(PS2-A21) Suicide Attempt Survivors Are More Likely to Exchange Text Messages Containing Negative Emotion Words with Peers and Significant Others Relative to Family

Background: Strong interpersonal relationships are crucial for young adults’ socio-emotional functioning and can buffer against adverse mental health outcomes. Thus, elucidating young adults’ relationship dynamics may inform prevention and intervention initiatives for outcomes such as depression and suicide. Much of young adults’ social connection occurs in the context of digital communications (e.g., via text message). Understanding to whom young adults choose to communicate emotionally is valuable in that it may reflect which relationships are characterized by a greater level of emotional intimacy and disclosure. The present study examined how the emotional tone of young adults’ (who have made a prior suicide attempt) personal text messages vary based on the relationship partner. We tested whether the emotional tone of texts differs when individuals are texting their: 1) family members, 2) friends, or 3) significant others. The analytic plan and hypotheses were pre-registered: osf.io/sgk3m. It was hypothesized that young adults would exchange more text messages containing words classified as negative emotion, anxiety, and sadness with friends and significant others, relative to family. As an exploratory analysis, we also tested whether young adults would exchange more text messages containing themes of anger or positive emotion among friends, significant others, and/or family.

Method: Participants were young adults (N = 26; Mage = 20.89; SDage = 2.53) with a history of at least one prior, non-lethal suicide attempt. Linguistic Inquiry Word Count (LIWC) was used to label the text messages (N = 414,032), and then a series of multilevel generalized linear models were run to test whether participants exchange more text messages containing negative emotion/anxiety/sadness words with friends and significant others, relative to family.

Results: As hypothesized, the likelihood of a text containing negative emotion, anxiety, or sadness-related words was greater when participants were texting friends (ORs > 1.66; ps < .001) and significant others (ORs > 1.60; ps < .001), compared to when they were texting family members. Results for exploratory analyses were consistent with the directional hypotheses, such that participants were also more likely to exchange texts containing themes of anger or positive emotions when texting friends (ORs > 1.25; p</span> < .001) and significant others (ORs > 1.23 p < .001), relative to family.

Conclusion: This study is the first to provide objective evidence that young adults tend to communicate more negative emotions with their peers, relative to their family members, in the context of personal text messages. These findings reinforce the importance of peer relationships and underscore the need to consider these relationships when conceptualizing risk for suicide or other mental health difficulties.