A comparison of retrospectively reported and ecological momentary assessment-reported perceived social support in predicting ecological momentary assessment-reported non-suicidal self-injury.

Introduction: Non-suicidal self-injury (NSSI) urges and behaviors are associated with lower perceived social support and related constructs (e.g., perceived rejection). However, no studies have examined the concordance of retrospective (baseline) and ecological momentary assessment (EMA) perceived social support assessments. Retrospective and EMA reports are often only weakly to moderately correlated; measurement approaches may, therefore, impact observed associations between variables. We tested whether average EMA-reported perceived emotional social support uniquely predicts EMA-reported NSSI urges and behaviors above baseline-reported retrospective self-report of perceived emotional social support alone.

Methods: 93 young adults (ages 18-34) with past-month NSSI urges or behaviors and lifetime NSSI behaviors completed a semi-structured interview, self-report surveys, and a 2-week EMA protocol.

Results: Baseline- and EMA-reported perceived emotional social support were positively correlated (Kendall's tau-b = 0.51). Average EMA-reported social support was uniquely associated with EMA-reported NSSI urges but not NSSI behaviors.

Conclusions: EMA-reported perceived emotional social support captured information not represented by baseline reports alone, but improvement in model fit was modest. EMA-reported social support may further improve the estimation of EMA-reported NSSI urges if modeled as a proximal predictor of NSSI. Further work is needed to clarify temporal directions between social support and NSSI urges. Limitations are discussed.