Limited reproducibility of hourly blood pressure values obtained by ambulatory blood pressure monitoring: implications for studies on antihypertensive drugs.

OBJECTIVE: To assess the reproducibility of average hourly blood pressure values obtained by 24-h non-invasive ambulatory monitoring. PATIENTS: Fifteen outpatients with essential hypertension. In all subjects antihypertensive treatment was withdrawn for 4 weeks before and during the 4 weeks of the study. METHODS: The 24-h blood pressure was monitored by a SpaceLabs 5300 device (four readings per hour during the day and three readings per hour during the night) twice, at a 4-week interval. Systolic (SBP) and diastolic blood pressure (DBP) were averaged for each hour and for the whole 24-h period, and hourly and 24-h reproducibility was quantified by the standard deviation of the mean difference (SDD) between the values obtained in the two recordings. RESULTS: The SDD of hourly SBP and DBP was much greater than that of the 24-h values and ranged widely between the hours of recording. The SDD of hourly SBP and DBP were also variably greater than the SDD of the 24-h value in another 14 untreated essential hypertensives in whom 24-h ambulatory blood pressure was monitored intra-arterially twice at a 4-week interval to calculate hourly average blood pressure on thousands rather than on three or four values per hour. CONCLUSION: Reproducibility is less for hourly than for 24-h average blood pressure. This feature (which probably depends on behavioural differences between two recordings) suggests that ambulatory blood pressure measurement partly loses its advantages for reproducibility and reduction in trial size if the results are analysed over hourly periods.