In This Story
Any pregnant woman in her third trimester can attest to feeling short of breath. Pregnancy in a low-oxygen environment, then, presents a challenge. A study in the Proceedings of the National Academy of Sciences of the United States of America (PNAS), however, demonstrates how Tibetan women living at high altitudes have evolved to carry oxygen. A multidisciplinary research team comprising anthropologists, doctors, and statisticians found that those women who carried oxygen more efficiently had more lifetime reproductive success.
While the study was led by Case Western Reserve University biological anthropologist Cynthia Beall and George Mason University statistician Jiayang Sun, statistics PhD candidate Shenghao Ye is the first author on the paper—a significant feather in his cap.
"Being part of the study published in PNAS is both exciting and humbling,” said Ye. “Under the guidance of my supervisor [Sun] and through collaboration with her and other professors, I’ve learned a great deal throughout this study. This research also lays an important foundation for my doctoral dissertation."
The collaborative effort, which took advantage of the natural laboratory among women on the Tibetan Plateau, provided valuable insights into human adaptation to environmental challenges and has potential applications in medical research, particularly for diseases associated with low oxygen levels, said Sun.
The researchers gathered data through interviews and specimens from 417 Tibetan women aged 46 to 86 living at altitudes of 11,500 feet or higher in Upper Mustang, Nepal. By analyzing demographic, genetic, environmental, and sociological factors, they examined how various physiological traits related to oxygen delivery were associated with reproductive success among the women, which were measured by the women’s number of live births. An important finding is that that those with higher reproductive success had hemoglobin cells that carried more oxygen. Their findings suggest multiple ways to maintain homeostasis and achieve many live births, explained Sun.
Statisticians will find the study especially exciting for inspiring the creation of a new explainable model learning strategy, Sun explained. Rather than a traditional linear model, which would be too simple to account for the changes in childbearing rates over the course of a woman’s life, or a fully nonparametric model, which would not be sufficiently transparent for identifying important factors influencing reproductive success, their new method of model development bridges between these two. This amalgamation helps fit biologically sensible models to data relating to human health, which is crucial for identifying accurate associations.
Ye conducted much of the study’s comprehensive statistical analyses, including explainable model learning and tree-based analyses, in collaboration with and under Sun’s guidance. Their new learning strategy has broad applications beyond this study and will constitute a significant portion of Ye’s doctoral dissertation.
This publication represents a significant achievement in both evolutionary biology and statistics. By combining interdisciplinary collaboration with rigorous data collection and a new statistical analysis technique, the study offers valuable insights into how human populations adapt to challenging environments. The team’s research not only enhances our understanding of evolutionary processes but also opens new avenues for scientific exploration and practical applications.