Abstract
Context: Rectus capitis posterior (RCP) muscles have physical attachments to the pain-sensitive spinal dura. Atrophy of these muscles is associated with chronic headache in some patients. The authors suspect that the significance of atrophy in the RCP muscles has been undervalued because the functional role of these muscles is not well defined.
Objective: To determine whether a statistically significant change in normalized levels of electromyographic activity in RCP muscles occurs when the head is voluntarily moved from a self-selected neutral head position to a protruded head position.
Methods: Fine wire, intramuscular electrodes were used to collect electromyographic data as asymptomatic participants moved their head from a neutral head position into a forward head position and back into the neutral head position. This sequence was repeated 4 times. Normalized levels of electromyographic activity were quantified using a 2-head position × 2 sides of the body repeated measures design that incorporated mixed-effects β regression models.
Results: Twenty participants were studied. Electromyographic activity collected from RCP muscles was found to increase as the head was voluntarily moved from a self-selected neutral head position (11% of maximum voluntary isometric contraction [MVIC] in RCP minor, 14% of MVIC in RCP major) into a protruded head position (35% of MVIC in RCP minor, 39% of MVIC in RCP major) (P<.001).
Conclusion: Rectus capitis posterior muscles may contribute to segmental stabilization of the occipitoatlantal and atlantoaxial joints by helping to maintain joint congruency during movement of the head.
The rectus capitis posterior (RCP) muscles comprise the RCP minor (RCPm) and the RCP major (RCPM). The RCPm muscles arise from the posterior tubercle of the posterior arch of the atlas (C1)—the only muscles that attach to the posterior arch of C1—and insert into the occipital bone inferior to the inferior nuchal line and lateral to the midline.
1 The RCPM muscles arise from the spinous process of the axis (C2) and insert into the lateral part of the inferior nuchal line of the occipital bone.
1
Bilateral contraction of RCP muscles has been reported to contribute to extension of the head at the occipitoatlantal (OA) and the atlantoaxial (AA) joints, and unilateral contraction contributes to ipsilateral rotation and sidebending of the head.
1 However, considering their small size, relative to larger muscles acting in parallel with them, RCP muscles would not be expected to substantially contribute to gross movements of the head.
2,3 The high spindle density in these muscles suggests that they provide feedback to the central nervous system that is related to motion and position of the head.
4-7 Thus, the morphologic and biomechanical properties of RCP muscles seem to be optimized to contribute to segmental stability of the OA and AA joints as a function of head position through control of the angular position of the posterior arch of the atlas and the posterior process of the axis relative to the occiput and to one another.
8,9
Atrophy of the RCP muscles observed on magnetic resonance images has been reported in patients with chronic headache associated with a nontraumatic event.
10,11 A forward head posture is commonly seen in patients with headache and has been reported to be more pronounced in patients with chronic tension-type headache than in controls without headache.
12 It has been hypothesized that forward head posture in patients with chronic tension-type headache may be responsible for disuse atrophy of the RCP muscles.
11 Physicians have long recognized that structure and function are reciprocally related and that compromising the functionality of one vertebral component can have a large impact on the stability of other vertebral motion segments.
Voluntary protrusion of the head within the sagittal plane without rotation is a physiologic motion (
Figure 1). Protrusion of the head would appear to increase anterior loading of the articular facets, resulting in extension of the OA and AA joints and shortening of RCP muscles.
13,14 Voluntary retraction of the head would appear to increase posterior loading of the articular facets, resulting in flexion of the OA and AA joints and lengthening of RCP muscles. Normal levels of electromyographic (EMG) activity in RCP muscles have been reported to statistically significantly increase as asymptomatic participants moved their head from a self-defined neutral head position (NHP) to a retracted head position.
15,16 Levels of EMG activity in RCP muscles with the head in a protruded head position (PHP) have not previously been published, to the authors’ knowledge. Therefore, we tested the null hypothesis that normalized activation levels of RCP muscles would be equal regardless of whether a participant’s head was in an NHP or a PHP.
An e-mail advertisement was used to recruit potential participants from the second-year student population of the Michigan State University College of Osteopathic Medicine in East Lansing. Participants were required to be free of head and neck pain, to be free of major motion restrictions, and to have no history of surgical procedures in the region of the upper cervical spine. Participant age was limited to between 20 and 40 years because muscle atrophy has been found to increase with age.
17 The study was approved by Michigan State University’s institutional review board.
The research protocol was reviewed with each potential participant. Potential participants willing to proceed with the study were required to give their informed consent by signing a form approved by the institutional review board. Participants were compensated $150. Compensation was not conditional on completion of the study.
On arrival at the Center for Orthopedic Research, the protocol was reviewed with each participant, and routine demographic and physical data (eg, sex, height, weight) were collected. The participant was then seated in a PRseries EWC-40, cervical flexion/extension device (Atlantis Engineering). A detailed explanation describing how the device was configured for each participant has been described in a previous publication.
15 After exiting the device, the participant positioned himself or herself prone on a standard examination table. Ultrasonography was used to confirm that no congenital blood vessel anomalies were present. The suboccipital region was prepared with rubbing alcohol. A ground electrode was placed just above the middle of the spine of the right scapula, and 25-gauge disposable bipolar stainless steel fine-wire hooked EMG electrodes were inserted into the right and left RCP muscles.
15 The participant then returned to the cervical flexion/extension device where viability of the inserted electrodes was assessed. Lack of signal or excessive noise was noted and corrected when possible.
Raw EMG data were simultaneously collected from the RCPm and RCPM muscles on both sides of the participant’s body. The electromyographic signals were amplified, passed through a bandpass filter (20-1500 Hz), and digitized at 3000 samples per second. Matlab (The Mathworks) software was used to digitally rectify the data, remove offset voltages and QRS waves, and filter at a low-pass cutoff frequency of 0.5 Hz (dual pass, fourth-order Butterworth filter).
A self-selected NHP, described in detail in a previous publication,
3 was used as a reference position for all tests. Participants performed a series of 3 MVIC efforts against a fixed resistance pad for a period of 5 seconds while data were collected. A 10-second rest period was allowed between each of the efforts. The largest EMG voltage observed for a given muscle (RCPm or RCPM) on a given side of the body (left or right) across these 3 efforts was defined as MVIC. A value of MVIC for each participant was used to normalize NHP and PHP data, thereby facilitating comparison of muscle activation between muscles and among participants by measuring it as a percentage of MVIC.
15,18,19
The position of the resistance pad was adjusted so that it would allow unrestricted protrusion of the head. Each participant held the NHP for 5 seconds, glided their head into a “chin-out” PHP position for 5 seconds, and then returned to the NHP for 5 seconds. This motion cycle was repeated 4 times while data were continuously collected. Participants were instructed to protrude their head as far as was possible without rotation or sidebending. Because we were interested in quantifying individual levels of muscle activation relative to an individual MVIC, participants served as their own control. After completion of all trials, the fine-wire electrodes were removed.
Dividing the EMG voltage observed during a trial by the relevant MVIC voltage (from same person, muscle, and side of the body) yielded the raw value for muscle activation as a percentage of MVIC. These raw values ranged from 0.000% to 96.250% for the RCPm muscles, and 0.000% to 97.500% for the RCPM muscles, but the values were rescaled using Zimprich Equation 320 to place them on the open-unit interval (0, 1) before analysis for consistency with the methods used in previous studies.
15,16 Rescaled muscle activation values ranged from 0.206% to 96.060% for the RCPm muscles and 0.206% to 97.305% for the RCPM muscles.
Head position and side of the body were dummy-coded binary variables, with NHP and the left side of the body serving as reference levels. Head position was the key predictor. Side of the body was modeled to account for potential asymmetry in muscle activation and to assess whether any head position effect depends on which side was observed. Sex (0, female; 1, male) and body mass index were used as covariates. The latter variable was centered around its grand mean to simplify interpretation21 and avoid numerical estimation problems.
22
Our data screening criteria followed those applied in previous papers.
15,16 The research design planned for each participant to contribute 36 observations (2 muscles × 2 sides of the body × 9 trials), which would yield a sample of 720 possible EMG observations (20 participants × 36 observations). However, the available sample was reduced by missing or unusable data. All data for an electrode were excluded if the MVIC voltage approached 0, the PHP voltages were all greater than the MVIC, or all NHP voltages were high (>30% of the MVIC voltage). Individual trials were excluded if PHP voltages were greater than MVIC voltage or if the NHP voltages were greater than 40% of the MVIC voltage.
We analyzed the data with 2 mixed-effects β regression models,
15,16,20,22,23 1 for RCPm data and the other for RCPM data. This method appropriately models outcomes measured as proportions and accounts for correlated observations arising from the repeated measures design, the skewed distribution of muscle activation, and the fact that PHP trials exhibit greater variance in muscle activation than NHP trials. Mean levels of muscle activation were predicted by a location submodel containing fixed effects of head position, side of the body, sex, and body mass index; a head position by side interaction; and a participant-level random intercept. The dispersion submodel used fixed effects of head position, side, and a head position × side interaction to model the skewness and heteroscedasticity shown in
Figure 2.
The 20 participants (13 men and 7 women) comprising the study cohort were the same participants used in 2 other studies,
15,16 which both reported analyses of data collected as part of a larger study. The mean (SD) age of participants was 25 (2) years; height, 177 (12) cm; weight, 74 (15) kg; and body mass index, 24 (4).
Participant 2 could not tolerate insertion of the electrodes, so none of the 36 planned trials for that participant could be collected. For the RCPm muscles, all data from participant 19 were excluded because he or she had high NHP voltage values. For participant 9, all right-side trials were excluded because of high NHP voltages; for participants 5, 6, 7, 8 and 10, all left-side trials were excluded because of high NHP voltages. Overall, 82 trials were excluded, as well as 1 participant from whom we collected EMG data, leaving 260 trials from 18 participants in the analysis (76% of the 342 RCPm trials actually collected).
For the RCPM muscles, all data from participants 7 and 10 were excluded because they either had high NHP voltage values or their PHP voltages consistently exceeded the MVIC voltage. For participant 9, all right-side trials were excluded because of high NHP voltages, for participants 11 and 19, all left-side trials were excluded because of high NHP voltages. Overall, 69 trials and 2 participants from whom we collected EMG data were excluded, leaving 273 trials from 17 participants in the analysis (80% of the 342 RCPM trials actually collected).
Head position strongly affected normalized levels of EMG activity observed in the RCPm and RCPM muscles of participants (OR, 5.04;
P<.001) (
Figure 3). In the RCPm muscles, mean activation was about 10% to 12% of MVIC in a self-defined NHP, but it increased to around 34% to 36% of MVIC in a PHP (OR, 5.04; 95% CI, 3.33-7.62). In the RCPM muscles, mean activation was about 11% to 18% of MVIC in a self-defined NHP, but it increased to around 35% to 42% of MVIC in a PHP (OR, 4.65; 95% CI, 3.51-6.16).
The effects of side of the body, sex, and BMI on mean RCPm muscle activation were not statistically significant. In addition, the head position × side of the body interaction terms were not statistically significant in either model, indicating that there is no evidence of lateral asymmetry in the size of the head position effect.
Many studies have explored the role of ligaments, joint capsules, and the odontoid process in the maintenance of OA-AA joint stability.
24,25 For example, the amount of posterior translation of the head with respect to C1 is by necessity constrained to less than 3 mm to avoid injury to the spinal cord.
26,27 The morphology of the occipital condyles and the superior articular facets of the atlas are critical components that contribute to OA joint stability. The anteroposterior curvature of the superior articular facet surfaces of the atlas transitions from an angle of approximately 11° at age 1 year to an angle of approximately 44° in adults, with the curvature reaching 90% of the final value by approximately age 8 years.
28 The relative flatness of these joint surfaces in infants and children younger than 6 years have been thought to reduce joint stability and increase the risk of nonphysiologic motion, such as OA dislocation, to occur even at relatively low intensities of impact.
29,30
In the lower cervical spine, muscles function to stabilize the head and neck, maintaining posture and guarding against movements caused by unexpected external forces.
5 It has also been suggested that in the upper cervical spine, the suboccipital muscles promote stability.
31 Biomechanical models have shown that short, deep muscles are necessary to maintain stability of the lumbar spine.
32
The atlas has been characterized as a passive element whose movements are essentially a function of head position.
33 Voluntary protrusion of the head has been postulated to increase anterior loading of the articular facets, resulting in extension of both the OA and AA joints with simultaneous shortening of the RCP muscles. Furthermore, voluntary retraction of the head may increase posterior loading of the articular facets, resulting in flexion of both the OA and AA joints with simultaneous lengthening of the RCP muscles. This model is incomplete in that it does not explain the increased levels of EMG activity in both sets of muscles that we have observed during protrusion of the head (35% of MVIC in RCPm and 39% of MVIC in RCPM in the present study) and retraction of the head (33% of MVIC in RCPm15; 31% of MVIC in RCPM16). An activity level of 30.0% of MVIC has been reported to be consistent with sustained, static contraction of a muscle for approximately 1.0 to 2.5 minutes.
34,35 Our data show increased EMG activity with both retraction and protrusion of the head, indicating that activation of RCP muscles is not dependent on the final position of the head in the sagittal plane. Active response that is independent of the direction of joint motion is characteristic of the active response of deep lumbar multifidus muscles, which have been classified as joint stabilizers.
36
Our research suggests that RCP muscles may promote joint stability by increasing joint congruency. A previous report by Hallgren et al15 demonstrates that voluntary head retraction results in a statistically significant increase in EMG activity in RCP muscles as they lengthen, resulting in eccentric contractions. The current study demonstrates that voluntary head protrusion also results in a statistically significant increase in EMG activity in RCP muscles as they shorten (rather than lengthen), resulting in concentric contractions. Both eccentric and concentric contractions are known to strengthen muscle.
A key tenet of osteopathic medicine is that structure and function are reciprocally related and that disease in one part of the body can have a major effect on the whole body. Atrophy of RCP muscles has been seen in some patients who have idiopathic head and neck pain.
10 Muscle atrophy will result in muscle weakness and has the potential to destabilize the OA-AA complex, resulting in the forward head posture that is commonly seen in patients with chronic tension-type headache.
12 It also has the potential to compromise the functional relationship between the RCP muscles and the spinal dura. A functional connection has been reported to exist between the RCP muscles and the pain-sensitive spinal dura of the posterior cranial fossa.
37 Nociceptive activity from the dura, ascending through the trigeminocervical complex, is responsible for headache pain in some patients.
We have shown that both voluntary retraction and protrusion of the head selectively activate RCP muscles. Future research will test the hypothesis that atrophied RCP muscles can be strengthened by using these head postures as treatment protocols. We expect to achieve an increase in muscle strength that can be quantified by an increase in the functional cross-sectional area of these muscles seen on magnetic resonance images.
38
The study has several weaknesses. First, we discarded 20% to 24% of the data collected because of dislodged or noisy electrodes or implausible values. Second, we reported on levels of muscle activation while participants maintained static head positions but not during dynamic head motions. Third, we had a small sample size, which limits the generalizability to a larger population.
Electromyographic activity in RCP muscles was found to increase significantly as the head was voluntarily moved within the sagittal plane without rotation from an NHP to a PHP. Previous work has shown a statistically significant increase in EMG activity of RCP muscles during retraction of the head within the sagittal plane (33% of MVIC in RCPm15; 31% of MVIC in RCPM16). Considered together, these findings would suggest that both RCPM and RCPm muscles may contribute to stabilization of the OA and AA joints by helping to maintain joint congruency. Future work will test the hypothesis that we can strengthen atrophied RCP muscles by using voluntary protrusion and retraction exercises as treatment protocols.