Elsevier

Journal of Affective Disorders

Volume 207, 1 January 2017, Pages 1-8
Journal of Affective Disorders

Research paper
The role of bullying in depressive symptoms from adolescence to emerging adulthood: A growth mixture model

https://doi.org/10.1016/j.jad.2016.09.007Get rights and content

Highlights

  • Bully/cyberbully victimization, not perpetration, predict depression trajectories.

  • Findings assist school personnel in identifying students' depression trajectories.

  • Four depression trajectories identified across adolescence and emerging adulthood.

Abstract

Background

The present study sought to identify trajectories of depressive symptoms in adolescence and emerging adulthood using a school-based sample of adolescents assessed over a five-year period. The study also examined whether bully and cyberbully victimization and perpetration significantly predicted depressive symptom trajectories.

Method

Data from a sample of 1042 high school students were examined. The sample had a mean age of 15.09 years (SD=.79), was 56.0% female, and was racially diverse: 31.4% Hispanic, 29.4% White, and 27.9% African American. Data were examined using growth mixture modeling.

Results

Four depressive symptoms trajectories were identified, including those with a mild trajectory of depressive symptoms, an increasing trajectory of depressive symptoms, an elevated trajectory of depressive symptoms, and a decreasing trajectory of depressive symptoms. Results indicated that bully victimization and cyberbully victimization differentially predicted depressive symptoms trajectories across adolescence, though bully and cyberbully perpetration did not.

Limitations

Limitations include reliance on self-reports of bully perpetration and a limited consideration of external factors that may impact the course of depression.

Conclusions

These findings may inform school personnel in identifying students' likely trajectory of depressive symptoms and determining where depression prevention and treatment services may be needed.

Introduction

The World Health Organization estimates that Major Depressive Disorder (MDD) and Dysthymia are among the leading sources of mental health burden worldwide (Üstün et al., 2004). The lifetime prevalence rate of MDD and Dysthymia in adolescence has been estimated at 11.7% by age 18 years (Merikangas et al., 2010) and as many as 29.9% of high school students report feeling so sad or hopeless almost every day for two or more weeks in the past year that they stopped doing some usual activities (Centers for Disease Control and Prevention, 2014). These data demonstrate the substantial unmet mental health burden of depression symptoms and disorders in youths – and additional data demonstrate the high impact of adolescent depressive symptoms on academic, social, and physical functioning (e.g., Jaycox et al., 2009).

Adolescent-onset depressive symptoms are associated with a high rate of recurrence and may be indicative of a chronic course (Harrington and Dubicka, 2001, Dunn and Goodyer, 2006). Even so, depressive symptoms show considerable heterogeneity over time during adolescence and emerging adulthood, as evidenced in several recent studies (e.g., Brendgen et al., 2005; Hill et al., 2014; Rodriguez et al., 2005; Stoolmiller et al., 2005; Yaroslavsky et al., 2013). Yaroslavsky et al. (2013) identified three depressive symptom trajectories among adolescents, in which some adolescents reported elevated symptoms from mid-adolescence through age 30 years, while others reported moderate or mild symptoms that decreased across adolescence and emerging adulthood. Stoolmiller et al. (2005) identified four trajectories of depressive symptoms across adolescence and emerging adulthood, in which some adolescents reported elevated symptoms, others very few symptoms and some reported moderate or elevated symptoms that decreased over time. These studies identify significant variability in depressive symptoms across adolescence, with some adolescents demonstrating stable symptoms over time but others showing changing symptom trajectories. Previous work has identified a history of psychopathology, parental psychopathology, gender, and social variables as predictors of depressive symptom trajectories (Stoolmiller et al., 2005, Yaroslavsky et al., 2013). Despite these findings, additional work is needed to better understand factors associated with the heterogeneity of depressive symptoms across adolescence and emerging adulthood.

A more thorough understanding of depressive symptom trajectories in adolescent and emerging adult populations, as well as the factors that predict those trajectories, can assist mental health service providers in identifying adolescents at risk for increased depressive symptoms or major depressive episodes. The ability to predict likely symptom trajectories a priori would afford mental health service providers an opportunity to implement preventive interventions more efficiently. For example, identifying adolescents whose symptoms are likely to escalate over time may afford providers the opportunity to intervene early and provide prevention services closer to the onset of symptom course (e.g., Hill et al., 2014). Further, identifying which adolescents are most likely to maintain their depressive symptoms over time, as opposed to showing a more transitory symptom pattern, may allow service providers to selectively deliver more intensive prevention programs to adolescents in greatest need of them (Hill et al., 2015).

Increasingly, high schools are being utilized as a point of intervention for mental health services and often provide services directly (Wells et al., 2003). As such, school counselors and other school-based mental health professionals have an enormous opportunity to impact student mental health via early identification and treatment. School counselors and personnel are actively involved with students, are aware of social dynamics within the school setting, and therefore may be particularly adept at identifying adolescents at risk for depression. Identifying factors relevant to school personnel as well as their relation to adolescent depressive symptom trajectories is therefore critical to optimizing school-based intervention efforts.

A growing body of research suggests that various forms of bullying and victimization are associated with adolescent depressive symptoms both cross-sectionally (Klomek et al., 2007, Seals and Young, 2003) and longitudinally (Kaltiala-Heino et al., 2000, Kaltiala-Heino et al., 2010). Bullying is defined as “any unwanted aggressive behavior(s) by another youth or group of youths…that involves an observed or perceived power imbalance and is repeated multiple times or is highly likely to be repeated (Gladden et al., 2014: pp. 7).” Craig and Pepler (2003) consider bullying as a relational problem between perpetrators and vulnerable victims. Both bully victimization and perpetration have been linked with elevated depressive symptoms in high school students (e.g., Klomek et al., 2007, Klomek et al., 2013a Hong et al., 2014), though some studies have failed to identify a link between bully perpetration and depressive symptoms (i.e., Perren et al., 2010).

Aside from traditional forms of bullying, such as physical and verbal bullying, cyberbullying has also been linked to depressive symptoms (Wang et al., 2011). Cyberbullying refers to aggressive, intentional acts using electronic forms of contact, such as messages sent via text, email, or social media, repeatedly and over time, against a victim who cannot easily defend him or herself (Smith et al., 2008; Völlink et al., 2013). Though cyberbullying may occur less frequently than traditional bullying, it appears to have similar adverse effects as traditional forms of bullying (Smith et al., 2008). Evidence demonstrating a link between cyberbullying and depressive symptoms has emerged in recent years, with a study by Perren et al. (2010) providing evidence that cyberbully victimization may predict depressive symptoms even after controlling for other forms of bully victimization. Evidence also indicates that individuals who are both cyberbully victims and perpetrators report greater depressive symptoms than those who are either perpetrators or victims alone (Kaltiala-Heino et al., 2000). Thus, cyberbullying may have especially profound effects on adolescents as compared to more traditional forms of bullying (Smith et al., 2008, Völlink et al., 2013).

With the increased focus on the impacts of bullying in schools, knowledge of the impact of various forms of bully victimization and perpetration on depressive symptom trajectories may help school personnel better identify at-risk individuals and provide prevention and intervention services more efficiently. To our knowledge, the potential impact of bully victimization and perpetration on depressive symptom trajectories has not yet been evaluated in the empirical literature. Existing studies provide evidence of overall linear associations between depressive symptoms bully victimization and perpetration and depressive symptoms, but have not examined potential subgroups of adolescents who may be differentially impacted by bully victimization and perpetration, while examining symptom trajectories over extended longitudinal follow-ups. A more nuanced understanding of the impact of bully victimization and perpetration on changes in depressive symptoms over time may be useful for informing clinicians and researchers in providing and developing efficacious prevention services. For example, the ability to predict which adolescents with elevated depressive symptoms are likely to have consistently elevated symptoms over time, as opposed to time-limited elevations in symptoms, would allow clinicians to direct more intensive treatments to those whose symptoms are likely to be maintained. Alternatively, if it is possible to identify adolescents with minimal depressive symptoms but whose symptoms are likely to increase, clinicians could direct those adolescents to preventive interventions. Finally, examination of the differential impact of types of bullying on depressive symptom trajectories may direct researchers to the most salient forms of bullying for the development of targeted prevention efforts.

The present study sought to identify trajectories of depressive symptoms in adolescents and emerging adults using a school-based sample of adolescents assessed over a five-year period. Consistent with previous research, multiple trajectories were expected (e.g., Stoolmiller et al., 2005; Yaroslavsky et al., 2013). The present study then sought to assess whether bully victimization and perpetration – both in-person and cyberbullying – significantly predicted depressive symptom trajectories. Finally, given the known relations between hostility, gender, race, and depressive symptoms (Felsten, 1996, Nolen-Hoeksema et al., 1999, Saluja et al., 2004), these factors were included as covariates in all analyses. Both victims and perpetrators of bullying and cyberbullying express greater hostility than children not involved in bullying (Ireland and Archer, 2004, Völlink et al., 2013). In addition, hostility has been linked to depressive symptoms (e.g., Mao et al., 2003) and thus hostility represents a potential confounding variable. Thus it was important to control for potential impacts of hostility on depressive symptom trajectories. Moreover, covarying for hostility (i.e., trait anger), helps differentiate between the adolescents' affective state (anger) and bullying behaviors. Utilizing covariates emphasizes the unique variance accounted for by bullying-related variables in the analyses.

Section snippets

Participants and procedures

Participants were 1042 adolescents from multiple public high schools serving a large diverse metropolitan region (response rate 62%; the generally accepted response rate is 60% (Johnson and Wislar, 2012)). A majority of participants were 9th and 10th graders (75.0% and 24.0%, respectively). The sample had a mean age of 15.09 years (SD=.79), was 56.0% female, and identified their race/ethnicity as Hispanic (31.4%), White (29.4%), African American (27.9%), Asian/Pacific Islander (3.6%) and

Results

Descriptive statistics for all study variables, as well as the correlations between them are provided in Table 1. Participants were categorized into groups based on the presence of any self-reported history of bully and cyberbully victimization and perpetration. Three mutually exclusive groups were defined, separately for bullying and cyberbullying: perpetrators, victims, and joint victim-perpetrators. Table 2 presents differences across bullying and cyberbully groups with regard to demographic

Discussion

The present study examined differences between bully and cyberbully victims, perpetrators, and joint victim-perpetrators. This study then identified trajectories of depressive symptoms in adolescence and emerging adulthood using a school-based sample of adolescents assessed over a five-year period. The study also examined whether bully and cyberbully victimization and perpetration significantly predicted depressive symptom trajectories. Consistent with previous research, multiple trajectories

Conflicts of interest

The authors have no conflicts of interest to report.

Funding source

This research was supported by Award Number K23HD059916 (PI: Temple) from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD) and 2012-WG-BX-0005 (PI: Temple) from the National Institute of Justice (NIJ). The content is solely the responsibility of the authors and does not necessarily represent the official views of NICHD or NIJ.

References (45)

  • K.L. Bradley et al.

    Factorial validity of the center for epidemiological studies depression 10 in adolescents

    Issues Ment. Health Nurs.

    (2010)
  • M. Brendgen et al.

    Relations with parents and with peers, temperament, and trajectories of depressed mood during early adolescence

    J. Abnorm. Child Psychol.

    (2005)
  • Centers for disease control and prevention, 2014. Youth risk behavior surveillance – United States, 2013. Morbidity and...
  • L.M. Collins et al.

    A comparison of inclusive and restrictive strategies in modern missing data procedures

    Psychol. Methods

    (2001)
  • W.M. Craig et al.

    Identifying and targeting risk for involvement in bullying and victimization

    Can. J. Psychiatry

    (2003)
  • L.R. Derogatis et al.

    Symptom checklist‐90–revised

    Corsini Encycl. Psychol.

    (2010)
  • V. Dunn et al.

    Longitudinal investigation into childhood-and adolescence-onset depression: psychiatric outcome in early adulthood

    Br. J. Psychiatry

    (2006)
  • C.K. Enders

    Analyzing longitudinal data with missing values

    Rehabil. Psychol.

    (2011)
  • X. Ge et al.

    Pubertal transition, stressful life events, and the emergence of gender differences in adolescent depressive symptoms

    Dev. Psychol.

    (2001)
  • Gladden, R.M., Vivolo-Kantor, A.M., Hamburger, M.E., Lumpkin, C.D., 2014. Bullying surveillance among youths: Uniform...
  • R. Harrington et al.

    Natural history of mood disorders in children and adolescents

    Depress. Child Adolesc.

    (2001)
  • J.S. Hong et al.

    Pathways from bullying perpetration, victimization, and bully victimization to suicidality among school-aged youth a review of the potential mediators and a call for further investigation

    Trauma, Violence, Abus.

    (2014)
  • Cited by (30)

    • The mental health correlates of cybervictimisation against ethnic minority young people: A systematic review

      2023, Aggression and Violent Behavior
      Citation Excerpt :

      More specifically, cybervictimised ethnic minority young people, including young people of Asian descent were likely to show higher levels of anxiety and depression compared to young people of European descent (Shapka & Maghsoudi, 2017). Previous research has suggested that both these ethnic groups face many stereotypes, with Asian Americans being viewed as “perpetually foreigner”, i.e., not truly Americans (Lee, Wong, & Alvarez, 2009), and Latinx people being viewed as “criminals” (Reny & Manzano, 2016). Belonging to an ethnic group that has been traditionally stigmatised may constitute an additional risk factor for developing mental health problems among cybervictimised adolescents.

    • Vicious cycle of emotional maltreatment and bullying perpetration/victimization among early adolescents: Depressive symptoms as a mediator

      2021, Social Science and Medicine
      Citation Excerpt :

      According to diathesis-stress models (Swearer and Hymel, 2015), bullying involvement is a stressful life event for both individuals who bully and who are victimized, serving as a catalyst for a diathesis-stress connection between bullying perpetration, bullying victimization, and psychosocial problems. A growing body of research indicates that different forms of bullying perpetration/victimization are associated with adolescents' depressive symptoms both cross-sectionally (e.g., Casper and Card, 2017; Humphreys et al., 2020) and longitudinally (e.g., Da Silva et al., 2020), although some studies have failed to find a link between bullying perpetration and depressive symptoms (i.e., Hemphill et al., 2011; Hill et al., 2017). Moreover, children's mental health problems represent a risk factor for child maltreatment in the family context.

    • Risk and resource factors for depressive symptoms during adolescence and emerging adulthood – A 5-year follow-up using population-based data of the BELLA study

      2021, Journal of Affective Disorders
      Citation Excerpt :

      By means of the variables included in our multivariate regression analyses, a proportion of 22.0 % (girls) and 16.0 % (boys) of the variance in depressive symptoms was explained. These findings may indicate that the development of depressive symptoms is associated with important factors that we did not consider in our model, which may include genetic risks (Kwong et al., 2019), environmental aspects (Lehtinen et al., 2003), the exposure to physical and sexual abuse (Gonzalez et al., 2012) or the experience of being bullied (Hill et al., 2017). Future studies may take these aspects into account.

    • Being a cybervictim and a cyberbully – The duality of cyberbullying: A meta-analysis

      2020, Computers in Human Behavior
      Citation Excerpt :

      The obtained effect size was moderate-high (r = 0.428) (Cohen, 1988) and significant (p<0.01). All the data were highly significant (p<0.01), although only one study (Hill et al., 2017) fell within the confidence intervals. Two groups were formed: The first comprised those countries that did not reach the minimum value (mainly Central European countries, along with two Mediterranean and two Asian countries).

    View all citing articles on Scopus
    View full text