Effect of COVID-19 infection related experiences on social behaviors when a state of emergency is declared: a cohort study

BACKGROUND: Restricting the movement of the public to gathering places and limiting close physical contact are effective measures against COVID-19 infection. In Japan, states of emergency have been declared in specific prefectures to reduce public movement and control COVID-19 transmission. We investigated how COVID-19 infection related experiences including people with a history of infection, people with a history of close contact, and people whose acquaintances have been infected, affected self-restraint from social behaviors during the second state of emergency in Japan. METHODS: A prospective cohort study was conducted among workers aged 20–65 years using data from an internet survey. The baseline survey was conducted on December 22–25, 2020, and a follow-up survey was on February 18–19, 2021. There were 19,051 participants who completed both surveys and were included in the final analysis. We identified eight social behaviors: (1) eating out (4 people or fewer); (2) eating out (5 people or more); (3) gathering with friends and colleagues; (4) day trip; (5) overnight trip (excluding visiting home); (6) visiting home; (7) shopping for daily necessities; and (8) shopping for other than daily necessities. We set self-restraint regarding each social behavior after the second state of emergency was declared in January 2021 as the dependent variable, and COVID-19 infection related experiences as independent variables. Odds ratios were estimated using multilevel logistic regression analyses nested in the prefecture of residence. RESULTS: Significant differences by COVID-19 infection related experiences were identified: compared to people without COVID-19 related experiences, people with a history of COVID-19 were less likely self-restraint from most social behaviors. People whose acquaintance had been diagnosed with COVID-19 were significantly more likely to refrain from most social behaviors. There was no significant difference in any social behaviors for people with a history of close contact only. CONCLUSION: To maximize the effect of a state of emergency, health authorities should disseminate information for each person in the target population, taking into account potential differences related to the infection related experiences.