3D.09

Abstract

Objective: Increased blood pressure (BP) variability is a possible independent risk factor for cardiovascular events. BP variability has been assessed with several methods of BP measurement in recent literature, although it is unclear whether these measurements of variability with varying timeframes reflect the same phenomenon. The aim of our study was to compare the agreement between ambulatory, home and office BP variability. Design and method: The study population consisted of 509 participants randomly drawn from the population register or recruited by general practitioners on the basis of newly diagnosed untreated hypertension. Ambulatory 24-h blood pressure monitoring, 28 home BP measurements (twice every morning and evening during 7 consecutive days) and 8 office BP measurements (duplicate measurements on 4 visits) were performed in all participants. 3 log-transformed variability indices (SD, standard deviation; CV, coefficient of variation and ARV, average real variability) were calculated for all measurement methods and Pearson's correlations between them were calculated. The agreement of different methods on the diagnoses of extreme BP variability (participants with variability above the highest decile) was also assessed with kappa coefficients. Results: Systolic/diastolic BP variability was greater in 24-h ambulatory (CV: 12.6 ± 2.8 /15.1 ± 3.4) than home (CV: 4.4 ± 1.8/4.7 ± 2.0, p < 0.001 versus ambulatory CV for both) and office (CV: 4.8 ± 2.6/5.3 ± 2.6, p < 0.001 versus ambulatory CV for both) measurements. Ambulatory daytime variability was greater than night-time variability (CV: 11.0 ± 2.8/13.1 ± 3.5 vs. 9.5 ± 3.8/12.8 ± 4.6, p < = 0.001 for both). Pearson's correlation coefficients for systolic/diastolic variability indices between different measurement methods were 0.08–0.34/0.03–0.26, indicating only negligible to weak positive relationship (Table). The agreement of ambulatory, home and office BP variability measures on diagnoses of extreme systolic/diastolic BP variability was only slight, with the kappa coefficients varying between 0.00–0.20/0.03–0.16. Extreme variability was diagnosed in only two persons with all three methods.Conclusions: Shorter-term and longer-term BP variability assessed with different methods of BP measurement seem to correlate only weakly with each other. Our study suggests that BP variability assessed by different methods and timeframes reflects various phenomena, not a single entity.