Alum Mary Regina Boland Connects the Dots Between Climate Data and Disease

This post is part of the People of DBMI series.

If you’re thinking about having a baby, you might want to try to steer clear of giving birth in the autumn.

At least in New York City, where Mary Regina Boland, PhD, and her Columbia University colleagues studied the ties between birth month and disease, finding that October and November babies are most correlated with increased risk of neurological, reproductive, and respiratory illnesses.

Today an assistant professor of informatics at the University of Pennsylvania, Boland conducted the research as part of her dissertation as a PhD candidate of Columbia’s Department of Biomedical Informatics. She and her team examined more than 1.7 million patient records at Columbia University Medical Center, looking at climate and seasonality as one environmental factor that plays an important role in pregnancy conditions.

The birth-month study is one in a string of research Boland undertook at Columbia looking at how clinical data like dental and medical records are connected. Also as a DBMI student, she found that periodontitis, or gum infection, is linked to, interestingly enough, diabetes, hypertension, and prostate inflammation.

Boland earned her MA from DBMI in 2012, her MPhil in 2016, and her PhD in 2017.

Throughout her career, Boland has been dedicated to assessing, engaging with, and making sense of data. She’s curious about the effects of different types of environmental exposures on not only prenatal and perinatal development, but also other health outcomes.

In July, Boland published a study of climate impact on hospital performance metrics, which found that, even after adjusting for socioeconomic factors like income, exposure to colder climates resulted in higher 30-day mortality rates. The research gathered scores from more than 4,500 hospitals in over 2,300 counties across the nation.

The challenge now is to apply the data, going beyond mere collection and understanding to use them for actual treatment. It’s about “closing the loop,” as Boland puts it. “In informatics, you try to build models to understand latent factors,” she says, and design methods for integration.

Research like Boland’s could be key to devising better health, as the field of biomedical informatics works increasingly more on the consumer end to develop health tracking and analysis systems. And such evaluations will be better, more robust, and more holistic, taking into account an individual’s entire environment.

“We’re going to have a better assessment of things that contribute to disease risk,” Boland says. “As we continue to move toward the future, there’s going to be less of an emphasis on genetics being the way to solve everything.”