Recent societal challenges, such as the COVID-19 pandemic, economic crises, wars, and natural disasters, have significantly impacted public mental health. These events affect large populations simultaneously and have long-lasting consequences, increasing stress levels and the risk of mental disorders. This review focuses on resilience as a favorable adaptation to stress, defined as maintaining or quickly recovering mental health during or after stressor exposure. Resilience factors, from individual psychological traits to societal resources, are seen as protective resources that increase the likelihood of resilient responses. A multisystemic approach considers individual (e.g., optimism, self-efficacy), social (e.g., social support, family climate), and societal (e.g., access to natural spaces, perceived safety) resilience factors. Recent research suggests that these factors might influence resilient outcomes through a smaller number of mediating resilience mechanisms (e.g., emotion regulation, positive appraisal). This review aims to systematically assess the predictive value of these multilevel resilience factors for resilient responses to societal challenges, considering their incremental validity beyond sociodemographic variables and other resilience factors.

Previous research on resilience factors has largely focused on individual factors like dispositional optimism and self-efficacy. However, these studies often lack conceptual clarity and fail to examine the incremental validity of individual factors beyond other resilience factors or sociodemographic variables. Research on social resilience factors, primarily focusing on perceived social support, has also shown some inconsistencies. Meta-analyses consistently show a strong link between perceived social support and stress-related mental symptoms, but longitudinal associations have been challenged due to potential statistical artifacts. Studies examining societal resilience factors are even rarer, lacking a consensus on categorization and rarely examining their influence on individual responses to major stressors. Existing reviews highlight the need for a more multilevel approach that considers the interplay between individual, social, and societal resources in fostering resilience.

This systematic review followed PRISMA guidelines. Five databases (APA PsycNet, Embase, PTSDPubs, Scopus, Web of Science) were searched until August 2, 2023, using keywords related to stress exposure, mental health, and trajectory modeling. Eligible studies were longitudinal observational studies examining adult (≥18 years) individuals from the general population in OECD countries exposed to various societal challenges (pandemics, wars, climate crisis, natural disasters). Studies had to use growth mixture modeling (GMM) or comparable approaches to identify mental health trajectories and investigate resilience factors as predictors. A modified Newcastle-Ottawa Scale assessed study quality. A rating scheme categorized evidence for each resilience factor based on its association with favorable or unfavorable trajectories, controlling for sociodemographic variables and other resilience factors. Effect sizes were also considered. Qualitative data synthesis was supplemented with non-parametric statistical tests to examine associations between evidence ratings and study, participant, and contextual factors.

The review included 50 studies (15 OECD countries, 360–65,818 participants). Most studies (84%) used GMM, with a median study quality rating of 66.67%. Pandemics were the most frequently studied stressor (58%), followed by environmental/natural disasters (18%) and terrorist attacks (14%). Mental health outcomes included PTSD symptoms (32%), depressive symptoms (44%), general distress (32%), and anxiety symptoms (28%), with fewer studies examining positive mental health (10%). Seventy-two percent of studies investigated individual resilience factors, 86% social factors, and 26% societal factors. The most frequently studied factors were education and income (individual), social support (social), and rural residence (societal). A significant association was found between the level of resilience factors and types of societal challenges: societal factors were more often examined in studies on environmental/natural disasters, while individual factors were more prevalent in studies on pandemics and terrorist attacks. Analysis of 478 effect estimates revealed the strongest evidence for the incremental validity of individual income, household income, socioeconomic status, perceived social support, and (cognitive) emotion regulation and psychological flexibility. Effects were more favorable in samples with more women and younger participants. Many other well-established factors showed mixed results. Effect sizes were mostly very small to small. Analyses examining recovery trajectories versus less favorable responses yielded less consistent results. There was no significant association between study quality and evidence ratings or the timing of assessments. A larger number of variables in regression models correlated with less favorable evidence ratings.

This review provides a comprehensive overview of multilevel resilience factors in response to societal challenges. While individual factors like income and emotion regulation showed consistent positive associations with resilient outcomes, the majority of effect sizes were small. The mixed findings for many well-established resilience factors highlight the importance of considering the interaction between factors and the situational context, supporting the concept of regulatory flexibility. The limited research on social and societal resilience factors underscores the need for future studies to address this gap. The lack of research on resilience mechanisms further emphasizes the need for more in-depth investigations into the underlying processes mediating the relationship between resilience factors and outcomes. The findings suggest that multilevel interventions addressing factors at individual, social, and societal levels may be more effective than solely focusing on individual-level interventions.

This systematic review highlights the importance of multilevel resilience factors in responding to societal challenges, particularly income, social support, and emotion regulation. However, the predominantly small effect sizes emphasize the need for future research to focus on social and societal resilience factors, resilience mechanisms, and more sophisticated statistical modeling to account for complex interactions and temporal dynamics. Large-scale, international studies are needed to inform the development and evaluation of effective multilevel resilience interventions.

Several limitations should be considered. The reliance on a specific list of stressors might have introduced bias. The findings reflect correlations, not causal relationships. The use of GMM might have inflated prevalence rates for resilient responses. The rating scheme, while innovative, relies on vote counting and is affected by limitations of statistical significance testing. The relatively small number of studies for some stressor types and outcomes limited the depth of analysis. The lack of pre-stressor data in most studies may have also influenced the results.