Emotion Regulation, Coping and Alcohol Use as Moderators in the Relationship Between Non-Suicidal Self-Injury and Psychological Distress
- First Online:
- Cite this article as:
- Williams, F. & Hasking, P. Prev Sci (2010) 11: 33. doi:10.1007/s11121-009-0147-8
Non-suicidal self-injury is a risk factor for more severe self-injury and later suicide, yet is relatively under-researched in non-clinical populations. In order to prevent more severe self-injury and later suicide, understanding of non-suicidal self-injury is imperative. This study aimed to examine whether coping skills, emotion regulation and alcohol use moderate the relationship between psychological distress and non-suicidal self-injury. Two hundred eighty-nine young adults completed self-report questionnaires assessing the variables of interest. Of the sample, 47.4% reported a history of non-suicidal self-injury. Adaptive coping strategies protected those who were psychologically distressed from severe self-injury. However for those who reported greater distress, this protective effect was negated by heavy alcohol use. Coping skills training may serve to protect young people from self-injury, although those who are severely distressed may also benefit from strategies to limit alcohol use.
KeywordsSelf-injuryCopingEmotion regulationPsychological distress
Non-suicidal self-injury (NSSI), the deliberate destruction or alteration of one’s body tissue without conscious suicidal intent (Favazza 1996), is a risk factor for future self-injury and completed suicide (eg. Nock et al. 2006). Most of our knowledge about self-injury is based on research in clinical samples; however, many individuals who engage in NSSI do so in private and rarely come to the attention of health services (Murray et al. 2005). Consequently, NSSI may be more prevalent than commonly recognised (Gratz and Chapman 2007). Due to the potential for NSSI to lead to suicide in some individuals (Nock et al. 2006; RANZCP 2004), further understanding of the correlates and risk factors for NSSI within community populations is crucial to informing effective prevention and early intervention strategies.
Despite increasing evidence of the need to make a distinction between suicidal and non-suicidal self-injury (Groholt et al. 2000), continued definitional confusion undermines efforts to further research the nature, extent and correlates of NSSI. Despite the inconsistent nomenclature for self-injurious behaviour, recent research suggests the incidence of self-injury (with or without suicidal intent) among young people is increasing (Fortune and Hawton 2005). Studies of young adults report that between 8.6% (Andover et al. 2007) and 44% of this population may engage in NSSI (Hasking et al. 2008; Paivio and Mcculloch 2004). These high rates of NSSI in community samples emphasise the emerging importance of understanding factors related to NSSI and developing effective prevention and early intervention programs. Such understanding is imperative to preventing more severe self-injury and later suicide.
Factors associated with NSSI in community samples are poorly understood. Yet, psychological distress (Patton et al. 1997), alcohol consumption (Poorasl et al. 2007), avoidant coping strategies (Evans et al. 2005), and emotional suppression and dysregulation (Brown et al. 2002; Zlotnick et al. 1996) are related to NSSI, at least among adolescents. Likewise, the rates of Axis I disorders in those who self-injure are two to four times higher than those who do not self-injure, and those of schizophrenia, bipolar disorder, eating disorder and substance dependence are six to ten times higher (Kessler et al. 1994; McLennan 1998). Accordingly, depression is commonly associated with NSSI in both clinical (Nixon et al. 2002; Nock et al. 2006) and community samples (Murray et al. 2005; Wong et al. 2005). Similar associations have also been found between anxiety symptoms and NSSI in community samples (Patton et al. 1997). These findings suggest that self-injury (with and without suicidal intent) may be utilised to cope with intense emotions and may be associated with deficits in emotional regulation (Kirkcaldy et al. 2004).
Emotional inexpressivity (Zlotnick et al. 1996), affect intensity/reactivity (Gratz et al. 2006) and emotion dysregulation (Haines and Williams 1997) have been identified as factors related to emotional responding in those who self-injure. Emotional regulation can involve two related processes: expressive suppression, in which emotion-expressive behavior is modified or inhibited and cognitive reappraisal, which involves construing a potentially emotional situation to change its emotional impact (Gross and John 2003). Although there is little research on the relationship between these emotion regulation processes and NSSI, cognitive reappraisal has been associated with better interpersonal functioning and well-being, in comparison to expressive suppression, which is related to poorer interpersonal functioning and well-being (Gross and John 2003). Given the potential for NSSI to relieve intense emotions, the way in which individuals who self-injure utilize these two emotion regulation processes warrants investigation. Likewise, avoidant coping strategies have been associated with NSSI in adolescents (Evans et al. 2005) and incarcerated young male adults (Haines and Williams 1997). Others have found individuals who self-injure report poor problem-solving skills (Hawton et al. 1999). Overall, NSSI appears to be a form of coping for those who lack more adaptive ways of dealing with intense emotions. However, few studies have examined other coping strategies utilised by those who self-injure.
Finally, alcohol use has also been associated with NSSI (Kerfoot 1996). Adolescent research in England has consistently found an association between alcohol consumption and self-injury (Hawton et al. 2002; Scott and Powell 1993). Data from hospital presentations suggest alcohol dependence is the second most common psychiatric diagnosis, after depression, among those who self-injure (Haw et al. 2001).
The study of NSSI has historically been atheoretical, as there is insufficient data to inform theory development. However, both risk and protective factors are likely to play a role in NSSI and in any subsequent escalation of suicidal behavior. It has been suggested that predisposing factors, such as a psychological distress, lead to internal distress and that NSSI is one way individuals with poor emotion regulation cope with this distress. Among those who self-injure, alcohol is recognised as a factor which exacerbates negative affect, and minimizes the ability to inhibit urges to self-injure (Sinclair and Green 2005). As such, consideration of coping skills, emotion regulation and alcohol use is important to furthering our understanding of NSSI. Although an established relationship has been demonstrated between psychopathology and NSSI, the way in which alcohol use, emotional regulation and coping strategies impact upon this association has not been investigated. This study examined the role these factors play in the relationship between psychological distress and NSSI. It was hypothesised that the relationship between psychological distress and NSSI will be stronger for those who utilize avoidant coping strategies, have poorer emotion regulation and report risky drinking behavior.
Recruitment and Procedure
Young adults were recruited by placing posters on university notice boards and in various businesses around Melbourne, Australia. The posters briefly outlined the purpose of the study (to examine the nature and extent of NSSI in young adults), emphasized that participation was voluntary (no incentives were provided for participation) and anonymous and invited interested participants to contact the researchers for a copy of the questionnaire. Posters explicitly invited both those who had a history of self-injury and those with no history of self-injury to participate. Interested participants provided their postal address so that researchers could mail an information sheet, the questionnaire, an information pack which provided information about depression, stress, alcohol use and self-injury, and appropriate counselling resources and a reply paid envelope. The information sheet included a definition of NSSI, the objectives of the study, right to withdraw and contact details of the researchers. Participants were asked to complete the questionnaire in their own time and return anonymously in the reply paid envelope. Consequently, it must be acknowledged when interpreting the results that the sample was self-selected and not a true probability sample. Ethical approval was obtained from the Monash University Human Research Ethics Committee before commencement of the study.
Of the 650 questionnaires distributed to potential participants, 289 were completed and returned (44.46%), by 78 male and 211 female young adults aged between 18 and 30 years (mean = 22.52 years, SD = 3.92). The majority of young adults were currently studying (72.3%). Most participants were born in Australia (77.5%), 40.8% lived with both parents, and 45% reported a previous diagnosis of a mental illness. The most frequently reported diagnosis was depression (80%).
Participants were asked to indicate the nature and extent of NSSI. To ensure that participants understood the questions did not refer to suicidal self injury, the instructions began with “Sometimes people hurt themselves on purpose, without trying to kill themselves. Please read the following questions and tick the response that best matches your own actions. Please answer the questions about whether you have hurt yourself without trying to kill yourself.” Participants were asked to indicate if they had ever engaged in deliberate cutting, burning, severe scratching, wound interference and self-nominated ways of self-injuring. In addition, participants were asked to indicate the frequency, location (on the body), recency, and severity of each form of self-injury. Usual frequency of self-injury and recency were assessed using a six-point scale (0 = never, to 5 = daily/today), while severity was assessed by asking participants to consider how serious their self-injury usually was (not at all serious, requiring first aid, requiring medical attention, or life threatening). This measure has been administered to over 850 young adults and adolescents and appears to be a reliable indicator of NSSI in these samples (see Hasking et al. 2008).
Emotional Regulation Scale (Gross and John 2003)
The emotional regulation scale is a ten-item questionnaire designed to assess individual differences in two emotion regulation strategies: expressive suppression and cognitive reappraisal. Items are measured on a seven-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree). This scale exhibits solid reliability, with alpha reliabilities averaging 0.79 for reappraisal and 0.73 for the suppression dimension. Test–retest reliability across 3 months was 0.69 for both scales (Gross and John 2003). Cronbach’s alpha in this sample were: cognitive reappraisal = 0.80; emotional suppression = 0.76.
COPE (Carver et al. 1989)
The COPE is a 60-item inventory that assesses the ways people respond to stress. Items are rated on a four-point Likert scale ranging from 1 (‘I usually don’t do this at all’) to 4 (‘I usually do this a lot’). These items are summated to provide 15 subscales that form three higher-order factors: problem focused coping, emotion focused coping and avoidant coping (Lyne and Roger 2000). Confirmatory factor analysis has revealed this scale to be a reliable measure of coping in community samples (Hasking and Oei 2002). In addition, test–retest reliability, ranging from 0.46 to 0.86 and high convergent and discriminant validity with personality measures have been reported (Carver et al. 1989). In the present sample the Cronbach’s alphas were: problem focused = 0.86; emotion focused = 0.90; avoidant = 0.79.
Brief Symptom Inventory (BSI; Derogatis and Melisaratos 1983)
The BSI is a 53-item measure of psychopathology on which participants are asked to indicate how much a problem has distressed them in the past 7 days. Items are rated on a five-point Likert scale ranging from 0 (‘not at all’) to 4 (‘extremely’). The BSI has demonstrated high internal reliability, with the subscales ranging from 0.71 to 0.85, high test–retest reliability of the GSI of 0.90 and high convergent validity (Derogatis and Melisaratos 1983). In order to assess the relationship between general psychological distress and NSSI, the Global Severity Index (GSI), a measure of overall psychological distress, was utilized in this study. The Cronbach’s alpha in this sample for the GSI was 0.97.
Australian Alcohol Use Disorders Identification Test (AusAUDIT; Degenhardt et al. 2001)
The AusAUDIT is a ten-item scale designed to screen individuals for harmful or hazardous drinking behavior. Items reflect volume and frequency of alcohol consumption as well as consequences which might arise from drinking alcohol. Items are summated to provide an overall measure of drinking behavior, with higher scores indicating more harmful drinking. Good internal consistency (α = 0.76) and discriminant validity have been demonstrated (Degenhardt et al. 2001). The Cronbach’s alpha for this sample was 0.86.
The Nature and Extent of NSSI
Nature and extent of NSSI
Percentage of total sample reporting self-injury
Percentage of those reporting each form of self-injury
Yearly or less
Daily, weekly or monthly
More than a year ago
Within the last year
Today, last week or last month
Not at all or first aid
Medical or life threatening
For the purpose of analysis, scores regarding frequency, recency, and self-reported severity of NSSI for each form of NSSI were averaged and weighted by the number of methods utilized. This provided a continuous score with a range from 0 to 70. For example, a person who reported both cutting and burning, engaged in cutting within the last week and burning within the last month, reported a usual frequency of monthly for each behavior and inflicted injuries that required home first aid would receive a score of 6.8 (total score for cutting = 9 + total score for burning = 8 + 0 for all other methods, average score = 3.4, multiplied by number of methods used = 6.8). Our method of classification and the creation of a composite score have previously been used with this measure (Hasking et al. 2008, under review), and appear to discriminate between clinical and non-clinical samples (Anderson et al., unpublished manuscript).1
In this sample, scores ranged from 0 to 48 (total sample mean = 3.28, SD = 6.70). Scores for those who self-injured were normally distributed (mean = 6.91, SD = 8.85). These results are comparable to those obtained from an adolescent sample utilizing the same measure (Hasking et al., under review). In order to create categories reflecting the severity of NSSI, the bottom 25th percentile who self-injured were classified as exhibiting mild NSSI (n = 36), those between 25% and 75% were classified as exhibiting moderate NSSI (n = 69) and those in the top 25th percentile (n = 32) were classified as exhibiting severe NSSI. These categories provide an index of severity relative to other participants in this sample, rather than an objective assessment of severity due to the absence of clinical data and objective criteria for rating severity.
The four groups did not differ in age, F (3, 285) = 0.21, p = 0.89, gender, χ2 (3, N = 289) = 0.34, p = 0.95, or education, χ2 (3, N = 289) = 4.03, p = 0.26. However those who reported engaging in NSSI were more likely to report a previous diagnosis of a mental illness, χ2 (3, N = 289) = 43.73, p = 0.000, with the severe NSSI group more likely to report a history of depression (standardised residual = 4.5).
Descriptive data and correlations
2. Psychological distress
3. Alcohol consumption
4. Emotional reappraisal
5. Emotional suppression
6. Problem focused coping
7. Emotion focused coping
8. Avoidant coping
The relationships between psychological distress, coping, emotional regulation, drinking behavior and NSSI were examined by performing a hierarchical multiple regression, with the continuous NSSI score (described above) as the criterion. Previous diagnosis of a mental illness was statistically controlled by entering this in the first step of the regression equation. Psychological distress (GSI) was entered in the second step, while the unique effects of coping strategies, emotional regulation and drinking behavior were examined in the third step of the equation. In the fourth step, two-way interactions between psychological distress and each of the coping, emotional regulation and alcohol factors were examined. All three-way interactions between psychological distress, drinking behavior and each of the coping and emotion regulation factors were entered in the last step.
Hierarchical multiple regression predicting self-injury
Psychological distress (GSI)
Problem focussed coping
Emotion focused coping
GSI × emotion coping
GSI × avoidant coping
GSI × alcohol × problem
GSI × alcohol × emotion
GSI × alcohol × avoid
The present study examined the nature and extent of NSSI in young adults and the way in which emotional regulation, coping strategies and drinking behavior moderated the relationship between psychopathology and NSSI. Of the total sample, a high percentage reported a history of NSSI. Our results are similar to other studies of young adults in the community, where a lifetime prevalence between 41% and 44% was found (Hasking et al. 2008; Paivio and Mcculloch 2004), and studies of university students which report a prevalence of 44% (Gratz and Chapman 2007), but much higher than the 17% (Whitlock et al. 2006) and 8.6% (Andover et al. 2007) found in other samples of young adults. These findings may indicate an increasing prevalence of NSSI among young adults or may be due to the encompassing nature of the NSSI measure, and selection bias in the sample. Overall severity of NSSI was relatively low; thus, the NSSI observed in these studies cannot be compared to NSSI usually seen in clinical samples, and may not generalize to other community samples. However as NSSI is a recognized risk factor for more severe self-injury and later suicide (Nock et al. 2006), even mild NSSI in a self-selected sample cannot be ignored.
Emotion focused and avoidant coping were both directly related to NSSI and moderated the relationship between psychological distress and NSSI. Psychological distress was positively related to NSSI for young adults who did not rely on avoidant coping strategies; however, this relationship was stronger for those who did utilize avoidant coping. In addition, psychological distress was associated with NSSI for those who did not use emotion focused coping but not for those who did rely on emotion focused strategies.
While these two-way interactions appear to support previous work highlighting the role of avoidant and emotion-focused strategies in NSSI (Evans et al. 2005; Muehlenkamp and Gutierrez 2004, 2007), these variables were also observed to interact with drinking behavior, indicating the complex interplay between psychological distress, alcohol use and coping strategies. The role of risky drinking in NSSI is evident for individuals with greater levels of psychological distress. Notably, although problem-focused and a lack of avoidant coping appeared to be protective for distressed individuals who did not drink heavily, those who reported riskier drinking patterns reported similar levels of NSSI even if they utilized problem-focused strategies, or reported not relying on avoidant coping. A lack of emotion-focused coping was also associated with NSSI, particularly for those who reported greater alcohol use. Thus, the combination of psychological distress, poor coping skills and risky drinking poses a significant risk for NSSI.
It was interesting to note that although bivariate relationships between emotion regulation and NSSI were observed, emotion regulation was not related to NSSI in the regression model. NSSI is commonly believed to be a method of coping with intense emotion, especially among those who have a poor ability to regulate emotions (Gratz et al. 2002; Klonsky 2007; Linehan 1993). Specifically, emotional suppression has previously been related to NSSI in both adolescent and adult samples (Crowell et al. 2005; Hasking et al. 2008). However, in this sample coping strategies appear to be more salient predictors of NSSI. The failure to find a significant relationship between emotion regulation and NSSI may be due, in part, to the high correlation between emotional suppression and emotion focused coping observed in this study. Emotion regulation and coping are theoretically similar; thus, it is not surprising that a strong correlation between these variables was observed. However, future research would benefit from an attempt to disentangle these constructs in an effort to determine which is more salient in predicting NSSI, and which should be the focus of prevention and early intervention programs.
Our results support previous research suggesting that alcohol use is common among those who self-injure (e.g., Hawton et al. 2002; Scott and Powell 1993; Sinclair and Green 2005), and highlight the role of adaptive coping in protecting individuals from NSSI. Taken together, the findings suggest that screening for harmful alcohol use and maladaptive coping strategies in young adults with significant psychological distress may assist in identifying those at greatest risk of NSSI. Further, given the high rate of NSSI observed in this sample, routine screening from NSSI in mental health services is warranted.
It should be noted, however, that given the correlational nature of this study the causal relationships between the variables cannot be ascertained. Of note, it is not known whether alcohol use exacerbates negative affect and thus leads to NSSI as suggested by Sinclair and Green (2005), or whether alcohol is used to cope with psychological distress. It is equally likely that both alcohol use and NSSI are associated with poor impulse control. Likewise, the relationships between alcohol use and coping require further exploration. Future research would benefit from examination of the drinking motives among those who self-injure in order to determine whether alcohol is used as a coping strategy, or whether alcohol use is associated with a general pattern of high risk behavior. Similarly, illicit drug use may also be prevalent among those who engage in NSSI. Future research would benefit from an examination of illicit drug use as well as alcohol use.
Prevention and early intervention initiatives designed to minimize suicidal behaviour may benefit from a focus on coping skills training and efforts to minimize high-risk drinking behavior. The results of this study suggest coping skills training could provide a greater focus on the development of positive emotion-focused coping skills and ways in which to minimize avoidant coping. Adaptive coping strategies already play a role in prevention and treatment programs (e.g., Klingman and Hochdorf 1993); however, the high rate of NSSI observed in this, and other studies of young adults, call the efficacy of these programs into question. Early intervention and treatment programs need to be tailored to the individual due to the complex nature of NSSI; however, these findings suggest that alcohol use, maladaptive coping strategies and emotion related factors may be important targets for such interventions.
The few prospective studies, based on hospital data, which show a progression in severity of NSSI suggest a negative psychological trajectory for those who self-injure, but the form this path takes, and the role resilience plays in protecting against further NSSI is yet to be established. In order to prevent NSSI escalating to the point where it requires medical attention or hospitalization, and to prevent subsequent suicide, an understanding of these relationships in community samples is needed. Our results suggest that inclusion of coping skills and emotion regulation in such studies may begin to answer questions about the most effective way to prevent more severe suicidal behavior.
Although this study aimed to shed light on NSSI in community samples, and we statistically controlled previous psychiatric diagnosis, some participants may have been receiving outpatient treatment or counselling for self-injury or related mental health concerns, contributing to the high rate of NSSI observed in this study. However, as the majority of previous research has focussed upon clinical samples of adolescents, this study provides preliminary data concerning the nature of NSSI in a self-selected sample of young people. Although our advertisements explicitly invited participation from both individuals with, and without, a history of NSSI, those with a history of NSSI may have been more inclined to participate, limiting the external validity of the findings. However our observed rates were similar to other estimates using similar methodology (e.g., Gratz and Chapman 2007; Hasking et al. 2008). If we were to take the response rate into account, and assume that the non-responding participants did not have history of NSSI, the incidence of reported NSSI would fall to 21.08%, a rate still higher than typically reported in non-clinical samples. Recruitment from a broader, more representative sample is required to confirm the high rates of NSSI seen in this sample.
Our sample comprised predominantly female participants. Consideration of gender differences in NSSI, and its relationship to associated morbidity and coping is warranted in further research. Future research utilizing longitudinal designs is also necessary to determine the direction of the observed relationships. Specifically, psychological distress may precede NSSI, which may serve as a coping strategy. Alternatively, symptoms of psychological distress may be induced as a result of self-injurious behaviour. As previously noted, the relationship between alcohol use and NSSI also requires further exploration. Longitudinal studies are imperative to clarify these relationships and inform development of theoretically based prevention and early intervention initiatives. Finally, although a significant number of participants reported NSSI within the last year, the majority reported their last episode of NSSI was more than a year ago, necessitating examination of NSSI in younger community samples.
Despite these limitations, the findings suggest there is a need for clinicians to be aware of the way in which coping strategies and alcohol use interact in the relationship between psychological distress and NSSI. Future investigation is needed on less severe NSSI in community populations, as it appears to be highly prevalent and associated with psychiatric morbidity. In addition, the high rates observed among young adults recruited from universities suggest that university counselling centers may be a prime source of assistance for those who self-injure, and that these services may be ideally placed to offer prevention and early intervention programs for young adults.
Readers are encouraged to contact the authors for further information regarding this scale.