Investigating the (in)stability of resting baseline measures of cardiac activity

Standard practice for collecting and analyzing cardiac activity data involves establishing a baseline of cardiac activity and assessing deviations from this baseline. However, empirical data on how these baseline measurements may vary within individuals over time and across contexts remains sparse. Here, we assess resting baseline measures of interbeat interval (IBI) and high-frequency heartrate variability (HF-HRV) for 124 participants during four separate 5-minute baseline sessions over a 6-week period. Analyses suggest that only about half of the variance in the data for each measure was attributable to between-subject differences (IBI=59.86%; HF-HRV: 51.49%) and a third or more of the variance was attributable to between-session differences (IBI: 35.93%; HF-HRV: 30.36%). Minute-to-minute variability was minimal (IBI: 4.20%; HF-HRV: 18.14%), particularly for IBI which did not change significantly over the course of the 5-minute baselines on average; HF-HRV was significantly higher in the first minute of the baseline than the others on average. Investigation of sources of person-level and session-level variability revealed that both IBI and HF-HRV were significantly higher during the first session than latter sessions, and HF-HRV was significantly higher when the baseline was collected after instead of before a 20-minute task. Older age, higher BMI, and less reported weekly exercise were also associated with lower HF-HRV at the between-subjects level. Other person- and session-level variables examined (e.g., experimenter characteristics, weather, time of day) did not significantly account for any observed variability. Methodological implications are discussed.