Obsessive-compulsive symptoms and information seeking during the Covid-19 pandemic

The Covid-19 pandemic created widespread uncertainty and isolation with disruptions across society, posing major challenges to mental health. Early reports showed worsening symptoms across various psychiatric conditions, including anxiety, depression, bipolar disorder, schizophrenia, and OCD, and elevated anxiety and depression in the general public. OC symptoms are expected to be disproportionately affected due to their focus on contamination and harm. Prior epidemics (e.g., HIV/AIDS) suggest public health campaigns can influence epidemic-related OCD-like behaviors even in individuals without psychiatric history, raising concerns about similar effects during Covid-19. Although psychiatric symptoms typically elevate following stress and then adapt back toward baseline, failures in this adaptation may lead to chronic problems. This study investigates, in the general UK public, how OC, anxiety, and depression symptoms evolved across the first pandemic wave and how these symptoms relate to Covid-related information seeking and adherence to governmental guidelines. The authors hypothesized that OC symptoms would be particularly elevated and associated with increased information seeking, which could influence guideline adherence.

The paper reviews evidence that patients with anxiety, depression, bipolar disorders, schizophrenia, and OCD experienced pandemic-related symptom increases, with mixed findings in OCD likely due to heterogeneity and comorbidities. In non-patient populations, the trajectory of OC symptoms during Covid-19 was largely unknown. Historical evidence from the HIV/AIDS epidemic indicates that health campaigns can precipitate OCD-like behaviors in individuals without prior psychiatric diagnoses. The broader literature shows mental health commonly worsens after stressful events but often adapts back, attributed to coping and reappraisal; failure to adapt can result in chronic issues. These insights motivate examining long-term trajectories and potential pandemic-specific effects on OC symptoms and related behaviors such as information seeking.

Design: Longitudinal online study of the general UK public with two time points during the first Covid-19 wave. T1: April 24–May 7, 2020 (peak of first wave in the UK). T2: July 15–August 15, 2020 (after major easing of restrictions, e.g., reopening of restaurants and pubs). Participants were recontacted at T2 to assess adaptation of symptoms to environmental change. Participants: Recruited via Prolific; adults (18+) residing in the UK; informed consent obtained; ethics approval by UCL REC. T1: 446 completed; exclusions: 30 for missing data or failed attention checks; 10 with self-reported OCD diagnosis excluded to maintain a non-clinical spectrum sample. Final T1 N=406 (233 females; Mage=34, SD=12.613). T2: Of T1, 315 completed; exclusions: 19 for missing/attention failures; final T2 N=296 (164 females; Mage=35, SD=12.623); retention 78%. No T1 mental-health differences between those who did vs. did not return at T2. Very few positive Covid-19 tests (T1: n=1; T2: n=2), so not excluded or controlled. Measures:

• OC symptoms: Padua Inventory—Washington State University Revision (PI-WSUR), chosen for test-retest reliability and detailed subscales. A pandemic-neutral OC subscore was created excluding items likely biased by pandemic context (e.g., contamination/hygiene items) to ensure effects were not driven by adaptive pandemic behaviors.
• Anxiety and depression: Hospital Anxiety and Depression Scale (HADS), focusing on psychological symptoms and minimizing physical-symptom confounds; two subscales evaluated separately; good internal consistency and test-retest reliability.
• Covid-19 information seeking: Newly developed 5-item questionnaire (5-point Likert) assessing information exchange and seeking across social/media channels. Dimensionality assessed by PCA; internal consistency via Cronbach’s alpha; test-retest via Pearson correlation between T1 and T2.
• Baseline media use: Retrospective average weekly news and social media consumption prior to the pandemic (Nov 2019) to control for pre-existing differences; used as covariate in robustness checks.
• Guideline adherence (T2): Degree of adherence to UK governmental Covid-19 recommendations; total adherence score used.
• Covariates: Cognitive ability (IQ estimate via a cognitive ability assessment at T1), age, gender, education, essential worker status. Statistical analysis: Data preprocessing in MATLAB 2020a; analyses in SPSS 26 and R 3.6.2. Psychiatric scores z-scored across time points. Repeated-measures ANOVA with within-subject factors: psychiatric dimension (OC, anxiety, depression) and time (T1, T2), with z-scored covariates IQ, age, gender; Huynh-Feldt correction applied. Paired two-sided t-tests (and permutation tests) compared T1 vs. T2 for each dimension and information seeking. Associations between information seeking and psychiatric scores at each time point tested with robust multiple regression (rlm, bisquare; p-values via rob.pvals), covarying age, gender, education, essential worker status, IQ. Longitudinal changes in associations examined with random-intercept mixed-effects model using lme4: Information seeking ~ time * (OC + anxiety + depression) + gender + age + IQ + education + essential worker + (1|subject). Mediation analysis tested whether T1 OC predicted T2 guideline adherence via T1 information seeking using the Canlab Mediation Toolbox with 10,000 bootstrap iterations; interpretation followed current mediation typology. Multicollinearity assessed via VIF (<5 for all models).

• Elevated symptoms at peak of first wave (T1): OC (PI-WSUR) mean 30.775 (SD=22.103), higher than pre-pandemic population samples; anxiety mean 7.797 (SD=4.786); depression mean 6.768 (SD=4.465). At T1, 48% (195/406) scored ≥8 on HADS anxiety and 41% (166/406) ≥8 on HADS depression.
• Longitudinal trajectories (T1→T2): Significant time by psychiatric domain interaction (F(1.970, 575.248)=19.848, p<0.001) and main effect of time (F(1, 292)=4.957, p=0.027). OC symptoms increased from T1 to T2 (paired t(295)=5.294, p<0.001); anxiety remained stable; depression decreased. These patterns were also evident in distributional plots and boxplots.
• Behavioral linkage: OC symptoms were directly associated with higher Covid-related information seeking, which in turn was associated with greater adherence to governmental guidelines. Mediation analysis (as described) indicated that information seeking mediated the relationship between OC symptoms and guideline adherence (full mediation interpretation when indirect effect significant and direct effect not significant).

Findings show that while general psychiatric distress was elevated during the first Covid-19 wave, symptom trajectories diverged over time: depression decreased and anxiety plateaued, consistent with partial adaptation to stress, whereas OC symptoms increased further even after restrictions eased. This domain-specific rise suggests that pandemic features (contamination, threat of causing harm) disproportionately affect OC symptomatology in the general public. The observed link between OC symptoms and increased Covid-related information seeking, and the association of information seeking with higher guideline adherence, highlight a pathway by which heightened OC symptoms can influence adaptive public health behavior. However, persistent or rising OC symptoms may carry risks for longer-term mental health burden. These results underscore the importance of monitoring OC-related symptoms beyond the acute crisis and considering targeted supports to mitigate potential chronicity while maintaining beneficial behaviors such as adherence to guidelines.

This longitudinal study of the UK general public demonstrates that OC symptoms not only were elevated during the first pandemic wave but further increased over time, in contrast to anxiety (stable) and depression (decreasing). OC symptoms were linked to greater Covid-related information seeking, which promoted higher adherence to governmental guidelines, suggesting a mediating behavioral pathway. The work contributes evidence of domain-specific, potentially persistent impacts of the pandemic on mental health and highlights the need for continued surveillance and targeted interventions to prevent chronic OC-related problems. Future research should examine longer-term trajectories post-pandemic, mechanisms linking OC symptoms to information behaviors, the impact of public health messaging, and potential interventions to support adaptation without exacerbating OC symptomatology.

• Contextual comparability: Direct quantitative comparisons with pre-pandemic samples are challenging due to different contexts (pandemic vs. non-pandemic), limiting clinical inference from score levels.
• Sampling and generalizability: Non-clinical UK sample; individuals with self-reported OCD diagnosis were excluded, which may limit generalizability to clinical populations.
• Self-report measures: Reliance on self-report questionnaires may introduce response biases; pandemic context may have influenced item interpretation (especially contamination/hygiene content).
• Information-seeking measure: Newly developed scale, although validated (PCA, internal consistency, test-retest), may require further external validation.
• Covid-19 status: Very few participants reported positive Covid-19 tests; status not controlled due to low numbers.
• Retrospective baseline media use: Pre-pandemic media consumption was assessed retrospectively, which may be subject to recall bias.
• Limited time points: Only two measurement points during the first wave, restricting the ability to model more nuanced trajectories.