Determinants of social distancing adherence

Introduction: Governments and public health authorities across many jurisdictions implemented social (physical) distancing measures to contain the spread of the 2019 novel coronavirus disease (COVID-19). Adherence to these measures is variable and likely influenced by various factors. This study aimed to 1) identify the individual sociodemographic, COVID-19 and social distancing related, and psychological determinants of social distancing adherence, and 2) explore regional differences in social distancing adherence in the United States (U.S.) and English-speaking Canada based on each region's discrepant response to social distancing restrictions.Methods: A web-based repeated cross-sectional survey was conducted in 4,942 English-speaking participants from the four most populous U.S. states, specifically New York, California, Texas, and Florida, and Canada (www.covid19-database.com). The study was conducted at two timepoints, from May 1 to 5, 2020 (n = 1,019, Canadian participants only) and from July 6 to 10, 2020 (n = 3,923). Separate univariate models were computed for individual sociodemographic, COVID-19 and social distancing related, and psychological determinants of social distancing adherence. To determine the total variance explained, a univariate analysis including all of the determinants was performed. Regional differences in social distancing were compared between the four U.S. states and Canada, and between the U.S. as a whole and Canada.Results: Adherence to social distancing was higher in May (mean = 4.4/5.0 +/- 0.7) compared to July (mean = 4.3/5.0 +/- 0.7) [t((4940)) = 6.96, p < 0.001], likely a reflection of relaxing restrictions. There were no regional differences in adherence. Sociodemographic, COVID-19 and social distancing related, and psychological determinants explained 10, 36, and 23% of the variance of social distancing adherence, respectively. Higher perceived seriousness of COVID-19 [beta (SE) = 0.39 (0.01), p < 0.001, partial eta(2) = 0.22], lower risk propensity [beta (SE) = -0.15 (0.01), p < 0.001, partial eta(2) = 0.06], germ aversion [beta (SE) = 0.12 (0.01), p < 0.001, partial eta(2) = 0.03], age [beta (SE) = 0.01 (0.00), p < 0.001, partial eta(2) = 0.02], and greater social support [beta (SE) = 0.03 (0.00), p < 0.001, partial eta(2) = 0.02] had the largest effects on social distancing adherence.Conclusion: Public service initiatives to emphasize the serious consequences of infection and targeted interventions toward certain sociodemographic groups, such as younger adults and vulnerable individuals in greater need of social support, may help enhance the public's adherence to social distancing measures during subsequent waves of COVID-19 and future pandemics.