Testing period for single cosinor: extent of human 24-h cardiovascular 'synchronization' on ordinary routine.

A clinically healthy woman monitored her systolic and diastolic blood pressure and pulse for 26 days at approximately 10-min intervals, with interruptions. Data over subspans of varying length were analyzed by single cosinor. A method for testing a period assumed to characterize data analyzed by single cosinor was introduced and programmed to compute a 95% confidence interval for the circadian period. Statistically significant deviations from precisely 24 h were found. The likelihood of their occurrence was found to vary as a function of the length of the interval analyzed. Under ordinary conditions in health, the circadian period may vary around 24 h, indicating that strict frequency-synchronization with the environment need not occur on a short-term basis. Deviations from an average circadian period may be brought about in part by infradian components modulating the circadian rhythm and by day-to-day changes in waveform.