(SSCP1-2) SSCP Submission: Inconsistent Reporting of past Self-injurious Thoughts and Behaviors in Children: A Longitudinal Analysis of Data from the Adolescent Brain Cognitive Development Study

In longitudinal research with adolescents and adults, approximately one-third of individuals who report self-injurious thoughts and behaviors (SITBs) fail to report this history in the future. This pattern has critical implications for the utility of self-reported histories of SITBs for safety assessments in clinical and research contexts. However, very little is known about this phenomenon in children or its implications. Thus, this study aimed to 1) estimate the prevalence of inconsistent reporting of SITBs in an epidemiologically-informed sample of children in the United States; 2) examine potential sociodemographic, clinical, neurocognitive, and brain structure correlates of inconsistent SITB reporting; and 3) examine developmental differences in inconsistent reporting by comparing two partially-overlapping samples reporting SITBs at different points in time. 

Data are from the baseline (BL, participant ages 9-10), Year 1 (Y1, ages 10-11), and Year 2 (Y2, ages 11-12) assessments of the Adolescent Brain Cognitive Development Study. Lifetime SITBs were assessed using a computerized clinical interview at each assessment. To address the study aims, 1) the prevalence of inconsistent reporting was estimated across a one-year interval from BL to Y1 among those who reported any SITB at BL (n BL-Y1 = 1350); 2) logistic regressions modeled candidate sociodemographic, clinical, neurocognitive, and neurobiological (e.g., brain structure) correlates at BL with inconsistent reporting from BL-Y1 correcting for multiple comparisons; and 3) inconsistency rates from Y1 to Y2 were estimated among those who reported any SITB at Y1 (n Y1-Y2 = 1086) and were compared to inconsistency rates from BL-Y1 using a Z8 test for partially-overlapping samples.

At Y1, 67% of SITB reporters from BL did not endorse past SITBs (inconsistency range = 68-80% when examining suicide ideation, suicide attempts, and non-suicidal self-injury separately). Younger age (OR = 0.977, p = 0.044), lower parent-reported clinical problems, including history of depressive disorder (OR = 0.667, p = 0.030), internalizing problems (OR = 0.956, p < 0.001), and total clinical problems (OR = 0.988, p < 0.001), and lower general neurocognitive ability (OR = 0.790, p = 0.008) were significantly associated with higher odds of inconsistent reporting. While differences between reporting periods were not found for SITBs in general, inconsistency in reporting of non-suicidal self-injury was significantly lower from Y1-Y2 (73%) compared to BL-Y1 (80%) (Z = -3.101, p = 0.002).

In late childhood, inconsistency in reported SITBs is the norm among those who would have reported at a previous point in time, which may hinder accurate risk assessment for youth. Our results suggest that features of development (e.g., age, neurocognitive abilities) as well as clinical severity are related to patterns of inconsistent reporting, though whether these factors are causal is not known. Thus, to improve our ability to accurately measure youth SITB histories in a clinical and research context, a task that is critical for effective suicide prevention efforts, further examination of the mechanisms that drive inconsistent reporting, such as forgetting, reconceptualizing, or concealment, is needed.