(Converted as ePoster, See Poster Gallery) COVID-19 Mental Health and Resilience

Background and Purpose: The mental health implications resulting from this ongoing pandemic are still to be determined, yet past studies suggest the potential for longer-term consequences (Wing Chit Mak, 2010; Hansel et al., 2020). However, COVID-19 mental health outcomes are likely to vary among subgroups within the larger population. Further worsened mental health does not mean the absence of resilience and there are limited studies on the ability to bounce back after pandemics. The purpose of this presentation is to deepen the understanding of mental health and resilience during the COVID-19 pandemic. We aim to identify profiles of adaptation and identify the psychosocial factors that predict membership within these groups.

Methods: An exploratory cross-section study was conducted (N=334) shortly after initial lockdowns begin happening in the US April 7, 2020 and ran through December 16, 2020. Participants answered questions regarding anxiety, depression, and resilience along with measure COVID and psychosocial experiences. Latent profile models were estimated to identify unobserved groups in the data based on depression, anxiety and resilience scores. The second phase of multivariate analysis included a step-wise multinomial logistic regression to predict class membership. Consistent with other studies (Osofsky et al. 2015) we hypothesize a 4 factor solution, with at least one classification predominately resilient.

Results: A combination of interpretability and comparative model fit indices were used to select the final four class solution model. The Akaike’s information criterion and Adjusted Bayesian Information Criterion both decreased from the three class solution. Entropy, that is the certainty of classification, increased from 0.81 to 0.90 between the three and four class model. Finally, both the Lo-Mendell Rubin likelihood ratio test and bootstrapped likelihood ratio test were significant (ll = 74.19, p<0.001; ll = -1781.93, p<0.001 respectively) indicating that the four class solution improved over the three classes. Explorations into predictors of group membership revealed COVID specific findings such as disruption and community rates decreased membership in the resilience grouping. Psychosocial factors such as demographics (females, younger age and non-white affiliation), trauma history, information access, social support and financial resources (food security, income, ability to meet basic needs) predicted poorer mental health outcomes (p^s<.05).

Conclusions and Implications: The current biological disaster highlights that psychosocial factors exacerbate poorer mental health. Through identification of subgroups and factors, services and outreach can be targeted toward specific needs of each group. Specific to this study identifying as female, experiencing higher levels of COVID-19 disruption and previous trauma placed respondents at greater risk for symptomology. Across all class, reporting closeness with community consistently decreased the relative risk of belonging to less favorable adaptation profiles suggesting that community cohesion and social support plays an important role in building resilience during the COVID-19 pandemic. Identifying factors associated with these groupings is the first step to better understanding the strengths and challenges within the population groups for within the larger population (Osofsky et al. 2017). Policy and recovery services are immensely needed to buffer COVID-19 mental health outcomes and improve overall resilience.