Trapezius activity of fibromyalgia patients is enhanced in stressful situations, but is similar to healthy controls in a quiet naturalistic setting: a case-control study

Twenty-six female patients with FM and 25 age-matched (±3 years) female HCs participated in the study (Table 1). The patients were mainly recruited through the local FM association while HCs were recruited among donors to the hospital blood bank. Inclusion criteria were age between 35 and 64 years. Upon inclusion to the study, eligible FM patients underwent a clinical examination to verify the FM diagnosis as defined by the American College of Rheumatology[30]. Number of years since first symptoms and number of years since confirmed diagnosis were retrieved from each participant’s medical record. Patients were excluded if they had: a) cardiorespiratory, cerebrovascular, neurologic, neuromuscular, endocrine, infectious, metabolic, lung, or cancer disease, b) injury that affected function, c) connective tissue disorder, d) tendinitis or capsular affection of the shoulder joint, or e) high blood-pressure (i.e., systolic pressure >140 mmHg or diastolic pressure >90 mmHg) or were taking anti-hypertensive medication. Participants were also excluded if they were taking medication that may interact with neural, vascular, or muscular function or the physiological measurements to be performed (e.g., antidepressants, antiepileptics, β-blockers). FM patients that used analgesics and/or sleep medicine on a regular basis were instructed to cease medication two days prior to the experiment. The study protocol was approved by the Regional Committee for Ethics in medical research (project no. 4.2005.2728) and all participants signed an informed consent before inclusion. The study was carried out in the premises of a hotel used by outpatients, with easy access to university laboratories. It is part of a comprehensive study including sleep recordings and the collection of samples for cortisol and catecholamine analyses[31‐34].

The order and time schedule for data collection is shown in Figure 1. Participants met in the laboratory at around 4.45 pm. After mounting the electrophysiological recording equipment, a session consisting of isometric maximal voluntary contractions (MVCs) and four test conditions was performed (Figure 1): a) controlled arm movement test (see below), b) Laboratory relaxation: participants were comfortably seated in an arm chair and watched a cartoon movie for 30 min, c) mental stress: four 6-min periods alternating between the Stroop test[35] and an arithmetic test with backward counting (mean duration 28 min, range 26–29 min), and d) activation of the sympathetic system during standing: intrathoracic pressure was increased by maximal inspiration and breath holding with epiglottis closed for ~15 s[36, 37].

In the controlled arm movement test the participants moved either the dominant or non-dominant hand continuously, holding a pen, between three circles forming an equilateral triangle on a horizontal table[38]. The dominant hand moved in clockwise direction and the non-dominant hand in anti-clockwise direction. Circle diameter was 70 mm in the first trial and 4 mm in the second trial. Each trial lasted 2 min and a metronome paced the arm movement at 88 beats/min, i.e., participants were required to set a mark within the circles following the beat of the metronome. In Laboratory relaxation, participants were instructed to relax to the best of their ability. The experimenter that mounted the recording equipment and performed the tests was blinded to the diagnosis of the participants.

After the laboratory session the participants had an evening meal (bread, salad, fruits) at around 8.30 pm and were thereafter free to choose activity (e.g., reading, watching TV, playing solitaire with cards) but were instructed to stay inside the hotel. The participants filled in a short diary with description of evening activities and time period for each activity. Physiological responses were determined for “quiet seated activity” (reading or watching TV), for the evening meal, and for the 5-min period with lowest surface electromyographic (sEMG) activity of upper trapezius.

The participants scored pain level (shoulder/neck, low back), perceived general tension, and perceived stress on a VAS (0–100 mm) upon arrival in the laboratory, after Laboratory relaxation, after the mental stress test, and at bedtime (Figure 1). Pain intensity was scored after first indicating whether they at all felt pain (yes/no). All VAS scales were anchored by “very high” and “very low” at endpoints. Participants also scored their level of effort associated with the stress test and how stressful they perceived the test.

A portable recording system (Myomonitor IV, Delsys Inc, Boston MA) was used to record force, sEMG, a modified lead II electrocardiogram (ECG), and respiratory frequency by a strain gauge embedded in a flexible belt placed around the chest just below the sternum.

For the ECG recordings, the QRS complex was detected, and the R-R intervals were derived on a beat-by-beat basis to determine heart rate (HR) during the stress test, during Laboratory relaxation in the laboratory, and during the 5-min evening period with lowest trapezius sEMG. Variables indicating heart rate variability (HRV) were determined[39, 40]. Time domain measures included the standard deviation of the NN interval (SDNN) and the root mean square successive difference (RMSSD), based on 5-min periods of recordings. SDNN is thought to reflect total variability while RMSSD is mainly thought to reflect vagal modulation of heart activity. In case of the mental stress test and Laboratory relaxation in the laboratory, the second last 5-min period was selected. Frequency domain measures included the power of the low frequency component (LF; 0.04-0.15 Hz) and the high frequency component (HF; 0.15-0.4 Hz), as well as the ratio of the two components of the heart period power spectrum (LF/HF ratio). The recordings were visually inspected for ectopic heart beats and artifacts. Heart beats classified as non-normal beats were excluded from further analysis.

Bipolar sEMG (bar-shaped contact surfaces 1 by 10 mm, 10 mm spacing between surfaces) was recorded from the clavicular and acromial fibers of the dominant upper trapezius, middle deltoid, and biceps brachii. The midpoint of the acromial trapezius electrode was placed at a point 2 cm distal to the midpoint of a line from the spinous process of the 7^th cervical vertebra (C7) toward the lateral edge of the acromion[41, 42]. The clavicular electrode was placed in parallel to the acromial electrode, 2 cm in the ventral (clavicular) direction[25]. For the middle deltoid the electrode was placed on a line from the acromion to the lateral epicondyle of the elbow, corresponding to the greatest bulge of the muscle. For the biceps brachii the electrode was placed on the line between the medial acromion and the fossa cubit at 1/3 from the fossa cubit. All signals were sampled at 1 kHz. The sEMG signal was band-pass filtered at 10–450 Hz and root-mean-square (RMS) values were calculated using a 100 ms non-overlapping window. Three isometric MVCs were performed for each muscle with simultaneous recording of sEMG and force. The highest sEMG response was used to normalize the sEMG signal (% EMG_max). For each MVC, participants were instructed to develop maximal force within 1–2 s and thereafter hold the force for 3–5 s. A 1-min break was applied between each MVC. MVCs for trapezius and middle deltoid were performed with participants in an erect seated posture, with arms 90° abducted in the scapular plane. Resistance was applied just proximal to the elbow joint by adjustable straps connected to two strain-gauge force transducers secured to the floor. MVCs for the dominant biceps brachii were performed with participants in seated position, elbow flexed at 90°, and the elbow supported against the side of the body. Resistance was applied unilaterally to the dominant arm just proximal to the wrist joint by the adjustable strap that was connected to one of the force transducers. The force signal was low-pass filtered (10 Hz, Butterworth, 6^th order) and downsampled to 10 Hz before further analysis. Maximal force was determined as the median of the highest 0.5 s interval (i.e., median of 5 samples) for each MVC. Recordings of FM patients were inspected for indication of systematic force reduction with successive MVCs, but no such effect was found.

Table 2 presents maximal force and sEMG_max during MVCs. FM patients generated lower force than HC during shoulder abduction (not significant) and elbow flexion. This is an anticipated consequence of their patient status, but may represent a source of error if “true” force capacity of FM patients is masked by sub-maximal force generation in the calibration contractions due to, e.g., pain-associated inhibition. Correlation analyses showed no association of maximal force with indicators of present pain (pain in the laboratory, pain last 24 hrs, pain last week). Alternative sEMG calibrations were nevertheless explored by upward adjustment of patient sEMG_max values by 20% to compensate for group differences in maximal force. sEMG activity during the evening and in tests was quantified by median activity level and by rest time (sEMG amplitude <0.5% EMG_max)[43].

Two questionnaires were administered to assess subjective health complaints and personality traits. The Subjective Health Complaints inventory consists of 29 questions concerning somatic and psychological complaints last 30 days[44]. For each item, severity of the complaint is rated on a 4-point scale (0=none, 3=severe) and duration by number of days during the last 30 days. The Karolinska Scales of Personality (KSP) is a self-rating questionnaire constructed to measure stable personality traits[45]. KSP consists of 135 items grouped into 15 subscales. Three subscales were used in this study: muscular tension, psychic anxiety, and somatic anxiety. Each item is scored on a 4-point Likert scale (0=does not apply at all, 3=applies completely).

Four indexes of potential relevance to influence physiological responses[24] were constructed from items in the two questionnaires: 1) musculoskeletal symptoms (pain in neck, shoulders and upper back, perceived stiffness in upper and lower body), 2) non-musculoskeletal symptoms (irregular HR, hot flushes, headache, abdominal pain, dizziness), 3) cognitive/psychological symptoms (anxiety, forgetfulness, depression), 4) muscle tension (tension in jaw muscles, feeling of tenseness when trying to sleep, difficulty in relaxing). The index scores (average of the included items) were considered long-term traits, to be distinguished from VAS scores of pain, stress, and perceived general tension on the experimental day[27].

The Kolmogorov-Smirnov test was used to explore normality of dependent variables. All HR variables, all symptom indexes, most force variables (three of four), and most sEMG_max variables (six of eight) were normally distributed. None of the sEMG variables or VAS-score variables was normally distributed. The independent samples t-test was used to test differences between groups for normally distributed data while the Mann–Whitney U test was used to test group differences for non-normally distributed data. A mixed design repeated measures ANOVA was used to test differences between repeated recordings of HR and HRV. If Mauchly’s test indicated violation of sphericity, a Greenhouse-Geisser correction was applied[46]. Friedman’s ANOVA was used to test differences between repeated recordings of sEMG and VAS-score variables. Wilcoxon signed-rank test with a Bonferroni correction was used for post-hoc comparisons. Linear regression coefficients and Spearman’s ρ were used for correlation analyses. Significance level was set to p<0.05.

Figure 2 shows sEMG responses in the laboratory tests. Clavicular trapezius sEMG activity was significantly higher for FM patients than HCs in sustained inspiration with breath holding (A). Both trapezius responses were markedly higher for FM patients vs. HCs in the mental stress test (C) and in Laboratory relaxation (D). sEMG rest time was correspondingly lower for FM patients (e.g., mental stress test: clavicular trapezius: 14 vs. 75%, p<0.011; acromial trapezius: 14 vs. 76%, p<0.001). Muscle activity during Laboratory relaxation was unchanged, comparing the first and last 5-min interval of the 30-min observation period. Trapezius sEMG activity of FM patients and HCs was similar in the test with dynamic arm movement (B). Statistically significant, but only moderately higher deltoid and biceps responses were observed for FM patients in several of the tests with higher trapezius responses. Downward adjustment of FM patient sEMG responses by 20% to compensate for putative submaximal effort in the calibration contractions did not alter the statistically significant differences of trapezius responses, comparing FM patients and HCs.

Figure 3 shows sEMG activity for quiet seated activity during the evening (A), for the evening meal (B), and for the 5-min period with lowest upper trapezius sEMG activity (C; always occurring during quiet seated activity). FM patients were not distinguished from HCs in quiet seated activity and lowest 5-min period, but showed higher trapezius and biceps activity during the evening meal. sEMG rest time during the evening did not distinguish FM patients from HCs (clavicular trapezius: 20 vs. 23%, acromial trapezius 30 vs. 23%; data not shown).

Table 3 presents HR responses. Repeated measures ANOVA showed a significant main effect of condition (F=41.1, df=1.7, p<0.001), but no main group effect (F=1.3, df=1, p=0.25), and no interaction between group and condition (F=2.3, df=1.7, p=0.12). Paired statistics showed elevated HR of FM patients in Laboratory relaxation relative to lowest 5-min during the evening (71 vs. 66 bpm), and further elevation in the mental stress test (76 bpm; Table 3). HR of HCs was unchanged in Laboratory relaxation vs. lowest 5-min period during the evening (67 vs. 65 bpm), and was elevated to the same level as for FM patients in the mental stress test (both groups at 76 bpm).

HRV analysis showed reduced variability for FM patients by the time domain variables SDNN (main group effect: F=8.51, df=1, p=0.006; no main effect of condition, no interaction between group and condition) and RMSSD (main group effect: F=5.05, df=1, p=0.03; main effect of condition: F=5.02, df=1.74, p=0.012; no interaction between group and condition). A non-significant tendency of elevated low-frequency HRV of FM patients in the stress test using time domain analysis (i.e., lower RMSSD) was very clear by frequency domain analysis (HF main condition effect: F=20,82, df=1.54, p<0.001; LF main condition effect: F=16.69, df=1.63, p<0.001; LF/HF ratio: F=13.43, df=1.68, p<0.001). Conversely, the frequency domain analysis did not show lower HRV for FM patients in the stress test.

Shoulder/neck pain of FM patients was relatively high upon arrival in the laboratory, with a non-significant reduction after Laboratory relaxation (Table 4). The stress test provoked a statistically significant increase in shoulder/neck pain that stayed at the same elevated level till bedtime (38%, CI: 25–45). Low back pain of FM patients was lower than shoulder/neck pain (significant following Laboratory relaxation), and was unchanged following the stress test. Perceived general tension was higher for FM patients than HCs upon arrival and after Laboratory relaxation, and was elevated and equally high for the two groups following the stress test. FM patients and HCs scored similarly on perceived stressfulness of their day, and effort invested in the stress test.

Most items used to construct the indexes of musculoskeletal symptoms; muscle tension symptoms, non-musculoskeletal symptoms, and cognitive and psychological symptoms (cf. Methods) were scored significantly higher by FM patients vs. HC. The only significant correlation found between index scores and sEMG activity or HR was a negative correlation between musculoskeletal symptoms and clavicular trapezius activity for FM patients during the stress test (R=−.58, p=0.002).