Abstract
BACKGROUND/OBJECTIVE: Loneliness is a risk factor for cognitive decline in aging and other clinical populations, but its role in long COVID (LC) remains poorly understood. Individuals with LC may be particularly vulnerable to loneliness due to debilitating, persistent symptoms and reduced functioning. We examined associations between overall loneliness and cognition in LC versus recovered controls, and whether loneliness subtypes (social, emotional) differentially relate to cognitive function. METHODS: Individuals meeting 2024 National Academy of Science, Engineering and Medicine criteria (NASEM) for LC and reporting at least one neuropsychiatric symptom (n = 120), along with recovered controls (n = 51), completed the 6-item De Jong Gierveld Loneliness Scale and a cognitive test battery. Correlation analyses, corrected for false discovery rate, identified bivariate loneliness-cognition associations. Significant correlations were followed by age-adjusted regressions using residualized loneliness scores, which excluded variance shared with depression and social isolation. RESULTS: LC participants reported higher overall, emotional, and social loneliness than controls. In bivariate analyses, greater overall and emotional loneliness were associated with more subjective cognitive complaints and poorer verbal fluency in the full sample and LC group, and with cognitive complaints and poorer verbal memory in controls. In adjusted models, residual overall and emotional loneliness remained significantly associated with fluency in LC and the full sample, and with memory in controls. Associations with cognitive complaints did not persist. Social loneliness showed weaker and inconsistent associations. CONCLUSION: Overall and emotional loneliness are independently linked to objective cognitive difficulties. Findings highlight emotional loneliness as a potential target for cognitive intervention in LC and recovered individuals.