Housing damage resulting from great east Japan earthquake and all-cause mortality

The work is published in the Journal of Epidemiology and Community Health on January 15, 2025.

On March 11, 2011, the GEJE and tsunami that followed devastated the northeastern coast of Japan. By March 1, 2021, there were 22,318 people dead or missing because of the earthquake and tsunami. The team wanted to understand the negative health effects the disaster had on the survivors, with a focus on the relationship between housing damage the people experienced and all-cause mortality.

For their research, the team used data from community-based cohort study conducted by the TMM Project in disaster-stricken areas. The TMM Project conducted their baseline survey from May 2013 to March 2016 using questionnaires, blood and urine tests, and detailed examinations. More than 59,000 participants took part in the survey. The GEJE survivors had experienced a wide range of physical and mental health challenges including depression, post-traumatic stress disorder, anxiety, liver dysfunction, diabetes, hypertension, and chronic kidney disease, as already revealed by many previous reports.

The research team analyzed their data using the Cox proportional hazards model. This method allows researchers to examine the survival time of patients, where there are several variables present.

"The results of this study showed that the degree of housing damage caused by the GEJE did not affect the overall mortality," said Naoki Nakaya, a professor at Tohoku University. These results may have occurred because the elapsed time until death was short or because the public health activities of each local government may have suppressed the increase in mortality. "For example, diseases may have been prevented by eliminating the burden of medical expenses in disaster-stricken areas and ensuring access to medical care by strengthening the medical system. This finding may also provide evidence that public health efforts do not increase mortality," said Nakaya.

Earlier studies had examined the association with all-cause mortality after natural disasters through observational studies with limited follow-up periods. What sets this study apart is that the researchers adopted a prospective cohort design and followed up approximately 60,000 participants over a long-term period of 6.5 years. They analyzed over 1,700 deaths and considered possible confounding factors.

The research team notes some limitations in their study. The team assessed the degree of damage to houses caused by the GEJE using self-administered questionnaires. The baseline survey period was three years, from May 2013 to March 2016, and misclassification may have occurred because of the participants' recall errors. Also, their baseline survey began approximately two years after the GEJE. Early deaths after the GEJE could not be included in this cohort study. If early deaths were higher among residents who experienced large-scale housing damage, the team notes that this study might have underestimated their mortality risk.

Another limitation is that this study was conducted at two institutions that used different follow-up methods for mortality up until 2021. These different methods could have allowed some deaths to have been overlooked. The team also notes that the social, physical, and lifestyle factors

investigated in this study may have changed after the survey. These factors could have changed because of the damage caused by the GEJE. And because participation was voluntary, this study might have been biased toward healthier people in the community. Although this bias might not affect the internal validity, the team urges caution when generalizing the results regarding the degree of housing damage or mortality.

"Further long-term follow-up studies are needed to examine the long-term health effects of natural disasters on survivors," said Nakaya. The team also notes the need for analysis of the causes of death as outcomes, including cardiometabolic diseases and mental illnesses.