Within-person increases in crisis-related distress track worsening adolescent depression, anxiety, and health-related quality of life – A post-COVID-19 longitudinal multilevel change analysis

Technical Summary: Within-person Increases in Crisis-Related Distress Track Worsening Adolescent Depression, Anxiety, and Health-Related Quality of Life

Problem Statement
Adolescence is a critical developmental period for mental health, with a significant proportion of mental disorder onsets occurring before age 18. While extensive research has documented the prevalence of mental health issues during and immediately after the COVID-19 pandemic, there is a gap in understanding how within-person changes in crisis-related distress correlate with changes in mental health over time. Most existing studies focus on between-person differences (comparing distressed adolescents to non-distressed peers) or average symptom prevalence. This study addresses the unresolved question of whether time-varying distress within the same adolescent is associated with concurrent fluctuations in depressive symptoms, anxiety symptoms, and health-related quality of life (HRQoL) as the world transitions from the acute pandemic phase to a post-emergency period. Furthermore, it investigates whether these associations persist after controlling for sociodemographic factors, individual life stressors, and protective resources like self-efficacy.

Methodology
The study utilized a two-wave longitudinal design with data collected from 1,117 adolescents (grades 7–10) attending 49 secondary schools in Saarland, Germany. Participants were assessed in 2022 (late pandemic period) and again in 2023 (post-emergency phase).

• Analytical Approach: The researchers employed random-intercept multilevel change models with within-between decomposition. This statistical technique separated pandemic-related distress into two components:
  1. Between-person differences: The average level of distress for an adolescent across both waves.
  2. Within-person deviations: The deviation of a specific wave's distress score from that adolescent's own mean.
• Measures:
  • Exposure: Pandemic-related distress (a composite of perceived stress, negative emotional responses, and perceived negative impact on daily coping; Cronbach's α = .90).
  • Outcomes: Depressive symptoms (PHQ-9-M), generalized anxiety symptoms (SCARED), and HRQoL (KIDSCREEN-10).
  • Covariates: Age, sex, socioeconomic status (SES), individual life stressors (ZLEL), and self-efficacy (ASKU).
• Controls: Models were adjusted for age, sex, SES, individual life stressors, and self-efficacy. Chi-square tests were used to compare the proportion of adolescents meeting clinically relevant cut-off criteria between measurement points.

Key Results
The fully adjusted models accounted for a substantial portion of the variance in mental health outcomes (approx. 70–72%).

1. Within-Person Associations: Increases in pandemic-related distress within an individual adolescent were significantly associated with:
   • Higher depressive symptoms (Standardized β = 0.11, p < .001).
   • Higher anxiety symptoms (β = 0.12, p < .001).
   • Lower HRQoL (β = −0.09, p < .001).
2. Between-Person Associations: Adolescents with generally higher average distress across the study period reported poorer mental health outcomes. These between-person associations were larger in magnitude than the within-person effects (e.g., β = 0.23 for depression; β = 0.32 for anxiety).
3. Protective Factors: Self-efficacy was consistently associated with lower depressive and anxiety symptoms and higher HRQoL. Individual life stressors were strong positive predictors of symptom burden.
4. Temporal Trends:
   • Depressive and anxiety symptom burdens increased significantly from 2022 to 2023.
   • The proportion of adolescents meeting clinical cut-offs for depression and anxiety rose between measurement points.
   • No significant change was observed in the proportion of adolescents with clinically low HRQoL between 2022 and 2023.
5. Demographics: Female adolescents and older adolescents reported higher symptom burdens and lower HRQoL.

Key Contributions

• Methodological Distinction: The study explicitly disentangles between-person and within-person effects, demonstrating that crisis-related distress is not merely a marker of stable interindividual risk but also a dynamic, time-varying factor that covaries with mental health fluctuations within the same individual.
• Post-Acute Trajectories: The findings challenge the notion of a simple "recovery" by the third pandemic year. Instead, they indicate that pandemic-related distress remains a relevant predictor of mental health burden in the post-emergency phase, with symptom levels continuing to rise in 2023.
• Comprehensive Burden Assessment: By analyzing depressive symptoms, anxiety, and HRQoL in parallel, the study provides a broader view of adolescent functioning, showing that while symptom prevalence increased, the proportion of clinically low HRQoL remained stable, suggesting complex shifts in perceived well-being.

Significance and Claims
The paper claims that pandemic-related distress serves as a valid model case for understanding how adolescents respond psychologically to large-scale, overlapping global crises. The primary significance lies in the identification of within-person increases in distress as a dynamic warning signal.

The authors argue that clinical and preventive approaches should move beyond static risk profiling (identifying only those with chronically high distress). Instead, longitudinal monitoring of intraindividual changes in crisis-related distress is necessary to identify periods of heightened vulnerability, even in adolescents whose average distress levels are not exceptionally high.

The study supports the integration of crisis-informed support in schools and communities. It suggests that interventions should not focus solely on symptom reduction but also address the specific burden of crisis-related distress, including strategies for coping with uncertainty and perceived uncontrollability. While self-efficacy is identified as a complementary protective resource, the authors caution against viewing it as a sole explanatory mechanism or placing the burden of coping entirely on the individual, emphasizing the need for broader systemic support.

The authors maintain a modest stance regarding causality, acknowledging that the two-wave design cannot establish temporal ordering (i.e., whether distress causes symptom worsening or vice versa). They also note that while pandemic distress is a useful model, findings should not be overgeneralized to all forms of crisis (e.g., climate or war) without further comparative research.