Geographical and socioeconomic factors relating to the distribution of Schistosoma mansoni infection in an urban area of north-east Brazil.

Geographical and socioeconomic factors relating to the distribution of Schistosoma mansoni infection in an urban area of north-east Brazil M.L. Barreto1 A study was carried out in Santo Antonio de Jesus, a town in Bahia State, north-east Brazil, to determine the relationship between various biological, socioeconomic, behavioural, and geographical factors and the prevalence and intensity of Schistosoma mansoni infection. The town's population was around 45000 and the study was targeted at all children born in 1970-71 who were living in the town at the time of the survey (August-November 1984). An extensive questionnaire was used to collect information on each child and on family and household conditions; samples of stools were also taken for examination (Kato-Katz method). A survey of water snails was also carried out and information on the distribution of their breeding sites was plotted on a map of the area. The overall prevalence of S. mansoni infection was 31.0%. Several variables that reflected different aspects of the population's way of life were strongly associated with the prevalence and the intensity of infection. Some of the findings are valuable for understanding the mechanisms involved in the occurrence of schistosomiasis and its distribution in urban locations as well as for defining high-risk groups, all of which are important for planning control strategies. Introduction The diversity and complexity of the mechanisms involved in the transmission of schistosomiasis are well recognized (1). The role played by various socioepidemiological factors in the transmission of the disease and on levels of infection has been studied in various locations using a variety of methods. Among such factors are the following: distance from the transmission site, migration and the emergence of new foci, urbanization, socioeco- nomic status, sanitation, water supply patterns, and level of faecal contamination of water sources. Information about these factors permits generaliza- tions to be made about the epidemiology of schisto- somiasis and identifies the control measures that can be made for specific conditions. However, only a few epidemiological studies have investigated schis- tosomiasis as a multifactorial phenomenon, focusing on the role and activities of human populations. Such studies are those in the Philippines (2) and in Egypt (3, 4), where emphasis was given to holistic explanations for the occurrence of the disease. The objective of the present study was to assess the relationship between certain biological, socioeco- nomic, behavioural, and geographical factors and the prevalence and intensity of Schistosoma mansoni infection in an urban environment in Brazil. ' Associate Professor, Departament of Preventive Medicine, Universidade Federal da Bahia, Rua Padre Feijo 29, 40 andar- Canela, 40.140 Salvador, Bahia, Brazil. Requests for reprints should be sent to Dr Barreto at this address. Reprint No. 5150 Methods Study site and population The study was carried out in Santo Antonio de Jesus, a town situated in Bahia State in the North-east Region of Brazil. The climate is hot and semihumid (5). In 1980 the total population of the municipal district was 51580. Of these, 33741 (65.4%) lived in the town itself and the remainder in the surrounding rural areas (6). Selection of the study children The children selected for the study were born between January 1970 and December 1971 and lived in San Antonio de Jesus at the time of the survey. The survey itself was conducted between August and November 1984, when the children were aged between 12 years 8 months and 14 years 11 months. This sampling scheme was chosen for the following reasons: in north-east Brazil schistosomiasis man- soni has its peak prevalence among 10-20-year-olds; and the peak intensity of infection occurs among 10-14-year-olds (7, 8). Since the prevalence and intensity of S. mansoni vary widely with age, use of a narrow age group reduces the confounding effect of age on the results. Field work and questionnaire With the help of a 1: 4000 map every building in the town was identified and visited to check whether it comprised a household. The names of any children born during the defined period were recorded for each household. For each child identified, a specific Bulletin of the World Health Organization, 69 (1): 93-102 (1991) © World Health Organization 1991 93 M. L. Barreto questionnaire was completed. This consisted of two pairs-the first dealing with the family's living conditions and the second with the children themsel- ves. The family questionnaire collected data on the household possessions, the income, education level and occupation of the head of the household, as well as on the latrine and water supply facilities in the household. The individual questionnaire collected the following information about each child: age, sex, position in the household, migratory history, school- ing, jobs, history of stool examinations, history of S. mansoni infection and antischistosomal treatment, and water contact activities. The distribution of S. mansoni infection was plotted cartographically. Stool examination When a target child was identified, a marked container was supplied for the collection of the faecal sample. Stool examination was performed using the thick smear (Kato) technique (9) with a commercially available kit.a Two thick smears were prepared from each sample and each was read by a different microscopist. The individual egg count was taken to be the mean of that on the two slides, while the intensity of infection was defined as the geomet- ric mean egg excretion of infected individuals. Snail survey Surveys were conducted at two different times to investigate the presence of biomphalaria snails in open bodies of water. Streams were surveyed, starting at their source and thereafter at every 100 m. At each collection point a sieve was dragged along the bottom and towards the shore, searching for snails (10). All the snails found were collected, classified according to species, and examined for the presence of cercarial infection (11). The distribution of snails in the study area was plotted cartographic- ally. Analysis of data Cross-tabulations and x2 tests, geometric means and their related one-way analysis of variance or linear- ity (trend) tests (12) were carried out using the SPSS-X software package (13). Trend tests for proportions (14) and calculation of odds ratios and 95% confidence limits (15) were performed using EPISTAT epidemiological analysis software. One- tailed P-values were used whenever there was evidence to suggest that an association was in a particular direction, e.g., in tests for trend. Other- wise, two-tailed tests were carried out. a Boehringer Mannheim Bioquimica SA, Rio de Janeiro, Brazil. Maps were drawn using a GIMMS geocarto- graphic system package (27). Results General characteristics of the study population A total of 1765 children who were born between January 1970 and December 1971 and who lived in San Antonio de Jesus were identified through a door-to-door search. Of these children, 1701 (96.4%) provided one stool sample for examination (797 males and 904 females). S. mansoni eggs were identified in 527 of the children (prevalence, 31.0%). Among those who were positive, the geometric mean number of eggs excreted per gram of stool was 93. The prevalence of eggs among males was higher than that among females (38.8% and 24.1%, respectively), but the mean egg counts were similar for both sexes (91 and 96 eggs per gram of stool, respectively). The data shown in Table 1 indicate that the prevalence of S. mansoni infection was lower among children who were members of a nuclear family than those who were not. The peak prevalence occurred among children who were working as household servants (74.6%). Similarly, the mean egg count was lowest among children from nuclear families and highest among those employed as servants. Migration Our data confirm that the rate of migration to San Antonio de Jesus was high: only 22.9% (342) of the heads of household who were interviewed were born in the town, and 54.2% (610) were of rural origin. Information about the children's places of birth and their relationship to the prevalence and intensity of S. mansoni infection is shown in Table 2. There was a clear association between the time lived in a rural environment and the prevalence and intensity of infection. The prevalence and intensity of infection among children who had always lived in San Antonio de Jesus were significantly lower than those who had lived previously in rural areas. Socioeconomic characteristics of the study families The minimum wage in Brazil is around US$ 60 per month. In the study population 34.8% of the families had a monthly wage lower than this. Family income was strongly associated with the risk of becoming infected with S. mansoni, since the prevalence increased with decreasing income, from 17.2% among families whose income was greater than or equal to seven minimum wages, to 34.8% 94 WHO Bulletin OMS. Vol 691991. Distribution of S. mansoni infection in north-east Brazil Table 1: Prevalence and intensity of Schistosoma mansoni infection according to the relationship of the study children to the head of the household Mean number of Relationship No. of children % prevalence eggs/g stool Odds ratio Child, stepchild, or grandchild 1498 (88.1)0 29.7 84: 75-95 1.0 Other relative 92 (5.4) 30.4 188: 108-326 1.0: 0.6-1.7 Without kinship 63 (3.7) 37.1 88: 48-160 1.3: 0.8-2.4 Household servants 48 (2.8) 74.6 206: 115-369 7.2: 3.6-14.7 P < 0.0001 P < 0.001 Total 1701 (100.0) 31.0 93: 87-98 a Figures in parentheses are percentages. b Figures in italics are 95% confidence limits. Table 2: Prevalence and intensity of Schistosoma mansoni infection among study children according to their place of birth and time lived in rural areas Mean number of No. of children examined % prevalence eggs/g stool Odds ratio Place of birth: San Antonio de Jesus 919 (54.Q)a 27.4 74: 64-86 b 1.0 Another town 338 (19.9) 28.5 89: 66-120 1.1: 0.8-1.4 Rural areas 444 (26.1) 40.5 130: 104-162 1.8: 1.4-2.3 P < 0.0001 P < 0.0001 No. of years lived in rural areas: 0 1083 (63.7) 25.7 77: 66-89 1.0 0-4 214 (12.6) 32.2 97: 71-133 1.4: 1.1-1.9 ¢5 404 (23.7) 44.6 122: 98-153 2.3: 1.8-3.0 P < 0.0001 P < 0.001 a Figures in parentheses are percentages. b Figures in italics are 95% confidence limits. among those earning less than one minimum wage. A similar trend was observed for mean egg excre- tion, but this was not as consistent as that exhibited by the prevalence of infection (Table 3). An increase in the education level of the head of the household (father, grandfather or stepfather) was strongly associated with a decrease in the prevalence and intensity of infection. Sanitation and water supply The majority of the children lived in houses equipped with a flush (66.2%) or pit latrine (19.1%), and 92.6% lived in households with piped water. Children from households with piped water and sewage disposal systems had similar prevalences of S. mansoni infection but a lower mean egg excretion rate than those from households that were served by ponds; however, the prevalence among children from households with piped water but no wastewater disposal facilities was greater than both the groups. The worst conditions were associated with households that were supplied from open water sources. Children from such households had a prevalence of infection of 52.5% and mean egg excretion of 107 eggs per gram of stool (Table 4). Water contact activities Table 5 shows the relationship between the four variables involving contact with water or defecation activities, together with the prevalence and intensity of S. mansoni infection. Among the other variables covered in the questionnaire were those associated with washing and occupational activities, but their frequency was too low to be considered in the analysis. Almost 92% of the children reported having played or swum in bodies of water at some time. The few who denied this kind of activity had a very low prevalence of S. mansoni infection (9.2%). A lower mean egg excretion was also observed for this group, but the difference was not significant at WHO Bulletin OMS. Vol 69 1991. 95 M. L. Barreto Table 3: Prevalence and intensity of Schistosoma mansoni infection according to the education level of the head of the household and family income Mean number of No. of children examined % prevalence eggs/g stool Odds ratio Family income:a <1 290 (20.1)b 34.8 113: 86-140' 2.6: 1.4-4.9 1 452 (31.3) 32.3 88: 72-108 2.3: 1.3-4.3 2-3 408 (28.3) 28.4 62: 50-78 1.9: 1.0-3.6 4-6 199 (13.8) 25.6 81: 56-116 1.7: 0.9-3.3 a7 93 (6.4) 17.2 95: 48-189 1.0 P < 0.001 P < 0.025 Education level:d None 444 (29.6) 35.8 99: 80-123 5.0: 1.4-16.3 Low 865 (57.8) 28.6 81: 69-95 3.6: 1.0-12.5 Medium 158 (10.6) 22.8 59: 39-90 2.7: 0.7-11.1 High 30 (2.0) 10.0 42: 27-64 1.0 P < 0.0001 P < 0.025 a Expressed as multiples of the minimum wage in Brazil (about US$ 60) b Figures in parentheses are percentages. c Figures in italics are 95% confidence limits. d None = never been to school; low = primary school; medium = secondary school; high = high school or university. Table 4: Prevalence and intensity of Schistosoma mansoni infection according to sanitation and water supply conditions in the study households No. of children Mean number of examined % prevalence eggs/g stool Odds ratio Latrine: Flush with sewage disposal 936 (62.7)8 26.1 79: 67_93b 1.0 Flush without sewage disposal 53 (3.5) 32.1 60: 30-119 1.4: 0.7-2.5 Pit 285 (19.1) 34.4 94: 72-121 1.9: 1.3-2.4 None 220 (14.7) 38.2 92: 70-122 1.8: 1.3-2.4 P< 0.0001 P> 0.10 Water supply: Piped with wastewater disposal 671 (44.8) 24.3 81: 73-89 1.0 Piped without wastewater disposal 716 (47.8) 33.8 82: 76-90 1.6: 1.2-2.0 Pond 49 (3.3) 22.4 134: 99-181 0.9: 0.4-1.9 Open bodies of water 61 (4.1) 52.5 107: 89-128 3.4: 2.0-6.1 P<0.0001 P> 0.10 a Figures in parentheses are percentages. Figures in italics are 95% confidence limits. the 5% level. Children who reported playing in water or swimming during the previous month exhibited twice the prevalence and a considerably higher intensity of S. mansoni infection than other children. Fishing was closely associated with an increased risk of infection with S. mansoni. The prevalence and intensity of S. mansoni infection decreased sharply as the frequency of day-to-day contact with water fell off. For children who reported frequent contact with water, the preval- ence of S. mansoni infection was 60.6% and the mean egg count was 177 eggs per gram of stool. Previous stool examinations and antischistosomal treatment In San Antonio de Jesus there were four labora- tories at the time of the study: one state owned, a FUNRURAL (state insurance scheme for agricul- tural workers), and two privately owned. A total of 54.5% of the study population stated that a sample of their stools had been examined at least once (Table 6). A higher prevalence of infection with S. mansoni was observed for those who had not had their stools examined, but the mean egg count was WHO Bulletin OMS. Vol 69 1991.96 Distribution of S. mansoni infection in north-east Brazil Table 5: Prevalence and intensity of Schistosoma mansoni infection among study children according to their activities involving contact with water Mean number of No. of children examined % prevalence eggs/g stool Odds ratio Playing in water or swimming (ever): No 141 (8.3)a 9.2 53: 32-89 1.0 Yes 1559 (91.7) 33.0 94: 84-106 4.8: 2.6-8.3 P < 0.0001 Playing in water or swimming (past month): No 1314 (86.0) Yes 338 (20.5) P < 0.0001 Fishing: No 1442 (86.0) Yes 234 (14.0) P < 0.0001 Frequency of contacts with water: None 1264 (74.9) Sporadic 264 (15.6) Frequent 160 (9.5) P < 0.0001 Figures in parentheses are percentages. Figures in italics are 95% confidence limits. 26.3 51.2 28.0 49.6 P> 0.10 87: 75-100 109: 88-134 p < 0.05 84: 74-95 136: 103-181 P < 0.001 25.3 39.8 60.6 78: 73-84 83: 74-92 177: 132-273 P < 0.0001 1.0 2.9: 2.3-3.8 1.0 2.5 1.9-3.4 1.0 1.9: 1.5-2.6 4.5: 3.2-6.5 Table 6: Prevalence and intensity of Schis examinations and antischistosomal treatmi No. of Previous stool examination: Yes No No. of previous stool examinations: 1 2-4 --5 Didn't know Time since last stool examination (years): <1 1-2 3 Didn't know History of S. mansoni infection: Yes No Didn't know Antischistosomal treatment: Yes No Didn't know a Figures in parentheses are percentages. b Figures in italics are 95% confidence limits. WHO Bulletin OMS. Vol 69 1991. stosoma mansoni infection among the study children according to previous stool ent children examined % prevalence Mean number of eggs/g stool 924 (54.5)a 27.1 96: 81-114 770 (45.5) 35.7 91: 78-107 347 (37.6) 368 (39.8) 98 (10.6) 111 (12.0) P < 0.001 31.7 26.3 16.3 P< 0.01 237 (25.6) 325 (35.2) 353 (38.2) 9 (1.0) P < 0.05 80 (8.7) 698 (75.5) 146 (15.8) P < 0.025 71 (88.8) 8 (10.0) 1 (1.3) P> 0.10 P> 0.10 P> 0.10 P> 0.10 P> 0.10 P> 0.10 113: 88-146 100: 76-132 39: 23-65 100: 69-143 99: 71-139 92: 73-116 116: 68-198 87: 72-106 113: 65-197 137: 42-444 24.9 23.7 31.7 40.0 26.2 39.4 50.0 97 M. L. Barreto not significantly different from that of children whose stools had been examined previously. Only 80 (8.7%) of those whose stools had been examined previously reported that the result had been positive for S. mansoni eggs. This group was diagnosed to have a prevalence of 40.0% compared with 26.2% for individuals who had had no S. mansoni infection in the past. Use of antischisto- somal drugs was reported by 71 (88.0%) of those with a history of positive infection with S. mansoni and 8 (10.0%) reported receiving no treatment at all. Because of the small number of cases involved, however, comparison of these two sets of individuals was not possible. Distribution of snails and schistosomiasis prevalence The topography surrounding San Antonio de Jesus consists of flat valleys and small hills. Considerable inequalities of wealth distribution exist, with better- off areas often juxtaposed with very poor areas. In the town the only snail that is host to S. mansoni is Biomphalaria glabrata. These were found at various places; however, some degree of aggregation in their distribution was observed (Fig. 1). A total of 625 snails were collected and 19.5% of these were infected. The rate of infection, depending on collec- tion point, varied from 0% to 100%. The prevalence and intensity of S. mansoni infection varied from 70.5% in census tract 1 to 13.6% in census tract 14. The spatial pattern of the prevalence in the various tracts is shown in Fig. 2. There were two geographically defined groups of tracts where the prevalence was greater than 30%, which made it possible to define three zones where the prevalence of infection and mean egg counts were as follows: 50.4% and 100 eggs/g of stool (zone 1); 34.4% and 77 eggs/g of stool (zone 2); and 23.8% and 80 eggs/g of stool (zone 3). In zone 1, where 15.0% of the study population lived, 29.9% of subjects excreted more than 400 eggs/g of stool. Discussion The results presented here show that schistosomiasis mansoni was moderately prevalent in children in the study area. While the study was not intended to test a model of the multicausal relationships leading to infection, the results suggest that the distribution of the disease follows a pattern related to the complex- ity of the urban infrastructure. Several variables that are related to different aspects of the population's way of life were strongly associated with the prevalence and the intensity of infection. The results are useful for understanding the mechanisms in- volved in the occurrence and distribution of schisto- somiasis in an urban setting and for identifying high-risk groups. One important finding was the role played by migration in determining the levels of schistosomia- sis infection. The prevalence and intensity of infec- tion were significantly lower among children born in San Antonio de Jesus and those who had lived there for a long time. The duration of residence in rural areas was strongly associated with the level of infection (Table 2). B. glabrata snails were found that had a high rate of schistosome infection in Fig. 1. Map of Santo Antonio de Jesus, showing census tracts, water sources and snail breeding sites; * = infected snails; * = uninfected snails. WHO Bulletin OMS. Vol 69 1991. 'K -I -< 0 V 98 Distribution of S. mansoni infection in north-east Brazil Fig. 2. Map of Santo Antonio de Jesus, showing the three zones formed by the aggregation of the census tracts according to the prevalence of Schistosoma mansoni infection; zone 1, prevalence <30%; zones 2 and 3, prevalence >30%. several bodies of water in various parts of the study area (Fig. 1). Taken together, these findings emp- hasize that S. mansoni infection is transmitted locally, but at a low level of intensity and that immigration, mainly from rural areas, is an impor- tant and permanent source of new stocks of infected subjects. In Brazil, such migration is a direct consequence of an enduring land tenure system that deprives most people of land and agricultural income, while leaving large rural areas unused (16). On the other hand, the development of industries, businesses, and services in some urban settings attracts landless individuals from rural areas. Cities and towns have grown more rapidly than compatible infrastructures, creating unhealthy peripheral areas that favour the focalization of many transmissible diseases, whether or not they are carried by rural migrants. For example, epidemics of malaria and leishmaniasis have occurred in this way (17). The negative association reported here between the prevalence and the intensity of infection and socioeconomic status is consistent with the findings of other workers (4, 18).b Although such variables have no direct bearing on the transmission of S. mansoni, they are closely linked with characteristics such as migration status, sanitation, water supply facilities and contact with water, all of which are more directly associated with the transmission. b Loureiro, S. [Epidemiological variables in schistosomiasis mansoni.] Thesis. Federal University of Bahia, Brazil, 1973 (in Portuguese). Other variables, such as sanitation and water supply conditions, should play a major role in the transmission of schistosomiasis and, as might be expected, the absence of a safe source of water supply in households was associated with high levels of infection. Also, a high prevalence of infection occurred among children from households that had no kind of sewage disposal system (Table 4). Our data suggest that at the individual level the lack of a supply of water is more serious than the absence of a latrine in the household. While a small number of people reported that there was no sewage disposal system in their household, there was no guarantee that some of the existing cesspits were safe enough to prevent faecal contamination of the environment. The high percentage of snails that were infected with S. mansoni, especially in certain locations, sug- gested that such contamination was very frequent (19, 20). Among the human activities identified, those associated with water were the most immediate link in the causal network of schistosomiasis transmis- sion. Association between such activities and the level of infection has been shown previously, not only qualitatively but also quantitatively through dose-response relationships (21, 22).C Because our survey was restricted to a specific age group, the strong associations between activities related to c Loureiro, S. Schistosomiasis mansoni in children-an epide- miological study of patterns of water exposure using path analy- sis. PhD thesis, University of Texas, 1978. WHO Bulletin OMS. Vol 691991. 99 M. L. Barreto contact with water and the prevalence and intensity of S. mansoni infection may have been favoured, and are similar to findings for Brazilian schoolchild- ren (23). The results of the present and several other studies emphasize the importance of playing in water or swimming on the level of S. mansoni infection in the population, in general, particularly among schoolchildren (24). The process of self- evaluating the extent of contact with open bodies of water proved to be useful for identifying individuals with high levels of infection, since those who reported frequent contact with water on a day-to- day basis had high prevalences and intensities of infection with S. mansoni. Fishing was also associ- ated with the highest prevalence and intensity of infection (25). The health services infrastructure in San Anto- nio de Jesus is considerable and if properly used it could have an impact on several aspects of the health status of the population. Although a high proportion of children reported that they had previously under- gone a stool examination, this had a low impact on the level of infection. A check on the validity of the findings from these examinations was not possible; however, the following evidence offers some safe- guards: the prevalence and intensity of S. mansoni infection decreased as the number of stool examina- tions reported in the past increased; and the prevalence was higher among those who reported that their last stool examination had been at least 3 years ago. Despite the availability of free drugs for treatment, the high prevalence of intestinal hel- minths in the study population must be due to very high rates of reinfection, perhaps associated with an inefficient local health system. As far as S. mansoni infection is concerned, 8.7% of those examined reported a prior infection, while 15.8% reported that they did not know their infection status. A total of 88.8% of those who reported a positive stool examination also reported having received antischis- tosomal treatment, but these individuals had a higher prevalence and intensity of infection than those who reported no previous history of infection. Antischistosomal drugs are expensive and not freely available in the study area. Comparison of the census tracts shows that there were large variations in the prevalences and mean egg excretion. The geographical pattern of S. mansoni infection in San Antonio de Jesus indicates that the tracts with the highest prevalence had a tendency to form clusters; by using a prevalence of 30% as a cut-off point three zones could be delimited (Fig. 2). Superposition of Fig. 1 and 2 shows that the two zones with high prevalence of S. mansoni infection have a large number of streams flowing through them. This observation is supported by other findings. For example, of the 25 collection points where B. glabrata were found, 18 were situated in the two high prevalence zones, while the remainder were situated close to their borders. These findings highlight the important role played by snail foci in the transmission of schistosomiasis and their constraint on the geographical distribution of infection in San Antonio de Jesus (26). Acknowledgements Dr E Veloso and her colleagues are thanked for their support and for the provision of facilities in Santo Antonio de Jesus. The hard work and dedication of Fatima, Marivaldo, Lucia, and Viviane in collecting the data is acknowledged. Dr S. Ribeiro and Dr F. Vasconcelos are thanked for backing the project. The technical staff from Superintenden- cia das Campanhas de Saude Publica (SUCAM) are thanked for their work in carrying out the stool examinations and the survey. Dr G. Cummper, Dr D. Ross, Dr B. Kirkwood, Dr S. Atkinson, Dr P. Phillips-Howard, and Dr K. Mott are thanked for reading drafts of this paper and for their helpful comments. This investigation was supported financially by the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases. Resume Facteurs geographiques et socio-economi- ques associes a la distribution des infections par Schistosoma mansoni dans une region urbaine du nord-est du Bresil Une 6tude a ete effectuee a Santo Antonio de Jesus, une ville de l'Etat de Bahia, dans le nord-est du Bresil, pour conna^itre les relations entre divers fac- teurs biologiques, socio-economiques, comportemen- taux et g6ographiques, et la pr6valence et l'intensite des infections dues a Schistosoma mansoni. La population de la ville etait d'environ 45000 habitants et l'tude 6tait centree sur tous les enfants nes en 1970 et 1971 et qui vivaient dans la ville au moment de l'enquete (ao0t-novembre 1984). Un questionnaire tres complet a ete utilise pour recueillir les informations sur chaque enfant et sur ses condi- tions de vie familiale; des 6chantillons de selles ont 6t6 egalement examines (m6thode Kato-Katz). Une enquete sur les mollusques dans la region a 6t6 egalement effectu6e et des informations sur la distri- bution de leurs sites de reproduction ont ete mar- quees sur une carte. Un total de 1765 enfants qui satisfaisaient aux conditions de l'6tude ont ete identifi6s par une visite porte a porte. La prevalence totale des infections WHO Bulletin OMS. Vol 69 1991.100 Distribution of S. mansoni infection in north-east Brazil dues a S. mansoni etait chez eux de 31,0%. Plu- sieurs variables qui refl6taient diff6rents aspects du mode de vie de la population etaient fortement asso- ciees a la prevalence et a l'intensite de l'infection: I'immigration-les immigres constituaient 46,0% de la population et ce groupe avait les taux les plus eleves de prevalence et d'intensite des infections; la situation socio-economique-34,8% des familles avaient un revenu mensuel inferieur au salaire mini- mum national (US$60), et la prevalence chez les enfants de ce groupe etait deux fois superieure a celle des enfants de families aisees; I'alimentation en eau-92,6% des habitants avaient une alimentation en eau par canalisation; cependant, les enfants des 7,4% d'habitants restants avaient un taux eleve d'infection; I'hygiene-66,2% des foyers avaient des toilettes avec chasse d'eau mais certains de ces foyers n'avaient aucune installation d'evacuation des eaux usees, ce qui favorise la contamination des etendues d'eau locales; les activites supposant un contact avec l'eau-ces activites, deja signalees, sont fortement associees a la prevalence et a l'inten- site de l'infection par S. mansoni, et l'auto-6valuation des contacts avec l'eau a e utile pour identifier les enfants ayant des taux 6lev6s d'infection; services de sante-bien que 54,5% des enfants aient signale que leurs selles avaient ete examin6es au moins une fois, les donn6es obtenues laissent a penser que cela n'a qu'un faible impact sur les taux d'infection. Les resultats de cette etude ont ete utiles pour comprendre les mecanismes impliques dans l'appari- tion et la distribution de la schistosomiase dans un milieu urbain, et aussi pour identifier les groupes a haut risque. Les r6sultats de l'etude font egalement penser que la distribution de la schistosomiase est influenc6e par la complexite de l'infrastructure ur- baine. References 1. Feachem, R. et al. Sanitation and disease: health aspects of excreta and wastewater management. Chi- chester, John Wiley, 1983. 2. Pesigan, T.P. et al. Studies on Schistosoma japonicum infection in the Philippines. 3. Preliminary control experi- ments. Bulletin of the World Health Organization, 19: 223-261 (1958). 3. Farooq, M. et al. The epidemiology of Schistosoma haemotobium and S. mansoni infections in the Egypt-49 project area. 2. Prevalence of bilharziasis in relation to personal attributes and habits. Bulletin of the World Health Organization, 35: 293-318 (1966). 4. Farooq, M. et al. The epidemiology of Schistosoma haematobium and S. mansoni infections in the Egypt-49 project area. 3. Prevalence of bilharziasis in relation to certain environmental factors. Bulletin of the World Health Organization, 35: 319-320 (1966). 5. Clima, N.E. In: Geografia do Brasil 2. Regiao Nordeste. Rio de Janeiro, FIBGE, 1977 (in Portuguese). 6. Fundacao Instituto Brasileiro de Geografia e Estat- istica. IX. General Census of Brazil 1980. Vol. 2, Tome 3. Bahia, Rio de Janeiro, FIBGE, 1983 (in Portuguese). 7. Barbosa, F.S. [Schistosomiasis morbidity.] Revista Bra- sileira de malariologia e doencas tropicais, (special issue) 3-159 (1966) (in Portuguese). 8. Lehman, J.S. Jr. et al. The intensity and effects of infection with Schistosoma mansoni in a rural commun- ity in northeast Brazil. American journal of tropical medicine and hygiene, 25: 285-294 (1976). 9. Katz, N. et al. A simple device for quantitative determi- nation of Schistosoma mansoni eggs in faeces exa- mined by the thick-smear technique. Revista do Instituto de Medicina Tropical de Sao Paulo, 14: 394-400 (1972). 10. Olivier, I.T. Techiques. In: Ansari, N., ed. Epidemology and control of schistosomiasis (bilharziasis). Basel, Karger, 1973. 11. Paraense, W.L. et al. [Distribution of planorbids and the prevalence of Schistosoma mansoni in Estado do Es- pirito Santo.] Memorias do Instituto Oswaldo Cruz, 78: 373-384 (1983) (in Portuguese). 12. Blalock, H.M. Jr. Social statistics. Auckland, McGraw Hill, 1981. 13. SPSS-X user's guide. New York, McGraw Hill, 1983. 14. Armitage, P. Statistical methods in medical research. Oxford, Blackwell, 1971. 15. Breslow, N.E. & Day, N.E. Statistical methods in cancer research. Volume 1: the analysis of case-control studies. Lyon, International Agency for Research in Cancer, 1980. 16. Singer, P. [Political economy of urbanization.] Sao Paulo, Brasiliense, 1976 (in Portuguese). 17. Marques, A.C. [Internal migrations and major ende- mics.] Revista Brasileira de malariologia e doencas tropicais, 31: 137-158 (1979) (in Portuguese). 18. Costa, M.F.F.L. et al. Water contact patterns and socioeconomic variables in the epidemiology of schisto- somiasis mansoni in an endemic area in Brazil. Bulletin of the World Health Organization, 65: 57-66 (1987). 19. Bayer, F.A.H. Schistosome infection of snails in a dam traced to pollution with sewage. Transactions of the Royal Society of Tropical Medicine and Hygiene, 48: 347-350 (1954). 20. Paulini, E. [Contributions by sanitary engineering to the proliferation of schistosomiasis and their value from the epidemiological point of view.] Revista Brasileira de malariologia e deoncas tropicais, 18: 163-174 (1964) (in Portuguese). 21. Dalton, P. A sociological approach to the control of Schistosoma mansoni in St. Lucia. Bulletin of the World Health Organization, 54: 587-595 (1976). 22. Dalton, P.R. & Pole, D. Water contact patterns in relation to Schistosoma haematobium infection. Bulletin of the World Health Organization, 56: 417-426 (1978). 23. Guimaraes, M.D.C. et al. [Clinical-epidemiological study of schistosomiasis mansoni in schoolchildren in lIha, Arcos county, Mines Gerais, Brazil.] Revista de saude WHO Bulletin OMS. Vol 69 1991. 101 M. L. Barreto publica de Sao Paulo, 19: 8-17 (1985) (in Portuguese). 24. Barreto, M.L. [Cause versus prediction: bathing as a risk factor and predictor of Schistosoma mansoni infec- tion.] Revista de saude publica de Sao Paulo, 21: 305-309 (1987) (in Portuguese). 25. Tayo, M.A. et al. Malumfashi endemic disease research project, XI. Water contact activities in the schistosomia- sis study area. Annals of tropical medicine and parasito- logy, 74: 347-354 (1980). 26. Kloos, H. ot al. Water contact behaviour and schistoso- miasis in an upper Egyptian village. Social science and medicine, 17: 545-562 (1983). 27. Waugh, T.C. & McCalden, J. GIMMS reference manual, release 4.5. Edinburgh, Gimms, 1983. 102 WHO Bulletin OMS. Vol 69 1991.