Self-harm among children and adolescents has become more prevalent in the general population over the last decades and is a major public health concern. Furthermore, self-harm episodes are one of the strongest risk factors for future suicide attempts and suicide which highlights the importance of understanding who are the children and adolescents at an increased risk to self-harm. This doctoral thesis focused on assessing: 1) the prevalences of self-harm among children and adolescents, 2) describing the characteristics of children and adolescents who self-harm in comparison to those who do not and 3) exploring a variety of environmental, sociodemographic, and mental health factors as well as developmental challenges, neurodevelopmental conditions, and genetic predispositions to investigate who are the children and adolescents particularly vulnerable to self-harm. In this thesis, we used data from the Norwegian Mother, Father and Child Cohort Study (MoBa), together with diagnostic and sociodemographic data linked from national registries. The MoBa cohort is one of the largest longitudinal cohort studies worldwide and this doctoral thesis includes one of the largest explorations to date of self-harm behaviours in children as young as 3 years.
In the first paper (paper I), 35,494 mothers responded to the questionnaire item from the Social Communication Questionnaire whether their child self-injured at the ages of 3 and 8 years: “Does your child ever injure himself/herself deliberately, such as by biting his/her arm or banging his/her head?”. We investigated patterns across family sociodemographic factors, occurrences of developmental difficulties in the children, rates of mental health diagnoses or neurodevelopmental conditions among children and parents, adverse life experiences of children and mothers and parental health-related behavioural factors. For the comparison, we created four distinctive groups: children who did not self-injure neither at 3 nor 8 years (no self-injurious behaviour), children who self-injured at 3 years only (transient self-injurious behaviour), children who self-injured at 8 years only (emergent self-injurious behaviour) and children who self-injured at 3 and 8 years (persistent self-injurious behaviour). The results revealed patterns across the characterisation variables, such as mental health and neurodevelopmental factors, which were more prevalent across all the self-injury groups compared to the no self-injurious behaviour group. Specifically, we saw that the emergent and persistent self-injurious behaviour groups differed the most from the no self-injurious behaviour group, including rates of variables which were standing out the most: paternal depression diagnoses and reports of children's motor developmental delay and sleep problems.
In the second paper (paper II), self-harm behaviour questionnaire data was available for up to 42,453 genotyped children and 24,731 genotyped family-trios in MoBa. In paper II, we assessed associations between polygenic indices for depression, anxiety, schizophrenia, posttraumatic stress disorder (PTSD), attention-deficit hyperactivity disorder (ADHD), substance use disorder, autism, bipolar disorder, anorexia nervosa, neuroticism, cognitive functioning and suicide attempts on self-harm behaviour measures in childhood and adolescence. Family trio analyses were applied to isolate direct genetic effects on self-harm behaviour. The results for the total and sex stratified genetic effects revealed, with varying magnitude, consistent patterns of higher genetic liabilities to ADHD, PTSD, depression and suicide attempts to self-harm behaviour across ages and measures. Some genetic liabilities were particularly pronounced among boys at 8 years, e.g., to substance use disorder or neuroticism, while others were associated with higher odds in girls at 16 years, e.g. to substance use disorder on non-suicidal self-harm and to ADHD on suicidal self-harm. In the within-family trio analyses, we found only evidence for direct genetic effects for genetic liabilities to suicide attempts, PTSD, neuroticism, depression and ADHD with non-suicidal self-harm at 16 years. We did not find any evidence for parental indirect genetic effects, though these finding might be limited by sample size.
In the third paper (paper III), we investigated the direct association between bullying victimisation and later self-harm in adolescence as well as potential moderators of these associations. In this study, 12,752 adolescents reported on bullying victimisation at 14 years and non-suicidal and suicidal self-harm at 16 years. We selected moderator variables based on previous research and contributions from user organisations. The assessed moderators were sex registered at birth, sexual minority status, mental health conditions (depression, anxiety and eating disorders), neurodevelopmental conditions (ADHD and autism) and social relationship factors (parent-child relationship closeness, parent-child conflict, pro-social behaviour and having close friend(s)). Almost one third (28.4%) of the adolescents in our study sample reported to have self-harmed in the past, including both, suicidal and non-suicidal self-harm. Among those, 5.5% were reporting that they self-harmed with suicidal intent (suicidal self harm). Over 40% of those who self-harmed, also reported that they experienced to be bullied at least sometimes, and almost 1 in 10 of adolescents who reported to self-harm were bullied at least weekly. Our findings revealed significant associations between bullying victimisation and self-harm outcomes. The moderation analyses showed that the impact of bullying victimisation on suicidal self-harm was significantly lower in females (OR, 4.06 [95% CI, 2.97 7.04]) compared to males (OR, 12.05 [95% CI, 4.91-29.55]). The models did not reveal any other significant interaction effects for the remaining moderators. The study also revealed increased rates of self-harm among vulnerable groups. Almost three-quarters of adolescents diagnosed with depression reported to self-harm. We also found high rates among those diagnosed with anxiety or eating disorders. About half of adolescents with ADHD and autistic adolescents reported to self-harm. Among sexual minority youth, self-harm was prevalent broadly in about 50% to 70% per sub-category, and in just under 40% of those who were questioning their sexual orientation. Paper III shows that self-harm is common among adolescents and highlights the increased self-harm rates among females, neurodivergent and sexual minority adolescents and those who were diagnosed with a mental health condition. Our evidence suggests that if the relationship between bullying victimisation and self-harm is causal, interventions to prevent bullying victimisation will reduce the risk of self-harm across many different groups of adolescents.
Overall, this thesis assessed the relevance of self-harm in childhood and adolescence by investigating its prevalences, describing the characteristics of those who self-harm and exploring specific vulnerabilities to self-harm. The inclusion of self-harm measures in children as young as 3 years old, in combination with measures at 8 and 16 years, helps to increase our understanding of how prevalence and risk factors change over age. The results also point out the impact of adverse life experiences such as being bullied and the increased genetic liability to suicide attempts, PTSD, neuroticism, depression and ADHD as risk factors for self-harm, particularly in adolescence. Knowledge of risk factors and vulnerabilities from early age onwards, is important to guide future research and improve interventions to decrease self-harm rates and prevent future mental health challenges associated with self-harm among children and adolescents.