Self-harm with and without suicidal intent in adolescence: causes and consequences in the ALSPAC cohort

Recent community studies indicate that up to 1 in 7 adolescents have self-harmed at some point in their lives, with the incidence rising rapidly between 10 and 16 years of age. These findings have led to concerns that more should be done to identify and treat young people who self-harm, but the clinical significance of community presenting self-harm is unclear. One quarter of self-harm episodes are carried out with expressed suicidal intent; whilst such episodes are a risk marker for later suicide attempts and suicide, most cases never reach medical attention and later health and social outcomes amongst adolescents who self-harm are uncertain.

To date, most studies of self-harm in adolescence have been based on small clinical samples - these are highly select groups, already known to medical services. The few population-based studies have generally measured risk factors and self-harm at one point in time, with few follow-up studies. The investigations that have been carried out are generally limited by small sample sizes and tend to have studied suicide attempts, rather than inter-relationships between self-harm with and without suicidal intent. Examining these latter relationships is important both to determine the clinical significance of non-suicidal self-harm in adolescence and to guide prevention and treatment strategies. Furthermore, knowing the health and social outcomes amongst adolescents whose self-harm does not come to medical attention will help determine the need for interventions in this group of people.

The proposed research is based on the Avon Longitudinal Study of Parents and Children (ALSPAC). ALSPAC is a long term study of over 13,000 children born in Bristol and the surrounding area in 1991-1992 who have been followed up since birth with regular questionnaires to parents, teachers and, more recently, the participants themselves. Study members have also been attending clinics every 1-2 years since the age of 7 years. They have provided information on suicidal thoughts and self-harm with and without suicidal intent at ages 11 and 16. The lifetime risk of self-harm was 1.6% at age 11, this had risen to 19% 16 years of age.

Our research will investigate whether there are differences in the characteristics and medium term outcomes between children who self-harm with and without suicidal intent in adolescence and in those who do and do not come to medical attention. Findings will be used to determine the clinical relevance of adolescent self-harm and whether there are sub-groups of self-harming adolescents who warrant particular attention. The principle outcomes we will investigate are a) mental health (assessed using the Clinical Interview Scale-Revised (CIS-R) at age 18) b) alcohol and cannabis abuse; c) educational performance (age 14: SATS; age16: GCSEs and age 18: A-levels and d) early employment record.

Two important issues when studying self-harm in ALSPAC are that it is possible that self-harm is under-reported by participants and that its prevalence (and risk factors) may differ between survey responders and non-responders. We will use recently established linkages between ALSPAC and national Hospital Episode Statistics data (all participants with relevant permissions) and the General Practice Research Database (GPRD) (5% of participants) to investigate these issues and assess their impact on the analyses in our proposed research and their implications for future research.

We have identified no other prospective studies that are in a position to answer these research questions.

The prevalence of self-harm in adolescence is between 13 and 18%. Most adolescent self-harm is carried out with no suicidal intent and the majority of cases never come to medical attention. Current understanding of the causes and consequences of community presenting self-harm is limited. Furthermore, there is international uncertainty, exemplified by recent debates during the Diagnostic and Statistical Manual (DSM) revisions, of the diagnostic and clinical utility of distinguishing self-harm with and without suicidal intent. We propose investigating these issues in the ALSPAC cohort of over 13,000 children followed up since birth. Participants completed detailed interview/ questionnaires about self-harm at ages 11 and 16 years with n=905 participants reporting episodes of self-harm. Through regular follow-up they continue to provide information on their health and social circumstances.

We will investigate differences in risk factors (e.g. depression, substance misuse) and medium term health and social outcomes (to age 19 years) for community presenting self-harm with and without suicidal intent and for those cases known/not known to services, to test the hypotheses that these two behaviours differ in aetiology and outcome. The principle outcomes we will investigate in our prospective analyses are a) mental health (clinical interview scale-revised (CIS-R) at age 18) b) alcohol and cannabis abuse at ages 18 and 19; c) educational performance (ages 14, 16 and 18 and d) early employment record. We will perform a series of regression models to test the main hypotheses, and use multiple imputation to assess the impact of missing covariate data on the associations we identify. We will also use recently established linkages with GPRD and HES to carry out sensitivity analyses taking account of any random or systematic error associated with self-report data and non-response (n=6,400 respondents age 11 and n=5,000 age 16).

Scientific advancement and impact upon the research community:

The research will lead to new knowledge concerning the appropriate classification of suicidal behavior and the health impact of different forms of suicidal behavior. This will inform future revisions of the Diagnostic and Statistical Manual (DSM) and the International Classification of Diseases (ICD) as well as contributing to the clinical management of children and adolescents who self-harm. Longitudinal studies of self-harm rarely distinguish between self-harm acts carried out with and without intent; the proposed research will provide guidance for the research community on whether such distinction is valuable. Furthermore, our linkages with HES and GPRD will provide a unique opportunity to assess the potential for bias in other studies of self-harm that rely on self-report or hospital presenting cases alone. This will allow researchers to make judgments about the impact of missing data and under-reporting on their findings.


Guidance for parents, the voluntary sector and other professional groups:

Parents, friends, teachers, school counsellors and healthcare professionals may all be affected by a young person's self-harming behaviour and want advice about how best to help the young person, the medium term risk associated with their child's self-harm and likely outcomes. Evidence in this area is currently limited. People often turn to organisations such as PAPYRUS (http://www.papyrus-uk.org/) or Samaritans; the proposed research will strengthen the evidence base for the advice these organisations can provide.

Contributing to increased public awareness about self harm and its risk factors will, we anticipate, reduce the stigma associated with self-harm / mental illness and increasing the likelihood that individuals will seek help and receive support from those from whom they seek help.


Impact on children's health and education:

Self-harm, and difficulties associated with it, are a cause of time off school in adolescence and this is likely to impact on children's academic achievement (we will investigate this in our proposed study). Improved knowledge to inform better prevention and treatment of this condition will likely improve children's educational performance and strengthen the UK's future workforce.


Clinical community:

There is a gap in knowledge concerning the causes and consequences of community-presenting self-harm and whether these differ for adolescents who self-harm with and without suicidal intent. Such information may improve risk stratification during clinical assessment and in prioritising referrals to Child and Adolescent Mental Health Services (CAMHS) teams. Improved knowledge in this area will therefore benefit both front-line clinicians / CAMHS teams as well as the NHS by allowing services to focus on those at greatest need. Further, there have been calls to screen for community presenting self-harm; by clarifying the medium-term health consequences of self-harm presenting and not presenting to health services our research will clarify the potential benefit of screening programmes.


Development of capacity and capability:

ALSPAC is one of the largest prospective studies capable of assessing the causes and consequences of self-harm in the current generation of young adults and the only such study to our knowledge in the UK. Improved knowledge about the aetiology and prognosis of self-harm with differing underlying motivations within ALSPAC will considerably enhance the value of ALSPAC for future studies of self-harm as well as informing future data collections and use of the ALSPAC resource by other researchers. The researcher employed on the study will develop skills through their analysis, write-up and presentation of study findings and workshop organization.