From Data to Discovery: Applying CDE Best Practices in AD/ADRD Research

Common Data Elements (CDEs) are data elements (or variables) that are defined and used the same way across multiple studies, to standardize the way data is collected. The use of CDEs in research offers a powerful way to enhance data quality, enable system interoperability, promote standardization and harmonization, and improve research accuracy and efficiency. By providing standardized definitions and formats for key variables, CDEs facilitate cross-study comparisons, streamline data aggregation, and support robust analyses.

To help researchers harness this potential, the NIA Common Data Elements (CDE) Working Group convened the inaugural NIA CDE Speaker Series in 2024. This five-part, web-based series featured NIA-funded investigators who shared practical strategies for consistent data practices, enabling more reliable, comparable, and reproducible research results. By focusing on standardized, precisely defined questions paired with systematic responses, the series highlighted how CDEs — especially when aligned with accepted standards — support open science and adhere to FAIR (findable, accessible, interoperable, reusable) principles.

While CDEs provide a consistent framework for data collection, the series emphasized their broader role in enabling best practices across research studies. The use of CDEs promotes consistency in data collection, harmonization, and reporting. CDEs also support robust, unbiased experimental designs and facilitate replication in surveys, registries, and other research projects.

The five-part speaker series featured the following topics and speakers:

• Data Harmonization in Research on Epidemiology of Aging — Dr. Michelle Shardell (April 2024): Framed data harmonization to include missing data, measurement and misclassification errors, and the use of proxy respondents during data collection.
  Watch here
• Systematic Data Harmonization Enabled by Generative AI, ML, and LLMs — Dr. Mike Nalls (May 2024): Explored how artificial intelligence and large language models can support data harmonization and outlined a four-step process for sharing brain-related datasets among researchers.
  Watch here
• Application of Computational Data Harmonization Approaches to Study AD/ADRD — Dr. Tim Hohman (June 2024): Showcased harmonizing genetic and molecular data across cohort studies to inform therapeutic research for patients with AD/ADRD.
  Watch here
• Data Harmonization of Large Digital Technology Datasets for Aging and Dementia Research — Dr. Rhoda Au (July 2024): Highlighted effective approaches for using common digital tools in longitudinal research and the importance of monitoring tool evolution over time.
  Watch here
• Harmonization of Cognition Assessment Measures — Dr. Jinkook Lee (December 2024): Discussed the Gateway to Global Aging Program and its surveys, which address the need for nationally representative cognitive assessment measures.
  Watch here