Degree of housing damage not a factor in mortality risk following Great East Japan Earthquake: Follow-up survey of 60,000 people by Tohoku University

A research group has found no statistically significant correlation between the extent of housing damage in the Great East Japan Earthquake and mortality risk. The group used data from the community-based cohort study of the Tohoku Medical Megabank (TMM) Project. This study was conducted by Tohoku University Tohoku Medical Megabank Organization (ToMMo) and Iwate Medical University Iwate Tohoku Medical Megabank Organization (IMM). It followed up on approximately 60,000 people aged 20 years or older in Miyagi and Iwate prefectures for an average of 6.5 years. The research team included Professors Naoki Nakaya and Atsushi Hozawa of ToMMo, and Professor Kozo Tanno and Lecturer Yuka Kotozaki of IMM. There had been concerns that the greater the housing damage, the more severe the health effects of long-term stress would be. Public health activities conducted in the disaster-stricken areas may have reduced the mortality risk. The research group commented that a cause-of-death study is needed to further analyze the impact. The group's findings were published in the January 15 issue of the Journal of Epidemiology and Community Health, an international journal.

After a major natural disaster, medical and financial assistance can help mitigate physical injury and financial loss for survivors. However, social isolation because of relocation after a disaster is expected to increase stress reactions. In 2022, the ToMMo research group compiled a summary of previous papers on the health effects of the Great East Japan Earthquake and reported various types of health exacerbations in the Japanese Journal of Cardiovascular Disease Prevention. Nine previous studies on natural disaster damage and all-cause mortality risk have been reported. However, all of them focused on the first year after the disaster. Thus, there was a lack of research and studies examining long-term effects.

In this study, the research group analyzed the association between the degree of housing damage and mortality risk. They used data from a cohort survey of residents in Miyagi and Iwate prefectures. The survey was conducted under the TMM project. The survey's follow-up has continued since the initial study (baseline study) to the present. Of the approximately 80,000 who agreed to participate in the baseline survey (conducted in 2013-16) of the community-based cohort study conducted under the TMM project, 58320 were included in the analysis. Deaths were confirmed by viewing the Basic Resident Registry, and 1,763 deaths were identified and analyzed with an average of 6.5 years (maximum 8 years) of follow-up until December 31, 2009.

Data on the extent of housing damage were obtained from responses to the baseline survey questionnaire. The survey form asked respondents to choose from the following options: "did not live in the disaster area"; "no damage"; "partial damage"; "medium-scale partial destruction"; "large-scale partial destruction"; or "complete destruction or completely washed away" at the time of the disaster.

A multivariate analysis using the Cox proportional hazards model was conducted. The "no damage" group was used as the base line (mortality risk: 1.00). The mortality risk for those "did not live in the disaster area" was 0.96. For those in the small- to medium-scale damage group ("partial damage" or "medium-scale partial destruction"), it was 0.98. Mortality risk was 0.98 for those in the large-scale damage group ("large-scale partial destruction" or "complete destruction or completely washed away"). The degree of housing damage and mortality risk showed no statistically significant correlation. Stratified analysis by sex, age, and location of implementation (ToMMo, IMM) showed no significant associations in any strata.

Nakaya said, "In this study, we hypothesized that long-term stress exposure would activate stress mechanisms and cause fatal health effects. However, this hypothesis was not supported. There are three possible reasons for this. First, this study was a long-term study, up to 8 years, but even longer-term observation may show negative health effects. Second, this study looked at all deaths. We believe it is necessary to analyze specific causes of death, such as stress-related cardiovascular disease, heart failure, or suicide. This study has disadvantages, including a lapse of approximately two years after the Great East Japan Earthquake. However, we believe that the public health activities of the national and local governments may have reduced the mortality risk."

Journal Information
Publication: Journal of Epidemiology and Community Health
Title: Degree of housing damage caused by the Great East Japan Earthquake and all-cause mortality in the community-based cohort study of the Tohoku Medical Megabank Project
DOI: 10.1136/jech-2024-223084