Predictors and mediators of self-harm in females with ADHD

A generation ago, ADHD was thought to be an almost exclusively male disorder.^{Reference Hinshaw2} Like nearly all neurodevelopmental conditions, ADHD is more common in boys than girls (the ADHD male/female ratio is 2.5:1, although this is lower by adulthood). When ADHD does present in females, it is associated with significant lifelong impairments, including poor health-related outcomes, low achievement (especially in mathematics), key problems with employment beyond secondary education and high rates of unplanned pregnancy.^{Reference Hinshaw2} Further, many girls with ADHD (particularly those with high impulsivity in childhood) follow a heterotypically continuous developmental pathway, such that they are at a strikingly high risk for self-harm, including both NSSI and suicide attempts.^{Reference Hinshaw2,Reference Beauchaine, Hinshaw and Bridge3}

Given the high rates of suicide globally, escalating in many countries (and especially among adolescent girls), a serious public health issue is at hand. Of even further concern, internalising disorders and suicide rates may well spike in the next few years, given increases in economic stress, social isolation, and other toxic psychosocial stressors associated with the COVID-19 pandemic. Herein, we focus on factors related to the pronounced rates of self-harm in girls with ADHD as they mature into adolescence and early adulthood.

The Berkeley Girls with ADHD Longitudinal Study represents the largest prospectively followed sample of girls with ADHD (plus an age- and ethnicity-matched comparison sample) in existence.^{Reference Hinshaw2} The investigation began during the sample's childhood and features 92–95% retention rates throughout three follow-up evaluations, extending through the age range of mid-to-late 20s.

At every evaluation, girls with ADHD showed core impairments, often in areas commensurate with those of boys with ADHD followed over time (e.g. increased need for special school services, high rates of comorbidity, excess parental distress and peer rejection). Specifically regarding self-harm, by the end of adolescence, participants who had entered the study with the combined presentation of ADHD (i.e. displaying clinically significant inattention plus hyperactivity and/or impulsivity) had rates of attempted suicide (22%) and engagement in moderate-to-severe NSSI (51%) 2.5–3.5 times the rates found in the comparison sample (for whom rates were statistically equivalent to rates within the childhood inattentive presentation). Crucially, girls with ADHD who had also experienced childhood maltreatment (physical abuse, sexual abuse and/or neglect) had a substantially higher risk of attempted suicide (33% had made an attempt) by early adulthood than those without a maltreatment history (13%) or than the matched comparison sample (6%).^{Reference Hinshaw2,Reference Beauchaine, Hinshaw and Bridge3} Thus, even for a condition as heritable as ADHD, exposure to early abuse experiences substantially compounds risk for self-harm. Such findings parallel another condition with strong heritability, bipolar disorder. The core message is that heritable risk is compounded by psychosocial adversity in predicting self-harm-related outcomes.

Additional childhood predictors of later self-harm included severity of ADHD symptoms, low self-esteem, comorbidity with internalising and externalising conditions, negative parent–child interactions with fathers (but not mothers) and poor executive functioning.^{Reference Meza, Owens and Hinshaw4} With respect to adolescent mediator processes, pathways to self-harm were relatively specific to self-harm type. That is, adolescent externalising symptoms and difficulties with response inhibition mediated the link between childhood ADHD and NSSI, whereas adolescent internalising symptoms mediated the link from childhood ADHD to suicide attempts. In addition, peer victimisation mediated the link between childhood ADHD and NSSI by early adulthood, whereas peer rejection mediated the link to suicide attempts. Overall, beyond intraindividual factors (impulsivity, comorbidity), contextual processes related to families (maltreatment) and peer groups (rejection, victimisation) are crucial regarding developmental pathways. We hasten to add, however, that adolescent NSSI is a potent risk factor for later suicide attempts, so that any distinction between non-suicidal behaviours and frank suicide attempts is not absolute.

Evidently, females with ADHD show significant impairments with respect to a cascading set of problems from childhood through adulthood. Moreover, evidence is converging on the finding that vulnerabilities conferred by high levels of (heritable) impulsivity, and negative family and peer-related environments and experiences, can precipitate significant impairment and psychopathology, particularly in the realm of self-harm.

We highlight, as well, that females with ADHD are at increased risk for unplanned pregnancy. The key adolescent mediator underlying the link between childhood ADHD and eventual unplanned pregnancy is low academic performance.^{Reference Owens and Hinshaw5} Thus, in addition to such factors as comorbid psychopathology, executive functions, family factors and peer-related processes, promotion of academic achievement appears to be a core target for preventive intervention.

Still, too little is known about later-adult outcomes of females with ADHD. Indeed, more distal end-points, such as the development of borderline personality disorder (e.g. affective instability, chronic suicidality) in adulthood, require additional longitudinal research to supplement existing cross-sectional research.^{Reference Beauchaine, Hinshaw and Bridge3} Moreover, the intergenerational transmission of risk for similar negative outcomes in the offspring of females with ADHD remains a priority area for future research efforts. Overall, we contend that the extraordinarily high risk for self-harm incurred by girls with ADHD as they mature requires a shift in clinical perspective.

Clinical value

Despite the growing body of evidence that females with ADHD are at high risk for self-harm, no evidence-based prevention programmes for self-harm currently exist for preadolescents.^{Reference Beauchaine, Hinshaw and Bridge3} Once established, self-harmful behaviours are self-reinforcing and difficult to treat. Access to specialised treatment (e.g. dialectical behaviour therapy) remains low because of the limited number of providers, expense and significant time requirements.

Recent research suggests that the age of initiation of NSSI may occur before the teenage years.^{Reference Beauchaine, Hinshaw and Bridge3} We therefore contend that preventive interventions should occur in late childhood, before the initiation of self-harm. The extreme pain and suffering associated with self-harm, along with cascading problems related to untreated psychopathology, support the development of early self-harm prevention programmes as a public health priority.

Existing interventions for ADHD include behavioural, pharmacologic and multi-modal treatments that typically yield improvements in core symptoms but are often time-limited and tend not to alter fundamental developmental trajectories. We believe that treating trait impulsivity and inattention, the core components of ADHD, is needed. Further, emotion dysregulation—the inability to modulate negative affective states such as anger, fear and sadness, in the service of goal-directed behaviour—should be a core preadolescent prevention target for individuals at risk.^{Reference Beauchaine, Hinshaw and Bridge3} Emotion dysregulation is shaped through negative environmental experiences such as maltreatment—a critical prevention target within itself—along with peer victimisation/rejection and negative parenting practices, including invalidating family environments. When coupled with underlying vulnerabilities (e.g. trait impulsivity), such contextual factors promote high levels of self-harm.^{Reference Beauchaine, Hinshaw and Bridge3} By focusing on emotion dysregulation as a prevention target, intervention must emphasise increasing emotion-related skills, along with improving parent–child relationship dynamics and enhancing peer relationships. Furthermore, by increasing emotion regulation capacities in preadolescence, cascading problems such as the risk for the development of internalising/externalising comorbidities (and thus subsequent self-harm) may be attenuated.

Recommendations for assessment

It is critical to educate clinicians to recognise ADHD and its different presentations, particularly for girls. Accurate, early diagnoses may prevent long-term, detrimental and cascading deficits, including self-harm. Medical practitioners prescribing stimulants for ADHD, as well as psychologists providing parent management, should make screening for emotion dysregulation and early signs of self-harm a common practice, following similar steps taken toward universal depression screening for adolescents recommended in 2018 by the American Academy of Pediatrics.^{Reference Zuckerbrot, Cheung, Jensen, Stein and Laraque6}