Interdepartmental Colloquium: From Wearable Data to Health Insights: Understanding Human Physiology Through Distributions, by Sujit K Ghosh(North Carolina State University)

October 28, 2026

Title: From Wearable Data to Health Insights: Understanding Human Physiology Through Distributions

Speaker: Sujit K Ghosh,Professor of Statistics, North Carolina State University.

Time: Friday, October 30, 11:00 a.m. - 12:00 p.m,  Austin 340.

Research Article: https://www.tandfonline.com/doi/full/10.1080/01621459.2025.2460232

Link: https://teams.microsoft.com/meet/210928975502383?p=leljKWWrUALInhshdo

Abstract: Wearable devices provide unprecedented opportunities to understand how human physiology and behavior evolve in everyday life. By collecting continuous streams of measurements, these devices capture not just average levels but the full range of daily experiences. However, analyzing such complex data presents new statistical challenges.

We develop Distributional Outcome Regression via Quantile Functions (DORQF), a framework that models entire distributions rather than reducing observations to single summary measures. The method accommodates both conventional predictors, such as age and gender, and distributional predictors, while ensuring coherent, interpretable relationships and providing tools to quantify uncertainty and assess statistical significance.

We apply DORQF to data from 781 older adults in the Baltimore Longitudinal Study of Aging, using a week of minute-by-minute heart rate and physical activity measurements. Our analysis investigates how daily heart rate patterns vary with physical activity, age, gender, and body composition, offering new insights into age-related changes in cardiovascular function. This work demonstrates how modern statistical methods can unlock richer scientific insights from wearable health data.

Bio of Sujit K Ghosh: Sujit K Ghosh is a Professor of Statistics at North Carolina State University and a Fellow of the American Statistical Association. His research spans Bayesian inference, biostatistics, spatial statistics, survival analysis, and statistical data science. He has authored over 155 refereed publications and the textbook Bayesian Statistical Methods, and has held leadership roles at the National Science Foundation and SAMSI. He received the 2023 UConn Statistics Department Distinguished Alumni Award.