62
NATURE’S IMPACT ON THE HEART: 
UNDERSTANDING ENVIRONMENTAL TRIGGERS
Out-of-hospital cardiac arrest (OHCA) is 
influenced by a complex interplay of factors, 
with the environment emerging as an important 
and often underrecognized contributor. 
Environmental conditions, including weather 
patterns, air pollution, and wildfire smoke, 
can influence cardiovascular physiology and 
contribute to acute health events. While these 
exposures are often studied in the context of 
chronic disease, a growing body of evidence 
suggests they may also act as triggers for life-
threatening events such as cardiac arrests. 
Understanding how environmental factors 
interact with cardiovascular risk is essential for 
advancing prevention strategies and improving 
population health outcomes.
The ability to study these relationships at scale 
has historically been limited by challenges in 
linking nationwide cardiac arrest data with 
environmental exposure data. Over the past 
several years, the CARES team has been able 
to securely link environmental datasets, including air quality metrics and weather data, to the CARES 
registry at the geographic level, while de-identifying the dataset before it is made available for research. 
This honest-broker approach enables precise investigation of how environmental exposures influence 
the distribution of cardiac arrest events, opening opportunities for identifying vulnerable populations 
and informing evidence-based mitigation strategies.
One example of linking environmental data with cardiac arrest outcomes is the study by Takahiro 
Nakashima and colleagues (2025), Development and Evaluation of a Machine Learning Model Predicting Out-
of-Hospital Cardiac Arrest Using Environmental Factors. The study aimed to predict daily OHCA incidence at 
a regional level by combining environmental exposures with national cardiac arrest records. 
To capture environmental influences, the study integrated high-resolution meteorological data, including 
temperature, humidity, barometric pressure, and wind speed, obtained from NASA’s North American 
Land Data Assimilation System (NLDAS), along with sociodemographic information such as population 
density and age distribution. Linkage was performed at the county level, matching NLDAS meteorological 
data and sociodemographic datasets to CARES OHCA records by incident location and date. This created 
a large-scale, de-identified dataset suitable for machine learning while protecting patient privacy. 
Researchers analyzed more than 420,000 EMS-treated non-traumatic OHCA cases from the CARES 
registry collected between 2013 and 2019.
Using gradient boosting machine learning, the researchers identified 17 predictors most closely linked 
to daily OHCA events. Extreme temperatures, rapid changes in weather, and seasonal trends were 
consistently associated with higher rates of cardiac arrest. Sudden shifts in temperature or atmospheric 
pressure can challenge the body’s ability to adapt, particularly among individuals with underlying heart 
disease. Seasonal patterns further reinforced these findings; winter months were often associated with 
increased cardiac events, potentially due to colder temperatures, higher rates of respiratory illness, 
reduced physical activity, and elevated blood pressure.
Weather and OHCA: Predictive Modeling

View this content as a flipbook by clicking here.