Predictors of Loneliness and Psychological Distress in Older Adults During the COVID-19 Pandemic: National Cross-Sectional Study.

Background: The COVID-19 pandemic has significantly affected the mental health of older adults, particularly through increased loneliness and psychological distress. While various contributing factors have been studied, the role of preference for solitude as a potential predictor and mediator remains poorly understood. Objective: This national cross-sectional study aimed to examine predictors of loneliness and psychological distress among older adults during the pandemic, with a specific focus on preference for solitude and its mediating role between pandemic-specific stressors and self-efficacy. Methods: A total of 2053 Croatian residents aged 65 years and older were recruited using snowball sampling. Validated instruments were used, including the UCLA Loneliness Scale, Preference for Solitude Scale, Clinical Outcomes in Routine Evaluation-Outcome Measure, General Self-Efficacy Scale, and Pandemic-Specific Stressors Questionnaire for Older Adults. Hierarchical regression and path analysis were used, with statistical significance set at P<.05. Results: For loneliness, the final model explained 30.7% of the variance, with a coefficient of determination (R²) of 0.307 and a root mean square error of 0.708 (P<.001). Significant predictors included marital status (eg, never married: B=0.390, P<.001; psychological problems: B=0.020, P<.001; functionality: B=-0.037, P<.001; and social distancing: B=0.014, P<.001). Preference for solitude was also a significant predictor of loneliness (B=0.011; P<.001). For psychological distress, the final model explained 30.8% of the variance (R²=0.308; root mean square error=14.872; P<.001). Self-efficacy emerged as the strongest negative predictor of distress (B=-1.066; P<.001), whereas preference for solitude was a positive predictor (B=2.403; P<.001). The variable "spending several hours alone per day" was associated with lower levels of distress (B=-3.509; P=.003), while "frequent or superficial interactions with acquaintances (eg, at least once a week)" were related to higher distress (B=4.321, P=.002). Path analysis revealed that both social distancing and exposure to infection had a significant direct effect on self-efficacy: negative in the case of social distancing (β=-.391; P<.001) and positive for exposure to infection (β=.386; P<.001). However, preference for solitude did not significantly mediate either relationship, as indicated by nonsignificant indirect effects (β=-.006; P=.09 and β=.002; P=.48, respectively). Conclusions: Psychological problems and reduced functionality emerged as the strongest predictors of loneliness among older adults during the pandemic. Self-efficacy was the most important protective factor against psychological distress. Although preference for solitude may have adaptive benefits, in the context of this study, it was associated with increased loneliness and distress during enforced isolation. These findings suggest that public health interventions should balance respect for individual preferences with the provision of active support for vulnerable populations during crises.