Psychiatry Research 215 (2014) 75–81

Contents lists available at ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Evaluative beliefs as mediators of the relationship between parental bonding and symptoms of paranoia and depression Carmen Valiente a,n, Nuria Romero a, Gonzalo Hervas a, Regina Espinosa a,b a b

Department of Clinical Psychology, School of Psychology, Complutense University of Madrid, Campus de Somosaguas, 28223 Madrid, Spain Department of Psychology, School of Health Sciences, Camilo Jose Cela University, Spain

art ic l e i nf o

a b s t r a c t

Article history: Received 6 December 2012 Received in revised form 2 October 2013 Accepted 16 October 2013 Available online 23 October 2013

This study was aimed to explore the distinct pathways that lead to depression and paranoia. We first examined the association of dysfunctional parenting experiences and negative self-evaluations in depression and paranoia. Furthermore, we also examined whether different self-evaluative beliefs could mediate the relationships between dysfunctional parenting experiences (i.e. parental overprotection or lack of care) and the development of depression and paranoia. A sample composed of 55 paranoid patients, 38 depressed patients and 44 healthy controls completed the Parental Bonding Instrument (PBI), the Evaluative Beliefs Scale (EBS) and some clinical scales. Our analyses revealed that lack of parental care and negative self–self evaluations were associated with depression symptoms. Analyses also revealed that parental overprotection and negative other–self evaluations were associated with paranoid symptoms. Furthermore, negative self–self and other–self evaluations fully mediated the relationship of parental overprotection and paranoia, whereas negative self–self evaluations partially mediated the relationship between lack of parental care and depression. These findings suggest that distinct patterns of parental practices may contribute to the development of different dysfunctional schemas which in turn may lead to either depression or paranoia. & 2013 Elsevier Ireland Ltd. All rights reserved.

Keywords: Parental practices Self-schemas Paranoia Depression Mediational effects

1. Introduction Psychotic spectrum disorders, especially for paranoid symptoms, have shown strong connections to depression (e.g., Zigler and Glick, 1988). For example, some studies have found elevated scores of paranoia in depressed participants (e.g., Bentall et al., 2008) which again suggests a phenomenological overlap. Furthermore, depression has been identified repeatedly in factor analytical studies of psychotic spectrum disorders (e.g., Emsley, et al., 2003). Both depression and paranoia seem to share psychological mechanisms such as negative self-esteem and expectation of negative events (e.g., Bentall et al., 2008) that could be rooted in parental practices. The analysis of the developmental pathways of these forms of psychopathology can highlight common and distinct factors and provide us with a deeper understanding of the processes involved. With this aim, the present study examined parental bonding and negative evaluative beliefs in two clinical samples with paranoid and depressive symptomatology. Theoretical approaches have linked early parental rearing behaviors with psychopathology (e.g., Bowlby, 1977). In general, empirical

n Corresponding author. Tel.: þ 34 91 394 3135x636 708 210; fax: þ34 91 394 3189. E-mail address: [email protected] (C. Valiente).

0165-1781/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.psychres.2013.10.014

studies investigating parental practices confirm the relevance of early relationships in psychopathology (e.g., Mickelson et al., 1997). Recent models of psychosis have drawn attention to the significance of early interpersonal experiences and trauma in the development and maintenance of psychotic symptoms (Garety et al., 2001). It is not surprising to find, given the suspiciousness and high level of interpersonal difficulties associated with paranoia, that studies have shown a particular relationship between insecure attachment and paranoid beliefs (Pickering et al., 2008). Some studies have found that people with schizophrenia rated their parents as less caring and more overprotective (see review in Berry et al., 2007). More specifically, Rankin et al. (2005) found that both currently ill and remitted paranoid patients reported lower parental care or higher overprotection during childhood in comparison to healthy control groups. Additionally, a twin study found that the difference in paternal overprotection was the most important variable to discriminate between schizophrenic and non-schizophrenic twins (Onstad et al., 1994). Likewise, the development of depression has been related to early disruptive parental practices by different theoretical approaches (Beck, 1967; Blatt et al., 1979). Extensive research using different methods seem to indicate that parent–child interaction is essential to understand the origins of depression (see review in by Blatt and Homann, 1992). Early disruptive parental practices (lack of care, overprotection or both) seem to be

76

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81

associated to depressive symptomatology during adulthood (Enns et al., 2000; Avagianou and Zafiropoulou, 2008) and are able to predict the presence of depression (Eberhart et al., 2006; Grotmol et al., 2010). Some studies have found that the bonding dimension of maternal care is a particularly strong predictor of severe depression (Grotmol et al., 2010). Psychopathology may be the by-product of negative schema of self and others as a result of early disruptive parental practices. Thus, self-evaluations may be a relevant link between parental practices and the development of psychopathology (Parker, 1993). Many theoretical approaches pose that early family factors, the subsequent sense of internalized self and the development of psychopathology are interconnected (e.g., Beck, 1967; Bowlby, 1980). Moreover, interactions with significant others seem to be the base of self–other representations (Bartholomew, 1990). These self and self–other representations are internalized and give rise to a set of beliefs about the self and others that are central determinants of psychological and behavioral functioning. Research indicates that higher levels of parental acceptance and care are associated with more positive self-evaluations, whereas higher levels of parental rejection, restraint, and inconsistent affect are associated with more negative self-evaluations (Liu, 2003). Persecutory delusions and depression have been associated with negative evaluative self-beliefs. According to the cognitive model (Beck, 1967), depressed individuals exhibit dysfunctional negative self-related schemas. Numerous studies have shown that depression is associated with increased negative thinking, enhanced accessibility of negative information, and negative biased information processing (see review in Ingram et al., 1998). The relationship between negative self-schemas and paranoia is not so clear, but disturbances in self-perceptions are clearly present in paranoia (Bentall et al., 2008) and have been found to be both a consequence of the illness as well as a maintaining (Freeman et al., 1998) and predisposing component (Krabbendam et al., 2002). Moreover, self-evaluation discrepancies (Valiente et al., 2011) and instability (Thewissen et al., 2008) have been identified as typical features of people who experience paranoia. In addition, individuals suffering from paranoia have strong negative evaluations of others (Gracie et al., 2007). In non-clinical population, there is some support for the role of self-related schemas in mediating negative parental practices and psychopathology. Studies have found that dysfunctional parenting increases the risk of depression by negatively impacting self-esteem (e.g., Avagianou and Zafiropoulou, 2008; Restifo et al., 2009; Grotmol et al., 2010). Campos et al. (2010) established a differential meditational pattern for care and overprotection practices. They found that the relationship between care and depression was mediated by self-criticism, while the relationship between overprotection and depression was mediated by neediness. Furthermore, McGinn et al. (2005) reported in a clinical sample that the association between dysfunctional parenting styles and depression was mediated by dysfunctional cognitive styles, as measured by the Young's Schema Questionnaire (Young and Brown, 1990). In the area of paranoia, Gracie et al. (2007) found, in a student sample, that negative beliefs about the self and others mediated the relationship between trauma and paranoia. Despite this finding, there is a lack of empirical studies addressing this question in a clinical population. Additional research is necessary to improve our understanding of the associations between developmental pathways and different psychological disturbances. 1.1. Aim of the study This study investigated the pathways that lead to depression and paranoia. First, we examined the association of dysfunctional parenting experiences and negative self-evaluations in depression

and paranoia. It was hypothesized that depression will be associated with low parental care (Grotmol et al., 2010), and that paranoia will be associated with both low parental care and particularly overprotection (Onstad et al., 1994). Secondly, we examined whether negative evaluative beliefs mediated the relationships between early parental practices, depression and paranoia. Depression usually involves negative self-esteem and, accordingly, negative self–self evaluative beliefs could mediate the relationship between parental bonding and depressive symptoms. On the contrary, paranoia has not been consistently associated with low self-esteem (e.g., Valiente et al., 2011), but a negative representation of others may be more significant to paranoid ideation (e.g., Lincoln et al., 2010). Thus, we predicted that negative self–other and other–self evaluative beliefs will be mediators of the relationship between a negative parental style and paranoid symptoms.

2. Methods 2.1. Participants and procedure The inclusion criteria for all participants included a signed informed consent and ages within 18–65. All participants were from the same geographical area. Depressed participants were recruited from an outpatient mental health center. Paranoid participants were recruited from an acute psychiatric inpatient unit from the same hospital. The healthy controls were recruited via the ‘snowball’ technique in which psychology students invited their acquaintances to voluntarily participate in the study. Three groups of participants were formed. The persecutory beliefs group (PG) included 55 participants (28 males), who were treated in an inpatient psychiatric unit. All participants were currently suffering persecutory beliefs as assessed by the Present State Examination (PSE10, SCAN, Sections 18 and 19, WHO, 1992) and had a score of Z 4 (i.e. level of severity) on the suspiciousness item of the Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987). None of the patients had brain disorders. Patients were selected through hospital records and diagnoses were confirmed with a clinical structured interview (MINIPLUS, Sheehan and Lecrubier, 2002). According to the DSM-IV criteria (APA, 1994), patients met diagnostics for the following categories: paranoid schizophrenia (n¼ 28), delusional disorder (n¼ 8), schizophreniform disorder (n¼ 9), schizoaffective disorder (n ¼6), brief psychotic disorder (n¼ 2), residual schizophrenia (n¼ 1) and non-specific psychotic disorder (n¼ 1). All patients were receiving antipsychotic medication at the time of participation in the study. The mean age of the paranoid sample was 34.6 years (S.D. ¼ 11.3; Table 1). The mean age of illness onset for this group was 28.1 years (S.D. ¼7.8) whereas the average mean illness duration was 73.4 months (S.D. ¼ 105.2). The depression group (DG) included 38 participants (nine males), who met DSM-IV criteria for a current depressive disorder. Participants were primarily outpatients who had never experienced persecutory delusions, brain disorders, or any other schizophrenia spectrum diagnoses. Diagnoses were confirmed in a clinical structured interview (MINIPLUS, Sheehan and Lecrubier, 2002). According to DSM-IV criteria, patients met diagnostic criteria for the following categories: major depressive disorder (single episode) (n¼ 11), major depressive disorder (recurrent episode) (n¼ 22) and bipolar I disorder (n¼ 5). All patients but four were currently receiving anti-depressive medication at the time of participating in the study. The mean age of the depression group was 42.8 (S.D. ¼ 11.5; Table 1). The mean age of illness onset for this group was 34.8 years (S.D. ¼ 9.7) whereas the average mean illness duration was 92.6 months (S.D. ¼ 113.5). The control group (CG) included 44 participants (20 males), who were screened for the absence of any clinical syndrome by a trained research assistant using an ‘ad hoc’ structured interview based on the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I; First et al., 2002). The controls had never required psychological assistance for any mental disorder, or had any concurrent medical condition. These participants did not report any previous history of mental health problems and were matched with the paranoid and depression groups in terms of sex and age. The mean age of the control group was 37.4 years (S.D. ¼13.0; Table 1). 2.2. Measures All participants were evaluated with the following measures: Evaluative Beliefs Scale (EBS; Chadwick and Birchwood, 1995). This scale contains 18 items that measure global and stable negative evaluative beliefs. Participants are asked to indicate their agreement with each statement on a 5 point scale ranging from strongly agree to strongly disagree. This scale is composed of three subscales, with six items for each subscale. The range of possible scores for each scale is 0–18. The first subscale measures negative beliefs about the

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81 participant's self (e.g., ‘I think I am totally bad’). The second subscale measures negative beliefs that the participant has about other people's judgment about him or her (e.g., ‘Other people think I am totally bad’). The third subscale measures negative beliefs that the participant may have about others (e.g., ‘Other people are totally bad’). In this study, the reliability was high for all subscales (α¼ 0.83, α ¼ 0.88 and α ¼0.85, respectively). Parental Bonding Instrument (PBI; Parker et al., 1979). This instrument retrospectively evaluates perceived parental style until the participant reached 16 years. The measure has 25 items, 12 items related to care and 13 items related to overprotection. The participants were required to complete separate maternal/ paternal forms. The reliability of the version used in this study was high for the two subscales (α¼ 0.90 for parental overprotection and α ¼ 0.87 for parental care). A composite parental care variable was created by summing maternal and paternal care scores, and a composite parental overprotection variable was created by summing maternal and paternal overprotection scores. Beck Depression Inventory II (BDI-II; Beck et al., 1996). This is a 21-item selfreport questionnaire to assess the severity of cognitive, affective and somatic symptoms of depression. The reliability of the version used in this study was high (α¼ 0.94). The Persecution and Deservedness Scale (PaDS; Melo et al., 2009). This is a brief measure to assess the severity of paranoid thinking and the perceived deservedness of persecution. In this study we used 10 statements of the persecution subscale (e.g., ‘There are times when I worry that others might be plotting against me’). Each statement was rated on 5-point Likert scales, ranging from certainly false to certainly true. In our study the internal consistency for the persecution ideation subscale was α ¼0.91.

77

3. Results 3.1. Demographic characteristics Table 1 shows demographic characteristics in the study. Analyses showed significant differences among groups in sex (χ2 ¼ 7.25, p ¼0.027). DG had fewer male participants than PG and CG. Analyses also showed significant differences among groups in age (F(2,136) ¼ 5.39, p ¼0.004). Post-hoc tests showed that DG was older than PG. Furthermore, we found significant differences in educational level between the PG and DG groups (χ2 ¼ 13.04, p ¼0.001) and between the PG and CG groups (χ2 ¼ 21.64, po 0.001). PG participants had a lower education level compared to both DG and CG participants. DG and CG participants did not significantly differ in their educational level. Our results revealed significant differences in employment status between the PG and CG groups (χ2 ¼ 9.13, p¼ 0.02) and between the DG and CG groups (χ2 ¼ 12.26, p ¼0.007). There were no significant differences between PG and DG participants in employment status.

2.3. Data analyses

3.2. Correlations among assessed variables

Demographic variables were analyzed by Pearson chi-square tests for qualitative variables and by one-way ANOVAs for quantitative ones. To analyze group differences in perceived parental bonding and evaluations about self and other we conducted one-way ANCOVAs with sex, age, level of education, and employment status as covariates. Bonferroni adjusted p-values were used to control for Type I error. A series of multiple regression analyses were conducted to examine the relationships among independent variables (i.e., parental overprotection and parental care), the mediator variables (i.e., negative evaluations about self and others), and the dependent measures of paranoid and depressive symptoms. Mediational analyses were used following Baron and Kenny (1986) guidelines.

Correlations between the variables assessed in the study are showed in Table 2. Depression severity was significantly related to negative self–self evaluations (r ¼ 0.56, p o0.001), negative other– self evaluations (r ¼0.34, po 0.001) and parental care (r ¼  0.27, p¼ 0.001). Paranoia severity was significantly related to negative self–self evaluations (r¼ 0.41, po0.001), negative other–self evaluations (r¼0.58, po0.001), negative self–other evaluations (r¼ 0.44, po0.001) and parental overprotection (r¼0.24, po0.001).

Table 1 Differences in demographic characteristics among paranoid, depression and control groups. Characteristics

Persecutory beliefs group (55)

Depression group (38)

Control group (44)

χ2

F

p

Age, mean (S.D.)

34.6 (11.1)

42.8 (11.5)

37.4 (13.0)

–

5.39

0.006nn

Education n (%) Primary education Secondary school University education

29 (70.7) 13 (38.2) 13 (21)

7 (17.1) 10 (29.4) 21 (33.9)

5 (12.2) 11 (32.4) 28 (45.2)

26.1

–

0.001nnn

Employment, n (%) Never employed Unemployment 41 year Unemployment o 1 year Employed

13 (44.8) 12 (52.2) 9 (56.3) 21 (30.4)

2 (6.9) 8 (34.8) 5 (31.3) 23 (33.3)

14 (48.3) 3 (13) 2 (12.5) 25 (36.2)

16.6

–

0.01n

n

po 0.05. p o0.01. nnn p o 0.001. nn

Table 2 Pearson correlations among psychological variables. Variables

(1)

(2)

(3)

(4)

(5)

(6)

(1) (2) (3) (4) (5) (6) (7)

–  0.39nn 0.25nn 0.18n 0.08 0.08 0.24nn

–  0.21n  0.08  0.01  0.27nn  0.09

– 0.64nn 0.20n 0.56nn 0.41nn

– 0.35nn 0.34nn 0.58nn

– 0.161;┼ 0.44nn

– 0.32nn

Parental overprotection (PBI) Parental care (PBI) Negative self–self evaluations (EBS) Negative other–self evaluations (EBS) Negative self–other evaluations (EBS) Depression (BDI-II) Paranoia severity (PADS)

Note: PBI: Parental Bonding Instrument. EBS: Evaluative Beliefs Scale. BDI-II: Beck Depression Inventory II. PADS: The Persecution and Deservedness Scale. n

po 0.05. p o0.001. po 0.10.

nn

┼

78

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81

Table 3 Differences in parental bonding subscales, negative evaluations subscales, and symptomatology among paranoid, depressive and control groups. Groups of participants

Self–self (EBS) Other–self (EBS) Self–other (EBS) Care (PBI) Overprotection (PBI) Depression (BDI-II) Paranoia (PADS)

Paranoia group (n¼ 55)

Depression group (n¼ 38)

Control group (n¼ 44)

ANCOVA overall (F)

p-value

1.95 3.00 1.59 66.6 61.0 13.85 1.66

3.61 2.44 1.24 63.4 58.9 31.08 1.46

0.23 0.16 0.14 73.4 51.4 6.13 0.39

14.62 10.40 2.97 5.52 4.77 62.5 18.42

0.0001nnn 0.0001nnn 0.055 0.005nn 0.01n 0.0001nnn 0.0001nnn

(3.04) (3.81) (3.09) (11.3) (13.5) (12.20) (1.16)

(3.65) (3.16) (2.31) (13.6) (12.7) (9.43) (1.06)

(0.80) (0.65) (0.67) (11.9) (11.9) (5.57) (0.35)

EBS: Evaluative Beliefs Scale. PBI: Parental Bonding Instrument. BDI-II: Beck Depression Inventory-II. PADS: The Persecution and Deservedness Scale. n

po 0.05. p o0.001. nnn p o0.0001. nn

3.3. Statistical differences among paranoia, depression and control groups One-way ANCOVAs were conducted separately for each variable to analyze group differences (Table HYPERLINK \l "MEP_L_tbl3" \o "Table 3 Differences in Parental Bonding Subscales, Negative Evaluations Subscales, and Symptomatology among Paranoid, Depressive and Control groups Groups of participants Paranoia Group (n¼ 55) Depression Group (n¼ 38)Control Group (n¼ 44)ANCOVA overall (F)p -v"3). Sex, age, level of education, and employment status were used as covariates in all analyses. First, we explored statistical differences in parental bonding subscales among groups. Covariates did not show significant effects for any of the analyses. Our analyses showed significant differences in parental care (F(2,130)¼ 5.52; po0.01; η2 ¼0.07). Bonferroni post-hoc tests showed that CG participants had significantly higher scores on parental care than DG participants (po0.01). PG participants did not show significant differences in parental care compared to both CG and DG participants. Furthermore, analyses revealed significant differences in parental overprotection (F(2,130)¼4.77; p¼0.01; η2 ¼0.06). Post-hoc tests showed that CG participants had significantly lower scores on parental overprotection than both PG (p¼0.02) and DG participants (p¼ 0.03). DG and PG participants did not show significant differences. Next, we examined statistical differences in negative evaluative beliefs among groups. Analyses revealed a significant group effect for negative self–self evaluations (F(2,130)¼ 14.62; p o0.0001; η2 ¼ 0.18). Post-hoc tests showed that CG participants had significantly lower scores on negative self–self evaluative beliefs than both PG (p o0.01) and DG participants (p o0.001). DG participants had significantly higher scores on negative self–self evaluative beliefs than PG participants, po 0.05. Regarding negative other–self evaluations, analyses revealed a significant group effect (F(2,130) ¼10.40; p o0.001; η2 ¼0.13). Post-hoc tests showed that CG participants had significantly lower scores in negative other–self evaluations than both PG (p o0.001) and the DG participants (p o0.01). DG and PG participants did not show significant differences in negative other–self evaluations. There were no significant group differences for negative self–other evaluations (F(2,130) ¼ 2.97; p ¼0.055; η2 ¼ 0.04). Finally, we examined statistical differences in paranoid and depressive symptoms. Analyses revealed a significant group effect for depressive symptoms (F(2,130)¼62.5; po0.0001; η2 ¼0.49). Post-hoc tests showed that DG participants had significantly higher scores in depressive symptoms than both PG (po0.001) and the CG participants (po0.001). Also, PG participants showed higher scores in depressive symptoms than CG participants (po0.01). Regarding paranoid symptoms, analyses revealed a significant group effect (F(2,130) ¼17.60; p o0.001; η2 ¼ 0.21). Post-hoc tests showed that CG participants had significantly lower scores in

PARENTAL CARE

-0.22*

NEGATIVE SELF-SELF EVALUATIONS

0.51***

DEPRESSION

-0.15* (-0.27*** ) Fig. 1. The role of negative evaluative beliefs about self and others (EBS) as mediators of the relationship between parental bonding and depression. Note: Numbers in the figure are referred to β's (i.e., standardized Beta coefficients). Value in parentheses represents the relationship between parental care and depression severity prior to controlling for negative self–self evaluations. Multicollinearity of the predictors was also assessed by computing tolerance statistics (no problems were found). npo 0.05; nnp o0.01; nnnpo 0.001.

paranoid symptoms than both PG (p o0.001) and the DG participants (p o0.001). However, DG and PG participants did not show significant differences in paranoid symptoms. 3.4. Negative evaluations as predictors of depressive and paranoid symptoms Furthermore, a series of regression analyses were conducted with depressed and paranoid severity measures as dependent variables. The three subscales of negative evaluative beliefs (i.e. self–self, other–self and self–other) were entered as predictors in each regression model to examine the relative prediction of unique variance. For depression severity, the regression model accounted for 32% (adjusted R2) of the variance, where only negative self–self evaluation emerged as a significant predictor (β¼0.58; p o0.001), whereas negative other–self and self–other evaluations did not predict significant amounts of variance (β¼  0.05; p ¼0. 57, and β¼ 0.06; p ¼0.41, respectively). For paranoia severity, the regression model accounted for 40% (adjusted R2) of the variance, where other–self (β¼ 0.43, po 0.001) and self–other evaluations (β¼0.27, po 0.001) emerged as significant predictors, whereas self–self evaluation did not predict a significant amount of variance (β¼0.08; p¼ 0.38). 3.5. The mediational role of negative evaluative beliefs in depression severity Since parental overprotection did not predict depressive symptoms, we only tested the mediational model using parental care as the independent variable. As commented above, only negative self–self evaluation predicted unique variance of depression symptoms. Thus, we examined whether negative self–self evaluations would mediate the association between parental care and depression severity. First, regression analyses showed that lower care was predictive of higher negative self–self evaluations (β¼  0.22; p¼0.01). Second, analyses showed that

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81

PARENTAL OVERPROTECTION

NEGATIVE SELF-SELF EVALUATIONS

NEGATIVE OTHER-SELF EVALUATIONS

79

PARANOIA

Fig. 2. The role of negative evaluative beliefs about self and others (EBS) as mediators of the relationship between parental bonding and paranoia. Note: Numbers in the figure are referred to β's (i.e., standardized Beta coefficients). Value in parentheses represents the relationship between parental overprotection and paranoia severity prior to controlling for negative other–self evaluations. Multicollinearity of the predictors was also assessed by computing tolerance statistics (no problems were found). npo0.05; nnpo0.01; nnnpo0.001.

lower care was predictive of higher depression severity (β¼  0.27, po0.001). Third, analyses also showed that higher negative self–self evaluations were predictive of higher depression severity (β¼ 0.55, po0.001). A mediation analysis was conducted with the depression severity as dependent variable with care and negative self–self evaluations entered as predictor and mediator variables, respectively. Negative self–self evaluations significantly predicted depression severity (β¼  0.51, po0.001), whereas the relationship between care and depression severity decreased in strength with the presence of the mediator, but remained significant (β¼  0.15, p¼ 0.04) (Fig. 1). The indirect effect was  0.129. Testing the standard error of the indirect effect (Sobel, 1982) indicated a significant mediation effect (z¼  2.44, p¼ 0.007). Thus, negative self–self evaluations partially mediated the relation between care and depression severity.

3.6. The mediational role of negative evaluative beliefs in paranoia severity Since only overprotection was related with paranoia severity, we tested the mediational role of negative evaluative beliefs in paranoia severity using overprotection as the independent variable. In the regression analysis reported above, both negative self– other and other–self evaluations significantly predicted paranoia. However, since negative self–other evaluation did not show any significant association with parental overprotection it was excluded for the mediation analyses. The case of self–self evaluation was more complex. This variable did not predict a significant amount of variance in the regression analysis, which might suggest that self–self evaluation is not a relevant predictor of paranoia. However, the pattern of correlations among parental overproduction, self–self-evaluation, other–self evaluation and paranoia (i.e., all p's40.05), and specifically the significant correlation between parental overproduction and self–self evaluation, may also suggest a mediational model with two sequential mediators (i.e., self–self and other–self evaluations). In this model, parental overprotection would predict self–self evaluation which, in turn, would predict other–self evaluations, which finally would predict paranoia (Fig. 2). This model may explain why self–self evaluations did not predict paranoia when other–self and self– other were in the equation. That is, if other–self evaluations fully mediated the relationship between self–self evaluations and paranoia, when other–self and self–self were included in the regression analysis at the same time, only other–self evaluations would remain significant (i.e., fully mediation). In fact, when we examined this hypothesis and included only self–self and other– self evaluations at the same time, with paranoia as the dependent variable, self–self evaluations indeed became non-significant (β¼ 0.09, ns.). Moreover, from a conceptual point of view this model with two sequential mediators is coherent with defensive explanations of paranoia (e.g., Bentall et al., 2008) as we will argue in Section 4.1 1 Since parental care was not significantly related to other–self or self–other evaluations, this strategy (i.e., sequential mediators) was not useful when testing the mediational model with depressive symptoms as the dependent variable.

Therefore, taking into consideration theoretical and empirical arguments the sequential mediators model is the pattern of results that we will test below. To test the sequential mediators model, we followed Cohen and Cohen (1983‚ pp. 352–378) guidelines for path analysis. To begin, we tested the first part of the model (i.e., self–self evaluation as a mediator between overprotection and other–self evaluation). Regression analyses showed that overprotection significantly predicted self–self evaluation (β¼ 0.25, p ¼0.02) and other–self evaluations (β¼0.17, p¼ 0.05). They also showed that self–self evaluations significantly predicted other–self evaluations (β¼0.64, p o0.001). Next, using other–self as the dependent variable, we included both overprotection and self–self evaluations. Parental overprotection failed to significantly predict paranoia (β¼ 0.01, ns.), but self–self evaluations remained significant (β¼0.64, p o0.001). Thus, negative self–self evaluations fully mediated the relationship between overprotection and other–self evaluations. Then, we tested the second part of the model (i.e., negative other– self evaluations as a mediator between parental overprotection and paranoia severity). First regression analyses showed that overprotection significantly predicted negative other–self evaluations (β¼0.17, p¼0.05) and paranoia severity (β¼0.21, p¼0.02). They also showed that negative other–self evaluations significantly predicted paranoia severity (β¼0.56, po0.001). Finally, using paranoia severity as the dependent variable, we entered, at the same time, parental overprotection and other–self evaluations. Parental overprotection failed to significantly predict paranoia (β¼0.11, ns.), but other–self evaluations remained significant (β¼0.54, po0.001) (Fig. 2). Thus, negative other– self evaluations fully mediated the relationship between overprotection and paranoia severity.

4. Discussion Our results indicate that parental bonding may play an important role in the development of psychopathology in adults. Moreover, our results showed that different patterns of parental bonding are involved in depression and paranoia severity. Although both clinical groups showed a mixed pattern of parental bonding dysfunction, correlation analyses showed that individuals who recalled their parents as being more overprotective reported greater paranoia, whereas individuals who recalled their parents as less caring reported greater depression. These results suggest that developmental trajectories leading to depression or paranoia may be different starting from an early stage. Moreover, they appear to show that paranoia is associated with parental overprotection, whereas depression is associated with a lack of parental care. This pattern may help to clarify inconsistent findings (Berry et al., 2007; Onstad et al., 1994; Rankin et al., 2005). Interestingly, parental overprotection has been previously associated with a defensive style which is, as discussed below, very relevant to paranoia. For example, previous research has found that avoidant coping and dissociation, which are two forms of defensive reactions, are related to parental overprotection but not to parental care (Yoshizumi et al., 2007).

80

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81

Regarding evaluative beliefs, our results indicate that both depressed and paranoid patients showed negative self–self and other–self evaluative beliefs when compared to controls. Our results are only partially consistent with initial reports by Chadwick and Trower (1997) because we did not find significantly higher negative self–other evaluations in the paranoid group in comparison to depressed participants. However, regression analyses indicated that self–other was predictive of paranoia severity. It is important to note that depressive severity was significantly increased in the paranoia group when compared to the control group (Table 3). Similarly, paranoia severity was elevated in the depression group. This fact can explain the incongruent results regarding both parental practices and negative evaluations found when comparing clinical groups. As such, our results support the analysis strategy of using depressive and paranoia severity as dependent variables instead of clinical groups. Although showing a partial mediation, the final model was consistent with our hypothesis in regards to depression. These results support previous research showing that, besides a consequence of depression, negative and unstable self-esteem is a vulnerability factor to depression (Liu, 2003; Grotmol et al., 2010; Roberts and Gotlib, 1997). Our results also supported that some – not all – roots of self–self evaluations are connected to early parenting practices in depression. Regarding paranoia, the resulting model seems to be more novel. Negative self–self and other–self evaluations fully mediated the impact of overprotection in paranoia. Our results are consistent with a model in which the negative self–self evaluations evolve into negative other–self evaluations which appear to activate paranoid thinking. This interpretation fits with defensive explanations of paranoia (e.g., Bentall et al., 2008). Therefore, other–self evaluations can represent the individual's attempts to maintain self-esteem in a self-threatening scenario. Consistent with this reasoning, prior research has shown that a decrease in self-esteem is associated with exacerbation of paranoid thinking (Thewissen et al., 2008). Moreover, Provencio et al. (2012) found that only depressive primes activated an attentional bias towards threatening interpersonal information in paranoia-prone individuals. In light of our results, this attentional bias could be a defensive reaction where self-threats activate an anxiety-based interpersonal schema. Likewise, our results suggest that negative self-schemas are related to paranoia through schemas about others (i.e., others–self evaluations). Lincoln et al. (2010) found that the perception of not being accepted by others was associated with psychotic symptoms. Our results point out that the perception of not being accepted by others may be a consequence of previous negative self–self appraisals. In contrast, although self–other evaluations correlated with paranoid symptoms, they did not correlate with overprotection and, as a consequence, were not suitable as a mediator. This is not surprising since paranoia is related to suspiciousness about others intentions more than to an extreme negativity about others. Thus, it is reasonable that other–self concerns, as opposed to self–other evaluations, fit better in the final mediational model of paranoia. At the same time, self–other evaluations might act more as a consequence of paranoia rather than a cause, which would explain correlations with other–self evaluations and with paranoia symptoms. Overall, results converge with recent research that emphasizes the particular significance of interpersonal schemas in psychosis and paranoia (Lincoln et al., 2010; Perry et al., 2011). Research has focused more on the intrapersonal schemas in paranoia (Fowler et al., 2006). However, studies examining interpersonal schemas are still scarce (Valiente et al., 2010).

4.1. Limitations Although the findings are promising, the PBI measure relies on retrospective recalls of parental behaviors, which leaves it open to possible influences of mood congruent or recall biases. Some studies indicate that although these biases remain a concern, their impact has been underscored (Brewin et al., 1993). In addition, the present study used averaged indexes of PBI from both parents. Further research should be aimed at clarifying the association of dysfunctional parenting experiences and negative self-evaluations in depression and paranoia, by considering different clusters of perceived paternal/ maternal bonding. This would help to better clarify the complex paternal vs. maternal dynamics that lead to depression and paranoia. Furthermore, the amount of variance in psychopathology explained by factors examined in the present study was modest, and there may be other important independent factors, such as adverse events, that could be linked to negative evaluative beliefs and psychopathology beyond parental factors. In addition, other mediational variables may play an important role. For instance, as several authors have pointed out, not only is the content of the schema relevant, but also other selfrelated processes (e.g. certainty, discrepancy or stability) (Petty et al., 2012). Further research into other relevant self-related processes that may influence individual vulnerability to depression and paranoia should be conducted. Finally, despite the fact that our results provide support for a mediational role of evaluative beliefs in the relationship between parenting and psychopathology, our research design does not allow for testing this causal hypothesis. Therefore, it would be important to evaluate the relationship between negative evaluative beliefs and increased paranoia with experimental designs directed to induce different cognitive evaluative styles and to examine their role in paranoid ideation. 4.2. Clinical implications In sum, this study provides support for the role of cognition in mediating the link between dysfunctional parenting and psychopathology. It also feeds into the literature that shows how depression and paranoia have a phenomenological overlap with shared but also separated vulnerability factors. Our study found a distinct pattern of dysfunctional parenting and negative beliefs when comparing depression to paranoia. Moreover, by assessing beliefs about different perspectives (other–self vs. self–self), we aimed to capture a wider spectrum of the cognitive style tapping inter- and intra-personal facets of evaluative beliefs. Consistent with previous theoretical and empirical evidence, a more complex view on the source of paranoid thinking, integrating distal and proximal factors, emerged. Future research should aim for methods of studying paranoia over time in relation to changes in underlying self-evaluations and interpersonal factors. Psychological interventions can be improved if there is an increased understanding of psychological mechanisms involved in the development of paranoid beliefs.

Acknowledgments This study was supported by a grant from the Iþ DþI Spanish Ministry of Education and Science, ref. PSI2009-13472, PSI2012-31494 and PSI2012-38298. We thank Hanna Song for her collaboration proofreading the final manuscript. References American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders (DSM-V), 4th ed. American Psychiatric Press, Washington DC. Avagianou, P.A., Zafiropoulou, M., 2008. Parental bonding and depression: personality as a mediating factor. International Journal of Adolescent Medicine and Health 20, 261–269, http://dx.doi.org/10.1515/IJAMH.2008.20.3.261.

C. Valiente et al. / Psychiatry Research 215 (2014) 75–81

Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology 51, 1173–1182, http://dx.doi. org/10.1037/0022-3514.51.6.1173. Bartholomew, K., 1990. Avoidance of intimacy: an attachment perspective. Journal of Social and Personal Relationships 7, 147–178, http://dx.doi.org/10.1177/ 0265407590072001. Beck, A.T., 1967. Depression: Clinical, Experimental and Theoretical Perspectives. Hoeber, New York (NY). Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for the Beck Depression Inventory, second edition The Psychological Corporation, San Antonio, TX. Bentall, R.P., Rowse, G., Kinderman, P., Blackwood, N., Howard, R., Moore, R., Cummins, S., Corcoran, R., 2008. Paranoid delusions in schizophrenia spectrum disorders and depression: the transdiagnostic role of expectations of negative events and negative self-esteem. Journal of Nervous and Mental Disease 196, 375–383, http://dx.doi.org/10.1097/NMD.0b013e31817108db. Berry, K.K.L., Barrowclough, C., Wearden, A.J., 2007. A review of the role of adult attachment style in psychosis: unexplored issues and questions for further research. Clinical Psychology Review 27, 458–475, http://dx.doi.org/10.1016/j. cpr.2006.09.006. Blatt, S.J., Homann, E., 1992. Parent–child interaction in the etiology of dependent and self-critical depression. Clinical Psychology Review 12, 47–91, http://dx.doi. org/10.1016/0272-7358(92)90091-L. Blatt, S.J., Wein, S.J., Chevron, E., Quinlan, D.M., 1979. Parental representations and depression in normal young adults. Journal of Abnormal Psychology 88, 388–397, http://dx.doi.org/10.1037//0021-843X.88.4.388. Bowlby, J., 1977. The making and breaking of affectional bonds: Part 1. Etiology and psychopathology in light of attachment theory. British Journal of psychiatry 130, 201–210, http://dx.doi.org/10.1192/bjp.130.3.201. Bowlby, J., 1980. Attachment and Loss, Vol. 3: Loss: Sadness and Depression. Basic Books, New York (NY). Brewin, C.R., Andrews, B., Gottlib, I.H., 1993. Psychopathology and early experience: are appraisal of retrospective reports. Psychological Bulletin, 113; , pp. 82–89, http://dx.doi.org/10.1037//0033-2909.113.1.82. Campos, R.C., Besser, A., Blatt, S.J., 2010. The mediating role of self-criticism and dependency in the association between perceptions of maternal caring and depressive symptoms. Depression and Anxiety 27, 1149–1157, http://dx.doi. org/10.1002/da.20763. Chadwick, P., Trower, P., 1997. To defend or not to defend: a comparison of Paranoia and Depression. Journal of Cognitive Psychotherapy: An international Quaterly 11, 63–71. (Available from: 〈http://www.ingentaconnect.com/content/springer/ jcogp/1997/00000011/00000001/art00005〉). Chadwick, P., Birchwood, M., 1995. The omnipotence of voices: II The Beliefs About Voices Questionnaire (BAVQ). British Journal of Psychiatry 166, 773–776, http: //dx.doi.org/10.1192/bjp.166.6.773. Cohen, J., Cohen, P., 1983. Applied Multiple Regression/Correlation Analyses for the Behavioral Sciences, 2nd ed., Lawrence Erlbaum; Hillsdale NJ. Eberhart, N., Shih, J., Hammen, C., Brennan, P.A., 2006. Understanding the sex difference in vulnerability to adolescent depression: an examination of child and parent characteristics. Journal of Abnormal Child Psychology 34, 495–508, http://dx.doi.org/10.1007/s10802-006-9020-4. Emsley, R., Rabinowitz, J., Torreman, M., 2003. The factor structure for the Positive and Negative Syndrome Scale (PANSS) in recent-onset psychosis. Schizophrenia Research 61, 47–57, http://dx.doi.org/10.1016/S0920-9964(02)00302-X. Enns, M.W., Cox, B.J., Larsen, D.K., 2000. Perceptions of parental bonding and symptom severity in adults with depression: mediation by personality dimensions. Canadian Journal of Psychiatry 45, 263–268. (Available from: 〈https:// ww1.cpa-apc.org/Publications/Archives/CJP/2000/April/April2000.asp〉). First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 2002. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P). Biometrics Research, New York State Psychiatric Institute, New York. Fowler, D., Freeman, D., Smith, B., Kuipers, E., Bebbington, P., Bashforth, H., Coker, S., Gracie, A., Dunn, G., Garety, P., 2006. The Brief Core Schema Scales (BCSS). Psychometric properties and associations with paranoia and grandiosity in non-clinical and psychosis samples. Psychological Medicine 36, 749–759, http: //dx.doi.org/10.1017/S0033291706007355. Freeman, D., Garety, P., Fowler, D., Kuipers, E., Dunn, G., Bebbington, P., Hadley, C., 1998. The London East Anglia randomised controlled trial of cognitivebehaviour therapy for psychosis IV: Self esteem and persecutory delusions. British Journal of Clinical Psychology 37, 415–430, http://dx.doi.org/10.1111/ j.2044-8260.1998.tb01399.x. Garety, P.A., Kuipers, E.K., Fowler, D., Freeman, D., Bebbington, P.E., 2001. A cognitive model of the positive symptoms of psychosis. Psychological Medicine 31, 189–195, http://dx.doi.org/10.1017/S0033291701003312. Gracie, A., Freeman, D., Green, S., Garety, P.A., Kuipers, E., Hardy, A., Ray, K., Dunn, G., Bebbington, P., Fowler, D., 2007. The association between traumatic experiences, paranoia and hallucinations: a test of predictions of psychological models. Acta Psychiatrica Scandinavica 116, 280–289, http://dx.doi.org/10.1111/ j.1600-0447.2007.01011.x. Grotmol, K.S., Ekeberg, Ø., Finset, A., Gude, T., Moum, T., Vaglum, P., Tyssen, R., 2010. Parental bonding and self-esteem as predictors of severe depressive symptoms: a 10-year follow-up study of Norwegian physicians. Journal of Nervous and Mental Disease 198, 22–27, http://dx.doi.org/10.1097/NMD.0b013e3181c8189c. Ingram, R.E., Miranda, J., Segal, Z.V., 1998. Cognitive Vulnerability to Depression. Guilford Press, New York.

81

Kay, S., Fiszbein, A., Opler, L.A., 1987. The Positive and Negative Syndrome Scale for Schizophrenia. Schizophrenia Bulletin 13, 261–276, http://dx.doi.org/10.1093/ schbul/13.2.261. Krabbendam, L., Janssen, I., Bak, M., Bijl, R.V., de Graaf, R., van Os, J., 2002. Neuroticism and low self-esteem as risk factors for psychosis. Social Psychiatry and Psychiatric Epidemiology 37, 1–6, http://dx.doi.org/10.1007/s127-002-8207-y. Lincoln, T.M., Mehl, S., Ziegler, M., Kesting, M.L., Exner, C., Rief, W., 2010. Is fear of others linked to an uncertain sense of self? The relevance of self-worth, interpersonal self-concepts, and dysfunctional beliefs to paranoia. Behavior Therapy 41, 187–197, http://dx.doi.org/10.1016/j.beth.2009.02.004. Liu, Y., 2003. The mediators between parenting and adolescent depressive symptoms: dysfunctional attitudes and self-worth. International Journal of Psychology 38, 91–100, http://dx.doi.org/10.1080/00207590244000205. McGinn, L.K., Cukor, D., Sanderson, W.C., 2005. The relationship between parenting style, cognitive style, and anxiety and depression: does increased early adversity influence symptom severity through the mediating role of cognitive style? Cognitive Therapy and Research 29, 219–242, http://dx.doi.org/10.1007/s10608005-3166-1. Melo, S., Corcoran, R., Shryane, N., Bentall, R.P., 2009. The Persecution and Deservedness Scale (PaDS). Psychology and Psychotherapy: Theory, Research and Practice 82, 247–260, http://dx.doi.org/10.1348/147608308  398337. Mickelson, Kristin D., Kessler, Ronald C., Shaver, Phillip R., 1997. Adult attachment in a nationally representative sample. Journal of Personality and Social Psychology 73 (5), 1092–1106, http://dx.doi.org/10.1037/0022-3514.73.5.1092). Onstad, S., Skre, I., Torgersen, S., Kringlen, E., 1994. Family interaction: parental representation in schizophrenic patient. Acta Psychiatrica Scandinavica 90 (Suppl), 67–70, http://dx.doi.org/10.1111/j.1600-0447.1994.tb05893.x. Parker, G., 1993. Parental rearing style: examining for links with personality vulnerability factors for depression. Social Psychiatry and Psychiatric Epidemiology 28, 97–100, http://dx.doi.org/10.1007/BF00801738. Parker, G., Tupling, H., Brown, L.B., 1979. A parental bonding instrument. British Journal of Medical Psychology 52, 1–10, http://dx.doi.org/10.1111/j.20448341.1979.tb02487.x. Perry, Y., Henry, J.D., Sethi, N., Grisham, J.R., 2011. The pain persists: how social exclusion affects individuals with schizophrenia. British Journal Clinical Psychology 50, 339–349, http://dx.doi.org/10.1348/014466510  523490. Petty, R.E., Briñol, P., Johnson, I., 2012. Implicit ambivalence. In: Gawronski, B., Strack, F. (Eds.), Cognitive Consistency: A Unifying Concept in Social Psychology. Guilford Press, New York, pp. 178–201. Pickering, L., Simpson, J., Bentall, R.P., 2008. Insecure attachment predicts proneness to paranoia but not hallucinations. Personality and Individual Differences 44, 1212–1224, http://dx.doi.org/10.1016/j.paid.2007.11.016. Provencio, M., Vázquez, C., Valiente, C., Hervás, G., 2012. Depressive primes stimulate initial avoidance of angry faces: an eye-tracking study of paranoid ideation. Cognitive Therapy and Research 36, 483–492, http://dx.doi.org/ 10.1007/s10608-011-9388-5. Rankin, P., Bentall, R.P., Hill, J., Kinderman, P., 2005. Perceived relationships with parents and paranoid delusions: comparisons of currently Ill, remitted and normal participants. Psychopathology 38, 16–25, http://dx.doi.org/10.1159/000083966. Restifo, K., Akse, J., Guzman, N.V., Benjamins, C., Dick, K., 2009. A pilot study of selfesteem as a mediator between family factors and depressive symptoms in young adult university students. Journal of Nervous and Mental Disease 197, 166–171, http://dx.doi.org/10.1097/NMD.0b013e318199f790. Roberts, J.E., Gotlib, I.H., 1997. Temporal variability in global self-esteem and specific self-evaluation as prospective predictors of emotional distress: specificity in predictors and outcome. Journal of Abnormal Psychology 106, 521–529, http://dx.doi.org/10.1037//0021-843X.106.4.521. Sheehan, D.V., Lecrubier, Y., 2002. MINI International Neuropsychiatric Interview for DSM-IV (English Version 5.0.0). University of South Florida, Tampa. Sobel, M.E., 1982. Asymptotic confidence intervals for indirect effects in structural equation models. In: Leinhardt, S. (Ed.), Sociological Methodology. American Sociological Association, Washington DC, pp. 290–312. Thewissen, V., Bentall, R., Lecomte, T., van Os, J., Myin-Germeys, I., 2008. Fluctuations in self-esteem and paranoia in the context of daily life. Journal of Abnormal Psychology 117, 143–153, http://dx.doi.org/10.1037/0021-843X.117.1.143. Valiente, C., Espinosa, R., Vázquez, C., Cantero, D., Fuentenebro, F., 2010. World assumptions in psychosis; do paranoid patients believe in a just world? Journal of Nervous and Mental Disease 198, 802–806, http://dx.doi.org/10.1097/ NMD.0b013e3181f9807b. Valiente, C., Cantero, D., Vázquez, C., Sánchez, A., Provencio, M., Espinosa, R., 2011. Implicit and explicit self-esteem discrepancies in paranoia and depression. Journal of Abnormal Psychology 120, 692–699, http://dx.doi.org/10.1037/ a0022856. World Health Organization, 1992. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, Switzerland. Young, J.E., Brown, G., 1990. Schema Questionnaire. Cognitive Therapy Center of New York, New York. Yoshizumi, T., Murase, S., Murakami, T., Takai, J., 2007. Dissociation as a mediator between perceived parental rearing style and depression in an adult community population using college students. Personality and Individual Differences 43, 353–364, http://dx.doi.org/10.1016/j.paid.2006.12.010. Zigler, E., Glick, M., 1988. Is paranoid schizophrenia really camouflaged depression? American Psychologist 43, 284–290, http://dx.doi.org/10.1037//0003066X.43.4.284.