777 Sleep and non-suicidal self-injury in college students
Abstract Introduction Non-suicidal self-injury (NSSI) can increase suicide risk and is highly prevalent among young adults, including college students. While there is mounting evidence that disrupted sleep increases suicide risk, it is unclear how sleep influences NSSIs. Therefore, the present study explored how sleep variables were associated with NSSIs in a college sample. Methods Data from N=506 respondents were collected as part of the Assessing Nocturnal Sleep/Wake Effects on Risk of Suicide (ANSWERS) Survey of college students. The primary outcome, lifetime NSSI, was assessed using a self-report question derived from the Columbia Suicide Severity Rating Scale. The predictors were weekday short sleep (≤ 6h; assessed by retrospective sleep diary), Insomnia Severity Index (ISI) score, Brief Inventory of Sleep Control (BRISC) score, and Disturbing Dreams and Nightmares Severity Index (DDNSI) score. Binomial logistic regression models estimated the associations between NSSI and sleep variables in models that were unadjusted, adjusted for age, sex, race, and ethnicity, and additionally adjusted for thwarted belongingness or perceived burdensomeness from the Interpersonal Theory of Suicide. Results A total of N=142 (28.1%) respondents endorsed lifetime non-suicidal self-injury. Individuals with NSSI were more likely to be female (p=0.015), in poorer health (p<0.001), and have more severe depression (p<0.001) and anxiety (p<0.001) than those without NSSI. In unadjusted models, higher BRISC scores were associated with lower odds of NSSI (OR 0.55 [0.43–0.71]), DDNSI scores of >=10 increased the odds of NSSI (OR 2.65 [1.70–4.11], and ISI scores of >=8 increased the odds of NSSI (OR 2.05 [1.38–3.08]), while short sleep was not associated with NSSI. Adjusting for age, sex, race, ethnicity, and thwarted belongingness did not eliminate any of these relationships but adjusting for perceived burdensomeness rendered the association between insomnia and NSSI non-significant. Conclusion Individuals with significant insomnia symptoms or nightmares were more likely to report a history of NSSI, while individuals with greater perceived control of sleep had lower odds of NSSI. These findings were generally independent of the Interpersonal Theory of Suicide. Further research is needed regarding the timing of NSSI (i.e., do they occur more often during nocturnal wakefulness) and whether sleep interventions can reduce the risk of NSSI. Support (if any):