Events

A Practical Guide to Analysis of Continuous Glucose Monitoring (CGM) Data

Continuous glucose monitors (CGMs), small wearable devices that automatically measure interstitial glucose levels at a 5-minute frequency, are widely used in clinical practice and research. Multiple digital biomarkers derived from CGM data have been proposed to quantify various aspects of glycemic control, with the resulting CGM metrics used both as digital endpoints in clinical trials and as explanatory variables for other outcomes of interest. 

In this talk, Dr. Gaynanova will review the complete CGM data processing pipeline based on the open-source R package iglu developed by her group, starting from raw data visualizations to evaluating data sufficiency and missing data patterns, and computing and comparing consensus metrics of glycemic control. She will use publicly available CGM data for illustration and also point to University of Michigan resources with pre-computed CGM metrics for U-M patients. Dr. Gaynonova will also discuss additional data quality control measures that may be needed with real-world data from data warehouses. She will conclude with a brief overview of her group's efforts in CGM Data Analysis 2.0, which introduces new methods based on functional and distributional representations of CGM data. These methods move beyond traditional metrics to enable more personalized diabetes management and a more granular examination of glycemic patterns.

Irina Gaynanova, Ph.D., is an associate professor of biostatistics at the University of Michigan School of Public Health. 

Register to attend

This internal event is open to U-M faculty, staff, trainees and learners only. Register for our 2025-26 DIGIT-MI season to receive session details and the Zoom link.

REGISTER

DIGIT-MI is an affinity group where University of Michigan faculty and staff meet to learn and share resources that encourage the use of digital and mobile technologies, as well as mental wellness measures, in research. Virtual sessions are held from 1 to 2 p.m. on the second Monday of the month from October through May.