Original Article

Journal of Human Hypertension (2010) 24, 807–813; doi:10.1038/jhh.2010.18; published online 4 March 2010

Regular slow-breathing exercise effects on blood pressure and breathing patterns at rest

D E Anderson1, J D McNeely1 and B G Windham1

1Clinical Research Branch, Intramural Research Program, National Institute on Aging, National Institutes of Health, Baltimore, MD, USA

Correspondence: Dr DE Anderson, Osher Center for Integrative Medicine, Campus Box 1726, University of California at San Francisco, San Francisco, CA 94143-1726, USA. E-mail: Andersod@mail.nih.gov

Received 11 June 2009; Revised 11 September 2009; Accepted 5 November 2009; Published online 4 March 2010.

Top

Abstract

Previous studies reported that a device-guided slow-breathing (DGB) exercise decreases resting blood pressure (BP) in hypertensive patients. This study investigated the effects of daily practice of DGB on (a) 24-h BP and breathing patterns in the natural environment, as well as (b) BP and breathing pattern during clinic rest. Altogether, 40 participants with pre-hypertension or stage 1 hypertension were trained to decrease breathing rate through DGB or to passively attend to breathing (control, CTL) during daily 15-min sessions. The participants practiced their breathing exercise at home for 4 weeks. The DGB (but not the CTL) intervention decreased clinic resting BP, mid-day ambulatory systolic BP (in women only) and resting breathing rate, and increased resting tidal volume. However, 24-h BP level was not changed by DGB or CTL interventions, nor was overnight breathing pattern. These findings are consistent with the conclusion that a short-term, autonomic mechanism mediated the observed changes in resting BP, but provided no evidence that regular DGB affected factors involved in long-term BP regulation. Additional research will be needed to determine whether 24-h BP can be lowered by a more prolonged intervention.

Keywords:

BP; breathing; tidal volume

Top

Introduction

Several previous studies have reported that regular practice of device-guided breathing (DGB) decreases resting blood pressure (BP) of hypertensive patients.1, 2, 3, 4, 5, 6, 7, 8 However, only three of the studies reported to date involved randomized clinical trials. Two of those did not measure BP in the natural environment,3, 8 and the third was unable to confirm greater effects than in a control group.7 Whether the effects of DGB observed in the clinic setting extend to the natural environment is an important issue that needs to be investigated in a randomized clinical trial with 24-h BP monitoring preceding and following the DGB intervention.

A study with 24-h BP monitoring would also be relevant to the mechanism by which the enduring BP changes are generated. Performance of DGB itself is accompanied by increased cardiopulmonary stretch receptor stimulation that reduces sympathetic efferent fibre discharge, resulting in peripheral vasodilation.9 Thus, acute decreases in BP during DGB are mediated, at least in part, by decreases in sympathetic and increases in parasympathetic nervous system activity.10 Long-term changes in BP are mediated, however, not by changes in autonomic nervous system activity, but by factors that change the set point for BP.11

Renal sodium regulation is sensitive to changes in breathing pattern that alter blood gas concentrations. Previous studies of DGB have found not only that tidal volume increases as breathing rate falls, but also that end tidal CO2 (PetCO2) decreases, apparently because of improved gas transfer associated with deeper breathing.12 To date, however, no studies have examined possible changes in resting breathing rate, tidal volume, minute ventilation or PetCO2 in response to regular practice of DGB. If PetCO2 chronically decreased in response to regular DGB, long-term effects on BP could involve alterations in blood gases and acid–base balance that alter total body sodium levels.

The effectiveness of DGB depends on the frequency of practice. For example, Elliott et al.6 found that a 180-min practice of DGB over an 8-week interval was a threshold for the occurrence of significant decreases in systolic BP. Although previous interventions were each of 8 weeks duration, several found that BP decreased during the first 3–4 weeks before levelling off.3, 4, 5

This study was designed to determine whether a daily, 4-week DGB intervention decreased resting and 24-h BP in a series of subjects with pre-hypertension or stage 1 hypertension. This study also investigated whether the effects of DGB on BP are accompanied by concurrent changes in breathing patterns at rest. The effects of DGB on BP and breathing patterns were compared with those observed in a randomized control group who attended to the breathing rhythm without attempting to control its rate.

Top

Materials and methods

Participants

In total, 102 men and women from the surrounding community responded to local advertising for a clinical trial (Clinicaltrial.gov ID # NCT00328016), and were screened over the telephone. In all, 72 respondents were invited to the National Institute on Aging Clinical Research Unit, where the purpose of the study was explained, and informed consent was obtained. A physical examination was performed, and blood and urine samples were collected to ensure that the participants were free of respiratory, cardiovascular and renal diseases. Additional exclusion criteria were diabetes or the use of tobacco, steroids, hormone-replacement therapy, angiotensin II receptor blockers, angiotensin-converting enzyme inhibitors, β-blockers or any other medications that would interfere with central nervous system activity. The protocol was approved by the Institutional Review Board of the Medstar Research Institute.

Subjects with pre-hypertension or mild hypertension were studied to eliminate the complicating effects of antihypertensive medication. A total of 30 subjects were eliminated who reported hypertensive BP over the telephone, but were found to have BP below criterion levels in the clinic. Two subjects who completed the study were removed from the data analysis because of noncompliance with intervention instructions (one in each group). The remaining 40 participants were assigned to either DGB or control (CTL) group using an open randomization procedure, and introduced to their respective breathing exercise. Table 1 shows the effectiveness of this randomization procedure in terms of the nonsignificance of differences between groups in the various pre-study measures.


Experimental design, randomization and interventions

This design had three basic phases: a period of pre-intervention monitoring, a 4-week intervention period for each of the two groups (DGB and CTL), and a period of post-intervention monitoring. Two screening sessions were scheduled to determine eligibility and establish pre-intervention baseline data, after which qualifying participants were randomized to their intervention condition. Each group practiced their respective breathing exercise daily throughout the 4-week intervention phase, after which they returned to the clinic for post-intervention monitoring.

During the pre-intervention monitoring period, each participant visited the clinic on two occasions within a 1-week interval during which breathing pattern and PetCO2 were monitored continuously, and BP was measured every 6min for 25min, as described below. Candidates were declared eligible for participation if mean systolic BP of the 10 measurements during the two sessions was >130 and <160mmHg and mean diastolic BP <100mmHg. Between the two pre-intervention sessions, BP was recorded for 24h in the natural environment, and breathing pattern was recorded continuously during overnight sleep.

Eligible participants were randomized to the DGB or CTL condition. The DGB condition involved the use of a commercially available device (RESPeRATE, Lod, Israel) that guides breathing by auditory stimulation.12 The device includes a microcomputer that is connected to a band worn around the torso and a set of earphones. The band senses individual breathing rhythm, and calculates an initial rate to which the subject entrains breathing cycle in accord with a series of ascending and descending tones presented over the earphones. Over time, the duration of each tone increases, and breathing rate is systematically slowed, usually <10 breaths per min, and often as low as less than or equal to6 breaths per min. Each participant was encouraged to breathe as comfortably and effortlessly as possible, while keeping the lungs moving in accord with the tones. The device contains software that enabled subsequent assessment of adherence and quantification of performance.

The control (CTL) group performed a meditative relaxation exercise.13 Participants were instructed to sit comfortably with eyes closed and legs and arms uncrossed, and to observe their natural breathing rhythm for 15min, without consciously controlling it. They were instructed to silently repeat the word ‘one’ during each exhalation interval. Daily performance at home was documented in a diary. All participants were called at home weekly to record adherence data.

Participants were instructed to practice their breathing exercise daily throughout the 4-week intervention phase. Both groups reported excellent adherence to the task instructions. Data recorded in the device showed that DGB was practiced for 27.9±0.3 of the 28 days, during which mean breathing rate was <10 per min for 11.5±1.0 of the 15min. The DGB group synchronized their breathing with the tones 81.9±15.3% of the time. The CTL group reported practicing on 27.9±0.3 of the 28 days for an average of 17.52±0.8min per day.

During the post-intervention period, each participant returned to the clinic for a single session of 25min of monitoring of breathing pattern, PetCO2, and BP, as during the pre-intervention period. In addition, BP was monitored for 24h, and overnight breathing pattern was recorded.

BP monitoring

BP was recorded in the clinic and in the natural environment using an inflatable arm cuff attached to an oscillometric device (Spacelabs, Redmond, WA, USA). During screening sessions, BP was recorded every 6min for a total of five measurements per session. During 24-h monitoring in the natural environment, BP was recorded every 30min for 16-h (daytime), and every 60min for 8-h (nighttime). The BP recorder was regularly calibrated in the laboratory using sphygmomanometer.

Breathing pattern and end tidal CO2 (PetCO2)

Breathing rate, tidal volume and minute ventilation were recorded from an elasticized vest that summed chest and abdominal expansion on a breath-to-breath basis through inductive plethysmography (Lifeshirt, Vivometrics, Ventura, CA, USA).14 The vest was used for both clinic sessions and for overnight recording at home. Tidal volume was calibrated before each monitoring session by exhaling a fixed volume of air into an inflatable bag. Data from the recorder were downloaded onto a desktop computer (Dell Computer, Round Rock, TX, USA).

PetCO2 was monitored continuously during clinic sessions using a nasal cannula connected to a respiratory gas monitor (Datex-Ohmeda, Fairfield, CT, USA). Data were recorded over successive 10-s intervals on a desktop computer (Dell Computer).

Data analysis

Independent two-tailed t-tests were used to determine whether there were any differences in biometric and physiological measurements between the two groups at baseline. The significance of the mean differences between screening sessions, the pre-intervention and post-intervention sessions was analysed for each clinic resting cardiovascular and respiratory measure through repeated measures two-way analysis of variance using Bonferroni multiple comparison tests. The significance of the differences between the pre-intervention and post-intervention measures of 24-h BP, night time BP (midnight–0800h), ambulatory daytime BP (0800–1600h), and evening breathing measures (1600h–midnight) for all participants, for men and women, were also determined by repeated measures two-way analysis of variance using Bonferroni multiple comparison tests.

Top

Results

Intervention effects on clinic resting BP

Figure 1 shows means and s.d. of resting clinic systolic BP during the two screening sessions (weeks 1 and 2) and the post-intervention session (week 6). Significant effects for time (F2,38=6.37; P<0.002) and a significant interaction between group and time were observed, indicating a significantly greater decrease in systolic BP following the intervention for DGB than in the CTL group (F1,38=3.72; P<0.029). The systolic BP of the DGB group was significantly lower after the intervention compared with both pre-intervention levels, which were not significantly different from each other.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Means and s.e. of systolic and diastolic BP for DGB and control (CTL) groups during two screening, pre-intervention sessions (weeks 1 and 2) and after 4 weeks of the intervention (week 6).

Full figure and legend (76K)

Figure 1 also shows means and s.e. of resting clinic diastolic BP during the two screening sessions and the post-intervention session. A significant effect for time (F2,38=9.85; P<0.002), but no interaction between time and group (F1,38=0.22; P<0.803) was observed. Multiple comparison tests showed that resting diastolic BP of the DGB (but not CTL) group was significantly lower after the intervention than during both pre-interventions session.

Intervention effects on clinic respiratory measures

Figure 2 shows means and s.e. of resting clinic breathing rate during the two screening sessions (weeks 1 and 2) and the post-intervention session (week 6). A significant interaction between time and group was observed (F1,38=16.24; P<0.001), indicating that breathing rate of the DGB group decreased more than in the CTL group. Breathing rate of the DGB group was significantly lower after the intervention than during both pre-intervention sessions.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Means and s.e. of breathing rate, tidal volume, minute ventilation and end tidal CO2 for DGB and control (CTL) groups during two screening, pre-intervention sessions (weeks 1 and 2) and after 4 weeks of the post-intervention session (week 6).

Full figure and legend (93K)

Figure 2 also shows means and s.e. of resting clinic tidal volume during the two screening sessions and the post-intervention session. Significant effects of time (F2,38=6.73; P<0.002), and a significant interaction between time and group (F1,38=7.83; P<0.001) were observed, indicating that tidal volume of the DGB group increased more than in the CTL group. Tidal volume of the DGB group was significantly larger after the intervention than during both pre-intervention sessions.

Figure 2 also shows means and s.e. of resting clinic minute ventilation during the two screening sessions and the post-intervention session. No significant effect of time (F2,38=3.07; P<0.052), or interaction between time and group (F1,38=0.02; P<0.976) were observed, indicating no difference between DGB and CTL groups in change in minute ventilation.

Figure 2 also shows means and s.e. of resting clinic PetCO2 during the two screening sessions and the post-intervention session. A significant effect of time (F2,38=5.31; P<0.007), but no significant interaction between time and group (F1,38=1.97; P<0.146) was observed, indicating no differences between groups in change in PetCO2. Multiple comparison tests showed that PetCO2 of the DGB (but not CTL) group was significantly lower following the intervention than during the first pre-intervention session.

Intervention effects on 24h, overnight, daytime and evening BP

Table 2 shows means and standard errors of 24h, daytime and night time systolic and diastolic BP of the DGB and CTL groups before and after the intervention. No significant differences between groups, over time, or in the interaction of group and time were observed for 24-h systolic (F1,38=0.12; P<0.73) or diastolic (F1,38=0.55; P<0.46) BP.


Figure 3 shows the diurnal variation in hourly mean systolic BP preceding and following the DGB and CTL interventions, plotted separately for women and for men. Figure 3 and Table 3 show that for women in the DGB group, the post-intervention daytime systolic (t=6.11; P<0.001) and diastolic (t=3.47; P<0.05) BP mean were significantly lower than the pre-intervention daytime levels. Table 3 also shows that no significant pre- to post-intervention differences were observed for overnight or evening systolic BP for women, or overnight, daytime or evening systolic BP for men. Similarly, no significant pre- to post-intervention differences in systolic BP for women or men in the CTL group were observed in night, day, or evening segments of the 24-h BP cycle (Table 3).

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Mean hourly systolic BP over the 24-h day (beginning at midnight) preceding and following the DGB and CTL interventions for men and women.

Full figure and legend (146K)


Intervention effects on overnight breathing patterns

Table 2 also shows means and s.e. of overnight breathing rate, tidal volume, and minute ventilation for DGB and CTL groups before and after the intervention. No significant differences between groups, over time, or in the interaction between group and time were found for overnight breathing rate (F1,38=0.48; P<0.49), tidal volume (F1,38=0.01; P<0.91) or minute ventilation (F1,38=0.21; P<0.65).

Top

Discussion

The DGB (but not the CTL) intervention decreased clinic resting BP, mid-day ambulatory systolic BP (in women only) and resting breathing rate, and increased resting tidal volume. However, 24-h BP level was not changed by the DGB or CTL interventions, nor was overnight breathing pattern.

The findings on resting BP are consistent with those in previous studies of longer duration,1, 2, 3, 4, 5, 6, 7, 8 and support the view that the antihypertensive effects of DGB on BP operate over a pathway in which a prolonged expiratory phase of the respiratory cycle exert reflex effects that decrease peripheral vasoconstriction and increase peripheral blood flow.9 As the magnitude of BP effects of regular DGB has been shown previously to depend on pre-intervention BP,4 the magnitude of the effects reported here may also reflect the borderline hypertensive condition of the sample. The findings are also consistent with the effects in a previous study that daytime BP in the natural environment is decreased by regular practice of DGB.2 In this study, this finding was confined to female subjects, but as discussed below, is consistent with the results of other research, which suggests that the associations of breathing pattern with BP may be more marked in women than in men. That the effects of DGB on daytime BP in women were related to decreases in autonomic arousal is also consistent with the results of a previous study with women showing that acute increases in BP in the workplace were accompanied by increases in perceived stress and urinary catecholamines.15

The findings in this study do not, however, support the view that regular practice of DGB has effects on mediators of long-term BP level. According to Guyton's formulation of hypertension pathogenesis, neither changes in cardiac output or peripheral resistance that accompany changes in sympathetic nervous system activity can result in long-term changes in BP because of the ‘infinite’ ability of the kidneys to adjust blood volume up or down to maintain BP around its set point.16 The development of chronic hypertension necessarily involves changes in the set point around which BP fluctuates, the locus of which has variously been ascribed to the kidneys11 or central nervous system.17

Previous research found that sustained salt-sensitive hypertension could be generated in laboratory animals by intermittent suppression of breathing that increased pCO2 and generated a cascade of responses involving transient changes in acid–base balance and renal regulation of sodium.18 From those findings, it was speculated that regular practice of DGB might alter renal regulation of sodium and long-term BP through a respiratory mechanism that includes sustained decreases in pCO2.19 In this study, resting PetCO2 was measured before and after the DGB and CTL interventions as an index of possible changes in pCO2. In fact, a significant decrease in resting PetCO2 was observed from the first screening session to the post-intervention session in the DGB group. Positive associations between resting PetCO2 and resting BP have been observed in previous studies with women. Higher resting PetCO2 has been associated with higher resting systolic BP in women over 50 years of age,20 and especially if they scored low on trait anger.21 It has been hypothesized that high resting PetCO2 might have a permissive role in the development of sodium-sensitive form of hypertension.22

Although 24-h BP did not decrease in this study, it would be premature to conclude that regular DGB could not decrease 24-h BP under other conditions. For example, the present intervention was 4 weeks in duration, and at least two of the previous studies of DGB continued to show a progression of decreases in BP after the first 4 weeks.3, 5 The adherence in this study was excellent, but 15min per day is only a small fraction of the waking hours during which deeper, slower breathing might exert its effects. It is conceivable that an intervention that encouraged deeper breathing over a greater percentage of the day could affect the set point for BP.

In summary, the finding of no changes in 24-h BP in response to repeated practice of DGB in the context of significant changes in resting BP is potentially important because of its implications for the pathways mediating short- and long-term BP regulation. It remains to further research to clarify the extent to which retraining in breathing patterns is capable of producing salutary effects on mechanisms involved in long-term BP regulation.

Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Top

Conflict of interest

The authors declare no conflict of interest.

Top

References

  1. Grossman E, Grossman A, Schein MH, Zimlichman R, Gavish B. Breathing control lowers blood pressure. J Hum Hypertens 2001; 15: 263–269. | Article | PubMed | ChemPort |
  2. Rosenthal T, Alter A, Peleg E, Gavish B. Device-guided breathing exercises reduce blood pressure: ambulatory and home measurements. Am J Hypertens 2001; 14: 74–76. | Article | PubMed | ChemPort |
  3. Schein MH, Gavish B, Herz M, Rosner-Kahan D, Naveh P, Knishkowy B et al. Treating hypertension with a device that slows and regularizes breathing: a randomized, double-blind controlled study. J Human Hypertens 2001; 15(4): 271–278. | Article | ChemPort |
  4. Viskoper R, Shapira I, Priluck R, Mindlin R, Chornia L, Laszt A et al. Nonpharmacologic treatment of resistant hypertensives by device-guided slow breathing exercises. Amer J Hypert 2003; 16: 484–487. | Article
  5. Meles E, Giannattasio C, Failla M, Gentile G, Capra A, Mancia G. Nonpharmacologic treatment of hypertension by respiratory exercise in the home setting. Am J Hypertens 2004; 17: 370–374. | Article | PubMed
  6. Elliott WJ, Izzo JL, White WB, Rosing DR, Snyder CS, Alter A et al. Graded blood pressure reduction in hypertensive outpatients associated with use of a device to assist with slow breathing. J Clin Hypertens 2004; 6: 553–559. | Article
  7. Logtenberg SJ, Kleefstra N, Houweling ST, Groenier KH, Bilo HJ. Effect of device-guided breathing exercises on blood pressure in hypertensive patients with type 2 diabetes mellitus: a randomized clinical trial. J Hypertens 2007; 25: 241–246. | Article | PubMed | ChemPort |
  8. Schein MH, Gavish B, Baevsky T, Kaufman M, Levine S, Nessing A et al. Treating hypertension in type II diabetic patients with device-guided breathing: a randomized controlled trial. J Hum Hypertens 2008; 23: 325–331. | Article | PubMed
  9. Parati G, Izzo JL, Gavish B. Respiration and blood pressure. In: Izzo JL, Sica DA, Black HR (eds). Hypertension Primer. American Heart Association: Dallas, TX, 2008, pp 136–138.
  10. Joseph CN, Porta C, Casucci G, Casiraghi N, Maffeis M, Rossi M et al. Slow breathing improves barorefex sensitivity and decreases blood pressure in essential hypertension. Hypertens 2005; 46(4): 714–718. | Article | ChemPort |
  11. Cowley AW. Long-term control of arterial blood pressure. Physiol Rev 1992; 72: 231–300. | PubMed | ISI |
  12. Anderson DE, McNeely JD, Windham BG. Device-guided slow breathing effects on end tidal CO2 and heart rate variability. Psychol Health Med 2009; 14: 667–679. | Article | PubMed | ChemPort |
  13. Benson H, Klipper MZ. The Relaxation Response. Wings Books: New York, NY, 1992.
  14. Wilhelm FH, Roth WT, Sackner MA. The LifeShirt. An advanced system for ambulatory measurement of respiratory and cardiac function. Behav Modif 2003; 27: 671–691. | Article | PubMed
  15. Kario K, James GD, Marion RM, Ahmed M, Pickering TG. The influence of work- and home-related stress on the levels and diurnal variation of ambulatory blood pressure and neurohumoral factors in employed women. Hypertens Res 2002; 25: 499–506. | Article | PubMed | ChemPort |
  16. Guyton AC. Dominant role of the kidneys and the accessory role of whole-body autoregulation in the pathogenesis of hypertension. Amer J Hypertens 1989; 2: 575–585. | ChemPort |
  17. Julius S, Weder AB. Brain and the regulation of blood pressure: a hemodynamic perspective. Clin Exper Hypertens 1989; 11(Suppl): 1–19. | Article
  18. Anderson DE. Behavior analysis and the search for the origins of hypertension. J Exp Anal Behav 1994; 61: 255–261. | Article | PubMed | ChemPort |
  19. Honig A. Peripheral arterial chemoreceptors and reflex control of sodium and water homeostasis. Am J Physiol Integ Comp Physiol 1989; 257(26): R1282–R1302. | ChemPort |
  20. Anderson DE, Parsons D. Association of end tidal CO2 with longitudinal blood pressure elevation. Am J Hypertens 1999; 17: 1073–1079. | Article | ChemPort |
  21. Scuteri A, Parsons DJ, Chesney MA, Anderson DE. Anger Inhibition potentiates the association of high end tidal CO2 with blood pressure of women. Psychosom Med 2001; 63: 470–475. | PubMed | ChemPort |
  22. Anderson DE, Scuteri A, Agalakova N, Parsons DJ, Bagrov AY. Racial differences in resting end tidal CO2 and sodium pump inhibitor. Amer J Hypert 2001; 14(8 part 1): 761–767. | Article | ChemPort |
Top

Acknowledgements

This research was solely supported by the Intramural Research Program of the NIH, National Institute on Aging. The authors are indebted to Dr Deborah Grady for calling our attention to the guided breathing device, and to Drs Benjamin Gavish, S Mortesa Farasat and Margaret A Chesney for review of the paper.

Extra navigation

.

natureevents

ADVERTISEMENT