Pica and the obsessive­compulsive spectrumdisordersDan J. Stein. Colin Bouwer. Ben van Heerden

Background. The concept of a spectrum of obsessive­

compulsive related disorders may have clinical and

research heuristic value in the approach to disorders

similar to obsessive-compulsive disorder (OCO) in respect

of phenomenology and psychobiology. Uke other

repetitive and rituaJistic behaviours, pica may be

postulated to fall at times on this spectrum.

Methods. Five cases of pica seen at our clinics are

presented here in order to test this hypothesis.

Phenomenology. neurobiology (where available) and

pharmacotherapy data are provided in order to consider a

possible relationship with aGO and QCD spectrum

disorders.

Results. In 2 of the cases, pica appeared to be a

compulsion and patients had additional symptoms which

met diagnostic criteria for QGD. In 2 of the cases, the

clinical picture and neurobiological data were reminiscent

of an impUlse control disorder. Four of the 5 patients

responded to treatment with a serotonin re-uptake

inhibitor (SRI).

Conclusion, These results are consistent with a

hypothesis that at least some cases of pica may usefully

be conceptualised as lying within a compulsive-impulsive

spectrum of symptoms and disorders.

S Afr Med J 1996; 86; 1586-1592_

Recent advances in the study of obsessive-compulsivedisorder (QCD) include rigorous characterisation of itsphenomenology, gradual delineation of the mediating neuro­anatomy and neurochemistry, and the development ofeffective pharmacotherapeutic and psychotherapeuticinterventions.\1- Such successes have also led to increasedattention to a range of disorders (including Tourette'ssyndrome,J hypochondriasis~ and trichotillomania5) wherephenomenology is reminiscent of QCD, and whereneurobiology and treatment response might also be thoughtto overlap with QCD. The concept of an obsessive­compulsive spectrum of disorders may have heuristic valuein both clinical and research settings.&.1

The notion of an obsessive-compulsive spectrum ofdisorders has, for example, contributed to current

Departments of Psychiatry and Nuclear Medicine, University ofStellenbosch, Tygerber9, W. Cape

Can J. Stein. M.a CH.B" FR.C.PC.

Colin BouWE!I'. M_B. CH,B.• "'-""EO (PSYCH.)

Ben van Heenten. M.B. CH.B_ M.MED (l/'IT. MEDJ. M.SC.

perspectives on a number of previously under-researchedconditions. Body dysmorphic disorder, which ischaracterised by obsessive concerns about somatic featuresand by compulsive checking of the body, has thereforerecently been postulated to respond selectively to treatmentwith serotonin re-uptake inhibitors (SRls), as does QCD.8

Similarly, hair-pulling and severe nail-biting are self-injuriousbehaviours in which there may be ritualistic features, andcontrolled trials have now demonstrated that these respondselectively to the SRls,i.,O giving many patients with chronicsymptoms hope for an effective treatment.

Pica has been defined as the persistent eating of non­nutritive substances. l1 The Diagnostic and Statistical Manualof Mental Disorders (DSM-IV) states that the typicalsubstance ingested tends to vary with age. Infants andyounger children may eat paint, plaster, string, hair or cloth;older children may eat animal droppings, sand, insects,leaves or pebbles; and adolescents and adults may eat clayor soil. Pica has been documented in association withpregnancy '2 as well as in association with mentalretardation,13 schizophrenia, 1~ autism,15 and other psychiatricdisorders. '8 In addition, pica may be either a manifestation ofiron deficiency, or a factor predisposing thereto. '0.17

Pica has not, however, often been linked to QCD. Zeitlinand POlivy'8 presented a patient with coprophagia, andstated that this patient's symptoms met DSM-JV criteria forQCD and that these responded to formal behaviour therapy.However, the patient had no other symptoms of QCD andindeed the pica symptoms were closely related to a pasthistory of abuse. Szabo et al. '9 recently presented a case ofpica thought to be secondary to iron-deficiency anaemia,and made the interesting speculation that pica may lie onthe QCO spectrum of disorders. They did not, however,provide any neurobiological or pharmacological data fromtheir case to support or reMe this contention.

In this paper, we review the case material of 5 picapatients seen at our clinics in order to shed light on thehypothesis that pica lies within the obsessive-compulsivespectrum of disorders.

MethodsFive patients who presented to our outpatient anxietydisorder and QCD clinics with pica are described.Phenomenology, neurobiology (where available) andpharmacotherapy data are provided in order to consider apossible relationship with OCD and aCD spectrumdisorders.

Case 1An 18-year-old single woman presented with a history ofQGD symptoms that had begun in early childhood.Symptoms included intrusive sexual thoughts, ritualisticmental reviewing and concerns with symmetry. Co-morbiddisorders at the time of presentation included panic disorderwith agoraphobia, social phobia and pica. Pica comprisedeating of sand, which the patient described as one of therepetitive behaviours that she felt compelled to carry out.The patient's sister also had a history of panic disorder withagoraphobia, social phobia and pica. There was no evidenceof anaemia or iron deficiency on laboratory investigation.

1586 Volumt86 No.12 Dtwnber /996 SAMJ

Bromazepam 3 mg or 6 mgEffective anxiolytic therapy.Rome PnxlUets IPtyl t.ra Reg No. 05n7044107

PO. Box 4589 .Jo/'IanfleSOUrg 2000

~]-'"!lor6mg~~oI~diISOIOI!'I"S~ou.~l11e"~ 1.5 3mgUPlOawerameGil#f~~~6 11mgl'\OlOor~llI'rIed;Jlly ElOerly.-.l_CI\'j~lowt'roases , 'lC.rvM'I~...ryIO~~-.~~~~.-.lmff"CiJl~ o.tnn ~gr-~1n~oI~ern:m

"leOICJnel; or CIo'\.Igl.~ 01 0l00C1 co.n:..-.:I 01 hoer It.ncoon ~ IP<~ .... lDng-tIemlll'll!f~~~ snoud De~ l~~ loI"I(Jl aI~~ NIIuor'IoIDCll' ..-.ru or eIIClaDOI"l

-~ Taae!s] .... 105. 9Ol. T.1OIm 6 mg. 3D\. 90'1.~lD~.-r.:Jor I'IAprooua~ Ji,Reg I'0OI 3 mg. G/1.~ 6 mg.G/2 6Al1:leQa_~ on ~/l'o'TlJllxJ'W I"focIl.ot;n II'lyl t.m iel 10111 974-5]]5 or (rem 1lDcne£:r-;.ilJWaQnCe I.Jne IIl£AU 101 ~0800 112115

The patient was treated with sertraline 200 mg daily. Panicattacks decreased rapidly; other symptoms responded moreslowly. However, after 12 weeks of treatment there was alsosignificant improvement in QGD, pica and social phobia.

Case 2A 38-year-old man presented with QGD symptoms that hadbegun after puberty. The patient had obsessions aboutfaecal contamination, and his rituals included washing andcleaning. Surprisingly, the patient also admitted to an urgeto eat certain sorts of dry dog faeces, followed by purging.The patient described this as a behaviour for which he couldgive no sensible explanation, but which he felt compelled todo. There was no family history of QCD or pica. There wasno evidence of anaemia or iron deficiency on investigation.

The patient was treated with citalopram 80 mg daily. Therewas a gradual response in QeD and pica symptoms.However, significant improvement in both was seen at 12weeks.

Case 3An 8-year-old boy presented for evaluation of symptoms ofaGO that had begun 3 years earlier. History revealed that inaddition to the classic obsessions and compulsionsassociated with counting, touching and symmetry, there wasrepetitive chewing of paper and crayons. The patientdescribed this symptom as subjectively dissimilar to hiscounting, touching and symmetry rituals. He stated that thechewing {unlike the QCD symptoms} was not preceded byan irresistible urge or by increased tension, but simply thathe enjoyed chewing paper and crayons. He did, however,note that once he began chewing, he would have to chew aset number of items - in the same way that he might haveto touch particular objects a certain number of times. Therewas no family history of QGD or pica.

The patient was treated with clomipramine 75 mg daily.There was a gradual and significant decrease in QCOsymptoms. However, the pica continued as before.

Case 4A 34-year-old single woman presented with symptoms ofQeD that had begun at age 17. These included washing andchecking compulsions, as well as ego-dystonic sexualobsessions. There was also a history of hair-pulling from thescalp, eyebrows and pubic area. On presentation, the patientalso acknowledged co-morbid symptoms of depression andpica. The pica comprised eating up to fifteen sticks of chalkper day. The patient described the hair-pulling and pica insimilar terms, stating that both were preceded by an increasein tension, and followed by tension reduction and evengratification. There was no family history of QCD or pica, butthe patient's brother had chronic motor tics. There was noevidence of anaemia or iron deficiency on laboratoryinvestigation. The patient consented to a lumbar puncture inorder to determine cerebrospinal 5-hydroxyindoleacetic acid(CSF5-HIAA) levels, and these proved to be abnormally low.

The patient was treated with clomipramine up to 300 mgdaily for 12 weeks with no response in obsessions andcompulsions, hair~pulJing or pica, but with someimprovement in depression. Medication was switched tocitalopram 60 mg daily, Which resulted in significant

improvement in depression by week 4, with subsequentsignificant improvement in OCD, hair-pulling and pica.

Case 5A 45-year-old woman presented with a history of recurrenteating of newspapers. The patient stated that this symptomhad begun during adolescence, and had continued tomanifest itself dUring periods of stress. At such times, thepatient felt an increased sense of tension, and sheconsciously chose to eat newspaper in order to decreasethis sense of tension. She described this as an essentiallypleasurable activity, although part of her worried aboutpossible medical complications. Although the patient tendedto eat the same part of the newspaper each time (theborders), there was no specific ritualistic pattern to theeating that she felt compelled to carry out.

There was no history of obsessions or compulsions.Although there were some symptoms of depression andanxiety at the time of presentation, the patient did not meetDSM-IV criteria for a major depressive episode or for anyanxiety disorder. There was no evidence of anaemia or irondeficiency on laboratory investigation. There was no familyhistory of QCD or pica.

The patient consented to undergo single photon emissioncomputed tomography (SPEC1) of the brain beforebeginning a trial of an SRI. Five minutes prior to injection of555 MBq (15 mq Tc-99m hexamethylpropyleneamineoxime, the patient was instructed to allow herself to thinkabout chewing a newspaper and also to begin manipUlatinga piece of newspaper as she usually did prior to eating it.The patient continued with these thoughts and actions for 5minutes after injection.

SPECT data were acquired using a dual detector gammacamera (Elscint, Helix), equipped with fanbeam collimatorsusing the step-and-shoot mode and a circular orbit. A 128 x128 image matrix was used, acqUiring data in 3° steps, for15 s1step, through 360°. Data were reconstructed using aMetz. 5-14 filter to provide tomographic images in thetransaxial (parallel to the orbitomeatalline), coronal andsagittal planes. Each image set was nonnaJised to the meancerebellar counts.

Pre-treatment SPECT revealed multiple perfusion defects,including the frontal cortex bilaterally, but also the temporalcortex bilaterally, occipital cortex bilaterally, left parietalcortex, left caudate and both sides of the thalamus (morepronounced on the left).

The patient was then treated with citalopram 20 mg daily.She showed a gradual improvement in mood and anxietyafter beginning medication. Concomitantly, she noted adecreased need for tension relief by pica, with significantimprovement by week 8. The symptoms remained betterthroughout the subsequent course of treatment, althoughthere was exacerbation at times of increased stress.

After 12 weeks of pharmacotherapy, the patientconsented to repeat SPECT She was given the sameinstructions as on the pre-medication SPECT, and againcomplied with these. SPECT data acquisition andprocessing were performed in the same way as in the pre­treatment study. Post-treatment SPECT revealedimprovement in frontal cortical pertusion, as well as in mostother preViously abnormal areas. The perfusion defect in theright occipital area remained unchanged.

1588 Volume 86 .'Jo. 12 DUi'mber J996 SAMJ

For depression sufferers,relief can be as profound

as rebirth...

___CipramitThe most selective approach

to treating depression~ressionCare

V lundbecl. ,~[h .\In.:,,, ,Pr. LIJ '~O.JJ,."'~...nu~· F"mJ.ll..~Irrarnd:! m.:; R.-...:. ,,"0:?U 1: :!3:!. E.Kh t.l~ld .:~'nUIru- (lI.l1""Fram h\Jn>!;o",mhi ... "'l.ul\J.~,nll ...' :,' mf: awl"r1.l.m

~DI',lIilIe~~:mllllml600 ...~~ £picpIy~ponIiRd airodoaic lInZlIra.....m CIf n!Iout IICOIllbry­~o-r=.-\cluIIr;!OO.3000.lllll/cli ... ~il~ldedclo.:L:-u - :- ri==~Ill~~1id.~~~IIaiIiaL~"""of..,......IiambdLs.-.,

omI~~~rLu-.ero-II..rrlll~

IImbal. ChI~ Cami:Ja iD rta!l1llIdIiaay~""'-~1~

~~druoosinaI, d.a-.d ..m.. bIoodal COUIlI!~ 1n:lSInll

w-a-~"DftC-.lIDlOaf ilIaDIioLLea~~""~~::b:UL4b::) ; . ~:w-..)lIaoa~hcb:Tliq:GI:m.1600 --wfir:d" p;lC:b<l6l1.1km~__ fill~ i:Ib-.:J'IJI205; tt for &IJ prac:i:IirIs~DIlIllIllI pKt bfIo:I CIf

~·1JII.' IJ29-9111. ca Gcirr 'Ptv I..a:I.. 72 Sla:l Roid,S;:an:c.~ hrL

:d-=oa::s an lile.~ lilrtber inbnwion, pIcax CODlXI tbc TriepW'rodua),b.n.gu III 011 929-!H1I..

sO[§xm,ver evolving solutions to C:\S disorders

The pathway to more comfortable controlfor your epileptic patient

TRI L EPTA L®300/600

Stops the seizure~

not the thinking

~riqxal300mg R.tg. .' X12.5/20:5; T~ 600 mg Rq. :-;0. XI2.5/206

ResultsIn 2 of the 5 cases presented here. pica appeared to be asymptom of QCD. These patients had classic QGDsymptoms which met DSM-IV diagnostic criteria for QGD,and they described their pica as subjectively similar to suchsymptoms. There was no evidence to suggest that the picawas the resutt of iron-deficiency anaemia. Finally, classicsymptoms of QGD and pica responded simultaneously totreatment with an SRI, and the gradual time course of thisresponse was similar to that typically seen in QGD. In a thirdpatient with QGD, the patient denied that the pica itself wasan QeD symptom (although it followed certain rules), andthe QeD and the pica responded differentially to an SRI.

In the remaining 2 patients, pica was more closelydescribed by the diagnostic criteria for the DSM-IV categoryof Impulse Control Disorders Not Otherwise Classified."These patients thus described increasing tension prior to thepica and relief of tension and gratification after the pica, with1 patient comparing pica to hair-pulling symptoms. Inaddition, 1 patient had decreased levels of CSF 5-HIAA,while in a second, functional brain imaging revealed hypo­activity in the frontal cortex, left basal ganglia and otherareas, which was normalised by pharmacotherapy.Response to an SRI was relatively rapid, and in 1 case thesymptom continued to wax and wane in response toexternal stressors.

DiscussionThe main findings of this small series of cases are: (I) thatpica may be a symptom of aGO and may respond gradually,together with classic QCD symptoms, to treatment with anSRI; and (iI) that pica may be phenomenologicallyreminiscent of an impulse control disorder, and that in suchcases pharmacotherapeutic data (relatively rapid responseto an SRI) may be consistent with that seen in impulsedisorders.2C Although limited neurobiological data wereavailable, these (decreased CSF serotonin metaboliteconcentration, frontal hypo-activity) were also consistentwith that seen in impulse control disorders.20

A phenomenologicaJ and neurobiological contrastbetween compulsive and impulsive disorders has beenconsidered in some detail elsewhere. '2' An overty simplisticschema contrasts the phenomenology of compulsivedisorders (increased harm-avoidance) and impulsivedisorders ~ncreased risk-seeking), as well as theneurobiology of compulsive disorders (serotonergichypersensitivity, frontal hyperactivity) and impulsivedisorders (serotonergic hyposensitivity, frontal hypo-activity).Similarly, while both compulsive and impulsive disordersmay respond to treatment with SRls, optimal dosage andtypical response time may differ.

Such a schema provides one way of conceptualising thefindings of this paper. Particular cases of pica may lie ateither the compulsive or the Impulsive end of thecompUlsive-impulsive spectrum of disorders. Such anhypothesis may be useful in guiding clinical Interventions ­suggesting, for example, that high-dose SAls may be usefulin the treatment of patients with QCD and compUlsive pica,but that more caution may be necessary in

SAMJPSYCHIATRY

pharmacotherapy for patients with impulsive pica. It mayalso provide a useful heuristic for guiding further research onpica, suggesting, for example, that further studies of picaneed to examine phenomenology, neurobiology andtreatment response more closely in relation to aCD andQCD spectrum disorders.

Our finding that pica may occur in association with classicsymptoms of DCD has not previously been emphasised inthe psychiatric literature. However, a historical reviewindicates that in many cases of pica, a crucial clinicalvariable is impulse dyscontrol, whether secondary tounderlying neuropsychiatric impairment,:l-,e or secondary toan impulsive personality structure.'~lB For example, pica maybe seen in mental retardation,U where there is diminishedimpulse control. Similarly, in the 19th century, pica was seenas a symptom of a general condition known as chlorosis"e inwhich there was a range of symptoms consistent with apathological or impulsive personality structure.

Szabo et al. 's suggest a link between iron deficiency andthe neurobiology of QGD. The relationship between irondeficiency and pica may not, however, be a simple or directone. '6.17 However, it is notable that iron deficiency has beenfound in some cases of trichotillomania.:z2 Nevertheless. datasupporting a relationship between iron levels and function ofthe neurotransmitters known to mediate OCD are currentlyscanty. Further work to delineate the neurobiology of bothcompUlsive and impulsive pica is clearly necessary.

The concept of a spectrum of OCO disorders has beencriticised for being at times overty inclusive and insufficientlyspecific.n Indeed, the view of pica taken here may apply toonly a subgroup of patients, and even then the approachmay be only partly valid. The QCD spectrum concept is bestunderstood as having heuristic value in both clinical andresearch settings, rather than as an ultimate delineation ofthe relationships between different psychiatric disorders. Inthe case of pica, the notion of a compulsi\ie-impulsivespectrum does seem to provide a useful fnmework for thecase material presented here, and will hopefully encouragefurther investigation of this interesting symptom.

Or Stein is supported by a grant from the Medical ResearchCouncil of South Africa and by the Lundbeck Fellowship Award.

REFERENCES

1 Jen o;e MA.. Boler LB. M'11OCh,e'1Q VIE OORss.....-C"mpujs. ..e o.sore1eni Theory.,,0 M""'g&mllnf 2n.a eo CnIColgo Yea. 800_ Meo,cal Pubhsners. 1990

2 AolPOPOt1 JL Aecerll aovances ,n obsess,~e·compulS'VIId.so.derNevropsychopharmacoJogy 1991. 5; 1·10

3 Hollandll' E. !.JllOOw,tz MA. OilCanol C Conceptual ana metrtodologocal ISSUes In

studies 01 obsess,v..compulSlve and Touletttfs o.so.oe.s Psycn,a~nc

Developments 1989.4; 267-296Fallen BA Javrtcn J, llebowrU MA HyOOChol\Onas,s ana OCD OverlapS ,no.agnos,s ana t"eatment J Cbn Psycnlllnry 1991, 52: 457 ·4.60

5 Ste'" DJ. 5lfTl«lll D, COI'I<en W. HOllal\Oer E TflChe! l1orTlan,a anc obsess,ve­comOUI5' •• o'SOl"Cel J Om Ps~c"~U'y 1995. 4: 28·34-

6 Je........ MA.. 1I1nl!$~ ••latee ~o obSllS$' ..e·comtluISlve O,.sot~er In Jef"ke MABaltl LB 'A." Ch.e!IO WE. ees OOse$S>ve-Compulslve O,so,aen; rreury anaManagement 2no ed Chtcago Veal 600k Meo cal Publosners 1990

7 Stetn DJ, Hollander E The spect'vm 01 obsess,v,-compuls,ve .elated olso.dersIn. Hollance. E.!!Cl ObSesslve·Compu/s,ve RelateO D"oraers Wasrllnglon DCAmerican PsyCh,atflC Press 1993

8 HOllanCe> E. lIebo... 11 MR. ,"""ncn,l R. ef al Tleal",.nt of boOy oysmo.ph,cCSOf"cer WIth ~otO",n ret;pta..e oloc..ers An J PsyCf1<iJ!ry 1989 146: 768-770

9 LeonalO ....L lenane MC. Sweeo SE. et AI J. cou~·OI'l\O COl'"oans.on otc:om o.a-"e and des plar- "le "r.at1T'enl of se_ere onycncgnagJa [n<ll' p'1.n\l'A.-cr Gen Ps.c~.arry 1991. 48: 821·827

10 S",ecQ 5E Lecnarc"'L RaPOoort JL er al /4 COI.IOe·OI,no co..-p¥"son ctciorllJ(a,... ..e ana ees o.al"l,re n •..e ~ream"ent 0' ·flCno: Ioman,a 'l.a.tr IJ'lI ng. N£ngl J MfK11989 321: ':97-501

11 Ar-encan Psyc~ amc A$,$OC,al,on D.agncs:'c ;]Ino Stat'srrcal ManUaJ of Menr.'D>scraen 4th eo Wa:snlngton. DC Amencan Psych,attlc P.IIS5 1994

12 Eowaros CH McDonalO S M lchell MD. er al Clay ana cornsta.ch e,lting womenJ Arl D et A.$soc 1959 35: 810-815

13 Kar.nerL C"ICPs/f;h,atry 2ndec 5p''''9'e·c CC-"oma:s 1955

AMI \olumt So Sl). f! Deamba 1990 , 1591

14. McLoughlin IJ. The picas. Br J Hosp Med 1987; 37: 286-290.15. Kinnell HG. Pica as a feature of autism. BrJ Psychiatry 1985; 147; 80-82.16. Parry-Janes B. Parry-Jones WLL Pica; Symptom or eating disorder? A historical

assessment. Br J Psychiatry 1992; 160~ 341-35417 Miller OR, Baehner AL. Blood Diseases of Infancy and Childhood. St LOUIS: CV

Mosby. 199018. ZeitHn S6. Polivy J. Copmphagia as a manifestation of obsessive-compulsive

disorder; A case report. J Behav Ther Exp Psychiatry 1995: 26: 57-6319 Szabo CP. Van Aooy W, AJlwood CW. Pica - is it a variant of obsessive­

compulsive disorder'! A case report. S Air Med J 1995; 85: 1390-1391.20. Stein OJ, HOllander E. Ueoowitz MA. Neurobiology of impulsivity and impulse

control disorders. J Neuropsychia£ry CUn Neurosci 1993; 5: 9-17.21. SH~in DJ. HOllander E. Impulsive aggression and obsessive-compulsive disorder.

PsychiatriC Annals 1993; 23; 389-39522 McGehee FT Jun. Buchanan GR. Trichophagia and \richobezaar: EtialogicaJ role

at Iran deficiency. J Pedialr 1980; 97: 946-948.23. Rasmussen SA. Obsessive-compUlsive spectrum disorders. J C/in Psychiarry

1994; 55: 89-91

Accepted 26 June 1996.

Obsessive-compulsivedisorder in black SouthAfricans - a case seriesPrakash S. Gangdev, Dan J. Stein,

Jean B. Ruzibiza

Background. Obsessive-compulsive disorder (OGD) has

been shown to be highly prevalent in both developing and

developed countries. Nevertheless. data on QGD in

blacks, and black South Africans in partiCUlar, are limited.

Method. Records of patients presenting with QGD at a

tertiary hospital serving a predominantly black population

were reviewed. Patient data, including demographic

information, presenting symptoms and clinical course,

were collated.

Results. Six black South Africans had presented with

QGD in the previous year. Phenomenology and

psychopharmacology of the disorder were largely

reminiscent of those previously reported in the

international literature.

Conclusion. Not surprisingly, black South Africans may

suffer from QGD. Nevertheless, it is likely that such

patients do not present for treatment or areunderdiagnosed. Future rigorous epidemiological research

on QCD in South Africa is necessary.

S Atr Med J 1996; 86: 1592-1598.

Department of Psychiatry, Medical University of Southern Africa,PO Medunsa, 0204

Prakash S. Gangdev. M.S. 8.S.. M,D. (BOMBAY). F.F.PSYCH.(SA)

Jean B. Ruzibiza. MD. {KINSHASA/

Department of Psychiatry, University of Stellenbosch, Tygernerg, W.C<tpe

Dan J. Stein. M.8. Cti.8.. ER-C.P.C.

Although Obsessive-compulsive disorder (QGD) has beenshown to be extremely common, with a lifetime prevalenceof 2 - 3% in both developed and developing countries,u'

data on QGD in Africa are extremely sparse. A number ofreviews of psychiatric disorders in Africa make no mentionof aCD,3.• while others have suggested that the incidence ofobsessive-compulsive symptoms in blacks is low.~·6

In individual studies documenting diagnoses of psychiatricdisorders in African settings, anxiety and neurotic disordershave unfortunately often either been classified together7

'16 or

have not included aCD as an anxiety disorder. lT•ls A number

of studies have, however, found that aGO in black(particularly from sub-Saharan Africa) psychiatric patientsl~2!;

is low [fable I). Nevertheless, many of these studies sufferfrom significant methodological problems, such as failure touse structured clinical interviews,

Table I: Prevalence of QeD in African studies

QGDAuthors Diagnosis Country No. (%)

Lambo'~ Clinical obsessive- Nigeria 7946 0.004

compulsive neurosisLeighton et al.20 Clinical obsessive- Nigeria 0.0

compulsive neurosisElsarrag~' Clinical obsessive- Sudan 2160 0.5

compulsive neurosisOkasha22 Clinical obsessive- Egypt t 000 2.6

compulsive disorderGerman Clinical obsessive- Uganda 121 0.0and Arya"Z$ compulsive neurosisQrleyand Present state Uganda 206 2.4

Winif4 examinationobsessions

Gureje et al,25 Clinical obsessionai Nigeria 1 914 0.1

neurosisBertschy Clinical obsessive- Benin 351 1.4and Ahyi~ compulsive disorder

There is no mention of QCD in South Africa in a number ofgeneral reviews of the South African psychiatric literature.m'sSouth African psychiatrists have, however, taken part inmultinational clinical trials of aco, which suggests that OCDpatients are seen and treated here.29 Nevertheless, we areunaware of any reports on black South Africans with QCD.In this paper, we therefore provide retrospective clinical dataon a series of black South African patients treated forsymptoms of this disorder.

MethodsRecords of 6 patients who had presented to a large tertiaryhospital with QCD during the last year were reviewed.The hospital serves a predominantly black popUlation, andall patients were black South Africans. All patients metdiagnostic criteria for QeD, as outlined in the currentDiagnostic and Statistical Manual of Mental Disorders(DSM-I\I).~

Patient data, including demographic information,presenting symptoms and course, were collated andtabulated (Table 11).

1592 Volume 86 No. 12 December 1996 SAMJ