Circadian period of human blood pressure and heart rate in clinical health under ordinary conditions

Putative point-and-interval estimates of the circadian period τ of cardiovascular variables were obtained from 95 clinically healthy nurses who measured their blood pressure and heart rate about every 15 minutes for 48 hours in November or June, using a Colin Medical Instruments ABPM-630 ambulatory monitor. Linear-nonlinear least-squares rhythmometry yields estimates of the circadian period, with a single-component (24-h cosine curve) and a two-component (24- and 12-h cosine curves) model. For systolic blood pressure, in an individualized fashion, analysis based on a single 24-h component allows the determination of the circadian period with 95% confidence limits in all but three of the 95 series. On a population average, 95% confidence intervals bracket anticipated periods of 24 and 12 h; but on an individualized basis, the 95% confidence interval of τ does not cover precisely 24 h in almost 50% of the cases. A reference standard is provided for the assessment by linear-nonlinear rhythmometry of circadian cardiovascular periods under usual conditions of life with a 48-h record