New research examining 17 years of data from Medicare hospitalization claims and major flooding events finds increased rates of skin diseases, nervous system diseases, and injuries or poisonings among adults aged 65 and older following major floods. Researchers from Harvard T.H. Chan School of Public Health, Ohio State University, University of Wisconsin–Madison, and Columbia University Mailman School of Public Health published the findings in the journal Lancet Planetary Health .
While past research has been limited to a single flood or a small set of health outcomes, with little information on older individuals, the new study provides robust, generalizable findings that can inform flood preparation in high-income countries.
The researchers matched Medicare hospitalization data from 2000 to 2016, representing adults aged 65 years and older, with satellite-based, high-resolution historical flood maps from the Global Flood Database. They estimated relative percentage changes in hospitalization rates for 13 disease categories within four weeks following flood exposure. These included 72 significant flood events and over 4.5 million hospitalizations.
They observed elevated rates of hospitalization on average during and following flood exposure for skin diseases (3⋅1%), nervous system diseases (2.5%), musculoskeletal system diseases (1.3%), and injuries or poisoning (1.1%). Communities with lower proportions of Black residents experienced worsened effects for nervous system diseases (7.6%), whereas skin diseases (6.1%) and mental health-related impacts (3.0%) were more pronounced for areas with larger percentages of Black residents during flood exposure.
Older adults are particularly vulnerable to flood-related health impacts because of their weakened immune systems, restricted mobility, limited ability to cope with climate hazards due to pre-existing conditions such as dementia, and poor access to medical services for routine appointments.
Skin diseases may be the result of exposure to polluted water sources, crowded shelters, and poor sanitation. Seizures could arise from traumatic brain injuries sustained during floods. For individuals with epilepsy, flood events can induce stress and fatigue that might result in poor seizure control. Musculoskeletal system hospitalizations may result from underlying conditions and delays in seeking care, as well as injuries from flood clean-up. Distinct patterns of pre-existing conditions, housing quality, and access to emergency resources may explain why some health impacts of flooding are more acute for Black communities. Another possible explanation is differences in access to care and implicit biases in the coding of disease conditions between racial groups.
"The findings of this study provide crucial new insights into the diverse, and previously underappreciated, health consequences of floods in older adults and can guide flood-specific resilience-building efforts to protect public health under climate change," the researchers write. "Targeted outreach and robust evacuation planning for vulnerable populations, such as older individuals, along with community-based alert systems, are crucial to minimizing health impacts. Hospital infrastructure should be equipped to function during flood events by moving essential components above flood levels, and mobile medicine units and telemedicine can serve as effective alternatives if access to hospitals is temporarily eliminated. Drones can also deliver essential medical supplies to flood-affected hospitals or help identify safe evacuation routes in real time to guide emergency responders," they conclude.
The study's lead author is Sarika Aggarwal, a PhD candidate at Harvard Chan School. Additional authors include Jie K. Hu ( Ohio State University), Jonathan A. Sullivan (University of Wisconsin–Madison), Robbie M. Parks (Columbia University Mailman School of Public Health), and Rachel C. Nethery (Harvard Chan School).
Funding for the study was provided by the Harvard Data Science Initiative, Harvard Graduate Prize Fellowship, Sloan Foundation (grant number G20201394), and National Institutes of Health (grants ES032458, ES000002, ES033742, ES007142, ES7069).
The authors declare no competing interests.
