, 22(10), 1343–1352. http://doi.org/10.1177/0956797611417255
Kessler, R. C., McLaughlin, K. A., Green, J. G., Gruber, M. J., Sampson, N. A., Zaslavsky, A. M., … Williams, D. R. (2010). Childhood adversities and adult psychopathology in the WHO world mental health surveys. British Journal of Psychiatry, 197(5), 378–385. http://doi.org/10.1192/bjp.bp.110.080499
L., S., & T.S.S., R. (2011). Parental divorce and the stress process: Does divorce increase the risk for childhood mental disorders? Indian Journal of Psychiatry. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed10&NEWS=N& AN=70733006
Laporte, L., Paris, J., Guttman, H., Russell, J., & Correa, J. A. (2012). Using a Sibling Design to Compare Childhood Adversities in Female Patients With BPD and Their Sisters.
Child Maltreatment, 17(4), 318–329. http://doi.org/10.1177/1077559512461173 Magnus, K., Diener, E., Fujita, F., & Pavot, W. (1993). Extraversion and Neuroticism as
Predictors of Objective Life Events: A Longitudinal Analysis. Journal of Personality and Social Psychology, 65(5), 1046–1053. http://doi.org/10.1037/0022-3514.65.5.1046 McHale, S. M., Updegraff, K. A., & Whiteman, S. D. (2012). Sibling Relationships and
Influences in Childhood and Adolescence. Journal of Marriage and Family, 74(5), 913– 930. http://doi.org/10.1111/j.1741-3737.2012.01011.x
Nanni, V., Uher, R., & Danese, A. (2012). Childhood maltreatment predicts unfavorable course of illness and treatment outcome in depression: A meta-analysis. American Journal of Psychiatry, 169(2), 141–151. http://doi.org/10.1176/appi.ajp.2011.11020335 Newbury, J. B., Arseneault, L., Moffitt, T. E., Caspi, A., Danese, A., Baldwin, J. R., & Fisher,
H. L. (2018). Measuring childhood maltreatment to predict early-adult
psychopathology: Comparison of prospective informant-reports and retrospective self-reports. Journal of Psychiatric Research, 96, 57–64.
http://doi.org/10.1016/j.jpsychires.2017.09.020
Norman, R. E., Byambaa, M., De, R., Butchart, A., Scott, J., & Vos, T. (2012). The Long-Term Health Consequences of Child Physical Abuse, Emotional Abuse, and Neglect: A Systematic Review and Meta-Analysis. PLoS Medicine.
http://doi.org/10.1371/journal.pmed.1001349
Ormel, J., Rosmalen, J., & Farmer, A. (2004). Neuroticism: A non-informative marker of vulnerability to psychopathology. Social Psychiatry and Psychiatric Epidemiology, 39(11), 906–912. http://doi.org/10.1007/s00127-004-0873-y
P., J., & E., I. (2006). The relationship of neuroticism and extraversion to symptoms of anxiety and depression in the general population. Depression and Anxiety. Retrieved from
http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed7&NEWS=N& AN=2006461006
Pearlin, L. I., & Schooler, C. (1978). The Structure of Coping. Journal of Health and Social Behavior, 19(1), 2. http://doi.org/10.1109/DigitalHeritage.2013.6744795
Penninx, B. W. J. H., Beekman, A. T. F., Smit, J. H., Zitman, F. G., Nolen, W. A., Spinhoven, P., … van Dyck, R. (2008). The Netherlands Study of Depression and Anxiety
(NESDA): Rationale, objectives and methods. International Journal of Methods in Psychiatric Research, 17(3), 121–140. http://doi.org/10.1002/mpr.256
Pike, A., & Plomin, R. (1996). Importance of nonshared environmental factors for childhood and adolescent psychopathology. Journal of the American Academy of Child and Adolescent Psychiatry, 35(5), 560–570. http://doi.org/10.1097/00004583-199605000-00010
Plomin, R., Asbury, K., & Dunn, J. (2001). Why are children in the same family so different? Nonshared environment a decade later. Canadian Journal of Psychiatry.
http://doi.org/10.1177/070674370104600302
Powers, A., Ressler, K. J., & Bradley, R. G. (2009). The protective role of friendship on the effects of childhood abuse and depression. Depression and Anxiety, 26(1), 46–53. http://doi.org/10.1002/da.20534
Pudrovska, T., Schieman, S., Pearlin, L. I., & Nguyen, K. (2005). The sense of mastery as a mediator and moderator in the association between economic hardship and health in late life. Journal of Aging and Health, 17(5), 634–660.
http://doi.org/10.1177/0898264305279874
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs. http://doi.org/10.1037/h0092976
Sander, J. B., & McCarty, C. A. (2005). Youth depression in the family context: Familial risk factors and models of treatment. Clinical Child and Family Psychology Review, 8(3), 203–219. http://doi.org/10.1007/s10567-005-6666-3
Spinhoven, P., Elzinga, B. M., Hovens, J. G. F. M., Roelofs, K., van Oppen, P., Zitman, F. G., & Penninx, B. W. J. H. (2011). Positive and negative life events and personality traits in predicting course of depression and anxiety. Acta Psychiatrica Scandinavica, 124(6),
462–473. http://doi.org/10.1111/j.1600-0447.2011.01753.x
Spinhoven, P., Elzinga, B. M., Van Hemert, A. M., De Rooij, M., & Penninx, B. W. (2016). Childhood maltreatment, maladaptive personality types and level and course of psychological distress: A six-year longitudinal study. Journal of Affective Disorders, 191, 100–108. http://doi.org/10.1016/j.jad.2015.11.036
Spinhoven, P., Elzinga, B., Roelofs, K., Hovens, J. G. F. M., Van Oppen, P., Zitman, F. G., & Penninx, B. W. J. H. (2011). The effects of neuroticism, extraversion, and positive and negative life events on a one-year course of depressive symptoms in euthymic
previously depressed patients versus healthy controls. Journal of Nervous and Mental Disease, 199(9), 684–689. http://doi.org/10.1097/NMD.0b013e318229d21f
Spinhoven, P., Roelofs, K., Hovens, J. G., Elzinga, B. M., van Oppen, P., Zitman, F. G., & Penninx, B. W. (2011). Personality, life events and the course of anxiety and depression. European Journal of Personality, 25(6), 443–452. http://doi.org/10.1002/per.808 Struijs, S. Y., Lamers, F., Spinhoven, P., van der Does, W., & Penninx, B. W. J. H. (2018).
The predictive specificity of psychological vulnerability markers for the course of affective disorders. Journal of Psychiatric Research, 103, 10–17.
http://doi.org/10.1016/j.jpsychires.2018.04.017
Sullivan, S. A., Thompson, A., Kounali, D., Lewis, G., & Zammit, S. (2017). The longitudinal association between external locus of control, social cognition and adolescent psychopathology. Social Psychiatry and Psychiatric Epidemiology, 52(6), 643–655. http://doi.org/10.1007/s00127-017-1359-z
Uliaszek, A. A., Hauner, K. K. Y., Zinbarg, R. E., Craske, M. G., Mineka, S., Griffith, J. W., & Rose, R. D. (2009). An examination of content overlap and disorder-specific
predictions in the associations of neuroticism with anxiety and depression. Journal of Research in Personality, 43(5), 785–794. http://doi.org/10.1016/j.jrp.2009.05.009 van Dam, D. S., Korver-Nieberg, N., Velthorst, E., Meijer, C. J., & de Haan, L. (2014).
Childhood maltreatment, adult attachment and psychotic symptomatology: a study in patients, siblings and controls. Social Psychiatry and Psychiatric Epidemiology, 49(11), 1759–1767. http://doi.org/10.1007/s00127-014-0894-0
Weisberg, Y. J., De Young, C. G., & Hirsh, J. B. (2011). Gender differences in personality across the ten aspects of the Big Five. Frontiers in Psychology, 2(AUG).
http://doi.org/10.3389/fpsyg.2011.00178
Widiger, T. A. (2011). Personality and psychopathology. World Psychiatry. http://doi.org/10.1002/j.2051-5545.2011.tb00024.x
Wittchen, H. U. (1994). Reliability and validity studies of the WHO-Composite International Diagnostic Interview (CIDI): A critical review. Journal of Psychiatric Research, 28(1), 57–84. http://doi.org/10.1016/0022-3956(94)90036-1
Yu, X., & Fan, G. (2016). Direct and indirect relationship between locus of control and depression. Journal of Health Psychology, 21(7), 1293–1298.
http://doi.org/10.1177/1359105314551624
