Effects of COVID-19-related stress and fear on depression in schizophrenia patients and the general population

Infectious disease outbreaks have caused psychosocial distress and emotional problems due to changes in daily life and fear of infection. The COVID-19 pandemic has increased social isolation, economic stress, fear of infection, and constraints on daily activities, worsening mental health and increasing the risk of depression in the general population. Individuals with mental illness, including those with schizophrenia, are more vulnerable during such disasters. People with schizophrenia may be more susceptible to COVID-19 transmission and have higher mortality rates. They often have functional impairments and limited social interactions; pandemic-related social distancing and fear of infection can further restrict social activity and reduce access to medical resources and community mental health services, increasing loneliness and risk of symptom exacerbation. While many studies have focused on the general population, few have examined COVID-19-related depression in patients with schizophrenia. The study aims to compare psychosocial stress and mental health patterns associated with COVID-19 between patients with schizophrenia and the general population, and to identify differential pathways to depression. Hypotheses: (1) psychological distress and depression are high among schizophrenia patients during the pandemic; (2) the pathway to depression differs between patients with schizophrenia and the general population. Structural equation modeling (SEM) and latent mean analysis were used to examine COVID-19-related stress, fear of infection, loneliness, and depression in both groups.

Prior literature notes that infectious outbreaks (e.g., MERS, SARS) trigger psychosocial distress, fear, and depression in the general population. Early COVID-19 research indicates widespread worsening of mental health, with increased depression risk. Vulnerable groups, including individuals with mental illness, have elevated susceptibility to infection and adverse outcomes, and face reduced access to services during pandemics. Preliminary reports suggest patients with schizophrenia experienced declines in hospital visits and may be particularly affected by social isolation, with heightened loneliness and depression. Studies in the general population have linked fear of COVID-19 with depression, but evidence in schizophrenia populations is limited, underscoring the need for comparative analyses of pathways linking COVID-19-related stressors, loneliness, and depression across clinical and nonclinical groups.

Design and setting: Cross-sectional comparative survey conducted in Korea during the COVID-19 pandemic (April–July 2020). Two cohorts were studied: patients with schizophrenia spectrum disorders and members of the general population. Participants: Patients (n=1340) aged 19–65 years, receiving treatment at community mental health centers or psychiatric outpatient clinics, diagnosed with a schizophrenia-spectrum disorder (DSM-5) by treating psychiatrists, capable of completing questionnaires and providing informed consent. Exclusions: clinically unstable/uncontrolled physical or mental illness; inability to understand or consent due to severe impairment in reality testing. Data were collected primarily via paper-and-pencil face-to-face surveys; <2% completed online. General population (n=2000) aged 19–65 years from three metropolitan areas, recruited via an anonymous online panel (Macromill Embrain) using quota sampling by age and sex. First wave April–May 2020 (n=1500); second wave July 2020 (n=500) matched by age/sex. Electronic informed consent obtained. Approvals: Chonnam National University Hospital IRB (CNUH-2020-092) and respective institutional approvals. Measures: Sociodemographics: sex, age, marital status (married vs single), education level, employment status (regular/temporary employed vs not). Depression: Patient Health Questionnaire-9 (PHQ-9); Cronbach’s α=0.897. Loneliness: 3-item UCLA Loneliness Scale (1=Rarely to 3=Often; total 3–9); Cronbach’s α=0.837. COVID-19-specific measures: Author-developed, validated questionnaire assessing fear of infection (7 items) and COVID-19-related stress (6 items), 5-point Likert scale; Cronbach’s α=0.838 (fear), 0.921 (stress). Psychometric properties referenced from prior work. Statistical analysis: Conducted using SPSS 20.0 and AMOS 20.0. Reliability (Cronbach’s α), frequency, normality (skewness<2, kurtosis<7), and descriptive statistics computed. Latent mean analysis performed after establishing measurement invariance (configural, metric, scalar; and factor variance invariance) between groups using χ2 difference tests and fit indices: TLI, CFI, RMSEA. Cohen’s d calculated using common SD under factor variance invariance. SEM compared direct and indirect effects across groups, controlling for sociodemographic covariates. Model fit evaluated via χ2, CFI, TLI, RMSEA. Mediation tested with bootstrapping. Data availability upon reasonable request to corresponding author.

Sample: 1340 schizophrenia-spectrum patients and 2000 general population; groups did not differ in age or sex. Patients were more likely to be single, unemployed, and less educated. Measurement invariance: Configural, metric, scalar, and variance invariance were supported across groups (e.g., configural: χ2=3827.93, df=538, TLI=0.914, CFI=0.923, RMSEA=0.061). Latent mean analysis (patients vs general population, reference mean=0 for general population): - COVID-19 Fear: patients lower (latent mean −0.506, p<0.01), Cohen’s d=0.60. - COVID-19 Stress: patients lower (latent mean −0.793, p<0.001), d=0.97. - Loneliness: patients higher (latent mean 0.371, p<0.01), d=0.57. - Depression: patients higher (latent mean 0.264, p<0.05), d=0.29. SEM model fit: Patients: χ2=1667.87 (df=269), TLI=0.923, CFI=0.931, RMSEA=0.062. General population: χ2=2160.06 (df=269), TLI=0.905, CFI=0.915, RMSEA=0.059. Direct effects: - Fear → Depression: not significant in patients (β=0.021, p>0.05); significant in general population (β=0.070, p<0.05). - Fear → Loneliness: significant in patients (β=0.218, p<0.001); not significant in general population (β=0.035, p>0.05). - Stress → Depression: significant in both patients (β=0.111, p<0.001) and general population (β=0.238, p<0.001). - Stress → Loneliness: significant in both patients (β=0.220, p<0.001) and general population (β=0.274, p<0.001). - Loneliness → Depression: significant and strong in both patients (β=0.656, p<0.001) and general population (β=0.578, p<0.001); loneliness had the largest effect on depression. - Fear → Stress: strong in both patients (β=0.645, p<0.001) and general population (β=0.568, p<0.001). Mediations (bootstrapped): - Fear → Loneliness → Depression: significant in patients (β=0.144, 95% CI 0.085–0.197, p=0.002); not significant in general population (β=0.020, 95% CI −0.023–0.056, p=0.420). - Stress → Loneliness → Depression: significant partial mediation in both patients (β=0.144, 95% CI 0.093–0.208, p=0.001) and general population (β=0.158, 95% CI 0.121–0.201, p=0.001). Covariates (selected): Older age increased fear in both groups (β≈0.30, p<0.05); in general population, younger age associated with higher loneliness (β=−0.310, p<0.01) and higher depression (β=−0.523, p<0.001). In patients, being unmarried and unemployed associated with greater loneliness (β=−0.302 and −0.468, both p<0.001) and greater depression (job β=−0.678, p<0.001; education level inversely related to depression β=−0.179, p<0.001).

Patients with schizophrenia exhibited lower COVID-19-related fear and stress than the general population, possibly due to relatively smaller lifestyle changes, preoccupation with preexisting intrinsic issues, or attribution of stress to other causes. Despite lower fear/stress, patients showed higher loneliness and depression, likely exacerbated by reduced access to medical resources and weakened community mental health services during the pandemic. SEM findings indicate different depression pathways: in the general population, fear of infection directly increased depression; in patients, fear did not directly affect depression but increased loneliness, which in turn strongly increased depression. COVID-19-related stress elevated both loneliness and depression in both groups, with loneliness showing the largest effect on depression overall. These results underscore loneliness as a key target for intervention, especially among patients with schizophrenia, and highlight the need for sustained social connection and tailored mental health support (including non-contact services) during prolonged pandemic conditions. Demographic moderators suggest prioritizing older patients for interventions targeting fear reduction, and unmarried, unemployed, or less educated patients for loneliness and depression mitigation.

Loneliness associated with COVID-19-related stress and fear of infection is a critical factor influencing depression, with a greater impact in patients with schizophrenia than in the general population. Mental health interventions for schizophrenia should prioritize prevention and reduction of loneliness, maintain social and emotional connections, and ensure access to services (including remote modalities) during prolonged public health crises.

- The analysis focused on loneliness as the sole mediator; other influential factors (e.g., overall health status, severity of psychopathology) were not modeled. - Different survey modes were used (face-to-face for patients vs online for general population), which may have influenced responses to sensitive questions; future work should use consistent methods across groups. - The study was conducted in Korea, where strict lockdowns were uncommon; generalizability to other countries and contexts may be limited. - Cross-sectional design (first year of the pandemic) precludes causal inference; patients with schizophrenia may be more lonely/depressed irrespective of the pandemic. Longitudinal studies, including multivariate latent growth models, are needed to assess trajectories and causal relationships.