Spectral Analysis of Heart Rate Variability During Trigger-Point Acupuncture of Various Muscles

Abstract

Objective: The aim of this research was to clarify changes in cardiovascular autonomic nervous system function due to trigger-point (TP) acupuncture; the current author evaluated differences in responses among acupuncture at TPs of various muscles using spectral analysis of heart rate variability.

Materials and Methods: Subjects were 48 healthy men. Before measurements began, subjects were assigned to a TP acupuncture or control group based on presence/absence of referred pain on applying pressure to a taut band within the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, or masseter muscle. Measurements were conducted in a room with a temperature of 25°C, with subjects in a long sitting position after 10 minutes of rest. Acupuncture needles were retained for 10 minutes at 1 site on the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, or masseter muscle. Electrocardiography was performed simultaneously with respiratory-cycle measurements. Based on the R–R interval on the electrocardiograms, frequency analysis was performed, low-frequency (LF) and high-frequency (HF) components were extracted, and the ratio of LF to HF components (LF/HF) was evaluated.

Results: All subjects in the TP acupuncture group showed a transient increase in the HF component, but no significant changes in heart rate (HR) or LF/HF. In the control group, no significant changes were observed in HR, HF component, or LF/HF.

Conclusions: These data suggest that acupuncture stimulation of TPs of the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, and masseter muscle increases parasympathetic nerve activity transiently.

Introduction

While noxious stimulation is expected to increase sympathetic activity, acupuncture stimulation has frequently been reported to induce a decrease in heart rate (HR).^1–5 This is considered to require either an increase in parasympathetic activity or inhibition of sympathetic activity, or both. There are various theories regarding the neurologic mechanism of this transient decrease in HR as a response to acupuncture. In a previous study, the current author and colleagues evaluated differences in responses between acupuncture stimulation of trigger points (TPs) that are clinically considered to be associated with the autonomic nervous system (ANS) and that of other sites using parameters of spectral analysis of HR variability (HRV).⁶ This group of researchers found that acupuncture stimulation of sites other than TPs did not cause significant changes. In contrast, acupuncture stimulation of TPs caused a transient decrease in HR and an increase in the high-frequency (HF) component.

Analysis of R–R interval variations on electrocardiography (ECG) was first used as a method to evaluate autonomic activity by Wheeler and Watkins.⁷ They reported a decrease in—or disappearance of—respiratory sinus arrhythmia (measured in terms of R–R interval variation) in patients with diabetes-induced autonomic neuropathy. In the same year, Sayers⁸ introduced spectral analysis using fast Fourier transformation (FFT) to analyze HRV, suggesting the possibility of frequency analysis of HRV. Spectral analysis of HRV is a method of evaluating autonomic activity by frequency analysis of time-series data on the R–R interval per unit of time using FFT to isolate the frequency components of cardiac sympathetic or parasympathetic activity. However, spectral analysis of HRV is also affected by arrhythmias, such as premature beats, and by the respiratory rate. In particular, when the respiratory rate is ≤9 breaths/minute, frequency components of cardiac sympathetic or parasympathetic activity cannot be isolated. Therefore, the evaluation of autonomic activity using spectral analysis of HRV is complicated.

TPs are considered to be treatment points for chronic musculoskeletal pain such as that encountered in myofascial pain syndrome.^9–13 Few studies have examined changes in cardiovascular ANS function induced by TP stimulation. To clarify the effects of TP acupuncture on the ANS, changes in cardiac sympathetic and parasympathetic activities were compared among TP acupuncture in various muscles with nonstimulation, using spectral analysis of HRV.

Materials and Methods

Subjects

The subjects were healthy men (N = 48; ages, 20–28) with no histories of cardiac disease or arrhythmias, and not currently experiencing any psychologic stressors, such as examinations or relationship breakups. Before measurement, it was identified whether or not subjects experienced referred pain when points on a taut band within the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, and masseter muscle were pressed manually. Based on the results, the each subject was assigned to 1 of 5 groups: (1) an extensor digitorum muscle group; (2) a tibialis anterior muscle group; (3) a gluteus medius muscle group; (4) a masseter muscle group; or (5) a control group. The TP acupuncture groups had subjects who had referred pain as described above, while the control group had those who did not have such referred pain.

After an adequate explanation of the purpose and contents of the study, written informed consent was obtained from all subjects, and the experiments were conducted. The institutional review board of Kansai University of Health Science, Osaka, Japan, approved all protocols and informed-consent documents (Approval number, 13–49) in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.

Measurements

ECG was performed with bipolar chest leads, and time-series R–R interval data were recorded. HR (beats/minute) was calculated from the obtained R–R interval. Paced breathing was performed with cycles of 2.5 seconds of inhalation followed by 2.5 seconds of exhalation, using a metronome. Simultaneously, this respiratory cycle was confirmed based on rib-cage movements using the chest impedance method.¹⁴ Based on the obtained time-series R–R interval data (sampled every 5 minutes), frequency analysis was performed with HRV analysis software (HRV Module for Chart; AD Instruments, Australia) using FFT. The low-frequency (LF; 0.04–0.15 Hz) and HF (0.15–0.40 Hz) components were extracted. The HF component was evaluated as a parameter of cardiac parasympathetic activity and LF/HF was evaluated as a parameter of cardiac sympathetic activity. To evaluate relative changes in cardiac sympathetic/parasympathetic activities for the HF component, normalized units (n.u.; i.e., the relative value of the HF-power component in proportion to the total power minus the very low frequency [VLF] component) were used.

Protocol

Measurements were performed at least 2 hours after a meal or caffeinated beverage, between 15:00 and 19:00. Subjects adopted a long sitting position (sitting with the legs extended straight in front of the torso with the hips flexed to 90°), in a room maintained at a temperature of 25°C, and measurement was initiated after a 10-minute rest. Five minutes after the initiation of measurement, acupuncture needles were inserted in the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, or masseter muscle and retained for 10 minutes (see detailed description below). The needles were then removed, and the subjects remained in the long sitting position for 10 minutes (total measurement time, 25 minutes).

Acupuncture Stimulation

For acupuncture, stainless-steel disposable needles (50 × 0.2 mm; SEIRIN, Japan) were used with an insertion depth of ∼15 mm. Previous studies have reported that acupuncture stimulation of acupoints on the limbs induces a decrease in HR.^2–5 Therefore, in this study, points were stimulated in the extensor digitorum muscle on the upper limb, in the tibialis anterior muscle on the lower limb, in the gluteus medius muscle on the trunk, or in the masseter muscle on the face. TP on the right extensor digitorum muscle was located ∼50 mm distal to the lateral epicondyle. TP on the right tibialis anterior muscle was located ∼50 mm distal to the lateral condyle of the tibia. TP on the right gluteus medius muscle was located ∼30 mm dorsal to the anterior superior iliac spine. TP on the right masseter muscle was located ∼15 mm from the cranial gonial angle.

TPs have been defined as “sites that are present in rope-like indurations of muscle tissue, and cause referred pain when stimulated by compression, needle insertion, or heating.”¹⁵ Therefore, in this study, TPs were identified as points that induced referred pain when a taut band within the right extensor digitorum muscle, tibialis anterior muscle, gluteus medius muscle, or masseter muscle were pressed. In the 4 TP acupuncture groups, an acupuncture needle was inserted into TPs on the right extensor digitorum muscle (extensor digitorum muscle group), tibialis anterior muscle (tibialis anterior muscle group), gluteus medius muscle (gluteus medius muscle group), or masseter muscle (a masseter muscle group), confirming the presence of referred pain, and the needle was retained for 10 minutes. The subjects used hand signals to indicate whether or not they experienced referred pain. In the control group, no acupuncture needles were inserted: These subjects did not receive any verum needles or sham needles, and remained in the long sitting position for 25 minutes.

Statistical Analysis

For statistical analysis, the Kruskal–Wallis test was used to compare each parameter among the TP acupuncture and control groups, and the Mann–Whitney-U test was performed when significance was observed. To compare time-series changes, repeated-measure analysis of variance was used, and a multiple comparison test (Tukey's honestly significant difference test) was performed when significance was observed. P-values <0.05 were regarded as significant. All measurements were expressed as mean ± standard deviation (SD).

Results

Comparison During the 5-Minute Period After Initiation of Measurement (Pre-Acupuncture Comparison; Table 1)

Table 1.

Comparison of the Prestimulation Value of Each Parameter

Parameters	Masseter muscle group (n = 9)	Extensor digitorum muscle group (n = 10)	Gluteus medius muscle group (n = 9)	Tibialis anterior muscle group (n = 10)	Control group (n = 10)	P-value
Mean age (yrs)	22.8 ± 3.0	22.7 ± 2.4	23.2 ± 2.3	22.8 ± 1.9	22.5 ± 1.8	0.921
HR (bpm)	72.6 ± 13.5	66.0 ± 8.2	66.2 ± 10.5	70.5 ± 11.3	69.2 ± 13.7	0.665
HF (n.u.)	57.8 ± 11.6	53.9 ± 9.3	50.2 ± 13.5	48.3 ± 14.3	58.6 ± 15.6	0.499
LF/HF	0.7 ± 0.4	0.8 ± 0.3	1.1 ± 0.6	1.2 ± 0.7	0.8 ± 0.7	0.448

Values are mean ± standard deviation.

yrs, years; HR, heart rate; bpm, beats per minute; HF, high-frequency component; n.u., normalized units; LF, low-frequency component.

Subjects were assigned to 1 of 5 groups prior to measurements: an extensor digitorum muscle group (n = 10: age = 22.7 ± 2.4 years); a tibialis anterior muscle group (n = 10: age = 22.8 ± 1.9 years); a gluteus medius muscle group (n = 9: age = 23.2 ± 2.3 years); a masseter muscle group (n = 9: age = 22.8 ± 3.0 years); and a control group (n = 10: age = 22.5 ± 1.8 years). Before acupuncture stimulation, no significant differences were observed among the 5 groups in HR. Spectral analysis also revealed no significant differences among the 5 groups in terms of HF component or LF/HF.

Changes in Heart Rate and Spectral Analysis During Acupuncture Stimulation (Figs. 1–3)

FIG. 1.

Changes in heart rate (HR) in the trigger-point acupuncture stimulation groups (masseter muscle: circles; extensor digitorum muscle: diamonds; gluteus medius muscle: triangles; and tibialis anterior muscle: squares) and the control group (crosses). No consistent changes were observed in all groups (P > 0.05). bpm, beats per minute; min, minute.

FIG. 2.

Changes in the high-frequency (HF) component in the trigger-point (TP) acupuncture stimulation group (masseter muscle: circles; extensor digitorum muscle: diamonds; gluteus medius muscle; triangles, and tibialis anterior muscle: squares) and the control group (crosses). In each stimulation group the mean values of the HF component at 5-minute intervals are shown. In all of the TP acupuncture groups, the HF component increased significantly, compared with the prestimulation value during the prior 5 minute period of acupuncture stimulation and recovered to a value close to the prestimulation value after removal of the acupuncture needle (*P < 0.05). n.u., normalized units; min, minutes.

FIG. 3.

Changes in the ratio of low-frequency (LF) to high-frequency (HF) components (LF/HF) in the trigger-point (TP) acupuncture stimulation group (masseter muscle: circles; extensor digitorum muscle: diamonds; gluteus medius muscle: triangles; and tibialis anterior muscle: squares) and the control group (crosses). In each stimulation group, the mean values of LF/HF at 5-minute intervals are shown. LF/HF did not change significantly in all the TP acupuncture groups and in the control group (P > 0.05). min, minutes.

In all TP acupuncture-stimulation groups and the control group, HR did not change significantly (P = 0.082).

In the TP acupuncture-stimulation groups, the HF component increased significantly during the first 5-minute period of acupuncture stimulation, compared with the prestimulation value (extensor digitorum muscle group: P = 0.005; tibialis anterior muscle group: P = 0.021; gluteus medius muscle group: P = 0.049; and masseter muscle group: P < 0.001), and recovered to close to the prestimulation value after removal of the needles. In the control group, the HF component showed no significant changes (prestimulation versus the first 5-minute period of acupuncture stimulation: P = 0.397).

In all TP acupuncture-stimulation groups and the control group, there were no significant changes in LF/HF (P = 0.14).

Comparison of the HF Component Between the First 5-Minute Period of Acupuncture Stimulation and Before Stimulation (Fig. 4)

FIG. 4.

Comparison of the high-frequency (HF) component for the first 5-minute period of acupuncture stimulation from the stimulation previous value ( HF). In each stimulation group, the mean values of HF are shown. Significant differences of HF were found between all TP acupuncture stimulation groups and the control group (*P < 0.05). Significant differences of HF was not found between the all trigger-point groups.

A significant difference in the HF component between the first 5-minute period of acupuncture stimulation and before stimulation (ΔHF) was found between all TP acupuncture-stimulation groups and the control group (extensor digitorum muscle group versus. the control group: P < 0.001; tibialis anterior muscle group versus the control group: P = 0.002; gluteus medius muscle group versus the control group: P < 0.001; and masseter muscle group versus the control group: P < 0.001). No significant difference in the HF component for the first 5-minute period of acupuncture stimulation and before stimulation (ΔHF) was found between any of the TP groups (extensor digitorum muscle group versus tibialis anterior muscle group: P = 0.796; extensor digitorum muscle group versus gluteus medius muscle group: P = 0.604; extensor digitorum muscle group versus masseter muscle group: P = 0.652; tibialis anterior muscle group versus. gluteus medius muscle group: P = 0.905; tibialis anterior muscle group versus masseter muscle group: P = 0.315; and gluteus medius muscle group versus masseter muscle group: P = 0.340).

Discussion

While noxious stimulation is expected to increase sympathetic activity, acupuncture stimulation decreases HR. In the present study, changes in autonomic nervous activity were evaluated by spectral analysis of HRV during acupuncture stimulation of TPs in various muscles with no acupuncture stimulation as a control. Spectral analysis showed a significant transient increase in the HF component in all 4 TP acupuncture-stimulation groups; this increase did not occur in the control group.

The results observed in the TP acupuncture groups suggest a significant increase in parasympathetic activity and inhibition of sympathetic activity. LF/HF is a parameter of relative sympathetic activity, and its decrease does not always represent inhibition of sympathetic activity. Therefore, the interpretation of this parameter requires caution. In the present study, the TP acupuncture groups showed an increase in the HF component but no significant changes in LF/HF. This constancy in LF/HF might, in fact, represent changes in both sympathetic and parasympathetic activity (i.e., an increase in LF tending to balance out the increase in HF). Therefore, it is possible that the present findings indicate that a rise in parasympathetic activity surpassed any increase in sympathetic nerve activity due to acupuncture stimulation, resulting in the apparent inhibition of sympathetic activity evaluated in terms of LF/HF as a relative parameter. The LF component and LF/HF ratio have been reported as parameters of sympathetic activity on spectral analysis of HRV, but their usefulness has not been confirmed and should be clarified by further studies.

The advantage of spectral analysis of HRV is the noninvasive quantitative evaluation of cardiovascular autonomic function. However, HRV reflects changes in various physiologic functions, such as thermoregulation, and is also affected by age and body position. In addition, the HF component, reflecting respiratory-sinus arrhythmia, is also affected by the respiratory rate.^16,17 In the present study, no changes in the respiratory rate (as measured by the strain-gauge method) were observed during acupuncture stimulation. Thus, the increase in the HF component during needle retention appears to, instead, represent a transient increase in parasympathetic activity due to TP acupuncture.

There are various theories regarding the neurologic mechanism of a transient decrease in HR due to acupuncture stimulation.^1–5 Uchida et al. found that the decrease in HR induced by acupuncture stimulation disappeared with cardiac sympathectomy but not with vagotomy in anesthetized rats.² Another study by Uchida et al. found that the reflex pathway involved in the decrease of HR with acupuncture stimulation comprises groups III and IV muscle-afferent nerves whose activation stimulates gamma aminobutyric acid–ergic neurons in the brainstem and inhibits sympathetic outflow to the heart.³

Yet, in the present study, parasympathetic activity increased during TP acupuncture, but no significant changes were observed in sympathetic activity. Therefore, the present findings are considered to be similar to those of the studies by Bäcker et al.⁴ and Haker et al.⁵ The influence of anesthesia on the ANS and differences in the stimulation conditions have been suggested as being responsible for these differences in findings. Studies examining the effect of TP stimulation by compression have also shown an increase in parasympathetic activity,^18,19 confirming that stimuli other than acupuncture also cause this response. These results suggest that TPs are sites causing an increase in parasympathetic activity when stimulated, and that this increase is not a response specific to the types of stimuli. In the present study, the subjects were healthy volunteers. TPs in healthy volunteers who are not experiencing pain may be latent rather than active in nature.⁹ Therefore, the results might have been different in patients with myofascial pain syndrome. Moreover, the design of the present study meant that subjects were preselected and assigned to groups rather than being randomized, which might have introduced bias.

With respect to the mechanism underlying TPs, the hypothesis integrating the motor endplate and energy-crisis theories is widely known.²⁰ This integrated hypothesis states that excessive release of acetylcholine from the motor endplate induces excessive contraction of localized muscle fibers, thereby compressing the blood vessels, and that the resulting relative oxygen insufficiency leads to an energy crisis, causing tissue injury. As a result, various sensitizers, such as metabolites and inflammatory products, are produced, sensitizing nociceptors and causing pain. However, Kawakita et al.²¹ suggested the occurrence of polymodal receptor sensitization and associated deep-tissue edema (polymodal receptor sensitization hypothesis), based on results of observation in a delayed muscle-pain model.

Thus, various hypotheses have been proposed, but the detailed mechanism of TP formation is still unclear. In each hypothesis, pain develops due to receptor sensitization and stimulation of the sensitized area. Therefore, for pain relief to occur as a result of TP stimulation, receptor desensitization or removal of stimulating factors—such as taut bands, metabolites, and inflammatory products—is necessary. Given that an increase in parasympathetic nerve activity increases peripheral blood flow^22,23 accumulated metabolites and inflammatory products are circulated, reducing congestion. This is an important factor for relief of chronic pain by receptor desensitization and removal of stimulating factors^23,24 The present study suggests that the ANS (in particular, the parasympathetic component) is involved in the mechanism of relief of chronic motor pain, such as myofascial pain syndrome, due to TP stimulation. Future studies should have a greater sample size to study the effects of TP stimulation further.

Conclusions

To clarify changes in cardiac autonomic nervous function due to TP acupuncture stimulation, differences in responses among acupuncture of TPs in various muscles were evaluated, using parameters of spectral analysis of HRV. Acupuncture stimulation of TPs in all muscles investigated caused a transient decrease in HR and an increase in the HF component. These results suggest that TPs are sites that tend to cause an increase in parasympathetic activity when stimulated.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research did not receive any specific grant from any funding agency in the public, commercial and non-profit sectors.