A Mixed-Methods Analysis of Self-Harm in a Longitudinal Study of Young Women with Childhood ADHD

Objective: Despite over fifty years of research on self-harming behavior, including both non-suicidal self-injury (NSSI) and suicidality, the need for continued research and interventions is imperative, as suicide rates have increased for adolescents and young adults over the past decades. In addition to cross sectional and longitudinal research, laboratory-based and micro-longitudinal studies have furthered our knowledge regarding more proximal, contextual factors that lead to self-harm behaviors. An additional approach, which is well-suited to discovering such proximal and contextual factors, involves qualitative interviews. A small but important literature supports the benefits of integrating qualitative with quantitative methods for mental health care research and practice – an approach termed “mixed-methods” strategies. My overall aim is to perform a mixed-methods investigation to better understand contextual/proximal factors associated with self-harm as well as its onset, persistence, and desistance. In addition, I use existing longitudinal, quantitative data to predict themes and codes generated from qualitative, self-harm interviews of young women with and without childhood attention-deficit/hyperactivity disorder (ADHD). An important goal is to capture participants’ unique perspectives, allowing them to describe their self-harm experiences in their own words. Method: The sample emanates from the Berkeley Girls with ADHD Longitudinal Study (BGALS), comprising the largest sample of girls with childhood-ascertained ADHD in existence. It contains four waves of quantitative data (mean ages of 9.5, 14.5, 19.5, and 26) regarding the sample of 228 women, both with and without childhood diagnosed ADHD. From this larger sample, a subsample of 57 young women, with a range of ADHD symptoms and self-harming behaviors, were recruited to take part in qualitative, semi-structured, conversational interviews. Interviews ranged from 90 minutes to 3 hours, covering a range of topics. Interview questions for this dissertation focused primarily on self-harm behavior. Interviews were electronically recorded, transcribed, and then indexed for thematic content before being coded. These data were then analyzed on the basis of existing quantitative data, including ADHD diagnostic status, self-harm behavior, and a measure of behavioral impulsivity. Results: Several self-harm themes were identified a priori, driven by existing research, and additional themes emerged from the qualitative interviews. Inter-rater reliability for coded transcripts was high (kappa > .80 across all coders). Excerpts of interviews are provided to further support richness of participant responses. Of the 39 participants whose interviews were fully coded and analyzed, 22 reported engaging in self-harm behavior. Almost all (20 out of 22) reported engaging in repetitive NSSI, and nine reported making a suicide attempt. Consistent with existing theories, several themes emerged regarding functions of self-harm behavior, including emotional dysregulation, attempts to gain attention from others, personal agency, and a belief that they deserved pain/punishment. Additional themes included reasons for desisting in self-harm, the relation between impulsivity and self-harm, as well as themes more specifically related to suicide. Several patterns emerged regarding the onset and course of self-harm behavior. Four self-harm codes were also identified, whereby numerical values were assigned to participant responses based on extent of response/level of detail provided. Exploratory, mixed-methods analyses were also employed, revealing that adolescent impulsivity was associated with greater severity/impact of later qualitatively-reported self-harm behavior. Conclusions: Qualitative, mixed-methods approaches offer an important mechanism through which individuals can provide rich context about important life experiences, such as self-harm. Several themes discussed by participants in the present study were complementary to existing research and theories of self-harm. In addition, the present study also describes several patterns in relation to self-harm onset, maintenance, and desistance that are novel. Recent research demonstrates the increased risk of self-harm behavior for girls with ADHD, and the present study underscores the need for earlier screening and intervention, given the nature and course of such behavior in individuals with ADHD and comorbid psychopathology. The present study also supports the crucial need for additional screening and school-based prevention programs, in addition to increased parent-education, suggesting that such programs might be indicated at the middle school-level