ORIGINAL ARTICLES T H E EFFECTS OF THORACIC MANIPULATION O N HEART RATE VARIABILITY: A CONTROLLED CROSSOVER TRIAL Brian Budgell, DC, PhD,^ and Barbara Polus, PhD''
ABSTRACT
Objective: The objective of this study was to measure the effects of thoracic spinal manipulation on heart rate variability (HRV) in a cohort of healthy young adults. Methods: A controlled crossover trial that was conducted on 28 healthy young adults (23 men and 5 women; age range, 18-45 years; mean age, 29 ± 7 years) measured HRV before and after a sham procedure and a thoracic spinal manipulation. Results: In healthy young adults, thoracic spinal manipulation was associated with changes in HRV that were not duplicated by the sham procedure. The ratio of the powers of the low-frequency and high-frequency components increased from 0.9562 ± 0.9192 to 1.304 ± 1.118 ( P = .0030, Wilcoxon signed rank test). In subjects undergoing sham spinal manipulation, there was no statistically significant change in the low-frequency or the high-frequency component of the power spectrum; neither was there any in the ratio of the two regardless of whether the comparison was made using the paired / test or the Wilcoxon signed rank test. Conclusion: High-velocity and low-amplitude manipulation of the thoracic spine appears to be able to influence autonomic output to the heart in ways that are not duplicated by a sham procedure or by other forms of somatic/physical therapies. (J Manipulative Physiol Ther 2006;29:603-610)
Key Indexing Terms: Spinal manipulation; Autonomic nervous system; Heart rate
V
arious forms of innocuous somatic stimulation have been shown to modulate such aspects of cardiovascularfianctionas heart rate (HR), blood pressure, and regional blood flow. In some instances, these effects are accompanied by, and perhaps attributable to, alterations in autonomic output to the cardiovascular system. For example, it has previously been shown that cervical manipulation results in changes in HR and HR variability (HRV) that are not achieved with a sham manipulation.' At present, it is uncertain whether autonomic responses to spinal manipu" Associate Professor, School of Health Sciences, Faculty of Medicine, Kyoto University, Kyoto, Japan. '' Associate Professor, Division of Chiropractic, School of Health Sciences, Royal Melbourne Institute of Technology University, Bundoora West Campus, Bundoora, Victoria, Australia. Submit requests for reprints to: Brian Budgell, DC, PhD, School of Health Sciences, Faculty of Medicine, Kyoto University, Kawahara-cho 53, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan (e-mails: Budgell@hs.med.kyoto-u.ac.jp, Barbara.polus@rmit.edu.au). Paper submitted May 3, 2006; in revised form June 24, 2006; accepted July 2, 2006. 0161-4754/$32.00 Copyright © 2006 by National University of Health Sciences. doi:I0.1016/j.jmpt.2006.08.011
lation are limited to the cervical spine or whether similar results can be achieved by manipulation of other regions of the spine. Thus, the present study was undertaken to measure the effects of thoracic spinal tnanipulation on HR and HRV in a cohort of healthy young adults.
METHODS
Twenty-eight healthy adults (23 men and 5 women) participated in a controlled crossover trial of the effects of upper thoracic spinal manipulation and sham spinal manipulatioti on HR and HRV. For each subject, the two treatment procedures were performed 1 week apart at approximately the same time of day. A minimum cohort size of 25 subjects was chosen with consideration that statistically significant alterations in cardiovascular outcome measures had been shown iti a pilot study with 17 data sets^ and in a study on the effects of cervical spinal manipulation with 24 complete data sets.' Thirty-one subjects were initially recruited. One subject had frequent premature ventricular contractions, another subject failed to attend the second treatment session, and yet another subject displayed a systolic pressure higher than 140 mm Hg at both treatment sessions; therefore, the cohort had 28 complete data sets for analysis. 603
604
Budgell and Polus Heart Rate Variability
Subjects were recruited on the basis that they did not have current neck and upper back pain. Before each intervention, the subjects were asked to assess their level of cervicothoracic spinal discomfort using a visual analogue scale (VAS) at the full extent of active left and right cervical rotations. On the VAS, 0 represented complete comfort and 10 represented "the worst pain imaginable." On this basis, the levels of pain (mean Âą SD) on full active left and right rotations before the sham and authentic stimulations were 0.7 + 1.5, 0.3 Âą 0.8, 0.2 + 0.8, and 0.6 Âą 1.4, respectively. In other words, on the day of the trials, subjects had, at most, trivial levels of cervicothoracic discomfort at the extremes of active cervical rotation. Subjects were excluded if they had (1) a history of cervicothoracic surgery, fracture, or dislocation; (2) a known anatomical abnormality in the cervicothoracic region; (3) a history of cervicothoracic trauma within the previous 3 months or persistent symptoms from an earlier trauma; (4) a history of cancer; (5) a history of stroke; (6) a history of positional vertigo; or (7) a history of chronic or recurrent inflammatory disease. In addition, subjects were excluded if they were currently receiving an anticoagulant or a steroid therapy or if they were currently engaged in litigation for spinal injury. The study was approved by the human research ethics committee of the Royal Melbourne Institute of Technology. Subjects were examined for the presence of carotid bruits as well as positional vertigo and nystagmus (with the neck held in extension and rotation) as contraindications to cervical manipulation and involvement in this study because attempted upper thoracic manipulation may result in inadvertent manipulation of the cervical spine. The validity of the screening tests used has been questioned,^ and the fact that no contraindication to manipulation was found was not taken to indicate zero risk. Similarly, immediately before each trial, blood pressure was measured with a sphygmomanometer, and subjects with a systolic pressure of 140 mm Hg or higher or a diastolic pressure of 90 mm Hg or higher were excluded from the trial. Subjects were encouraged to report any discomfort immediately. Exit questionnaires were provided but subjects did not report unpleasant effects from treatment. The mechanical stimuli were applied to the upper thoracic spine (first to fourth vertebral levels). The order of presentation of the thoracic and sham manipulations was determined by a coin toss immediately before the first trial for each subject. The sham and thoracic manipulations were performed with the subjects in the prone position. With regard to the selection of the site for manipulation in the thoracic spine within the studies described, conventional clinical indicators from the literature were used. Different authors may have used slightly different terminologies, but essentially they all described much the same set of palpable indicators of tissue quality and motion. By way of example, Cleveland'* suggested certain defining characteristics of the
Journal of Manipulative and Physiological Therapeutics October 2006
manipulable lesion: vertebral malposition, abnormal vertebral motion, abnormal joint play or end feel, soft-tissue abnormalities, and muscle contraction or imbalance. McPortland et al^ used a similar set of criteria: tenderness, asymmetry of joint position, restriction of range of motion, and tissue texture abnormality. Tenderness was ranked by each subject in response to palpation over the spinal joints. Restriction of range of motion was defined as abnormality of resistance and abnormal end feel. Tissue texture abnormality was defined as a sense of flillness over the joint space. Faye and Wiles^ described the use of static palpation in the detection of abnormal tissue texture, temperature, contour, and tenderness while advocating the use of motion palpation to determine normal active range of motion, hypermobile or aberrant motion, and capsular end feel. End feel has been used clinically to identify the location of segmental restriction of cervical motion.^ In addition, Vemon and Gitelman^ attempted to validate the use of algometry and tissue compliance in the identification of the manipulable lesion, albeit with the use of mechanical devices. The validity of static and motion palpation procedures in identifying the manipulable lesion has not been well investigated and certainly has been questioned.^ Nonetheless, motion palpation has been advocated as the assessment of choice in identifying the site of the manipulable lesion and in selecting the manipulative procedure best suited to correct the lesion.^ Recent work by Humphreys et al'" suggests that cervical motion palpation is a valid and reliable method of assessing hypomobility. Motion palpation alone or in combination with static palpation has been used for this purpose in several clinical trials involving cervical spinal manipulation."''^ In this study, sites for the application of thoracic manipulation were chosen on the basis of relatively restricted joint play and relative paraspinal hypertonicity, which have been found to have good intraexaminer and interexaminer reliabilities in this region of the spine.'^ One of two types of manipulation was used, depending on the location of restriction of vertebral motion on the day of the trial. These manipulations are of the types commonly referred to as "cross-bilateral adjustment' and "combination adjustment." The cross-bilateral adjustment was applied as described by Gitelman and Eligg.'^ Specifically, the clinician stood on the side of the subject's upper thoracic region with the ipsilateral hand reaching across the subject to apply pressure to a contralateral thoracic transverse process. The contralateral hand was used to brace the ipsilateral transverse process of the first vertebral level below. (Thus, with the hands crossed, the clinician makes contact on either side of the subject's spine, hence the name of the technique.) Pressure was applied to separate the hands until tissue resistance was detected. Then, a high-velocity and low-amplitude thrust was applied, resulting in an audible sound. This technique has been used previously in
Journal of Manipulative and Physiological Therapeutics Volume 29, Number 8
Budgell and Polus Heart Rate Vat^ability
Table 1. Measures of normal distribution of data Kolmogorov-Smimov test
Shapiro-Wilk test
Parameter
Skewness
Kurtosis
Statistic
df
P
TS HR TS HFab TS HFn TS LFab TS LFn TS LF/HF Sham HR Sham HFab Sham HFn Sham LFab Sham LFn Sham LF/HF
-0.152 3.495 -0.302 0.545 0.422 2.013 0.119 3.866 -0.089 0.551 -0.223 1.568
-0.497 12.40 -0.744 -1.036 -0.714 4.739 -0.482 16.71 -0.202 -0.768 -0.554 2.934
0.087 0.357 0.099 0.171 0.108 0.198 O.IU 0.362 0.080 0.169 0.082 0.149
28 28 28 28 28 28 28 28 28 28 28 28
.200 .000 .200 .035 .200 .006 .200 .000 .200 .039 .200 .116
Statistic 0.976 0.478 0.971 0.906 0.962 0.789 0.967 0.492 0.988 0.931 0.980 • 0.862
df
P
28 28 28 28 28 28 28 28 28 28 28 28
.753 .000 .620 .016 .380 .000 .510 .000 .984 .065 .848 .002
For skewness and kurtosis, the z scores are shown. Values between +1.96 and -1.96 were considered to be indicative of a norrnal distribution. All values printed in boldface suggest nonnomial distribution of data. TS, Values before thoracic spinal manipulation; Sham, values before sham manipulation; HFah, absolute high-frequency component; HFn, normalized high-frequency component; LFab, absolute low-frequency component; LFn, normalized low-frequency component; LFIHF, the ratio of high-frequency component to low-frequency component.
clinical investigations of thoracic spinal The combination adjustment is a variation of the crossbilateral adjustment and is a convenient technique for use in the upper thoracic region. In the combination adjustment, the clinician stood to the side of the subject's shoulder and the subject's head was cradled in the clinician's ipsilateral hand such that the palm of the hand covered the ear, the fmgers extended over the temporal region, and the thumb extended around the base of the occiput. The clinician's contralateral hand was positioned over an ipsilateral upper thoracic transverse process. For the manipulation, pressure was applied to separate the hands until tissue resistance was detected. Then, a high-velocity and lowamplitude thrust was applied, resulting in an audible sound. The sham manipulation involved standing on the side of the subject's upper back and placing the hands over the scapulae bilaterally, with the clinician's left hand on the subject's left scapula and the right hand on the subject's right scapula. As the patient reached the end of exhalation, the clinician applied a single light brief impulse simultaneously with both hands. In no instance was this sham procedure accompanied by an audible sound from the patient's back. The entire process of making contact with the subject, applying the stimulation, and withdrawing the hands would not normally take more than 5 seconds. The electrocardiogram (ECG) recording during that period of manipulation was excluded from the data analysis because of movement artifacts that occurred on some recordings. Throughout the course of each experiment, the ECG was monitored using disposable electrodes (Blue Sensor, Medicotest, Olstykke, Denmark), with the negative electrode immediately beneath the suprasternal notch, the positive electrode at the left anterior-axillary line in the fifth intercostal space, and an
indifferent electrode on the right mid axillary line (CM5 lead). The ECG signal was fed through a PowerLab/4ST bioamplifier (ADlnstruments, Castle Hill, NSW, Australia) and stored in a personal computer (Macintosh Powerbook G4, Apple Japan, Tokyo, Japan) for offline analysis. During the course of the experiments, subjects were asked to coordinate their respiratory rate with a metronome set at 0.25 Hz. The entire ECG recording for each subject was reviewed for anomalies and to determine that all R waves exceeded the triggering threshold for automated HRV analysis. Five-minute prestimulation and poststimulation blocks of the ECG records were analyzed using the software Chart v4.1 with HRV extension vl.O.I (ADlnstruments). Pretreatment and posttreatment values were compared via the paired t test, in which data were distributed normally (Table 1), and via the Wilcoxon signed rank test, in which data were not normally distributed. A P value of .05 was used as the threshold for statistical significance.
RESULTS
Of the 28 subjects who provided complete data sets, 5 were women and 23 were men. They ranged in age from 18 to 45 years, with an average age (mean ± SD) of 29 ± 7 years. All subjects were normotensive, with mean sitting systolic and diastolic blood pressures of 122 ± 12 and 72 ± 14 mm Hg, respectively, immediately before the sham procedure and 123 ± 12 and 73 ± 14 mm Hg, respectively, immediately before the thoracic manipulation. The differences between the values before the sham and thoracic manipulations were not significant (P > .75 in both cases) when compared using a two-tailed paired t test.
605
606
Budgell and Polus Heart Rate Variability
Journal of Manipulative and Physiological Therapeutics October 2006
Premanipulation values of HR and HRV were examined Table 2. Premanipulation and postmanipulation HR and measures for skewness and kurtosis (Table 1) as indicators of of HRV (mean ± S.D.) nonnally distributed data. The z scores for skewness and Premanipulation Postmanipulation P Parameter kurtosis were calculated by dividing the calculated values of skewness and kurtosis by the standard error of the mean for Sham manipulation the respective data set, with an acceptable measure of HR (bpm) 72.71 ± 12.60 70.01 ± 11.87 <.OOO1'''* 199.4 ± 137.0 228.4 ± 243.7 .69737.9534'' LFab normality set between +1.96 and —1.96. In addition, the LFn 45.00 ± 19.23 46.43 ± 18.76 .6175° Kolmogorov-Smimov test and the Shapiro-Wilk test were HFab 480.3 ± 919.1 412.0 ± 873.9 .0703'' applied to the data sets to test for normal distribution. As HFn 50.72 ±21.63 47.67 ± 17.33 .5259° indicated in Table 1, 3 of the premanipulation data sets LF/HF 1.159 ± 0.9717 1.249 ± 1.019 .6083" departed markedly from a normal distribution. These data Thoracic spinal manipulation HR (bpm) 71.73 ± 10.93 69.27 ± 10.58 <.OOO1'''* sets were as follows: (1) TS HFab, absolute high frequency LFab 195.6 ± 146.8 275.1 ± 202.9 .0098"'* before thoracic spinal manipulation; (2) TS LF/HF, the ratio LFn 40.25 ± 18.55 46.66 ± 20.35 .0201"'* of low frequency to high frequency before thoracic spinal HFab 604.6 ±1168 414.1 ± 599.7 .2795'' manipulation; and (3) Sham HFah, absolute high frequency HFn 57.34 ± 18.45 49.86 ± 18.90 .0043'''* before sham spinal manipulation. Consequently, two-colLF/HF 0.9562 ± 0.9192 1.304 ± 1.118 .0030''-* umn statistical comparisons involving these data sets were " Obtained via two-tailed ; test. performed using the Wilcoxon signed rank test. Where data '' Obtained via Wilcoxon signed rank test for the null hypothesis were normally distributed, based on skewness and kurtosis, that there was no difference between premanipulation and postmanipulathe data were initially analyzed using a t test, with P < .05 tion values. * P < . 0 5 ; n = 28. considered to be significant. In as much as normality, based on skewness and kurtosis, was not always confirmed by the Kolmogorov-Smimov test and Shapiro-Wilk test, where or the Wilcoxon signed rank test as described; subseeither of these tests suggested a nonnormal distribution, it quently, data sets with skewness and kurtosis within normal was decided to also perform a second comparison of limits but significant P values as per the Kolmogorovprestimulation and poststimulation measures using the Smimov test or the Shapiro-Wilk test were reassessed using Wilcoxon signed rank test. the Wilcoxon signed rank test. Using a cutoff point of P = Where multiple outcome measures are used, it is not .05, in no instance did this second analysis change the unusual to apply a correction factor to P values to decision about whether responses were statistically significompensate for the increased probability of an apparently cant. As shown in Table 2, in subjects undergoing thoracic significant association occurring by chance. This is most spinal manipulation, there were significant increases in appropriate in cases in which outcome measures are absolute (LF) and normalized (LF/total power) levels of the independent,^' which would not be the case in this study. LF component of the power spectrum as well as in the LF/ Hence, no correction was applied to account for multiple HF ratio. outcome measures, but the reader should be alert as to The power of the absolute LF component increased from the possibility that this may result in a less-stringent 195.6 ± 146.8 to 275.1 + 202.9 {P = .0098, Wilcoxon statistical analysis. signed rank test). The power of the normalized LF With respect to the effectiveness of the randomization of component increased from 40.25 ± 18.55 to 46.66 ± presentation of thoracic vs sham spinal manipulation, there 20.35 (P = .0201, paired t test). In addition, the power of was no significant difference between the two groups in the the normalized HF component decreased from 57.34 ± prestimulation levels of HR, LF/HF, and normalized or 18.45 to 49.86 ± 18.90 [p = .0043, paired t test). Hence, absolute LF and HF. That is to say, in all instances using the the ratio of the powers of the LF and HF components two-tailed paired t test (HR, normalized HF, and normalized increased from 0.9562 ± 0.9192 to 1.304 ± 1.118 (P = LF) or the Wilcoxon signed rank test (absolute HF, absolute .0030, Wilcoxon signed rank test). In subjects undergoing LF, and LF/HF) to compare premanipulation levels for sham sham spinal manipulation,, there was no statistically sigand thoracic manipulations, P was greater than .05. nificant change in the LF or H F component of the power As shown in Table 2, in the trials of both sham and spectrum; neither was there any in the ratio of the two thoracic manipulations, HR declined from the first to second regardless of whether the comparison was made using the 5-minute intervals {P < .0001, two-tailed paired t test), and paired t test or the Wilcoxon signed rank test. there was no significant difference {P > .05, two-tailed One subject rated the discomfort of the sham treatment as paired t test) in the absolute degree of decline between the 3.8 cm on the VAS. All other subjects rated the sham two trials. manipulation as completely comfortable. Two subjects rated For both sham and thoracic manipulations, premanipu- the discomfort of the thoracic manipulation as 1.3 and lation and postmanipulation measures of HRV were 1.4 cm each on the VAS. All other subjects rated the compared by first using either the paired two-tailed t test maneuver as completely comfortable (0/10 on the VAS).
Journal of Manipulative and Physiological Therapeutics Volume 29, Number 8
Hence, both the sham and thoracic manipulations could be considered as generally innocuous. Before the experiments, all subjects had received a written explanation allowing them to anticipate an upper thoracic manipulation accompanied by an audible sound. In addition, as first-year students in a health sciences program, they may have had some familiarity with manipulative techniques. Before the experiments, they were not aware that they might receive a sham manipulation. On questioning after each experiment, in 25 of 34 trials (including trials for which data were not included in the fmal analysis because of the outlined reasons), subjects correctly identified when they had received thoracic manipulation. Eight subjects thought that they had not received thoracic manipulation, and 1 subject was unsure. In 3 of 31 trials, subjects incorrectly guessed that they had received thoracic manipulation on the occasion that they had received a sham manipulation. One subject was unsure, and the remaining 27 subjects correctly guessed when they had received the sham procedure. In summary, the results indicate that thoracic spinal manipulation was associated with changes in HRV that were not duplicated by a sham procedure in healthy young adults. The thoracic manipulation and sham procedures were essentially innocuous according to the volunteers. However, the volunteers were not effectively blinded as to which procedure was thoracic manipulation and which was the sham procedure.
DISCUSSION
With both the thoracic manipulation and sham procedures, there was a statistically significant (P < .05) decline in HR when comparing the average values for the 5-minute periods immediately before and after the intervention. The decline was approximately 2.45 Âą 2.5 bpm for thoracic manipulation, whereas it was approximately 2.7 Âą 2.4 bpm for the sham procedure. Comparing these declines with a two-tailed paired t test, the difference is not significant (P = .7450). This suggests that the decline in HR was more likely attributable to the general experimental conditions (eg, lying prone for 10 minutes) and that the thoracic manipulation had no specific effect distinguishable from the effects of sham manipulation on HR. Much research has investigated HRV in the supine position, and cardiovascular adaptations to movement from the supine to the upright position are sufficiently stereotypical to serve as clinical measures of autonomic function; for a review, see Reference.^^ Not surprisingly, HRV measures indicate increased cardiac sympathetic output in the standing position vs the supine position.'^^ The prone position is of less clinical interest because medical patients would generally recline in the supine or the lateral position. It has been observed that in patients with spinal anesthesia, movement from the supine to the prone position is
Budgell and Polus Heart Rate Variability
associated with an increase in HR.^'* In this regard, subjects in this study on thoracic manipulation did have higher HRs than similar subjects in a previous study on cervical manipulation.' However, there may have been a variety of differences between the groups to account for this. Patients with chronic heart failure show a preference for reclining in the right lateral recumbent position, and this position appears to attenuate cardiac sympathetic output.'^^ In healthy young adult volunteers, the left lateral recumbent position results in lower HR and blood pressure than the sitting, supine, and right lateral recumbent positions.^* Collectively, these studies suggest that apparently small changes in the recumbent position can have measurable effects on cardiovascular flinction, and these effects may be sufficient to have clinical importance. In this study, both the sham and thoracic manipulations were applied with subjects in the prone position. Whether there is an interaction between spinal manipulation and posture with regard to autonomic effects remains unexplored. The immediate challenge is to explain the different effects achieved by sham and thoracic manipulations despite the common posture of the subjects. As discussed, the thoracic manipulation was accompanied by significant increases in the absolute and normalized LF component of the power spectrum and an increase in the LF/HF ratio. This may be interpreted as indicating an increase in sympathetic output to the heart and a shift in the balance of sympathetic to parasympathetic cardiac output in favor of the sympathetic component (particularly because, on this occasion, there was also a decrease in the normalized HF component that reflects cardiac parasympathetic output). Both thoracic and sham manipulations involved compression of the upper thoracic region, including the upper thoracic spine, to approximately the physiologic limit. Both applied a rapid innocuous mechanical stimulation to the skin. Both also probably involved a very low-amplitude movement of the thoracic spine into extension with the thrust, although this movement might have been greater with the thoracic manipulation. However, the sham manipulation had no effect on measures of HRV, whereas the thoracic manipulation did. The thoracic manipulation and sham thrusts were delivered at the end of expiration, when the glottis was still open. Consequently, pulmonary mechanoreceptors would have been essentially unloaded. On the other hand, the thrust maneuvers may have had a direct mechanical effect on the heart and great vessels, transiently increasing blood pressure and activating cardiovascular mechanoreceptors. Among the distinguishing features of the thoracic manipulation is the very high-velocity and low-amplitude force applied to the spinal column. As discussed, the biomechanical features of the thoracic manipulations used in this study have been well characterized. However, the sham procedure has not been studied. Although every effort was made to mimic the velocity and force of the thoracic manipulation, no precise comparison can be offered.
607
608
Budgell and Polus Heart Rate Variability
Furthermore, it has been proposed that relatiyely small changes in the point of application or direction of a force applied to the thorax could have disproportionately large effects on autonomic responses.^' Experiments in anesthetized animals have shown that autonomic regulation of cardiovascular flinction is modulated by segmentally organized reflexes in response to noxious stimulation of somatic tissues, including the skin, appendieular skeletal muscle, and various paraspinal tissues. ' Also, in conscious animals, even innocuous somatic stimuli, including a spinal manipulation mimic^"* and massage,^^ have been shown to alter visceral function. Human physiologic experiments of somatoautonomic reflexes are rather limited. Lumbopelvic soft-tissue therapies have been shown to augment HRV measures of parasympathetic output in younger adults^^'^^ but not in older adults.''^ Of more relevance to our study on thoracic manipulation, one article reported that noxious electrical stimulation within the distribution of the thoracic spinal nerves attenuates gastric motility,''' suggesting augmentation of sympathetic output to the stomach. One study on electroacupuncture stimulation of the forearm, hence within the distribution of the lower cervical or upper thoracic spinal nerves, showed a reduction in blood pressure and an increase in HR."*" The authors interpreted these results as indicating reduced sympathetic outflow to the peripheral vasculature, with reduced cardiac vagal output and therefore increased HR. Conversely, another study on conventional acupuncture, using a different forearm insertion point, showed decreased HR."*' In this study, modulation of cardiovascular responses by atropine and propranolol suggested that the effects of acupuncture were achieved by facilitation of cardiac vagal activity and suppression of cardiac sympathetic activity. Myofascial trigger-point massage to the head, neck, and shoulder has been shown to decrease HR, systolic blood pressure, and diastolic blood pressure while augmenting HRV measures of parasympathetic tone.''^ A study on the effects of shoulder and back massage on elderly patients hospitalized for stroke revealed decreases in HR, systolic blood pressure, and diastolic blood pressure as well as decreases in levels of anxiety and pain.'*'' Collectively, these studies show that innocuous somatic stimulation is most often associated with changes in HR, HRV, and blood pressure that are strongly suggestive of augmentation of vagal output to the heart or attenuation of sympathetic output. These are quite the opposite of the effects that we observed with thoracic spinal manipulation. Thus, in seeking to explain the differences in effects between thoracic manipulation and sham manipulation, one is necessarily drawn to the consideration of the effects of that unique quality of the manipulationâ&#x20AC;&#x201D;the high-velocity and low-amplitude (innocuous) thrustâ&#x20AC;&#x201D;on spinal and paraspinal mechanoreceptors. Studies on anesthetized animals'*'*"'*^ and human beings'*^ suggest that the thrust component of the sort of spinal manipulative maneuver
Journal of Manipulative and Physiological Therapeutics October 2006
used in this experiment particularly activates muscle spindles. In particular, it has been suggested that spinal manipulation may particularly access sensory beds within deep and intersegmental paraspinal muscles whose primary physiologic function may be as sensors rather than as actuators of vertebral motion.'*^ Hence, notwithstanding numerous studies on anesthetized animals that suggested that muscle spindle stimulation has little impact on autonomic nervous system function,'*^ one must entertain the possibility that activation of muscle spindles in paraspinal muscles may modulate cardiac autonomic output in conscious human beings. This possibility has also been raised earlier in reference to the effects of cervical manipulation. The cervical muscles may represent a special case because they have an important role in head movement relative to the body and they are invested with a high density of muscle spindleâ&#x20AC;&#x201D;Golgi tendon organ complexes, which have not yet been described in other deep paraspinal muscles. Comparable histologic studies have yet to be performed with human deep thoracic paraspinal muscles. Nonetheless, it appears that, in general, deeper muscles tend to have more muscle spindles,^** and it may be argued that the very deep intervertebral muscles are particularly adapted to flinction as sensory receptors rather than as actuators.^''^^ If these muscles do have a particularly important role in signaling postural change, then, from an adaptive point of view, it is reasonable that they would also have some input into reflex regulation of cardiac function and regional blood flow. This may then explain why a short-amplitude and high-velocity thrust to the thoracic spine- is capable of influencing the balance of autonomic output to the heart.
CONCLUSION
In this study, thoracic manipulation was accompanied by short-term changes in HRV, whereas these changes were not seen with the sham procedure. The changes in HRV seen with thoracic manipulation imply changes in the balance of sympathetic and parasympathetic output to the heart. Hence, some effects of thoracic manipulation, including clinical effects, may be mediated in part by alterations in autonomic nervous system function. REFERENCES
1. Budgell B, Hirano F. Innocuous mechanical stimulation of the neck and alterations in heart-rate variability in healthy young adults. Auton Neurosci 2001;91:96-9. 2. Fujimoto T, Budgell B, Uchida S, Suzuki A, Meguro K. Arterial tonometry in the measurement of the effects of innocuous mechanical stimulation of the neck on heart rate and blood pressure. J Auton Nerv Syst 1999;75:109-15. 3. Thiel H, Rix G. Is it time to stop functional pre-manipulation testing of the cervical spine? Man Ther 2005;10:154-8. 4. Cleveland C. The high-velocity adjustment. In: Haldeman S, editor. Principles and practice of chiropractic. 2nd ed. Norwalk (CT): Appleton and Lange; 1992. pp. 459-81.
Journal of Manipulative and Physiological Therapeutics Volume 29, Number 8
5. McPortland J, Brodeur R, Hallgren R. Chronic pain, standing balance, and suboccipital muscle atrophyâ&#x20AC;&#x201D;a pilot study. J Manipulative Physiol Ther 1997;20:24-9. 6. Faye L, Wiles M. Manual examination of the spine. In: Haldeman S, editor. Principles and practice of chiropractic. 2nd ed. Norwalk (CT): Appleton and Lange; 1992. pp. 301-18. 7. Sobel J, Winters J, Groenier K, Arendzen J, Meyboom de Jong B. Physical examination of the cervical spine and shoulder girdle in patients with shoulder complaints. J Manipulative Physiol Ther 1997;20:57-62. 8. Vemon H, Gitelman R. Pressure algometry and tissue compliance measures in the treatment of chronic headache by spinal manipulation: a single case/single treatment report. J Can ChiroprAssoc 1990;34:141-4. 9. Fjellner H, Bexander C, Faleij R, Strender L. Interexaminer reliability in physical examination of the cervical spine. J Manipulative Physiol Ther 1999;22:5-11. 10. Humphreys B, Delahaye M, Peterson C. An investigation into the validity of cervical spine motion palpation using subjects with congenital block vertebrae as a 'gold standard'. BMC Musculoskelet Disord 2004;5. 11. Rogers R. The effects of spinal manipulation on cervical kinesthesia in patients with chronic neck pain: a pilot study. J Manipulative Physiol Ther 1997;20:80-5. 12. Nilsson N, Christensen H, Hartvigsen J. The effect of spinal manipulation in the treatment of cervieogenic headache. J Manipulative Physiol Ther 1997;20:326-30. 13. Nilsson N. A randomized controlled trial of the effect of spinal manipulation in the treatment of cervieogenic headache. J Manipulative Physiol Ther 1995; 18:435-40. 14. Boline P, Kassak K, Bronfort G, Nelson C, Anderson A. Spinal manipulation vs. amitriptyline for the treatment of chronic tension-type headaches: a randomized clinical trial. J Manipulative Physiol Ther 1995;18:148-54. 15. Bove G, Nilsson N. Spinal manipulation in the treatment of episodic tension-type headache. JAMA 1998;280:1576-9. 16. Van Schalkwyk R, Parkin-Smith G. A clinical trial investigating the possible effect of the supine cervical rotatory manipulation and the supine lateral break manipulation in the treatment of mechanical neck pain: a pilot study. J Manipulative Physiol Ther 2000;23:324-31. 17. Christensen H, Vach W, Vach K, Manniche C, Haghfelt T, Harvigsen L, et al. Palpation of the upper thoracic spine: an observer reliability study. J Manipulative Physiol Ther 2002;25:285-92. 18. Gitelman R, Fligg B. Diversified technique. In: Haldeman S, editor. Principles and practice of chiropractic. 2nd ed. Norwalk (CT): Appleton and Lange; 1992. pp. 483-501. 19. Haas M, Peterson D, Rothman E, Panzer D, Krein R, Johansen R, et al. Responsiveness of leg alignment changes associated with articular pressure testing to spinal manipulation: the use of a randomized clinical trial design to evaluate a diagnostic test with a dichotomous outcome. J Manipulative Physiol Ther 1993;16:306-ll. 20. Schiller L. Effectiveness of spinal manipulative therapy in the treatment of mechanical thoracic spine pain: a pilot randomized clinical trial. J Manipulative Physiol Ther 2001; 24:394-401. 21. Feise R. Do multiple outcome measures require p-value adjustment. BMC Med Res Methodol 2002;2. 22. Low P. Laboratory evaluation of autonomic function. In: Low P, editor. Clinical autonomic disorders. 2nd ed. Philadelphia: Lippincott-Raven; 1997. pp. 179-208. 23. Pomeranz B, Macaulay R, Caudill M, Kutz 1, Adam D, Gordon D, et al. Assessment of autonomic function in humans by heart
Budgell and Polus Heart Rate Variability
rate spectral analysis. Am J Physiol Heart Circ Physiol 1985; 248:H151-3. 24. Tetzlaff J, O'Hara J, Yoon HAS. Heart rate variability and the prone position under general versus spinal anesthesia. J Clin Anesth 1998;10:656-9. 25. Fujita M, Miyamoto S, Tambara K, Budgell B. Trepopnea in patients with chronic heart failure. Int J Cardiol 2002;84:115-8. 26. Jones A, Dean E. Body position change and its effect on hemodynamic and metabolic status. Heart Lung 2004;33: 281-90. 27. Kullok S, Mayer C, Baekon J, Kullok J. Interactions between non-symmetric mechanical vector forces in the body and the autonomic nervous system: basic requirements for any mechanical technique to engender long-term improvements in autonomic fijnction as well as in the fiinctional efficiency of the respiratory, cardiovascular, and brain systems. Med Hypotheses 1990;31:91-7. 28. Sato A, Sato Y, Schmidt R. Heart rate changes reflecting modifications of efferent cardiac sympathetic outflow by cutaneous and muscle afferent volleys. J Auton Nerv Syst 1981;4:231-47. 29. Sato A, Sato Y, Schmidt R. Changes in heart rate and blood pressure upon injection of algesic agents into skeletal muscle. PflugersArch 1982;393:31-6. 30. Sato A, Sato Y, Schmidt R. Changes in blood pressure and heart rate induced by movements of normal and inflamed knee joints. Neuroscience Lett 1984;52:55-60. 31. Sato A, Swenson R. Sympathetic nervous system response to mechanical stress of the spinal column in rats. J Manipulative Physiol Ther 1984;7:141-7. 32. Kimura A, Ohsawa H, Sato A, Sato Y. Somatocardiovascular reflexes in anesthetized rats with the central nervous system intact or acutely spinalized at the cervical level. Neurosci Res 1995;22:297-305. 33. Kimura A, Sato A, Sato Y, Suzuki H. A- and C-reflexes elicited in cardiac sympathetic nerves by single shock to a somatic afferent nerve include spinal and supraspinal components in anesthetized rats. Neurosci Res 1996;25:91-6. 34. DeBoer K, Schutz M, McKnight M. Acute effects of spinal manipulation on gastrointestinal myoelectric activity in conscious rabbits. Man Med 1988;3:85-94. 35. Lund I, Lundeberg T, Kurosawa M, Uvnas-Moberg K. Sensory stimulation (massage) reduces blood pressure in unanesthetized rats. J Auton Nerv Syst 1999;78:30-7. 36. Cottingham J, PiDrges S, Lyon T. Effects of soft tissue mobilization (Rolfmg pelvic lift) on parasympathetic tone in two age groups. Phys Ther 1998;68:352-6. 37. Cottingham J, Porges S, Richmond K. Shifts in pelvic inclination angle and parasympathetic tone produced by Rolfing sott tissue manipulation. Phys Ther 1998;68:1364-70. 38. Reed B, Held J. Effects of sequential connective tissue massage on autonomic nervous system of middle-aged and elderly adults. Phys Ther 1988;68:1231-4. 39. Camilleri M, Malagelada J, Kao P, Zinsmeister A. Effects of somatovisceral reflexes and selective dermatomal stimulation on postcibal antral pressure activity. Am J Physiol Gastrointest Liver Physiol 1984;247:G703-8. 40. Lin C, Liao J, Tsai S, Chiang P, Ting H, Tang C, et al. Depressor effect on blood pressure and flow elicited by electroacupuncture in normal subjects. Auton Neurosci 2003; 107:60-4. 41. Nishijo K, Mori H, Yosikawa K, Yazawa K. Decreased heart rate by acupuncture stimulation in humans via facilitation of cardiac vagal activity and suppression of cardiac sympathetic activity. Neurosci Lett 1997;227:l65-8.
609
610
Budgell and Polus Heart Rate Variability
42. Delaney J, Leong K, Watkins A, Brodie D. The short-term effects of myofascial trigger point massage therapy on cardiac autonomic tone in healthy subjects. J Adv Nurs 2002; 37:362-71. 43. Mok E, Woo C. The effects of slow-stroke back massage on anxiety and shoulder pain in elderly stroke patients. Complement Ther Nurs Midwifery 2004;10:209-16. 44. Pickar J, Kang Y. Short-lasting stretch of lumbar paraspinal muscle decreases muscle spindle sensitivity to subsequent muscle stretch. J Neuromusculoskelet Syst 2001;9: 88-96. 45. Pickar J, Wheeler J. Response of muscle proprioceptors to spinal manipulative-like loads in the anesthetized cat. J Manipulative Physiol Ther 2001;24:2-ll. 46. Sung P, Kang Y, Pickar J. Effect of spinal manipulation duration on low threshold mechanoreceptors in lumbar paraspinal muscles. Spine 2004;30:l 15-22.
Journal of Manipulative and Physiologieal Therapeutics October 2006
47. Symons B, Herzog W, Leonard T, Nguyen H. Reflex responses associated with activator treatment. J Manipulative Physiol Ther2000;23:155-9. 48. Bolton P, Budgell B. Spinal manipulation and spinal mobilization influence different axial sensory beds. Med Hypotheses 2006;66:258-62. 49. Sato A, Sato Y, Schmidt R. The impact of somatosensory input on autonomic functions. Rev Physiol Biochem Pharmacol 1997;130:l-328. 50. Kokkorogiannis T. Somatic and intramuscular distribution of muscle spindles and their relation to muscular angiotypes. J Theor Biol 2004;229:263-80. 51. Gandevia S, Mahutte C. Joint mechanics as a determinant of motor unit organization in man. Med Hypotheses 1980;6:527 - 33. 52. Bogduk N, Twomey L. The lumbar muscles and their fascia. In: Bogduk N, Twomey L, editors. Clinical anatomy of the lumbar spine. 2nd ed. Melbourne: Churchill-Livingstone; 1991. pp. 3 -105.