Psychological well-being in Europe after the outbreak of war in Ukraine

The February 24, 2022, Russian invasion of Ukraine marked a significant escalation of the conflict that began in 2014. Beyond the devastating impact on Ukraine, the invasion triggered a major refugee crisis, global food shortages, and economic turmoil. The UN estimates that 1.6 billion people across 94 countries are affected. While the humanitarian and economic consequences are readily apparent, the psychological repercussions are less understood. Existing research on major life events often suffers from retrospective biases due to post-event data collection. Cross-sectional studies lack the time resolution to capture immediate effects and longitudinal studies often have low temporal resolution and are usually confined to single nations. This study addresses these limitations by leveraging data from a global experience-sampling study to analyze the impact on well-being in Europe during the weeks surrounding the invasion.

Previous research on the psychological effects of major life events highlights the challenges of retrospective bias and the limitations of cross-sectional studies. Longitudinal studies, while offering a temporal perspective, often lack the necessary time resolution or geographical scope to comprehensively assess the immediate and widespread impact of sudden events like war. Studies on well-being adaptation to life events suggest a process of adjustment, but the extent and individual variability of adaptation remain areas of ongoing investigation. This study builds on this literature by utilizing a high-resolution, international dataset to examine well-being changes around the Ukrainian war outbreak.

This study uses data from the "Coping with Corona" (CoCo) project, a global experience-sampling study conducted from October 12, 2021, to August 16, 2022. The analysis focuses on a sample from 17 European countries (N=1341 participants, 44,894 assessments). Participants completed daily experience-sampling surveys, reporting their well-being multiple times a day. Pre- and post-surveys gathered data on personality (Big Five Inventory-2, assessing Stability and Plasticity meta-traits) and sociodemographic variables (age, gender, subjective social status, political orientation). The daily number of tweets containing "Ukraine" served as a measure of the war's social media salience. Multilevel models analyzed well-being trajectories around the invasion day, examining the influence of personality, sociodemographic factors, and war salience. Eight multilevel models, representing different well-being trajectories, were compared using AIC. Model 2d, characterized by independent linear trends before and after the invasion and a sudden baseline shift on the invasion day, best fit the data. Further analyses explored moderators (personality, sociodemographic factors) and the within- and between-subjects effects of war salience. Bayesian estimations were used as a robustness check. Supplementary analyses examined country-specific effects, global sample comparisons, different time frames, lagged effects of war salience, and associations between well-being and Ukraine-related behaviors/emotions.

The invasion led to an acute, significant decline in well-being (b=-0.200, 95% CI [-0.278, -0.122], p < 0.001) on the day of the invasion. No significant pre-invasion anticipation effects were observed. Post-invasion recovery was slow and positively associated with the personality trait Stability (b=0.161, 95% CI [0.066, 0.255], p=0.001). Individuals high in Stability recovered faster than those low in Stability. Daily well-being was negatively associated with daily war salience on social media (b=-0.070, 95% CI [-0.096, -0.044], p<0.001), both within and between subjects. While the initial decline in well-being wasn't significantly associated with sociodemographic factors, recovery was strongly linked to personality. The effect size of the well-being decline (0.2 SD) is substantial compared to other population-level global events, although smaller than that observed for personal life events like bereavement. Supplementary analyses showed some country-specific differences in recovery patterns.

The findings indicate a widespread, albeit temporary, negative impact of the Ukrainian war on European well-being. The lack of association between initial decline and sociodemographic factors suggests a potent, shared situational impact. The strong association between recovery and personality highlights individual differences in coping. The negative correlation between well-being and war salience on social media suggests a potential mechanism for the observed effects. Set-point models of well-being are partially supported, with a general adaptation effect observed, although individual differences in adaptation speed are significant. Exploratory analyses suggest a link between decreased well-being and dysfunctional psychological indicators but did not find consistent support for a positive association with pro-Ukraine behavior.

The study demonstrates the significant, albeit temporary, impact of the Russo-Ukrainian war on European well-being. Personality traits, specifically Stability, play a key role in the recovery process. The relationship between social media salience and well-being highlights the potential for information exposure to influence emotional response. Future research could investigate the interaction between personality, coping mechanisms, and well-being in greater detail, and also explore the specific effects across different national contexts and age groups.

The study has several limitations. The uneven distribution of participants across countries limits the generalizability of findings to specific nations. The sample’s overrepresentation of female, educated, and younger participants affects its representativeness of the general population. The lack of data from Ukraine and Russia prevents a complete understanding of the psychological effects on those most directly affected. The reliance on self-reported data introduces potential bias. Finally, the study is correlational and cannot establish causal relationships.