Report

Tinea capitis in south-western Ethiopia: a study of riskfactors for infection and carriageJose Ignacio Figueroa, MD, PhD, Thomas Hawranek, MD, Aynalem Abraha, MD, andRoderick James Hay, DM, FRCP, FRCPath

From the St. John's Institute ofDermatology (UMDS), Guys' Hospital,London, England; DermatologischeAbteilung, LKA Salzburg, Salzburg,Austria; and Jimma Institute forHealtti Sciences, Jimma, Ettiiopia

CorrespondenceJose Figueroa, MDSt. John's institute of DermatologyGuy's HospitalLondon SEt 9RTUK

Supplied in part by The LeverhulmeTrust

AbstractBackground Tinea capitis is a common dermatophyte infection which constitutes an ^

important public health problem among children worldwide. The endemic nature of scalp

ringworm in Africa is perpetuated mainly by the lack of knowledge about the prevalence

and carrier status, and the absence of control measures.

Methods Two hundred and nineteen schoolchildren from urban and rural communities of

the Illubabor district, south-western Ethiopia, were examined, and scalp samples were

taken. Children were classified according to clinical signs and mycologic findings.

Results Physical examination revealed that 29% of the children had clinical lesions

compatible with tinea capitis. Dermatophytes were isolated from 33% of the children's

scalp samples; of these, 16% had clinical lesions and 17% were identified as carriers.

Trichophyton vioiaceum was responsible for 97% of infections.

Conclusions Tinea capitis was the second most prevalent cutaneous finding in these

children, with a higher prevalence in the urban community; the predictive value of the

clinical diagnosis was low and a high proportion of children were identified as carriers in

these communities. No relationship between household overcrowding and scalp infection

was found.

Tinea capitis is a common superficial fungal infection; it ishighly contagious and represents a significant public healthproblem, particularly among schoolchildren.^ The epidemi-ology of tinea capitis is poorly understood. Infections-mayremain undiagnosed; the clinical manifestations range fromnonspecific self-limited scaling lesions to widespread scar-ring alopecia. The causative agent of tinea capitis varieswith geography and time,̂ >3 perhaps as a result of changesin the living conditions of affected communities^ or fungaladaptations. Tinea capitis is endemic in many Africancountries; its prevalence varies between io% and 30%,and it is estimated that there are approximately 20 millionactive infections with an even higher carriage rate.' Endemicfoci of infection caused by distinct organisms have beendescribed, e.g. Microsporum audouinii in Nigeria*̂ andTricbophyton vioiaceum in North Africa.̂ Often outbreaksof infection appear to be due to indirect spread via externalagents, combs, or hairdressers' equipment, or person toperson transmission under overcrowded conditions, suchas in schools or refugee camps.

There are no data on the prevalence of dermatophyte

infection in rural Ethiopia. In a previous study, 53% ofsamples from 51 patients with clinical lesions showeddermatophyte infections: 50% were T. vioiaceum, 27%were T. schoenleinii, and 23% were mixed infections withthe two fungi (J. Figueroa, unpublished data). A total of219 randomly selected children from three different schoolswere examined. The presence of clinical lesions suggestiveof tinea capitis was noted. Scalp samples of both clinicallysuspected tinea capitis cases and normal children weretaken using the scalp brush technique, and the prevalenceof dermatophyte infection was described.

Background

Illubabor province (south-western Ethiopia) has an esti-mated population of 3 million inhabitants (only 10% livingin urban settlements) and an area of 35,000 km';^ it isEthiopia's most important coffee growing region. Theweather is temperate and the rainy season extends fromJuly to September. Jimma, the administrative capital of theprovince, is located 1700 m above sea-level; Sebeka-Boke, 661

© 1997 Blackwell Science Ltdinternationai Journai of Dermatoiogy 1997, 36, 661-666

662 Report Tinea capitis in south-western Ethiopia Figueroa et al.

at 1500 m, is a small roadside rural community andKishe-i, at 1200 m, is a small riverside settlement. Themajority of the inhabitants of these two rural communitiesare subsistence farmers, and their dwellings are round,single-bedroomed, mud-floored houses. Sebeka-Boke'spopulation belong to the Oromo ethnic group, whilstKishe-i, which has recently been created, has a moreheterogeneous ethnic mix (mainly Amhara, Kaffa andOromo).

Patients and methods

In February 1995, during a week-long survey, three schools,

including one in an urban area, were selected. A total of 219

children were examined; children attending Grades 1 and 2 at

one of Jimma's primary schools in the outskirts of the town

(128); all Grade 1 children at Sebeka-Boke school (33); and all

Grade 1 and 2 children at Kishe-1 school (58). The children's

ages ranged between 5 and 15 years (median, 10 years). The

sex distribution of the sample was 141 (64.4%) boys and 78

(35.6%) girls. After explaining to the teachers the purpose of

the study and obtaining their consent, children were examined

in their classrooms. Our team consisted of three dermatologists

and a medical officer-interpreter attached to the Shebe Health

Center. A questionnaire designed to identify self-awareness of

skin disease, type of housing, overcrowding, living conditions,

and habits was administered to all the children. Data regarding

clinical symptoms (itching) and signs (scaling, hair loss, black

dots, kerion, scutulum, pustules, and id reaction) were noted. A

presumptive clinical diagnosis of tinea capitis was made on the

basis of the presence of at least two clinical signs. Scalp

samples were taken from all the children irrespective of the

clinical symptoms. Statistical analysis was performed using the

EPI-INFO 6^ epidemiology program.

Scalp samples were taken by gentle brushing with sterile

toothbrushes, and transported to the Mycology Laboratory, St.

John's Institute of Dermatology, St. Thomas' Hospital, London.

Samples were examined by direct microscopy and cultured on

Sabouraud's agar (Oxoid), supplemented with 0.05 g/L

chloramphenicol and 0.4 g/L cycloheximide. Cultures were

incubated at 28 °C for between 2 and 6 weeks. Dermatophytes

were identified by colony morphology and microscopic

appearance.

Definitions

Overcrowding was defined as more than five individualsper room in the household. Fungal growth was reportedas negative, moderate, and profuse according to the numberof colonies present at the end of the incubation period:negative, in the absence of any fungal colony, regardlessof any bacterial growth; moderate, if between one and six

colonies were present in the sample; and profuse, if morethan six colonies were present.

After mycologic analysis, children were classified accord-ing to their clinical status and the isolation of dermatophytesfrom their scalps as confirmed and possible cases, carriers,and healthy: confirmed cases involved children with bothclinical diagnosis and positive dermatophyte isolation; pos-sible cases involved children with clinical diagnosis, butnegative cultures; carriers involved children without clinicalevidence of infection, but with positive dermatophyte isola-tion; and healthy involved children withovit clinical lesionsor dermatophyte growth. ,, ,

Results

A total of 382 cutaneous findings were recorded in 212children (97%); only seven children (3%) were found tobe clear of any skin disorder. Infestations were the mostcommon conditions (61% of all the diagnoses), followedby skin infections (33%). Dermatophytoses were the secondmost common condition, affecting 33% of the childrenand comprising 18% of all the diagnoses; 63 children(29%) had clinical evidence of tinea capitis, three (1.4%)had favus, and four (2%) had tinea corporis. Positivedermatophyte cultures were obtained from 72 (33%) chil-dren. It is noteworthy that, despite the addition of anti-biotics to the media, there was a high degree of bacterialcontamination in the samples; this may be due to the highprevalence of pediculosis capitis, cultural practices such asthe application of butter to the scalp (which can also makescaling difficult to assess), and a general lack of hygiene.T. vioiaceum was identified as the most prevalent dermato-phyte species amongst these children, comprising 96% ofthe cultures (see Table i); T. schoenleinii was isolated in1.4%; mixed T. violaceum-T. schoenleinii infection wasfound in 1.4%; and T. rubrum was observed in 1.4%.Growth was profuse in 39 (53%) of the cultures, andmoderate in 33 (46%). ,:,;;-;

A pie representation of the correlation between clinicalcases and mycologic examination is shown in Fig. iA. Ofthe 63 children with a positive clinical diagnosis, 35 (56%)were confirmed cases, representing 16% of the sample; theremaining 28 (44%) yielded negative cultures and werecategorized as possible cases (13% of the sample). Underthese conditions, the positive predictive value of the clinicaldiagnosis of tinea capitis was 56%. In contrast, of the 156children with no apparent clinical evidence of tinea capitis,37 (24%) yielded positive dermatophyte cultures and weredistinguished as carriers (17% of the sample); 117 (76%)yielded negative cultures and were reported as healthy(54% of the sample). The lack of clinical evidence of tineacapitis had a negative predictive value of 76%. A significantcorrelation was observed between the amount of fungal

International Journai of Dermatoiogy 1997, 36, 661-666 1997 Blackwell Science Ltd

Figueroa et al. Tinea capitis in south-western Ethiopia Report 663

Table 1 Isolation of dermatophytes according to sex and school (A). Sex and school stratification of the children with positiveclinical diagnosis, confirmed cases, possible cases, carriers, and healthy children (B). The percentages of boys and girls withineach category are shown

A "'

T. vioiaceumT. schoenleiniiT. rubrumMixed infection*

School

Kishe-1

S

6001

9

4000

Sebeka-Boke

S

0100

9

1010

Jimma

2600

9

3200

- 0

Total

e

3210

• - 1

9

37010

Total 26 32 34

Clinical diagnosisConfirmed casesPossible cases

CarriersHealthy

11562*

26

0004 a

15

.7071S

19

22+004

2615

. 1138

17 .,.13 :,

;, 4. .19

n

44 (70%)20 (57%)24 (86%)14(38%)83 (70%)

19 (30%)

15 (43%)4 (14%)

23 (62%)36 (30%)

"T. violaceum-T. schoenleinii. +One T. vioiaceum and one T. rubrum culture. *One T. vioiaceum and one mixed infection(T. violaceum-T. schoenleinii). ^One T. schoenleinii culture.

growth and the classification as either confirmed cases orcarriers. Fungal growth was profuse in 23 (66%) of the35 confirmed cases and moderate in 12 (34%); similarly,it was profuse in 16 (42%) of the 37 carriers and moderatein 21 (58%). Children with clinical evidence of tinea capitiswere three times more likely to yield profuse fungal growth(Odds ratio 3.15, 95% CI i.07-9.41, P=o.oi).

Tinea capitis was diagnosed in 29% of the children fromKishe-i, 27% from Sebeka-Boke, and 34% from Jimmaschool. Children from rural schools, Kishe-i and Sebeka-Boke, were grouped together and compared with childrenfrom Jimma (Fig. iB). The prevalence of clinical diagnosisof tinea capitis was higher in children attending the urbanschool (Odds ratio 1.8, 95% CI 0.92-3.51, P=o.O4).Dermatophyte isolation rates were also higher in the urbanthan in the rural schools (45% of the children at Jimmaschool had positive cultures versus 15% for the ruralschools. Odds ratio 5.13, 95% CI 2.46-10.86, P<o.O5).Indeed, 80% of the positive fungal cultures came fromchildren attending Jimma school. Children attending theurban school had a significantly higher risk for clinicaltinea capitis (Odds ratio 4.36, 95% CI i.zi-16.3, P<o.O5).Similarly, the presence of carriers was also significantlyhigher in the urban compared with the rural schools (Oddsratio 7.2, 95% CI 2.42-23, P<o.O5). There were nosignificant differences in housing or conditions of hygienebetween the two rural settlements.

There were no statistically significant differences in thepercentages of boys and girls with a positive clinical

diagnosis of tinea capitis (see Table i); 38% of girls had apositive clinical diagnosis versus 31% of boys. Samplesfrom girls attending Jimma school accounted for 44% ofthe total fungal isolations; 60% of the scalp samplesfrom girls at Jimma school yielded positive dermatophytecultures. Cirls exhibited a higher isolation rate of T.vioiaceum: 53% of positive cultures (Odds ratio 2.99, 95%CI 1.58-5.66, P = O.OOO2).

Younger children (aged 10 years or under) were morelikely to present with clinical lesions compatible withtinea capitis; they represented 63% ofthe positive clinicaldiagnoses, and 94% had multiple clinical lesions. In addi-tion, 45 (38%) of the scalp samples from children under10 years yielded positive dermatophyte cultures, and theseaccounted for 62% ofthe total positive cultures. Similarly,25 (21%) children under 10 years were classified as con-firmed cases, compared with only 10 (10%) older children(see Fig. iC). No significant differences were observed inthe proportion of possible cases, carriers, and healthychildren according to age group.

Overcrowded conditions existed in 109 (50%) of thechildren's households: 59% of Kische-i, 52% of Sebeka-Boke, and 46% of Jimma. No significant association wasidentified between overcrowding and the clinical diagnosisof tinea capitis or fungal isolation. Overcrowding wasnoted in 33 (52%) of the houses of children with clinicaldiagnosis of tinea capitis and in 37 (52%) of the dwellingsof children with positive isolations. v-.;:.*•-;, -;a i

Scahng was the most common clinical lesion; it was

© 1997 Blackwell Science Ltd international Journal of Dermatology 1997, 36, 661-666

664 Report Tinea capitis in south-western Ethiopia Figueroa et al.

Possible diagnosis ,. ;: ; :; . . , . : „ •-

° Confirmed diagnosis16%

Healthy

5 4 %

Rural schools Urban schoolPossiblediagnosis Confirmed

] 5o/̂ diagnosis7%

Possiblediagnosis

12%

Confirmeddiagnosis

22%

Healthy71%

Carriers7%

c.<10 years old

Possible Confirmeddiagnosis diagnosis

13% 21%

Healtby49%

Healthy, 42%

>10 years

- • ' - '

^ - — - IHealthy

60%

old

Possiblediagnosis

13%

Carriers24%

Confirmeddiagnosis

^ ^ 10%

Carriers17%

Figure 1 Pie representation of the rcorrelation between the clinicalevidence of tinea capitis anddermatophyte isolation in thelaboratory: (A) overall; (B) schoolstratification according to urban andrural; (C) age stratification. Youngerchildren were those aged lo years orunder (119 children) and olderchildren were older than 10 years(100 children) • •

present in 95% of the positive clinical diagnoses, in 97%of the confirmed cases, and in 93% of the possible cases.Scaling was also noted in 10 children classified as carriers(27%), although it was minor and was considered bythe team to be nondiagnostic. The second most frequentsymptom noted was itching; it was present in 57% ofpatients with clinical diagnoses, 63% of confirmed cases(P<o.oo5), 50% of possible cases, and 32% of carriers;however, itching is commonly reported (3 5 % of the generalpopulation), and is perhaps associated with the high preval-ence of pediculosis capitis. Hair loss was identified in 48%of clinical diagnoses, 49% of confirmed cases, 46% ofpossible cases, and 3% of carriers; it was identified in 13%of the general population, where traction alopecia is acommon finding. Crusting was only present in 9% ofconfirmed cases and black dots in 3%; pustules, kerion,scutulum, and id reaction were not present in mycologicallyconfirmed cases.

D i s c u s s i o n " • " " ' • - ; ' " ' • • • ' - ' • - •- ••• •••.'•'•-

The prevalence of skin disease identified in this study ishigh (97%). Bechelli et aL,^° in 1981, described a variable

prevalence of dermatoses between 21% and 87% amongstschoolchildren in different rural communities in Brazil.Our results show that dermatophyte infections (particularlytinea capitis) constitute the second most prevalent dermato-logic condition in children in the urban and rural communi-ties studied, only after ectoparasite infestations. There is ahigh prevalence of dermatophyte infection and carriage inthis community. The prevalence of tinea capitis, based onclinical diagnosis, is 29%; this is in agreement with previousreports from rural communities in Brazil and CentralAmerica.̂ '̂̂ ^ Romiti et al.^^ reported dermatophyte infec-tions as the second most important cause of skin morbidityin a sample of schoolchildren from rural Brazil, with aprevalence of 20.4%; more recently, Kottenhan etal.^^reported a prevalence of dermatophyte infection of 25%amongst schoolchildren in Honduras. Dermatophytes wereisolated in 33% of the total scalp samples studied (16%confirmed cases and 17% carriers); it is also importantthat 13% of the children were classified as possible cases.

In Butare, Rwanda, 65% of 100 children with clinicaldiagnosis of tinea capitis yielded positive cultures, with T.vioiaceum isolated in 42% of cases.'3 Similarly, Hussain

internationai Journal of Dermatology 1997, 36, 661-666 1997 Blackweli Science Ltd

Figueroa et al. Tinea capitis In south-western Ethiopia Report 665

et al.^'* isolated T. vioiaceum from 69.4% of 180 individualswith clinical diagnosis of tinea capitis in Lahore, Pakistan.In our study, T. vioiaceum was isolated in 64% of thesamples from children with clinical lesions (confirmedcases). The index of fungal isolation from samples fromchildren with a positive clinical diagnosis is similar to theisolation index of these previous reports, despite the highlevel of bacterial contamination observed in our samplesand the fact that the samples were transported to theUK prior to culture inoculation. The degree of growthcorrelated with the different clinical stages. These resultspose the question of whether it is the burden of fungi thatdetermines the development of the disease; however, thefact that growth was profuse in 43% of the carriersindicates the delicate balance between previous immuneexperiences, fungal burden, and clinical symptoms.

It is not possible to predict what percentage of thechildren, here identified as carriers, will develop clinicallesions, become culturally negative, or continue as carriers.Ive,'' in 1966, followed up 77 healthy children in Nigeria;19 children (25%) were M. audouinii carriers; 4 monthslater, four children (21%) had developed clinical lesions,eight (42%) continued to carry M. audouinii, and seven(37%) had become culturally negative.

The organisms responsible for tinea capitis exhibit avariable geographic distribution; thus whereas T. tonsuranscontinues to be the most common agent of tinea capitis inthe USA,' T. vioiaceum predominates in North Africa andAsia."* T. vioiaceum is the most prevalent dermatophyte inthis part of Ethiopia; T. schoenleinii and T. rubrum-^eitidentified as the second and third most prevalent agents oftinea capitis. In the Central African Republic, tinea capitisis caused by M. langeronii, and boys were found to beinfected more often than girls;'^ similarly, infection withM. canis, in Qatar, was more common amongst boys thangirls;''' Moore et al.'^ reported similar observations inTrinidad, where the main causative agent is T. tonsurans.Our data suggest that, although there are no significantdifferences in the presence of clinical lesions, girls are morecommonly carriers of T. vioiaceum, perhaps as a result ofcultural patterns, such as the sharing of combs or theelaborate hair braiding practices commonly employed.Tinea capitis is a disease of childhood. In our study, themajority of children with a confirmed diagnosis wereyounger than 10 years of age, although there were nodifferences in the percentages of carriers.

The prevalence of tinea capitis was higher in the urbanthan in the rural schools; however, this is a pointprevalence study and it is not possible to detect unstablerates of disease occurretice. The different distribution offungal isolation amongst the three communities studiedis highly suggestive of child to child transmission ofinfection at the school, by direct contact or by exposure

to contaminated combs or hair pieces. Interestingly, wedid not find an association between overcrowding in thehome and either tinea capitis or carrier conditions,although the fact that the isolation rate for T. vioiaceumwas higher in the urban school may suggest that theimportant factor in the transmission of tinea capitis isschool overcrowding rather than overcrowding in thedwelling. In Ive's report,'' M. audouinii was isolatedfrom adult members of the household of childrenidentified as carriers, suggesting that adults may also beinstrumental in the dissemination and perpetuation ofdermatophyte passage.

Acknowiedgments

Assistance was provided by the teachers and children fromthe three schools surveyed, and the staff of the MycologyDepartment of the St. John's Institute of Dermatology,London.

References

1 Williams JV, Honig PJ, McGinley KJ, LeydenJJ. Semiquantitative study of tinea capitis and theasymptomatic carrier state in inner-city school children.Paediatrics 1995; 96: 265-267.

2 Korstanje MJ, Staats CCG. Tinea capitis inNorthwestern Europe 1963-1993: etiologic agents andtheir changing prevalence, hit J Dermatol 1994; 33:548-549.

3 Rippon JW. The changing epidemiology and emergingpatterns of dermatophyte species. In: McGinnis MR, ed.Current Topics in Medical Mycology. New York:Springer-Verlag, 1985: 208-234.

4 MacDonald E, Smith EB. The geography of thedermatophytes. Dermatol Clin 1995; 2: 85-89.

5 Verhagen AR. Distribution of dermatophytes causingtinea capitis in Africa. Trop Geogr Med 1974; 26:101-120.

6 Soyinka E Epidemiological study of dermatophyteinfections in Nigeria. Mycopathologia 1978; 63: 99-103.

7 Hay RJ. Cutaneous mycoses. In: Strickland GT, ed.Hunter's Tropical Medicine. Philadelphia: W.B. SaundersGo., 1985: 429-438.

8 Zerihun N. Ophthalmology in Illubabor region. SouthWestern Ethiopia. Community Eye Health 1994; 7(13):9-11.

9 Dean AG, Dean JA, Goulombier D, et al. Epi Info,Version 6: A Word Processing, Database, and StatisticsProgram for Epidemiology on Microcomputers. Atlanta:Centers for Disease Control and Prevention, 1994.

10 Bechelli LM, Haddad N, Pimienta WP, et al.Epidemiological survey of skin diseases in school

© 1997 Blackwell Science Ltd international Journal of Dermatology 1997, 36, 661-666

666 Report Tinea capitis in south-western Ethiopia Figueroa et al.

children living in the Purus valley (Acre State, mo^, .•Amazonia, Brazil). Dermatologica 1981; 163: 78-93.

11 Kottenhan RK, Heck JE. Prevalence of paediatric skindiseases in rural Honduras. Tropical Doctor 1994; 24:87-88.

12 Romiti N, Paes de Almeida JR, Lopes Mattos S.Recenseamento dermiatrico no municipio de Santos.An Brasil Dermat 1978; 53: 385-406.

13 Bugingo G. Causal agents of tinea of the scalp in theregion of Butare (Rwanda). Ann Soc Beige de Med Trop1993; 73: 67-69.

14 Hussain I, Aman S, Haroon TS, et al. Tinea capitis inLahore, Pakistan. Int J Dermatol 1994; 33: 255-257. •

15 Ive EA. The carrier stage of tinea capitis in Nigeria.B J Derm 1966; 78: 219-221.

16 Testa J, Kaimha C, Georges A, et al. Epidemiology oftinea capitis in Bangui (Gentral African Republic). BullSoc de Pathologie Exotique 1992; 85: 395-396.

17 El-Benhawi MO, Fathy S, Mouhasher AH, etal. IntJ Dermatol 1991; 30: 204-205.

18 Moore MK, Suite M. Tinea capitis in Trinidad. / Trop •Med Hyg 1993; 96: 346-348.

•"I'bii-. : !'•• i ;

Skin Cancer

Attempts to establish the specific character of a particular disease, howeverfruitless they may prove, are attended with the advantage of promoting accuracyof observation, and exciting minute inquiry. With the hope that such may, insome degree, be the case in the present instance with respect to the obscuresubject of tumors and ulcers, I am induced to call the attention of surgeons toa disease, which, although probably observed by many, has never, I believe,been accurately described. I allude to a destructive ulceration of peculiarcharacter which I have observed to attack and destroy the eyelids, and extendto the eye-ball, orbit, and face. The characteristic features of this disease are,the extraordinary slowness of its progress, the peculiar condition of the edgesand surface of the ulcer, the comparatively inconsiderable suffering producedby it, its incurable nature unless by extirpation, and its not contaminating theneighbouring lymphatic glands. The slowness with which this disease proceedsis very remarkable; of three cases which have come under my observation, one,that which is represented in the annexed engraving, had existed for four years,and now presents no remarkable difference when compared with the drawing,which was executed six months ago: the eye-ball, exposed and dissected out asit has been by the ulceration, remains precisely in the same state, and the edgesoccupy the same situation as at that period. In another case, now also undermy observation, the patient, an unmarried woman aged fifty-five, states, thatthe disease has existed for twenty-three years without having ever healed; hereye-ball also has been exposed by the ulceration for nearly a year, and has notyet been totally destroyed. In the third case, that of a gentleman about sixtyyears of age, the disease existed for about nine years previous to his deatli,which took place from a different cause.

From Jacob A. Observations respecting an ulcer of peculiar character whichattacks the eye-lids and other parts ofthe face. Dublin Hosp Rep 1827; 232-239.

Internationai Journal of Dermatology 1997, 36, 661-666 © 1997 Blackwell Science Ltd