Predicting physical symptoms through expressions of loneliness and anxiety in individuals utilizing social media during SARS-CoV-2

Abstract

The effect of the novel coronavirus (SARS-CoV-2) pandemic has produced significant health concerns negatively impacting individuals. As the ongoing and constantly changing nature of SARS-CoV-2 continues, the unique characteristics of this pandemic trend toward anxiety and loneliness as significant behavioral health outcomes. Furthermore, the SARS-CoV-2 pandemic has significantly impacted the utilization of social media platforms such as Twitter. Utilizing over 1.6 million tweets from approximately 988,760 Twitter users geolocated in Washington state from the University of Pennsylvania’s publicly available Twitter database (from March 2020 to March 2021), this study evaluated the impact of SARS-CoV-2 by using expressions of loneliness and anxiety to predict mental and physical symptoms. Bivariate correlations revealed expressions of loneliness were correlated to trouble breathing while expressions of anxiety were correlated to skin lesions, body aches, flu-like, seasonal cold, trouble breathing, nausea or diarrhea, fever, chills, and cough. Multiple multivariate linear regressions were completed, and a significant regression equation was found in predicting trouble breathing symptoms on expressions of loneliness and anxiety, however, the proportion of variance was 8% of the observed variation in the model. Further implications revealed the importance of understanding mental and physical well-being during a public health crisis as well as the use of social media platforms as primary and supplementary stimuli.