The ecology of trachoma: an epidemiological study in southern Mexico

Bulletin of the World Health Organization, 63 (3): 559-567 (1985) © World Health Organization 1985 The ecology of trachoma: an epidemiological study in southern Mexico HUGH R. TAYLOR,' FRANCISCO M. VELASCO,2 & ALFRED SOMMER3 A total of 1097 people in two communities in Chiapas, Mexico, were examinedfor trachoma, and information was obtained about personal and family hygiene. Trachoma was hyperendemic; approximately 25% of those under 10 years old were found to have significant inflammatory trachoma and almost 100% of those aged over 40 years had cicatricial trachoma, although the prevalence of trachoma differed significantly between the two communities. Riskfactor analysis wasperformed by contingency table analysis and x2 testing. The most important parameter associated with the occurrence and severity of inflammatory trachoma in children was the frequency of face washing. Children who washed theirfaces 7 or more timesper week had significantly less trachoma than those who washed less often (X2 2df = 28.7; P < 0.001). This effect was independent ofage, use ofclean water and soap, or use of clothes to dry the face. Children who washed infrequently and who used clothes to dry the face or clean the nose were more at risk for trachoma. No parameters of family hygiene or socioeconomic status correlated with the amount of trachoma in a family. These data confirm and quantifyfor thefirst time the long-held beliefthat trachoma is associated with poor personal hygiene and suggest potentially effective and efficient inter- vention strategies. Trachoma used to be a major blinding disease in Europe and North America during the last century. In this century, it gradually disappeared through a multitude of changes in living conditions, personal hygiene, health care delivery, and many other factors, or, as expressed by Reinhards, through "material and cultural progress" (1) rather than specific chemo- therapeutic intervention programmes. Trachoma is, however, a major blinding disease in many devel- oping areas, especially in rural communities in the tropical and subtropical zones. For many years people have tried to identify the factors responsible for the occurrence of severe blinding trachoma. In 1962 the report of a WHO Expert Committee (2) listed the following possibi- lities: race; climate (including temperature, rainfall, altitude, and ultraviolet light); insect vectors; population density; diet and nutrition; cultural and ' Associate Director, International Center for Epidemiologic and Preventive Ophthalmology, Dana Center, The Wilmer Ophthalmo- logical Institute, and Associate Professor, The Johns Hopkins University, The Johns Hopkins Hospital, Baltimore, MD 21205, USA. Requests for reprints should be addressed to this author. 2 Chief, SanitaryJurisdiction Number II, Coordinated Services of Public Health, Prodesch, San Cristobal de Las Casas, Chiapas, Mexico. 3 Director, International Center for Epidemiologic and Preven- tive Ophthalmology, Dana Center, The Wilmer Ophthalmological Institute, and Associate Professor, The Johns Hopkins University, Baltimore, MD, USA. social customs (including household structure, religion, occupation, and water utilization); general economic levels; previous population movements; educational status; and the presence of other ocular and general diseases. Under this broad umbrella, almost every conceivable aspect of life was covered without materially advancing our understanding or fostering the development of specific intervention strategies. Some authors have identified certain general factors associated with trachoma, such as intimate family contact (3), poor community hygiene (4), and poverty and general lack of industrial development (5, 6). Other authors have looked more critically at specific factors associated with trachoma. For example, Mann (7) identified the importance of the introduction of textiles in the absence of good personal hygiene; Assaad et al. (8) noted that crowding in sleeping quarters was important; Marshall (9) identified the introduction of piped water and the sharing of towels; and Jones drew attention to the presence of flies, which had pre- viously been suggested by Wilson (10) and Weir (11). Hollows and coworkers (12) found clear correlations between the prevalence of trachoma and four climatic variables and six compound community hygiene variables. Blinding trachoma is known to occur in Indian 4554 -559 H. R. TAYLOR ET AL. villages in the highlands of Chiapas in southern Mexico (13). The present article describes an epidemi- ological study whose aim was to identify the factors (a) that differentiated individuals with and without trachoma in communities where the disease was endemic, and (b) that could serve as a basis for an effective intervention programme. MATERIALS AND METHODS Two communities, Naranja Seca and Chaonil, were selected from the municipality of Oxchuc, which is an area of high endemicity for trachoma. Both villages were within 40 minutes walk of the nearest road and each had a community health auxiliary working from a community health centre. They were thought to have a comparable housing density, most houses being relatively close together and of similar general appearance. The field team went from house to house, taking a census and mapping each village. A family was defined as all the people whose meals came from the same cooking pot, and a family questionnaire was completed for each household. The questionnaire included demographic data on all persons normally living in that house, a general description of the house, and specific questions relating to water usage and sanitation. Completion of these forms required information obtained by both observation and inter- rogation. A second questionnaire was completed for all children aged 10 years or less. This questionnaire included questions on personal hygiene, especially relating to washing practices. All family members had an ocular examination at a centrally located point. Each person was examined for trachoma with a 2.5 x loupe and a flashlight. Trachoma was graded according to the following criteria. Inflammatory trachoma. Mild -the presence of more than 5 follicles on the superior tarsal conjunc- tiva (equivalent to WHO follicles grade F2) (14); moderate -more than 5 follicles in the central tarsal plate (equivalent to WHO follicles grade F3); severe- papillae obscuring more than half the tarsal vessels plus follicles (equivalent to WHO papillae grade P3 and follicles grade F, to F3. Results are presented for all inflammatory trachoma (WHO intensity catego- ries -mild, moderate, and severe) and severe inflammatory trachoma (WHO intensity category- severe). Cicatricial trachoma. Mild -the presence of definite scars in the tarsal conjunctiva (equivalent to WHO conjunctival scarring C, and C2); moderate- sufficient scarring of the tarsal conjunctiva to cause distortion of the tarsal plate (equivalent to WHO conjunctival scarring C3); severe-trichiasis in the presence of tarsal conjunctival scarring. Results are presented for all cicatricial trachoma and severe cicatricial trachoma. Photographs of the superior tarsal conjunctiva of the left eye were taken on all children between the age of 1 and 10 years in every fourth family. These photo- graphs were used to verify the clinical grading in this sub-sample of children in a masked fashion. Conjunc- tival scrapings were collected from the superior tarsal conjunctiva of the right eye of the same children. The scrapings were stained with Giemsa stain and examined for inclusions, bacteria, and inflammatory cells. RESULTS The census indicated that the total population of the two communities, Naranja Seca and Chaonil, was 1289, of whom 1097 (85.1 7%) were examined. Tra- choma was prevalent in both communities (Table 1). Inflammatory trachoma was most common in people under the age of 20 years (Fig. 1); under the age of 10 years it was more common in girls than in boys. In both sexes, mild to moderate inflammatory trachoma persisted into adult life. Severe inflammatory trachoma persisted only in adult women. Inflam- matory trachoma was significantly more prevalent and more severe in Naranja Seca than in Chaonil (for children aged 10 years or less, X2ldf trend = 46.3; P < 0.001). The prevalence of cicatricial trachoma increased with age (Fig. 2). Although the overall prevalence of scarring was comparable in males and females, trichiasis was most common in elderly females. Conjunctival scrapings were obtained from 171 children. Characteristic chlamydial inclusions were seen in Giemsa-stained smears from 19 (2407o) of the 81 children diagnosed clinically as having trachoma. Personal hygiene A number of parameters on personal hygiene were examined for their association with inflammatory trachoma in the 469 children aged 10 years or less. Trachoma occurred significantly more frequently and was more severe in those who washed their faces infre- quently (Table 2). The relative risk for trachoma in those who washed their faces an average of less than once per day, compared with those who washed one or more times per day, was 3.1. Overall, face washing had a population attributable risk of 5707o. Because of the striking association between face 560 EPIDEMIOLOGY OF TRACHOMA IN MEXICO Table 1. Percentage prevalence of inflammatory and cicatricial trachoma by age and sex for Naranja Seca and Chaonil Percentage prevalence Inflammatory trachoma Cicatricial trachoma Age (years) Naranja Seca Chaonil Naranja Seca Chaonil Females: 0-2 26(8)' 0(0) 3)0) 910) 3-4 77(14) 17(0) 9(0) 0(0) 5-10 47(6) 20(2) 15(0) 22(2) 11 -14 50(8) 3210) 29(4) 24(0) 15- 19 39(6) 17(0) 56(0) 50)0) 20-39 25(3) 15(1) 83(8) 83(3) 40-59 30(3) 3(3) 97(20) 100(19) ) 60 23(0) 0(0) 100(62) 91(45) Males: 0-2 35(9)" 3(0) 9(0) 18)0) 3-4 300) 8(0) 11 (0) 8(0) 5-10 36(6) 6(2) 14)0) 19(0) 11 -14 41 (9) 30(3) 32(0) 24)0) 15- 19 50101 31(10) 40(0) 45)0) 20-39 27(4) 14(2) 79(3) 70(0) 40-59 35)0) 3(0) 88)12) 100(6) > 60 17)0) 9(0) 94(17) 100(45) Figures in parentheses give the percentage prevalence of severe disease. 100.0 - MALES -- FEMALES 80.0 z < 60.0 'Ji a. Li 20.0 - A GRADES "~~~~~~SEVERE 20.0 40.0 60.C AGE Fig. 1. Age-specific prevalence of inflammatory tra- choma and severe inflammatory trachoma for males and females in Naranja Seca and Chaonil. 100.0 80.0 w z -J > 60.0 w cr 0- zW 40.0 cr 0 . 20.0 - MALES - - FEMALES -. ALL GRADES / SEVERE / 20.0 40.0 60.0 AGE Fig. 2. Age-specific prevalence of cicatricial trachoma and severe cicatricial trachoma (trichiasis) for males and females in Naranja Seca and Chaonil. 561 o H. R. TAYLOR ET AL. Table 2. Prevalence of trachoma in children aged 10 years or less in Naranja Seca and Chaonil, by the frequency with which their faces were washed Frequency of face washing No. of cases No. of trachoma cases: Percentage (per week) without trachoma with trachoma" Moderate Severe Both communities: 0 11 8 2 48 1-2 77 36 7 36 3-6 107 35 5 27 ,> 7 154 17 1 10 Naranja Seca: 0 8 7 2 53 1-2 49 29 6 42 3-6 47 31 4 43 >,7 28 8 1 24 Chaonil: 0 3 1 0 25 1-2 28 7 1 22 3-6 60 4 1 8 ,> 7 126 9 0 7 U Comparison of the prevalence of trachoma in the groups with infrequent (0-6 times per week) and more frequent (7 or morc times per week) face washing. Both communities: X22dr= 28.7, P < 0.001; X2 trend = 27.8, P < 0.00 1. Naranja Seca: X22df= 4.62, p= 0.099; x2 trend= 4.33, P < 0.04. Chaonil: X2 Idf= 3.03, P= 0.08. washing and trachoma, the relationship between face washing and other hygiene variables was examined in children under the age of 10 years. The following variables were not associated with the frequency of face washing: age, sleeping habits, or use of pre- viously used water. However, children who washed their faces more frequently (7 or more times per week) were also more likely to use soap and wash their bodies and clothes more often, and were less likely to use their clothes to dry the face or blow the nose. They were also more likely to live farther from their school and to live in a house with a separate kitchen, and their families were less likely to dispose of rubbish by leaving it on the ground. However, on their own, none of these variables was associated with trachoma. There was a major difference in the frequency of face washing between the two communities. In Naranja Seca, 84Gb of children washed their faces less than 7 times per week compared with 44%o in Chaonil (X2 = 18.0; P < 0.0001), which corresponds well with their related rates of trachoma. There was still an association between the frequency of face washing and trachoma within each community (Table 2). When stratified by community, the source of water was not correlated with either the occurrence of trachoma or with the frequency of face washing. Also, when stratified by community, there was no relationship between how long it took to get to the water source and how often children washed their faces in Chaonil; but paradoxically, in Naranja Seca, those who lived farther from their water source washed their faces more frequently (X2 = 20.6; P< 0.001). The importance of personal hygiene parameters was examined in children who washed their faces frequently and those who did not. Children who washed infrequently and who dried their faces on their own clothes or their mother's clothes had a higher rate of trachoma than those who used a towel or let their faces dry in the air (Table 3). Similarly, children who washed infrequently and who blew their noses on their own clothes or their mother's clothes had a higher rate of trachoma than those who used a handkerchief or a separate cloth, or used the back of their hand, or blew their noses onto the ground. These 562 EPIDEMIOLOGY OF TRACHOMA IN MEXICO Table 3. The relationship between the frequency of face washing, the methods used to dry or clean the face and nose, and the occurrence of trachoma in children aged 10 years or less Faces washed less than 7 times per week Faces washed at least 7 times per week Trachoma absent Trachoma present Trachoma absent Trachoma present Method of drying the face: Clothes 114 71"(38)' 54 8 (1 3)' Towel 39 11 (22) 52 2 (4) Air 42 11 (21) 48 8 (14) Method of nose blowing: Clothes 103 63"(38)' 34 3 (8)' Handkerchief 46 13 (22) 61 3 (5) Hand or ground 46 17 (27) 59 12 (17) X22df= 8.78, P < 0.01. X 2dl = 6.08, P < 0.05. Figures in parentheses are percentages. factors were not important in those who washed their faces frequently. The use of soap or clean water (water not previously used by anyone else) was not important in either group (faces washed infrequently or frequently). The variation in relative risk, by methods of nose blowing, is at most 1.5: 1; but within each category of face drying or nose blowing, the relative risk associated with face washing is of a different order of magnitude entirely, namely, 3.5:1. There was no association between the presence of trachoma and the following factors: child sleeping alone, use of soap to wash the face, use of water that had previously been used by others to wash the face, frequency of bathing (including steam baths), frequency or site of clothes washing, or frequency with which the bed linen was washed. Family risk factors To assess the importance of family risk factors, families were grouped by the percentage of children aged 10 years or younger who had inflammatory trachoma. In the study population, there were 214 such families, 109 in Naranja Seca and 105 in Chaonil. In 34 families, there were no children aged 10 or younger, and these were omitted from our analysis; 45 families had only one child aged 10 or younger. Trachoma occurred more frequently and in a higher proportion of children in families from Naranja Seca than in those from Chaonil (Table 4). Further analysis of family parameters were therefore stratified by community. The following factors were examined and found not to be associated with the Table 4. Distribution of families with at least one child aged 10 years or less by the prevalence of trachoma in the children in Naranja Seca and Chaonil Distribution of families by the prevalence of trachoma in children:' Total Community 0% 1-50% 51-99% 100% Naranja Seca 36(41)b 24(27) 16(18) 12(14) 88 Chaonil 74(80) 12(13) 2(2) 4(4) 92 Total 110(61) 36(20) 18(10) 16(9) 180 a X23df= 31.9, P < 0.001; x2 trend = 25.9, P < 0.001. b Figures in parentheses are percentages. 563 H. R. TAYLOR ET AL. frequency of inflammatory trachoma in children: size of the family; distance to the nearest house (although 75% of houses were within a 5-minute walk of the next); principal source of water for bathing, washing, or cooking; distance to the principal water source; level of education of parents; initial or final method of fecal or garbage disposal; and construction of the house, including type of walls, roof, or floor, the number of rooms, and whether there was a separate kitchen. There was no consistent evidence for family clustering of trachoma in either community. This was true for those with severe as well as less severe inflammatory trachoma. For example, the 39 indivi- duals with severe inflammation came from 31 dif- ferent families. The risk of infection in children aged 10 years or less was examined to see whether it was influenced by the presence or absence of trachoma in the mother or a sibling. In Naranja Seca, there was a slightly increased risk of children having inflam- matory trachoma if the mother had inflammatory trachoma (relative risk 1.4 for any inflammatory trachoma, 1.7 for severe trachoma). There was no such association in Chaonil. The presence of cicatricial trachoma in a mother did not influence the risk of trachoma in her children. The risk of a child having inflammatory trachoma if one of its sibs had trachoma was also assessed. In Chaonil where the overall prevalence of trachoma in children aged 10 years or less was low (1001o), the rate of trachoma in siblings of affected children was appreciably higher (21%/). In Naranja Seca where the prevalence in children was high (37%), no such increase was seen (rate in siblings of affected children, 390%o). DISCUSSION This study confirmed that there is a high prevalence of trachoma and its sequelae in villages in the high- lands of southern Mexico. It also identified a simple aspect of personal hygiene, face washing, that was strongly associated with a higher prevalence of inflammatory trachoma in young children. The study employed well-recognized epidemiological tech- niques of sample definition- a house-to-house census and a centrally located place of examination. Over 850/o of the defined population was examined. A simplified schema for grading of trachoma was used, which agreed with the WHO criteria (14) but was simpler to use in the field. Our method may, however, have underestimated the overall prevalence of inflammatory trachoma by excluding occasional cases that would have been classified by WHO as trivial trachoma (F,). The grading of trachoma was verified by having a second, experienced ophthalmo- logist review photographs from a systematic sub- sample of the children in a masked fashion (un- published data); the agreement was very close. The diagnosis was further confirmed by finding chlamy- dial inclusions in Giemsa-stained smears in a similar percentage of clinical cases as previously reported from other endemic areas (4). Inflammatory trachoma mostly affected the younger people, as is true in most endemic areas (12, 15-18). This is attributed to the ease of transmission of infection in this age group, resulting in frequent reinfection. Children are usually regarded as the major reservoir of chlamydial infection in an endemic area. The persistence of inflammatory trachoma in approximately one-fifth of adults suggests that trans- mission continued in this older group. Inflammatory trachoma has been noted in adults in some endemic areas such as Burma (19), Ethiopia (20), and Iran (17); but it is not common in others, such as Australia (12). There was a progressive increase in the prevalence and severity of cicatricial trachoma with increasing age so that, by the age of 40 years, almost everyone had at least some trachomatous scarring. Severe cicatricial scarring with trichiasis was most common in elderly women; almost half of the women over the age of 60 years had trichiasis. This is a particularly high rate of trichiasis, although in many endemic areas women have a higher rate of severe trachoma than men (5, 18, 21, 22). The increased severity of trachoma in women has been attributed to continuing transmission between children and women who have to take care of the children and are, therefore, closely associated with them, whereas the men are usually separated from the children for most of the day. To the casual observer, the two communities studied appear identical; they certainly had in common most of the physical characteristics specifically examined in this study. Both had the same racial, cultural, and socioeconomic characteristics although they differed significantly in the prevalence of trachoma. Those characteristics not in common that were identified by this study (e.g., use of well water as a major water source) did not correlate with the occurrence of trachoma. Detailed analysis of the community-level variables was not possible because only two communities were studied. Interestingly, practically no flies were found in these communities in January, a time when climatic conditions would be expected to favour fly breeding (Milan Trpi's, personal communication). Many have stressed the important role that flies are thought to play in the transmission of trachoma (5, 17, 23), although others have considered them to be not so important (18, 22, 24). The finding of severe trachoma in Chiapas in the absence of flies indicates 564 EPIDEMIOLOGY OF TRACHOMA IN MEXICO that flies are not a prerequisite for intense trans- mission of trachoma. This finding is in accord with that of Reinhards (22) who found fly control alone had little influence on the occurrence or severity of trachoma. Trachoma is often attributed to crowding (4, 6, 23, 25, 26). Many authors have indicated the importance of transmission of trachoma within the family (3, 5, 10, 21, 27, 28), although others have stressed the importance of community transmission of trachoma (8). No single factor of family hygiene was associated with an increased risk of trachoma in this study. An increased risk of trachoma in large families has been reported in some areas (29) but not others (27, 30). Taylor and coworkers (18) have pointed to a corre- lation between infection in a mother and infection in her children. Although we found no consistent family clustering of trachoma as such, the children in Naranja Seca did have a small increased risk of having trachoma if the mother had inflammatory trachoma. Just as no single family-level factor of hygiene was identified with an increased risk of trachoma, no indicator of family socioeconomic status or housing was associated with trachoma. The most important and exciting association in this study was the relationship between face washing and trachoma. A clean face and clean hands have often been advocated for the reduction of trachoma, but until now no study has specifically investigated the role this sort of intervention might play. Unless some- one washes the child's face, it is impossible for a child to wash its face without using its hands. Hence, we cannot distinguish between the relative importance, if any, of these two attributes. Washing of the rest of the body did not appear to be important. Face washing has been incorporated along with other factors (personal and environmental hygiene and chemotherapy) in trachoma control programmes in Australia (31), Malta (32), and South Africa (33). In the latter two countries, trachoma became less prevalent; but in each case this was ascribed to a general improvement in living conditions. In the Ryukyu Islands, the implementation of compulsory face washing in schools occurred as piped water became freely available. This led to a paradoxical increase in the amount of trachoma, which was ascribed to the sharing of towels (9). From the first modern descriptions of trachoma, the sharing of towels or water for washing has been implicated in the transmission of infection (26, 34, 35). In the present study, neither the use of shared towels nor shared water was associated with an increased risk of infection, although the use of clothes to dry the face or to clean the nose was important among children who were infrequent face washers. It is postulated that chlamydia are transmitted on the clothes that are used to dry or clean a child's face. Infection could also be spread among members of the same family when the mother's clothes are used; this would also lead to infection of the mother. Auto-reinfection may be important when the child's clothes are used to clean or dry its own face. The frequent isolation of chlamydia from the nasopharynx of children in a trachoma area (36) high- lights the potential importance of the spread of trachoma by nasal secretions as well as the well- recognized ocular secretions (17). Attention has also been drawn to the role dirty clothes and bed linen could play in the transmission of trachoma (7, 17, 27, 37), although we found no correlation with the frequency with which these items were washed. The sharing of sleeping space might increase the transmission of trachoma (27, 34); in some areas, at least, it has been shown to be associated with a higher prevalence of trachoma (8) but it was not found to be important in this study. From the public health viewpoint, the identifi- cation of the importance of face washing is most fortuitous. It is probably one of the very few hygiene practices amenable to change and for which expensive intervention is not required. The change of other practices might require extensive capital outlay; continuing high, recurring costs; or fundamental cultural or social alterations. Frequent face washing does require water, but this can be a minimal amount. Trials in the United Republic of Tanzania have shown that up to 50 children can have their faces washed with a pint of water (Alan Foster, personal communica- tion, 1984). The present study indicates that soap, and quite possibly clean water, are not necessary and that air drying of the face is not only acceptable but ideal. Therefore, with a single motivational message and minimal amounts of water, it should be possible to significantly reduce the prevalence of inflammatory trachoma in the highland villages of Mexico. For this area, the next step must be to mount a campaign to promote face washing and determine its impact on trachoma. It is not known whether frequent face washing is equally central to trachoma control in other cultures and areas. The use of carefully designed and executed risk factor studies will be needed to guide the development of effective and efficiently targeted intervention programmes for the control and ultimate eradication of trachomatous blindness. 565 566 H. R. TAYLOR ET AL. ACKNOWLEDGEMENTS A field study such as this is impossible without the active support and effort of a great many people. The authors wish to specifically acknowledge the contributions of Dr Francisco Cifuentes Castellanos, Dr Freddy Aguilar de la Torres, Mrs Virginia Calhoun de Millan, Dr Enrique Castro Munoz, and other members of the field team; Dr Jorge Fernandez de Castro, Dr Luis Cabrera Coello, Dr Victor M. Cardenas Ramirez, Dr Salomon Gallegos Fernandez, and members of the General Direction of Epidemiology, Secrctaria de Salubridad y Asistencia; Dr Fernando Beltran Hernandez and the staff of Centro de Investigaciones Ecologicas del Sureste, San Cristobal; Miss Vesna Bonsjak, representative of UNICEF; and Mrs Cheryl Auer and Mrs Alice Flumbaum of The Johns Hopkins University. This study received support from the World Health Organization (Prevention of Blindness Programme), Geneva; the Pan American Health Organization, Washington, DC; UNICEF, Mexico; the Federal Government of Mexico; the Governor of the State of Chiapas; and NIH Grant EY01 765. RESUME L'ECOLOGIE DU TRACHOME: UNE ENQUETE EPIDEMIOLOGIQUE DANS LE SUD DU MEXIQUE Une enquete epidemiologique sur les facteurs de risque Uies au trachome a ete executee dans deux communautes de 1'etat de Chiapas, au Mexique. Le depistage a porte sur 1097 individus et des renseignements ont et recueillis sur l'hygiene personnelle et familiale. Le trachome avait un caractere hyperendemique et environ 250/o des sujets de moins de 10 ans presentaient des signes evidents de la maladie au stade inflammatoire, et pres de 100% des sujets ages de plus de 40 ans portaient des cicatrices, mais la prevalence de la maladie differait sensiblement entre les deux communautes. L'analyse des tables de contingence et le test du x2 ont et appliques A l'etude des facteurs de risque. Le parametre le plus important associe a l'apparition et A la gravite de la maladie a son stade inflammatoire chez les enfants etait la frequence des lavages du visage. Le trachome etait nettement moins frequent chez les enfants qui se lavaient le visage 7 fois, ou davantage, par semaine que chez ceux qui le faisaient moins souvent (x22di = 28,7; P < 0,001). Cet effet etait independant de l'age, de l'utilisation d'eau claire et de savon, ou de l'emploi des vetements pour s'essuyer. Le risque etait plus eleve chez les enfants qui se lavaient peu souvent et qui utilisaient leurs vetements pour s'essuyer le visage ou se nettoyer le nez. Aucun parametre d'hygiene familiale ou de condition socio- economique ne s'est revele correle avec le degre d'activite du trachome dans une famille. Ces donnees confirment, sous une forme quantitative pour la premiere fois, la vieille croyance selon laquelle le trachome est associe a une mauvaise hygiene personnelle et elles permettent de suggerer des strategies d'intervention susceptibles d'etre a la fois efficaces et efficientes. REFERENCES 1. REINHARDS, J. Trachoma control in the European region. WHO Chronicle, 21: 57-61 (1967). 2. WHO Technical Report Series, No. 234, 1962 (Trachoma: third report of an Expert Committee). 3. GRAYSTON, J. T. ET AL. Pathogenesis and immunology of trachoma. Trans. Assoc. Am. Physicians, 85: 203-211 (1972). 4. NICHOLS, R. L. ET AL. Immunofluorescence studies of the microbiological epidemiology of trachoma in Saudi Arabia. Am. j. ophthalmol., 63: 1371-1442 (1967). 5. THYGESON, P. Epidemiologic observations on trachoma in the United States. Invest. ophthalmol., 2: 482-489 (1963). 6. DAWSON, C. R. ET AL. Severe endemic trachoma in Tunisia. Br. j. ophthalmol., 60: 245-252 (1976). 7. MANN, 1. Correlation of race and way of life in Australia and the Territory of Papua and New Guinea with incidence and severity of clinical trachoma. Am. j. ophthalmol., 63: 1302-1309 (1967). 8. ASSAAD, F. A. ET AL. Use of local variation in trachoma endemicity in depicting interplay between socio- economic conditions and diseases. Bull. Wld Hlth Org., 41: 181-194 (1969). 9. MARSHALL, C. L. The relationship between trachoma and piped water in a developing area. Arch. environ. health, 17: 215-220 (1968). 10. WILSON, R. P. Ophthalmia aegyptiaca. Am. j. ophthalmol., 15: 397-406 (1932). 11. WEIR, J. M. ET AL. An evaluation of health and sanitation in Egyptian villages. J. Egypt. Pub. Hith Assoc., 27: 55-114 (1952). 12. ROYAL AUSTRALIAN COLLEGE OF OPHTHALMOLOGISTS. The national trachoma and eye health program. Sydney, 1980, pp. 37-150. 13. TAYLOR, H. R. ET AL. Trachoma in Chiapas, Mexico. Rev. int. Trach. Pathol. ocul. trop. subtrop., 60: 19-27 (1983). EPIDEMIOLOGY OF TRACHOMA IN MEXICO 567 14. DAWSON, C. R. ET AL. Guide to trachoma control. Geneva, World Health Organization, 1981, Ch. 2, pp. 12-20. 15. ASSAAD, F. A & MAXWELL-LYONS, F. The use of catalytic models as tools for elucidating the clinical and epidemiological features of trachoma. Bull. Wld Hith Org., 34: 341-355 (1966). 16. NICHOLS, R. L. ET AL. Trachoma isolation studies in Saudi Arabia from 1957 through 1969. In: Trachoma and related disorders. Amsterdam, Excerpta Medica, 1971, pp. 518-528. 17. JONES, B. R. The prevention of blindness from trachoma. Trans. Ophthalmol. Soc. UK, 95: 16-33 (1975). 18. TAYLOR, C. E. ET AL. Eye infections in a Punjab village. Am. j. trop. med. hyg., 7: 42-50 (1958). 19. KYAW, T. A. ET AL. Control of trachoma and preven- tion of blindness in rural communities in Burma. Bull. Wld. HIth Org., 56: 945-955 (1978). 20. CERULLI, L. ET AL. Epidemiological study on trachoma in two regions of Ethiopia. Rev. int. Trach. Pathol. ocul. trop. subtrop., 68: 147-155 (1981). 21. TREHARNE, J. D. Seroepidemiological studies of trachoma. In: Chlamydial infections. Amsterdam, Elsevier Biomedical Press, 1982, pp. 83-90. 22. REINHARDS, J. ET AL. Studies in the epidemiology and control of seasonal conjunctivitis and trachoma in southern Morocco. Bull. WId Hlth Org., 39: 497-545 (1968). 23. DHIR, S. P. ET AL. The role of environmental factors in cataract pterygium and trachoma. Am. j. ophthalmol., 64: 128-135 (1967). 24. NICHOLS, R. L. ET AL. Epidemiological data derived from immunotyping of 338 trachoma strains isolated from children in Saudi Arabia. In: Trachoma and related disorders. Amsterdam, Excerpta Medica, 1971, pp. 337-358. 25. DETELS, R. ET AL. Trachoma in Punjabi Indians in British Columbia: a prevalence study with comparisons to India. Am. j. epidemiol., 84: 81-91 (1966). 26. THYGESON, P. & DAWSON, C. R. Trachoma and fol- licular conjunctivitis in children. Arch. ophthalmol., 75: 3-12 (1966). 27. BARENFANGER, J. Studies on the role of the family unit in the transmission of trachoma. Am. j. trop. med. hyg., 24: 509-515 (1975). 28. BALLARD, R. C. ET AL. Trachoma in South Africa. Soc. sci. med., 17: 1755-1765 (1983). 29. KOK, P. W. The epidemiology of trachoma blindness in southern Africa. Soc. sci. med., 17: 1709-1713 (1983). 30. ASSAAD, F. A. ET AL. The household pattern of trachoma in Taiwan. Bull. Wid Hith Org., 44: 605-615 (1971). 31. HARDY, D. ET AL. The cytology of conjunctival smears from Aboriginal schoolchildren at Yalata, South Australia. Am. j. ophthalmol., 63: 1538-1540 (1967). 32. DAMATO, F. J. The fight against trachoma in the island of Malta. Br. j. ophthalmol., 45: 71-74 (1961). 33. SUTTER, E. E. & BALLARD, R. C. Community partici- pation in the control of trachoma in Gazankulu. Soc. sci. med., 17: 1813-1817 (1983). 34. FEIBEL, R. M. John Vetch and the Egyptian Ophthalmia. Surv. ophthalmol., 28: 128-134 (1983). 35. MANN, 1. & ROUNTREE, P. A report on a recent survey of Australian Aboriginals with an addendum on bacteriology. Am. j. ophthalmol., 66: 1020-1034 (1968). 36. MALATY, R. ET AL. Extraocular infections in children in areas with endemic trachoma. J. infect. dis., 143: 853 (1981). 37. BOBB, A. A. & NICHOLS, R. L. Influence of environ- ment on clinical trachoma in Saudi Arabia. Am. j. ophthalmol., 67: 235-243 (1969).