Examining Relationships Between Behavior, Social Environment, and Trait Factors on Mood and Anxiety State During Social Isolation and Pseudo-Confinement

Extended duration in isolated, confined, and extreme (ICE) environments can have negative impacts on behavioral health in spaceflight and terrestrial environments. During the early months of the COVID-19 pandemic, large portions of the general public were required to socially distance and isolate at home (a pseudo-confined environment). This period presented unique circumstances to examine how a more diverse population weathers isolated and confined conditions and to explore relationships between psychological traits, behaviors, and social environment on mood and anxiety state. An internet-based survey was conducted from April to December 2020 of individuals social distancing during the early pandemic. The survey captured background measures of mental health and trait characteristics as well as repeated measures of lifestyle activities, social living environment, and anxiety and mood states over the 9-month period. Explanatory multi-level models of anxiety state and mood state were constructed to examine relationships between these state outcomes and explanatory variables of trait, behavioral, and environmental measures. Relationships were designated as either a protective relationship (decreased anxiety, increased positive mood) or risk relationship (increased anxiety, decrease mood), but these terms are not meant to signify causality. These terms are only used to refer to the directionality of change in well-being measures toward a more beneficial or detrimental state. Significant factors (p<0.1) with protective relationships to anxiety and mood included: mood level trait (Mood Survey level score), extroversion, hours of sleep, number of meals eaten, time spent outdoors, time spent exercising, time spent on video games, and living alone. Additionally, factors that had significant protective relationships to anxiety alone were introversion, being required to social distance, and time spent on cooking. Conversely, significant factors (p<0.1) with risk relationships to anxiety and mood included: time spent on schoolwork and living with individuals who were not family or friends. Additionally, number of adults in the residence had a significant risk relationship to anxiety alone. Effect sizes of factors were evaluated with intra-class correlation coefficients (ICC) for random effects and semi-partial R2 for fixed effects and revealed that trait measures had the largest effects (particularly extroversion could explain the largest percentage of variance in outcomes), while behavioral and social environment factors had relatively small effects. The significance of trait variables corroborates evidence for the use of trait measures in evaluating whether individuals may demonstrate greater resilience to stressors during exposure to extended isolation and confinement. The relationships observed for lifestyle activities and social environment corroborate the protective benefits of operational psychology support resources currently being used in spaceflight that may translate benefit for a more diverse population. The findings that deviated from prior literature included protective relationships from living alone and being required to social distance (due to policy rather than self-selecting to social distance). These deviations may be due to concerns of virus communicability and health risk during the pandemic that are not generally relevant to operational ICE environments. However, statistical significance should not be equated to practical relevance. The relationships examined, whether statistically significant or not, can inform the development, prioritization, and personalization of behavioral health countermeasures and assessment tools for supporting civilians in ICE environments on Earth and future spaceflight missions.