Overexposure to COVID-19 information amplifies emotional distress: a latent moderated mediation model

The study addresses how overexposure to COVID-19 information (via legacy and social media) relates to emotional distress during the pandemic. Motivated by the WHO-labeled infodemic and prior evidence that repeated crisis media exposure can exacerbate mental health issues, the authors propose: (1) COVID-19 information exposure increases emotional distress (anxiety and depression); (2) risk perception mediates this association; and (3) psychological resilience moderates the mediating process by buffering the impact of risk perception on emotional distress. Conducted in China during early 2020, the work aims to clarify mechanisms linking media exposure, risk perception, and resilience to inform mental health interventions and public communication strategies.

- COVID-19 information exposure and emotional distress: Both legacy and social media became primary COVID-19 information channels. Overexposure and information overload, particularly on interactive social media with less gatekeeping, are linked to fear, anxiety, depression, and insomnia. Studies indicate frequency and duration of exposure independently predict poorer mental health, with higher risk when exposure exceeds several hours daily. - COVID-19 information exposure and risk perception: Risk perception (perceived susceptibility and severity) is shaped by mass media during outbreaks (H1N1, avian flu, BSE). Limited exposure can elevate uncertainty and inflate perceived risk; overexposure increases contact with negative or false information, potentially amplifying risk perception. The authors hypothesize positive association between COVID-19 information exposure and risk perception in China. - Risk perception and emotional distress: Perceived danger triggers negative affect (worry, fear, anxiety, depression). Evidence during COVID-19 shows higher risk perception correlates with greater distress. The authors posit risk perception mediates the exposure–distress link. - Resilience and emotional distress: Resilience, a multidimensional capacity for adaptation, correlates positively with mental health and negatively with psychological distress. During COVID-19, higher resilience relates to lower worry, anxiety, and depression. The study examines resilience as a moderator that buffers the effect of risk perception on emotional distress and explores differential roles of its subcomponents (tenacity, tolerance, acceptance, control, spirituality).

Design and participants: Cross-sectional, cluster-sampled online survey administered Feb 2–Mar 3, 2020, at two universities in Beijing. Undergraduate and graduate students completed a teacher-distributed online questionnaire. Informed consent obtained; ethics approval from Beijing Normal University. Of 1713 approached, 1347 responded (78.6%); after exclusions (COVID-19 infection, attention check failure, too-short completion time <5 minutes), final N = 1313 (Mage = 19.76 ± 2.25; 73.1% female; 85.1% Han; 94.6% undergraduates). Measures: - COVID-19 information exposure: Two items capturing (a) frequency (“times per day” on 0–20 scale) and (b) duration (“hours per day” on 0–12 scale), covering both legacy and social media. Frequency–duration correlation r = 0.33, p < .001. - Perceived risk: Two items on likelihood of self and close others being infected (0%–100% each; composite score; Cronbach’s α = 0.872; higher scores = greater perceived risk). - Emotional distress: Assessed via anxiety and depressive symptoms. - Anxiety: GAD-7 (0–3 per item; α = 0.916); cutoff ≥5 for presence of anxiety symptoms. - Depression: PHQ-9 (0–3 per item; α = 0.885); cutoff ≥5 for presence of depressive symptoms. - Psychological resilience: Connor–Davidson Resilience Scale (CD-RISC-25; 1–5 Likert; α = 0.941). Subscales: tenacity (α = 0.865), tolerance (α = 0.860), acceptance (α = 0.798), control (α = 0.736), spirituality (α = 0.419). Analytic approach: Conducted in SPSS 26 and Mplus 8.3, two-tailed α = .05. Descriptives and correlations computed; Harman’s single-factor test for common method bias (single factor explained 33.9%). Structural equation models with demographics (age, sex, education) as covariates: (1) direct-effects model (latent exposure → latent emotional distress), (2) latent mediation model (exposure → perceived risk → distress), and (3) latent moderated mediation model testing interaction of perceived risk × resilience on distress and resilience’s moderation of the indirect effect. Model fit assessed via χ2/df, CFI, TLI, RMSEA, SRMR; nested models compared via log-likelihood ratio tests. Threshold analyses examined frequency/duration corresponding to symptom severity categories.

- Descriptives: Mean exposure frequency 5.73 times/day (SD = 4.18); mean duration 1.46 h/day (SD = 1.18). - Correlations: Exposure frequency correlated with anxiety r = 0.19 (p < .001) and depression r = 0.14 (p < .001). Exposure duration correlated with anxiety r = 0.17 (p < .001) and depression r = 0.12 (p < .001). - Direct effect model: Good fit (χ2/df = 1.858, CFI = 0.993, TLI = 0.987, RMSEA = 0.026, SRMR = 0.018). Higher COVID-19 information exposure predicted greater emotional distress (β = 0.30, 95% CI [0.20, 0.39], p < .001). - Thresholds: Moderate anxiety associated with ≥7 times/day or ≥2 hours/day exposure; moderate depression associated with ≥6 times/day or ≥1.64 hours/day (~38 minutes) exposure. - Mediation by perceived risk: Mediation model fit well (χ2/df = 2.180, CFI = 0.990, TLI = 0.985, RMSEA = 0.030, SRMR = 0.022). Exposure → perceived risk (β = 0.12, 95% CI [0.05, 0.19], p = .005); perceived risk → emotional distress (β = 0.23, 95% CI [0.18, 0.28], p < .001). Significant indirect effect (β = 0.03, 95% CI [0.01, 0.04], p = .005). - Moderated mediation by resilience: Model with interaction fit better than null (log-likelihood ratio D = 19.266, df = 1, p < .001). Interaction perceived risk × resilience predicting emotional distress was significant (β = -0.15, 95% CI [-0.22, -0.07], p = .002). Simple slopes: At low resilience (−1 SD), perceived risk strongly predicted distress (β = 0.34, 95% CI [0.25, 0.44], p < .001); at high resilience (+1 SD), the slope was weaker and nonsignificant (β = 0.06, 95% CI [-0.03, 0.14], p = .276). Indirect effects: low resilience β = 0.08 (95% CI [0.01, 0.13]); high resilience β = 0.01 (95% CI [-0.01, 0.02]). - Resilience subfactors: Tenacity (β = -0.15, 95% CI [-0.23, -0.07], p = .002), acceptance (β = -0.14, 95% CI [-0.25, -0.03], p = .033), control (β = -0.11, 95% CI [-0.19, -0.02], p = .048), and spirituality (β = -0.17, 95% CI [-0.27, -0.08], p = .004) significantly moderated the perceived risk → distress link; tolerance was not significant (β = -0.07, 95% CI [-0.19, 0.05], p = .365).

Findings support the hypothesis that overexposure to COVID-19 information is associated with elevated emotional distress among Chinese university students. The results elucidate a mechanism wherein increased exposure raises perceived risk, which in turn exacerbates anxiety and depressive symptoms. Psychological resilience buffers this pathway: individuals with higher resilience are less affected by perceived risk, substantially weakening the indirect effect of exposure on distress. Not all resilience components function equally; tenacity, acceptance, control, and spirituality showed protective moderation, whereas tolerance did not. Threshold analyses suggest relatively low exposure levels (6–7 checks/day or ~1.6–2 hours/day) already correspond to moderate symptom severity, potentially reflecting cultural or contextual sensitivities. The study advances understanding of mental health impacts during infodemics and highlights the importance of fostering resilience and risk-communication strategies to mitigate harms.

Overexposure to COVID-19 information increases perceived risk and amplifies emotional distress. Risk perception partially mediates the exposure–distress relationship, and psychological resilience attenuates this mediated effect, with specific resilience facets (tenacity, acceptance, control, spirituality) providing the strongest buffering. The work contributes a theoretically integrated model of exposure, risk perception, and resilience during a health crisis and identifies practical exposure thresholds relevant for assessment and guidance. Future research should include content characteristics of information exposure, additional psychosocial mediators, longitudinal designs, and objective biomarkers to better capture causal mechanisms and generalizability.

- Sample limited to university students comfortable with mobile phone use; may not represent older adults or legacy media users. - Exposure measured only by amount (frequency and duration); content and valence of information were not assessed. - Other potential mechanisms (e.g., perceived self-efficacy, information processing modes) were not modeled. - Cross-sectional, self-reported data preclude causal inference; objective biological/physiological measures were not included. - Data collected during outbreak-to-post-peak in China; robustness across later phases requires further study.