[Evaluation parameters of ambulatory blood pressure monitoring: mean values versus percent values].

In 1039 ambulatory blood pressure (BP) recordings, we studied if mean values or blood-pressure load (% of elevated readings) is superior for evaluation. There was a nonlinear correlation between the two parameters with high correlation coefficients (syst. r = 0.98, diast. r = 0.97). At very high (low) BP levels the differentiation of the percentage-scale is insufficient, whereas in the central BP range it is too detailed as compared with the accuracy and reproducibility. Due to the nonlinear correlation with BP, treatment effects cannot be easily compared by percentage-values, since identical BP differences correspond with varying percentage. Intervals between readings are usually shorter during daytime than during nighttime. In contrast to the BP load, the resulting problems for 24-h statistics can be easily avoided by calculation of weighted mean values (from hourly means). In summary, weighted mean values are rather suited for evaluation of ambulatory blood pressure than is BP load.