With expansion of the concept of bipolar disorder (BD), there has been concern about the potential for overdiagnosis. However, diagnostic errors in bipolar disorder are currently skewed towards underdiagnosis.
July 2006, Vol. XXIII, No. 8
Psychiatric diagnosis is currentlybased on a system of categories,much like the Linnaean classificationin biology. However, a seeminglyopposite diagnostic system is emerging,usually dubbed the dimensional view,in which related psychiatric conditionssuch as major depressive disorder (MDD)and bipolar disorder (BD) are seen asend points of a spectrum.
With this expansion of the conceptof BD to include intermediate forms (ie,manifestations between clearly unipolarand clearly bipolar cases), concernshave arisen about the potential for overdiagnosisof BD1 and the blurring of ourunderstanding.2 However, diagnosticerror thus far has been skewed in thedirection of underdiagnosis.3
The concept of overdiagnosis or underdiagnosispresumes some naturaltypes that we diagnosticians arecorrectly or incorrectly classifying. Indeed,the term classify implies that weare looking at native classes. This is,for example, the process used by anornithologist to identify a particularbird: scientists use visible traits such assize, color, beak shape, wing bars, eyerings, and other field signs. They mayalso observe preferences such as habitat(ground or high trees), interactions(flocking or pairing), diet (seeds, worms,or fish), and even behavior over time(eg, migration or wintering-over).
The DSM system is like that of theornithologist. The purpose of the diagnosticinterview is to gather observations,for which the DSM serves as afield guide, organizing the observedfindings by category of illness. But justas in ornithology, over time an experiencedclinician relies less and less onthe specific DSM rules and progressivelymore on pattern recognition. Agood bird watcher distinguishes awoodpecker from a hawk not so muchby specific field signs but from an overallgestalt--rapid synthesis of a wholecomplex of findings (eg, location in thetree, position on the branch, size, shape, color, behavior, call). Likewise, an experiencedclinician might stronglysuspect that a patient has BD based onobservations on an inpatient unit, evenbefore an interview.
Unlike birding, however, experiencegives clinicians the ability to recognizeintermediate cases that blur the boundariesof the classification system.Indeed, this lack of discrete boundarieshas been studied quantitatively in atleast 2 research settings, both of whichshowed a continuum of symptoms inpatients with DSM-diagnosed MDDand BD. Some patients with MDDhave no symptoms suggestive of BD,others have a few such symptoms, andstill others have many with no gaps inthis progression that might serve as anatural point of cleavage between theconditions.4
The same result was obtained byBenazzi5 in a study designed specificallyto look for such gaps. Similarly,Mackinnon and Pies6 recently presentedan elegant model for rapid-cycling andmixed-state BD, demonstrating thatmood, energy, and creativity/speed-ofthoughtsymptoms fluctuating asynchronouslycan create a continuum offorms of the illness. Although thismodel has not yet been subjected torigorous study, it accords well with clinicalimpressions, just as Jamison7summarized over a decade ago: Theclinical reality of manic-depressiveillness is far more lethal and infinitelymore complex than the current psychiatricnomenclature . . . would suggest.
Is the DSM system incorrect, then? IsBD really just one end of a mood spectrumcontinuum? The key word here isreally: notice that again we are presumingto model reality. We are assumingthere is a correct way of explaining,such that there could be overrecognitionand underrecognition of bipolarity.Yet we still lack biologic underpinningsfor these models. A truly validnosologic system will ultimately beanchored by understanding etiology--from genetic susceptibilities throughmolecular differences to anatomicand/or physiologic changes. We arewitnessing exciting progress in thisregard. For example, recent work byTsankova and colleagues8 on regulationof brain-derived neurotrophic factortranscription by histone methylationand acetylation in response to socialdefeat stress further calcifies neuroplasticity,an important bone in theskeleton of mood disorder pathology.
Must we wait for the rest of the skeletonto form to see how bipolar and unipolarmood disorders really are related?(Is there truly a joint in there somewhere?)One hopes not--critical clinicaldecisions rest on these distinctions.Many patients do not respond, or do notsustain their response, to antidepressanttreatment. What percentage of themhave an underlying bipolar componentas the basis for that failed treatment?Who is at risk for a dangerous reactionto antidepressants? How can we recognizesubthreshold bipolarity if, as shownin the most recent iteration of the UnitedStates National Comorbidity Survey, itcan cause as much disability as diabetesor asthma?9
Many have suggested that the DSMshould include a broader range of bipolarity,including Klerman's 10 categoriesI through VI; Akiskal and Pinto's11 Ithrough IV, including 1½, 2½, 3½; andthe recent bipolar spectrum disorderproposed by Ghaemi and colleagues.12A review for the International Societyfor Bipolar Disorders' position paperseries on bipolar diagnosis13 concludesthat 3 groups of patients should beconsidered for such an extension of thecategoric system, including:
Or, might it be better to view bipolarityas an illness dimension, one thatcan be present to varying degrees? Thisapproach has been adopted, at least asa working approach awaiting furtherstudy, by the bipolar clinic at Harvard'steaching hospital, where instead ofasking, Does this patient have bipolardisorder or not? clinicians aretaught to ask How much bipolarity might this patient have?15 Amongother diagnostic tools (available atwww.manicdepressive.org), the clinicuses the Bipolarity Index, a 100-pointscale with 20 points for hypomania ormania, but an additional 80 points forother features (eg, family history, ageat onset, course of illness, response tomedications).16
As we consider revisions to the currentdiagnostic system, we should rememberthe problem we are trying to solve.At present, the DSM system workswell at the extremes of diagnosis:mania identifies BD well when recognizedas such. (Numerous factors canimpede this recognition, such as lackof insight on the part of the patient or lack of sufficient search on the partof the clinician; neither is the faultof the DSM system.) Conversely, anadult with a single episode of majordepression following severe psychosocialstress with no family history ofBD will be widely recognized as havingMDD. These fully expressed, recognizablecases--shown in the Figure (See Psychiatric Times, July 2006, p. 76) aspoints E and A--will be treated withmood stabilizers and antidepressants,respectively.
Point D in the Figure (See Psychiatric Times, July 2006, p. 76) representsidentifiable BD II, with fully expressedhypomania. The DSM system of diagnosisworks reasonably well for such apatient. Our problem comes with intermediateforms. Point C represents subthresholdcases per the NCS-R. Somepsychiatrists would call this BD nototherwise specified (BD NOS; ie, bipolarbut not meeting BD II criteria).However, others would regard this as aform of depression, perhaps atypical butnot bipolar because DSM criteria forhypomania are not met.
The same disagreement arises--andperhaps with more strength--at pointB, where only a few bipolar-minded cliniciansmight persist in calling this BDNOS, and most psychiatrists would callit MDD. Obviously, from a spectrumviewpoint there are almost an infinite variety of other intermediate cases (butlet us not digress into Zeno's paradox!).
Some clinicians are already using aspectrum view of diagnosis, but manyare not. The resulting division of opinionwas demonstrated dramatically ina 2005 Webcast presented by Harvardfaculty, with an audience of hundredsof psychiatrists.17 The program focusedon a patient with clear symptoms of depression,as well as anger, anxiety, sleepproblems, and alcohol use. His case wasclearly selected to present features ofMDD yet suggests possible BD II. Hissymptoms were presented in considerabledetail to focus on this differential.After the Harvard experts concluded infavor of a unipolar diagnosis, the audiencewas asked to voice their opinion--bipolar or unipolar? Two thirds chosebipolar. Not only did they disagree withthe experts, they were substantially dividedamong themselves. What an embarrassmentfor our field.
Notice that on this disagreementhinges a critical, relatively immediatedecision. Suppose a patient presentswith depression. Suppose the importantelements in a standard differential diagnosishave been excluded: substance useor hypothyroidism and other suborganiccauses such as Parkinson disease or B12deficiency (some would include personalitydisorders, but that is a complexissue18). Suppose the patient stronglyprefers a medication approach topsychotherapy and is unlikely to adoptan exercise program as the primary intervention.If the patient is electing apharmacologic approach for managing depression, should the clinician beginwith an antidepressant or consider a moodstabilizer with antidepressant clout?
We can debate whether the linebetween unipolar and bipolar should bedrawn farther to the left on the continuum.We can debate whether there reallyis a continuum. These debates have persistedfor years. (The term soft bipolarspectrum was first used in 1987.19) Perhapsin a few years the DSM-V will includeone more intermediate form (eg,BD III). Yet here we sit in the field ofpsychiatry with a system of diagnosisthat forces us to disagree with oneanother in public. This system mayeven be doing harm by forcing misclassificationof patients as unipolar, leadingto dangerous adverse effects fromantidepressants. All of this, because wepersist in clinging to a categoric systemfor a continuum of symptoms.
To address this dilemma now, insteadof waiting for the DSM-V, we couldreframe our current division of opinionin new, more manageable terms. Thus,when considering a medication approachto depression, which findingswarrant changing our current defaultassumption from antidepressants tomood stabilizers? This reframing alonewill not solve our problem; some psychiatristswould require clear-cut mania orhypomania, setting the bar for this shiftin strategy between points C and D inthe Figure (See Psychiatric Times, July 2006, p. 76). Others would extend use ofmood stabilizers as far as C and otherseven to B. Several well-known researchershave even suggested that somepatients at Point A (ie, no observablesigns or history of hypomania) mightwarrant consideration for mood stabilizers.The bipolar spectrum disorderproposed by Ghaemi and colleagues12suggests that even patients who havenever had hypomania, only multiple softsigns--such as a first degree relativewith BD, depression onset betweenages 15 and 19, and subsequent repeatedbrief episodes--may still have enoughbipolarity to present some risk if treatedwith antidepressants alone.
By comparison, another researchgroup has shown up to 6 months ofimprovement treating BD II with antidepressantmonotherapy, which wouldappear to contradict the concerns aboutthis practice. However, Amsterdam andShults20 did indeed see an increase inmania scores (Young Mania RatingScale [YMRS]) in patients treated withfluoxetine. Although they dismiss it asnot clinically meaningful, the YMRSis not very sensitive to subtle hypomania;thus finding any increase with thisinstrument may warrant concern.
In general practice, however, mostphysicians are lumpers--when usinga medication approach, they treatdepression with antidepressants. Indeed,physicians verge on lumping even bipolar depression into the same approach:as of 2003, as many as 80% of communitypsychiatrists' patients with BDwere taking antidepressants.21 Soperhaps a more operational reframingof the diagnostic dilemma might besummarized by a single question:presuming, as discussed above, that aproper differential diagnosis of depressionhas been conducted and a medicationapproach is now to be initiated, howmany antidepressants should be triedin a patient lacking DSM indications ofBD (and, therefore, unipolar), beforetrying a mood stabilizer?
In current practice, we routinely trymultiple antidepressants, do we not?But for a patient with a family historyof BD, perhaps the transition in strategyshould come earlier. And for apatient with many soft signs of bipolarity,perhaps the transition shouldcome after a single antidepressant trial--or even before. This is the principle valueof the spectrum concept: it makes youthink. It adds bipolarity to the differentialof apparently unipolar depression,even when a thorough search for hypomaniahas yielded little. (Credit for thisconcept goes to Jack Katzow, MD, mycolleague on the International Societyfor Bipolar Disorders' spectrumsubcommittee.)
Should we simply expand the rangeof patients for whom one can rationallyconsider mood stabilizers? All of theatypical antipsychotics have been usedwith some evidence of efficacy,22 butfew formal randomized trials have beenconducted. Interestingly, 2 importantrandomized trials were not published.Data from these studies show that lamotriginefailed to improve the course oftreatment-resistant patients with unipolardisorder,23 even though one wouldthink such a group would represent anenriched sample of soft bipolarity. Thesedata obviously give us pause in consideringthe breadth of applicability of lamotrigine.Furthermore, olanzapine in thispopulation caused the same weight gainproblems we know all too well.24
Clinicians also consider medicationrisks when choosing among options. Inmy opinion, we may be underestimatingthe risk of antidepressants relativeto those of mood stabilizers (a very broadand complex group, none of which meetthe most rigorous definition of thisterm25). When, for a given patient, antidepressantslack efficacy as evidencedby a series of failed but presumably wellconductedtreatments, what should wedo next (besides reconsidering thedifferential yet again)? In my opinion,clinicians--and patients--may be slowto consider mood stabilizers because ofthe greater perceived risks of suchmedications. This deserves study,certainly. In the meantime, let's exam-ine what we know so far about theserelative risks.
The Table compares the short- andlong-term risks (significant implications,not just side effects) of theseagents, focusing on mood stabilizers thathave at least some data supporting antidepressantefficacy.
Every mood stabilizer can, at leastrarely, cause death. Most have severalother significant long-term risks. Bycomparison, antidepressants carry lessdirect risk of death, unless one countsthe hotly debated risk of increasingsuicidality. Instead, a variety of destabilizingeffects have been associatedwith antidepressants, although the dataare scant and debated (for a review,see Phelps26).
Of even greater concern, at leasttheoretically, is the potential for the antidepressantsto worsen an underlyingBD--a process generally referred to askindling. Readers are likely well awarethat this is a model, a framework forthought, not a research-tested concept.Nonetheless, it is an important conceptto consider. Yet kindling concern israrely raised when discussing the useof antidepressants for BD, perhapsbecause the data by which to evaluatethis risk are almost nonexistent.
For example, Grunze27 criticized theAmerican reticence to use antidepressantsin BD, but his review does notmention kindling as a potential risk.By comparison, El-Mallakh andKarippot28 concluded a review on antidepressant use in BD: Long-term usemay destabilize the illness, leading toan increase in the number of bothmanic and depressed episodes; inducerapid cycling (at least 4 episodes ayear); and increase the likelihood of amixed state.
A case report of an apparent antidepressant-induced mood instability in a patient with unipolar disorder following7 years of euthymic response to anantidepressant raises broader concerns,29akin to the acid syndrome (antidepressant-associated chronic irritabledysphoria.)30 Using nationwide registerstudies, Kessing and Andersen31 havedemonstrated that the average risk ofrecurrence increases with the numberof episodes in depressive and bipolaraffective disorders.
But can precipitating mania orhypomania with an antidepressantkindle an underlying bipolarity so asto make the patient's long-term courseworse than it otherwise would havebeen? We must ask ourselves this question,even though at present we haveno systematic studies of such long-termeffects, and such research would betremendously difficult to conduct. Fornow and likely for the foreseeable future(perhaps until we have a biomarkerto follow over time), this leaves usreliant on case reports and clinical suspicions--a very weak basis for complextreatment decisions.
Therefore, a full appraisal of the risksof antidepressants is not possible atpresent. A biomarker of antidepressantexacerbation that could be used to projectrisk would be extremely helpful.Already there is a potential candidate,since inheriting 2 short versions of the serotonin transporter gene may substantiallyincrease this risk (not all studieshave supported the connection, but 2are strongly positive32,33). In the meantime,differences in opinion regardinghow much risk antidepressants pose inpatients with BD might be a significantfactor explaining why some BD specialistsare inclined to avoid them as muchas possible, whereas others feel theyare being underused. Could some--much?--of the debate about the socalledbipolar spectrum boil down toour different perception of antidepressantrisks?
Perhaps we could address that questionthrough attitudinal research.Ironically, psychiatrists' understandingof bipolarity may not be the most importantdeterminant of just which patientswill receive antidepressants--becausewe are not the most frequent antidepressantprescribers.
Rather than psychiatrists or even psychotherapists,primary care providers(PCPs) are clearly the front line ofdepression treatment in the UnitedStates. In a further irony, while psychiatristsare increasingly considering aspectrum approach to bipolar diagnosis,PCPs are being encouraged bycolleagues,34 psychiatrists,35 and thepharmaceutical industry36 to use a categorictool, the Mood Disorders Questionnaire(MDQ).
Although most physicians knew thisonce, many have forgotten that accurateuse (the predictive value) of atest like the MDQ depends heavily onclinicians' pretest hunch (ie, the priorprobability of BD).37 Yet developing adiscriminating hunch comes from understandingbipolar presentations, fromBD I mania through hypomania tothe 11 soft signs that form the criteriafor bipolar spectrum disorder and theBipolarity Index, as previously discussed.Such training is rare for PCPs,although an online primer has beenavailable and refined over the past 5 years.37 PCPs are besieged by variousefforts to educate them on many topics,and thus the extent to which evencontinuing medical education can reachthem is limited. When we have theirear, as when discussing a consultation,we can teach them to systematicallyassess 4 areas they already query, inaddition to the fifth dimension of bipolaritythey will see reflected in theMDQ (hypomania):
To summarize for them the spectrumperspective in concrete, applicableterms, we could suggest (after reviewingwith them the full differential) thatthe more bipolar soft signs one detects,the earlier one should consider switchingfrom an antidepressant-basedstrategy to a mood-stabilizer-basedstrategy. However, we must admit thatthere is no agreed-on or empiricallyestablished cutoff point to guide thisdecision, and indeed there may not besuch a point of natural cleavage: bipolardisorder may not be a species, as inornithology, despite the implications ofthe DSM.
Despite the advantages of a spectrumview, particularly in matching clinicalexperience, the DSM's, categoric approach,with its greater internal structureand validated diagnoses, is also avaluable part of our clinical thinking.We can adopt an approach long used inphysics, where 2 apparently exclusiveways of viewing the electromagneticspectrum are used simultaneously: lightis a wave and a particle.
For psychiatrists, this means usingthe DSMand the spectrum views in complementaryfashion, like bifocals in apair of glasses--carrying both lenseswith us, switching back and forth to findthe one that best brings an individualpatient into focus. Our patients haveillnesses that are both waves--varyingcontinuously over a spectrum--andparticles--with discrete characteristicsthat can help us identify who islikely to do well with an antidepressantand who merits caution. As in physics,complementarity helps manage uncertainty.This dual-lens perspective mightallow the focus of our debate to shifttoward further identifying characteristicsthat predict adverse reactions toantidepressants (away from debatesabout the breadth of bipolarity). So farthese factors include 11 soft signs.11
Many of our primary care colleagues,however, are struggling to identify anynonmanic version of BD, such as BDII. For them, the MDQ as a screeningtool is probably better than no screeningtool at all. However, there will besignificant misclassification unless theyare quickly educated on bipolar phenomenology,which must be used to anchor their interpretation of the MDQ results.Other PCPs are ready to move beyonda categoric view; they need a cautiousde-emphasis of the MDQ, with increasedemphasis on recognizing intermediatebipolar spectrum variations.
This discussion will be continuedin a second article to be published in afuture issue of Psychiatric Times.
Dr Phelps has been practicing psychiatry formore than 15 years and specializes in treatingbipolar disorder. He recently published WhyAm I Still Depressed? Recognizing andManaging the Ups and Downs of Bipolar IIand Soft Bipolar Disorder.
Dr Phelps states that he has received grantsand honoraria from GlaxoSmithKline, Astra-Zeneca, and Abbott Laboratories.
Divalproex sodium (Depakote)
Lithium (Eskalith, Lithobid)
1. Patten SB. Does almost everybody suffer from abipolar disorder? Can J Psychiatry. 2006;51:6-8.
2. Baldessarini RJ. A plea for integrity of the bipolardisorder concept. Bipolar Disord. 2000;2:3-7.
3. Ghaemi SN, Sachs GS, Chiou AM, et al. Is bipolardisorder still underdiagnosed? Are antidepressantsoverutilized? J Affect Disord. 1999;52:135-144.
4. Cassano GB, Rucci P, Frank E, et al. The moodspectrum in unipolar and bipolar disorder: argumentsfor a unitary approach. Am J Psychiatry. 2004;161:1264-1269.
5. Benazzi F. Bipolar II disorder and major depressivedisorder: continuity or discontinuity? World JBiol Psychiatry. 2003;4:166-171.
6. Mackinnon DF, Pies R. Affective instability as rapidcycling: theoretical and clinical implications forborderline personality and bipolar spectrum disorders.Bipolar Disord. 2006;8:1-14.
7. Jamison KJ. Touched With Fire. New York: Simonand Schuster; 1993.
8. Tsankova NM, Berton O, Renthal W, et al. Sustainedhippocampal chromatin regulation in a mouse modelof depression and antidepressant action. NatNeurosci. 2006;9:519-525.
9. Kessler R. Prevalence and effects of mood disorderson role performance in the United States.Presented at: International Society of Bipolar DisorderAnnual Meeting; June 2005; Pittsburgh.
10. Klerman GL. The spectrum of mania. ComprPsychiatry. 1981;22:11-20.
11. Akiskal HS, Pinto O. The evolving bipolar spectrum.Prototypes I, II, III, and IV. Psychiatr Clin NorthAm. 1999;22:517-534.
12. Ghaemi SN, Ko JY, Goodwin FK. "Cade's disease"and beyond: misdiagnosis, antidepressant use, anda proposed definition for bipolar spectrum disorder.Can J Psychiatry. 2002;47:125-134.
13. Phelps JR, Angst J, Katzow J, Sadler JZ. Validityand utility of the bipolar spectrum model. BipolarDisorders. In press.
14. Angst J, Gamma A, Benazzi F, et al. Toward a redefinitionof subthreshold bipolarity: epidemiologyand proposed criteria for bipolar-II, minor bipolardisorders and hypomania. J Affect Disord. 2003;73:133-146.
15. Saenger E. The Bipolarity Index as a toolfor assessment and creating rapport: an expertinterview with Gary Sachs, MD. MedscapePsychiatry & Mental Health. 2005. Available at:http://www.medscape.com/viewarticle/503893.Accessed March 28, 2006.
16. Massachusetts General Hospital, Bipolar ClinicWeb site. Affective Disorders Evaluation. Availableat: http://www.manicdepressive.org/images/blankade.pdf, Accessed March 28, 2006.
17. Massachusetts General Hospital, PsychiatryAcademy. Mood Disorders at the Interface of ClinicalPractice and Emerging Genetics. Module 1; Jan2005. Available at: http://www.mghcme.com/show.event.php?s=psychlink_module1_011205_details&id=5. Accessed March 28, 2006.
18. Phelps JR. Bipolar "Versus" Borderline. Availableat: http://www.psycheducation.org/depression/borderline.htm. Accessed April 28, 2006.
19. Akiskal HS, Mallya G. Criteria for the "soft" bipolarspectrum: treatment implications. PsychopharmacolBull. 1987;23:68-73.
20. Amsterdam JD, Shults J. Fluoxetine monotherapyof bipolar type II and bipolar NOS major depression:a double-blind, placebo-substitution,continuation study. Int Clin Psychopharmacol. 2005;20:257-264.
21. Ghaemi SN, Hsu DJ, Soldani F, Goodwin FK.Antidepressants in bipolar disorder: the case forcaution. Bipolar Disord. 2003;5:421-433.
22. Barbee JG, Conrad EJ, Jamhour NJ. The effectivenessof olanzapine, risperidone, quetiapine, andziprasidone as augmentation agents in treatmentresistantmajor depressive disorder. J Clin Psychiatry.2004;65:975-981.
23. GSK Trial Register, Lamotrigine Studies #SCA20022,SCA20025. Available at: http://ctr.gsk.co.uk/Summary/lamotrigine/studylist.asp. Accessed April30, 2006.
24. Andersen SW, Clemow DB, Corya SA. Long-termweight gain in patients treated with open-label olanzapinein combination with fluoxetine for majordepressive disorder. J Clin Psychiatry. 2005;66:1468-1476.
25. Bauer MS, Mitchner L. What is a "mood stabilizer"?An evidence-based response. Am J Psychiatry.2004;161:3-18.
26. Phelps J. Antidepressants in Bipolar Disorder:The Controversies. 2006. Available at: http://www.psycheducation.org/bipolar/controversy.htm.Accessed May 11, 2006.
27. Grunze H. Reevaluating therapies for bipolardepression. J Clin Psychiatry. 2005;66:17-25.
28. El-Mallakh RS, Karippot A. Use of antidepressantsto treat depression in bipolar disorder. PsychiatrServ. 2002;253:580-584.
29. Phelps JR. Agitated dysphoria after late-onsetloss of response to antidepressants: a case report.J Affect Disord. 2005;86:277-280.
30. El-Mallakh RS, Karippot A. Antidepressant-associatedchronic irritable dysphoria (acid) in bipolardisorder: a case series. J Affect Disord. 2005;84:267-272.
31. Kessing LV, Andersen PK. Predictive effects ofprevious episodes on the risk of recurrence in depressiveand bipolar disorders. Curr Psychiatry Rep.2005;7:413-420.
32. Masoliver E, Menoyo A, Perez V, et al. Serotonintransporter linked promoter (polymorphism) in theserotonin transporter gene may be associated withantidepressant-induced mania in bipolar disorder.Psychiatr Genet. 2006;16:25-29.
33. Mundo E, Walker M, Cate T, et al. The role ofserotonin transporter protein gene in antidepressantinducedmania in bipolar disorder: preliminary findings.Arch Gen Psychiatry. 2001;58:539-544.
34. Kaye NS. Is your depressed patient bipolar?J Am Board Fam Pract. 2005;18:271-281.
35. Citrome L, Goldberg JF. The many faces of bipolardisorder. How to tell them apart. Postgrad Med.2005;117:15-16, 19-23.
36. AstraZeneca Pharmaceuticals. Resources foran Accurate Diagnosis. Available at: www.IsItReallyDepression.com. Accessed March 28, 2006.
37. Phelps JR, Ghaemi SN. Improving the diagnosisof bipolar disorder: Predictive value of screeningtests. J Affect Disord. 2006;92:141-148.
Cassano GB, Rucci P, Frank E, et al. The mood spectrumin unipolar and bipolar disorder: argumentsfor a unitary approach. Am J Psychiatry. 2004;161:1264-1269.
Ghaemi SN, Ko JY, Goodwin FK. Cade's diseaseand beyond: misdiagnosis, antidepressant use, anda proposed definition for bipolar spectrum disorder.Can J Psychiatry. 2002;47:125-134.