Researchers have created a more precise model to understand how genetics and air pollution contribute to the development of diseases.
Diseases arise from a combination of genetic factors and environmental influences such as air pollution, climate, and social conditions. However, it’s not entirely clear how much each of these factors contributes to the risk of developing a disease, which often leaves individuals uncertain about how to lower their disease risk.
A research team from Penn State College of Medicine discovered a method to differentiate the impacts of genetics and environmental elements on disease risks using a substantial, nationwide sample. They found that some previous studies had exaggerated the influence of genetics on disease risk, revealing that lifestyle and environmental aspects have a bigger role than previously thought. Unlike genetic factors, environmental influences—such as air pollution—can often be changed, presenting more opportunities to reduce disease risk. Their findings were shared in Nature Communications.
“Our goal is to clarify how much genetics and how much the environment contribute to disease development. A clearer understanding of each factor’s role can enhance our ability to forecast disease risk and formulate targeted interventions, especially as we move into the realm of precision medicine,” stated Bibo Jiang, assistant professor at Penn State College of Medicine and the study’s senior author.
The researchers highlighted that measuring environmental risk factors has been challenging in the past due to their broad scope, including elements like diet, exercise, and climate. Without accounting for these environmental factors, analyses may mistakenly assign family disease risks primarily to genetics.
“Neighbors tend to face similar air pollution levels, socio-economic status, access to healthcare, and food availability,” explained Dajiang Liu, distinguished professor and co-senior author of the study. “If we can separate these common environmental factors, what remains can provide a more accurate indication of genetic heritability related to diseases.”
In their study, the team developed a spatial mixed linear effect (SMILE) model that takes both genetic information and geographical data into account. This geolocation data helped represent community-level environmental risks.
Utilizing data from IBM MarketScan, a health insurance claims database covering the electronic health records of over 50 million individuals with employer-based health insurance in the United States, the team analyzed over 257,000 nuclear families and examined disease outcomes for 1,083 different diseases. They also integrated publicly available environmental information, such as climate data, socio-economic demographics, and pollution levels (specifically particulate matter 2.5 (PM2.5) and nitrogen dioxide (NO2)).
Their analysis yielded more accurate estimates of the influences on disease risk. For example, previous studies indicated that genetics accounted for 37.7% of the risk for developing Type 2 diabetes. However, their model, which accounted for environmental impacts, revised this estimate downward to 28.4%, indicating that environmental factors had a more significant influence. A similar revision was observed for obesity risk, where the genetic contribution decreased from 53.1% to 46.3% after accounting for environmental conditions.
“Earlier studies suggested that genetics had a more dominant role in predicting disease risk, but our findings suggest otherwise,” Liu noted. “This means that individuals with a family history of Type 2 diabetes can remain optimistic, as there are various actions they can take to lower their personal risk.”
The research team also assessed the specific effects of two air pollutants—PM2.5 and NO2—on disease risks. Historically lumped together, the study found that these pollutants exert different causal influences on health. NO2, for example, was found to directly lead to conditions such as high cholesterol, irritable bowel syndrome, and both types of diabetes, while PM2.5 was associated with lung function issues and sleep disorders.
Ultimately, the researchers believe this model will facilitate a deeper exploration of why certain diseases are more common in specific geographic regions.
Other contributors to the research from Penn State include: Havell Markus and Austin Montgomery, both dual-degree medical students; Laura Carrel, a professor of biochemistry and molecular biology; Arthur Berg, a public health sciences professor; and Qunhua Li, a statistics professor. The study was also co-led by Daniel McGuire, a former doctoral student in biostatistics, along with doctoral students Lina Yang and Jingyu Xu.
This research was partially funded by the National Institutes of Health and the Penn State College of Medicine’s artificial intelligence and biomedical informatics pilot funding program, with some materials provided by the Center for Applied Studies in Health Economics at the Penn State College of Medicine.
