The COVID-19 pandemic presented a significant threat to mental health due to rapidly rising case and death numbers, overwhelmed healthcare systems, economic strain, and social distancing measures. Prior research has documented the mental health consequences of past pandemics, including increased suicide rates during the 1918 influenza pandemic. While worsening mental health conditions have been reported during COVID-19, human emotional adaptability and resilience remain under-investigated in the context of prolonged crises. This study aimed to examine longitudinal changes in depression and anxiety during the initial weeks of the COVID-19 pandemic in the United States, exploring both static (demographic) and dynamic (evolving over time) factors influencing mental health.

Existing research highlights the mental health impacts of past pandemics and the current COVID-19 pandemic, showing increased rates of depression and anxiety. Studies have identified individual-level factors associated with resilience and emotional adaptation, such as social support, family support, education levels, gender, and socioeconomic status. While some research has explored the relationship between specific behaviors (e.g., outdoor activities, social media use) and mental well-being during the pandemic, the longitudinal trajectory of emotional adaptation and the contributing factors during a prolonged crisis remained unclear. This study aimed to fill this gap by investigating the dynamic interplay of demographic and time-varying factors influencing mental health during the pandemic's initial phase.

This web-based longitudinal study enrolled 1512 US adults (aged 18-64) via Prolific.co from April 2nd to June 4th, 2020. Participants completed weekly surveys for 10 weeks, assessing depression and anxiety using the Zung Self-Rating Depression scale and State-Trait Anxiety Inventory (state subscale), respectively. Demographic data (sex, age, income, prior psychiatric diagnoses, marital status) and COVID-related data (economic impact, informedness, social media use, pandemic duration projection) were also collected. After removing duplicate responses and failed attention checks, 1456 participants' data remained at the first time point. Two linear mixed-effects models analyzed the longitudinal changes in depression and anxiety, considering both static and dynamic variables. Model comparisons were used to assess the contribution of additional variables, including interactions between time and demographic factors. The analysis was conducted using MATLAB and R.

Both depression and anxiety scores were initially high but significantly declined over the 10 weeks (repeated-measures ANOVA, p<0.001 for both). While average depression scores remained below the clinical cutoff, the average anxiety score exceeded the clinical cutoff at the first time point. Linear mixed-effects models revealed that younger age, being female, lower income, and having a prior mood or anxiety disorder diagnosis were associated with higher levels of both depression and anxiety. Being married was associated with lower depression scores. Dynamic factors showed that worsening COVID-related economic impact and longer projected pandemic duration significantly exacerbated both depression and anxiety. Increased social media use and the 7-day change in COVID-19 cases (not the total case count) were positively associated with anxiety. Higher levels of informedness were negatively associated with depression. An age-by-time interaction indicated that while older individuals were less depressed overall, their depression levels did not decrease as much over time compared to younger individuals.

This study provides evidence of emotional adaptation during the initial phase of the COVID-19 pandemic in the US, with depression and anxiety levels declining over time. The findings highlight overlapping yet distinct factors contributing to depression and anxiety. The negative association between worsening economic impact and mental health aligns with previous research on economic recessions and mental health. The differential effects of dynamic factors on anxiety and depression suggest unique mechanisms influencing each disorder. For example, the association between anxiety and social media use and changes in COVID-19 cases may reflect the impact of uncertainty and fear. The negative relationship between informedness and depression may be due to hopefulness. The findings also confirm existing research on demographic factors' influence on mental health vulnerability.

This study demonstrates emotional adaptation during the initial phase of the COVID-19 pandemic, with a decline in depression and anxiety over time, but also highlights the impact of economic hardship, pandemic duration projections, and information access. Future research should investigate causal relationships between these factors and mental health outcomes, using larger representative samples, and exploring potential interventions to mitigate the negative effects of crises on mental well-being.

The study's correlational nature limits causal inferences. The absence of baseline mental health data and the lack of complete representation of the US population restrict the generalizability of the findings. The reliance on self-reported data may introduce bias. Future research should address these limitations by employing longitudinal designs with baseline data, focusing on diverse and representative samples, and incorporating objective measures of mental health.