Longitudinal association between time-varying social isolation and psychological distress after the Great East Japan Earthquake

Background: The association between social isolation and psychological distress among disaster survivors is inconclusive. In addition, because these previous studies were cross-sectional in design, the longitudinal association between time-varying social isolation and psychological distress was not clear. The present study examined the longitudinal association between social isolation and psychological distress after the Great East Japan Earthquake. Methods: We analyzed longitudinal data for 959 adults who had responded to the self-report questionnaires about Lubben Social Network Scale-6 (LSNS-6) and K6 in both a community-based baseline survey (2011) and a follow-up survey (2014) after the disaster. Participants were categorized into four groups according to changes in the presence of social isolation (<12/30 of LSNS-6) at two time points (2011 and 2014): "remained socially isolated", "became not socially isolated", "remained not socially isolated", and "became socially isolated". We defined a K6 score of ≥10/24 as indicating the presence of psychological distress. We used multiple logistic regression analysis to estimate the adjusted odds ratios (ORs) and 95% confidence intervals (CIs) to indicate how the change in social isolation was related to changes in psychological distress over 3 years. Results: Among the participants who had not shown psychological distress at the baseline, the rates of deterioration of psychological distress were significantly lower in participants who "became not socially isolated" (multivariate OR = 0.26, 95% CI = 0.08-0.70) and "remained not socially isolated" (multivariate OR = 0.49, 95% CI = 0.27-0.91), compared with participants who "remained socially isolated". Among the participants who had psychological distress at the baseline, the rate of improvement of psychological distress was significantly higher in participants who "remained not socially isolated" (multivariate OR = 2.61, 95% CI = 1.08-6.44). Conclusion: The present findings suggest that prevention of social isolation may be an effective public health strategy for preventing psychological distress after a natural disaster.