Understanding Self-Harm in Youth: Risks, Consequences, and Feasibility of Targeted Family Treatment

Abstract in English

Self-harm is common in youth and has significant psychosocial impacts on both young people and their parents. While some engage in self-harm only once, others repeat the behaviour. Although most cease during adolescence, a substantial minority persist into adulthood. The overall objectives of this doctoral research project were to examine risks and consequences associated with youth self-harm for both young people and their caregivers, and to evaluate the feasibility of a targeted psychosocial treatment. Study I was a population-based cohort study including 77,647 individuals with a first hospital-treated self-harm episode, i.e., treated in inpatient or outpatient specialised healthcare, between ages 10 and 24 years. Data were collected from national registers. We estimated cumulative incidence and incidence rates of subsequent self-harm, suicide attempt, and suicide at 1 month, 3 months, and 1 year after discharge. Within one year, cumulative incidence was 17.25% (95% confidence interval 16.98–17.52) for subsequent self-harm, 8.33% (8.14–8.53) for suicide attempt, and 0.26% (0.23–0.30) for suicide. Risks were highest during the first month after discharge. Rates of subsequent self-harm and suicide attempts peaked within the first 15 days and declined sharply thereafter, while suicide trends were less clearly defined. Study II was also a population-based cohort study, that included 18,111 individuals with a first hospital-treated self-harm episode between age 10 and 18 years. Data were collected from national registers. Participants were followed for six years from age 19 years to examine persistence of self-harm into early adulthood and associated risk factors. Overall, 2,639 individuals (14.6%) experienced persistent self-harm. Psychiatric history was the strongest risk factor, with hazard ratios for any psychiatric disorder of 2.88 (2.58–3.22) for females and 2.73 (2.26–3.28) for males; for prior psychiatric hospitalisation, hazard ratios were 2.90 (2.67–3.16) for females and 2.44 (2.00–2.97) for males. Violent crime and poor school performance were associated with higher risk of persistence in both sexes, particularly among males. Assault victimisation increased risk for both sexes, whereas parental suicide was a statistically significant risk factor only for females. Among females, a first self-harm episode involving tissue damage was associated with higher persistence risk, while poisoning carried the greatest risk among males.

Study III was a matched cohort study estimating work absence in parents of youth who self-harm, in comparison both with parents of youth without selfharm and themselves before and after the first hospital-treated self-harm event. The study included 176,472 mothers and 161,833 fathers of 17,726 youths who self-harmed at least once during adolescence and 177,260 matched youths without evidence of self-harm. Data were collected from national registers. Differences in work absence, including family leave to care for a sick child and sick leave, were estimated using conditional Poisson regression. Offspring selfharm was associated with increased family leave for both mothers and fathers, with rate ratios of 3.47 (3.25–3.72) and 2.71 (2.47–2.98), respectively. Sick leave was also more common, with rate ratios of 1.25 (1.20–1.31) for mothers and 1.25 (1.17–1.33) for fathers. Parents of self-harming youth took more family leave during the self-harm year compared with the previous year (mothers: 1.65 [1.55– 1.75], fathers: 1.41 [1.29–1.54]), with no corresponding increase in sick leave. Study IV was a randomised feasibility trial of the suicide attempt prevention family treatment Safe Alternatives for Teens and Youths (SAFETY). Thirty youths with recent suicidal behaviour were randomised to 12 weeks of SAFETY or Supportive Therapy, an active control treatment. Both treatments showed high compliance, satisfaction, credibility, and session completion, with few adverse events, dropouts, and low attrition. At 3-month follow-up, two participants (14%) in SAFETY and four (27%) in Supportive Therapy had attempted suicide. Nonsuicidal self-injury decreased by 95% (rate ratio 0.05 [0.01–0.20]) in SAFETY and 69% (0.31 [0.11–0.83]) in Supportive Therapy. Participation in SAFETY, but not Supportive Therapy, was associated with improvements in anxiety and depression, quality of life, and emotion dysregulation. In summary, among young people with a first hospital-treated self-harm episode, the risks for subsequent self-harm, suicide attempt, and suicide were highest during the first month after discharge but remained elevated throughout the first year. For most individuals, self-harm was limited to adolescence. However, about one in seven continued into early adulthood. Persistence was strongly associated with prior psychiatric vulnerability and showed sex-specific risk patterns, with stronger links to educational difficulties and externalising behaviours among males. Parents of youth who self-harm had higher rates of work absence due to family leave and sick leave compared with other parents, and a first self-harm episode was associated with a peak in family leave. Finally, SAFETY appears to be a feasible and acceptable treatment for youth with suicidal behaviour and their parents.