This page was translated automatically from Turkish. Read the original

This page is an AI-assisted Turkish summary of the study presented at the 26th World Congress of Psychiatry. The study belongs to the presenting authors; please refer to the original source when citing it.

EV0838

The relationship of stress and mental health symptoms with working memory performance and response bias

STRESS AND MENTAL HEALTH SYMPTOMS IN RELATION TO WORKING MEMORY PROCESSES UNDER COGNITIVE LOAD

Perceived stress, anxiety, and depressive symptoms showed no consistent association with working memory accuracy or sensitivity and were associated with the response criterion only through a three-way interaction.

The study aimed to use signal detection theory measures to examine whether the relationship between perceived stress, anxiety, and depressive symptoms and working memory under cognitive load reflects sensitivity in distinguishing target from non-target stimuli or the response criterion, while also accounting for adverse childhood experiences. Following ethics approval, 169 Chilean university students completed questionnaires and a cognitive task via a mobile application, and 154 were included in the analysis (83 women, 64 men, 7 other; mean age 22.3). Measures included the 10-item Adverse Childhood Experiences (ACE) Questionnaire, PSS-14, GAD-7, and PHQ-9; the cognitive task was a two-choice letter n-back task at the 1-, 2-, and 3-back levels, with trials presented every 3 seconds. Linear mixed models were fitted for sensitivity (d'), criterion (c), accuracy, and reaction time, with participant-specific random intercepts, load as a within-participant variable, and gender and ACE as covariates; p-values were adjusted using the Holm correction. As load increased, accuracy decreased to 96.7%, 90.1%, and 81.2%, and sensitivity d' decreased to 3.45, 2.56, and 1.77, while reaction times increased to 935, 1108, and 1120 ms; the criterion remained conservative at all levels (c = 0.18; 0.21; 0.25). No consistent main effects or two-way interactions involving symptom scales were observed for sensitivity, accuracy, reaction time, or criterion; PSS (p = 0.036) and the GAD by PHQ interaction (p = 0.047) emerged as conditional effects only in the model including the higher-order interaction and did not survive correction. The three-way PSS x GAD x PHQ interaction was not associated with sensitivity (b = -1.83 × 10⁻⁴; p = 0.307), accuracy (p = 0.304), or reaction time (p = 0.416), but was associated with criterion (b = 1.84 × 10⁻⁴; SE = 6.28 × 10⁻⁵; p = 0.004) and improved model fit (χ²(1) = 8.99; p = 0.003); this association remained significant after Holm correction. The interaction pattern indicated that, among participants with high anxiety but relatively low depression scores, increasing perceived stress corresponded to a more liberal response threshold (b = -0.018; 95% CI -0.035 to -0.0005), with no comparable association found for other symptom combinations; a one-standard-deviation change in perceived stress corresponded to Δc = -0.18, equivalent to an approximately five-percentage-point increase in the hit rate and a four-percentage-point increase in the false alarm rate at the 3-back level. Because the symptom scales were strongly correlated with one another (r = 0.72 to 0.75) and the interaction weakened when influential observations were excluded, the authors considered the finding hypothesis-generating and suggested that symptom combinations may be associated with response strategy rather than discriminability.

All WCP 2026 summaries