Perceived stress, emotion regulation, sleep quality, and executive dysfunction among operating room nurses: a cross-sectional structural equation modeling study

Operating room (OR) nurses routinely perform complex tasks under considerable pressure. This study examined how perceived stress, emotion regulation, and sleep quality were associated with self-reported executive dysfunction in this population. In this cross-sectional study, 216 operating room nurses from tertiary hospitals in Chongqing, China, completed the Dysexecutive Questionnaire, Perceived Stress Scale, Emotion Regulation Questionnaire, and Pittsburgh Sleep Quality Index. Structural equation modeling was used to estimate theory-informed direct and indirect associations, with 5,000 bias-corrected bootstrap resamples. A post hoc, model-dependent Monte Carlo simulation assessed estimation performance at the observed sample size. The specified model showed acceptable overall fit (χ²/df = 1.90, CFI = 0.93, TLI= 0.92, RMSEA = 0.06). Perceived helplessness showed a significant total association (β = 0.58, 95% CI 0.39–0.70) and direct association (β = 0.41, 95% CI 0.21–0.58) with self-reported executive dysfunction. Its indirect association through expressive suppression was statistically significant (β = 0.15, 95% CI 0.08–0.26), but the direction was opposite to the prespecified hypothesis. Low perceived self-efficacy had no significant total or direct association with executive dysfunction, although an indirect association through expressive suppression was observed (β = −0.13, 95% CI −0.24 to −0.06). Indirect associations through cognitive reappraisal and sleep quality were not statistically significant. The Monte Carlo simulation indicated moderate but not optimal estimation performance at N = 216, with potentially limited power for smaller structural effects. Perceived helplessness was cross-sectionally associated with greater self-reported executive dysfunction among operating room nurses. The inverse association involving expressive suppression was unexpected and should be considered hypothesis-generating rather than evidence of a protective or adaptive effect. Because all variables were measured concurrently using self-report instruments, temporal ordering and causal mediation cannot be established.