(PS9-19) Perceived Burdensomeness and Social Contribution: Examining Word Choices When Speaking to a Driver of Suicidal Thoughts and Behaviors

Abstract Text: Perceived burdensomeness, a theory-based driver of suicidal thoughts and behaviors (SITBs), can be understood in a transactional model between perceived contributions and perceived liability to others. Relatedly, individuals with depression who are at elevated risk for developing SITBs have been known to exhibit general memory deficits such as greater abstraction and lesser richness, while simultaneously recalling negative memories with greater phenomenological specificity than positive memories. The context in which memory deficits associated with clinical depression are reflected through word choice and whether suicidal language is unaffected by these deficits has yet to be investigated. The current study aimed to evaluate the differences in language patterns when speaking about a positive memory focused on protective content compared to a negative memory focused on risk content. Undergraduates (N = 114; Mage = 19.41, SD = 1.56; 62.28% female) all scoring above the cutoff for clinical depression (M = 22.31, SD = 3.55) recounted episodic autobiographical memories of events in which they perceived themselves to be burdensome on (PB) or socially contributing (SC) to others. Frequencies of words reflective of sensory details, time orientation, and suicidal language were quantified through Linguistic Inquiry and Word Count (LIWC) software. Accordingly, a PERMANOVA was run to examine frequency mean differences, after controlling for word count, between the PB and SC prompts. While the overall PERMANOVA was nonsignificant, F(2, 221) = 1.59, p = 0.193, R² = 0.014, indicating that complete linguistic categories of sense, time, and suicide were not distinct in either prompt, Wilcoxon signed-rank tests found facets of sensory details and suicidal language, but not time orientatation, differed significantly between prompts. Specifically, attention (Cohen’s d = 0.41, p < .05, MSC = 0.64, MPB = 0.21) and visual (Cohen’s d = 0.43, p < .05, MSC = 1.03, MPB = 0.53) words were significantly more frequent in the SC prompt, while feeling (Cohen’s d = -0.76, p < .001, MPB = 2.14, MSC = 1.08) words were significantly more frequent in the PB prompt. Further, the SC prompt included significantly more frequent collective (“we”: Cohen’s d = 0.62, p < 0.001, MSC = 1.68, MPB = 0.74), social (Cohen’s d = 1.28, p < 0.001, MSC = 18.63, MPB = 11.70), and positive (Cohen’s d = 0.63, p < 0.001, MSC = 1.40, MPB = 0.56) words, while the PB prompt contained significantly more frequent self-referential (“I”: Cohen’s d = -1.26, p < 0.001, MPB = 11.68, MSC = 7.60), generally negative (Cohen’s d = -0.38, p < 0.05, MPB = 1.50, MSC = 1.01), and anger (Cohen’s d = -0.40, p < .001, MPB = 0.31, MSC = 0.09) words. These results are the first to indicate the language patterns involved with protective and suicide risk content among depressed young adults, and provide the impetus for more fine-grained language analysis among at-risk populations. Accounting for cognitive impairments associated with depression and incorporating how the encoding and rich recall differs between protective and risk content provides insight into how an individual may be interpreting and remembering psychotherapy content and a promising avenue toward strengthening the long-term effects of targeted treatments.