Effects of diaphragmatic breathing on ambulatory blood pressure and heart rate

Psychological procedures, such as self-hypnosis in the form of autogenic training, have been proposed for correcting a deviant, e.g., high blood pressure (BP). In view of the overwhelming evidence for the circadian (CD) stage dependence of any treatment effects, the CD stage dependence of the effects of diaphragmatic breathing (DB) on BP and heart rate (HR) was explored in data from a clinically healthy normotensive subject who, following 3 weeks of ambulatory monitoring as a reference standard., measured BP and HR with a manual monitor at 1-min intervals for 5 min before and after DB (three deep diaphragmatic breaths) and who performed DB for about 2 weeks at about 2-h intervals while awake. The 3-week data series were analyzed by cosinor, involving the least-squares fit of cosine curves with periods of 24, 12 and 168 h. A CD rhythm was detected for BP and HR (P < 0.001 in each case), peaking in the afternoon. Some about-weekly (circaseptan; CS) BP rhythms and 12-h (circasemidian) components were also statistically significant. DB was found to reduce systolic (S) BP. Overall, SBP decreased by 5.9 +/- 0.8 mmHg (P < 0.001) and diastolic (D) BP by 1.4 +/- 0.8 mmHg (P < 0.005), while HR remained at about the same average. The effect of DB on BP was CD-dependent, the largest response occurring in the afternoon, 2-3 h before the peaks in SBP and DBP found in the reference data of the same subject. There was also a 5-10% decrease in SBP around the weekend (Friday, Saturday and Sunday). The results on a single subject suggest the need to collect similar data on others for optimizing by clock-hour, day of the week, and eventually by the marker rhythms BP and/or HR the best times for DB and other procedures. The personalized best time for people on different work/rest schedules for relaxation may be several hours before their BP has reached its highest point in the 24-h span. HR may serve as a marker for DB timing, but the effect on HR of DB was only of borderline statistical significance in the subject investigated. (c) 2003 Editions scientifiques et medicales Elsevier SAS. All rights reserved.