Recognition of irrationality of fear and the diagnosis of social anxiety disorder and specific phobia in adults: Implications for criteria revision in DSM-5

Research Article

DEPRESSION AND ANXIETY 27 : 1044–1049 (2010)

RECOGNITION OF IRRATIONALITY OF FEARAND THE DIAGNOSIS OF SOCIAL ANXIETY DISORDER

AND SPECIFIC PHOBIA IN ADULTS: IMPLICATIONSFOR CRITERIA REVISION IN DSM-5

Mark Zimmerman, M.D.,1,2� Kristy Dalrymple, Ph.D.,1,2 Iwona Chelminski, Ph.D.,1,2 Diane Young, Ph.D.,1,2

and Janine N. Galione, B.S.1,2

Background: In DSM-IV, the diagnosis of social anxiety disorder (SAD) andspecific phobia in adults requires that the person recognize that his or her fear ofthe phobic situation is excessive or unreasonable (criterion C). The DSM-5Anxiety Disorders Work Group has proposed replacing this criterion becausesome patients with clinically significant phobic fears do not recognize theirrationality of their fears. In the present report from the Rhode Island Methodsto Improve Diagnostic Assessment and Services project we determined thenumber of individuals who were not diagnosed with SAD and specific phobiabecause they did not recognize the excessiveness or irrationality of their fear.Methods: We interviewed 3,000 psychiatric outpatients and 1,800 candidatesfor bariatric surgery with a modified version of the Structured ClinicalInterview for DSM-IV. In the SAD and specific phobia modules we suspendedthe skip-out that curtails the modules if criterion C is not met. Patients who metall DSM-IV criteria for SAD or specific phobia except criterion C wereconsidered to have ‘‘modified’’ SAD or specific phobia. Results: The lifetimerates of DSM-IV SAD and specific phobia were 30.5 and 11.8% in psychiatricpatients and 11.7 and 10.2% in bariatric surgery candidates, respectively. Lessthan 1% of the patients in both samples were diagnosed with modified SAD orspecific phobia. Conclusion: Few patients were excluded from a phobiadiagnosis because of criterion C. We suggest that in DSM-5 this criterion beeliminated from the SAD and specific phobia criteria sets. Depression andAnxiety 27:1044–1049, 2010. r 2010 Wiley-Liss, Inc.

Key words: social anxiety disorder; specific phobia; DSM-5; diagnostic criteria;classification

INTRODUCTIONClassification is the process by which the complexityof phenomena is reduced by arranging them intocategories according to some established criteria forone or more purposes. At present, the classification ofmental disorders consists of specific mental disordersthat are grouped into various classes on the basis ofsome shared phenomenological characteristics. Theultimate purpose of classification is to improve treat-ment and prevention efforts.[1] Ideally, a classificationof disorders is based on knowledge of etiology or

Published online 23 June 2010 in Wiley Online Library (wiley

onlinelibrary.com).

DOI 10.1002/da.20716

Received for publication 25 March 2010; Revised 5 May 2010;

Accepted 5 May 2010

�Correspondence to: Mark Zimmerman, Bayside Medical Center,

235 Plain Street, Providence, RI 02905.

E-mail: [email protected]

The authors report they have no financial relationships within the

past 3 years to disclose.

1The Department of Psychiatry and Human Behavior, Brown

Medical School, Providence, Rhode Island2The Department of Psychiatry, Rhode Island Hospital,

Providence, Rhode Island

rr 2010 Wiley-Liss, Inc.

pathophysiology because this increases the likelihoodof improving treatment and prevention efforts. In theabsence of knowledge of pathophysiology, a phenom-enology-based classification may nonetheless guidetreatment to the extent that it validly identifies clinicalentities. Another clinical purpose served by classifica-tion is communication, both with patients and betweenclinicians. To facilitate these goals, a classificationsystem should not be overly burdensome to apply.[2]

Published in 1980, DSM-III was the first officialdiagnostic system to specify inclusion and exclusiondiagnostic criteria. DSM-III followed the precedentsset forth in the Washington University criteria[3] andthe Research Diagnostic Criteria[4] for defining dis-orders, and brought the reliable diagnostic approachused by a few research groups to the clinical community.The specification of diagnostic criteria enabled study ofthe boundaries between disorders and between dis-order and no disorder. Thus, the validity of thediagnostic criteria could be evaluated scientifically. Infact, it was assumed when DSM-III was published thatchanges would be made, and it was anticipated thatthese changes would follow scientific study rather thanideological debate. Such has been the case, at least inpart, with the subsequent publications of DSM-III-Rand DSM-IV, and is expected to be the case in thesoon-to-be published DSM-5. This study was con-ducted in this spirit—to empirically examine one of theproposed changes delineated in the draft criteria forDSM-5.

According to DSM-IV, the diagnosis of social anxietydisorder (SAD) and specific phobia in adults requiresthat the person recognize that his or her fear of thephobic situation is excessive or unreasonable (criterion C).The DSM-5 Anxiety Disorders Work Group hasproposed replacing this criterion with a clinician’sjudgment that the fear is ‘‘out of proportion to actualdanger posed by the situation.’’ In discussing the reasonfor this change for the SAD criterion set, Bogels et al.[5]

indicated that it was ‘‘our clinical impressionythatsome adult SAD patients do not recognize the‘‘irrationality’’ of their fears.’’ Likewise, in discussinga similarly proposed change for specific phobia,LeBeau et al.[6] indicated that clinical judgment waspreferable to self-judgment because of evidence thatphobias are frequently underdiagnosed in elderlyindividuals who overattribute their fears to age-relatedconstraints, and clinical experience that some indivi-duals lack insight into the excessiveness or unreason-ableness of their fears. In neither review was empiricalevidence presented regarding the frequency of false-negative diagnoses resulting from this criterion therebywarranting the change.

The insight criterion for SAD and specific phobiahas changed through the iterations of DSM-III, DSM-III-R, and DSM-IV. In DSM-III, criterion B requiredthe presence of significant distress and recognition thatthe fear was excessive. Thus, the insight criterion wasnot a stand-alone criterion. Criterion C excluded the

diagnosis if it was due to another disorder, and indiscussing the differential diagnosis of specific phobia(called simple phobia in DSM-III), the text noted thatin patients with schizophrenia the diagnosis should beexcluded if activities are avoided in response todelusions.

In DSM-III-R the insight criterion stood alone as itsown separate criterion. For specific phobia (calledsimple phobia in DSM-III-R), the exclusion due toother disorders was changed to ‘‘The phobic stimulus isunrelated to the content of the obsessions of obsessivecompulsive disorder or the trauma of post-traumaticstress disorder.’’ The text, however, again noted thatthe diagnosis would be excluded in patients withschizophrenia who avoided certain activities inresponse to delusions. The implementation of exclu-sion due to psychosis was with the insight criterion.Thus, in DSM-III psychotic patients could be excludedfrom the diagnosis of specific phobia in two ways—failing to meet the insight criterion as well as exclusiondue to another mental disorder. In DSM-III-Rpsychotic patients could be excluded solely on thebasis of the insight criterion. Regarding SAD, in DSM-III-R the insight criterion was a stand-alone criterion,and there was a separate, differently worded criterionexcluding the diagnosis if the social anxiety was relatedto another psychiatric or medical disorder. There wasno discussion in the text of DSM-III-R SAD section ofhow the insight criterion would reduce false-positivediagnoses. Rather the text simply stated: ‘‘Invariablythe person recognizes that his or her fear is excessive orunreasonable.’’

In DSM-IV the insight criterion for specific phobiawas not changed from DSM-III-R. However, theexclusion criterion related to the presence of otherdisorders was broadened and thus similar to theDSM-III exclusionary rule, though the wording ofthe criterion was different. The text again discusseddifferential diagnosis in patients with psychotic dis-orders, and provided a specific example of a case inwhich specific phobia would not be diagnosed insomeone with delusional disorder. For SAD, theinsight criterion was not changed from DSM-III-R.The exclusion criterion was also similar to DSM-III-R,though reworded. For the first time, the text discusseddifferential diagnosis of SAD in psychotic patients, andprovided a specific example of a case in which SADwould not be diagnosed due to lack of insight.

Through the last three editions of the DSM some-what different approaches have been taken to excludethe diagnosis of SAD and specific phobia because thefears are attributable to another psychiatric disorder.The draft criteria for DSM-5 suggest yet anotherchange, though we are not aware of any researchexamining the impact of this criterion on diagnosis. Inthe present report from the Rhode Island Methods toImprove Diagnostic Assessment and Services (MIDAS)project, we determined the number of individuals whowere not diagnosed with SAD and specific phobia

1045Research Article: Recognition of Irrationality of Fear

Depression and Anxiety

because they did not recognize the excessiveness orirrationality of their fear. We examined false-negativerates attributable to criterion C in a large sample ofpsychiatric outpatients presenting for treatment and asample of candidates for bariatric surgery. We includedthe bariatric sample as a quasi-general populationsample because these individuals were not seekingpsychiatric care. It is possible that the insight criterionwould have a greater impact in nontreatment-seekingsamples. We also tested the hypotheses that lack ofinsight into the excessiveness or reasonableness of fearwould be associated with older age and a diagnosis of apsychotic disorder.

METHODS

The Rhode Island MIDAS project represents an integration ofresearch methodology into a community-based outpatient practiceaffiliated with an academic medical center.[7–9] To date, 3,000psychiatric outpatients have been evaluated with a semi-structureddiagnostic interview in the Rhode Island Hospital Department ofPsychiatry outpatient practice. This private practice group predomi-nantly treats individuals with medical insurance (including Medicarebut not Medicaid) on a fee-for-service basis, and it is distinct from thehospital’s outpatient residency training clinic that predominantlyserves lower income, uninsured, and medical assistance patients. Thesource of referral was recorded for the last 1,200 patients. Patientswere most frequently referred from primary-care physicians (31.6%),psychotherapists (15.8%), and family members or friends (17.6%).The Rhode Island Hospital institutional review committee approvedthe research protocol, and all patients provided informed, writtenconsent.

The data in Table 1 show that the majority of the patients werewhite, female, married or single, and graduated high school. The data

in Table 2 show that the most frequent lifetime DSM-IV diagnoseswere major depressive disorder, alcohol use disorders, and SAD.

The second sample consisted of 1,800 candidates for bariatricsurgery who were evaluated with the same interview schedule as thepsychiatric patients. The majority of the patients were white, female,married or single, and graduated high school (Table 1). The mostfrequent lifetime DSM-IV diagnoses were major depressive disorder,alcohol use disorders, and eating disorders (Table 2).

All patients were interviewed by a diagnostic rater who adminis-tered a modified version of the Structured Clinical Interview forDSM-IV (SCID).[10] Because we were interested in the psychometricperformance of different criteria sets, we modified the SCID andeliminated certain skip-outs. Of relevance to the current report, inthe SAD and specific phobia modules we suspended the skip-out thatcurtails the modules if criterion C is not met. Patients who met allDSM-IV criteria for SAD or specific phobia except criterion C wereconsidered to have ‘‘modified’’ SAD or specific phobia.

The diagnostic raters were highly trained and monitoredthroughout the project to minimize rater drift. Diagnostic ratersincluded Ph.D. level psychologists and research assistants withcollege degrees in the social or biological sciences. Research assistantsreceived three to four months of training during which they observedat least 20 interviews, and they were observed and supervised in theiradministration of more than 20 evaluations. Psychologists onlyobserved 5 interviews; however, they, too, were observed andsupervised in their administration of 15–20 evaluations. During thecourse of training the senior author met with each rater to review theinterpretation of every item on the SCID. Also during training everyinterview was reviewed on an item-by-item basis by the senior raterwho observed the evaluation. At the end of the training period theraters were required to demonstrate exact, or near exact, agreementwith a senior diagnostician on five consecutive evaluations. Through-out the MIDAS project ongoing supervision of the raters consisted ofweekly diagnostic case conferences involving all members of theteam. Written reports of all cases were reviewed by M. Z., who alsoreviewed the item ratings of every case.

TABLE 1. Demographic characteristics of psychiatric outpatients and surgical candidates

Psychiatric outpatients (n 5 3,000) Candidates for bariatric surgery (n 5 1,800)

Characteristic N % N %

GenderFemale 1,818 60.6 1,489 82.7Male 1,182 39.4 311 17.3

Educationo12 years 273 9.1 113 6.3High school graduate or GED 1,865 62.2 1,311 72.8College graduate 862 28.7 376 20.9

Marital statusMarried 1,231 41.0 957 53.2Living with someone 170 5.7 122 6.8Widowed 53 1.8 39 2.2Separated 159 5.3 47 2.6Divorced 439 14.6 264 14.7Never married 948 31.6 371 20.6

RaceWhite 2,622 87.4 1,455 80.8Black 135 4.5 138 7.7Hispanic 77 2.6 109 6.1Asian 28 0.9 5 0.3Other 138 4.6 93 5.2

Age (years) M 5 38.5 SD 5 13.0 M 5 42.3 SD 5 10.9

1046 Zimmerman et al.


As an ongoing part of the MIDAS project, joint-interviewdiagnostic reliability information was collected on 65 psychiatricpatients. Of relevance to the current study, diagnostic reliability washigh for both SAD (k 5 .84) and specific phobia (k 5 .93). In thecandidates for bariatric surgery, reliability was evaluated in 63patients. Again, reliability was high for both SAD (k 5 .79) andspecific phobia (k 5 .82).

RESULTSThe data in Table 3 show that less than 1% of the

3,000 psychiatric outpatients were diagnosed with alifetime history of modified SAD or modified specificphobia, much lower than the lifetime rates of DSM-IVsocial and specific phobia. Only 2.1% (20/936) of thepatients who met lifetime criteria of DSM-IV ormodified SAD did not meet criterion C. For specificphobia, 4.1% (15/369) of the combined group did not

meet criterion C. The results were similar for currentdiagnoses. That is, less than 1% of the patients werediagnosed with modified SAD or modified specificphobia. Again, only a minority of the patients meetingthe DSM-IV or modified criteria did not meet criterionC (SAD—2.0% (17/837); specific phobia—4.2% (14/336)). We tested the hypothesis that age would beassociated with the exclusion criterion by comparingthe mean age of patients who met the lifetime DSM-IVand modified criteria and did not find a significantdifference (SAD: 37.3712.2 versus 41.3712.1,t 5�1.5, not significant; specific phobia: 38.0712.2versus 39.7714.7, t 5�0.5, not significant).

In the text of DSM-III, DSM-III-R, and DSM-IVthe insight criterion has been discussed almostexclusively in reference to psychotic patients. Wetherefore compared the frequency of modified SADand modified specific phobia in patients with and

TABLE 2. Lifetime DSM-IV Axis I diagnoses of 3,000 psychiatric outpatients and 1,800 candidates for bariatric surgery


DSM-IV diagnosis N % N %

Major depressive disorder 1,933 64.4 524 29.1Bipolar disorder 226 7.5 27 1.5Dysthymic disorder 292 9.7 61 3.4Generalized anxiety disorder 555 18.5 55 3.1Panic disorder 745 24.8 190 10.6Social anxiety disorder 916 30.5 210 11.7Specific phobia 354 11.8 184 10.2Obsessive–compulsive disorder 269 9.0 35 1.9Posttraumatic stress disorder 609 20.3 145 8.1Adjustment disorder 215 7.2 175 9.7Psychotic disorder 82 2.7 5 0.3Eating disorder 389 13.0 212 11.8Alcohol abuse/dependence 1,205 40.2 328 18.2Drug abuse/dependence 768 25.6 154 8.6Somatoform disorder 247 8.2 18 1.0Attention deficit disorder 225 7.5 20 1.1Impulse control disorder 597 19.9 141 7.8

TABLE 3. Current and lifetime diagnoses of social anxiety disorder and specific phobia with and without criterion C in3,000 psychiatric outpatients and 1,800 candidates for bariatric surgery


N % N %

Lifetime diagnosisDSM-IV social anxiety disorder 916 30.5 210 11.7Modified social anxiety disorder 20 0.7 8 0.4DSM-IV specific phobia 354 11.8 184 10.2Modified specific phobia 15 0.5 5 0.3

Current diagnosisDSM-IV social anxiety disorder 820 27.3 135 7.5Modified social anxiety disorder 17 0.6 4 0.2DSM-IV specific phobia 322 10.7 152 8.4Modified specific phobia 14 0.5 5 0.3



without a lifetime history of psychotic disorder. Therewere no significant differences between groups (mod-ified SAD: 1.2% versus 0.7%, w2 5 .4, n.s.; modifiedspecific phobia: 1.2% versus 0.5%, w2 5 .9, n.s.).

The data in Table 3 show that the results were similarin the sample of 1,800 candidates for bariatric surgery.That is, for both lifetime and current diagnoses only asmall number of patients were diagnosed with modifiedSAD or specific phobia, and only a minority of patientsmeeting the DSM-IV or the modified criteria didnot meet criterion C (lifetime diagnoses: SAD—3.7%(8/218), specific phobia—2.6% (5/189); current diag-noses: SAD—2.9% (4/139), specific phobia—3.2%(5/157)). Too few patients were diagnosed with modifiedSAD or specific phobia to compare the mean age of thisgroup to the patients meeting DSM-IV criteria.

DISCUSSIONIn two large samples, one of psychiatric outpatients

and the other of candidates for bariatric surgery, fewpatients were excluded from the DSM-IV diagnosis ofSAD and specific phobia because they did notrecognize that their fear of the phobic situation wasexcessive or unreasonable. Thus, criterion C had aminimal contribution to diagnosis. In contrast to theobservations of other authors,[6] patients who met allcriteria except criterion C were not significantly olderthan patients who met the full DSM-IV criteria.

Through the last three versions of the DSM, theinsight criterion, and the exclusion criterion due otherpsychiatric disorders, has undergone repeated changeswith no discussion of the reason for these changes andno research demonstrating the extent of the problemthat the changes were designed to fix. The draft versionfor DSM-5 recommends yet another change—repla-cing criterion C with another criterion based onclinicians’ judgments of excessiveness or unreasonable-ness, operationalized as ‘‘the fear or anxiety is outof proportion to the actual danger posed by thespecific object or situation.’’ The suggested reason forthis change was the clinical observation that somepatients who should be diagnosed with a phobia did notmeet the diagnostic criteria because they did notrecognize the excessiveness or unreasonableness oftheir fear. That is, criterion C created a false-negativeproblem.

The results of this study found that few individualswere excluded from the diagnosis of SAD or specificphobia because of criterion C. Because of the lowfrequency of this phenomenon, one interpretation ofthese results is that the criterion should be left alone.However, this means that the low rate of false negativeswill persist. An alternative recommendation is tosimply delete the criterion. Our results indicate thatthe criterion has little impact on diagnosis and there-fore is not necessary. Because there is a risk that analternative wording of the criterion might haveunintended consequences on diagnostic prevalence,[11]

the most conservative approaches would be to simplyretain the criterion and accept a low frequency of falsenegatives or to eliminate the criterion.

It is not clear to us why a separate insight criterion Cis needed along with a criterion that excludesthe diagnosis of specific phobia or SAD when thefear might be accounted for by another disorder.A psychotic patient who expresses a phobic-like fearcould be excluded from the phobia diagnosis becausethe symptoms are attributable to another psychiatricdisorder. Elimination of the insight criterion willeliminate the false-negative problem, and should notcreate a false-positive problem because of the psychia-tric exclusion criterion. In this study we did not findevidence that modified SAD and specific phobia wasassociated with a psychotic disorder diagnosis.

We believe, however, that it is premature to change,or eliminate, the insight criterion on the basis of asingle study. The empirical basis for a change indiagnostic criteria should be based on replicatedresearch. On the other hand, elimination of thecriterion could be justified on conceptual grounds.That is, an appropriately worded psychiatric exclusionshould be sufficient to exclude the diagnosis in patientswhose phobic anxiety is related to the content ofdelusional thinking.

A limitation of this study is that it was based on asingle site in which the majority of the patients werewhite, female, and had health insurance. Additionalstudy in samples with different demographic andclinical profiles, and in general population communitysamples, is warranted. Particularly important will be alarge-scale study of the insight criterion in patientswith psychotic disorders.

REFERENCES1. Zimmerman M, Spitzer R. Psychiatric classification. In:

Sadock BJ, Sadock VA, editors. Kaplan and Sadock’s Compre-hensive Textbook of Psychiatry. Philadelphia, Lippincott:Williams, & Wilkins; 2005:1003–1034.

2. First M, Pincus H, Levine J, et al. Clinical utility as a criterionfor revising psychiatric diagnoses. Am J Psychiatry 2004;161:946–954.

3. Feighner JP, Robins E, Guze SB, et al. Diagnostic criteria for usein psychiatric research. Arch Gen Psychiatry 1972;26:57–67.

4. Spitzer RL, Endicott J, Robins E. Research diagnosticcriteria: rationale and reliability. Arch Gen Psychiatry 1978;35:773–782.

5. Bogels SM, Alden L, Beidel DC, et al. Social anxiety disorder:questions and answers for the DSM-V. Depress Anxiety2010;27:168–189.

6. LeBeau RT, Glenn D, Liao B, et al. Specific phobia: a review ofDSM-IV specific phobia and preliminary recommendations forDSM-V. Depress Anxiety 2010;27:148–167.

7. Posternak MA, Zimmerman M, Solomon DA. Integratingoutcomes research into clinical practice: a pilot study. PsychiatrServ 2002;53:335–336.

8. Zimmerman M. Integrating the assessment methods of research-ers in routine clinical practice: the Rhode Island Methods toImprove Diagnostic Assessment and Services (MIDAS) project.

1048 Zimmerman et al.


In: First M, editor. Standardized Evaluation in Clinical Practice.Vol. 22. Washington, DC: American Psychiatric Publishing, Inc.;2003:29–74.

9. Zimmerman M, Mattia JI, Posternak MA. Are subjects inpharmacological treatment trials of depression representative ofpatients in routine clinical practice? Am J Psychiatry2002;159:469–473.

10. First MB, Spitzer RL, Gibbon M, Williams JBW. StructuredClinical Interview for DSM -IV Axis I Disorders—patient edition(SCID-I/P, version 2.0). New York, Biometrics ResearchDepartment: New York State Psychiatric Institute; 1995.

11. First MB, Frances A. Issues for DSM-V: unintended conse-quences of small changes: the case of paraphilias. Am J Psychiatry2008;165:1240–1241.



Documents

Recognition of irrationality of fear and the diagnosis of social anxiety disorder and specific phobia in adults: Implications for criteria revision in DSM-5