Emotion reactivity, nonsuicidal self-injury, and regulatory emotional self-efficacy: A moderated mediation model of suicide ideation

Highlights

• The female adolescents reported higher prevalence of suicide ideation than male adolescents.

• Emotion reactivity was positively associated with an increased risk of suicide ideation.

• The effect of emotion reactivity on suicide ideation was mediated by nonsuicidal self-injury.

• Regulatory emotional self-efficacy had a buffer effect on the relationship of emotion reactivity to suicide ideation and that of emotion reactivity to nonsuicidal self-injury.

Abstract

Background

Adolescent suicide ideation (SI) has received increased attention in recent years, and emotion reactivity (ER) has been found to be a risk factor for SI. However, little is known about the mediating and moderating mechanisms underlying this relationship. The aim of this study is to test a moderated mediation model of SI.

Method

A total of 2,716 Chinese adolescents (53.2% males, mean age = 13.19 years, SD = 0.52) completed self-report questionnaires regarding SI, ER, nonsuicidal self-injury (NSSI), and regulatory emotional self-efficacy (RESE).

Results

After controlling for gender, ER was significantly associated with SI, and this association was mediated by NSSI. RESE buffered the relation between ER and SI, as well as the relation between ER and NSSI.

Limitation

This study was cross-sectional in design and relied upon self-report measures only. Almost all participants were recruited from a single city of China.

Conclusion

Findings of this study demonstrate a relationship among ER, NSSI, and SI, and expand our understanding of the development of SI. Further, RESE may be a protective factor against SI and NSSI in adolescence, which suggests a potential role of promoting RESE in the prevention and intervention of SI and NSSI.

Introduction

Almost 800, 000 people die by suicide each year across the world (World Health Organization, 2018). The suicide death among adolescents was highlighted in the report of World Health Organization (2018). In China, adolescent suicide ideation (SI) has also become an increasingly alarming public health problem. SI, defined as plans and wishes to commit suicide but without any overt suicide attempt (Beck et al., 1979), is regarded as an essential phase of suicide attempt or even suicide completion (Deykin and Buka, 1994). Around 16.1% of Chinese adolescents reported at least one episode of SI during the past 12 months (Guo et al., 2018). A better understanding of the risk factors and related mechanisms for SI is thus in need to enlighten prevention and intervention efforts. Of the many factors that contribute to the emergence of SI, emotion reactivity (ER) has been one of the most extensively examined risk factors (Nock et al., 2008c). In this study, we focus on the pathway from ER to SI, and the potential mechanisms underlying this pathway.

ER, referring to the sensitivity to and the intensity and the persistence of negative emotional experiences, is conceptualized as a component of temperament that affects why and how an individual responds to emotional stimuli (Nock et al., 2008c). Empirical research has provided extensive support for the role of ER in SI. Both cross-sectional and longitudinal studies have reported that ER (Polanco-Roman et al., 2018), or similar constructs, such as trait negative affect intensity (Lynch et al., 2004) and emotion vulnerability (Rajappa et al., 2012), significantly correlate with SI.

Several theories have considered ER as a risk factor for SI. For instance, the diathesis-stress model asserts that individual vulnerabilities confer an elevated risk for SI under the presence of negative life stressors (Schotte and Clum, 1987). One of the vulnerabilities was personality characteristics, such as ER (Rabinowitz et al., 2016). Similarly, in the integrated motivational-volitional model of suicide (O'Connor and Kirtley, 2018), ER is one of the diatheses to provide backdrops for the occurrence of SI. Moreover, the biosocial theory of borderline personality disorder underlines the key role of emotional vulnerability (similar to ER) on the emergence of negative and overflowing emotions, and dysfunctional behaviors, such as suicide, as maladaptive coping strategies to overwhelming negative affect (Linehan, 1993).

All these theoretical and empirical studies have pinpointed that ER contributes to the development of SI. The mediating (i.e., how ER relates to SI) and moderating mechanisms (i.e., when the relation is stronger or weaker) underlying the association, however, have remained largely unexplored. Confirming a risk-protective framework could be critical to advance our understanding of SI (Salami et al., 2015). The present study thus aims to investigate a conceptual model with one mediator, nonsuicidal self-injury (NSSI), and one moderator, regulatory emotional self-efficacy (RESE) in the relationship between ER and SI.

NSSI is the direct, deliberate, and socially unacceptable destruction of one's body tissues without suicidal intent (Nock, 2010). NSSI may act as a mediator in the relation between ER and SI, such that ER may precipitate a heightened risk for NSSI, and NSSI may also increase the risk of SI.

ER as a risk factor for NSSI has been proposed in many theoretical models. For example, the experiential avoidance model stipulates that individuals with high emotion intensity (one aspect of ER), difficulty regulating when aroused, and emotion regulation skill deficits tend to engage in NSSI when emotionally evocative events occur because NSSI functions as the actual means of escape from unwanted experiences (Chapman et al., 2006). Later, Nock (2010) identified high ER as a distal risk factor for NSSI in his integrated theoretical model of NSSI. ER induces intrapersonal vulnerability factors, such as high aversive emotions. NSSI, functioning as affect regulation (e.g., relieving stress, regaining control of mood), may be used to cope with negative aversive emotions triggered by stressful events. In addition to theoretical research, empirical studies also have supported the positive association between ER and NSSI (Glenn et al., 2011; Kleiman et al., 2014).

Longitudinal studies as well as meta-analytic studies have revealed that NSSI is an important predicting factor for SI (Hamza and Willoughby, 2016; Maciejewski et al., 2014; Zhang et al., 2017). Adolescents with NSSI history are vulnerable to experiencing feelings of thwarted belongingness and perceived burdensomeness (Chu et al., 2016), which are two necessary conditions for SI in the interpersonal psychological theory of suicide (Van Orden et al., 2010). It was because NSSI, mostly performed in isolation, might elicit feelings of disconnection from others (Glenn and Klonsky, 2009). Besides, engagement in NSSI may trigger shame or guilt (Klonsky, 2009).

The majority of the research has been in line with the notion that ER is related with NSSI, and NSSI is related with SI. To date, no previous studies have directly examined the mediating role of NSSI in the relationship between ER and SI.

Although some personal characteristics as risk factors for SI have been recognized (Batterham and Christensen, 2012; Neufeld and O'Rourke, 2009), few studies have explored how to protect those with such characteristics from SI. Therefore, this study tests the protective role of RESE in the paths from ER to SI and from ER to NSSI.

RESE is described as an individual's confidence level and perceived ability in regulating emotions (Caprara et al., 1999). According to Bandura's (1997) self-efficacy theory, individuals have incentives to undertake activities only if they believe that they will be able to attain desired results. Thus, RESE may play a vital role in adolescents’ selection of and actual engagement in adaptive and effective emotion regulation behaviors (Bigman et al., 2016). Three lines of empirical research have supported the protective role of RESE against SI and NSSI. First, RESE was negatively associated with suicide-related behaviors, with high RESE correlating with low SI in adolescents (Kwok and Shek, 2010; Valois et al., 2015). Second, RESE was linked to low levels of risk factors of SI and NSSI, such as depression (Caprara et al., 2010) and anxiety (Niditch and Varela, 2012). Third, RESE protected individuals from maladaptive coping strategies, such as internet addiction and smoking (Pan et al., 2014; Yuan et al., 2018). It is thus plausible that RESE acts as a protective factor in the relations between ER and NSSI and between ER and SI.

Although previous studies have established preliminary associations among ER, NSSI, RESE, and SI, they have been limited in several areas.

First, only a hand of studies has examined the correlation between ER and SI in China, even fewer of which considered mediators or moderators. Cultural differences may exist between Chinese and Western adolescents, for instance, physiological process in stress-induced cortisol response differed between Western children and Chinese children (Grabell et al., 2015). Chinese children showed higher degrees of emotion inhibition compared with their U.S. peers, and emotion inhibition was verified as a risk factor for self-injurious thoughts and behaviors (Najmi et al., 2007). It is of great significance to examine how SI is affected by ER, NSSI, and RESE among Chinese adolescence.

Second, fewer studies have centered on protective factors than risk factors. Protective factors played a pivotal role in decreasing SI (Johnson et al., 2011). RESE has been proved as a protective factor in some dysfunctional behaviors (e.g., smoking, internet addiction), but not in SI or NSSI. Hence, this study aims at protecting high-risk adolescents from SI in consideration of RESE.

Adolescence is a time of significant increases in ER due to puberty-induced changes in brain structure and function, and shifting social contexts as well (Casey et al., 2010; Yap et al., 2007; Zimmermann and Iwanski, 2014). In this study, we investigate a moderated mediation model among a large sample of Chinese adolescents to understand the relationship among ER, NSSI, RESE, and SI. Based on previous research, we put forward three hypotheses: (1) ER would positively relate to SI; (2) NSSI would mediate the association between ER and SI; and (2) RESE would moderate the mediation model, such that a higher level of RESE would mitigate the relationship of ER to SI and that of ER to NSSI.