Number of measurements required for the analysis of diurnal blood pressure profile

The aim of this study was to investigate how frequent blood pressure (BP) readings need to be obtained to reproduce the diurnal BP profile without loss or distortion of information. The subjects were 97 normotensives aged 23-84 years. Noninvasive ambulatory BP readings were programmed with an interval of 7.5 minutes during the day (from 8 am to 8 pm) and at 15 minutes intervals at night. Readings were stepwise omitted from the original recordings. For each step the diurnal BP profile was modelled with five different techniques. The concordance between original and reduced recordings was quantified using the repeatability coefficient, i.e. twice the standard deviation of the differences between these recordings (expressed as a percentage of the 5th to 95th percentile range of the parameter under investigation). The concordance between original and reduced recordings tended to be better for the level of pressure than for the parameters of the diurnal profile. If the sampling frequency was two readings per hour, concordance for SBP was < 10% for the BP level, 19% for the 24h standard deviation, 12% for the nocturnal fall in BP, 23% for the amplitude of the Fourier curve and 17% for the cusum derived circadian alteration magnitude. Concordance worsened to > 25% for most parameters of the diurnal BP curve when the interval between consecutive measurements exceeded 30 minutes.(ABSTRACT TRUNCATED AT 250 WORDS)