CHAPTER ONE: INTRODUCTION
To quote Corie Blount “How will you know I am hurting if you cannot see my pain? To wear it on my body tells what words cannot explain”.
Previously, non-suicidal self-injury (NSSI) was only mentioned in the DSM as a symptom of Borderline Personality Disorder, now it is included in section 3 of the DSM 5, listed as a condition for further study (Jacobson, Hill, Pettit & Grozeva, 2015). Even though NSSI is not yet a diagnosis, personal history of self-harm is listed as a V-code (V15.59) in the DSM-5 (Zetterqvist, 2015). NSSI occurs when an individual inflicts bodily harm on themselves without the intention of suicide (Bheamadu, Fritz & Pillay, 2012; Duggan, Heath & Hu, 2015). The most common forms of NSSI are skin cutting, scratching, hitting oneself and burning; other forms of NSSI include wound picking and inserting objects under the skin (Jacobson et al., 2015; Brausch & Girresch, 2012).
Despite negative consequences in the long run, NSSI is known to get rid of unpleasant emotional states, which is why it is greatly prevalent among adolescents with depressive symptoms, as it is also used as a means of social communication for those who find it difficult to express their emotions as it seems to have a greater effect than language (Garc´?a-Nieto, Carballo, Hernando, Leo ´n-Martinez & Baca-Garc´?a, 2015; Bheamadu et al., 2012). Adolescents in particular are more susceptible to engaging in self-injurious behaviours, as changes in their neurodevelopment affects how they would process distress and pain which in turn will put them at a greater risk. Their emotional instability could cause depression, which may lead those who are unable to cope with these emotions to engage in self-injurious behaviours as a coping mechanism (Bheamadu et al., 2012).
Although there seems to be studies on NSSI among adolescents worldwide, there is a need for more research on the worldwide impact depression has on NSSI among adolescents. The lack of research in the South African context, even though the rate of NSSI in South African adolescents is so high, is also a huge concern, as there is a need for more research in this field of study in South Africa.
1.1. Rationale for the study
Depression is a risk factor for NSSI, especially amongst adolescents since depressed adolescents are at a greater risk of developing negative interpretations about themselves and others which could lead to self-injurious behaviours (Baetens et al., 2013; Garc´?a-Nieto et al., 2015). There is not much research done on the relationship between NSSI and depression, especially in South Africa, hence very little is known about this issue. The purpose of this study therefore is to explore the relationship between NSSI behaviours and depression in a sample of South African high school students. Insights from this study could bring awareness of the effect depression has on NSSI among adolescents and hopefully encourage more research and studies on this matter including solutions to help curve this growing concern and decrease adolescent NSSI by assisting the treatment of depressive symptoms.
1.2. Research aims
-The overall aim of this study is to investigate the relationship between age, NSSI and depression in a sample of high school students.
-To investigate age differences in depression scores and engagement in NSSI.
-To explore the nature and prevalence of NSSI in a sample of adolescents.
– To explore the relationship between NSSI and depression in the sample.
-To explore age differences in depression scores and engagement in NSSI.
1.4. Research questions
-What is the nature and prevalence of NSSI in a sample of adolescents?
-What is the relationship between NSSI and depression in the sample?
-What are the age differences in depression scores and engagement in NSSI?
CHAPTER TWO: LITERATURE REVIEW
2.1. Non-suicidal self-injury as a global problem
Self-injurious behaviour is a huge mental health concern around the world. According to research “self-injury tends to first occur during adolescence, is associated with the range of psychiatric difficulties, serves multiple interpersonal and intrapersonal functions and is significantly associated with increased suicidality” (Muehlenkomp, Claes, Havertape & Plener, 2012, p. 1).
Since there are many terms used around the world to define self-injury, it is difficult to compare prevalence rates of NSSI between countries using the same assessment tool, as the term deliberate self-harm is used mostly in countries like Australia and European countries, where the definition includes the infliction of bodily harm with or without the intention of suicide with non-fatal outcomes, and the term NSSI which includes bodily harm without suicidal intent, is used in countries such as the United States and Canada. This causes confusion for users, clinicians and researchers (Muehlenkomp et al., 2012; Sandy, 2013).
One big concern is that of cultural bias in the diagnosis of NSSI, as according to Muehlenkomp et al. (2012), the DSM-5 is proposing a NSSI disorder based on data collection from the United states and Canada which are both countries that use the term non-suicidal self-injury, which leads to irrelevance with other countries that use the term deliberate self-harm instead of non-suicidal self-harm. This affects the advancement of research on self-injury since reliable and valid differences cannot be estimated, which therefore makes it difficult to make pronouncements on protective and risk factors of self-injury among adolescents (Muehlenkomp et al., 2012).
A study in Spain, of 239 participants, found the most frequently used methods of NSSI for adolescents to be “hitting oneself (72.4%), cutting oneself (25.9%), and scratching oneself (22.4%)” (Garc´?a-Nieto et al., 2015, p. 222). The study showed that out of all of the participants that engaged in NSSI, 27.6% engaged in more than one type of NSSI, and out of the 27.6%, only 1.5% required medical attention. It was reported that 98.3% engaged in NSSI impulsively or with less than an hour of planning, and none of the participants that engaged in NSSI were found to have used alcohol/drugs before engaging in NSSI (Garc´?a-Nieto et al., 2015). It was also found that NSSI was noticeable in adolescents with 19 % from the community sample and between 32% and 50% of hospitalized adolescents being self-injurious (Garc´?a-Nieto et al., 2015).
Self-harm is a huge problem in child care institutions as this behaviour is often repeated in these settings; it becomes a growing concern as the highest incidents of self-harm are reported to be in secure settings. However, statistics about the incidence of self-harm are inconclusive as they do not take into account the number of secretive acts of self-harm that have not been discovered or have been discovered and injurers have mislead health professional into believing that their injuries have not been self-inflicted (Sandy, 2013).
Due to the limited understanding that nurses have about the motives for self-harm, many nurses often refer to these self-injurious behaviours as being attention seeking and manipulative. This negative association of self-harm by nurses can result in a risk for further self-harm with self-harmers possibly wanting to avoid going to health care services if they feel worthless and fearful of treatment knowing what the nurses think of them. There is therefore a need for the identification and understanding of self-injurious behaviours from nurses to prevent or reduce these behaviours (Sandy, 2013).
In a sample of 25 nurses who had over 2 years of experience with self-injurers in secure settings, four themes of self-injury were identified; visibility of self-harm, a cry of pain, a cry for help, and detention and institutional factors (Sandy, 2013). Findings suggested that in terms of visibility of self-harm, some participants stated that many people self-harm privately and therefore could not be doing it to seek attention as they were doing it secretively which is a concern as this could lead to suicide if encouraged, while most participants stated that some people self-harm publicly to seek attention and be listened to. In terms of a cry for pain, some participants believed that people self-harm to regulate distress while others believed that people self-harm to punish themselves for wrongdoings. In terms of a cry for help, some participants stated that people may self-harm to punish their abusers which is common among victims of sexual abuse, while some participants said that it is a way of seeking attention and manipulating their care to attract nurses when they are unable to verbally ask for help, and others suggested that it could also be used to drive others away. In terms of detention and institutional factors, some participants said that people self-harm because they feel confined in a controlled environment, some said that the lack of control in self harmers lives is a factor, and some participants felt that stigmatisation and labelling has an impact on why adolescents self-harm (Sandy, 2013).
The internet can also play a big role in self injurious behaviour among adolescents as they have daily access and are more engaged in online social networking and video sharing than other age groups. Isolated adolescents who find it difficult to communicate with others tend to use the internet to express their psychological distress or emotional difficulties associated with NSSI as the internet provides an anonymous platform for them (Lewis, Heath, Michal & Duggan, 2012).
Research suggests that there are benefits but also several risks with online NSSI content and communication. A benefit of online NSSI activity is that users can find social or peer support through these platforms which can help them share their NSSI experiences and connect with other self-injurers. According to research, some people involved in online NSSI activity have had a reduction in non-suicidal self-injurious behaviours after joining these groups (Lewis et al., 2012). Some risks include shared NSSI experiences and strategies that may cause the reinforcement of NSSI. Many people may use shared experiences and strategies as a means of encouragement to find new ways of non-suicidal self-injuries, therefore resulting in either people being introduced to NSSI as an option or people continuing to use self-injurious behaviours because of what they are exposed to online (Lewis et al., 2012; Marshall, Tilton-Weaver &Stattin, 2013).
2.2. Non-suicidal self-injury in South Africa
There is a paucity of research on NSSI in South Africa. Research on NSSI in high income countries seem to be abundant, while research on NSSI in low and middle income countries is often scarce.
A study by Penning and Collings (2014) was conducted in 2011 in a secondary school in Kwa-Zulu Natal, Durban. The purpose of this study was to determine the risk factors for traumatic re-enactment of child sexual abuse experiences such as perpetration, revictimization, and self-injury. In this study, 249 of the 718 participants had been sexually abused during the past 12 months and 90 participants stated that in the past 12 months they have engaged in self-injurious behaviour on at least five different occasions (Penning & Collings, 2014).
Research shows that NSSI is associated with a history of child sexual abuse and is mediated or moderated by affective dysregulation, dissociation, depressed mood, self-criticism and the extent of polyvictimization. Some individuals use NSSI to show their strength to ward off victimization when all else fails in ending assaults. Some individuals tend to use self-harm as a means of punishing their abusers and alleviating feelings of self-hatred, self-blame or guilt that may have occurred after being sexually abused while others may use self-injurious behaviours to drive others away trying to make themselves appear less attractive to their sexual abusers or to drive everyone away from them to protect themselves from further victimization or emotional harm (Penning & Collings, 2014; Marshall et al., 2013; Sandy, 2013).
Non sexual forms of traumatic exposure such as exposure to domestic neglect and non-accidental domestic injury also lead to NSSI in participants. Children who were exposed to physical abuse, neglect, and emotional abuse were more likely to administer self-injurious behaviour. This study also showed that domestic child physical abuse predicted self-injurious behaviour in participants (Penning & Collings, 2014).
A study by Bheamadu et al. (2012) conducted with twelve university students who had self-injured from adolescence to young adulthood used the biopsychosocial framework to assist the understanding of self-injury within the South African context, since self-injurious behaviour is so complex. Research has shown that NSSI can be analysed in multiple domains such as; the biological, psychological, and social domain. Biological and neurological influences form part of the biological domain; cognitive and affective factors form part of the psychological domain; and social, cultural and familial influences form part of the psychological domain (Bheamadu et al., 2012).
The biological domain can be broken down into two themes, which are physical pain and psychological pain. Even though both of these themes are complex ones, they can be differentiated in the sense that physical pain is self-inflicted and can be controlled whereas psychological pain, according to many, is uncontrollable (Bheamadu et al., 2012).With physical pain, individuals may feel that cutting themselves becomes an addiction as they would develop some sort of craving to inflict pain on themselves as it is said to give many a euphoric feeling that makes them feel better afterwards. With psychological pain, for some, it may not be about the pain but rather the feeling of a release of negative emotions for a relaxing feeling, as it is a quick way of decreasing negative affect-states and increasing positive affect-states for that period of time (Bheamadu et al., 2012; Penning & Collings, 2014).
The psychological domain can be broken down into two themes, which are affective experience and cognitive experience. Anger and frustration are common affective experiences that lead individuals to inflict harm on themselves. With cognitive experience, since individuals claim to feel a euphoric and revitalized sensation after injuring themselves, they would associate this act with these good feelings giving themselves this impression that these self-injurious acts are their friends and that they now have a sense of control over their emotions through self-injury (Bheamadu et al., 2012).
The social domain can be broken down into three themes, which are family experiences, peer relations, and social roles. With family experiences, a lack of boundaries in the home and mistreatment of individuals in the family may lead many to self-injurious behaviour. Parental criticism and familial abuse could also lead to self-criticism which in turn could lead to self-injurious behaviours (Bheamadu et al., 2012; Penning & Collings, 2014; Muehlenkomp et al., 2012).With peer relations, many may feel as if it is difficult to make friends with the lack of friendship being a result of them being too shy or scared to make friends. This could lead to individuals not having positive support from friends, leaving them feeling alone most of the time to drown their feelings of loneliness in self-injury (Bheamadu et al., 2012). Some individuals may also feel that some peer relations may contribute to their self-injurious behaviours by being in contact with people who self-injure and having this behaviour influenced or encouraged, while others may use self-injury as a sense of bonding with peers (Penning & Collings, 2014).
With social roles, many individuals feel frustrated because they do not fit in with everyone else and this frustration of being different and wanting to fit in leads to self-injurious behaviour (Bheamadu et al., 2012). Many individuals also end up using NSSI as a way of showing their strength to those who are not accepting of them or people around them who are emotionally distressing (Penning & Collings, 2014).
2.3. Depression and non-suicidal self-injury
A study by Marshall et al. (2013), using 506 Swedish adolescents from the 7th grade, found that a year after the study began, the increase of NSSI was predicted by depression; thereafter, there was a significant correlation between depression and NSSI, where the co-occurrence showed no direction of effect, and neither of the variables predicted the other. Depressive symptoms could predict NSSI over time as NSSI tends to alleviate the negative emotions and numbness associated with depression. Findings have suggested that people with higher levels of depression tend to have a greater engagement in NSSI (Jacobson et al., 2015; Baetens et al., 2013).
Seeing that individuals tend to feel that NSSI reduces problems such as depression, it is therefore associated with positive feelings, which is why depression precedes NSSI. Although some may feel that being self-injurious may decrease their depressive symptoms, it is also possible that NSSI can also be the cause of depression. Another issue is that since it is assumed that NSSI is meant to reduce depression and cause relief, if it in turn causes depression to reoccur, it could lead to continuous self-injury, therefore both depression and NSSI are bidirectional with a back and forth cycle (Marshall et al., 2013).
In a sample of 449 undergraduate psychology participants conducted in the United States by Jacobson et al. (2015), it was found that even after emotional expressiveness and reactivity was controlled, participants with higher depressive symptoms were four times more likely to engage in self-injurious behaviours as compared to those without depressive symptoms. The study found that the ability to express your feelings had an impact on NSSI, where even if depressive symptoms were controlled, those who were unable to express their feelings were at a higher risk of self-injurious behaviour than those that are able to express their feelings.
Another factor associated with NSSI is negative life events, which seems to be associated with depression in adolescents in clinical samples. In accordance with the stress exposure model of psychopathology, adolescents that are confronted with stressful life events are at a great risk of using NSSI to cope with their difficulties with emotional regulation (Liu et al., 2014).
In a longitudinal study of hospitalized adolescents in an inpatient facility in the United States, it was found that greater depressive symptoms and greater rates of negative life events resulted in greater engagements of NSSI. Even when the effects of gender and concurrent depressive symptoms were taken into account, greater life events were still associated with greater engagement of NSSI (Liu et al., 2014).
In a study using a sample of high school students between the ages of 15 and 16, females with depression were found to be more at risk for self-injurious behaviours. Students who self-injured and adolescents in community samples had more depressive symptoms as compared to those who did not self-injure (Brausch & Girresch, 2012).
Parental criticism and lack of parental support can also lead to self-criticism and depressive symptoms. Depression is common among adolescents raised in an invalidating environment as they may not know how to regulate intense emotions in an adaptive way. The lack of parental support can lead to depression which in turn can increase self-criticism which can then increase the risk of NSSI (Baetens et al., 2013).
2.4. The relationship between age, non-suicidal self-injury and depression
NSSI can begin during early to middle childhood with the behaviour often occurring between the ages of 12 and 16 years old. The age- related increases of heterogeneity of adolescents involved in NSSI is a possible explanation for variations between depressive symptoms and NSSI (Marshall et al., 2013).
Duggan et al’s. (2015) study showed that prevalence rates of NSSI among the clinical adolescent populations ranges from 38% to 82%, and 21% to 65% among adults, showing that adolescents are more vulnerable to non-suicidal self-injurious behaviours than adults. Research also shows that prevalence rates of NSSI from community young adults ranges from 12% to 20% whereas ranges from 14% to 26% are shown among adolescents.
Duggan et al’s. (2015) study found that, most of the youth who engaged in NSSI reported that they began between the ages of 12 and 15 years of age. Although there are reports that non-suicidal self-injurious behaviours can stop after 5 years, it can often continue into adulthood. According to Liu et al. (2014), the median age of onset of NSSI is 14, with adolescents appearing to be “a particular high-risk group, with lifetime prevalence rates for NSSI ranging from 13% to 23.2%” (p. 251).
Research has found that NSSI begins at an early onset, with the majority of adolescents beginning to engage in non-suicidal self-injurious behaviours between the ages of 13 and 15 years. According to Brausch and Girresch (2012), “Laye-Gindhu and Schonert-Reichl (2005) found a lifetime prevalence of 15% for NSSI in a high school sample”, whilst “Nock and Prinstein (2004) identified a prevalence rate of 82.4% for NSSI in a sample of adolescents drawn from a psychiatric inpatient setting” (p.4). Brausch and Girresch (2012) also makes reference to , Ross and Heath’s findings obtained from interviews with high school students, that suggested that 13.9% of the high school students that were interviewed had engaged in NSSI at least once, with 25% of those reporting that they began engaging in self-harm before the age of 12.
The prevalent rates of NSSI among adolescents around the world indicate the importance of attention that needs to be paid on this topic especially among this particular age group (Brausch ; Girresch, 2012). The lack of research in this field of study, especially in the South African context is a concern, as more research would be beneficial to informing stakeholders on the impact depression has on NSSI among adolescents and the necessary steps that can be taken to curb these increasing rates of adolescent non-suicidal self-injurious behaviours can be established.
CHAPTER THREE: METHODOLOGY
3.1. Research approach
This study was based on a positivistic paradigm using a deductive approach, as it was an objective study that was used to explain why and how NSSI is related to depression. Qualitative research design generates exploratory data using a small sample size while quantitative research designs generates confirmatory data using a large sample size. Qualitative approaches use structured or semi structured response options with the use of focus groups, in-depth interviews, participant observations, field notes, and reflections; while quantitative research uses fixed response options with the use of surveys/questionnaires, structured interviews ; observations, and reviews of records or documents for numeric information. The most common research objectives of qualitative research is to explore, discover, and construct, while the most common research objectives of quantitative research is to describe, explain, and predict (Johnson ; Christensen, 2008; Lichtman, 2006). This study adopted a quantitative approach and analyses were performed using SPSS 24.
Since quantitative research requires large samples, this study included 233 participants for more accuracy and to minimise the chances of errors. A convenience sampling approach, which is a method of recruiting participants from wherever you find them, of high school students was used. The sample comprised both males and females from four high schools in the Durban area.
This study was conducted by administering questionnaires, and the scales that were used are the Inventory of Statements about Self-Injury (ISAS), the Beck Depression Inventory (BDI-II), and a demographics questionnaire. The ISAS and BDI have not been standardised for use in South Africa but both have been used in the South African context before.
Inventory of Statements about self-Injury
The ISAS is a self-reported questionnaire that assesses NSSI behaviours and functions. Section I in the ISAS “measures the lifetime frequency of 12 NSSI behaviours”, with the 13th item being other, of which have been performed either with or without the intent of suicide, indicating the number of times each type of self-harm has been performed from 0 to 500 being the highest amount of times (Glenn ; Klonsky, 2011, p.376). Question 2 to 7 of section I ask about the age of first and recent self-harm, whether physical pain was experienced during self-harm, whether self-harm occurs alone, the amount of time that elapses from the time you have an urge to self-harm to the time you act upon that urge, and whether the plan to stop self-harming (Hamza ; Willoughby, 2013). According to Glenn and Klonsky (2011) the behaviour scales, “have demonstrated good internal consistency (? =.84) and short-term (1-4 weeks) test–retest reliability (r = .85). Section II of the ISAS consists of 39 items and assesses 13 functions of NSSI that make up two subscales which are the Intrapersonal and Interpersonal subscales (Glenn & Klonsky, 2011).
According to Glenn and Klonsky (2011), the Intrapersonal subscale has a coefficient alpha of ?= .80, and the interpersonal subscale has a coefficient alpha of ?= .88. The ISAS behavioural and functional scales have exhibited good construct validity as indicated by theoretically consistent relationships to other NSSI and clinical variables” (p.375).
Beck Depression Inventory
The BDI-II is a 21 item self-report scale that measures depression in adolescents and adults. The BDI-II can be scored by adding up the scores of each of the 21 symptoms. Each symptom is ranked in terms of severity and scored on a 4-point scale ranging from 0 to 3 with total scores that can range from 0 to 63 (Steer, Ball ; Ranieri, 1999). An example of a group of statements that is seen in the BDI-II is sadness: 0(I do not feel sad), 1(I feel sad much of the time), 2(I am sad all the time), and 3(I am so sad or unhappy that I can’t stand it), where the participant is required to circle the number of that best describes them during the past 2 weeks for the statement.
Steer et al. (1999) recorded a coefficient alpha of .90 in a study sample of 120 clinically depressed outpatients. The coefficient alpha represents a high internal consistency which shows the reliability of the BDI-II.
The participants were also required to answer a questionnaire that contained items about their demographic characteristics. This questionnaire included questions about their age, sex, race, religion, with whom they lived, and their grade.
3.4. Data collection
A survey method, using paper and pen questionnaires was used in this study. These standardized self-report questionnaires were administered to the participants. Once all questionnaires were completed by the participants, they were collected and later analysed.
3.5. Data analysis
The analyses of the data were conducted using SPSS 24 software. The data was analysed using frequency analyses, followed by scale reliability analyses, a correlation analysis, Chi Square analysis and an ANOVA. The scale reliability analyses were used to analyse the reliability of the ISAS subscales and the BDI used in this study. The correlation analysis was used to explore the relationship between NSSI and depression, and whether the relationship resulted in a positive correlation or negative correlation. The ANOVA was used to compare the effects the different age groups had on depression. The Chi Square analysis was used to determine whether there was a significant difference between the different age groups and NSSI. The Chi square analysis was also run on each of the different types of self-harm to determine whether there was a significant difference between the different age groups and each of the different types of self-harm.
3.6. Ethical considerations
The study was ethically reviewed and an amendment of protocol to include high school learners was applied for to the UKZN Biomedical research Ethics Committee (REF: BE 138/14), by the project supervisor. Permission to conduct the study was sought from the Department of Education and the principals of the four high schools. Final ethical clearance was also obtained so that we had the go ahead to conduct our study. All data will be stored in a locked cabinet in the university for a period of 5 years and destroyed by shedding afterwards so as to maintain the confidentiality of all participants. Given that adolescents are a vulnerable group when it comes to NSSI, emotional harm from the study was a possibility; however, although the study was voluntary, participants were given the option of leaving the study if they felt distress, with the option of debriefing and follow up counselling.
3.7. Research procedure
Once final ethical clearance was obtained, we approached the schools and made the necessary arrangements with the educators to conduct the study with a few classes on specific dates that was convenient for them and would not disrupt their academic program. Thereafter the students were given a consent form, of which they had to get permission from their parent or guardian to participate in the study. The consent form stated that the study would respect the confidentiality and anonymity of the participant, participation was voluntary, and the nature and aims of the study was explained so that participants were not misled by the study and felt comfortable to participate. Before commencing the study, participants were given the opportunity to ask any questions about any confusion or concerns they may have had. Assent forms were then given to the participants, which were filled to indicate that they understood what the study entails and are willing to participate in the study. The class was then split with the learners who provided consents and assents sitting on one side of the class, for the participation of the study. To try and avoid harm in the study, the schools counsellor was present so that if at any point during the study any child experienced emotional distress, counselling would’ve been provided. This study was independent and impartial as there are no conflicts of interest of the researcher and the research was conducted objectively. Throughout the study, the learners could also ask questions about any uncertainties they encountered regarding what the questionnaires required from them. Thereafter, all of the questionnaires were collected and the participants were thanked for their participation. Once the data was obtained, it was recorded without tampering with the results and it was checked to ensure that no errors were made.
CHAPTER FOUR: RESULTS
This chapter presents the results obtained from the statistical analysis of the data for the present research study. It provides the results of the frequency analysis, reliability analyses, the correlation analysis, Chi square analysis and the ANOVA.
4.1. Socio-demographic characteristics
From table 1, we can see that from the total sample of 233 participants, the majority of which were females (n=176, 74.9%). There were more Blacks in the study with the percentage being 40.9% (n=96), and the minority of the participants (4.7%) indicating Other as their race (n=11).The majority of the participants (55.3%) in the study belonged to the age group of 15-16 years old (n=130), while the minority of participants (13.2%) belonged to the age group of 17-20 years old (n=31). The majority of 60.4% were Christians (n=142), while the minority of 3.4% were Muslims (n=8).
With regards to the percentage of participants that engaged in self-harm, as seen in table 2, a total of 60.5% of the 233 participants indicated that they have self-harmed before or are currently engaging in self-harming behaviours. Within each age group, the amount of participants that self-harmed outnumbered the amount of participants that did not engage in NSSI.
Table 1: Frequency analysis of age groups, sex, race, and religion
Items N (233) PERCENTAGE (100 %)
Black 96 40.9
Indian 94 40.0
White 15 6.4
Hindu 64 27.2
Christian 142 60.4
Muslim 8 3.4
Other 19 8.1
Table 2: Participants that engaged in self-harm
From the table below, it was found the majority of the participants (42.1%) indicated engaging in self-harm behaviours alone.
Table 3: Nature of NSSI
No 18 7.7%
Yes 99 42.1%
Sometimes 15 6.4%
Total 132 56.2%
From the table below, it was found that the majority of the participants indicated engaging in NSSI within less than an hour of their urge to self-harm (26.0%). A minority of the participants indicated engaging in NSSI within 12-24 hours of having an urge to engage in self-harm behaviours (1.3%).
Table 4: Time lapse from the urge to self-harm to the engagement of self-harm behaviours
Time lapse Frequency Percent
; 1 hour 61 26.0%
1-3 hours 21 8.9%
3-6 hours 4 1.7%
6-12 hours 4 1.7%
12-24 hours 3 1.3%
;1 day 31 13.2%
Total 124 52.8%
4.2. Reliability analyses
The Intrapersonal subscale has good internal consistency; with a Cronbach alpha coefficient reported .93. The Interpersonal subscale has good internal consistency, with a Cronbach alpha coefficient reported of .93. The BDI has good internal consistency, with a Cronbach alpha coefficient reported of .94.
4.3. The relationship between non-suicidal self-injury and depression
A bivariate correlation using the Pearson product-moment correlation coefficient was used to explore the relationship between the Intrapersonal and Interpersonal subscales, the Intrapersonal subscale and the BDI, as well as the relationship between the Interpersonal subscale and the BDI. To ensure that no violations of assumptions were made, preliminary analyses were performed. There was a strong positive correlation between the Intrapersonal subscale and Interpersonal subscale, r=.643, n=231, p;0.001, with a strong linear relationship showing increased levels of Intrapersonal functions of NSSI being associated with increased levels of Interpersonal functions of NSSI. There was a strong, positive correlation between the Intrapersonal subscale and the BDI, r=.551, n=231, p;0.001, with a strong linear relationship showing increased levels of NSSI being associated with increased levels of depression. There was a weak, positive correlation between the Interpersonal subscale and the BDI, r=.204, n=233, p;0.002, with a weak linear relationship showing increased levels of NSSI being associated with increased levels of depression.
Table 5: Pearson product-moment correlations between measures of self-harm behaviour and depression
Scale 1 2 3
Total Intrapersonal (ISAS) -.643**.551**
Total Interpersonal (ISAS).643** -.204**
Total Depression (BDI).551**.204** –
** p; 0.01 (2-tailed).
4.4. The relationship between age and non-suicidal self-injury
The Chi-square test for independence was used to explore the relationship between the different age groups and NSSI as measured by the ISAS. It was discovered that the assumptions for the Chi-square test for independence were not violated, as zero cells (0.0%) had an expected count less than 5, with the minimum expected count being 12.24. The Likelihood Ratio was used to interpret the significant value and it indicated no significant association between age and NSSI as measured by the ISAS, ?2 (1, n= 233) = 0.31, p=.86, phi= .036.
From the table below, we can see that among the participants between the ages of 13-14 years old, the most common type of NSSI behaviour was interfering with wound healing (n=24). Between the ages of 15-16, the most common type of NSSI behaviour was interfering with wound healing (n=41). Between the ages of 17-20 years, the most common type of NSSI behaviour was banging or hitting oneself (n=11). Overall, the most common type of NSSI behaviour was banging or hitting oneself (n=64).
Table 6: Prevalence of individual self-harm behaviour according to age group
13-14 years 15-16 years 17-20 years Total
Cutting 19 (8.2%) 30 (12.9) 9 (3.8%) 58 (24.9%)
Scratching 13 (5.6%) 23 (9.9%) 6 (2.6%) 42 (18.0%)
Biting 19 (8.2%) 21 (9.0%) 5 (2.1%) 45 (19.3%)
Banging or Hitting Oneself 17 (7.3%) 36 (15.5%) 11 (4.7%) 64 (27.5%)
Burning 4 (1.7%) 10 (4.3%) 8 (3.4%) 22 (9.4%)
Interfering with Wound Healing 24 (10.3%) 41 (17.6%) 8 (3.4%) 73 (31.3%)
Carving 7 (3.0%) 13 (5.6%) 3 (1.3%) 23 (9.9%)
Rubbing Skin Against Rough Surface 12 (5.2%) 16 (6.9%) 7 (3.0%) 35 (15.0%)
Pinching 20 (8.6%) 31 (13.3%) 4 (1.7%) 55 (23.6%)
Sticking Self with Needles 10 (4.3%) 8 (3.4%) 2 (0.9%) 20 (8.6%)
Pulling hair 19 (8.2%) 27 (11.6%) 8 (3.4%) 54 (23.2%)
Swallowing Dangerous Substances 13 (5.6%) 22 (9.4%) 5 (2.1%) 40 (17.2%)
Other 3 (1.3%) 8 (3.4%) 1 (0.4%) 12 (5.2%)
4.5. The relationship between age and depression
A one-way between groups analysis was conducted to compare the impact the different age groups had on depression as measured by the BDI. Participants were divided into three groups according to their age (Group 1: 13-14 years; Group 2: 15-16 years; Group 3: 17-20 years). There was no statistically significant difference at p;0.05 level in the BDI scores of the three age groups: F (2, 230) = .621, p= 0.54. Even though statistical significance was not reached, the actual difference in mean scores between the groups was large, as the effect size using eta squared was 5.37. The Post-hoc comparisons using the Tukey HSD test was therefore not needed as Group 1: 13-14 years (M=17.47, SD=12.18), Group 2: 15-16 years (M=19.64, SD=14.4), and Group 3: 17-20 years (M=19.77, SD=15.36) did not differ significantly from each other, seeing that p;0.05.
CHAPTER FIVE: DISCUSSION
In this chapter, I will be discussing the results according to the three objectives of this study, viz exploring the nature and prevalence of NSSI, the relationship between NSSI and depression, and the age differences in depression scores and engagement in NSSI in the sample. The frequency analysis will be used to discuss prevalence of depression and the engagement of NSSI in the sample of high school learners, and the objectives of this study will be discussed using the results obtained from the Frequency analysis, scale reliability score analyses, Correlation analysis, Chi square analysis and the ANOVA analysis. The results of this study will be discussed with reference to previous research, and thereafter the strengths and limitations for this study will be discussed, followed by recommendations for future research and a conclusion.
5.1. Socio-demographic characteristics
Using the frequency analysis we were able to categorize the participants of this study in accordance to the different socio-demographic characteristics. By categorizing each participant into the different socio-demographic groups, we were able to determine whether or not certain socio-demographic characteristics had an influence on depression and the engagement of NSSI. The demographic characteristics that we focused on was age, race, religion, and sex
The demographic variable that I have focused on in this paper is the different age groups of the high school participants (this sub study was part of a larger project), and whether or not age had an influence on their depression or engagement in NSSI. The three age groups obtained from this research sample are: 13-14 year olds, 15-16 year olds, and 17-20 year olds.
5.2. Scale reliability score analyses
Looking at the scale reliability score analysis of the Intrapersonal and the Interpersonal subscales, there is therefore strong internal consistency reliability of both subscales of the ISAS for this sample of high school learners. A comparison of the Cronbach alphas of the interpersonal and intrapersonal subscale from the study by Glenn and Klonsky (2011) with the present study indicates that the present study yielded a higher Cronbach alpha for the intrapersonal subscale and interpersonal subscale, suggesting a better model fit and a better construct validity for this sample. The results suggest that this is a valid and reliable instrument for use in our context.
In considering the scale reliability score analysis of the BDI, the Cronbach alpha value in the present study is above .8, and indicates strong internal consistency of the BDI for this sample of high school learners. If we had to compare the Cronbach alpha from the study by Steer et al. (1999) with the present study, we can see that the present study yielded a higher Cronbach alpha for the BDI scale, suggesting a better model fit and a better construct validity. However, even thought the Cronbach alpha for the present study is higher than the Cronbach alpha for the study by Steer et al. (1999), the difference between both alpha values is not large. A South African study conducted by Naidoo (2016) indicated a higher Cronbach alpha than the findings of the present study. Since this scale produced strong internal consistency reliability, the use of the BDI is recommended in our context as it will be useful in measuring self-reported depression.
5.3. Nature and prevalence of NSSI
The present study showed that the amount of participants that indicated engaging in self-harm significantly outnumbered the amount of participants that indicated not engaging in self-harm behaviours, showing that there seems to be a very high prevalence rate of NSSI among adolescents. This is consistent with previous research that suggests that the adolescent population is a high risk population when it comes to engaging in NSSI (Bheamadu et al., 2012; Duggan et al., 2015; Brausch ; Girresch, 2012).
Regarding the nature and prevalence of NSSI, the results found that the majority of participants indicated engaging in self-harm behaviours alone. This is consistent with previous research by Sandy (2013), which indicated that the majority of self-harm behaviours are usually private and secretive. It was also found that the majority of the participants indicated acting on their urge to self-harm within less than an hour. This is consistent with previous research by García-Nieto et al. (2015) that showed a majority of participants acting impulsively on their urge to self-harm (within less than an hour).
A possible reason for the majority of participants engaging in self-harm behaviours alone could be due to the fact that they do not want to be labelled as “manipulators or attention seekers” (Sandy, 2013, p.363). Some may self-harm alone due to them not wanting their self-harm behaviours discovered, so that they may continue engaging in NSSI (Sandy, 2013).
A possible reason for individuals engaging in self-harm impulsively could be due to these individuals wanting a quick way to deal with their emotional pain, as many claim that NSSI replaces these negative feelings with euphoric feelings (Bheamadu et al., 2012; García-Nieto et al., 2015). It is also possible that these individuals may engage in self-harm out of impulse due to their inability to cope with their emotional pain or their lack of problem-solving skills (Brausch ; Girresch, 2012).
5.4. The relationship between non-suicidal self-injury and depression
In order to explore the relationship between NSSI and depression, a bivariate correlation analysis was performed on the Intrapersonal subscale of the ISAS and the Interpersonal subscale of the ISAS, the Intrapersonal subscale of the ISAS and the BDI, as well as the Interpersonal subscale of the ISAS and the BDI. The Intrapersonal subscale includes self-focused functions and the Interpersonal subscale includes relational functions related to non-suicidal self-injury, while the BDI included items related to depression.
The Pearson product-moment correlation coefficient was used to determine the strength of the relationship between the Intrapersonal subscale and the Interpersonal subscale, which indicated a strong positive correlation between both variables. The strength of the relationship between the Intrapersonal subscale and the BDI indicated a strong positive correlation between both variables. The strength of the relationship between the Interpersonal subscale and the BDI indicated a weak positive correlation between both variables. Since there was a statistical significance between the Intrapersonal subscale and the BDI as well as between the Interpersonal subscale and the BDI, the null hypothesis was rejected and the alternate hypothesis stating that there was a relationship between NSSI and depression was accepted.
These correlation analyses could therefore be backed up by previous research that showed the link between NSSI and depression, showing that depressive symptoms were more likely to be seen among individuals who engaged in NSSI compared to individuals who did not engage in NSSI (Brausch ; Girresch, 2012). Research done by Turner, Chapman and Layden (2012), suggested that “findings from several studies further supported an association of Intrapersonal functions of NSSI with symptoms of depression” as well as one study supporting the association between Interpersonal functions and depression (p.3). Turner et al. (2012) suggested that depressive symptoms were “more strongly associated with intrapersonal rather than interpersonal functions”, which is consistent with findings of the present study that showed a weaker correlation between the interpersonal subscale and BDI. Even though it was found that depression seems to be more likely related to Intrapersonal functions of NSSI, these findings indicated that the relationship is not unique to Intrapersonal functions of NSSI but could also be seen between Interpersonal functions of NSSI and depression, which indicates that there is a relationship between the overall functions of NSSI and depression (Turner et al., 2012).
From the findings of the present study as well as previous studies, we can deduce that depressive symptoms seem to have an impact on the engagement of NSSI among adolescents. This could be due to adolescents not knowing how to correctly deal with their depressive symptoms and feelings of being alone, therefore engaging in NSSI to escape from the harsh realities of life. They may use the infliction of physical pain upon themselves through NSSI, as a coping mechanism to deal with the emotional pain and burden they may face from depression (Skegg, 2005). Another possible reason for the association between depression and NSSI among adolescents could be the fact that these adolescents may feel immense emotional pain due to their depression that may lead them to engage in NSSI to express their internal pain through their external scars (Skegg, 2005).
5.5. The relationship between age and non-suicidal self-injury
In order to explore the relationship between age and NSSI, a chi-square test for independence analysis was performed using the ISAS total and the age groups variable. This analysis helped determine which age group engaged in self-harm more than the others. Another analysis was performed using the individual ISAS total for each of the different types of NSSI and the age group variable, to determine which of the specific types of NSSI behaviours are most prevalent among the different age groups.
The results obtained from the present study showed that there was no significant difference in the relationship between age and NSSI. Therefore age has no effect on the engagement in NSSI.
Although there were no statistically significant differences, the present study found that the age group of 17-20 years had the highest prevalence of self-harm. This is inconsistent with previous studies that suggest that adolescents between the ages of 12-16 years of age have a higher prevalence of NSSI (Marshall et al., 2013; Duggan et al., 2015; Brausch ; Girresch, 2012). The fact that the 17-20 year age group indicated a higher prevalence in terms of engaging in NSSI, could be due to the immense stress they may be faced with, with the lack of support and feelings of social isolation resulting in them engaging in NSSI behaviours (Bheamadu et al., 2012; Skegg, 2005).
According to previous research, individuals between the ages of 12-16 years of age are most at risk for engaging in NSSI due to this age group often being the recorded onset of NSSI among adolescents (Marshall et al., 2013). However, the results from the present study indicated that the participants between the ages of 15-16 years showed the lowest prevalence in terms of NSSI. Even though this age group reported the lowest prevalence of self-harming behaviours as compared to the other three age groups, the percentage of participants that indicated engaging in NSSI from this particular age group was still high, which is still a cause for concern. With both the 13-14 years and 15-16 years age group having such a high prevalence rate of NSSI, even though it is not the highest among all three age groups of the sample, we cannot reject the notion made by Marshall et al. (2013) that these particular age groups are the most common age groups in terms of the onset of NSSI among adolescents.
With regards to the specific types of NSSI, according to the results obtained from the present study, the most common types of NSSI behaviour was interfering with wound healing and banging or hitting oneself. According to Glenn and Klonsky (2011), their findings have suggested that burning and cutting were the most common types of NSSI behaviours. There are therefore inconsistencies between the present study and previous studies with regards to the specific forms of NSSI.
The inconsistency between the present study and the previous study could be due to the fact that burning and cutting are less lethal traditional methods of suicide that can most likely be associated with engaging in self-harm behaviours with the intent of suicide whereas interfering with wound healing and banging or hitting oneself may cause tissue damage but are more likely associated with self-harm behaviours without the intent of suicide (Skegg, 2005). Another possible reason for interfering with wound healing and banging or hitting oneself being the most common forms of NSSI used in the present study, could be due to the fact that these behaviours are most commonly noticed among individuals with psychiatric illnesses; therefore there is a possibility that these behaviours could be due to the adolescents experiencing depression, or other psychiatric illnesses such as anxiety disorders or substance abuse (Skegg, 2005).
5.6. The effect of age on depression scores
In order to compare the relationship between depression and age, a one-way between groups ANOVA analysis was conducted using the BDI total and the age groups variable. This helped us to determine which age groups were more likely to show depressive symptoms.
The present study showed that there was no significant difference in mean scores between the three age groups on the dependent variable which in this case is the BDI. We can therefore accept our null hypothesis that the mean of the BDI is the same for all of the age groups. Although there were no statistically significant differences, if we had to look at the mean scores of the different age groups, we can see that the 13-14 year olds had the lowest depression score, whereas the participants belonging to the age group of 17-20 years old had the highest depression score. This indicates that learners between the ages of 17-20 years seem to have more depressive symptoms as compared to the learners from the other age groups. A similar finding was reported by Steer et al. (1999) which also showed a gradual increase in self-reported depression among people between the ages of 18-38 years old, indicating that depressive symptoms seem to be more prevalent in people during late adulthood.
A possible reason for the high depressive scores seen among adolescents between the ages of 17-20 years old could be due to individuals between these age groups being faced with more stressors and greater expectations and challenges that may lead to their depressive symptoms. Another reason could be that individuals from this age group are faced with greater responsibilities, which could lead to their depressive symptoms due to their feelings of being overwhelmed with such responsibilities. This may result in their feelings of depression, due to the difficulty they may feel from coping with life’s demands (Liu et al., 2014; Skegg, 2005). Since the sample includes high school learners, it is a possibility that higher depressive symptoms are seen among this age group, due to the stress and uncertainty they may feel about what the future holds for them in terms of post school, with regards to furthering their studies or getting a job (Skegg, 2005).
5.7. Summary of findings
The findings of this study suggested that there is a relationship between age, NSSI and depression. It was found that individuals who exhibit increased depressive symptoms are more likely to engage in NSSI. This can be attributed by many feeling that NSSI decreases their negative affect-states, therefore engaging in NSSI to cope with their overwhelming depressive symptoms (Marshall et al., 2013). It was also found that individuals between the ages of 17-20 years old seem to have the highest prevalence of self-harm and depressive symptoms. This increased rate of NSSI and depressive symptoms could be attributed to the increased negative life events or life stressors that this particular age group may constantly encounter (Liu et al., 2014).
6.1. Strengths and Limitations of the study
The present study obtained a sample size of 233 participants which can be seen as a strength, as this large sample size increased the accuracy of the study and minimized potential errors. However, a larger sample size would have been preferable. Since the research population of the current study included high school learners, we therefore had to give out consent forms seeing that most participants were below the age of 18, which lead to problems as many participants did not remember to give their parents or return the consent forms, which in turn affected the sample size of this study as we were unable to use most learners in the study.
The present study made use of self-report questionnaires; although the use of these questionnaires was easy to interpret, inexpensive, and a quick way to collect a large amount of data, there were a few limitations to the use of self-report questionnaires (McDonald, 2008). The credibility of self-report questionnaires could have been affected as there is a possibility of response bias, which affects the accuracy of the study. Response bias can take on the form of socially desirable responding which involves presenting oneself in a favourable light, acquiescent responding which involves agreeing with statements with a disregard of the content, and extreme responding which involves the tendency to give extreme rating on scales (Paulhus & Vazire, 2007). Other forms of response bias that could have influenced the credibility of the study are “pattern responding, random responding, and inconsistent responding” (Paulhus & Vazire, 2007, p.232).
According to Paulhus and Vazire (2007), “no one else has access to more information” than oneself, and individuals are usually motivated to reflect and talk about themselves (p. 227); this is advantageous as these self-report questionnaires become more valid, with information about the respondents being received directly from them. According to McDonald (2008), however, a “potential limitation is whether people know enough about themselves to be able to accurately portray what the self-report is attempting to determine” (p.79).
The fact that this study focused solely on learners across the Durban area could be a limitation as it may not be generalizable to other geographical areas in representing the wider school or adolescent population. However, the study demographically represented the high school learners in the Durban area of which the study took place.
6.2. Recommendations for future research
The study provides us with an understanding of the relationship between age, NSSI, and depression; however there is a need for further research especially with regards to the South African context. Although most of the results from the present study are consistent with the results obtained from previous research, there were inconsistencies with regards to the relationship between age and NSSI and age and depression, which needs to be looked at through further research, to address the possible reasons behind the inconsistencies of the findings. A larger sample size is recommended for future research, as it increases the validity of the study and provides clarification for the issues relating to NSSI and depression among adolescents.
The purpose of this study was to investigate the relationship between age, non-suicidal self-injury and depression in a sample of high school students. This study indicated that there is a relationship between NSSI and depression among adolescents. The findings indicate that NSSI is largely prevalent among adolescents, with the majority of participants indicating that they have self-harmed. This high prevalence rate therefore needs to be looked at to discover why these particular age groups are at a huge risk in terms of NSSI behaviour, to inform prevention programmes to reduce the problem. The prevalence of self-reported depressive symptoms among adolescents is also a concern, as a depressive mood could not only lead to NSSI but also lower performance in school and decreased self-esteem, therefore affecting the livelihood of these adolescents. As much as NSSI and depression is a problem among both adolescents and adults, there is a need for much more research on adolescent NSSI and depression, especially in the South African context, to create awareness on the prevalence of NSSI and depressive symptoms among adolescents and to encourage solution based strategies to decrease this prevalence. By creating awareness about these issues and finding solutions to adolescent NSSI and depression, we can maintain the mental and physical health among the youth, in terms of NSSI and depression. To quote Nelson Rolihlahla Mandela “The youth of today are the leaders of tomorrow”, therefore more attention needs to be paid to adolescent NSSI and depression to try and prevent long term mental health problems among the youth.