Sleep architecture and the absence of trapezius muscle atonia in women with chronic whiplash-associated disorder: a pilot study

Abstract

Sleep disturbances frequently occur in people with whiplash-associated disorder (WAD) and have been evaluated using questionnaires or actigraphy. It is not clear whether sleep architecture, as assessed by polysomnography (PSG), is altered in individuals with WAD. Additionally, in people with WAD, muscle dysfunction is observed during tasks performed during wakefulness. Thus, this study aimed to analyze the sleep architecture of patients with chronic WAD as well as to evaluate trapezius muscle activity during sleep. Nine women with chronic WAD and nine healthy age-matched women participated in the study. Two PSG recordings were conducted for each participant. Surface electromyography signal samples of the right and left trapezius, and mentonian muscles were obtained from N2, N3, and REM sleep stages for analysis. Significant differences were found in the percentages of total sleep time in the N1 and N2 stages between the two groups. While the muscle tone of the three muscles analyzed decreased progressively across the sleep stages in the healthy group, in the chronic WAD group, this decrement was observed only in the mentonian muscle, and the trapezius muscle continued to exhibit the same muscle tone throughout the sleep stages without atonia during REM sleep. The absence of trapezius muscle atonia during REM sleep in the WAD patients may be related to dysfunction of the mechanisms that regulate motor activity.

Introduction

Whiplash is produced by a sudden jerking motion of the muscles of the neck due to acceleration–deceleration forces, and can cause injury to various tissues and areas such as the articular facet, zygapophyseal joint, intervertebral discs, and neck and shoulder muscles [1]. Whiplash-associated disorder (WAD) encompasses the sequelae of the injury and is different from the mechanism that produced the injury [2].

The severity of WAD disability and symptoms such as pain decrease rapidly within the first 3 months following the injury; however, if there is no improvement after this time, the symptoms are likely to remain for a long period [3]. Prolonged recovery can occur in as many as 50% of cases [2]. Studies with extended follow-up data suggest that recovery is slower in individuals that have more severe initial symptoms [4]. Adverse prognostic factors include severe pain, numbness or pain in the arms, older age, being female, and severe trauma [2, 5]. However, the factors that accurately predict the chronification of this disorder have not been established [3].

Studies have shown that individuals with WAD have difficulty beginning and maintaining sleep, insomnia, and poor sleep quality [6–8], especially when experiencing acute pain [9, 10]. Previous research has evaluated sleep using actigraphy or questionnaires such as the Pittsburgh sleep quality index. However, polysomnographic (PSG) studies have not been conducted, which are necessary to determine whether sleep architecture is altered.

Muscle dysfunction contributes to the chronification of WAD [11–14]. However, while muscle dysfunction has been studied during wakefulness during the execution of tasks such as muscle contractions [11, 15, 16], it is not clear whether this type of dysfunction is modified during sleep. With some exceptions, muscle atonia regularly occurs in postural muscles during rapid eye movement (REM) sleep [17]. However, changes in activity across the sleep stages in damaged muscles in individuals with WAD have not been reported in the existing literature.

The objectives of this study were to determine whether sleep architecture differs between a group of women with chronic WAD and a healthy age-matched group, and to evaluate the activity of the trapezius muscle during sleep stages in both groups.

Materials and methods

A cross-sectional, case–control study was carried out.

Participants

The study participants were women over 18 years old who had been diagnosed with WAD over 3 months ago as the result of a car accident and had pain or discomfort in their neck and shoulders. Participants who reported frequent neck pain before the accident that caused WAD were excluded from the study. Other criteria for exclusion were loss of consciousness during the accident and diagnosis of cervical or lumbar radiculopathy, scoliosis, or fibromyalgia. Individuals who used pain medication daily were also excluded. The healthy group included women who had not been diagnosed with WAD or any chronic diseases, were not undergoing drug treatments, and did not have any sleep disorders. In addition, individuals with frequent neck and shoulder pain were excluded. Both groups were recruited through personal contacts of the research staff.

All participants were provided with information on the PSG procedure and subsequently signed an informed consent form. All study procedures were conducted in accordance with the Declaration of Helsinki [18].

Procedure

Two PSG recordings were conducted during consecutive nights at the Neurosciences Laboratory of the Psychology School of the National Autonomous University of Mexico. PSG was conducted using Cadwell Easy II equipment (Kennewick, Washington, USA). The first PSG was performed to acclimate the participants to the recording conditions and also served to detect the presence of any indicators of sleep disorders. In the second PSG, frontal (F3, F4), central (C3, C4), and occipital (O1, O2) electroencephalogram (EEG) leads with contralateral references to the mastoids were used. Electrooculogram (EOG), electrocardiogram (ECG), thoracic respiratory effort, and surface electromyography (EMG) of the mentonian muscle and the right and left trapezium muscles were also performed. In addition, sleep position was evaluated with a body position sensor attached to a respiratory effort belt. The start time of the PSG recording session was chosen based on the typical bedtime of each participant.

The electrodes were placed on the muscle belly of the trapezius muscle at a midpoint between the C7 spinous process and the acromial process, with an interelectrode distance of 2 cm. The sampling frequency of all the recorded signals was 400 Hz, and a notch filter was established at 60 Hz. Bandpass filters were used for all signals except the EMG according to the recommendations of the American Academy of Sleep Medicine (AASM) [19].

For this study, only the data obtained from the second PSG recording of participants who did not present indications of sleep disorders during the first night were considered.

WAD classification

WAD was classified based on the Quebec Task Force which grades symptoms as follows:

• Grade 0 indicates no neck complaints and no physical signs.

• Grade I indicates injuries resulting in complaints of neck pain, stiffness, or tenderness but no physical signs.

• Grade II refers to neck complaints accompanied by decreased range of motion and point tenderness (musculoskeletal signs).

• Grade III refers to neck complaints accompanied by neurological signs such as decreased or absent deep tendon reflexes, weakness, and/or sensory deficits.

• Grade IV refers to injuries in which neck complaints are accompanied by fracture or dislocation.

Pain evaluation

Pain intensity was evaluated with a numerical pain rating scale composed of 11 integers from 0 to 10, with 0 indicating the absence of pain and 10 indicating the highest pain intensity. Participants rated their pain by choosing the number that best fit their pain intensity.

Sleep analysis

Sleep was scored following the AASM criteria [19]. The following variables were quantified: total sleep time (TST) and the percentage of TST spent in the N1, N2, N3, and REM sleep stages. Additionally, wake after sleep onset, sleep efficiency, sleep latency, and REM sleep latency were assessed. Sleep latency was defined as the time between the beginning of the recording (light off) until the first epoch of stage N1, and REM sleep latency was defined as the time between the first epoch of sleep and the first epoch of REM sleep.

Analysis of muscle activity

EMG signal samples that did not include movement artifacts were selected from the right and left trapezius and mentonian muscles in the N2, N3, and REM sleep stages. All samples were collected in the supine position. In addition, during REM sleep, signals with bursts on the mentonian muscle, which are characteristic of the REM stage, were omitted. Because a band-pass filter was used in the 20–70 Hz band for the EMG signals, the trapezius muscle signals showed cardiac activity contamination, mainly of the R wave. To remove this contamination, MyoResearch-XP V.1.07 Master Edition software (Noraxon, U.S.A. Inc.) was used, which applies an algorithm that selectively detects and removes R waves from EMG signals to preserve the original power of the EMG signal. After cleaning the EMG signals, each sample was segmented into 5-s bins and processed by root mean square (RMS) in 100-ms windows. Finally, each 5-s bin was averaged for statistical analysis.

Statistical analysis

Groups were compared using the Mann–Whitney U test. Furthermore, in the chronic WAD group, Spearman correlation analysis was used to assess correlations between sleep variables and the perception of pain.

To evaluate the characteristics of muscle activity among the N2, N3, and REM stages, the Friedman test was used for each group. The significance level for all analyses was set at p < 0.05. For statistically significant results of the Friedman test, post-hoc analyses of the N2-N3, N2-REM, and N3-REM pairs were performed with the Wilcoxon signed-rank test, and the significance level was set at p < 0.017.

Results

Sixteen women diagnosed with WAD were initially contacted and assessed, but five of these women did not meet the inclusion criteria. Of these five women, three had comorbid conditions; two had fibromyalgia, and one had lumbar scoliosis. In two other women, the whiplash was not a consequence of a car accident. Of the remaining 11 participants, one did not complete the two PSG recordings, and another participant was excluded due to signal artifacts in the right trapezius muscle activity throughout the second PSG. Thus, a sample of nine women with chronic WAD was used, which had a mean (standard deviation) age of 37.7 (14.3) years and an occasional moderate self-reported pain level of 5.2 (1.6). The severity of WAD was grade I in three participants and grade II in six participants, according to the Quebec Task Force classification system. The time from the diagnosis of WAD was 3.9 (1.4) years. The healthy control group included nine women with an average age of 35.7 (14.6) years.

Considering only the statistically significant differences between the two groups, the healthy group had a lower percentage of TST in stage N2 than the chronic WAD group (Table 1). In the chronic WAD group, the percentage of TST in stage N1 was lower. No statistically significant Spearman correlations were found between the perception of pain and the following sleep variables: TST (rho = − 0.22, p = 0.57), wake after sleep onset (rho = 0.39, p = 0.31), sleep efficiency (rho = − 0.34, p = 0.37), sleep onset latency (rho = − 0.56, p = 0.12), REM sleep latency (rho = − 0.14, p = 0.73), percentage of TST in stage N1 (rho = 0.5, p = 0.17), percentage of TST in stage N2 (rho = 0.15, p = 0.69), percentage of TST in stage N3 (rho = − 0.09, p = 0.83), and percentage of TST in REM sleep (rho = 0.2, p = 0.61).

Table 1.

Comparison of sleep variables between the two groups

	Chronic WAD group (n = 9)	Healthy group (n = 9)	U	p
TST (min)	428 (426–464)	451 (392–456)	30	ns
N1 (%)	7 (6–10)	11 (8–17)	15.5	0.024
N2 (%)	61 (59–68)	56 (46–58)	11	0.009
N3 (%)	13 (5–16)	19 (17–20)	19	ns
REM (%)	20 (18–21)	16 (14–17)	19	ns
WASO (min)	26 (8–51)	21 (15–39)	36	ns
SE (%)	90 (89–96)	94 (88–96)	39.5	ns
SOL (min)	5 (4–10)	6 (4–9)	35	ns
RL (min)	112 (68–130)	74 (66–100)	32	ns

Median (Quartile 1–Quartile 3), ns not significant, TST total sleep time, WASO wake after sleep onset, SOL sleep onset latency, SE sleep efficiency, RL REM latency

In the raw EMG signals, we observed that the muscle tone of both trapezius muscles remained similar between sleep stages in WAD patients, whereas in the healthy group, the muscle tone decreased in the trapezius muscles during REM sleep (Fig. 1). The Friedman's test revealed statistically significant differences in muscle tone among the N2, N3, and REM sleep stages in the healthy group in the three muscles analyzed, while the chronic WAD group only showed these differences in the mentonian muscle (Table 2). A paired analysis performed with the Wilcoxon signed-rank test showed differences between the REM sleep stage and the other sleep stages in the mentonian muscle in both groups (Fig. 2) and in the right and left trapezius muscles only in the healthy group (Fig. 3).

Fig. 1.

Comparison of the raw EMG signal between sleep stages. a In the WAD patient, only the muscle tone of the mentonian muscle decreased during REM sleep. b The muscle tone of the healthy participant decreased in all three muscles during REM sleep

Table 2.

Comparison of muscle activity across sleep stages

Group	Muscle	N2	N3	REM	Chi-square	p
Chronic WAD	Right trapezius	3.3 (2.7–3.7)	3.4 (2.9–3.5)	2.9 (2.6–3.3)	1.75	ns
	Left trapezius	4.5 (2.5–5)	4.1 (2.5–4.8)	4.1 (2.2–4.6)	5.25	ns
	Mentonian	0.9 (0.7–1.4)	0.8 (0.5–0.9)	0.4 (0.3–0.6)	9.25	0.01
Healthy control	Right trapezius	2.1 (1.9–3.6)	2.3 (2–3.6)	1.9 (1.6–3)	13	0.002
	Left trapezius	2.5 (2.1–3.4)	3.3 (2.5–4.4)	2 (1.6–3.1)	12.25	0.002
	Mentonian	0.9 (0.7–1.9)	0.9 (0.4–1.8)	0.3 (0.2–0.4)	12.25	0.002

Median (Quartile 1–Quartile 3), ns not significant. Values in microvolts

Fig. 2.

Comparison of mentonian muscle activity across sleep stages in both groups. The lines show significant differences at p < 0.017

Fig. 3.

Comparison of trapezius muscle activity across sleep stages in the healthy group. The lines show significant differences at p < 0.017

Discussion

During the first days of WAD presentation, the initiation and maintenance of sleep is difficult for patients mainly due to pain [10, 15] and other factors, such as the use of neck braces at night, contributing to a poor or unrefreshing sleep perception [20]. In this study, we found no significant differences in sleep latency and sleep efficiency, which is consistent with a study that monitored sleep through actigraphy in WAD patients [8]. In addition, we found that only the percentage of TST in the N1 and N2 sleep stages were significantly different between the two groups of women participants. The participants with WAD had a higher percentage of TST in the N2 stage and a lower percentage of TST in the N1 stage, while the percentage of TST in the N3 and REM stages did not differ.

While there are normative values for sleep stage proportions of TST, the proportions may differ depending on the sleep classification method [21]. In the AASM criteria, the sleep stage N1 occupies 6% to 18%, the stage N2 occupies 41% to 58%, the stage N3 occupies 8% to 26%, and REM sleep occupies 17% to 25% [21]. Although we found that the WAD and the healthy groups had significant differences in time spent in stage N1, the percentages of TST in this stage were in the normative range in both groups. The percentage of TST in stage N2 in the WAD group (61%) was higher than the normative values. Although the healthy group had a lower percentage of TST (16%) in REM sleep, there was no significant difference between the WAD group and the healthy group.

The higher percentage of TST in the N2 sleep stage observed in the WAD group may indicate a homeostatic adaptation in the presence of moderate-level chronic pain. Pain is known to disrupt sleep patterns, but it has been suggested that several mechanisms may act as guardians of sleep; some of these mechanisms include K-complexes, sleep spindles, and a cyclic alternating pattern occurring during the N2 sleep stage [22]. Since sleep microstructure was not evaluated in this study, future studies should evaluate the number of sleep spindles and K-complexes as well as the alternating cyclic pattern.

The correlations between the perceived level of pain and sleep variables obtained from this series of cases were non-significant, which contrast with the significant correlations previously found between sleep quality, sleep duration, and sleep efficiency evaluated by the Pittsburgh sleep quality index [10]. The lack of significant correlations corroborates previously reported findings, demonstrating that sleep efficiency and sleep duration, assessed with self-reports and actigraphy, are not consistent between patients with WAD and healthy individuals [8]. The absence of agreement within data obtained through self-reported and PSG (or actigraphy) has also occurred when analyzing the relationship between pain and sleep in patients with musculoskeletal pain like fibromyalgia, rheumatoid arthritis, juvenile arthritis, osteoarthritis, or temporomandibular disorder [23].

Whiplash is associated with dysfunction in the neck and shoulder muscles [1, 14, 16]. In the upper trapezius muscle, this dysfunction has been assessed with electromyography [11, 15], which we conducted during sleep in this study. The absence of muscle atonia in both trapezius muscles during REM sleep in the chronic WAD group is relevant, since muscular atonia is an outstanding characteristic of REM sleep and is disrupted in certain disorders such as REM sleep behavioral disorder [24]. In contrast, we observed a decrement of the muscle tone in both trapezius muscles in the healthy group. This result is consistent with a previous study that assessed muscle tone through qualitative analysis (visual scoring) of raw EMG signal and found that the muscle tone in the trapezius muscle was reduced in REM sleep [25].

No previous studies were found assessing trapezius muscle activity during sleep in individuals with WAD to compare to our findings. In one study, the activity of the trapezius muscle was evaluated during sleep in people with neck and shoulder pain without considering the sleep stages [26]; another study on idiopathic cervical dystonia evaluated the trapezius and sternocleidomastoid muscles only during non-REM sleep [27].

The muscle atonia during REM sleep is regularly generated by inhibitory impulses exerted by supraspinal regions on spinal motoneurons, including glycinergic premotor neurons that hyperpolarize motoneurons [17]. The absence of muscle atonia in the trapezius muscle during REM sleep may be caused by dysfunction in the motor control system associated with chronic WAD. Previous research has suggested that pain induces changes in the systems that regulate motor activity, and such changes can occur at multiple levels. Thus, chronic pain contributes to the reorganization and implementation of new motor strategies [28]. Although sleep is not characterized by the presence of movements, changes in the motor control system occur during sleep. For example, in the reticular response-reversal phenomenon, stimulation of the nucleus reticularis pontis oralis during wakefulness and non-REM sleep leads to motor activation, while stimulation of this nucleus during REM sleep leads to motor inhibition [29]. Likewise, it has been reported that the antinociceptive behavior of the tail-flick latency in cats can be increased by cholinergic stimulation of the medial pontine reticular formation, a region involved in the regulation of REM sleep [30].

Individuals with WAD exhibit different motor control strategies by changing the movement patterns of the cervical spine [31]. In the chronic WAD group in this study, the change in motor control manifested as an absence of atonia in the trapezius muscles during REM sleep. This absence of atonia in these muscles may reflect a change in motor control at the spinal cord level, where the balance between the monosynaptic stretch reflexes and the inverse myotatic reflexes has been affected due to the presence of pain, directly affecting the motoneurons and/or the premotor neurons. This spinal interneuron system is known to be actively involved in various aspects of motor adaptation [32]. In addition, glycinergic/GABAergic interneurons of the spinal ventral horn also participate in regulating REM sleep muscle atonia [33]. This functional interrelationship may explain the absence of atonia during REM sleep in the trapezius muscles of the WAD participants.

Since previous research has proposed that postural immobility during sleep in the lateral position causes the upper body to initiate pressure on the shoulder that may influence shoulder pain [34], we included EMG samples only in the supine position. However, the body position sensor was placed at the level of the middle chest, without considering hands’ positions or slight right or left head turns. Hands’ positions are relevant, since the position of the hands has been found to influence different EMG activity from the trapezius muscle in the supine position [35]. Therefore, future studies should more accurately consider body and limb position when assessing muscle activity.

Although a 5–70 Hz band has been used to assess the activity of neck, shoulder, trunk, and limb muscles during sleep [25], a main limitation of this study was using the 20–70 Hz frequency band to record muscle activity. Another limitation was the limited number of participants with chronic WAD. Therefore, future research should include a larger sample of participants as well as record EMG signals over a wider frequency band. In addition, other muscles that are not directly related to whiplash injury should be assessed in the future. Pain should also be assessed through standardized pain assessment tools. If possible, the range of motion or muscle contractions during wakefulness should additionally be recorded, which may be helpful in determining whether impairment in motor control exists.

Declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Ethical committee permission

Not applicable. This study is part of a doctoral thesis and was approved by the doctoral committee assigned by the Psychology Graduate Program. At that time, there was no ethics committee assigned to evaluate the projects of Psychology Graduate students.

Informed consent

Informed consent was obtained from all participants in the study.