Background: The Great East Japan Earthquake of 11 March 2011 led to the relocation of 300 000 survivors. Studies following disasters focus primarily on data collected in the immediate aftermath and neglect the influence of wider community factors.
Aims: A three-level prospective study examining associations between survivors' psychological distress and individual- and social-level factors in the 6 years following a complex disaster.
Method: We drew on two multi-wave data collections in the 6 years after the earthquake, using residents from different forms of housing. Sample 1 included six waves of private-housing residents from 2011 to 2016 (n = 1084 per wave), sample 2 five waves of residents living in prefabricated housing from 2012 to 2016 (n = 1515 per wave). We analysed prospective associations between distress and time (level 1), pre-existing disorders and disaster experiences and behaviours (level 2) and city-wide measures of support and physical activity (level 3).
Results: Multilevel models with random coefficients demonstrated greater distress in earlier waves (samples 1 and 2 respectively, adjusted β = -15 and β = -0.16, P < 0.001), among female respondents (β = 0.58, P = 0.01 and β = 1.74, P = 0.001), in those with a previous psychiatric history (β = 2.76, β = 2.06, P < 0.001) with diminished levels of activity post-earthquake (β = 1.40, β = 1.51, P < 0.001) and those lacking in social support (β = 1.95, β = 1.51, P < 0.001). Support from spouses and friends was most protective of psychological health. City-level support was negatively associated with distress, but only among those in prefabricated housing.
Conclusions: Psychological distress diminished with time, but varied across gender, psychiatric history, housing, levels of activity and availability of social support. Practitioners should consider individual- and city-level factors when devising effective interventions.
Keywords: Disasters; Japan; multilevel analyses; psychological distress.