Association between smoking status and changes in health behaviours during the COVID-19 lockdown: A cross-sectional study in Israel

Objectives To explore the association between smoking status (current vs former), changes in smoking habits (among current smokers only) and negative changes in dietary and physical activity habits during the initial COVID-19 lockdown in Israel. Design A secondary analysis of an online cross-sectional study (6 April 2020 to 28 April 2020). Dependent variables included deteriorations in dietary habits score and odds of reporting worsening of physical activity habits during the first COVID-19 lockdown. The study analysed the association between these variables and € smoking status' (current vs former) and, within smokers, changes in smoking habits, using multivariate logistic and linear regression models. For changes in physical activity habits, there was a significant interaction with baseline physical activity levels (p=0.04), therefore, analysis was stratified accordingly. Setting Online data collection. Participants 660 participants, current or former smokers in Israel, Hebrew speakers and ≥18 years old. Primary outcomes Self-reported negative changes in dietary habits and physical activity during the first COVID-19 lockdown. Results The sample (n=660) included 66.2% (n=437) current smokers and 33.8% (n=223) former smokers. Among current smokers, 43.5% (n=190) indicated an increase in their smoking habits. Of all respondents, 25% (n=170) reported a negative dietary change and 48% (n=192) spent less time engaging in physical activity, with 66% (n=437) reporting increased levels of stress. No significant association was found between smoking status and the dietary habits score in the multivariate linear regression (B=-0.046, CI -0.493 to 0.401, reference group: former smokers). This indicates that being a current smoker was associated with a non-significant 0.046 point decrease in the dietary habits score compared with former smokers. Similarly, no significant association was found between smoking status and odds of reporting worsening of physical activity, even after stratifying by baseline physical activity levels. Among participants who currently smoke, no significant association was found between changes in smoking habits and the dietary habits score (B=0.391, 95% CI -0.061 to 0.843, p=0.090, reference group: those who smoke the same or less) or with odds of reporting worsening of physical activity (OR=1.16, 95% CI 0.688 to 1.956, p=0.577, reference group: those who smoke the same or less). Conclusion Among current and former smokers, high rates of negative health habit changes were found, emphasising the need for interventions during future crises. Smoking status and/or changes in smoking habits among current smokers were not associated with negative changes in dietary and physical activity habits.