Journal of Psychosomatic Research 72 (2012) 51–57

Contents lists available at SciVerse ScienceDirect

Journal of Psychosomatic Research

Catecholamines and heart rate in female fibromyalgia patients☆

Roberto Riva a,⁎, Paul Jarle Mork b, Rolf Harald Westgaard c, Tonje Okkenhaug Johansen d, Ulf Lundberg e,f

a Department of Psychology, Stockholm University, Swedenb Department of Human Movement Science, Norwegian University of Science and Technology, Trondheim, Norwayc Department of Industrial Economics and Technology Management, Norwegian University of Science and Technology, Trondheim, Norwayd Department of Neuroscience, Norwegian University of Science and Technology, Trondheim, Norwaye Department of Psychology, Stockholm University, Stockholm, Swedenf CHESS (Centre for Health Equity Studies), Stockholm University, Stockholm, Sweden

☆ The study was performed at the Department of Humgian University of Science and Technology, Trondheim, Ncatecholamines and the statistical analyses for the presenDepartment of Psychology, Stockholm University.⁎ Corresponding author at: Stockholm University, Dep

Stockholm, Sweden. Tel.: +46 8 163892; fax: +46 8 15E-mail address: roberto.riva@psychology.su.se (R. Ri

0022-3999/$ – see front matter © 2011 Elsevier Inc. Alldoi:10.1016/j.jpsychores.2011.09.010

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 19 January 2011Received in revised form 12 September 2011Accepted 27 September 2011

Keywords:Chronic musculoskeletal painStressNoradrenalineAdrenalineDopamineAutonomic imbalance

Background: Fibromyalgia syndrome is a disease of unknown pathogenesis characterised by widespreadchronic musculoskeletal pain. Fibromyalgia has been associated with dysregulation of the stress systems,but results are inconsistent.Purpose: To investigate autonomic nervous system activity (urinary noradrenaline, adrenaline, dopamine,and heart rate) of fibromyalgia patients and healthy controls.Methods: Urinary catecholamines and heart rate were assessed for a 24-hour period in a controlled hospitalsetting (including relaxation, a test with prolonged mental stress, and sleep), and during daily activity in 29female fibromyalgia patients and 29 age-matched female healthy controls.Results: With repeated measures ANOVAs, catecholamine levels were lower in patients than controls(P=.035 for noradrenaline; P=.005 for adrenaline; P=.001 for dopamine). One-way ANOVAs for the singleperiods showed that patients compared to controls had significantly lower adrenaline levels during the night

(P=.010) and the second day (P=.010), significantly lower dopamine levels during the first day (P=.008),the night (P=.001), and the second day (P=.004). However, single time point noradrenaline levels were notsignificantly different between the groups. Overall, heart rate was significantly higher in patients than con-trols (P=.014). Specifically, significant differences emerged during relaxation (P=.016) and sleep(P=.011), but not during stress provocation or daily activities.Conclusions: The results indicate an altered regulation of the autonomic nervous system in fibromyalgia pa-tients, with attenuated activity of both the sympathetic (adrenal medulla component) and the parasympa-thetic branch.

© 2011 Elsevier Inc. All rights reserved.

Introduction

Fibromyalgia syndrome is characterised by prolonged widespreadmusculoskeletal pain and multiple tender points [1]. Approximately3% of adults are affected, with a female-to-male ratio of about 5:1[2]. The incidence increases with age and peaks between 30 and50 years [3]. Comorbid symptoms commonly associated with fibro-myalgia include poor sleep quality, morning stiffness, fatigue, anxiety,depression, and psychosocial stress [1,4].

According to the allostatic load model, the pathogenesis of psy-chosomatic diseases is associated with the physiological responses

an Movement Science, Norwe-orway. The analysis of urinaryt paper were conducted at the

artment of Psychology; 106 919342.va).

rights reserved.

to psychosocial stress [5]. Studies indicate that women with fibromy-algia are characterised by abnormalities in the major stress systems,the hypothalamic–pituitary–adrenal axis, with reduced levels of cor-tisol [6–9], and the sympathetic and the parasympathetic branchesof the autonomic nervous system [10–14]. However, findings are in-consistent [15].

Assessment of catecholamines' release by the sympathetic ner-vous system is a commonly used method to investigate the regulationof sympathetic activity. Three primary catecholamines have beenidentified operating in the two main components of the sympatheticnervous system (the sympathoneural and the adrenomedullary),namely noradrenaline, adrenaline, and dopamine [13,16]. Previousresearch shows conflicting findings as regards catecholamine levels inpatients with fibromyalgia during both basal and stressful conditions.Several studies show no differences in basal levels of noradrenalineand adrenaline in blood [17–21], or in urine [19,22], and in plasma nor-adrenaline and adrenaline levels after a reaction-time stress provoca-tion between fibromyalgia patients and healthy controls [17]. Onestudy found normal levels of plasma noradrenaline and adrenaline

Table 1Descriptive statistics for demographic and other background variables of fibromyalgiapatients (N=29) and healthy controls (N=29).

Patients Controls

Mean (SD) Mean (SD)

Age (years) 52.1 (8.9) 52.7 (8.4)Body mass index (kg/m2) 27.1 (5.9) 25.0 (3.5)No. of tender points 15.7 (2.2) Not addressedYears since diagnosis 5.5 (6.0) Not addressedYears since first symptoms 13.1 (8.6) Not addressed

% (N) % (N)Employment fraction ≥50%* 24 (7) 90 (26)Smokers 24 (7) 21 (6)Exercise sessions per week

Once 7 (2) 10 (3)1–3 times 69 (20) 59 (17)N3 times 24 (7) 31 (9)

*Patients significantly differ from controls (Pb .001).

52 R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

after a postural challenge test, role-play simulating a situation of verbalaggression and ischemic pain stimulation [23], but low basal noradren-aline levels in fibromyalgia patients [23]. However, another studyshowed high noradrenaline levels in fibromyalgia patients both at base-line and after interleukin-6 injections, together with normal adrena-line levels [24]. Also, submaximal exercise, which mainly inducesa sympathoneural response, has been found to elicit attenuated nor-adrenaline and adrenaline secretion in blood in fibromyalgia patients[21]. Interestingly, other studies show normal noradrenaline levels,but attenuated adrenaline levels in the blood in fibromyalgia patientsduring standardised isometric contraction [20,25], and after hypoglyce-mia induction [18]. Finally, dopamine levels in blood or urine have beeninvestigated in a few studies showing normal values during daily activ-ities in fibromyalgia patients [19,24]. These inconsistencies may be dueto different types of stress provocation, i.e., mental or physical stress,which may have different physiological effects [17,18,23,24], or toparticipants being examined during or after a physical exercise ses-sion [21].

In addition to catecholamines, heart rate and particularly restingheart rate are considered indicators of autonomic balance [11,26].Few studies have investigated heart rate and catecholamines togetherin fibromyalgia patients, and the existing findings are inconsistent.One study showed increased sympathoneural activity, with high nor-adrenaline and heart rate, in fibromyalgia patients [24], while otherstudies reported an attenuated sympathoadrenal response, with lowadrenaline levels and heart rate, in fibromyalgia patients during phys-ical exercise [21], muscular contraction [20,25], and stress provoca-tion [17,23].

The aim of the present study was to compare autonomic nervoussystem activity in long-term female fibromyalgia patients and healthycontrols. This study combines 24-hour measurements of urinarycatecholamines and heart rate, including two experimental conditions(relaxation period andmental stress provocation), during the night andduring unconstrained daily activities. Moreover, participants providedsubjective ratings of pain in neck, shoulders and low back during thecorresponding timeperiods. The present studywas performed in a care-fully controlled hospital setting, inwhich patients and controls had sim-ilar activities, thus allowing a reliable comparison of autonomic activity.On the basis of themutual interaction between themajor stress systems[27] and on previous findings of attenuated hypothalamic–pituitary–adrenal axis activity in the same study sample [8], we hypothesisedthat fibromyalgia patients would show an altered activity of the auto-nomic nervous system compared with healthy controls.

Methods

Participants

Twenty-nine female patients with fibromyalgia and 29 age-matched (±3 years) female healthy controls took part in the study(Table 1). The patients were mainly recruited through the local fibro-myalgia association in Trondheim, Norway. The controls wererecruited among donors to the hospital blood bank. Inclusion criteriawere age between 35 and 67 years. Upon inclusion in the study, eligi-ble fibromyalgia patients underwent a clinical examination to verifythe fibromyalgia syndrome diagnosis as defined by the AmericanCollege of Rheumatology [1]. Number of years since the first symp-tom and number of years with confirmed diagnosis were retrievedfrom each participant's medical record (Table 1). Participantswere excluded if they had: a) cardiorespiratory, cerebrovascular,neurologic, neuromuscular, endocrine, infectious, metabolic, lung,or cancer disease, b) injury that affected function, c) connective tis-sue disorder, d) tendinitis or capsular affection of the shoulder joint,or e) high blood pressure (i.e., systolic pressure N140 mmHgor diastol-ic pressure N90 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 thephysiological measurements to be performed (e.g., antidepressants,antiepileptics, β-blockers). After the clinical examination, 11 eligi-ble fibromyalgia patients were excluded because they did not fulfilthe criteria. Fibromyalgia patients using analgesics and/or sleepmedicine on a regular basis were instructed to cease medication2 days prior to the experiment. All participants volunteered forthe study but they were compensated for travel costs and otherexpenses.

The study protocol was approved by the Regional Ethics Commit-tees in Norway (project no. 4.2005.2728) and Sweden (Dnr 2006/87-31/1). The study was carried out according to the Declaration ofHelsinki.

Procedure

Fig. 1 shows the order and time schedule of the data collection.Participants had dinner at the hospital-hotel next to the laboratorybefore meeting in the laboratory at around 4:45 p.m. After mountingthe electrophysiological recording equipment, a series of experimen-tal conditions was performed. First, participants were comfortablyseated in an armchair and watched a cartoonmovie for 30 min (relax-ation period). A previous study found that watching cartoons is ac-companied by few changes in bodily sensation and only involvestemporary heart rate decrease [28]. Second, participants underwentstress provocation consisting of four six-minute periods alternatingbetween the Stroop test and an arithmetic test with backward counting(mean duration 28 min, range 26–29 min), which provided high stressexposure [29]. Details of the experimental procedure have beenpublished elsewhere [8]. The experimenter performing the testswas blinded to the participants' diagnosis. After the laboratory ses-sion, all participants had an evening meal (bread, salad, fruit) ataround 8:30 p.m. and were then free to choose activities (e.g., read-ing, watching TV, playing cards), but were instructed to stay insidethe hospital-hotel. Before going to sleep in a comfortable room atthe hospital-hotel (mean time 11:33±0:29 p.m.) they completeda set of questionnaires including the Subjective Health Complaintsinventory, the Karolinska Scales of Personality, the Perceived StressScale, and the subscale Neuroticism of the Eysenck PersonalityQuestionnaire (for details see [8]). Upon awakening (mean time6:38±0:25 a.m.), the participants responded to a short question-naire on sleep quality and were then free to go home or to work.

Urinary catecholamines

Urine samples were collected over a 24-hour period to assesslevels of noradrenaline, adrenaline, and dopamine, and were divided

Fig. 1. Order and time schedule of the data collection. Urine samples were collected throughout the experimental session. For the statistical analysis the urine samples were dividedinto three periods: “Day1”,“Night”, and “Day2”. Recording equipment was mounted immediately after participants' arrival at the hospital (4:45 p.m., range 4:30–5:00 p.m.) fol-lowed by an experimental session with laboratory recordings. After an evening meal, the participants were free to choose activities (e.g., reading, watching TV, playing cards)until bedtime. One hour after awakening, participants were equipped with a heart rate monitor and a step counter before going home or to work.

53R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

into three periods for the statistical analysis (Fig. 1). The first peri-od comprises catecholamine levels from arriving at the hospital,when participants were asked to empty their bladders (meantimes 4:34±0:15 p.m. for controls and 4:43±0:12 p.m. for fibromy-algia patients; significantly different; P=.032), until the end of thestress provocation, when participants were asked to provide a urinesample (mean times 7:59±0:21 p.m. for controls and 8:01±0:17 p.m. for fibromyalgia patients; not significantly different;P=.354); This measure is labelled “Day1” in Fig. 1. The second peri-od includes catecholamine levels from the end of the stress provocationuntil waking up on the second day, when all participants were asked toprovide a urine sample (mean times 6:39±0:36 a.m. for controls and6:41±0:27 a.m. for fibromyalgia patients; not significantly different;P=.761); this measure is labelled “Night” in Fig. 1. The third period in-cludes catecholamine levels fromwaking up on the second day until thelast urine sample in the afternoon (mean times 2:09±1:51 p.m. forcontrols and 3:51±1:06 p.m. for fibromyalgia patients; significantlydifferent; Pb .001). This final measure is labelled “Day2” in Fig. 1.

To determine the excretion of urinary catecholamines, the volumeof urine was measured and its pH adjusted to 3.0 with 6 M HC1. Then,20 ml of each sample was frozen (−18 °C) until being analysed fornoradrenaline, adrenaline and dopamine by high-pressure liquidchromatography. Finally, catecholamine concentration was multi-plied by volume and divided by time, and expressed in terms ofpicomol/min.

Heart rate

To measure heart rate, a portable recording system (MyomonitorIV, Delsys Inc., Boston MA) was used to record a modified lead II elec-trocardiogram from the beginning of the laboratory session until 1 hafter the participants woke up the next morning. The QRS complexwas detected, and the intervals between the R peaks (R–R intervals)were derived on a beat-by-beat basis. Data were corrected for arte-facts and heartbeats with inter-beat interval ≤0.4 s were omittedfrom further analysis. Finally, a mean value of the heart rate was cal-culated for each participant during the relaxation period, the stressprovocation, and the sleep period.

After the electrocardiographic recording equipment was removedupon awakening the second day, participants were equipped with aPolar heart rate monitor S610. The mean R–R intervals of every five-second period were used to obtain a heart rate expressed as beatsper minute. Finally, a mean heart rate value during daily activitywas calculated for each participant. Additionally, a Silva step counter

was used to record the number of steps during the second day, as anindicator of physical activity.

Self-ratings

On arrival, after relaxation, after the mental stress provocationon the first day, and upon awakening on the second day, all partic-ipants were asked to rate their levels of general tension, pain(neck/shoulder, low back), and mental fatigue on a visual analoguescale (VAS, 0–100 mm) and physical fatigue on Borg's scale [30].Right before bedtime the participants were asked to fill in a shortdiary to describe their activities during the evening (after the labo-ratory session) and the number of cigarettes smoked and cups ofcoffee consumed during the day.

Statistical analysis

A repeated measures mixed ANOVA, using Greenhouse–Geissercorrection of degrees of freedom if the sphericity assumption was vi-olated, was conducted to investigate differences between patientsand controls. Periods of measurements for the catecholamine levelsand for the heart rate served as within-subject factor. The between-subject factor was the group (fibromyalgia patients vs. healthy con-trols). Additionally, separate one-way ANOVAs were conducted foreach single time point for noradrenaline, adrenaline, and dopaminelevels, heart rate values, and self-reported pain and fatigue, withgroups (fibromyalgia patients vs. healthy controls) as between-subjects factor. The significance level (alpha) was set to Pb0.05.

Pearson's r and Spearman's ρ values were used to explore possiblecorrelations between variables excluding cases list-wise. Correlationswere calculated separately for fibromyalgia patients and healthy con-trols when there was a significant difference between groups, whilein the other cases correlations were calculated including the wholesample (N=58). The two methods showed the same results andonly correlations based on the Pearson's r are reported. Variablesknown to influence catecholamines secretion, including age, bodymass index, number of cigarettes smoked, coffee intake, total amountof sleep, sleep efficiency (i.e., total amount of sleep divided by totaltime in bed), sampling time, and physical activity (i.e., average num-ber of exercise sessions during a week, and number of steps duringthe second day) [31,32], were introduced as covariates in the ANOVAsif these variables correlated with the dependent and the independentvariables.

Fig. 2. Bar plot showing the mean values in fibromyalgia patients (N=29) and healthycontrols (N=29) during the three periods of urinary measurements (“Day1”, “Night”,and “Day2”) for the levels of noradrenaline (A), adrenaline (B), and dopamine (C).Error bars represent 95% CI. *Pb .05, **Pb .01.

54 R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

Results

Catecholamine levels

Catecholamine levels were generally lower in fibromyalgia patients than inhealthy controls. For the noradrenaline levels, a repeated measures ANOVA showed asignificant effect of group (F(1,49)=4.70; P=.035), of period (F(2,98)=135.02;Pb .001), but no interaction effect (F(2,98)=2.26; P=.118). For the adrenaline levels,there was a significant effect of group (F(1,47)=8.81; P=.005), of period (F(2,94)=83.84; Pb .001), but no interaction effect (F(2,94)=2.11; P=.127). For the dopa-mine levels, there was a significant effect of group (F(1,49)=13.21; P=.001), of peri-od (F(2,98)=34.55; Pb .001), but no interaction effect (F(2,98)=1.49; P=.231).Separate one-way ANOVAs on the single periods showed significant differences dur-ing the night and the second day for adrenaline, during all the time periods for do-pamine, whereas the differences for noradrenaline failed to reach significance(Fig. 2). Table 2 shows means, standard deviations, confidence intervals, and Fvalues for the catecholamine levels during the three periods of measurements.

Heart rate

Mean heart rate was higher in fibromyalgia patients than in healthy controls. A re-peated measures ANOVA showed a main effect of group (F(1,35)=6.76; P=.014), ofperiod (F(3,105)=64.43; Pb .001), but no interaction effect (F(3,105)=1.25;P=.295). Separate one-way ANOVAs on the single periods showed significant differ-ences during the relaxation period and during sleep, but not during stress provocationor during daily activities (Fig. 3). In addition, a mixed between-within subjects ANOVAshowed a significant difference between the relaxation period and the stress provoca-tion (F(1,56)=60.83; Pb .001), no significant difference between patients and controls(F(1,56)=2.77; P=.101), and a significant interaction effect (F(1,56)=7.81;P=.007). Table 3 shows means, standard deviations, confidence intervals, and F valuesfor heart rate during the four periods of measurements.

Self-ratings

Patients with fibromyalgia had significantly higher subjective ratings of pain in theneck/shoulders and low back, of general tension, and mental and physical fatigue thandid the healthy controls (Table 4). Among fibromyalgia patients significant negativecorrelations were found between noradrenaline-day2 and low back pain during thelast 6 months (r=−.49; N=24; P=.016), between adrenaline-day2 and physical fa-tigue (r=−.42; N=24; P=.041), between heart rate during sleep and low back painduring the last 6 months (r=−.49; N=21; P=.025), and a significant positive corre-lation between heart rate during relaxation and low back pain during the last 24 h(r=.50; N=21; P=.022). In addition, among healthy controls significant negativecorrelations were found between noradrenaline-day2 and low back pain during thelast 6 months (r=−.42; N=23; P=.044), between adrenaline-day1 and general ten-sion (r=−.45; N=23; P=.032), and between heart rate during stress provocationand physical fatigue (r=−.56; N=16; p=.024).

Potential confounders

With an independent-samples t test, there was no difference between patients andcontrols for the variables that may influence the catecholamine secretion, such as age(t(56)=.26; P=.796), body mass index (t(55)=−1.554; P=.127), number of ciga-rettes smoked (t(56)=0; P=1), coffee intake (t(56)=.22; P=.830), average numberof exercise sessions during a week (t(56)=.23; P=.822), and number of steps duringthe second day (t(47)=−.118; P=.907). The only variables that showed a significantdifference between patients and controls were the total amount of sleep (t(46)=3.80;Pb .001), and the sleep efficiency (t(46)=3.16: P=.003), with lower values for the fi-bromyalgia patients. The correlations for this variable were calculated separately forthe two groups. Among healthy controls, a significant positive correlation was foundfor total amount of sleep versus dopamine-night (r=.65; N=17; P=.005), but no sig-nificant correlations were found with heart rate. After adjusting for the total amount ofsleep, the difference between fibromyalgia patients and healthy controls remained sig-nificant (Pb .05).

Discussion

In the present study, we found lower urinary catecholamine levelsin fibromyalgia patients than in healthy controls with significant dif-ferences for adrenaline and dopamine levels, but not for noradrena-line levels. The present study showed elevated resting levels ofheart rate in fibromyalgia patients compared with healthy controls.Significant differences were found during the relaxation period inthe laboratory and during sleep, but not during the stress provocationor daily activities on the second day. Finally, in line with expectations,fibromyalgia patients had significantly higher self-reported pain andtension as well as mental and physical fatigue than did healthy

controls and, as reported previously, more psychological problems[4,8,18,23]. In the present study, self-ratings (particularly back painand physical fatigue) correlated with noradrenaline and adrenalinelevels, and with heart rate.

The results of the current study indicate an altered regulation ofthe autonomic nervous system activity in fibromyalgia patients. Thepatients' high resting heart rate may suggest that fibromyalgia pa-tients have an attenuated parasympathetic activity with low heartrate variability [26]. The low heart rate variability has also beenreported in a forthcoming paper of our group for the fibromyalgia pa-tients from the same study sample [33]. However, the heart rate in fi-bromyalgia patients was similar to that of controls during mental

Table 2Mean (M), Standard Deviation (SD), 95% Confidence Interval (CI), F value, and significance level for the ANOVAs on the urinary catecholamines (noradrenaline, adrenaline, dopa-mine) at the three different points of the experimental session. Comparisons between fibromyalgia patients (N=29) and healthy controls (N=29). Values are expressed inpmol/min.

Catecholamines Patients Controls F value

M SD CI M SD CI

Noradrenaline: Day1 103.3 47.0 84.7–121.9 113.7 46.8 95.2–132.2 .66Night 28.4 14.5 22.8–34.0 36.6 23.7 27.5–45.6 2.45Day2 109.2 44.5 92.0–126.5 133.2 45.6 114.8–151.6 3.81

Adrenaline: Day1 17.2 11.0 12.9–21.6 22.1 9.4 18.4–25.8 3.04Night 3.1 1.9 2.3–3.8 5.1 3.4 3.8–6.4 7.16 ⁎

Day2 14.5 7.0 10.9–18.0 21.2 9.3 17.4–25.0 7.09 ⁎

Dopamine: Day1 660.3 409.0 498.5–822.1 1028 565.6 804.2–1251.7 7.49 ⁎⁎

Night 328.1 193.7 253.0–403.2 584.2 319.5 460.3–708.1 13.16 ⁎⁎

Day2 749.5 404.9 595.5–903.5 1183.2 635.2 926.6–1439.7 9.31 ⁎⁎

⁎ Pb .05.⁎⁎ Pb .01.

55R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

stress provocation and during daily activities. Additionally, partici-pants' higher heart rate on the second day compared with heartrate on the first day may be explained by differences in participants'physical activity between the 2 days. On the first day, participantsunderwent the stress provocation in a seated position, while duringdaily activities on the second day they were more physically active.Interestingly and in line with previous reviews [13], patients' lowlevels of adrenaline and dopamine during sleep and on the secondday suggest an attenuated activity of the adrenal medullary compo-nent of the sympathetic nervous system in fibromyalgia patients. Pre-vious research has shown that the synthesis and release ofnoradrenaline are constantly stimulated under chronic stress condi-tions [34]. Thus, a lack of sufficient synthesis of the neurotransmitterleads to a depletion of noradrenaline storage vesicles and preventstheir refilling [34]. Furthermore, an altered sensitivity of the cardiacβ-adrenoceptors or a reduction in the parasympathetic tone to theheart may occur in fibromyalgia patients in response to stimulationof the stress system [24]. Among the fibromyalgia patients studiedhere, noradrenaline levels did not differ significantly from those ofhealthy controls, but the lower adrenaline levels and the elevatedheart rate during rest may follow the pattern of altered cardiac sensi-tivity of the β-adrenoceptors previously described in fibromyalgia pa-tients [24], and in other diseases, such as diabetes and uraemia [35].Previous findings on the catecholamine levels show inconsistent re-sults. Among the studies that have examined urinary catecholaminesin fibromyalgia patients and healthy controls, one has shown normal24-hour adrenaline, noradrenaline and dopamine levels during dailyactivities, but responses to stress provocation were not measured

Fig. 3. Bar plot showing the mean heart rate values in fibromyalgia patients (N=29)and healthy controls (N=29) during the different periods of measurements (relaxa-tion, stress provocation, sleep, and day 2). Error bars represent 95% CI. *Pb .05.

[19]. Additionally, normal catecholamine values have been found,but only when collecting 12-hour overnight urine samples [22].

The differences in the catecholamine levels between the twogroups found in the present study might also be explained by differ-ences in work-related stress exposure (i.e., employment fractionwas higher in healthy controls than in fibromyalgia patients). However,work-related stress exposure could have influenced the catecholamineexcretion only during the second day, because during the first day andnight fibromyalgia patients and healthy controls were tested in thesame well-controlled hospital environment. In addition, on the secondday fibromyalgia patients provided their last urine sample later thanthe healthy controls. Because of the diurnal variation in catecholaminelevels, this difference might have contributed to the lower catechol-amine levels among fibromyalgia patients compared with healthy con-trols on the second day, but no differences in time of urine samplingoccurred after the stress provocation and upon awakening. Moreover,the autonomic nervous system is also influenced by genetic and envi-ronmental factors, and the interplay between such factors [36]. Differ-ences in catecholamine levels between groups might be explained bya genetic vulnerability in fibromyalgia patients linked to the expressionof catechol-O-methyltransferase (COMT) [36], an enzyme responsiblefor the degradation of the catecholamine neurotransmitters [37]. Fi-nally, it is still unclear whether the altered activity of the stress sys-tems in fibromyalgia patients results from decreased sympatheticactivity, decreased parasympathetic activity, or a combination ofboth, as pointed out by other authors in their neurovisceral integrationmodel [38]. Additional studies investigating heart rate variability in fi-bromyalgia patients may provide more detailed explanations.

The present study included an analysis of relevant confounders.Release of catecholamines, especially noradrenaline, and heart ratecan be influenced by physical demands and body posture [39,40].Thus, the higher heart rate in fibromyalgia patients compared to con-trols might also be explained by differences in nocturnal activities, asshowed in another study [41]. Furthermore, catecholamine secretion

Table 3Mean (M), Standard Deviation (SD), 95% Confidence Interval (CI), F value, and signifi-cance level for the ANOVAs on the heart rate during different periods on two days ofmeasurement. Comparisons between fibromyalgia patients (N=29) and healthy con-trols (N=29). Values are expressed in beats per minute.

Heart Rate Patients Controls F value

M SD CI M SD CI

Relaxation (TV) 71.9 8.1 68.9–75.1 66.7 8.2 63.6–69.8 6.13 ⁎

Stress provocation 75.8 8.0 72.8–78.8 74.6 6.9 72.1–77.4 .29Sleep 67.8 6.4 65.2–70.5 62.4 7.8 59.2–65.6 7.09 ⁎

Day2 88.2 9.8 84.1–92.4 82.8 10.6 77.7–87.9 3.03

⁎ Pb .05.

Table 4Mean (M), Standard Deviation (SD), 95% Confidence Interval (CI), F value, and signifi-cance level for the ANOVA on the self-ratings. Comparisons between fibromyalgia pa-tients (N=29) and healthy controls (N=29). All variables are expressed on a visualanalogue scale (VAS, 0–100 mm), except the variable physical fatigue expressed on aBorg's scale [30].

Self-ratings Patients Controls F value

M SD CI M SD CI

Neck/shoulders pain last24 h

45.1 19.7 37.7–52.6

7.0 10.5 3.1–11.0

84.84 ⁎⁎⁎

Neck/shoulders pain lastweek

50.2 15.9 44.1–56.2

7.8 9.7 4.1–11.5

150.48 ⁎⁎⁎

Neck/shoulders pain last6 months

67.3 15.0 61.6–73.0

13.5 16.4 7.3–19.7

170.64 ⁎⁎⁎

Low back pain last 24 h 24.1 24.3 14.9–33.4

7.1 10.9 3.0–11.3

11.84 ⁎⁎

Low back pain last week 32.2 24.9 22.8–41.7

5.7 9.0 2.2–9.1

29.21 ⁎⁎⁎

Low back pain last6 months

48.2 28.2 37.5–59.0

17.2 25.5 7.5–26.9

19.38 ⁎⁎⁎

General tension 32.8 22.1 24.4–41.3

16.1 17.2 9.6–22.7

10.27 ⁎⁎

Mental fatigue 36.1 28.8 25.2–47.1

20.7 16.9 14.3–27.1

6.20 ⁎

Physical fatigue 27.2 21.1 19.2–35.3

13.6 12.1 9.0–18.2

9.09 ⁎⁎

⁎⁎⁎ Pb .001.⁎⁎ Pb .01.⁎ Pb .05.

56 R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

in fibromyalgia patients can be influenced by sleep disturbances [14]and obesity [42]. However, in the present study, neither the level ofphysical activity (i.e., average number of exercise sessions during aweek, and number of steps during the second day), the amount ofsleep and sleep efficiency, nor body mass index influenced the results.Additionally, we performed an analysis of the catecholamine valuesexcluding participants with a body mass index that exceeded the obe-sity threshold (i.e., above 30 kg/m2). Comparing patients and controlswith separate ANOVAs on the single periods, noradrenaline levels onthe second day became significantly different (Pb .05), whereasadrenaline levels during the night were not significantly differentanymore. However, the results did not change for the other single pe-riods for the catecholamine levels. Moreover, no changes in the re-sults also analysing catecholamine levels with repeated measureANOVAs (data not presented), suggesting only a marginal influenceof body weight on catecholamine secretion in fibromyalgia patients.Finally, we also controlled for other relevant confounders such asage, smoking habits, coffee intake, and sampling time, but none ofthese variables influenced the results.

Several studies have investigated plasma catecholamines in fibro-myalgia patients, but, to date, this is the first study investigating 24-hour urinary catecholamine levels together with heart rate duringmental stress provocation, sleep, and normal daily activities in awell-controlled hospital-hotel environment. Urine sampling is con-sidered a non-invasive adequate method for assessing autonomicnervous system activity and the secretion of the adrenal medulla, par-ticularly relevant in the analysis of adrenaline and dopamine [13]. Al-though the same authors have suggested that urinary noradrenalinemay not be a reliable measure of sympathoneural activity becausepart of the noradrenaline secreted is taken up into the nerve termi-nals and does not reach the bloodstream or urine [13], other authorshave emphasised the role of noradrenaline as a reliable measure ofphysical activity and sympathetic activity [16,31,40]. Also, results re-lating to dopamine levels are difficult to interpret because dopamineis a precursor of adrenaline and noradrenaline, rather than a stresshormone [13,16,31]. However, almost half of the dopamine is secret-ed by the adrenal medulla and consequently it may be useful to mea-sure it in urine as an additional control for autonomic activity. In

addition, the participants were tested in a controlled hospital settingduring the first day. This has several advantages: compliance with theurine sampling procedure was easier to control; during all experi-mental procedures, investigators kept track of the participants; in-structions for the participants were always available; despite thecontrolled hospital conditions, participants were allowed to wakeup at their natural time in a hotel room avoiding stressful distur-bances. However, we did not control for menstrual cycle, menopausalstate, or oral contraceptives, but findings on the influence of menstru-al cycle are inconsistent. Some studies show decreased noradrenalinelevels in the luteal phase but no changes in the heart rate [43], whileothers report the influence of the menstrual cycle on catecholaminelevels and cardiovascular responses to be small or moderate [31], orabsent during acute stress [44] or when investigating pain symptomsin women during menopausal transition [45]. However, there is nosystematic effect of menstrual cycle on the difference betweengroups. Moreover, we did not specifically control for depressionwhich has been associated with deregulated noradrenaline secretion[46]. However, participants taking antidepressants were excludedand fibromyalgia patients were carefully examined to verify fibromy-algia syndrome diagnosis. Nonetheless, the relatively small samplesize of the present study might have influenced the results. The lackof power may explain the difference in the results between thesignificant main effect for the repeated-measures ANOVA and thenon-significant one-way ANOVAs for the noradrenaline levels.However, it should be underscored that considering the homogeneousgroup of long-term chronic fibromyalgia patients carefully recruited(i.e., average of 13 years since the first symptoms, with range 3–36 years), matched by age with the healthy controls, and the well-controlled hospital-hotel environment, the sample size is reason-able as reflected in the highly significant differences betweengroups.

In sum, the present findings suggest that fibromyalgia patientshave an attenuated activity of both branches of the autonomic ner-vous system. Specifically, fibromyalgia patients showed an attenuatedsympathetic activity of the adrenalmedullary component in terms oflower adrenaline and dopamine levels (although single time pointnoradrenaline levels were not significantly different), and a reducedactivity of the parasympathetic branch in terms of higher restingheart rate, compared with a matched group of healthy controls. Inview of previous findings of significantly lower cortisol levels [8],the results of the present study indicate a central attenuation of themajor stress systems in women suffering from long-term fibromyal-gia syndrome.

Conflict of interest statement

All authors declare that they do not have any conflicts of interest.

Acknowledgments

This research was supported by grants to Professor Ulf Lundbergfrom the Swedish Research Council and the Swedish Council forWorking Life and Social Research. Support was also obtained fromthe Centre for Musculoskeletal Research at the University of Gävle.

Thanks to Associate Professor Petra Lindfors for her comments onthe preliminary manuscript, to Eva Kosek, MD, for her comments onthe preliminary data, to Mrs Ann-Christine Sjöbeck for performingthe catecholamine analysis, and to Mr Håvard Wuttudal Lorås forassisting during data collection.

References

[1] Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, et al.The American College of Rheumatology 1990 Criteria for the Classification of Fi-bromyalgia. Report of the Multicenter Criteria Committee. Arthritis and Rheuma-tism 1990 Feb;33(2):160–72.

57R. Riva et al. / Journal of Psychosomatic Research 72 (2012) 51–57

[2] Kurtze N, Svebak S. Fatigue and patterns of pain in fibromyalgia: correlations withanxiety, depression and co-morbidity in a female county sample. The British Jour-nal of Medical Psychology 2001 Dec;74(Pt 4):523–37.

[3] Wolfe F, Anderson J, Harkness D, Bennett RM, Caro XJ, Goldenberg DL, et al. A pro-spective, longitudinal, multicenter study of service utilization and costs in fibro-myalgia. Arthritis and Rheumatism 1997 Sep;40(9):1560–70.

[4] Krag NJ, Nørregaard J, Larsen JK, Danneskiold-Samsøe B. A blinded, controlledevaluation of anxiety and depressive symptoms in patients with fibromyalgia, asmeasured by standardized psychometric interview scales. Acta Psychiatrica Scan-dinavica 1994 Jun;89(6):370–5.

[5] McEwen BS. Protective and damaging effects of stress mediators. The NewEngland Journal of Medicine 1998 Jan 15;338(3):171–9.

[6] Fries E, Hesse J, Hellhammer J, Hellhammer DH. A new view on hypocortisolism.Psychoneuroendocrinology 2005 Nov;30(10):1010–6.

[7] Heim C, Ehlert U, Hellhammer DH. The potential role of hypocortisolism in thepathophysiology of stress-related bodily disorders. Psychoneuroendocrinology2000 Jan;25(1):1–35.

[8] Riva R, Mork PJ, Westgaard RH, Rø M, Lundberg U. Fibromyalgia syndrome is asso-ciated with hypocortisolism. International Journal of Behavioral Medicine 2010Sep;17(3):223–33.

[9] Tak LM, Cleare AJ, Ormel J, Manoharan A, Kok IC, Wessely S, et al. Meta-analysisand meta-regression of hypothalamic-pituitary-adrenal axis activity in functionalsomatic disorders. Biological Psychology 2011 May;87(2):183–94.

[10] Buskila D, Press J. Neuroendocrine mechanisms in fibromyalgia-chronic fatigue.Best Practice & Research Clinical Rheumatology 2001 Dec;15(5):747–58.

[11] Martinez-Lavin M. Biology and therapy of fibromyalgia. Stress, the stress responsesystem, and fibromyalgia. Arthritis Research & Therapy 2007;9(4):216.

[12] Okifuji A, Turk DC. Stress and psychophysiological dysregulation in patients withfibromyalgia syndrome. Applied Psychophysiology and Biofeedback 2002 Jun;27(2):129–41.

[13] Petzke F, Clauw DJ. Sympathetic nervous system function in fibromyalgia. CurrentRheumatology Reports 2000 Apr;2(2):116–23.

[14] Sarzi-Puttini P, Atzeni F, Diana A, Doria A, Furlan R. Increased neural sympa-thetic activation in fibromyalgia syndrome. Annals of the New York Academyof Sciences 2006 Jun;1069:109–17.

[15] Tak LM, Riese H, de Bock GH, Manoharan A, Kok IC, Rosmalen JGM. As good as itgets? A meta-analysis and systematic review of methodological quality of heartrate variability studies in functional somatic disorders. Biological Psychology2009 Oct;82(2):101–10.

[16] Goldstein DS. Catecholamines and stress. Endocrine Regulations 2003 Jun;37(2):69–80.

[17] Nilsen KB, Sand T, Westgaard RH, Stovner LJ, White LR, Bang Leistad R, et al. Au-tonomic activation and pain in response to low-grade mental stress in fibromyal-gia and shoulder/neck pain patients. European Journal of Pain Oct. 2007;11(7):743–55.

[18] Adler GK, Kinsley BT, Hurwitz S, Mossey CJ, Goldenberg DL. Reduced hypothalam-ic-pituitary and sympathoadrenal responses to hypoglycemia in women with fi-bromyalgia syndrome. American Journal of Medicine 1999 May;106(5):534–43.

[19] Yunus MB, Dailey JW, Aldag JC, Masi AT, Jobe PC. Plasma and urinary catechol-amines in primary fibromyalgia: a controlled study. Journal of Rheumatology1992 Jan;19(1):95–7.

[20] Giske L, Vøllestad NK, Mengshoel AM, Jensen J, Knardahl S, Røe C. Attenuated ad-renergic responses to exercise in women with fibromyalgia—a controlled study.European Journal of Pain Apr. 2008;12(3):351–60.

[21] van Denderen JC, Boersma JW, Zeinstra P, Hollander AP, van Neerbos BR. Physio-logical effects of exhaustive physical exercise in primary fibromyalgia syndrome(PFS): is PFS a disorder of neuroendocrine reactivity? Scandinavian Journal ofRheumatology 1992;21(1):35–7.

[22] Loevinger BL, Muller D, Alonso C, Coe CL. Metabolic syndrome in women withchronic pain. Metabolism, Clinical and Experimental Jan. 2007;56(1):87–93.

[23] Light KC, Bragdon EE, Grewen KM, Brownley KA, Girdler SS, Maixner W. Adrener-gic dysregulation and pain with and without acute beta-blockade in women with

fibromyalgia and temporomandibular disorder. The Journal of Pain May 2009;10(5):542–52.

[24] Torpy DJ, Papanicolaou DA, Lotsikas AJ, Wilder RL, Chrousos GP, Pillemer SR. Re-sponses of the sympathetic nervous system and the hypothalamic-pituitary-adrenal axis to interleukin-6: a pilot study in fibromyalgia. Arthritis and Rheuma-tism Apr. 2000;43(4):872–80.

[25] Kadetoff D, Kosek E. Evidence of reduced sympatho-adrenal and hypothalamic-pituitary activity during static muscular work in patients with fibromyalgia. Jour-nal of Rehabilitation Medicine 2010 Sep;42(8):765–72.

[26] Thayer JF, Brosschot JF. Psychosomatics and psychopathology: looking up anddown from the brain. Psychoneuroendocrinology 2005 Nov;30(10):1050–8.

[27] Chrousos GP, Gold PW. The concepts of stress and stress system disorders. Over-view of physical and behavioral homeostasis. Journal of the American Medical As-sociation Mar 4 1992;267(9):1244–52.

[28] Hubert W, de Jong-Meyer R. Psychophysiological response patterns to positiveand negative film stimuli. Biological Psychology Aug 1991;31(1):73–93.

[29] Lundberg U, Kadefors R, Melin B, Palmerud G, Hassmen P, Engstrom M, et al. Psy-chophysiological stress and EMG activity of the trapezius muscle. InternationalJournal of Behavioral Medicine 1994;1(4):354–70.

[30] Borg G, Borg E. A new generation of scaling methods: level anchored ration scal-ing. Psychologia 2001;28:15–45.

[31] Kudielka BM, Wüst S, Kirschbaum C, Hellhammer DH. Trier Social Stress Test. En-cyclopedia of stress. 2nd revised ed. Oxford: Academic Press; 2007. p. 776–81.

[32] Lindfors P. Psychophysiological aspects of stress, health and well-being. VDMVerlag; 2008.

[33] Westgaard RH, Mork PJ, Lorås HW, Riva R, Lundberg U. Fibromyalgia patientsshow elevated trapezius muscle activity in tests with autonomic activation, butsimilar habitualtrapezius activity to healthy controls. Submitted for publication.

[34] Klingmann PO, Hellhammer DH. Noradrenergic and sympathetic disorders.Stress: the brain-body connection. Basel: Karger; 2008. p. 78–90.

[35] Ziegler MG, Ruiz-Ramon P, Shapiro MH. Abnormal stress responses in patientswith diseases affecting the sympathetic nervous system. Psychosomatic Medicine1993 Aug;55(4):339–46.

[36] Finan PH, Zautra AJ, Davis MC, Lemery-Chalfant K, Covault J, Tennen H. Genetic in-fluences on the dynamics of pain and affect in fibromyalgia. Health Psychology2010 Mar;29(2):134–42.

[37] Li PP, Warsh JJ, Godse DD. Formation and clearance of norepinephrine glycol me-tabolites in mouse brain. Journal of Neurochemistry 1984 Nov;43(5):1425–33.

[38] Thayer JF, Lane RD. A model of neurovisceral integration in emotion regulationand dysregulation. Journal of Affective Disorders 2000 Dec;61(3):201–16.

[39] Galbo H, Holst JJ, Christensen NJ. Glucagon and plasma catecholamine responsesto graded and prolonged exercise in man. Journal of Applied Physiology 1975Jan;38(1):70–6.

[40] Lundberg U. Stress hormones in health and illness: the roles of work and gender.Psychoneuroendocrinology 2005 Nov;30(10):1017–21.

[41] Korszun A, Young EA, Engleberg NC, Brucksch CB, Greden JF, Crofford LA. Use ofactigraphy for monitoring sleep and activity levels in patients with fibromyalgiaand depression. Journal of Psychosomatic Research 2002 Jun;52(6):439–43.

[42] Okifuji A, Bradshaw DH, Olson C. Evaluating obesity in fibromyalgia: neuroendo-crine biomarkers, symptoms, and functions. Clinical Rheumatology 2009 Apr;28(4):475–8.

[43] Hirshoren N, Tzoran I, Makrienko I, Edoute Y, Plawner MM, Itskovitz-Eldor J, et al.Menstrual cycle effects on the neurohumoral and autonomic nervous systemsregulating the cardiovascular system. Journal of Clinical Endocrinology and Me-tabolism 2002 Apr;87(4):1569–75.

[44] Mills PJ, Nelesen RA, Ziegler MG, Parry BL, Berry CC, Dillon E, et al. Menstrual cycleeffects on catecholamine and cardiovascular responses to acute stress in black butnot white normotensive women. Hypertension 1996 Apr 1;27(4):962–7.

[45] Mitchell ES, Woods NF. Pain symptoms during the menopausal transition andearly postmenopause. Climacteric 2010 Oct;13(5):467–78.

[46] Meyer SE, Chrousos GP, Gold PW. Major depression and the stress system: a lifespan perspective. Development and Psychopathology 2001;13(3):565–80.