انواع بدرفتاری با کودک و رفتارهای پرخطر: ارتباطات با سبک دلبستگی و ابعاد تنظیم احساسات

Abstract Child maltreatment is a robust psychosocial risk factor linked to the development of a wide range of risk behaviors among young adults. Adult attachment style and emotion dysregulation are two potential mechanisms through which maltreatment leads to risk behaviors. Yet, less is known about the specificity of the relations among different maltreatment types, attachment styles, emotion regulation strategies, and risk behaviors. The present study examined the relations among various forms of maltreatment and risk behaviors (e.g., substance use; risky sex) among 361 undergraduate students and tested whether attachment styles and emotion dysregulation might underlie these relations. Emotional, and sexual but not verbal abuse (although verbal abuse was directly related to alcohol use), were related to anxious and avoidant attachment styles, emotion dysregulation, and a variety of risk behaviors. Among the emotion regulation dimensions, impulsivity showed the strongest indirect effect from child maltreatment to risk behaviors. Results support a cross-sectional link between child maltreatment and risk behavior outcomes via attachment styles and emotion regulation. Implications for treatment and prevention of these risk behaviors in young adults are discussed.

1. Introduction Child maltreatment is associated with a variety of risk behaviors, yet the specific mechanisms that underlie this link are not fully understood. Difficulties with attachment style and emotion regulation (ER) are two factors that have received extensive empirical support in relation to both child maltreatment and risk behaviors (English and John, 2013, Gratz et al., 2009, Maughan and Cicchetti, 2002 and Wekerle and Wolfe, 1998). In fact, attachment quality and ER are primary developmental processes that can be affected by child maltreatment (Cicchetti & Valentino, 2006), turning these factors into putative mechanisms through which child maltreatment may lead to risk behaviors. More research is needed examining both maltreatment and ER in a less narrow (uni or bidimensional) manner (English and John, 2013 and Maughan and Cicchetti, 2002), which may obscure differences found when using more multifaceted articulated measures. The current study examines the specificity of linkages between child maltreatment types (e.g., emotional, sexual abuse), adult attachment styles (i.e., anxious and avoidant), ER dimensions (e.g., clarity, nonacceptance), and risk behaviors (e.g., alcohol use; antisocial behaviors) using a cross-sectional sample of young adults. Subsequently, indirect linkages between maltreatment and risk behaviors were tested via attachment styles and ER dimensions. 1.1. Child maltreatment and risk behaviors Child maltreatment poses challenges that can channel victims into maladaptive developmental pathways eventuating in risk behavior participation over time. For instance, adolescents and young adults with child maltreatment histories report greater participation in sexual risk behaviors (Hussey, Chang, & Kotch, 2006), cannabis use (Oshri, Rogosch, & Cicchetti, 2013), and alcohol use (Shin, Miller, & Teicher, 2013). The experience of chronic stress in childhood may permeate and disrupt multiple stage-salient tasks (e.g., relational and self-regulatory capacities) that continue to develop throughout adolescence and young adulthood (Cicchetti & Toth, 2005). Although research has shown specific associations from different types of maltreatment and risk taking behaviors (Oshri, Tubman, & Burnette, 2012), more research is needed on the specific relations of adult attachment styles and ER dimensions in explaining the connection between child maltreatment and participation in various risk behaviors. 1.2. Attachment style Attachment theory postulates that children develop the ability to regulate their emotions through supportive, sensitive parenting and secure attachment (Bowlby, 1969/1982). Attachment organization is theorized to be an evolutionary mechanism that serves the individual in coping with stressful situations. However, if parents are abusive or unavailable, children are likely to develop an insecure attachment style (e.g., anxiety) as attempts to seek out attachment figures do not provide relief from stress or fear (Mikulincer, Shaver, & Pereg, 2003). The impact of child maltreatment on attachment is established in childhood (Stronach et al., 2011), with this effect remaining relatively stable through adulthood (Weinfeld, Sroufe, & Egeland, 2000). Adult relationship researchers extended the concept of childhood attachment into adulthood using two insecure attachment dimensions (anxious and avoidant). Anxious individuals fear abandonment, are obsessive, and desire high levels of reciprocity with others whereas avoidant individuals fear intimacy and closeness and avoid committed relationships (Hazan & Shaver, 1987). Recent studies demonstrate that an insecure attachment style is a risk factor for engagement in antisocial behaviors (e.g., Allen, Porter, McFarland, McElhaney, & Marsh, 2007). Missing is research that can delineate the link between different types of child maltreatment, the two attachment styles and risk behaviors. 1.3. Emotion regulation ER is formed through a range of socialization experiences during development (Aldao, Nolen-Hoeksema, & Schweizer, 2010) and is composed of multiple facets reflecting the individual’s strategies to achieve emotional control. For example, attention to the occurrence of various emotions and the ability to correctly identify emotions are thought to be central to effective emotion modulation. Similarly, awareness, acceptance, and use of effective strategies to modulate emotions and subsequent arousal have been shown to play a key role in ER (Gratz & Roemer, 2004). The greatest plasticity of ER occurs during childhood and adolescence when cognitive and emotional capacities are rapidly forming (John & Gross, 2004). Parents and other caregivers serve as the primary source of emotion socialization in childhood and adolescence, and abusive parenting is related to poor ER strategies among children (Shipman & Zeman, 2001). For example, experiential avoidance has been hypothesized and recently shown as one behavioral characteristic that affects the risk for psychological difficulties linked to experiences of childhood abuse (Gratz, Bornovalova, Delany-Brumse, Nick, & Lejue, 2007). Theoretically, in the process of evaluating and interpreting their emotions, abuse victims may be reinforced via secondary emotional responses (e.g., anxiety) to avoid “aversive” emotions (Gratz et al., 2007). Maladaptive emotion regulation strategies are associated with multiple risk behaviors (Simons, Maisto, & Wray, 2010). Thus, child maltreatment may be linked with risk behaviors via ineffective ER strategies. An effective test of this hypothesis requires the use of a multivariate analytic strategy such as structural equation modeling that can parse relevant constructs into specific and smaller components while accounting for their shared method error variances. 1.4. The present investigation The aim of the present study is to examine the direct and indirect effects that account for the multivariate associations between child maltreatment types and risk behaviors in young adults. The specific hypotheses and aims of the present study are as follows. First, we examine the specificity of associations between child maltreatment types, attachment anxiety and avoidance, and with six dimensions of ER. Second, the study tests which adult attachment styles and emotional regulation dimensions are most strongly related with four risk behaviors: alcohol use, drug use, condom use, and antisocial behaviors. Lastly, the last aim is to identify direct and indirect linkages between child maltreatment types and risk behaviors via attachment styles and ER dimensions. We hypothesize significant indirect links between child maltreatment types and risk behaviors via insecure attachment styles and reduced ER.

3. Results 3.1. Descriptive analyses Table 1 summarizes the bivariate zero-order correlations among the variables included in this study. Table 1. Zero order correlations for study variables. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Abuse 1. Physical – 2. Verbal .54⁎⁎ – 3. Sexual .69⁎⁎ .37⁎⁎ – 4. Emotional .63⁎⁎ .75⁎⁎ .45⁎⁎ – Attachment 5. Anxiety .26⁎⁎ .31⁎⁎ .24⁎⁎ .35⁎⁎ – 6. Avoidance .22⁎⁎ .18⁎⁎ .24⁎⁎ .22⁎⁎ .52⁎⁎ – Emotion regulation 7. Clarity .27⁎⁎ .32⁎⁎ .27⁎⁎ .34⁎⁎ .44⁎⁎ .34⁎⁎ – 8. Strategies .26⁎⁎ .33⁎⁎ .24⁎⁎ .37⁎⁎ .51⁎⁎ .23⁎⁎ .53⁎⁎ – 9. Awareness .14⁎⁎ .04 .15⁎⁎ .05 .10 .27⁎⁎ .35⁎⁎ .04 – 10. Impulse .32⁎⁎ .34⁎⁎ .28⁎⁎ .35⁎⁎ .38⁎⁎ .24⁎⁎ .53⁎⁎ .70⁎⁎ .16⁎⁎ – 11 Goals .13⁎ .23⁎⁎ .10 .28⁎⁎ .31⁎⁎ .09 .28⁎⁎ .50⁎⁎ −.07 .44⁎⁎ – 12. Nonacceptance .23⁎⁎ .32⁎⁎ .26⁎⁎ .34⁎⁎ .45⁎⁎ .28⁎⁎ .52⁎⁎ .63⁎⁎ .10 .55⁎⁎ .42⁎⁎ – Risk behaviors 13. Alcohol Use −.01 .12⁎ .07 −.02 .12⁎ .12⁎ .14⁎⁎ .10 .12⁎ .17⁎⁎ .05 .11⁎ – 14. Drug Use .02 .11⁎ .10 .08 .05 .14⁎ .16⁎⁎ .13⁎ .11⁎ .23⁎⁎ .05 .08 .65⁎⁎ – 15. Condom Use −.13 −.15⁎ −.11 −.14⁎ −.01 −.01 −.10 −.09 −.14⁎⁎ −.17⁎⁎ .05 −.04 −.26⁎⁎ −.38⁎⁎ – 16. Antisocial behavior .14⁎⁎ .26⁎⁎ .15⁎⁎ .22⁎⁎ .12⁎ .16⁎⁎ .19⁎⁎ .13⁎ .14⁎⁎ .26⁎⁎ .11⁎ .15⁎⁎ .34⁎⁎ .51⁎⁎ −.31⁎⁎ – 17. Age .06 .05 .00 .12⁎ .01 −.01 .02 .08 .06 .05 .08 .07 .10 .21⁎⁎ −.08 .08 – 18. Sex .00 .09 .00 .17⁎⁎ .02 −.01 .10 .06 −.13⁎ −.02 .09 .09 −.12⁎ −.08 .11⁎⁎ −.33⁎⁎ −.02 – Mean 0.22 0.92 0.16 0.63 3.34 3.02 2.09 1.95 2.28 1.79 2.96 2.21 −0.04 1.47 4.33 0.69 19.11 0.62 Std. Dev. 0.51 0.87 0.55 0.90 1.18 1.09 0.71 0.82 0.68 0.73 1.00 0.93 0.67 1.07 1.32 0.58 1.68 0.49 Note: N = 361; Sex coded as 0 = male, 1 = female. ⁎ p < .05. ⁎⁎ p < .01. Table options 3.2. Evaluation of structural model 3.2.1. Child maltreatment, attachment, and risk behaviors An SEM model was run to evaluate specific paths between maltreatment types, attachment styles, and risk behaviors. Insignificant paths were trimmed from the final model, resulting in the removal of physical abuse (see Fig. 1). Verbal abuse was unrelated to either attachment style but was significantly, positively related to alcohol use (β = .20, p = .002) and antisocial behavior (β = .13, p = .027; see Table 2). Sexual abuse was positively related to anxious (β = .11, p = .013) and avoidant attachment (β = .19, p = .001), alcohol (β = .10, p = .035) and drug use (β = .08, p = .048), and negatively related to condom use (β = −.11, p = .011). Emotional abuse was significantly, positively related to anxious (β = .30, p = .001) and avoidant (β = .15, p = .010) attachment, antisocial behavior (β = .15, p = .012), and was negatively related to alcohol use (β = −.24, p = .001). Last, anxious attachment was positively related to alcohol use (β = .14, p = .001), and avoidant attachment was positively related to drug use (β = .09, p = .025) and antisocial behavior (β = .10, p = .028). Next, we tested and found that the indirect links from child maltreatment types to risk behaviors via both anxious and avoidant attachment styles were all significant ( Table 2). Model 1 with paths between child maltreatment, attachment, and risk behavior ... Fig. 1. Model 1 with paths between child maltreatment, attachment, and risk behavior observed variables. Note: Age and sex entered as covariates. Standardized betas presented. Figure options Table 2. Results and fit indices for model 1. Paths B (SE) β 95% CI Verbal abuse Alcohol use 0.15 (.05) 0.20 [0.057, 0.247]⁎⁎ Antisocial behavior 0.09 (.04) 0.13 [0.010, 0.169]⁎ Sexual abuse Anxious attachment 0.24 (.10) 0.11 [0.050, 0.436]⁎ Avoidant attachment 0.39 (.08) 0.19 [0.236, 0.551]⁎⁎ Alcohol use 0.13 (.06) 0.10 [0.009, 0.242]⁎ Drug use 0.16 (.08) 0.08 [0.001, 0.320]⁎ Condom use −0.20 (.08) −0.11 [−0.350, −0.046]⁎ Emotional abuse Anxious attachment 0.41 (.07) 0.30 [0.276, 0.543⁎⁎ Avoidant attachment 0.18 (.07) 0.15 [0.043, 0.319]⁎ Alcohol use −0.19 (.05) −0.24 [−0.283, −0.088]⁎⁎ Antisocial behavior 0.10 (.04) 0.15 [0.021, 0.172]⁎ Anxious attachment Alcohol use 0.08 (.02) 0.14 [0.033, 0.123]⁎⁎ Avoidant attachment Drug use 0.09 (.04) 0.09 [0.011, 0.171]⁎ Antisocial behavior 0.05 (.02) 0.10 [0.006, 0.100]⁎ Indirect effects Sexual abuse → Anxious attachment → Alcohol use 0.02 (.01) 0.02 [0.004, 0.046]⁎ Sexual abuse → Avoidant attachment → Drug use 0.04 (.02) 0.02 [0.006, 0.076]⁎ Sexual abuse → Avoidant attachment → Antisocial behavior 0.02 (.01) 0.02 [0.004, 0.042]⁎ Emotional abuse → Anxious attachment → Alcohol use 0.03 (.01) 0.04 [0.013, 0.058]⁎⁎ Emotional abuse → Avoidant attachment → Drug use 0.02 (.01) 0.01 [0.001, 0.044]⁎ Emotional abuse → Avoidant attachment → Antisocial behavior 0.01 (.01) 0.02 [0.001, 0.025]⁎ R2 Anxious attachment 0.131 Avoidant attachment 0.081 Alcohol use 0.075 Drug use 0.068 Condom use 0.029 Antisocial behavior 0.218 Fit indices CFI/TLI: .995/.940 RMSEA: .058 SRMR: .021 ⁎ p < .05. ⁎⁎ p < .01. Table options 3.2.2. Child maltreatment, emotion regulation, & risk behaviors An SEM model was run to evaluate specific paths between maltreatment types, ER dimensions, and risk behaviors (see Fig. 2). Insignificant paths were trimmed for the final model, resulting in the removal of physical abuse and four DERS subscales from the model: clarity, strategies, awareness, and nonacceptance. As shown in Table 3, verbal abuse was unrelated to the DERS subscales but was positively related to alcohol use (β = .21, p = .001). Sexual abuse was positively related to impulsivity (β = .17, p = .001). Emotional abuse was positively related to both impulsivity (β = .29, p = .001) and goals (β = .28, p = .001). Further, emotional abuse was negatively related to alcohol use (β = −.21, p = .002) and positively related to antisocial behavior (β = .21, p = .001). In relation to the ER dimensions, impulsivity was positively related to alcohol use (β = .17, p = .002), drug use (β = .22, p = .001), and antisocial behavior (β = .18, p = .001), and negatively related to condom use (β = −.22, p = .001). Last, goals was positively related to condom use (β = .13, p = .017). Significant indirect links from the child maltreatment types to risk behaviors via the DERS subscales of impulsivity and goals were found (see Table 3). Model 2 with paths between child maltreatment, emotion regulation, and risk ... Fig. 2. Model 2 with paths between child maltreatment, emotion regulation, and risk behavior observed variables. Note: Age and sex entered as covariates. Standardized betas presented. Figure options Table 3. Results and fit indices for model 2. Paths B (SE) β 95% CI Verbal abuse Alcohol use 0.17 (.05) 0.21 [0.070, 0.260]⁎⁎ Sexual abuse DERS impulsivity 0.23 (.07) 0.17 [0.088, 0.373]⁎⁎ Emotional abuse DERS impulsivity 0.24 (.05) 0.29 [0.147, 0.335]⁎⁎ DERS goals 0.32 (.05) 0.28 [0.217, 0.421]⁎⁎ Alcohol use −0.16 (.05) −0.21 [−0.259, −0.061]⁎⁎ Antisocial behavior 0.14 (.03) 0.21 [0.082, 0.200]⁎⁎ DERS impulsivity Alcohol use 0.15 (.05) 0.17 [0.057, 0.251]⁎⁎ Drug use 0.32 (.08) 0.22 [0.168, 0.462]⁎⁎ Condom use −0.30 (.07) −0.22 [−0.432, −0.157]⁎⁎ Antisocial behavior 0.14 (.04) 0.18 [0.064, 0.221]⁎⁎ DERS goals Condom use 0.13 (.05) 0.13 [0.023, 0.232]⁎ Indirect effects Sexual abuse → DERS impulsivity → Alcohol use 0.04 (.02) .03 [0.011, 0.080]⁎⁎ Sexual abuse → DERS impulsivity → Drug use 0.07 (.03) .04 [0.027, 0.154]⁎⁎ Sexual abuse → DERS impulsivity → Condom use −0.07 (.03) −.04 [−0.141, −0.023]⁎⁎ Sexual abuse → DERS impulsivity → Antisocial behavior 0.03 (.01) .03 [0.011, 0.072]⁎⁎ Emotional abuse → DERS impulsivity → Alcohol use 0.04 (.01) .05 [0.012, 0.069]⁎⁎ Emotional abuse → DERS impulsivity → Drug use 0.08 (.03) .06 [0.034, 0.137]⁎⁎ Emotional abuse → DERS impulsivity → Condom use −0.07 (.02) −.06 [−0.124, −0.035]⁎⁎ Emotional abuse → DERS impulsivity → Antisocial behavior 0.03 (.01) .05 [0.014, 0.064]⁎⁎ Emotional abuse → DERS goals → Condom use 0.04 (.02) .04 [0.008, 0.083]⁎ R2 DERS impulsivity 0.151 DERS goals 0.086 Alcohol use 0.074 Drug use 0.094 Condom use 0.056 Antisocial behavior 0.227 Fit indices CFI/TLI: 0.989/0.973 RMSEA: 0.036 SRMR: 0.033 Note: DERS – Difficulties in Emotion Regulation Scale. ⁎ p < .05. ⁎⁎ p < .01.