Child Abuse & Neglect 44 (2015) 130–139

Contents lists available at ScienceDirect

Child Abuse & Neglect

Research article

Emotional abuse in childhood and suicidality: The mediating roles of re-victimization and depressive symptoms in adulthood Min-Ah Lee Department of Sociology, Chung-Ang University, 84 Heukseok-ro Dongjak-Gu, Seoul 156-756, Republic of Korea

a r t i c l e

i n f o

Article history: Received 13 December 2014 Received in revised form 18 March 2015 Accepted 26 March 2015 Available online 16 April 2015 Keywords: Emotional abuse in childhood Re-victimization Childhood abuse Depressive symptoms Suicidality Mediation Korea

a b s t r a c t This study examined the effects of emotional abuse in childhood, along with physical and sexual abuse, on suicidality in adulthood, and whether and how emotional abuse and depressive symptoms in adulthood mediate the association between the childhood emotional abuse and suicidality. The data were drawn from the 2012 Korean General Social Survey, a nationally representative survey with a multistage area proportional probability sampling method. Random effects regression and the Sobel test were used to analyze the relationships between childhood emotional abuse and suicidality and the mediating effects. Random effects models showed that emotional abuse in childhood was positively associated with suicidality in adulthood, even after controlling for physical and sexual abuse in childhood. Emotional abuse and depressive symptoms in adulthood mediated the association between emotional abuse in childhood and suicidality. Depressive symptoms also mediated between emotional abuse in adulthood and suicidality. These findings suggest that emotional abuse in childhood has indirect harmful effects on suicidality in adulthood. It increases suicidality through higher occurrences of re-victimization and depressive symptoms in adulthood. Practitioners and policy makers should recognize that experiences of emotional abuse in childhood may result in re-victimization in adulthood, which, in turn, lead to suicidality. Early intervention programs to reduce the likelihood of experiencing re-victimization may be critical for people exposed to emotional abuse in childhood. © 2015 Elsevier Ltd. All rights reserved.

Introduction Childhood abuse has been identified as an important predictor of suicidality, including suicidal ideation and behaviors (e.g., Dube et al., 2001), as well as mental health outcomes such as depressive symptoms and anxiety in adulthood (Crow, Cross, Powers, & Bradley, 2014; see Lindert et al., 2014 for a review). Childhood abuse such as physical, emotional, and sexual abuse is positively associated with suicidality in adulthood (Dube et al., 2001; Jeon et al., 2009; Friestad, Ase-Bente, & Kjelsberg, 2014; Harford, Yi, & Grant, 2014) as well as prior to adulthood (Sigfusdottir, Asgeirsdottir, Gudjonsson, & Sigurdsson, 2013). The previous studies have suggested that the deleterious effects of childhood abuse can be sustained across decades of life. The long-term influence of childhood abuse on suicidality can be explained by the stress process paradigm (Pearlin, 1989) and life course perspectives (Dannefer, 1987, 2003; O’Rand, 1996, 2003). Childhood adversities, including highly stressful life events, may have deleterious effects that do not diminish with time, but are rather sustained through the life course by modifying and affecting developmental and psychological processes, which in turn may encourage psychological, physical, http://dx.doi.org/10.1016/j.chiabu.2015.03.016 0145-2134/© 2015 Elsevier Ltd. All rights reserved.

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

131

and/or behavioral manifestations of stress (Pearlin, 1989; Pearlin, Schieman, Fazio, & Meersman, 2005). Moreover, childhood adversity can lead to exposure to additional adversity later in life, with the effects of childhood adversity thereby sustained and accumulated over time (Dannefer, 1987, 2003; O’Rand, 1996, 2003; Pearlin et al., 2005). It is probable that childhood abuse leads to additional adversity such as interpersonal difficulties, decreased ability to form social attachments, and even re-victimization when abuse causes harmful effects on the developmental and psychological processes of exposed children. Thus, childhood abuse can be a lifelong predictor of suicidality. However, questions remain regarding the relationship between childhood abuse and suicidality. First, relatively little is known about the effects of emotional abuse in childhood on suicidality in adulthood and underlying mechanisms. Compared with physical and sexual abuse (e.g., Goldney, Grande, & Taylor, 2009; Joiner et al., 2007; Martin, Bergen, Richardson, Roeger, & Allison, 2004; Ystgaard, Hestetum, Loeb, & Mehlum, 2004), researchers have paid relatively little attention to emotional abuse (Crow et al., 2014), perhaps because emotional abuse has been regarded as less harmful and critical than physical or sexual abuse. This distinction is supported by studies showing nonsignificant effects of emotional abuse on suicidality across different life stages (Kwok, Chai, & He, 2013; Ystgaard et al., 2004). The mechanism between emotional abuse in childhood and suicidality has not been systematically examined. The pathway through which emotional abuse in childhood affects suicidality in adulthood should be explored. Childhood abuse may have long-term effects by modifying and affecting the psychological and developmental processes of exposed persons, and may lead to additional adversity over the life course. Investigating and identifying mediating factors between childhood emotional abuse and suicidality in adulthood would help understand how childhood emotional abuse can have long-lasting and far-reaching effects. Childhood abuse increases the likelihood of mental health problems in adulthood by damaging emotion regulation (Crow et al., 2014) and self-esteem (Stein, Leslie, & Nyamati, 2002) and increasing startle reactivity (Jovanovic et al., 2009), which may lead to problems in social functioning. Because childhood abuse is not only harmful to self-esteem (Stein et al., 2002), but also to the developmental processes of learning to trust and form stable relationships with others (Dimitrova et al., 2010), people who are exposed to childhood abuse may experience additional adversity regarding social relationships in adulthood. In this context, this study examines two potential mediators between emotional abuse in childhood and suicidality in adulthood: emotional abuse in adulthood (i.e., re-victimization) and depressive symptoms. Because childhood abuse can have harmful effects on social functioning as well as psychological processes, individuals exposed to emotional abuse in childhood may be more likely to be emotionally abused in adulthood as well. Although some studies have explored revictimization in adulthood in terms of physical and sexual abuse (Alexander, 2009; Coid et al., 2001; Irwin, 1999; Stein et al., 2002) and emotional abuse (Chiu et al., 2013), there have been, to the best of the author’s knowledge, no studies examining the re-victimization of victims of childhood emotional abuse and its mediating effects on suicidality. Depressive symptoms may also be an important mediator as a psychological outcome affected by childhood abuse (Brodsky & Stanley, 2001). If childhood abuse leaves indelible imprints on the psychological processes of exposed children, it likely also has a significant effect on suicidality through psychological status. Although depressive symptoms have been identified as an important predictor of suicidality (Chabrol, Rodgers, & Rousseau, 2007) and also as an outcome affected by childhood emotional abuse (Crow et al., 2014), the role of depressive symptoms as a mediator between childhood emotional abuse and suicidality remains unclear. Childhood sexual abuse has been shown to have significant effects on both depressive symptoms and suicidal ideation (Goldney et al., 2009), but the mediating role of depressive symptoms has not been examined. In contrast to Western societies, child abuse has relatively recently been recognized as a social problem in Korea. The Korean government ratified the Convention on the Rights of the Child in 1991 (Kim, 2007). Child abuse, mainly physical abuse by parents, has been publicly reported in Korea since the mid-1990s, and the first law prohibiting child abuse, the Special Law for Family Violence Criminal Prohibition and the Protection of the Victim Act, was established in 1998 (Hahm & Guterman, 2001). The Korean Child Welfare Act (article 2), which defines child abuse as doing “harm on a child’s health or welfare or physical, mental and sexual violence or cruel acts that are likely to impede normal growth of a child” including abandonment and nonintervention, was established in the year 2000. Due to the influence of Confucian family values, which give parents higher prestige regarding child discipline (Hahm & Guterman, 2001), physical punishment by parents has been often considered to be appropriate discipline for improving child behaviors, rather than abuse (Yang, 2009). Interestingly, parents are not the only perpetrators of childhood abuse in Korea. Abuse by teachers and peers are also frequently observed in Korea (Kim, Kim, & Park, 2000; Lee & Kim, 2011). According to Confucian values, teachers operate similarly to parents in terms of child discipline, leading to relatively high prevalence of physical and emotional abuse by teachers in Korea. Physical and emotional abuse by older male schoolmates and peers also often occur in Korean society (Lee & Kim, 2011), which might be associated with the age-based hierarchical nature of Korean culture. This study examines the effects of emotional abuse in childhood on suicidality in adulthood, and whether and how emotional abuse and depressive symptoms in adulthood mediate between childhood emotional abuse and suicidality. Although child abuse is now openly discussed more than ever in Korea, the discussion has focused on what are perceived to be more severe forms of child abuse, such as sexual abuse and physical abuse, rather than emotional abuse (Ju & Lee, 2010). However, emotional abuse in childhood has been shown to have critical effects on suicidality among Korean medical students, more so than physical or sexual abuse (Jeon et al., 2009). Little is known about whether emotional abuse in childhood has significant

132

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

effects among the Korean population. Because emotional abuse in childhood often occurs alongside physical or sexual abuse (Chiu et al., 2013), this study also control for physical and sexual abuse. Suicide has been one of the most serious social problems in Korea since the Asian financial crisis of 1997. According to a report by Statistics Korea (2014), the suicide rate among Koreans climbed from 14.6 per 100,000 in 2000 to 28.4 in 2009, and 31.7 in 2011. Korea’s suicide rate ranked first among OECD countries as of 2011 (OECD, 2014b). The average agestandardized suicide rate per 100,000 people of OECD countries was 12.4 in 2011 (OECD, 2014b). As of 2011 or the nearest year, age-standardized suicide rates per 100,000 people were 20.9 for Japan, 12.5 for the United States, 11.1 for Canada, and 3.1 for Greece (OECD, 2014b). Gender differences in suicide rates of Koreans are also observed, with suicide rates being 43.3 for men and 20.1 for women in 2011 (Statistics Korea, 2014). In the context of this lack of study regarding emotional abuse in childhood and suicidality in Korea and the fact that suicide rates in Korea are high, the findings of this study contribute to our understanding of the relationship between childhood emotional abuse and suicidality and also outline social and practical implications that may be helpful in decreasing suicide rates in Korea. It is important to examine whether emotional abuse in childhood has long-term effects on suicidality in Korean society, in which childhood abuse has been considered to be less serious than in Western societies. Methods Data Data were drawn from the 2012 Korean General Social Survey (KGSS). The KGSS is a nationally representative crosssectional survey that is conducted once per year. The sampling method (i.e., the multistage area proportional probability sampling method), interview protocols, and data-processing procedures used for the KGSS conform to those used for the General Social Survey (GSS) in the United States. Similar to the GSS, the KGSS includes special sets of questions every year in addition to core items such as questions about socio-demographic factors. The 2012 KGSS included questions about suicidal thinking, suicide attempts, and suicidal behavior for the assessment of suicidality. In addition, it uniquely included questions assessing childhood abuse and depressive symptoms. Measures Suicidality. Suicidality was measured using the standardized Korean version of the suicidality module in the Mini International Neuropsychiatric Interview (MINI) (Agoub, Moussaoui, & Kadri, 2006; Yoo et al., 2006), and the severity of suicidality was determined by summing points assigned to positive answers to the following six questions: In the past month, did you (1) (2) (3) (4) (5) (6)

think you would be better off dead or wish you were dead? (1 point) want to harm yourself? (2 points) think about suicide? (6 points) have a suicide plan? (10 points) attempt suicide? (10 points) In your life, did you ever attempt suicide? (4 points)

The higher the value, the higher the suicide risk. The logged variable after adding a value of 1 was used in analytic models for skewness. Emotional, physical, and sexual abuse in childhood. Childhood abuse was measured using the Korean version of the Early Trauma Inventory Self Report-Short Form (ETISR-SF) for the assessment of childhood abuse occurring by age 18 (Bremner, Vermetten, & Mazure, 2000). The Korean version of the ETSR-SF is a valid and reliable measure for Koreans (Jeon et al., 2012a) and is used for assessing abusive experiences among Koreans (Jeon et al., 2012b, 2009). Emotional abuse in childhood was measured by 5 items as follows: (1) often put down or ridiculed; (2) often ignored; (3) often told that one is no good; (4) treated in an uncaring way most of the time; (5) parents or caregivers failed to understand needs. The other two types of childhood abuse, physical and sexual abuse, were also measured. Physical abuse was measured with 5 items: (1) slapped in the face; (2) burned with hot water or cigarette as a punishment; (3) punched or kicked; (4) hit with thrown objects as a punishment; (5) pushed or shoved as a punishment. Sexual abuse was measured with only one item: forced/coerced to touch intimate parts, kiss, or perform sexual acts. Responses to all questions about childhood abuse were dichotomous (yes or no). Values for total numbers of emotionally and physically abusive experiences were used in this study. Sexual abuse was measured as a dummy variable. Emotional abuse in adulthood and depressive symptoms. Emotional abuse in adulthood as an indicator of re-victimization was measured using the same 5 items used for emotional abuse in childhood. These measures indicated the frequency of emotionally abusive events that respondents had experienced at the time of the survey. Depressive symptoms were measured using the Patient Health Questionnaire-9 (PHQ-9), which includes questions about 9 major depressive symptoms occurring in the past two weeks (Kroenke, Spitzer, & Williams, 2001). The PHQ-9 has been proven to be a valid and reliable diagnostic measure of depressive symptoms among Koreans (Choi et al., 2007; Han et al., 2008). The respondents rated

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

133

each symptom on a four-point scale from 1 (not at all) to 4 (nearly every day), and the higher the score, the more severe the depressive symptoms. The 9 symptoms are: (1) trouble falling or staying asleep, or sleeping too much; (2) feeling tired or having little energy; (3) poor appetite or overeating; (4) little interest or pleasure in doing things; (5) feeling down, depressed, or hopeless; (6) feeling that you are a failure or have let yourself or your family down; (7) trouble concentrating on things, such as reading the newspaper or watching television; (8) moving or speaking so slowly that other people notice, or the opposite - being so fidgety or restless that you have been moving around a lot more than usual; and (9) thoughts that you would be better off dead, or of hurting yourself. The reliability of the PHQ-9 was 0.838 (Cronbach’s ˛). Socio-demographic variables. Because socio-demographic factors such as household income and marital status are also important predictors of suicidality (Moon & Park, 2011), it is critical to examine the effects of emotional abuse in childhood and the two potential mediating factors after controlling for socio-demographic characteristics. The socio-demographic variables included gender (female = 1), age, educational attainment, monthly household income, and marital status. Age was measured in years, and household income was logged because it was skewed. Educational attainment was classified into three categories: less than high school; high school graduates; college or more. The reference group for educational attainment in the analytic models was respondents whose educational attainment was less than high school. Marital status was measured by asking the respondents whether they were married, widowed, separated/divorced, or never married. The reference group of marital status in the analytic models was married. Analytic strategy Random effects regression analysis was used to estimate parameters given that sampling for the KGSS was based on the multistage area proportional probability method. Random effects models are appropriate for clustered data because they do not assume that each observation is independent (Hedeker, Gibbons, & Faly, 1994). Four analytic models were tested to examine the effects of emotional abuse in childhood along with childhood physical and sexual abuse by adding other covariates in hierarchical order. Model 1 was a basic model including socio-demographic factors only. Model 2 included the three forms of childhood abuse (i.e., emotional, physical, and sexual abuse) to analyze the unique effects of emotional abuse along with those of physical and sexual abuse. Model 3 included emotional abuse in adulthood to examine whether it had a mediating effect on the relationship between emotional abuse in childhood and suicidality. Finally, depressive symptoms were added in Model 4 to examine their mediating effects. This set of random effects regression models was designed following Baron and Kenny’s (1986) criteria for mediation. For significant mediation, the mediator should have a significant effect on the dependent variable, and the significant effect of the predictor on the dependent variable also should substantially decrease or become nonsignificant after controlling for the mediating factor. In addition, a Sobel test (Baron & Kenny, 1986) was conducted to determine whether emotional abuse in adulthood and depressive symptoms had significant mediating effects on the association between childhood emotional abuse and suicidality. The Z score for the Sobel test was measured as Zab =



ab b2 SEa2

+ a2 SEb2

where a and b are unstandardized coefficients of an independent variable on a mediating factor and the mediating factor on the dependent variable (i.e., suicidality), respectively. SE indicates the standard error for each factor. Results Sample characteristics and t-test Table 1 presents the means/proportions and standard deviations for all variables and Table 2 summarizes t-test results comparing the means/proportions of variables depending on whether or not respondents were exposed to emotional abuse in childhood. For the t-tests reported in Table 2, respondents were divided into two groups: respondents with at least one emotional abuse event in childhood and those without any abusive experience. The mean of suicidality was significantly greater for respondents who had experienced emotional abuse than for those who had not. Among socio-demographic factors, the proportion of females was higher among respondents who had not experienced emotional abuse in childhood, suggesting that women were less likely to experience emotional abuse in childhood than men. The proportion of respondents whose educational attainment was less than high school was higher among respondents who had not experienced emotional abuse in childhood, whereas the proportion of respondents who had attended college or beyond was higher among respondents who had experienced emotional abuse in childhood. The mean household income was higher for respondents who reported at least one event of emotional abuse in childhood, which might be related to age given that the mean age of respondents who had experienced emotional abuse in childhood was lower than those who had not in Table 2. In fact, younger generations in Korea are more educated than older generations. The proportions of Koreans with more education (i.e., college or more) were 52% among those aged 35-44, 29% among those aged 45–54, and 14% among those aged 55–64 in 2012 (OECD, 2014a). The average income of Korean adults also decreased after about 47

134

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

Table 1 Descriptive statistics for all variables. N

Mean/proportion

Standard deviation

Suicidality (logged) Gender (Female = 1) Age

1,396 1,396 1,396

0.460 0.558 50.616

0.871 – 18.203

Educational attainment Less than high school High school graduate College or more Household income (logged)

1,392 1,392 1,392 1,392

0.338 0.291 0.371 5.260

– – – 1.297

Marital status Married Widowed Divorced/separated Never married

1,392 1,392 1,392 1,392

0.597 0.147 0.053 0.203

– – – –

Emotional abuse in childhood Physical abuse in childhood Sexual abuse in childhood Emotional abuse in adulthood Depressive symptoms

1,396 1,396 1,396 1,396 1,396

0.709 0.910 0.059 0.292 13.645

1.243 1.327 – 0.787 4.709

years of age as of 2009 (Park, 2011). In Table 2, the proportion of never married respondents was higher in the group who had not experienced emotional abuse in childhood. Regarding physical and sexual abuse in childhood, the mean of physically abusive events was higher among respondents who had experienced at least one event of emotional abuse than those who had not experienced abuse. Similarly, the proportion of respondents who experienced sexual abuse was higher among respondents who had experienced emotional abuse in childhood, suggesting that emotional abuse may co-occur with other forms of childhood abuse (Chiu et al., 2013). The mean of emotional abuse in adulthood was significantly higher for those who had experienced emotional abuse in childhood than for those without any abusive experiences. This suggests that respondents who were exposed to emotional abuse by the age of 18 years were more likely to experience emotional abuse in adulthood. The mean of depressive symptoms was also significantly higher for those who had experienced emotional abuse in childhood.

Table 2 Descriptive statistics for variables and t-tests comparing means/proportions of variables based on occurrence of emotional abuse in childhood. No emotional abuse in childhood (N = 923)a Mean/proportion Suicidality (logged) Gender (Female = 1) Age Educational attainment Less than high school High school graduate College or more Household income (logged) Marital status Married Widowed Divorced/separated Never married Emotional abuse in childhood Physical abuse in childhood Sexual abuse in childhood Emotional abuse in adulthood Depressive symptoms

Standard deviation

Had emotional abuse in childhood (N = 473)a Mean/proportion

t-Test

Standard deviation

0.335 0.586 52.763

0.762 – 18.402

0.704 0.503 46.427

1.010 – 17.066

***

0.380 0.288 0.332 5.201

– – – 1.313

0.256 0.297 0.447 5.377

– – – 1.258

***

0.623 0.164 0.049 0.165

– – – –

0.547 0.113 0.062 0.279

– – – –

**

0.455 0.022 0.105 12.908

0.969 – 0.460 4.449

2.093 1.797 0.131 0.655 15.082

1.289 1.476 – 1.104 4.873

Note: Standard deviations of binary variables are excluded. a The number of cases may vary due to missing data. * p < 0.05. ** p < 0.01. *** p < 0.001.

** ***

*** *

*

***

*** *** *** ***

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

135

Table 3 Bivariate correlations between childhood abuse, re-victimization of emotional abuse, depressive symptoms, and suicidality in adulthood. Suicidality Suicidality Emotional abuse in childhood Physical abuse in childhood Sexual abuse in childhood Emotional abuse in adulthood Depressive symptoms

Emotional abuse in childhood

Physical abuse in childhood

Sexual abuse in childhood

Emotional abuse in adulthood

1.000 0.523*** 0.235*** 0.428*** 0.236***

1.000 0.203*** 0.233*** 0.208***

1.000 0.198*** 0.117***

1.000 0.324***

1.000 0.232*** 0.205*** 0.131*** 0.287*** 0.435***

Depressive symptoms

1.000

Note: Sexual abuse in childhood is a binary variable. *** p < 0.001.

Childhood emotional abuse and suicidality: Examining potential mediating factors Table 3 presents bivariate correlations between the three childhood abuse variables, emotional abuse in adulthood, depressive symptoms, and suicidality. Emotional abuse variables, along with physical and sexual abuse, had significantly positive correlations with suicidality. Depressive symptoms also were significantly and positively correlated with suicidality. The correlation between emotional abuse in childhood and emotional abuse in adulthood indicates that respondents who were exposed to emotional abuse in childhood were more likely to report emotional abuse in adulthood. The two emotional abuse variables were also significantly correlated with depressive symptoms. Table 4 presents the results of random effects regression models examining the effect of emotional abuse in childhood on suicidality. Model 1 shows the results of a basic model including socio-demographic factors only. Models 2 to 4 added childhood abuse, the re-victimization of emotional abuse, and depressive symptoms in the hierarchical order. In Model 2, all three forms of childhood abuse were significantly associated with suicidality after controlling for socio-demographic factors. As the frequency of emotional or physical abuse in childhood increased, the level of suicidality increased. Respondents who experienced sexual abuse in childhood had a higher level of suicidality than their counterparts. In Model 3, the effect of emotional abuse in childhood remained significant, but its coefficient substantially decreased when emotional abuse in adulthood was included. This suggests that emotional abuse in adulthood has a partial mediating effect between emotional abuse in childhood and suicidality. The coefficient of sexual abuse decreased and became marginally significant after controlling for emotional abuse in adulthood, whereas the coefficient of physical abuse did not

Table 4 Random effects analyses of suicidality. Model 1

Model 2

Model 3

Model 4

Gender (Female = 1) Age Educational attainment High school graduate College or more Household income (logged) Marital status Widowed Divorced/separated Never married Emotional abuse in childhood Physical abuse in childhood Sexual abuse in childhood Emotional abuse in adulthood Depressive symptoms Constant

−0.091+ (0.049) −0.005* (0.002)

−0.009 (0.052) −0.002 (0.002)

−0.015 (0.051) −0.002 (0.002)

−0.092+ (0.048) −0.002 (0.002)

−0.117 (0.071) −0.226** (0.079) −0.059* (0.023)

−0.129+ (0.069) −0.250** (0.076) −0.058* (0.022)

−0.120+ (0.068) −0.242** (0.075) −0.054* (0.022)

−0.023 (0.064) −0.128+ (0.071) −0.038+ (0.021)

0.021 (0.081) 0.315** (0.106) 0.018 (0.082)

0.007 (0.079) 0.290** (0.103) 0.031 (0.079) 0.104*** (0.021) 0.083*** (0.021) 0.281** (0.100)

0.009 (0.077) 0.246* (0.101) 0.033 (0.078) 0.045* (0.022) 0.081*** (0.021) 0.194+ (0.099) 0.234*** (0.031)

1.171*** (0.220)

0.837*** (0.216)

0.799*** (0.212)

−0.016 (0.073) 0.182+ (0.095) 0.002 (0.073) 0.033 (0.021) 0.036+ (0.020) 0.162+ (0.093) 0.134*** (0.030) 0.066*** (0.005) −0.137 (0.211)

N Chi-squared Rho

1384 35.78*** 0.037

1384 138.4*** 0.027

1384 201.19*** 0.030

1384 411.02*** 0.041

R-squared Within Between Total

0.022 0.074 0.025

0.073 0.153 0.095

0.108 0.170 0.132

0.216 0.242 0.236

Note: Numbers in parentheses are standard errors. + p < 0.1. * p < 0.05. ** p < 0.01. *** p < 0.001.

136

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

Table 5 Results of Sobel tests for the mediating effects of emotional abuse and depressive symptoms in adulthood. Mediating effects of emotional abuse in adulthood and depressive symptoms Emotional abuse in childhood → emotional abuse in adulthood → suicidality Physical abuse in childhood → emotional abuse in adulthood → suicidality Sexual abuse in childhood → emotional abuse in adulthood → suicidality Emotional abuse in childhood → depressive symptoms → suicidality Physical abuse in childhood → depressive symptoms → suicidality Sexual abuse in childhood → depressive symptoms → suicidality Emotional abuse in adulthood → depressive symptoms → suicidality + *

Z scores 4.231*** −1.042 3.007** 1.408+ 5.523*** 0.936 7.688***

p < 0.1. p < 0.05. ** p < 0.01. *** p < 0.001.

substantially change. Emotional abuse in adulthood significantly increased suicidality. As the frequency of emotional abuse in adulthood increased, the level of suicidality increased. In Model 4, depressive symptoms had a significant and positive effect on suicidality. Interestingly, the coefficients of both emotional abuse variables decreased when depressive symptoms were included in Model 4, and emotional abuse in childhood became nonsignificant. Emotional abuse in adulthood remained strongly significant, but its coefficient decreased substantially. This suggests that depressive symptoms have mediating effects between emotional abuse in childhood, emotional abuse in adulthood, and suicidality. It should be also noted that physical abuse became marginally significant when depressive symptoms were controlled, suggesting that depressive symptoms mediated the relationship between physical abuse in childhood and suicidality as well. Regarding other independent variables, in Model 1, age, college education or more, household income, and being divorced or separated were significant. Respondents whose educational attainment was college or more had a lower level of suicidality than those who were the least educated. As household income increased, the level of suicidality decreased. Age increased suicidality in Model 1. Age, however, became nonsignificant after controlling for three forms of childhood abuse in Model 2. All other socio-demographic factors that were significant in Model 1 became marginally significant after controlling for depressive symptoms in the final model. Sobel test: The mediating effects of depressive symptoms and emotional abuse in adulthood Because the coefficient of emotional abuse in adulthood substantially decreased after controlling for depressive symptoms in Model 4 of Table 4, a Sobel test for the mediating effect of depressive symptoms on emotional abuse in adulthood was conducted. In addition, because coefficients of physical and sexual in childhood also changed after controlling for emotional abuse in childhood and depressive symptoms, mediating effects associated with physical and sexual abuse in childhood were also tested. Table 5 presents the results of the Sobel test to determine whether emotional abuse in adulthood and depressive symptoms had mediating effects for suicidality. Emotional abuse in childhood increased the likelihood of emotional abuse in adulthood, and emotional abuse in adulthood increased suicidality. Interestingly, emotional abuse in adulthood also had an indirect effect on suicidality through depressive symptoms. Emotional abuse in childhood had a marginally significant indirect effect on suicidality through depressive symptoms. Physical and sexual abuse in childhood also had indirect effects on suicidality. Physical abuse in childhood increased depressive symptoms, which ultimately led to suicidality, while sexual abuse in childhood increased suicidality through increased emotional abuse in adulthood. Discussion Emotional abuse in childhood was significantly associated with suicidality in adulthood. As the frequency of emotional abuse that respondents experienced before age 18 increased, the level of suicidality increased. The coefficient and significance of emotional abuse, however, substantially decreased after controlling for emotional abuse in adulthood. Emotional abuse in childhood appears to have an indirect effect on suicidality through emotional abuse in adulthood. This implies that emotional abuse in childhood has long-term effects on suicidality through re-victimization. Depressive symptoms, as well as emotional abuse in adulthood, had important mediating effects shaping suicidality. Depressive symptoms had a marginally significant mediating effect between emotional abuse in childhood and suicidality, and also significantly mediated between emotional abuse in adulthood and suicidality. This suggests that individuals who were exposed to emotional abuse in childhood have a higher likelihood of being exposed to emotional abuse in adulthood and a higher level of depressive symptoms, which in turn are associated with suicidality. These findings have identified a pathway through which emotional abuse in childhood results in suicidality in adulthood, and also underscore important implications regarding childhood emotional abuse.

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

137

Emotional abuse in childhood may lead to re-victimization in adulthood. Some studies have shown that people who have experienced physical and sexual abuse are more likely to be exposed to the same types of abuse in adulthood (Alexander, 2009; Chiu et al., 2013; Coid et al., 2001; Irwin, 1999; Stein et al., 2002). However, the re-victimization of individuals who have experienced emotional abuse and its effect on suicidality has not been previously investigated. The findings of this study emphasize the importance of re-victimization of emotional abuse and its harmful effects on suicidality. Why are people who were exposed to childhood emotional abuse more likely to experience emotional abuse in adulthood? Emotional abuse in childhood may affect the quality of relationships with family members and others. Because childhood abuse is harmful not only to self-esteem (Stein et al., 2002), but also to the developmental processes of learning to trust and form stable relationships with others (Dimitrova et al., 2010), people who have experienced emotional abuse in childhood have lower quality relationships in mid-life and later life (Savla et al., 2013). If emotional abuse in childhood is harmful to the development of social skills that are important for shaping the quality of interpersonal relationships, emotionally abused children are vulnerable to emotional abuse in adulthood. According to the life course perspective, exposure to childhood adversity leads to exposure to additional adversity in later life, which is how the effects of childhood adversity can be sustained and accumulated over time (Dannefer, 1987, 2003; O’Rand, 1996, 2003; Pearlin et al., 2005). Emotional abuse re-victimization, specifically as an additional form of adversity faced by people who have experienced emotional abuse in childhood, can have a significant and deleterious effect on suicidality. In Korea, the school as well as home environments are places where childhood abuse occurs. Physical and emotional abuse by teachers and peers are frequently observed in Korea (Kim et al., 2000; Lee & Kim, 2011). Emotional abuse by teachers or peers may have sustained effects on exposed people in terms of overall social functioning, because it can negatively affect social relationships with schoolmates by increasing discrimination and negative views on exposed people. The indirect effect of sexual abuse in childhood observed in the results of this study can be explained in terms of similar mechanisms. As it has been observed that sexual abuse results in lower levels of interpersonal closeness in relationships in adulthood (Dimitrova et al., 2010), sexual abuse in childhood may be harmful to social relationships throughout life. People exposed to sexual abuse may have difficulty with trust and forming stable relationships, and therefore be vulnerable to emotional abuse in adulthood. Both sexual and emotional abuse in childhood may have deleterious effects on the ability to trust and feel affection for others, leading to lower quality social relationships and functioning. Depressive symptoms also mediated the effects of emotional abuse and physical abuse in childhood on suicidality. A few studies have shown associations between emotional abuse in childhood and depressive symptoms (Courtney, Kushwaha, & Johnson, 2008; Crow et al., 2014) as well as between physical abuse in childhood and depressive symptoms (see Lindert et al., 2014 for a review) across different life stages. Childhood abuse may increase mental health problems in adulthood by damaging emotion regulation (Crow et al., 2014) and self-esteem (Stein et al., 2002) and increasing startle reactivity (Jovanovic et al., 2009). Consequently, the increased depressive symptoms associated with emotional and physical abuse in childhood may lead to increased risk of suicide in adulthood. Interestingly, depressive symptoms also exhibited a significant mediating effect between emotional abuse in adulthood and suicidality. People who were exposed to emotional abuse in adulthood had a higher level of depressive symptoms, which in turn increased suicidality. It is noteworthy that age became nonsignificant after controlling for the three forms of childhood abuse in Model 2 of Table 4, which suggests that reports of childhood abuse might vary by age. This is also consistent with the statistics summarized in Table 2, which showed that the mean age was higher among respondents who did not experience emotional abuse in childhood than those with at least one event of emotional abuse. Because child abuse has only been relatively recently recognized as a social problem in Korea, there might be generational differences in the perception of childhood abuse. This might be related to the fact that cultural and perceptional acceptance of certain forms of physical and emotional abuse may lead to a lower likelihood of reporting because people do not recognize whether they had such experiences (Lee & Kim, 2011). It is also possible that older people are less likely to recall their experiences in childhood due to cognitive aging. Supplementary analyses testing interaction effects between each form of childhood abuse and age have been conducted to examine whether childhood abuse has differential effects according to age. The interaction terms, however, were not significant (data not shown), suggesting that childhood abuse had similar effects across all ages. Some limitations of this study should be mentioned. First, because this study analyzed a cross-sectional data set, the possibility of reverse causation exists. Although childhood emotional abuse precedes other factors such as suicidality and depressive symptoms in adulthood, it is also possible that respondents who had higher levels of depressive symptoms or suicidality overestimated their childhood abuse experiences. Second, physical and sexual abuse in childhood still had marginally significant effects on suicidality in the final model shown in Table 4, suggesting that there might be other mediating factors underlying these effects. The present study, however, did not examine other potential mediating factors such as substance abuse due to the limitation of the data. Further research should be conducted to investigate and identify other potential mediating factors between childhood abuse and suicidality. In addition, although the purpose of the study was investigating the re-victimization of emotional abuse in adulthood, the re-victimization of other forms of abuse could not be considered because the KGSS survey did not include questions about physical and sexual abuse in adulthood. Whether sexual and physical abuse in childhood are associated with revictimization of the same types of abuse in adulthood in terms of associations with suicidality should be investigated. Finally, it is reasonable to include abusive experiences not only at the hands of parents, but also caused by others such as teachers or peers, in the Korean context. It would be, however, necessary to examine and compare the effects of childhood abuse according to the main perpetrators. The KGSS, unfortunately, did not include data specifying main perpetrators.

138

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

Despite these limitations, the results of this study indicate that social and academic attention should be paid to emotional abuse by demonstrating the substantive effects of childhood emotional abuse on suicidality in adulthood, as previous studies have mainly focused on physical and sexual abuse (e.g., Goldney et al., 2009; Joiner et al., 2007; Martin et al., 2004; Ystgaard et al., 2004). Although childhood abuse has only relatively recently been recognized as a social problem in Korea, these results show that emotional abuse as well as physical and sexual abuse have harmful long-term effects for Koreans for all generations. The results of this study also have important implications on the associations between emotional abuse in childhood and suicidality by identifying the mediating effects of emotional abuse in adulthood and depressive symptoms. Practitioners and policy makers should recognize that the experience of emotional abuse in childhood may result in re-victimization in adulthood, eventually leading to negative outcomes in terms of mental health and suicidality. In Korea, social and political attention paid to emotional abuse and the implementation of social policies against emotional abuse have been limited. Korean policy makers should recognize emotional abuse as a problem, and provide intervention programs designed to reduce the likelihood of re-victimization among those exposed to emotional abuse in childhood, because re-victimization is associated with depressive symptoms and suicidality. References Agoub, M., Moussaoui, D., & Kadri, N. (2006). Assessment of suicidality in a Moroccan Metropolitan Area. Journal of Affective Disorders, 90, 223–226. Alexander, P. (2009). Childhood trauma, attachment, and abuse by multiple partners. Psychological Trauma: Theory, Research, Practice, and Policy, 1(1), 78–88. 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. Bremner, J. D., Vermettern, E., & Mazure, C. M. (2000). Development and preliminary psychometric properties of an instrument for the measurement of childhood trauma: The early trauma inventory. Depression and Anxiety, 12, 1–12. Brodsky, B. S., & Stanley, B. (2001). Developmental effects on suicidal behavior: The role of abuse in childhood. Clinical Neuroscience Research, 1, 331–336. Chabrol, H., Rodgers, R., & Rousseau, A. (2007). Relations between suicidal ideation and dimensions of depressive symptoms in high-school students. Journal of Adolescence, 30(4), 587–600. Chiu, G. R., Lutfey, K. E., Litman, H. J., Link, C. L., Hall, S. A., & McKinlay, J. B. (2013). Prevalence and overlap of childhood and adult physical, sexual, and emotional abuse: A descriptive analysis of results from the Boston Area Community Health (BACH) Survey. Violence and Victims, 28(3), 381–402. Choi, H., Choi, J., Park, K., Joo, K., Ga, H., Ko, H., & Kim, S. (2007). Standardization of the Korean Version of Patient Health Questionnaire-9 as a screening instrument for major depressive disorder (in Korean). Journal of the Korean Academy of Family Medicine, 28(2), 114–119. Coid, J., Petruckevitch, A., Feder, G., Chung, W., Richardson, J., & Moorey, S. (2001). Relation between childhood sexual and physical abuse and risk of revictimization in women: A cross-sectional survey. Lancet, 358(9280), 450–454. Courtney, E. A., Kushwaha, M., & Johson, J. G. (2008). Childhood emotional abuse and risk for hopelessness and depressive symptoms during adolescence. Journal of Emotional Abuse, 8, 281–298. Crow, T., Cross, D., Powers, A., & Bradley, B. (2014). Emotion dysregulation as a mediator between childhood emotional abuse and current depression in a low-income African–American sample. Child Abuse & Neglect, 38, 1590–1598. Dannefer, D. (1987). Aging as intracohort differentiation: Accentuation, the Matthew effect and the life course. Sociological Forum, 2, 211–236. Dannefer, D. (2003). Cumulative advantage/disadvantage and the life course: Cross-fertilizing age and the social science theory. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 58, 327–337. Dimitrova, N., Pierrehumbert, B., Glatz, N., Torrisi, R., Heinrichs, M., Halfon, O., & Chouchena, O. (2010). Closeness in relationships as a mediator between sexual abuse in childhood or adolescence and psychopathological outcome in adulthood. Clinical Psychology and Psychotherapy, 17, 183–195. Dube, S. R., Anda, R. F., Felitti, V. J., Chapman, D. P., Williamson, D. F., & Giles, W. H. (2001). Childhood abuse, household dysfunction and the risk of attempted suicide throughout the life span. JAMA, 286(24), 3089–3096. Friestad, C., Ase-Bente, R., & Kjelsberg, E. (2014). Adverse childhood experiences among women prisoners: Relationships to suicide attempts and drug abuse. International Journal of Social Psychiatry, 60(1), 40–46. Goldney, R. D., Grande, E. D., & Taylor, A. (2009). Population-attributable risk of childhood sexual abuse for symptoms of depression and suicidal ideation in adulthood. International Journal of Psychiatry in Clinical Practice, 13, 285–291. Hahm, H. C., & Guterman, N. B. (2001). The emerging problem of physical child abuse in South Korea. Child Maltreatment, 6, 169–179. Han, C., Jo, S., Kwak, J., Pae, C., Steffens, D., Jo, I., & Park, M. (2008). Validation of the Patient Health Questionnaire-9(PHQ-9) Korean version in the elderly population: The AGE study. Comprehensive Psychiatry, 49, 218–223. Harford, T. C., Yi, H., & Grant, B. (2014). Associations between childhood abuse and interpersonal aggression and suicide attempts among U.S. adults in a national study. Child Abuse & Neglect, 38(8), 1389–1398. Hedeker, D., Gibbons, R. D., & Flay, B. R. (1994). Random-effects regression models for clustered data with an example from smoking prevention research. Journal of Consulting and Clinical Psychology, 62(4), 757–765. Irwin, H. (1999). Violent and nonviolent revictimization of women abused in childhood. Journal of Interpersonal Violence, 14(10), 1095–1110. Jeon, H. J., Roh, M., Kim, K., Lee, J., Lee, D., Yoon, S., & Hahm, B. (2009). Early trauma and lifetime suicidal behavior in a nationwide sample of Korean medical students. Journal of Affective Disorders, 119, 210–214. Jeon, H. J., Kang, E. S., Lee, E. H., Jeong, E. G., Jeon, J. R., Mischoulon, D., & Lee, D. (2012). Childhood trauma and platelet brain-derived neurotrophic factor (BDNF) after a three month follow-up in patients with major depressive disorder. Journal of Psychiatric Research, 46(7), 966–972. Jeon, J. R., Lee, E. H., Lee, S. W., Jeong, E. G., Kim, J. H., Lee, D., & Jeon, H. J. (2012). The early trauma inventory self-report-short form: Psychometric properties of the Korean version. Psychiatry Investigation, 9, 229–235. Joiner, T. E., Jr., Sachs-Ericsson, N. J., Wingate, L. R., Brown, J. S., Anestis, M. D., & Selby, E. A. (2007). Childhood physical and sexual abuse and lifetime number of suicide attempts: A persistent and theoretically important relationship. Behaviour Research and Therapy, 45, 539–547. Jovanovic, T., Blanding, N. Q., Norrholm, S. D., Duncan, E., Bradley, B., & Ressler, K. J. (2009). Childhood abuse is associated with increased startle reactivity in adulthood. Depression and Anxiety, 26, 1018–1026. Ju, S., & Lee, Y. (2010). Experiences of family maltreatment by Korean children in Korean National Protective Services. Child Abuse & Neglect, 34, 18–27. Kim, D., Kim, K., & Park, Y. (2000). Children’s experience of violence in China and Korea: A transcultural study. Child Abuse & Neglect, 24(9), 1163–1173. Kim, S. (2007). The status of Children’s rights and policies for its promotion in Korea. Health-Welfare Policy Forum, 6, 21–36 (in Korean). Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. Kwok, S. Y. C. L., Chai, W., & He, X. (2013). Child abuse and suicidal ideation among adolescents in China. Child Abuse & Neglect, 37, 986–996. Lee, Y., & Kim, S. (2011). Childhood maltreatment in South Korea: Retrospective study. Child Abuse & Neglect, 35, 1037–1044. Lindert, J., von Ehrenstein, O. S., Grashow, R., Gal, G., Braehler, E., & Weisskopf, M. G. (2014). Sexual and physical abuse in childhood is associated with depression and anxiety over the life course: Systematic review and meta-analysis. International Journal of Public Health, 59, 359–372. Martin, G., Bergen, H., Richardson, A. S., Roeger, L., & Allison, S. (2004). Sexual abuse and suicidality: Gender differences in a large community sample of adolescents. Child Abuse & Neglect, 28, 491–503.

M.-A. Lee / Child Abuse & Neglect 44 (2015) 130–139

139

Moon, S., & Park, S. (2011). Risk factors for suicidal ideation in Korean middle-aged adults: The role of socio-demographic status. International Journal of Social Psychiatry, 58(6), 657–663. O’Rand, A. M. (1996). The precious and the precocious: Understanding cumulative disadvantage and cumulative advantage over the life course. The Gerontologist, 36, 230–238. O’Rand, A. M. (2003). Cumulative advantage theory in life course research. In S. Crystal, & D. F. Shea (Eds.), Annual review of gerontology and geriatrics: Focus on economic outcomes in later life (pp. 14–30). New York, NY: Springer. OECD. (2014a). Education at a glance 2014. OECD. Retrieved from http://www.oecd.org/edu/Education-at-a-Glance-2014.pdf. OECD. (2014b). Society at a glance 2014-OECD. OECD. Retrieved from http://www.oecd.org/els/soc/OECD2014-SocietyAtAGlance2014.pdf. Park, S. (2011). Income and consumption of midlife and older Koreans. Statistical Research Institute. Retrieved from http://kostat.go.kr/edu/sri kor new/1/1/index.action?bmode=read&p idx=259&pageNo=1&search mode=&keyword=%EC%86%8C%EB%93%9D (in Korean). Pearlin, L. I. (1989). The sociological study of stress. Journal of Health and Social Behavior, 30, 241–256. Pearlin, L. I., Schieman, S., Fazio, E. M., & Meersman, S. C. (2005). Stress, health, and the life course: Some conceptual perspectives. Journal of Health and Social Behavior, 46, 205–219. Savla, J. T., Roberto, K. A., Jaramillo-Sierra, A. L., Gambrel, L. E., Karimi, H., & Butner, L. M. (2013). Childhood abuse affects emotional closeness with family in mid- and later life. Child Abuse & Neglect, 37, 388–399. Sigfusdottir, I. D., Asgeirsdottir, B. B., Gudjonsson, G. H., & Sigurdsson, J. F. (2013). Suicidal ideations and attempts among adolescents subjected to childhood sexual abuse and family conflict/violence: The mediating role of anger and depressed mood. Journal of Adolescence, 36, 1227–1236. Statistics Korea. (2014). Cause of death statistics. Statistics Korea. Retrieved from http://kosis.kr/statisticsList/statisticsList 01List.jsp?vwcd=MT ZTITLE& parmTabId=M 01 01 (in Korean). Stein, J. A., Leslie, M. B., & Nyamathi, A. (2002). Relative contributions of parent substance use and childhood maltreatment to chronic homelessness, depression, and substance abuse problems among homeless women: Mediating roles of self-esteem and abuse in adulthood. Child Abuse & Neglect, 26, 1011–1027. Yang, S. (2009). Cane of love: Parental attitudes towards corporal punishment in Korea. British Journal of Social Work, 39, 1540–1555. Yoo, S. W., Kim, Y. S., Noh, J. S., Oh, K. S., Kim, C. H., Namkoong, K., Chae, J. H., Lee, G. C., Jeon, S. I., Min, K. J., Oh, D. J., Joo, E. J., Park, H. J., Choi, Y. H., & Kim, S. J. (2006). The validation study of the Korean version of mini-international neuropsychiatric interview. Anxiety and Mood, 2(1), 50–55. Ystgaard, M., Hestetun, I., Loeb, M., & Mehlum, L. (2004). Is there a specific relationship between childhood sexual and physical abuse and repeated suicidal behavior? Child Abuse & Neglect, 28, 863–875.