Accelerated Age and the Electrocardiogram

Machine learning algorithms are increasingly being applied to electrocardiograms (ECGs) to extract diagnostic and prognostic information not readily apparent from traditional interpretations. One observation is that ECGs can predict age on average, and deviations may carry prognostic information. Investigators used data from the longitudinal Framingham Heart Study (FHS) to validate and extend the prior studies by determining whether differences between chronologic age and ECG age can predict long-term cardiovascular outcomes.

The researchers evaluated FHS cohorts from 1986 to 2021, including 9877 participants with a mean age of 55 years. Over half were women. They determined that the difference between ECG age and chronological age was assoc…