Original Article

Fibromyalgia and bipolar disorder:a potential problem?

Approximately 2% of the general populationsuffers from fibromyalgia, a syndrome character-ized by chronic widespread pain and fatigue,multiple tender points on examination, and avariable constellation of additional symptoms:

poor sleep, frequent headaches, morning stiffness,paresthesias, irritable bowel syndrome, urinaryurgency, and mood disturbances (1). A largenumber of patients with fibromyalgia have sub-stantial lifetime psychiatric comorbidity (2). About17% of the general population, and up to 90% offibromyalgia patients, suffer from depression dur-ing their lifetime (3, 4).Bipolar disorder affects about 1.5% of adults and

is characterized by a history of manic or hypomanicepisodes. The disease is often misdiagnosed as

Wilke WS, Gota CE, Muzina DJ. Fibromyalgia and bipolar disorder:a potential problem?Bipolar Disord 2010: 12: 514–520. ª 2010 The Authors.Journal compilation ª 2010 John Wiley & Sons A ⁄S.

Objective: To screen patients with fibromyalgia for bipolar disorderand to determine if there were any clinical clues, other than the MoodDisorders Questionnaire (MDQ), which might suggest a diagnosis ofcomorbid bipolar disorder.

Methods: A total of 128 consecutive new fibromyalgia patients referredto a tertiary care center rheumatology practice were enrolled andassessed using a standard clinical protocol that included the completionof four screening questionnaires: (i) MDQ for bipolar disorder, (ii) BeckDepression Inventory (BDI) for depression, (iii) Epworth SleepinessScale (ESS) for daytime sleepiness, and (iv) Fibromyalgia ImpactQuestionnaire Disability Index (FIQ-DI) to assess for functionalcapacity.

Results: A quarter of the fibromyalgia subjects, 25.19%, had a positivescreen for bipolar disorder (MDQ ‡ 7); 78.12% were clinically depressed(BDI ‡ 10); and 52.13% reported daytime sleepiness (ESS ‡ 10).Fibromyalgia subjects who screened positive for bipolar disorder hadmore severe depression than those with a negative screen [medianBDI: 26.0 (19.0, 32.0) versus 15.0 (9.0, 24.0), p < 0.001].

Conclusions: We report a high prevalence of positive testing for bipolardisorder in this fibromyalgia cohort. Clinical data and questionnaireinstruments other than nonspecific high depression severity failed toidentify these patients. Since the norepinephrine serotonin reuptakeinhibitors duloxetine and milnacipran have been recently approved bythe U.S. Food and Drug Administration for the treatment offibromyalgia, and because patients with bipolar disorder may experiencedestabilization of mood when treated with such agents, patients withfibromyalgia should be systematically screened for bipolar disorderprior to treatment.

William S Wilkea, Carmen E Gotaa

and David J Muzinab

aDepartment of Rheumatic and Immunologic

Diseases, Cleveland Clinic Orthopedic and

Rheumatologic Institute, bCenter for Mood

Disorders Treatment and Research, Department of

Psychiatry and Psychology, Cleveland Clinic

Neurological Institute, Lutheran Hospital,

Cleveland, OH, USA

doi: 10.1111/j.1399-5618.2010.00848.x

Keywords: Beck Depression Inventory (BDI) –

bipolar disorder – depression – fibromyalgia –

Mood Disorder Questionnaire (MDQ)

Received 25 September 2009, revised and

accepted for publication 3 May 2010

Corresponding author:

William S. Wilke, M.D.

Department of Rheumatic and

Immunologic Diseases, A50

Cleveland Clinic Orthopedic and

Rheumatologic Institute

9500 Euclid Avenue

Cleveland, OH 44195, USA

Fax: 216-445-7569

E-mail: wilkew@ccf.org

WSW has received consulting fees from Pfizer. DJM has served as a

speaker for Pfizer; and has received consulting fees from Pfizer,

AstraZeneca, and Bristol-Myers Squibb. CEG has no conflict of

interest to report.

Bipolar Disorders 2010: 12: 514–520ª 2010 The Authors

Journal compilation ª 2010 John Wiley & Sons A/S

BIPOLAR DISORDERS

514

unipolar depression, because depressive symptomsdominate the course of the disease (3, 5). For thisreason, an acute manic episode, which occurs veryfrequently with depression, can be very challenging,and treatment needs can be quite different fromunipolar depression.Inquiries regarding bipolar disorder in patients

with fibromyalgia are limited (2, 6). It has beensuggested that these patients have poor outcomesand do not respond well to fibromyalgia treatment(7). To date, the pivotal antidepressant treatmenttrials in fibromyalgia, reporting the efficacy of theU.S. Food and Drug Administration (FDA)-approved norepinephrine serotonin reuptakeinhibitors (NSRIs) duloxetine and milnacipran,excluded patients with psychiatric conditions otherthan major depressive disorder (8–10). While theseand other antidepressants have proven useful forfibromyalgia patients (11), their use in individualswith bipolar disorder comes with the followingimportant caveats: (i) concerns for lack of efficacy(12), (ii) increased risk for triggering manic epi-sodes or cycle acceleration (12, 13), and (iii)possible increase of suicidal ideation (14). Due tothese concerns, the FDA advises physicians toscreen for bipolar disorder prior to the prescriptionof antidepressants in the management of unipolarmajor depression (15).Given the possible comorbidity of bipolar dis-

order with fibromyalgia, and the potential negativeimpact of antidepressants on patients with unrec-ognized simultaneous bipolar disorder, it is imper-ative that this subset of patients be identified totailor efficacious and safe treatment.The goal of our study was to determine how

many of our patients with fibromyalgia screenpositive for bipolar disorder using the MoodDisorders Questionnaire (MDQ) (16). We werealso interested in determining if there were anyother clinical clues, besides the MDQ score, whichcould help identify this subset of patients.

Material and methods

Inclusion criteria

All new consecutive patients diagnosed withfibromyalgia that were seen in the RheumatologyDepartment at the Cleveland Clinic (Cleveland,OH, USA) by one physician (WSW) over asix-month period (January 2005–June 2005) wereenrolled in the study. The diagnosis of fibromyalgiawas made by one physician (WSW) based on thepresence of widespread pain, generalized tendernessto palpation, multiple tender points, and clinicalcriteria [‘‘the company fibromyalgia keeps’’ (17)].

The study was approved by the Cleveland ClinicInstitutional Review Board. The routine clinicalassessment of patients with fibromyalgia in ourclinic probably exceeds the standard evaluationperformed in most rheumatology practices. Wecollect the following data: demographic (gender,race, marital status, employment status), socialhistory, fibromyalgia symptom inventory, presenceof tender points, brisk deep-tendon reflexes andlarge-artery tenderness on palpation, number andkind of medications prescribed for fibromyalgia,and presence or absence of sleep apnea. Theprotocol includes the completion of four screeningquestionnaires: (i) MDQ for bipolar disorder, (ii)Beck Depression Inventory (BDI) (18) for depres-sion, (iii) Epworth Sleepiness Scale (ESS) (19) forsleep apnea, and (iv) Fibromyalgia Impact Ques-tionnaire Disability Index (FIQ-DI) (20) to assessfunctional capacity.

Bipolar disorder risk assessment

The MDQ is a self-report inventory that screens fora lifetimehistory of amanic or hypomanic syndromeby including 13 yes ⁄no items (Question 1).In addition to achieving the threshold number ofsymptom items, the subject must also have indicatedthat the symptoms clustered in the same period oftime (Question 2), and caused moderate or seriousproblems (Question 3) (16, 21, 22). A MDQ screen-ing score of ‡ 7, plus concurrence of symptoms andat least moderate problems reported, was foundto have a sensitivity of 29–91% and a specificity of67–94% for the diagnosis of bipolar disorder(16, 22–25). We employed a modified MDQ, usinga cutoff of ‡ 7 �yes� answers (Question 1), withouttaking into account symptom clustering and severity(Questions 2 and 3). This was found to yield asensitivity of 68–81% and specificity of 63–95%(21, 26). Limiting the MDQ to the first 13 items(Question 1), may increase the sensitivity to detecthypomania,whichmay cause little or no impairmentand thus escape detection (24).

Depression assessment

The BDI is a widely used 21-item self-reportinventory of depressive symptoms. Using a cutoffscore of 10, the BDI correctly identified 84% of thecases of depression, with a sensitivity of 89% and aspecificity of 82% (18, 27).

Daytime sleepiness assessment

The ESS is a self-administered questionnairedesigned to measure daytime sleepiness. It rates

Fibromyalgia and bipolar disorder

515

on a scale of 0–3 the chances for dozing in each ofeight different situations. The ESS score is the sumof the eight item scores (range 0–24). The clinically�normal� range of ESS scores is 2–9. ESS score ‡ 10is encountered in patients with various sleepproblems such as sleep apnea, narcolepsy, andrestless leg syndrome (19).

Functional assessment

The Fibromyalgia Impact Questionnaire wasdeveloped to capture the total spectrum of fibro-myalgia-related symptoms and has been extensivelyused as an outcome measure in fibromyalgia-related studies. The first section, with 10 items,quantitates activities of daily living, which involvelarge muscle groups, and is termed the FIQ-Disability Index (FIQ-DI). The responses arescaled from 0 = always able to do, to 3 = neverable to do. The FIQ-DI shows good correlationwith pain, the tender-point count, and the myalgicscore (28).

Statistical analysis

Continuous measures were described by means,standard deviations, and percentiles. Categoricalmeasures were summarized using frequencies andpercentages. For continuous outcomes of interest,two-sample t-tests were performed. The Wilcoxonrank sum test was used if the normality assumptionof the t-test was violated. For categorical outcomesof interest, Pearson�sChi-square testwas used unlessindicated otherwise. Fisher�s Exact Test was usedwhen expected cell counts were < 5. Spearman�scorrelation coefficients were estimated in order tostudy associations between FIQ-DI, MDQ, ESS,and BDI on a continuous scale. All tests wereperformed at a significance level of 0.05. Note thatfor each outcome, many hypothesis tests wereperformed. With additional tests, the probabilityof obtaining a statistically significant result bychance alone also increases. However, no p-valuecorrection was performed, in order to identify allpotentially important associations that maywarrant further study. SAS software (SAS Insti-tute, Cary, NC, USA) was used for all analyses.

Results

The demographic, social, and clinical characteris-tics of the group are presented in Tables 1 and 2.Symptomatic depression (BDI ‡ 10) was present in78.12% of the 128 fibromyalgia patients; BDImean ± SD score = 19.5 ± 11.2, indicating alevel of moderate to moderately severe depression.

A positive bipolar disorder screen (MDQ ‡ 7) wasfound in 25.19% of the 127 patients who com-

Table 1. Sociodemographic characteristics of the fibromyalgia patients

Female, % 85.2Age, mean ± SD 46.3 ± 13.6Weight (kg), mean ± SD 88.8 ± 23.4Height (inches), mean ± SD 65.5 ± 3.1Median body mass index (min, max) 29.8 (16.4, 55.2)Ethnicity, %

Caucasian 85.9African American 8.6Hispanic 0.8Other 3.1Unknown 1.6

Work history, %Employed 40.6Unemployed 18.0Receiving disability 9.4Retired 11.7Student 2.3

Marital status, %Married 53.9Divorced ⁄ separated 14.9Single 16.4Widowed 0.8

Family history of: (%)Fibromyalgia 18.8Depression 14.4Bipolar disorder 2.4Suicide 1.6

Table 2. Clinical characteristics of the fibromyalgia patients

Fatigue 89.1Poor sleep 93.0Headaches 35.9Irritability with lights, noises, smells 82.0Sensation of swollen glands ⁄ sore throat 57.8Memory impairment 4.7Concentration impairment 58.6Alternating constipation with diarrhea 78.9Orthostasis ⁄ feeling faint or weak after hot shower 60.2Urinary frequency 61.7Restless legs at night 7.8Intermittent paresthesias 25.8Tender carotid arteries 48.4Brisk deep-tendon reflexes 63.3Medications

Antidepressants: SSRIs, NSRIs 35.9Other antidepressants (bupropion, trazodone) 16.4Tricyclic and tetracyclic antidepressants 14.1Benzodiazepines and buspirone 24.2Analgesics and nonsteroidal anti-inflammatorydrugs

50.0

Narcotic analgesics 27.3Pregabalin or gabapentin 6.3

Patients with known sleep apnea by history ortesting

19.5

Total number of medications for fibromyalgia(mean ± SD)

2.0 ± 2.0

Values listed as percent except where noted otherwise.SSRIs = selective serotonin reuptake inhibitors; NSRIs =norepinephrine serotonin reuptake inhibitors.

Wilke et al.

516

pleted the questionnaire; nearly all of these subjects(96.9%) also had symptomatic depression. Amongthe 100 subjects with fibromyalgia and sympto-matic depression, nearly one third (32%) screenedpositive for bipolar disorder. Mean ± SD MDQscore for the entire fibromyalgia population was4.1 ± 3.4. Daytime sleepiness (ESS ‡ 10) wasreported by 52.13% of the 117 patients whocompleted the questionnaire, ESS mean ± SDscore = 10.4 ± 5.4. Of the 32 patients with anESS score ‡ 10 who underwent a sleep study, 60%were found to have sleep apnea.

Relationship between depression, bipolar symptoms,daytime sleepiness, and functioning

The MDQ and BDI scores showed a significantcorrelation, rho = 0.46 [95% confidence interval(CI): 0.31–0.62, p < 0.001]. Fibromyalgia subjectswho screened positive for bipolar disorder weresignificantly more depressed than those with anegative screen: median BDI = 26.0 (19.0, 32.0)versus 15.0 (9.0, 24.0), p < 0.001.The FIQ-DI correlated with the BDI score,

rho = 0.52 (95% CI: 0.37–0.67, p < 0.001). Themedian FIQ-DI score in subjects with symptomaticdepression (BDI ‡ 10) was significantly higher thanthe FIQ-DI score in nondepressed subjects (BDI <10), median FIQ-DI = 17.0 (10.0, 21.0) versus 8.0(0.0, 15.0), p < 0.001. No significant differenceswere noted in FIQ-DI scores between putativebipolar and nonbipolar disorder fibromyalgiapatients (data not shown).Significant, but small, correlations were observed

between the ESS and MDQ, rho = 0.27 (95% CI:0.10–0.45, p = 0.003), and the ESS and BDI,rho = 0.23 (95% CI: 0.05–0.41, p = 0.012).The median ESS score was significantly higher in

fibromyalgia patients with depression (BDI ‡ 10)versus fibromyalgia patients without depression(BDI < 10), median ESS = 12.0 (7.0, 15.0) versus8.0 (3.0, 12.0), p = 0.005.We used Pearson Chi-square test to compare

categorical demographic and clinical variablesgrouping fibromyalgia subjects by presence orabsence of symptomatic depression (BDI ‡ 10 andBDI < 10), probable bipolar disorder (MDQ ‡ 7and MDQ < 7), and daytime sleepiness (ESS ‡ 10and ESS < 10). Subjects with depressive symptomshad significantly more fatigue (p = 0.002), concen-tration impairment (p = 0.001), symptoms of uri-nary frequency (p = 0.02), and brisk reflexes(p = 0.036), and were more likely to take selec-tive serotonin reuptake inhibitors (p = 0.024)and other antidepressants [buspirone, trazodone(p = 0.043)] than nondepressed subjects. Fewer

fibromyalgia subjects with probable bipolar disor-der (MDQ ‡ 7) had concentration impairment(p = 0.002) and urinary frequency (p = 0.006),andweremore likely to be treatedwith pregabalin orgabapentin (p = 0.027), than those with a negativebipolar disorder screen (MDQ < 7). Subjects whoreported daytime sleepiness (ESS ‡ 10) had moreconcentration impairment (p = 0.004), briskreflexes (p = 0.006), and tender carotid arteries(p = 0.024) than subjects with ESS < 10.No association was observed between depressive

symptoms, symptoms suggestive of bipolar disor-der or daytime sleepiness, and age, weight, bodymass index, marital status, employment status,number of surgeries, and number of medicationstaken for fibromyalgia (data not shown).

Discussion

The most important finding of our study pertainsto the high prevalence of a positive screen forbipolar disorder (25.19%) in patients with fibro-myalgia. With the exception of more severe depres-sion, no other clinical clues were present to helpidentify patients with fibromyalgia and bipolardisorder.Only a few studies have investigated bipolar

disorder in fibromyalgia patients (6, 7, 29). Ourresults are similar to those reported by Carta andcolleagues (6), who found a positive screen forbipolar disorder (using a cutoff of 6 on the MDQ)in 29.7% of 37 Italian women with fibromyalgia. Inthe U.S., patients with fibromyalgia from twocommunity-based rheumatology practices had a12.8% lifetime prevalence of bipolar disorder (29),which is much higher than the 1.5% incidencereported in the general population (3, 5). Thehigher prevalence of a positive BD screen in ourstudy may reflect differences between fibromyalgiapatient populations seen in the community andthose seen in a tertiary care referral center, but alsodifferences in the diagnostic methods used.Almost a third of depressed fibromyalgia patients

(32%) had a positive bipolar disorder screen. Ofnote, almost all (96.9%) patients with fibromyalgiaand probable bipolar disorder were at least mod-erately depressed by self-report at the time ofevaluation, again highlighting the predominantlydepressive nature of bipolar disorder. Severalprevious studies reported that bipolar disorder iscommon in patients with presumed unipolardepression, with a prevalence of 18.6–61.2% (7,30). In our cohort, putative bipolar disorderpatients were more likely to be treated withpregabalin or gabapentin, which may indirectlysuggest that pain was a prominent complaint.

Fibromyalgia and bipolar disorder

517

The majority of fibromyalgia patients (78.12%)reported clinically significant levels of depression.Although we found higher rates of concurrentdepression in these patients than some havereported (20–46%) (31–34), our findings are notunique. Using a BDI cutoff of 9, Guven andcolleagues (4) reported a depression rate of 90%and mean BDI scores of 15.7 ± 9.7 among 53Turkish women with fibromyalgia. A BDI cutoff of10 for depression may be considered too low bysome authors, who have suggested that cutoffs of13 may be more appropriate (35, 36). It was arguedthat in patients with chronic pain, the BDI couldoverdiagnose depression by capturing somaticsymptoms, such as fatigue, sleep difficulty, andeffort required to get things done, which onlyreflect other comorbidities (35, 37). Using a cutoffof 13, 68.75% of our fibromyalgia patients wouldhave been considered depressed. The mean BDIscore of 15 in our patients lies well above mostaccepted thresholds for clinically relevant depres-sion. Therefore, we believe that overall in ourfibromyalgia population, the observed high prev-alence of depression is not an overestimate.The co-occurrence of fibromyalgia with mood

disorders may be explained in part through agenetic link. Alteration of the serotonergic systemis one of the factors underlying the pathogenesis offibromyalgia. Harm-avoidance temperament inindividuals with the S-allele and fibromyalgia hasbeen observed (38). These linked factors mayincrease perceived distress and play a role in thepathogenesis of fibromyalgia. This relationshipbetween fibromyalgia and mood disorders such asmajor depression and bipolar disorder becomesimportant when considered in the context offibromyalgia treatment with tricyclic antidepres-sants, selective serotonin reuptake inhibitors(SSRIs), and NSRIs (11, 39, 40). These agentsare associated with a variety of adverse outcomesin bipolar disorder, including affective switch andmixed rapid cycling at a rate of two or three timesthe spontaneous rate (41).Our study has several limitations. The study

sample was moderately sized and data werecollected from a naturalistic treatment settingwithout randomization or blinding. Formal psy-chiatric assessments or structured clinical inter-views were not conducted; subjective reporting ofmood symptoms using the BDI and MDQ does notequate to a diagnosis of major depression orbipolar disorder. It should also be emphasizedthat since our subjects were recruited from atertiary care setting, they may represent a subsetwith more severe and treatment-resistant illness, orwith a higher frequency of comorbidities. The

lower prevalence of bipolar disorder in patientswith fibromyalgia from community-based rheuma-tology practices supports this hypothesis (29). Forthese reasons, we do not know if our findings maybe generalizable beyond the tertiary care oracademic clinic setting. Additionally, it has previ-ously been suggested that a higher incidence ofpsychiatric comorbidity characterizes patients withfibromyalgia who actively seek medical care,resulting in ascertainment bias (42).In defense of our methods, the BDI and the

MDQ have been found to be reliable, and arewidely used as screening tests for depression andbipolar disorder (16, 18, 43). In addition, ourfibromyalgia patient population as a prospectivecohort was composed of unselected, consecutivepatients. Such naturalistic setting observationsutilizing basic screening and symptom-measuringtools provide valuable information to practicingclinicians in the �real world� setting. It is possiblethat adjustments for age and sex may have changedour results somewhat, but we believe the effect isnegligible. Although we did not use the AmericanCollege of Rheumatology diagnostic criteria for thediagnosis of fibromyalgia in all patients, all patientsexperienced diffuse pain of at least three months�duration, and a constellation of other symptomsthat are absolutely consistent with the diagnosis offibromyalgia, as the concept evolves (44).Despite these limitations, our findings should

add to the existing concerns regarding the prescrip-tion of antidepressants to patients with bipolardisorder (known or unknown) given the approvedtreatment options for fibromyalgia, which nowinclude two NSRIs with antidepressant proprieties,duloxetine andmilnacipran. Althoughmuch debateremains regarding the true risk of affective switch-ing caused by antidepressants, NSRIs may carry agreater risk for causing mood instability in patientswith bipolar disorder (13). Manic switching hasbeen reported in patients treated with duloxetine(45). Unsuspecting clinicians, believing that theNSRIs could address both the symptoms offibromyalgia and the commonly concurrent depres-sion, may be placing patients with bipolar disorderat unacceptable risk. Treating physicians withknowledge of potential bipolar disorder will beable to counsel patients about the risks of affectiveswitches and ⁄or provide mood stabilizers and ⁄orreferral to psychiatry when appropriate.

Conclusions

We report a high prevalence of potential bipolardisorder in a moderately sized cohort of fibro-myalgia patients seen by one physician using a very

Wilke et al.

518

comprehensive standard outpatient protocol.Despite all of the data collected, the provisionaldiagnosis of bipolar disorder could not be madewithout use of the MDQ. Comorbid depressionwas high, and greater severity of depression wasmoderately associated with bipolar disorder.Severe depression in fibromyalgia might be usedas an indication for further investigation forbipolar disorder.With FDA approval of duloxetine and milna-

cipran for the treatment of fibromyalgia, there isthe potential for naıve antidepressant treatment offibromyalgia patients with comorbid bipolar dis-order, resulting in significant and unexpectedtoxicity. We urge that bipolar disorder risk becarefully assessed in all patients with fibromyalgiaprior to initiation of drug therapy, particularlysince the NSRIs may carry elevated risk forinduction of mania. Alternative pharmacothera-pies for the comorbid fibromyalgia ⁄bipolar disor-der patients should be considered and warrantmore intensive future investigation.

Acknowledgement

Partial funding for the statistical analysis of this study wasreceived from the CCF Rackley Charitable Heritage Fund ofthe Baton Rouge Area Foundation.

References

1. Wolfe F, Ross K, Anderson J, Russell IJ. Aspects offibromyalgia in the general population: sex, pain thresholdand fibromyalgia symptoms. J Rheumatol 1995; 22: 151–156.

2. Arnold LM, Hudson JI, Keck PE Jr, Auchenbach MB,Javaras KN, Hess EV. Comorbidity of fibromyalgia andpsychiatric disorders. J Clin Psychiatry 2006; 67: 1219–1225.

3. Kessler RC, McGonagle KA, Shanyang Z et al. Lifetimeand 12-month prevalence of DSM-III-R psychiatric disor-ders in the United States. Arch Gen Psychiatry 1994; 51:8–19.

4. Guven AZ, Kul Panza E, Gunduz OH. Depression andpsychosocial factors in Turkish women with fibromyalgiasyndrome. Eura Medicophys 2005; 41: 309–313.

5. Post RM, Denicoff KD, Leverich GS et al. Morbidity in258 bipolar outpatients followed for 1 year with dailyprospective ratings on the NIMH life chart method. J ClinPsychiatry 2003; 64: 680–690.

6. Carta MG, Cardia C, Mannu F et al. The high frequencyof manic symptoms in fibromyalgia does influence thechoice of treatment? Clin Pract Epidemol Ment Health2006; 2: 1–5.

7. Wallace DJ, Gotto J. Hypothesis: bipolar illness withcomplaints of chronic musculoskeletal pain is a form ofpseudofibromyalgia. Semin Arthritis Rheum 2008; 37:256–259.

8. Russell IJ, Mease PJ, Smith TR et al. Efficacy and safety ofduloxetine for treatment of fibromyalgia in patients withand without major depression: Results from a 6-months,

randomized, double-blind, placebo-controlled, fixed-dosetrial. Pain 2008; 136: 432–444.

9. Arnold LM, Hess EV, Hudson JI, Welge JA, Berno SE,Keck PE Jr. A randomized, placebo-controlled, double-blind, flexible-dose study of fluoxetine in the treatment ofwomen with fibromyalgia. Am J Med 2002; 112: 191–197.

10. Mease PJ, Clauw DJ, Gendreau RM et al. The efficacy andsafety of milnacipran for treatment of fibromyalgia. Arandomized, double-blind, placebo-controlled trial.J Rheumatol 2009; 36: 398–409.

11. Hauser W, Bernardy K, Uceyler N, Sommer C. Treatmentof fibromyalgia syndrome with antidepressants: a meta-analysis. JAMA 2009; 301: 198–209.

12. Goldberg JF, Perlis RH, Ghaemi SN et al. Adjunctiveantidepressant use and symptomatic recovery amongbipolar depressed patients with concomitant manic symp-toms: findings from the STEP-BD. Am J Psychiatry 2007;164: 1348–1355.

13. Post RM, Altshuler LL, Leverich GS et al. Mood switch inbipolar depression: comparison of adjunctive venlafaxine,bupropion and sertaline. Br J Psychiatry 2006; 189: 124–131.

14. Baumer FM, Howe M, Gallelli K, Simeonova DI,Hallmayer J, Chang KD. A pilot study of antidepressant-induced mania in pediatric bipolar disorder: characteris-tics, risk factors, and the serotonin transporter gene. BiolPsychiatry 2006; 60: 1005–1012.

15. FDA.gov. Antidepressant use in children, adolescents, andadults. http://www.fda.gov/Drugs/DrugSafety/Informa-tionbyDrugClass/ucm096273.htm. Accessed July 11, 2010.

16. Hirschfeld RMA, Williams JBW, Spitzer RL et al. Devel-opment and validation of a screening instrument forbipolar spectrum disorder: the Mood Disorder Question-naire. Am J Psychiatry 2000; 157: 1873–1875.

17. Wilke WS. New developments in the diagnosis of fibrom-yalgia syndrome: say goodbye to tender points. Clev Clin JMed 2009; 76: 345–352.

18. Beck A, Ward C, Mendelson M, Mock J, Erbauch J. Aninventory for measuring depression. Arch Gen Psychiatry1961; 4: 561–571.

19. Johns MW. A new method for measuring daytime sleep-iness: the Epworth Sleepiness Scale. Sleep 1991; 14: 540–545.

20. Bennett RM. The Fibromyalgia Impact Questionnaire(FIQ): a review of its development, current version,operating characteristics and uses. Clin Exp Rheumatol2005; 23 (Suppl. 39): S154–S162.

21. Kim B, Wang HR, Son JI, Kim CY, Joo YH. Bipolarity indepressive patients without histories of diagnosis of bipolardisorder and the use of the Mood Disorder Questionnairefor detecting bipolarity. Compr Psychiatry 2008; 49: 469–475.

22. Hirschfeld RMA. The Mood Disorder Questionnaire: asimple, patient-rated screening instrument for bipolardisorder. Prim Care Companion J Clin Psychiatry 2002;4: 9–11.

23. Twiss J, Jones S, Anderson I. Validation of the MoodDisorder Questionnaire for screening for bipolar disorderin a UK sample. J Affect Disord 2008; 110: 180–184.

24. Miller CJ, Klugman J, Berv DA, Rosenquist KJ, GhaemiSN. Sensitivity and specificity of the Mood DisorderQuestionnaire for detecting bipolar disorder. J AffectDisord 2004; 81: 167–171.

25. Stang P, Frank C, Ulcickas Yood M, Wells K, Burch S.Impact of bipolar disorder: results from a screening study.Prim Care Companion J Clin Psychiatry 2007; 9: 42–47.

Fibromyalgia and bipolar disorder

519

26. Sanchez-Moreno J, Villagran JM, Gutierrez JR et al.Adaptation and validation of the Spanish version of theMood Disorder Questionnaire for the detection of bipolardisorder. Bipolar Disord 2008; 10: 400–412.

27. Berger A, Dukes E, Martin S, Edelsberg J, Oster G.Characteristics and healthcare costs of patients with fibrom-yalgia syndrome. Int J Clin Pract 2007; 61: 1498–1508.

28. Jensen B, Wittrup IH, Rogind H, Danneskiold-Samsoe B,Bliddal H. Correlation between tender points and theFibromyalgia Impact Questionnaire. J Musculoskel Pain2000; 8: 19–28.

29. Arnold LM, Hudson JI, Hess EV et al. Family study offibromyalgia. Arthritis Rheum 2004; 50: 944–952.

30. Calabrese JR, Muzina DJ, Kemp DE et al. Predictors ofbipolar disorder risk among patients currently treated formajor depression. Med Gen Med 2006; 15: 38–43.

31. Kirmayer LJ, Robbins JM, Kapusta MA. Somatizationand depression in fibromyalgia. Am J Psychiatry 1988; 145:950–954.

32. Epstein SA, Kay G, Clauw D et al. Psychiatric disorders inpatients with fibromyalgia. A multicenter investigation.Psychosomatics 1999; 40: 57–63.

33. Kurland JE, Coyle WJ, Winkler A, Zable E. Prevalence ofirritable bowel syndrome and depression in fibromyalgia.Dig Dis Sci 2006; 51: 454–460.

34. Offenbaecher M, Glatzeder K, Ackenheil M. Self-reporteddepression, familial history of depression and fibromyalgia(FM) and psychological distress in patients with FM.J Rheumatol 1998; 57: 94–96.

35. Burckhardt CS, O�Reilly CA, Wiens AN, Clark SR,Campbell SM, Bennett RM. Assessing depression infibromyalgia patients. Arthritis Care Res 1994; 7: 35–39.

36. Turner JA, Romano JM. Self-report screening measuresfor depression in chronic pain patients. J Clin Psychol1984; 40: 909–913.

37. Williams AC, Richardson PH. What does the BDI measurein chronic pain? Pain 1993; 55: 259–266.

38. Cohen H, Buskila D, Neumann L, Ebstein RP. Confirma-tion of an association between fibromyalgia and serotonintransporter promoter region (5- HTTLPR) polymorphism,and relationship to anxiety-related personality traits.Arthritis Rheum 2002; 46: 845–857.

39. Cipriani A, Furukawa TA, Salanti G et al. Comparativeefficacy and acceptability of 12 new-generation antidepres-sants: a multiple-treatments meta-analysis. Lancet 2009;373: 746–758.

40. Nakagawa A, Watanabe N, Omori IM et al. Milnacipranversus other antidepressive agents for depression. Cochra-ne Database Syst Rev 2009; 3: CD006529.

41. El-Mallakh RS, Karippot A. Use of antidepressants totreat depression in bipolar disorder. Psychiatr Serv 2002;53: 580–584.

42. Aaron LA, Bradley LA, Alarcon GS et al. Psychiatricdiagnoses in patients with fibromyalgia are related tohealth care-seeking behavior rather than to illness. Arthri-tis Rheum 1996; 39: 436–445.

43. Hirschfeld RMA, Calabrese JR, Weissman MM et al.Screening for bipolar disorder in the community. J ClinPsychiatry 2003; 64: 53–59.

44. Khasnis A, Wilke WS. Diagnosing fibromyalgia: movingaway from tender points. J Musculoskel Med 2010; 27:1–5.

45. de Dios C, Ezquiaga E. Manic switching in patientsreceiving duloxetine. Am J Psychiatry 2007; 164: 1121.

Wilke et al.

520