Reproducibility of 24-h ambulatory blood pressure and measures of autonomic function: methodological and statistical issue
I was interested to read the paper by Morrin et al. 1 published in January 2017 issue of the Blood Pressure Monitoring. The authors aimed to assess the reproducibility of 24-h ambulatory blood pressure, 24-h heart rate variability (HRV), and resting measurements of HRV and blood pressure variability 1. As the authors mentioned, accurate recordings of ambulatory blood pressure monitoring and autonomic function are prerequisite for the appropriate design of intervention studies 1. It is good to know that accuracy and reproducibility are different methodological issues 2. According to their results, ambulatory HRV was more reproducible than resting measures. The most reproducible markers were root mean square of successive differences [coefficient of variation (CV): 13.2–10%] and high-frequency normalized units (CV: 15.2–6.4%), with the percentage of adjacent NN intervals differing by more than 50 ms showing the poorest reproducibility (CV: 23.9–20.7%) 1.
It is crucial to know that in reproducibility analysis an individual-based approach instead of group–based one should be considered 2–7. Therefore, for quantitative variables, intraclass correlation coefficient agreement single measure instead of group measure (CV) should be reported 2–7.
As the authors pointed out in their conclusion, overall blood pressure variability (SD) was more reproducible than the frequency domain low-frequency component. Ambulatory HRV provide superior reproducibility to resting measurements. Such a conclusion can be misleading because of the global average approach being is considered to assess reproducibility.