Objective
As a public concern, the risk of non-suicidal self-injury (NSSI) is high among college students. Based on the benefit barrier model of NSSI and the autobiographical memory vulnerability/scarring hypothesis, positive memory are potentially associated with NSSI in college students. This study investigated concurrent and lagged associations between NSSI urge and positive memory characteristics (count, specificity, vividness, coherence, accessibility, time perspective, sensory details, visual perspective, emotional intensity, sharing, distance) in college students’ everyday life, as well as associations between baseline positive memory characteristics and NSSI urge over time.
Methods
70 college students (Mage = 19.51 years, 43.7% female) with NSSI behaviors within the last year participated in this study, who completed the baseline and 20 daily measures of NSSI urge and positive memory characteristics. Multilevel linear models and dynamic structural equation models examined concurrent and lagged associations between NSSI urges and positive memory characteristics. Multilevel linear growth models examined the association between baseline positive memory characteristics and NSSI urge over time.
Results & Discussion
The results indicated that days with higher NSSI urge were associated with lower vividness (β = -0.87, p ˂ 0.001), lower coherence (β = -0.92, p ˂ 0.001), lower accessibility (β = -1.34, p ˂ 0.001), lower time perspective (β = -1.01, p ˂ 0.001), lower sensory details (β = -0.89, p ˂ 0.001), lower visual perspective(β = -0.67, p ˂ 0.001), lower emotional intensity(β = -0.87, p ˂ 0.001), lower sharing(β = -1.04, p ˂ 0.001) and further distancing(β = -0.58, p ˂ 0.001). Cross-lagged analyses showed that lower vividness (β = -1.18, p ˂ 0.001), lower coherence (β = -1.05, p ˂ 0.001), lower accessibility (β = -0.73, p ˂ 0.001), lower sensory details (β = -0.83, p ˂ 0.001), and lower visual perspective (β = -0.75, p ˂ 0.001) of positive memory on the previous-day significantly predicted higher NSSI urges on the next-day, whereas none of the predictive effects of the the previous-day NSSI urge on the next-day positive memory characteristics were significant. There was a significant main effect of time (β = -0.20, p = 0.044) on NSSI urge, significant main effects of baseline positive memory count (β = -0.04, p = 0.026), specificity (β = -0.39, p = 0.042), vividness (β = -0.05, p = 0.044), accessibility (β = -0.06, p = 0.037), and distance (β = 0.04, p = 0.029) on NSSI urge, and a significant interaction between baseline positive memory distance and time on NSSI urges (β = 0.004, p = 0.045), with NSSI urges declining slower over the 20-day period when baseline positive memory distance was further (simple slop = -0.19, p = 0.042).
Conclusions
The study supports the NSSI benefit-barrier model and the autobiographical memory vulnerability hypothesis, that positive memory characteristics are associated with college students' everyday NSSI urge, and that some positive memory characteristics predict the occurrence of NSSI urge. Therefore, facilitating access to specific positive autobiographical memories, adopting a first-person perspective and increasing the level of detail and vividness in recall, and reducing psychological distance may be potential interventions for college students’ NSSI problems.
 

Non-suicidal self-injury,positive autobiographical memory,daily diary designs,college students