Depression, anxiety and stress during COVID-19: associations with changes in physical activity, sleep, tobacco and alcohol use in Australian adults

The novel coronavirus (COVID-19) has enforced dramatic changes to daily living including economic and health impacts. Evidence for the impact of these changes on our physical and mental health and health behaviors is limited. We examined the associations between psychological distress and changes in selected health behaviors since the onset of COVID-19 in Australia. An online survey was distributed in April 2020 and included measures of depression, anxiety, stress, physical activity, sleep, alcohol intake and cigarette smoking. The survey was completed by 1491 adults (mean age 50.5 +or- 14.9 years, 67% female). Negative change was reported for physical activity (48.9%), sleep (40.7%), alcohol (26.6%) and smoking (6.9%) since the onset of the COVID-19 pandemic. Significantly higher scores in one or more psychological distress states were found for females, and those not in a relationship, in the lowest income category, aged 18-45 years, or with a chronic illness. Negative changes in physical activity, sleep, smoking and alcohol intake were associated with higher depression, anxiety and stress symptoms. Health-promotion strategies directed at adopting or maintaining positive health-related behaviors should be utilized to address increases in psychological distress during the pandemic. Ongoing evaluation of the impact of lifestyle changes associated with the pandemic is needed.

Applying Machine Learning to Identify Anti-Vaccination Tweets during the COVID-19 Pandemic

Anti-vaccination attitudes have been an issue since the development of the first vaccines. The increasing use of social media as a source of health information may contribute to vaccine hesitancy due to anti-vaccination content widely available on social media, including Twitter. Being able to identify anti-vaccination tweets could provide useful information for formulating strategies to reduce anti-vaccination sentiments among different groups. This study aims to evaluate the performance of different natural language processing models to identify anti-vaccination tweets that were published during the COVID-19 pandemic. We compared the performance of the bidirectional encoder representations from transformers (BERT) and the bidirectional long short-term memory networks with pre-trained GLoVe embeddings (Bi-LSTM) with classic machine learning methods including support vector machine (SVM) and naive Bayes (NB). The results show that performance on the test set of the BERT model was: accuracy = 91.6%, precision = 93.4%, recall = 97.6%, F1 score = 95.5%, and AUC = 84.7%. Bi-LSTM model performance showed: accuracy = 89.8%, precision = 44.0%, recall = 47.2%, F1 score = 45.5%, and AUC = 85.8%. SVM with linear kernel performed at: accuracy = 92.3%, Precision = 19.5%, Recall = 78.6%, F1 score = 31.2%, and AUC = 85.6%. Complement NB demonstrated: accuracy = 88.8%, precision = 23.0%, recall = 32.8%, F1 score = 27.1%, and AUC = 62.7%. In conclusion, the BERT models outperformed the Bi-LSTM, SVM, and NB models in this task. Moreover, the BERT model achieved excellent performance and can be used to identify anti-vaccination tweets in future studies.