Borderline personality disorder and compulsive buying: A multivariate etiological model

Addictive Behaviors 60 (2016) 117–123

Contents lists available at ScienceDirect

Addictive Behaviors journal homepage: www.elsevier.com/locate/addictbeh

Borderline personality disorder and compulsive buying: A multivariate etiological model Aniko Maraz a,b,⁎, Róbert Urbán a, Zsolt Demetrovics a a b

Institute of Psychology, Eötvös Loránd University, Budapest, Hungary Doctoral School of Psychology, Eötvös Loránd University, Budapest, Hungary

H I G H L I G H T S • • • •

26% of those with CBD scored positive for BPD Contingent self-esteem is highly predictive of CBD in the presence of BPD among women. Psychiatric distress was predictive of CBD in the presence of BPD among men. Gender-dependent etiological risk factors support different theoretical models.

a r t i c l e

i n f o

Article history: Received 3 November 2015 Received in revised form 7 April 2016 Accepted 12 April 2016 Available online 19 April 2016 Keywords: Impulsivity Self-esteem Path analysis Etiology Contingent self-esteem

a b s t r a c t Compulsive buying disorder (CBD) and borderline personality disorder (BPD) are highly comorbid. Accordingly, the theories concerning the etiological factors of CBD and BPD overlap. The present study aimed to test a multivariate etiological model of CBD in the presence of BPD. Our sample comprised shopping mall customers (N = 1409, 63% female, mean age: 31.2 SD: 12.1). CBD was measured using the Edwards Compulsive Buying Scale Revised Version, and BPD was measured using the McLean Screening Instrument for BPD. Impulsivity, global self-esteem, contingent self-esteem, and psychiatric distress were also assessed. Overall, 8.5% (n = 120) and 7.7% (n = 109) of participants scored above the cut-off value on the CBD and BPD measures respectively. In total, 2.2% (n = 31) had both BDP and CBD 26% of those with CBD also had BPD, corresponding to a risk ratio of 5.4. Path analysis revealed that contingent self-esteem (β = 0.27 p b 0.001) and impulsivity (β = 0.27 p b 0.001) exerted the strongest mediating effects, followed by psychiatric distress (β = 0.15 p b 0.15) and global self-esteem (β = 0.12 p b 0.001) after controlling for the effect of BPD. Distress had higher predictive value for men and contingent self-esteem had higher predictive value for women. The data suggest that impulsivity, contingent self-esteem, and distress (to a lesser extent) account for the high comorbidity between BPD and CBD and that etiological risk factors differ between men and women. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction 1.1. General introduction The DSM-5 lists compulsive buying disorder (CBD) as an example of the impulsivity criteria of borderline personality disorder (BPD) (American Psychiatric Association, 2013). In spite of this, only few data are available on the comorbidity of CBD and BPD, which appears to be substantially high. Given that the theories about etiological factors of CBD largely overlap with those of BPD, the present study aimed to test the factors that predict CBD in the presence of BPD. These etiological factors may account for the high comorbidity between BPD and CBD. ⁎ Corresponding author at: Department of Clinical Psychology and Addiction, Institute of Psychology, Eötvös Loránd University, Izabella u. 46, Budapest, Hungary. E-mail address: [email protected] (A. Maraz).

http://dx.doi.org/10.1016/j.addbeh.2016.04.003 0306-4603/© 2016 Elsevier Ltd. All rights reserved.

Compulsive buying disorder is a behavioural addiction characterised by a chronic and repetitive purchasing behaviour that becomes a primary response to negative events and feelings (Faber, O'Guinn, & Krych, 1987). It is associated with significant psychological distress, and it may result in financial difficulties (Black, 2007; Maraz et al., 2015). About 5% (3.4–6.9%) of the general population has experienced compulsive buying in their lifetime, and women are somewhat more affected compared to men (Maraz, Griffiths, & Demetovics, 2016). Given the substantial overlap with other disorders (Mueller et al., 2010) and the challenges in determining the core symptom, there is an ongoing debate whether CBD is a compulsive–obsessive (e.g., Hollander, 1993), impulse control (e.g., Christenson et al., 1994), or mood (e.g., Lejoyeux, Tassain, Solomon, & Ades, 1997) disorder. Borderline personality disorder, on the other hand, is an official diagnostic entity which is “characterised by a pervasive pattern of instability in affect regulation, impulse control, interpersonal relationships, and self-image” (p. 453) (Lieb, Zanarini,

118

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123

Schmahl, Linehan, & Bohus, 2004). BPD affects about 2% of the general population, although females are more likely to be affected (Swartz, Blazer, George, & Winfield, 1990). Various theories regarding the etiology of CBD include low selfesteem, impulsivity, and distress. The objects we buy become a part of self; therefore, buying can be regarded as an attempt to control and strengthen the self and identity (Belk, 2000; Dittmar, 2007). From this perspective, possessions provide an opportunity to present the self to others in a way that triggers positive feedback and enhances selfesteem. Studies confirm this theory by reporting that compulsive buyers have lower self-esteem compared to non-compulsive buyers (Hanley & Wilhelm, 1992; O'Guinn & Faber, 1989). Although previous studies measured global self-esteem, we propose that contingent selfesteem may have higher predictive value in the behaviour than global self-esteem. As opposed to global self-esteem, which proposes that self-esteem is a trait-like attribute of the person, contingent selfesteem is conceptualised as the situation-dependent nature of selfesteem (Johnson & Blom, 2007). This corresponds well with the definition of compulsive buying which states that buying episodes are triggered by negative emotional states. On the other hand, CBD can also be regarded as a maladaptive attempt to self-regulate one's impulses to buy. From this perspective, impulsive buying is an under-regulation of the behaviour (the individual is unable to resist the buying impulses) and compulsive buying is a dysregulation of the behaviour (where buying is a maladaptive coping mechanism against distress and anxiety) (Faber & Vohs, 2004). Indeed, many studies have confirmed the positive association between CBD and impulsivity (Desarbo & Edwards, 1996; Lejoyeux, Mathieu, Embouazza, Huet, & Lequen, 2007). One aspect of impulsivity, urgency (the tendency to deliberately suppress automatized responses and to experience strong reactions, frequently in a context of negative affect) is an especially strong indicator of CBD (Billieux, Rochat, Rebetez, & Van der Linden, 2008). Compulsive buyers are more likely to experience negative affect states compared to non-compulsive buyers (Beatty & Ferrell, 1998; Miltenberger et al., 2003). For this reason, several authors have argued that buying may function as a self-regulatory mechanism to cope with negative feelings (Christenson et al., 1994; McElroy, Keck, Pope, Smith, & Strakowski, 1994; Miltenberger et al., 2003). From this perspective, compulsive buying, as any other addiction disorder, is an attempt to self-medicate against the negative mood states (Khantzian, 1997). Similarly to CBD, decreased self-esteem (Rüsch et al., 2007; Zeigler-Hill & Abraham, 2006), increased impulsivity (Berlin, Rolls, & Iversen, 2005; Dougherty, Bjork, Huckabee, Moeller, & Swann, 1999), negative affect and affect-regulation problems (Donegan et al., 2003) in individuals with borderline personality disorder have been described in the literature and included in the diagnostic criteria of BPD (American Psychiatric Association, 2013). Persons with BPD typically experience these negative traits chronically, for a long period, sometimes for decades (Zanarini, Frankenburg, Hennen, & Silk, 2003). However, etiological factors should be interpreted as interacting factors rather than separate systems. For example, impulsivity may be a result of an attempt to avoid emotional distress through behavioural means (Anestis, Selby, Fink, & Joiner, 2007). This emotional distress may arise from the guilt of an unnecessary purchase (Black, 2007). For this reason, etiological models, which incorporate multivariate variables, are a fundamental tool to investigate prognostic factors in epidemiological research (Tripepi, Jager, Dekker, & Zoccali, 2008). Studies propose that common factors mediate the development of substance use disorders (SUD) in the presence of BPD pathology. In a review by Trull, Sher, Minks-Brown, Durbin, and Burr (2000), personality traits of affect instability and impulsivity were found to be central to the development of both BPD and SUD; thus, they were found to account, at least partly, for the comorbidity between these disorders. Although longitudinal data have not yet confirmed these assumptions, there is substantial reason to believe that personality traits mediate a

large part of the association between BPD and substance abuse. To our knowledge, no previous studies explored a multivariate etiological model to test the relationship between BPD and another behavioural addiction. Compulsive buying and borderline personality disorder co-occur frequently. For example, Sansone and Wiederman (2012) and Sansone, Chang, Jewell, Sellbom, and Bidwell (2013) reported that excessive spending and borderline symptomatology are strongly associated (r = 0.49 and r = 0.43, respectively). In other studies, 15% of selfselected compulsive buyers (7 out of 46) (Schlosser, Black, Repertinger, & Freet, 1994) and 20% of treatment-seeking compulsive buyers (6 out of 30) (Mueller et al., 2009) had BPD. Due to their high comorbidity, the DSM-5 lists CBD as an example of the impulsivity criteria of BPD (American Psychiatric Association, 2013). Despite this fact, it is still unclear which factors contribute to the development of CBD in the presence of BPD symptoms. Specifying mediating factors would contribute to the prevention of CBD as a symptom of BPD as well as to determine treatment directions in an effort to reduce excessive buying tendencies. Studies assessing the degree of functional impairment indicate that those with a BPD diagnosis or features exhibit a greater degree of functional impairment than participants with other, non-BPD personality disorder features (Hueston, Mainous, & Schilling, 1996; Nakao et al., 1992). Given that having more than one diagnosis is associated with higher functional impairment than having „only” one diagnosis (Trull et al., 2000), it is paramount to explore etiological factors behind the comorbidity in an effort to increase functionality. To our knowledge, this is the first study to investigate a multivariate etiological model that considers BPD along with a behavioural addiction, such as CBD. 1.2. Aims of the study The aim of the present study was to explore the comorbidity between borderline personality disorder and compulsive buying disorder. More specifically, we aimed to investigate a multivariate etiological model with impulsivity, self-esteem and psychiatric distress as candidate mediators of the relation between borderline personality disorder and compulsive buying disorder. We hypothesised that high impulsivity, deficit in self-esteem and high psychiatric distress each increase the probability of having CBD in the presence of BPD. Finally, we aimed to test gender differences in the model. 2. Methods 2.1. Participants and procedure The study used the data from participants recruited at three different shopping malls in Budapest and one in Győr (Western-Hungary) between April and November 2012. We utilized this sample in previous, unrelated studies (Maraz, Király and Demetrovics, 2015, Maraz, van den Brink and Demetrovics, 2015), none of which examined the mediating factors between BPD and CBD. Overall, 37,469 people passed the entrance to the mall at the time of data collection. Customers who were excluded from data collection (N = 8840) included those who (i) were younger than 18 years of age, (ii) did not have an email address, and/or (iii) did not speak Hungarian. Of the 28,629 individuals approached, 5068 persons agreed to participate (17.7%) by providing an email address. Participants provided written consent and provided their e-mail address. An e-mail was sent containing study link and password within 24 h of providing consent. Overall, 1776 individuals began filling out the questionnaire, with 1409 of them completing and providing valid responses with no missing data on any of the variables used for this study (27.8%). Those who participated did not receive any financial remuneration for participating. However, all participants received brief feedback

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123

regarding their buying behaviour at the end of the study. The Institutional Review Board approved the study design. 2.2. Measures 2.2.1. Edwards Compulsive Buying Scale Revised Version (ECBS-R) Items in the original instrument were based on the compulsive buying criteria by Faber and O'Guinn (1992) and measured on a 5-point Likert scale (Edwards, 1993). The revised version (Maraz et al., 2015) contains 16 items loading on four subscales: Lack of Control, Mood Modification, Guilt and Unnecessary Buying. The compulsive spending score (grand total) represents the average of all items, with a cut-off score of 42 and above indicating the probable presence of compulsive buying disorder. The questionnaire was translated into Hungarian, then back-translated by an independent reviewer. Any discrepancies were resolved involving a third, English-speaking expert in the field of psychology. Cronbach alpha for the scale was 0.87. 2.2.2. McLean Screening Instrument for Borderline Personality Disorder (MSI-BPD) The instrument is based on the DSM-IV criteria for borderline personality disorder and is suitable to be used as an index of the severity of borderline symptoms (Zanarini et al., 2003). The ten items are rated on a dichotomous yes-or-no scale, with N7 affirmative answers indicating the probable presence of borderline personality disorder. The MSI-BPD has acceptable sensitivity (81%) and specificity (85%) when applying the BPD module of the Diagnostic Interview to DSM-IV Personality Disorders (Zanarini, Frankenburg, Sickel, & Yong, 1996) as external criterion. The scale was translated by the university's standard protocol (see the description of ECBS-R). Cronbach alpha was 0.75. 2.2.3. Brief symptom inventory (BSI) The Brief Symptom Inventory was used to assess psychiatric symptoms. The BSI is a 53-item self-report symptom inventory designed to assess psychiatric symptom patterns (Boulet & Boss, 1991; Derogatis & Melisaratos, 1983). Each questionnaire item is measured on a 5-point scale of distress ranging from 0 (not at all) to 4 (very much), for example “During the past seven days, how much were you distressed by nervousness or shakiness inside”. The BSI consists of nine primary symptom dimensions, somatization, obsessivecompulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. The Global Severity Index (GSI) is used as a summary of the nine symptom dimensions (49 items) and four additional items that are not included in any of the dimensions (Urbán, Szigeti, Kökönyei, & Demetrovics, 2014). Scale reliability was 0.96. 2.2.4. Global self-esteem (GSE) Self-esteem was assessed using the Hungarian version of the Rosenberg Self-Esteem Scale (Rosenberg, 1965). This scale contains five positively and five negatively worded items measured on a 4-point scale. This measure of global self-esteem has recently been confirmed in Hungarian population (Urbán et al., 2014). For clarity, in the present study, high GSE scores indicated lower self-esteem. Cronbach alpha for the scale was 0.86.

119

2.2.6. Barratt Impulsiveness Scale (BIS) The Hungarian version of the BIS-11, based on the original English version (Patton, Stanford, & Barratt, 1995), contains 21 items loading on three factors, Self-control, Impulsive behaviour, and Impatience (Varga et al., 2015). Items are measured on a scale ranging from 0 to 3, with higher scores indicating higher impulsivity. The scale was reliable (Cronbach alpha = 0.82). 2.3. Data analysis Skewness and Kurtosis were within the acceptable range (between −1 and 1) for all variables included in the final model (MSI-BPD, GSI, GSE, CSE, BIS total, and ECBS-R), except for the GSI, where S = 1.74 and K = 3.86. Path analysis within structural equation modelling (SEM) was used to test the proposed mediation models with maximum likelihood estimation (MLR), which is robust to non-normality. All models were fully saturated; therefore, the usual fit indices (χ2, CFI, RMSEA) of overall model fit were not applicable because of zero degrees of freedom. We also tested the gender differences in paths coefficients using Wald test, which provides information about the change in chisquare that results when free parameters are fixed (Hoyle, 1995). Descriptive analyses, t-test for gender differences, chi-square test for dichotomous and categorical variables, and Pearson product–moment correlation among continuous variables were calculated using SPSS 22.0 statistical software package, and all SEM analyses were performed using MPlus 7.3 (Muthen & Muthén, 1998–2012). Effect sizes were calculated using the risk ratio and 95% confidence interval (CI) for categorical variables and Cohen's d for continuous variables (Rosnow, Rosenthal, & Rubin, 2000). Cohen's d is defined as the difference between two means divided by the pooled standard deviation. According to Cohen (Cohen, 1988), effect size values of 0.2 to 0.3 indicate a “small” effect, values around 0.5 indicate a “medium” effect, and values above 0.8 indicate a “large” effect. Correlation coefficients were compared using Fisher's z test (Eden & Yates, 1933). Cases with missing data were deleted listwise. 3. Results 3.1. Sample description Two-thirds (63%, n = 886 out of 1409) of participants were female. The mean age was 31.2 years (SD = 12.1), with the youngest participant being 18 years and the oldest being 77 years old. Half of the sample (49%) had completed secondary and 43% completed graduate education. Overall, 42.8% of participants claimed to have an average standard of living, 41.9% claimed to live above the average, and 14.4% reported below average standard. About half of the participants were employed full time (48%), and 40% were unemployed. In total, 37% (N = 538) shopped at least monthly, 32% (N = 458) weekly, 6% every day (N = 87), and the rest shopped less than monthly. While 36% of non-CBD participants shop at least weekly, twice as many, 79% of CBD participants shop at least weekly out of which onefifth shop daily. This association is significant (χ2 = 21.85, p b 0.001). 3.2. Comorbidity between BPD and CBD

2.2.5. Contingent self-esteem (CSE) Refers to the external sources of a person's perceived self-worth, such as others' love and evaluation of competence (Johnson & Blom, 2007). Sample items are “I feel worthwhile only when I have performed well” or “I tend to suppress my own needs and emotions to make others feel good”. The 26 items assessing CSE are measured on a scale from 1 to 4, with higher scores indicating increased likelihood to base one's selfesteem on others' evaluation. CSE contains two sub-scales, Competencebased and Relation-based self-esteem. The scale was translated to Hungarian in the previously described way. Cronbach alpha was 0.94.

In total, 8.5% (n = 120) of the sample had CBD and 7.7% (n = 109) scored above the cut-off value for the MSI-BPD. In total, 2.2% (n = 31) had both BDP and CBD. This corresponds to a χ2 = 59.9 (p b 0.001). The risk ratio of developing CBD when someone has BPD was 5.41 (95% CI: 3.37–8.61). The correlation between MSI-BPD and ECBS-R was somewhat higher among men (r = 0.39 p b 0.001) than among women (r = 0.35 p b 0.001). This difference was non-significant (Fisher's z = 0.84, p = 0.40). On the other hand, the risk ratio for having CBD in the presence of BPD was higher among men (7.2 [CI: 2.9–17.9])

120

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123

than among women (4.7 [CI: 2.7–8.2]). Eight men (1.5%) and 23 women (2.6%) had both CBD and BPD. Women scored higher compared to men on all variables, except for the MSI-BPD for which no differences between genders emerged (see Table 1). All Cronbach's alphas were above the acceptable level, ranging from 0.75 to 0.96 (see Table 2). Significant low to medium correlations emerged among the variables (between 0.25 and 0.65). Zero-order correlation between ECBS-R and MSI-BPD was significant (r = 0.36, p b 0.001).

Table 2 Zero-order correlations between variables and Cronbach's alpha. 1 1. 2. 3. 4. 5. 6.

Borderline personality (MSI-BPD) Global severity (GSI) Global self-esteem (GSEa) Contingent self-esteem Impulsivity (BIS total) Compulsive buying (ECBS-R)

2

3

4

5

Cronbach's alpha 0.75

0.65 0.48 0.45 0.43 0.36

0.57 0.52 0.54 0.37 0.35 0.26 0.38 0.25 0.39 0.39

0.96 0.86 0.94 0.81 0.87

Notes: all correlations were significant on p b 0.001 level. MSI-BPD = McLean Screening Instrument for Borderline Personality Disorder. GSI = Global Severity Index. GSE =

3.3. Mediation Structural regression (path) analysis was carried out on the total sample of N = 1409. We estimated the fully saturated model. The final model with the significant paths is presented in Fig. 1. MSI-BPD significantly and strongly predicted all four mediating variables, with the standardized path coefficients ranging from 0.43 to 0.65. Each mediating variable significantly predicted compulsive buying tendencies, with standardized path coefficients between 0.12 and 0.27. Different routes (mediation pathways) between MSI-BPD and ECBS-R were also significant (via GSI: 0.09, GSE: 0.05, CSE: 0.11 and BIS: 0.12, all p b 0.05) while the direct route between MSI-BPD and ECBS-R had a significant although weak effect size (0.09, p = 0.016). Overall, the variables in the model had a standardized total effect size of 0.36 (p b 0.05) of which 75% was explained by the mediating variables (standardized indirect effect: 0.28 p b 0.05). The full model explained 27% of the total variance in ECBS-R. In the next step, the path analysis was conducted separately for men and women, and all paths were compared using Wald test. Overall, the results (see Fig. 2) were fairly similar to the first model (Fig. 1). Paths had a similar predictive effect in men compared to women except for two paths. The effect size between GSI and compulsive buying was 0.24 (p b 0.001) for men and 0.10 (p = 0.02) for women. This difference was significant (χ2 = 4.24 p = 0.039), indicating that GSI exerts stronger mediating effect among men compared to women. Furthermore, CSE mediated the effect of compulsive buying with an effect size of 0.17 (p b 0.001) for men and 0.23 (p b 0.001) for women, indicating that women who score high on the CSE are more likely to have CBD tendencies compared to men (χ2 = 4.17 p = 0.041). All other paths yielded similar effects within genders. The direct route between MSIBPD and ECBS-R was significant only for women (0.11 p b 0.05) and not for men (0.10, p N 0.05). Total standardized effect size was 0.39 (p b 0.001) for men and 0.35 (p b 0.001) for women, whereas the total indirect effect was 0.29 (p b 0.001) for men and 0.24 (p b 0.001) for women, indicating that the mediating variables explained 74% (men) and 68.6% (women) of the total effect. The full model explained a slightly higher proportion of the variance in ECBS-R for men (31%) than for women (24%).

Global self-esteem. BIS = Barratt Impulsiveness Scale. ECBS-R = Edwards Compulsive Buying Scale Revised Version. a High scores = low self-esteem.

4. Discussion In the current sample, 26% of those with compulsive buying disorder had also borderline personality disorder according to the screening tests. Individuals who score positively on BPD test are over five times more likely to develop CBD compared to those without BPD tendencies. The risk ratio of developing CBD when someone has BPD is almost twice as high in men (7.2) compared to women (4.7). The test of multivariate etiological models revealed that contingent self-esteem and impulsivity had the highest predictive effect on CBD, followed by psychiatric distress and (low) trait self-esteem in the presence of BPD tendencies and after controlling for overlap between variables. Psychiatric distress had a stronger predictive effect on CBD among men than among women, whereas high contingent self-esteem (the tendency to base one's self-esteem on others' evaluation) was highly predictive among women but had much lower value among men. These results provide support for the notion that common etiological factors mediate the development of CBD in the presence of BPD and that these factors differ by gender. The rate of comorbid BPD was somewhat higher in the present sample of shopping centre visitors compared to treatment seeking compulsive buyers (15–20%, Mueller et al., 2009; Schlosser et al., 1994). The prevalence rate of BPD in the current sample was higher (7.7%) compared to the prevalence of BPD in the general population (2%, Swartz et al., 1990). On the other hand, the association between CBD tendency and BPD symptoms in this sample (0.36) was similar to the associations reported in previous studies (0.43 and 0.49 in Sansone & Wiederman, 2012; Sansone et al., 2013, respectively). These results support the notion that CBD and BPD are highly comorbid and therefore warrant further research. Regarding the mediating factors, it appears that borderline symptoms are associated with increased psychiatric distress. However, increased psychiatric distress is not a strong predictor of compulsive

Table 1 Descriptive statistics and gender differences. Total

Borderline personality (MSI-BPD) Global severity (GSI) Global self-esteem (GSEa) Contingent Self-esteem Impulsivity (BIS total) Compulsive buying (ECBS-R)

Men

Women

Mean

SD

Mean

SD

Mean

SD

2.51 30.18 10.97 59.83 39.72 27.28

2.32 29.11 5.16 13.29 7.53 8.97

2.42 27.79 10.34 57.43 38.94 24.27

2.26 28.44 4.93 12.79 7.69 7.75

2.56 31.60 11.34 61.25 40.19 29.09

2.36 29.43 5.26 13.38 7.40 9.16

Gender difference (t=)

Cohen's d

1.10 2.41⁎ 3.53⁎⁎⁎ 5.24⁎⁎⁎ 3.02⁎ 10.67⁎⁎⁎

0.05 0.13 0.19 0.28 0.16 0.57

Notes: MSI-BPD = McLean Screening Instrument for Borderline Personality Disorder; GSI = Global Severity Index; GSE = Global self-esteem; BIS = Barratt Impulsiveness Scale; ECBS-R = Edwards Compulsive Buying Scale Revised Version. ⁎ p b 0.05. ⁎⁎⁎ p b 0.001. a High scores = low self-esteem.

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123

121

Fig. 1. Path analysis between MSI-BPD and ECBS-R. Notes: ***p b 0.001, p b 0.05, Total standardized direct effect: 0.36*; total standardized indirect effect: 0.28* The covariances between mediating variables are not presented here for the sake of clarity.

buying tendency after accounting for the overlap among mediating variables. Similarly, even though BPD was associated with low trait self-esteem (although to a lesser extent compared to distress), selfesteem was less likely to predict compulsive buying. Moreover, contingent self-esteem and impulsivity were both associated with BPD, and they remained strong predictors of CBD even after controlling for the overlap among mediators. This means that a tendency to enhance one's self-esteem via purchasing and possessing new items and an increased level of impulsivity are both risk factors for the development of CBD when borderline tendency is present. Furthermore, it is possible that a common underlying neurobiological vulnerability, such as deficiency in the serotonergic system, moderates the comorbidity between BPD and CBD (Coccaro, Silverman, Klar, Horvath, & Siever, 1994; Higley & Linnoila, 1997). On the other hand, robust differences emerged between men and women in the mediating paths. Men with borderline tendencies and low or insufficient distress regulation appear to be more likely to develop CBD compared to women for whom distress had a much lower predictive effect. Additionally, women with underlying borderline traits and with the need to enhance their self-esteem via possessions of new items (contingent self-esteem) seem to be more likely to develop CBD, whereas the same is not true for men. Thus, it is possible that men use compulsive buying as an attempt to cope with distress (Christenson et al., 1994; Khantzian, 1997; McElroy et al.,

1994), which has been observed in other addictive behaviours, such as alcohol use (Stoltenberg, Batien, & Birgenheir, 2008). Women, on the other hand, use buying activity to enhance their self-esteem via the possession of new items (Belk, 2000). Interestingly, in this study, borderline symptoms were equally strongly associated with distress, self-esteem (both contingent and trait-like) and impulsivity across genders, with no gender differences in the effects of trait-like selfesteem and impulsivity on CBD. This suggests that trait-like selfesteem is a low predictor of compulsive buying tendency, irrespective of gender, and that low self-esteem may be a consequence of impulsive behaviour (i.e., guilt and regret), rather than a risk factor of CBD by itself. Impulsivity, on the other hand, had equally strong effects in men and in women. The lack of gender differences in impulsivity (executive functions) is not unprecedented in the literature (Cross, Copping, & Campbell, 2011), but future studies should clarify the role of facets of impulsivity (i.e. urgency, see Billieux et al., 2008) in the mediation between CBD and BPD. The results are also in line with previous findings reporting that in female students, the lack of impulse control, and in male students depression emerged as independent significant predictors for CBD (Mueller et al., 2011). Contingent self-esteem partially overlaps with materialism. In this sense, compulsive buying is characterised by the motivation to move closer to the ideal self through material goods (Dittmar, 2005a). Indeed, studies have repeatedly confirmed the strong association between CBD

Fig. 2. Gender differences in path analysis between MSI-BPD and ECBS-R. Notes: Men (n = 523, first values, in blue), Women (n = 886, second values, in red); ***p b 0.001, **p b 0.01, *p b 0.05; n.s. = p N 0.05; Total standardized direct effect: 0.39*** (for men), 0.35*** (for women), Total standardized indirect effect: 0.33*** (for men), 0.24*** (for women). Dashed arrows indicate significant gender differences on the given route (Wald test, χ2 = 4.243 p = 0.039 for Global Severity – Compulsive Buying and χ2 = 4.173 p = 0.041 for Contingent Self-esteem – Compulsive Buying). The covariances between mediating variables are not presented here for the sake of clarity. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

122

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123

and materialism (Dittmar, 2005b; Mueller et al., 2011, 2014; Dittmar & Kapur, 2011), and found evidence that depression mediated the effect of materialism in women (Otero-López & Villardefrancos, 2013). Conceptually, materialism can be regarded as one aspect of contingent self-esteem. However, it remains a question whether self-esteem obtained via material goods (materialism) or a more general selfesteem earned through situational feedback (contingent self-esteem) play a dominant role in the development of CBD. Given that distress, impulsivity and self-esteem are highly overlapping concepts, it is possible that other variables, such as early trauma, dysfunctional family environment or common underlying neurobiological proneness, moderate the comorbidity (Trull, 1995). Furthermore, other variables, such as motivations (Desarbo & Edwards, 1996), urgency (Rose & Segrist, 2014) or cue-reactivity (Starcke, Schlereth, Domass, Schöler, & Brand, 2013), should also be considered in future studies. Finally, the comorbidity of CBD in a sample of BPD individuals remains unclear. BPD or CBD diagnosis rarely occurs in isolation. A growing body of evidence suggests that a comorbid BPD may serve as a negative prognostic factor in those suffering from a variety of Axis I conditions, such as anxiety disorders and major depression (e.g., Nurnberg et al., 1989; Shea, Widiger, & Klein, 1992). The current study proposed that interventions which target certain factors, such as impulsivity (for both genders), contingent self-esteem (for women), and improve coping with psychological distress (especially for men) are likely to be effective in the treatment and prevention of compulsive buying. Further research is needed to test whether targeting these factors reduces other Axis I conditions in the presence of BPD. The present study has several limitations. Despite the advantage of a large sample size, the self-selected and self-reported nature of the Hungarian-only data needs to be taken into consideration when generalizing the results to other samples. Furthermore, given that the positive predictive value (the proportion of true positive cases) of a screening test is low when the prevalence of the disease is low (Baldessarini, Finklestein, & Arana, 1983; Maraz, Király, & Demetrovics, 2015), future studies should consider using clinical interviews instead of screening tests when assessing a population with relatively low prevalence estimates. Given the cross-sectional nature of the data, causation and directionality of findings need to be confirmed in longitudinal and interventionbased studies. The question whether contingent self-esteem and/or psychiatric distress symptoms make individuals more vulnerable for compulsive buying, or whether compulsive buying leads to or increases contingent self-esteem and/or psychiatric distress, is unresolved. Future studies should also explore whether the proposed model (and the gender differences) extend to other Axis I disorders in the presence of BPD. Moreover, it remains a question as to what extent other factors such as trauma or abusive family environment contribute to the development of both disorders. 5. Conclusions Individuals with compulsive buying tendencies are also likely to have borderline traits. The data suggest that impulsivity, and to a lesser extent distress (among men) and contingent self-esteem (among women), may account for the high comorbidity between BPD and CBD and that etiological and risk factors differ between men and women. Acknowledgements Present work was supported by the Hungarian Scientific Research Fund (grant number: 111938).

References American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (DSM-5) (5th ed.). Arlington, VA: American Psychiatric Association.

Anestis, M. D., Selby, E. A., Fink, E. L., & Joiner, T. E. (2007). The multifaceted role of distress tolerance in dysregulated eating behaviors. International Journal of Eating Disorders, 40, 718–726. Baldessarini, R. J., Finklestein, S., & Arana, G. W. (1983). The predictive power of diagnostic tests and the effect of prevalence of illness. Archives of General Psychiatry, 40, 569–573. Beatty, S. E., & Ferrell, M. E. (1998). Impulse buying: Modeling its precursors. Journal of Retailing, 74, 169–191. Belk, R. (2000). Are we what we own. In A. L. Benson (Ed.), I shop therefore I am: Compulsive buying and the search for self (pp. 76–104). Berlin, H. A., Rolls, E. T., & Iversen, S. D. (2005). Borderline personality disorder, impulsivity, and the orbitofrontal cortex. American Journal of Psychiatry, 162, 2360–2373. Billieux, J., Rochat, L., Rebetez, M. M. L., & Van der Linden, M. (2008). Are all facets of impulsivity related to self-reported compulsive buying behavior? Personality and Individual Differences, 44, 1432–1442. Black, D. W. (2007). A review of compulsive buying disorder. World Psychiatry, 6, 14–18. Boulet, J., & Boss, M. W. (1991). Reliability and validity of the brief symptom inventory. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 3, 433–437. Christenson, G. A., Faber, R. J., de Zwaan, M., Raymond, N. C., Specker, S. M., Ekern, M. D., ... Eckert, E. D. (1994). Compulsive buying: Descriptive characteristics and psychiatric comorbidity. Journal of Clinical Psychiatry, 55, 5–11. Coccaro, E. F., Silverman, J. M., Klar, H. M., Horvath, T. B., & Siever, L. J. (1994). Familial correlates of reduced central serotonergic system function in patients with personality disorders. Archives of General Psychiatry, 51, 318–324. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, New Jersey: L: Erlbaum. Cross, C. P., Copping, L. T., & Campbell, A. (2011). Sex differences in impulsivity: A metaanalysis. Psychological Bulletin, 137, 97–130. Derogatis, L. R., & Melisaratos, N. (1983). The brief symptom inventory: An introductory report. Psychological Medicine, 13, 595–605. Desarbo, W. S., & Edwards, E. A. (1996). Typologies of compulsive buying behavior: A constrained clusterwise regression approach. Journal of Consumer Psychology, 5, 231–262. Dittmar, H. (2005a). A new look at” compulsive buying”: Self-discrepancies and materialistic values as predictors of compulsive buying tendency. Journal of Social and Clinical Psychology, 24, 832–859. Dittmar, H. (2005b). Compulsive buying-a growing concern? An examination of gender, age, and endorsement of materialistic values as predictors. British Journal of Psychology, 96, 467–491. Dittmar, H. (2007). Consumer culture, identity and well-being: The search for the 'good life' and the ‘body perfect’. New York: Psychology Press. Dittmar, H., & Kapur, P. (2011). Consumerism and well-being in India and the UK: Identity projection and emotion regulation as underlying psychological processes. Psychological Studies, 56, 71–85. Donegan, N. H., Sanislow, C. A., Blumberg, H. P., Fulbright, R. K., Lacadie, C., Skudlarski, P., ... Wexler, B. E. (2003). Amygdala hyperreactivity in borderline personality disorder: Implications for emotional dysregulation. Biological Psychiatry, 54, 1284–1293. Dougherty, D. M., Bjork, J. M., Huckabee, H. C., Moeller, F. G., & Swann, A. C. (1999). Laboratory measures of aggression and impulsivity in women with borderline personality disorder. Psychiatry Research, 85, 315–326. Eden, T., & Yates, F. (1933). On the validity of Fisher's z test when applied to an actual example of non-normal data. The Journal of Agricultural Science, 23, 6–17. Edwards, E. A. (1993). Development of a new scale for measuring compulsive buying behavior. Financial Counseling and Planning, 4, 67–84. Faber, R. J., & Vohs, K. D. (2004). To buy or not to buy?: Self-control and self-regulatory failure in purchase behavior. In R. F. Baumeister, & K. D. Vohs (Eds.), Handbook of self-regulation: Research, theory, and applications (pp. 509–524). New York, NY, US: Guilford Press. Faber, R. J., & O'Guinn, T. C. (1992). A clinical screener for compulsive buying. Journal of Consumer Research, 19, 459–469. Faber, R. J., O'Guinn, T. C., & Krych, R. (1987). Compulsive consumption. Advances in Consumer Research, 14, 132–135. Hanley, A., & Wilhelm, M. S. (1992). Compulsive buying: An exploration into self-esteem and money attitudes. Journal of Economic Psychology, 13, 5–18. Higley, J., & Linnoila, M. (1997). Low central nervous system serotonergic activity is traitlike and correlates with impulsive behavior. Annals of the New York Academy of Sciences, 836, 39–56. Hollander, E. (1993). Obsessive-compulsive spectrum disorders: An overview. Psychiatric Annals, 23, 355–358. Hoyle, R. H. (1995). Structural equation modeling: Concepts, issues, and applications. California: Sage Publications. Hueston, W. J., Mainous, A. G., & Schilling, R. (1996). Patients with personality disorders: Functional status, health care utilization, and satisfaction with care. The Journal of Family Practice, 42, 54–60. Johnson, M., & Blom, V. (2007). Development and validation of two measures of contingent self-esteem. Individual Differences Research, 5, 300–328. Khantzian, E. J. (1997). The self-medication hypothesis of substance use disorders: A reconsideration and recent applications. Harvard Review of Psychiatry, 4, 231–244. Lejoyeux, M., Mathieu, K., Embouazza, H., Huet, F., & Lequen, V. (2007). Prevalence of compulsive buying among customers of a Parisian general store. Comprehensive Psychiatry, 48, 42–46. Lejoyeux, M., Tassain, V., Solomon, J., & Ades, J. (1997). Study of compulsive buying in depressed patients. Journal of Clinical Psychiatry, 58, 169–173. Lieb, K., Zanarini, M. C., Schmahl, C., Linehan, M. M., & Bohus, M. (2004). Borderline personality disorder. The Lancet, 364, 453–461.

A. Maraz et al. / Addictive Behaviors 60 (2016) 117–123 Maraz, A., Király, O., & Demetrovics, Z. (2015). The diagnostic pitfalls of surveys: If you score positive on a test of addiction, you still have a good chance not to be addicted. A response to Billieux et al. 2015. Journal of Behavioural Addictions, 4, 151–154. Maraz, A., van den Brink, W., & Demetrovics, Z. (2015). Prevalence and construct validity of compulsive buying disorder in shopping mall visitors. Psychiatry Research, 228, 918–924. Maraz, A., Eisinger, A., Hende, B., Urbán, R., Paksi, B., Kun, B., ... Demetrovics, Z. (2015). Measuring compulsive buying behaviour: Psychometric validity of three different scales and prevalence in the general population and in shopping centres. Psychiatry Research, 225, 326–334. Maraz, A., Griffiths, M., & Demetovics, Z. (2016). The prevalence of compulsive buying: A meta-analysis. Addiction, 11, 408–419. McElroy, S. L., Keck, J. P. E., Pope, J. H. G., Smith, J. M., & Strakowski, S. M. (1994). Compulsive buying: A report of 20 cases. Journal of Clinical Psychiatry, 55, 242–248. Miltenberger, R. G., Redlin, J., Crosby, R., Stickney, M., Mitchell, J., Wonderlich, S., ... Smyth, J. (2003). Direct and retrospective assessment of factors contributing to compulsive buying. Journal of Behavior Therapy and Experimental Psychiatry, 34, 1–9. Mueller, A., Claes, L., Georgiadou, E., Möllenkamp, M., Voth, E. M., Faber, R. J., ... de Zwaan, M. (2014). Is compulsive buying related to materialism, depression or temperament? Findings from a sample of treatment-seeking patients with CB. Psychiatry Research, 216, 103–107. Mueller, A., Claes, L., Mitchell, J. E., Faber, R. J., Fischer, J., & de Zwaan, M. (2011). Does compulsive buying differ between male and female students? Personality and Individual Differences, 50, 1309–1312. Mueller, A., Mitchell, J. E., Black, D. W., Crosby, R. D., Berg, K., & de Zwaan, M. (2010). Latent profile analysis and comorbidity in a sample of individuals with compulsive buying disorder. Psychiatry Research, 178, 348–353. Mueller, A., Mühlhans, B., Silbermann, A., Müller, U., Mertens, C., Horbach, T., ... de Zwaan, M. (2009). Pathologisches Kaufen und psychische Komorbidität [Compulsive buying and psychiatric comorbidity]. Psychotherapie, Psychosomatik, Medizinische Psychologie, 59, 291–299. Muthen, L. K., & Muthén, B. O. (1998–2012). Mplus user's guide (7th ed.). Los Angeles, CA: Muthén & Muthén. Nakao, K., Gunderson, J. G., Phillips, K. A., Tanaka, N., Yorifuji, K., Takaishi, J., & Nishimura, T. (1992). Functional impairment in personality disorders. Journal of Personality Disorders, 6, 24–33. Nurnberg, H. G., Raskin, M., Levine, P. E., Pollack, S., Prince, R., & Siegel, O. (1989). Borderline personality disorder as a negative prognostic factor in anxiety disorders. Journal of Personality Disorders, 3, 205–216. O'Guinn, T. C., & Faber, R. J. (1989). Compulsive buying: A phenomenological exploration. Journal of Consumer Research, 16, 147–157. Otero-López, J. M., & Villardefrancos, E. (2013). Materialism and addictive buying in women: The mediating role of anxiety and depression. Psychological Reports, 113, 328–344. Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of the Barratt impulsiveness scale. Journal of Clinical Psychology, 51, 768–774. Rose, P., & Segrist, D. J. (2014). Negative and positive urgency may both be risk factors for compulsive buying. Journal of Behavioral Addictions, 3, 128–132. Rosenberg, M. (1965). Rosenberg self-esteem scale (RSE). Acceptance and commitment therapy. Measures package, Vol. 61, . Rosnow, R. L., Rosenthal, R., & Rubin, D. B. (2000). Contrasts and correlations in effect-size estimation. Psychological Science, 11, 446–453.

123

Rüsch, N., Lieb, K., Göttler, I., Hermann, C., Schramm, E., Richter, H., ... Bohus, M. (2007). Shame and implicit self-concept in women with borderline personality disorder. The American Journal of Psychiatry, 164, 500–508. Sansone, R. A., & Wiederman, M. W. (2012). Spending too much: Relationships with borderline personality symptomatology. International Journal of Psychiatry in Clinical Practice, 16, 316–318. Sansone, R. A., Chang, J., Jewell, B., Sellbom, M., & Bidwell, M. (2013). Compulsive buying and borderline personality symptomatology. Journal of Personality Disorders, 27, 260–268. Schlosser, S., Black, D. W., Repertinger, S., & Freet, D. (1994). Compulsive buying: Demography, phenomenology, and comorbidity in 46 subjects. General Hospital Psychiatry, 16, 205–212. Shea, M. T., Widiger, T. A., & Klein, M. H. (1992). Comorbidity of personality disorders and depression: Implications for treatment. Journal of Consulting and Clinical Psychology, 60, 857–869. Starcke, K., Schlereth, B., Domass, D., Schöler, T., & Brand, M. (2013). Cue reactivity towards shopping cues in female participants. Journal of Behavioral Addictions, 2, 17–22. Stoltenberg, S. F., Batien, B. D., & Birgenheir, D. G. (2008). Does gender moderate associations among impulsivity and health-risk behaviors? Addictive Behaviors, 33, 252–265. Swartz, M., Blazer, D., George, L., & Winfield, I. (1990). Estimating the prevalence of borderline personality disorder in the community. Journal of Personality Disorders, 4, 257–272. Tripepi, G., Jager, K. J., Dekker, F. W., & Zoccali, C. (2008). Testing for causality and prognosis: Etiological and prognostic models. Kidney International, 74, 1512–1515. Trull, T. J. (1995). Borderline personality disorder features in nonclinical young adults: 1. Identification and validation. Psychological Assessment, 7, 33. Trull, T. J., Sher, K. J., Minks-Brown, C., Durbin, J., & Burr, R. (2000). Borderline personality disorder and substance use disorders: A review and integration. Clinical Psychology Review, 20, 235–253. Urbán, R., Kun, B., Farkas, J., Paksi, B., Kökönyei, G., Unoka, Z., ... Demetrovics, Z. (2014a). Bifactor structural model of symptom checklists: SCL-90-R and brief symptom inventory (BSI) in a non-clinical community sample. Psychiatry Research, 216, 146–154. Urbán, R., Szigeti, R., Kökönyei, G., & Demetrovics, Z. (2014b). Global self-esteem and method effects: Competing factor structures, longitudinal invariance, and response styles in adolescents. Behavior Research Methods, 46, 488–498. Varga, G., Kapitány, M., Urbán, R., Paksi, B., Kun, B., Szekely, A., & Demetrovics, Z. (2015). Barratt impulsiveness scale. (Manuscript). Zanarini, M. C., Frankenburg, F. R., Hennen, J., & Silk, K. R. (2003a). The longitudinal course of borderline psychopathology: 6-year prospective follow-up of the phenomenology of borderline personality disorder. American Journal of Psychiatry, 160, 274–283. Zanarini, M. C., Frankenburg, F. R., Sickel, A. E., & Yong, L. (1996). The diagnostic interview for DSM-IV personality disorders (DIPD-IV). Belmont, MA: McLean Hospital. Zanarini, M. C., Vujanovic, A. A., Parachini, E. A., Boulanger, J. L., Frankenburg, F. R., & Hennen, J. (2003b). A screening measure for BPD: The McLean screening instrument for borderline personality disorder (MSI-BPD). Journal of Personality Disorders, 17, 568–573. Zeigler-Hill, V., & Abraham, J. (2006). Borderline personality features: Instability of self– esteem and affect. Journal of Social and Clinical Psychology, 25, 668–687.