Prevention factors for suicide ideation among abused pre/early adolescent youths
- Etiology and Surveillance Branch, Division of Violence Prevention, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
- Dr J E Logan, Division of Violence Prevention, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, 4770 Buford Highway, MS-F63, Atlanta, GA 30341-3724, USA;
- Accepted 4 March 2009
Suicide ideation is a problem among youths who have been previously abused. This study assesses whether three factors (ie, feeling connected to school, having parents who reward good behaviour, and feeling able to cope with peer conflict) are negatively associated with suicidal ideation for 2598 pre/early adolescents with various levels of prior abuse. For the entire youth population, those who reported all three factors were less than half as likely to have suicidal thoughts as those who did not report any of these factors (10.8% vs 30.3%, p<0.05). This pattern was similar and significant for youths who experienced peer abuse (10.2% vs 35.0%, p<0.05) and youths who experienced both early child abuse and peer abuse (21.6% vs 54.8%, p<0.05). Comprehensive programmes that improve school connectedness, parent–child relationships and coping skills to avoid violent peer conflicts might help decrease suicide ideation among youths, particularly those who have been previously abused.
During 2001–2005, suicide was the second leading cause of death each year for youths aged 12–16 years in the USA.1 Preventing suicidal thoughts in pre/early adolescence might not only reduce psychological suffering but also future suicidal behaviour. Among adolescents, experiences of early child abuse and/or peer abuse have been found to be risk factors for suicide ideation.2–5 Perceived parent, family and school connectedness have been associated with low risk of suicidality.236 Also, being able to cope with peer conflicts in a non-violent manner might even further reduce risk of suicide ideation.5 (These factors have often been referred to as “protective factors”, a term used for factors that buffer risk.7 As these factors could also benefit youths who are not considered “at risk”, they will be referred to as “prevention factors” in this report.)
To add to the literature in this area of research, this study assesses whether similar prevention factors (ie, feeling connected to school, having rewarding parents and feeling able to cope with peer conflicts) are negatively associated with suicide ideation among pre/early adolescent youths who live in communities that have high levels of crime and poverty. Furthermore, these associations are assessed for youths who have been abused in early childhood, abused by a peer, or both.
Data for the present analyses were drawn from the Centers for Disease Control and Prevention’s (CDC’s) Youth Violence Survey: Linkages among Different Forms of Violence (“Linkages”) study. This was a cross-sectional study that surveyed youth from a high-risk community on their exposure to violence. This study was approved by human subjects institutional review boards at the CDC and ORC Macro International and was conducted in 2004. Details of this study have been reported by Swahn et al.8
To select this population, a list of the 20 cities with the highest sum of ranks on poverty, unemployment, single-parent households and serious crimes was reviewed. A study site was selected on the basis of commitment to the study and the feasibility of conducting a suitable census of students in targeted grades. The participating school district was among the highest 25 nationally in poverty, the highest 15 in single-parent families, the highest 10 in serious crime rates, and the highest 35 in rate of unemployment.
The survey instrument was administered to English-speaking public school students enrolled in grades 7, 9, 11 and 12; however, to understand suicide ideation among pre/early adolescents, this study focused only on those in the 7th and 9th grade. Before data collection, written parental permission and student assent were obtained. Of the 3314 7th and 9th grade students eligible, 2598 (78%) participated.
The outcome variable for this study, suicide ideation, was dichotomised: did/did not seriously consider attempting suicide within the 12 months before the survey. This outcome was assessed for youths who experienced up to two risk factors: (1) early child abuse (ie, before the age of 10 years, suffered sexual abuse (by anyone) or physical abuse by a caregiver that resulted in bruises, cuts or broken bones); (2) peer abuse (ie, perpetrator inflicted emotional abuse (eg, said hurtful things), physical abuse (eg, hit, kicked, pushed or grabbed), or sexual abuse). The prevention factors analysed included: (1) feeling connected to school (ie, feeling part of the school, close to people at school, or happy to be at school); (2) having rewarding parents (eg, perceiving parents to almost always say something nice, or give a hug or pat on the back for good behaviour); (3) feeling able to cope with peer conflicts (ie, feeling very confident about talking out disagreements with other kids, ignoring someone who is teasing, or seeking an adult to help with a disagreement).
Logistic regression was used to describe the association (via odds ratio (OR)) between various characteristics (ie, demographic, mental health, abuse history, prevention factors) and the outcome variable. For the risk and prevention factors, the respective OR was adjusted for age, sex and race to reduce confounding by potential demographic, social and/or cultural influences. The ORs were also adjusted for level of mental distress; severe distress could potentially confound the relationship between the risk/prevention factors and the outcome variable. Mental distress was measured as feeling sad, irritable or hopeless, having difficulty concentrating on schoolwork or having unusual eating or sleeping patterns. Youths who reported having one or more symptoms “often” or “always” within the 30 days before the survey were considered to have severe mental distress. Youths who reported “sometimes” for at least one symptom were considered to have mild mental distress.
For each level of risk (ie, no prior history of abuse, early child abuse only, peer abuse only, early child and peer abuse), the outcome variable was assessed as a function of: (1) feeling connected to school only; (2) feeling connected to school and having rewarding parents; (3) feeling connected to school, having rewarding parents, and feeling able to cope with peer conflicts. (The sample sizes of the risk groups were insufficient to assess the associations for having rewarding parents only, coping with peer conflicts only, or any other combination of prevention factors.) Prevalence estimates for suicide ideation were provided for youths at each level of risk. Fisher's exact tests were used to compare the prevalence of suicide ideation between those with various combinations of prevention factors and those who did not report having any factors. Similar analyses have been used in previous studies.69
Among this youth population, 436 (17.0%) reported suicidal ideation (table 1). Suicide ideation was more common among older youths (14+ years), females, Hispanic and non-Hispanic whites, and those categorised as having severe mental distress.
Suicide ideation was also positively associated with each pattern of prior abuse and negatively associated with each of the prevention factors (table 2).
On the whole, those who felt connected to school were less likely to consider suicide than those who did not report any prevention factors (20.9% vs 30.3%, significant at 0.05 level) (table 3). The prevalence of suicide ideation was even lower for those who also felt they had rewarding parents (12.9%) and for those who reported all three factors (10.8%). These findings were similar for youths who experienced peer abuse only. Among youths who experienced both early child abuse and peer abuse, those who did not report any prevention factors had the highest prevalence of suicide ideation (54.8%); those who had all three prevention factors were less than half as likely to have suicide ideation (21.6%). Promising results were also found for youths who either did not experience prior abuse or experienced early child abuse only. For both categories of youths, the prevalence of suicide ideation was lowest among those who either had the combination of school connectedness and rewarding parents or all three prevention factors. However, these values were not considered significantly different from those of the referent youths.
What is already known on this topic
Suicide is a problem among youths, especially among those who have experienced prior abuse by a parent/caregiver or peer.
Parent, family and school connectedness are factors associated with low risk of suicidality among adolescents.
What this study adds
Pre/early adolescents who feel connected to their school, who have parents who reward good behaviour, and who feel able to cope with peer conflict were not even half as likely to have suicidal thoughts as their counterparts.
This relationship was also found among youths who experienced prior peer abuse and youths who experienced both peer abuse and early child abuse.
Early child abuse2 and peer abuse4510 are considered risk factors for suicide ideation, and an estimated 61.8% of the youths in this study experienced one form of abuse or both. Among youths who experienced these forms of abuse, those who felt connected to school, had parents who rewarded good behaviour, and who felt able to cope with peer conflicts were significantly less likely to have thoughts of suicide than their counterparts. Similar findings have been reported in other youth sample populations, such as a national sample,6 a sample of Minnesota youths3 and a sample of Native American youths.9
There are promising strategies that emphasise these prevention factors and that can be conducted in a school setting. For example, Houck and colleagues11 studied support groups that emphasised cognitive–behavioural therapy and psycho-educational skill-building strategies among a high-risk female student population. These strategies were designed to improve interpersonal communication, problem solving, and emotional health. The support groups reduced suicide ideation among the group participants by 55%. These strategies also helped the participants initiate mental health treatment. Each support group was administered by a school nurse and a mental health counsellor.11 Other family-based interventions that strengthen parent–child relationships have shown promise for reducing interpersonal violence among youths.6 These interventions might also be effective in reducing self-directed violence.
This study had several limitations. Firstly, this was a cross-sectional study, therefore causation cannot be inferred. Secondly, the small sample of youths who experienced early child abuse only and the low prevalence of suicide ideation reported among youths who had no prior abuse might have prohibited the ability to detect significant associations for these groups. Furthermore, youths who experienced early child abuse only might have been biased toward not having rewarding parents, which further made the associations difficult to detect for this group. Thirdly, demographic differences between those with the prevention factors and the referent groups might have partially accounted for the difference in suicide ideation. However, on the basis of between-group comparisons of demographic factors using Fisher's exact tests, there were only differences by sex. Groups with the prevention factors had significantly higher proportions of females than the referent groups. Because the females were more likely to have suicide ideation, the true reduction in suicide ideation attributed to the prevention factors might have been underestimated.
Strategies that increase school and parent connectedness, improve coping skills for handling peer conflicts, and reduce the risk of early child abuse and peer abuse might help reduce suicide ideation among youths.
I thank Lawrence Barker for providing statistical guidance for this article.
Competing interests: None.
Funding: Centers for Disease Control and Prevention.
Ethics approval: Obtained.
Contributorship: The “Linkages” survey was developed by the Centers for Disease Control and Prevention (CDC) and was implemented in 2004 by contractors at ORC Macro International. I accept full responsibility for the conceptualisation, design and data analysis of this particular study.