International Journal of Epidemiology
© International Epidemiological Association 1996
Vol. 25, No. 6
Printed in Great Britain
Urban Schistosomiasis: Morbidity,
Sociodemographic Characteristics
and Water Contact Patterns
Predictive of Infection
JOSELIA O A FIRMO,* MARIA FERNANDA LIMA E COSTA,*" HENRIQUE L GUERRA* AND
ROBERTO S ROCHA*
Firmo J O A (Laboratory of Epidemiology and Medical Anthropology and Laboratory of Schistosomiasis, Rene Rachou
Research Institute, Oswaldo Cruz Foundation, Av. Augusto de Lima 1715, 30.190.002, Belo Horizonte, Minas Gerais,
Brazil), Lima e Costa M F, Guerra H L and Rocha R S. Urban schistosomiasis: morbidity, sociodemographic characteristics
and water contact patterns predictive of infection. International Journal of Epidemiology 1996; 25: 1292-1300.
Background. Schistosomiasis (Schistosoma mansoni) is classically described as a rural disease that occurs in areas with
poor sanitary conditions. This cross-sectional study was undertaken in a suburban area of a large industrialized city in
Brazil (Belo Horizonte), aiming at examining epidemiological characteristics of schistosomiasis in an urban setting.
Methods. A simple random sample of 658/1896 dwellings was selected and 3049/3290 (92.7%) residents were submitted to stool examination. Of 518 eligible infected cases and 518 uninfected controls, 87.1% and 89.8% participated in the
study, respectively.
Results. The prevalence of S.mansoni infection was 20%, predominantly low egg counts in stools; no cases of splenomegaly
were found. Signs and symptoms associated with infection were bloody stools (odds ratio [OR] = 8.0) and hardened
palpable liver at the middle clavicular and at the middle sternal line (OR = 5.5 and 8.0, respectively). Sociodemographic
variables and water contacts predictive of infection were age (10-19 and s=20 yrs; OR = 7.1 and 3.3, respectively),
gender (male; OR = 3.1), contacts for swimming and/or playing (twice a month or less and more than twice a month;
OR = 2.2 and 3.0, respectively) and residence in Belo Horizonte (born in the City; OR = 2.5). Ninety per CRnt of dwellings
had a piped water supply; no association between water supply and infection was found.
Conclusion. Our results emphasize the need for schistosomiasis control measures focusing on water contacts for leisure
purposes in this industrialized urban area.
Keywords: urban schistosomiasis, epidemiology
included in long-term national strategies, after identification of both general and specific control measures.1
Effective control measures depend on knowledge of the
epidemiology of parasitic diseases in urban settings.
Epidemiological studies on S.mansoni in large cities are
few and concentrate, as a rule, on identification of foci of
transmission and/or on descriptive information related to
the prevalence of infection in selected populations.2"6'8"12
An exception is a study carried out during the early 1970s
in a district of Salvador City, northeast of Brazil, where
morbidity and sociodemographic variables predictive of
S.mansoni infection were investigated. A comparison of
infected cases and uninfected controls in this area showed
that infection was associated with low education and low
socioeconomic status, migration, and rural origin;
morbidity due to infection was low.7
Schistosomiasis is classically described as a rural disease that occurs in areas with poor sanitary conditions.
Nevertheless, owing to migration, urban areas in Africa
and South America are now foci of transmission.1 In
Africa, the disease is endemic in cities like Harare,2
Kinshasa 3 and Dar es Salaam.4 In Brazil, foci of transmission of Schistosoma mansoni were described in
large cities (over one million inhabitants) situated in the
north, 5 the northeast, 6 ' 7 and in the industrialized
southeast of the country.8"12
Recently, it has been proposed that programmes
for parasitic disease control in urban areas should be
* Laboratory of Epidemiology and Medical Anthropology and Laboratory of Schistosomiasis, Rene Rachou Research Institute, Oswaldo
Cruz Foundation, Av. Augusto de Lima 1715, 30.190.002, Belo
Horizonte, Minas Gerais, Brazil.
*• Department of Preventive and Social Medicine, School of
Medicine, Federal University of Minas Gerais, Belo Horizonte, Minas
Gerais, Brazil.
This paper describes a study of the epidemiology of
S.mansoni, which was undertaken in a suburban area of
a large industrialized city in Brazil (Belo Horizonte).
1292
1293
SCHISTOSOMIASIS IN AN URBAN AREA OF BRAZIL
RiodasVWhas
Naigbornarea
1 A
Lagoa dos Borges
(neigborn area)
FIGURE 1 Map of the studied area with points for monitoring snail population
(planorbides)
The objectives of the study were: (i) to examine the
existence of foci of transmission; (ii) to assess the prevalence and intensity of infection in a representative
sample of the total population; (iii) to assess the morbidity due to infection and (iv) to assess sociodemographic variables and water contact patterns predictive
of S.mansoni infection in the study area.
SUBJECTS AND METHODS
Study Area
Belo Horizonte is the capital of the State of Minas
Gerais which is situated in southeast Brazil. The city
has around two million inhabitants.13 The study was
undertaken during 1991-1992 in a district named
Gorduras, in the northeast of the city.
Foci of Transmission
All water sources within and surrounding the study area
were mapped. Snail surveys were carried out monthly
over one year along the mapped streams, and in a small
lake used by the population (Figure 1). Snails collected
were transported to the laboratory where they were
measured, classified and examined microscopically,
following squashing between two glass plates, for the
presence of cercariae or sporocysts of S.mansoni.1*
Study Population
The study area was mapped and all dwellings (n = 1896)
were numbered (Figure 1). A simple random sample of
658 dwellings was selected and 3290 residents were
chosen to participate in the descriptive study (sociodemographic survey and stool examinaton). Assumptions for the sample size calculation were the following:
number of inhabitants = 9480 (five per dwelling); prevalence of S.mansoni infection = 0.20; precision = 0.02;
type 1 error = 0.05; correction for design effect = 2;
losses = 0.20.l5-15 A signed informed consent was obtained from all participants or from their parents/adult
responsible if they were under 18 years of age.
1294
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Sociodemographic Survey
A sociodemographic survey was performed in all sampled households. One adult occupant in each household
was interviewed. All information was obtained by two
interviewers without knowledge of individuals' laboratory results. The questionnaire consisted of three categories of explanatory variables: (a) demographic (age,
gender and residence in Belo Horizonte), (b) socioeconomic (family monthly income and education), and
(d) sanitation (water supply and sewerage system in the
household).
Stool Examination and Selection of Cases
and Controls
All residents in the sampled households received a
container for collection of stools, labelled with their
name and identification number, as well as an identifiable code in case of illiteracy. Two separate stool
samples were collected within a 4-month interval;
they were examined for the presence and number of
S.mansoni eggs and for the presence of Ascaris lumbricoides, Trichiura trichiuris and ancilostomideos eggs.
Stool examinations were performed by the Kato-Katz
method. 17 The number of S.mansoni eggs per gram was
estimated from the mean of four slides. All examinations were done by two trained technicians. Agreement
between technicians, considering the presence or absence
of S.mansoni eggs, was assessed in a sample of 100 stool
examinations.
Cases were defined as those who presented S.mansoni
eggs on at least one of four slides. Controls were
defined as those without S.mansoni eggs on four slides.
A simple random sample of individuals over 2 years of
age who presented eggs in stools (518/609) and a
simple random sample of those over 2 years of age who
did not present S.mansoni eggs (518/2256) was selected. The sample size was sufficient for detecting an
odds ratio (OR) of 1.5 with 90% power at 5% level of
significance, if the prevalence of exposure was 50%. l 8
Clinical Examination
Clinical examination was undertaken by a single physician without prior knowledge of patients' infection
status. Both physician and patient were informed of the
infection status at the end of the clinical session, when
all cases received oxamniquine for treatment. All cases
and controls were eligible for clinical examination.
Patients were submitted to a questionnaire concerning
the occurrence of diarrhoea, abdominal pain or bloody
stools over the previous 30 days, previous treatment for
schistosomiasis, and history of splenectomy, haematemesis or melaena during their lifetime. Abdominal
examinations were done with the patients lying on their
back in the right lateral position. The liver or spleen
was considered palpable when detectable immediately
under the costal margin with the breath held.
Water Contact Interview
Water contact interviews were also double blind and
were done after the second stool collection. All cases
and controls were eligible for interviews. A pre-coded
questionnaire concerning frequencies and reasons
for water contacts in streams and lakes situated in
Belo Horizonte City during the previous 6 months
was used. Reasons for water contact included: laundry,
washing dishes, bathing (body hygiene), fishing,
swimming or playing and other. Water contact frequencies were classified as none (no contact in the previous 6 months), twice a month or less and more than
twice a month.
Analysis
Crude analysis was based on overall %2, y} for linear
trend (for proportions) and Kappa statistics with confidence interval (CI) (for interobserver agreement).19
Crude and adjusted OR and CI ( Woolf's method) were
also calculated.19 Multivariate logistic regression was
used to assess the independent effect of variables.20
Two logistic models were constructed: the first to assess
signs and symptoms and the second to assess sociodemographic variables and water contact patterns
associated with S.mansoni infection. The criterion for
inclusion of variables in each logistic model was association with infection beyond 0.20 significance level in
the univariate analysis.21 Abdominal pain and bloody
stools were adjusted for age, gender, previous treatment
for schistosomiasis and other helminths (Ascaris lumbricoides, Trichiura trichiuris and ancilostomideos).
Palpable liver at the middle clavicular line and palpable
liver at the middle sternal line were adjusted for age,
gender and previous treatment; other helminths were
not included in this analysis because there is no evidence that they can produce liver impairment. The initial
logistic model for sociodemographic variables and
reasons for water contact included age, gender, previous treatment, residence in Belo Horizonte, piped water
in the household and water contacts for laundry,
bathing, fishing and swimming or playing. The analysis
was done using Epi Info,15 Egret22 and SAS 23 software
packages.
RESULTS
Foci of Transmission
Biomphalaria glabrata was found at all points along
streams which were selected for monitoring snails.
SCHIST0S0M1ASIS IN AN URBAN AREA OF BRAZIL
B.glabrata infected with S.mansoni were found at
6 points points 1, and 7-11). No snails were found at
the monitored small lake. Sewerage draining to streams
was observed at points 9 and 10 (Figure 1).
Internal Validity and Quality Control of Stool
Examination
The investigation covered 658 households with 3290
eligible residents. Of these, 3049 (92.7%) were submitted to stool examination. Of 518 eligible cases and
518 eligible controls, 451 (87.1%) and 465 (89.8%)
participated, respectively. Regarding reliability of stool
examination, Kappa value between technicians 1 and 2
was 0.96 (95% CI : 0.90-1.00).
Descriptive Results
Of the 658 sampled households, 92.7% had a piped
water supply and 89.4% had a sewerage system. Prevalence of S.mansoni infection in the sampled population
was 20.0% (609/3049), predominantly low egg counts
(61.1% of infected eliminated <100, 25.8% 100-499,
and 13.1% 3=500 eggs per gram [epg] of stool); the
geometric mean of S.mansoni eggs was 70.8 epg (95%
CI : 7.4-679.7).
Signs and Symptoms
Table 1 shows the results of the univariate analysis of
signs and symptoms, previous treatment, other selected
helminths and S.mansoni infection. Bloody stools, palpable liver at the middle clavicular line, palpable liver
at the middle sternal line, previous treatment, Ascaris
lumbricoides, ancilostomideos and Trichiura trichiuris
were associated with S.mansoni infection (P < 0.05).
No cases of splenomegaly or histories of splenectomy,
haematemesis or melaena were found in the study area.
Table 2 shows the results of the multivariate analysis of selected signs and symptoms and S.mansoni
infection. Bloody stools (OR : 8.0; 95% CI : 1.8-36.6)
persisted associated with infection after adjustments
for age, gender, previous treatment and other
helminths. Palpable hardened liver at the middle
clavicular line (OR : 5.5; 95% CI : 1.1-27.7.) and
palpable hardened liver at the middle sternal line
(OR : 8.0; 95% CI : 1.5-42.0) persisted in association
with infection after adjustments for age, gender and
previous treatment.
Sociodemographic Variables and Reasons for
Water Contacts
Table 3 shows the results of the univariate analysis
of selected sociodemographic variables and S.mansoni
1295
infection. Age group, gender and residence in Belo
Horizonte were associated with infection (P < 0.05).
Swimming and/or playing was the only reason for water
contact associated with infection; the OR increased as
the frequency of contact increased (dose-response
relationship) (Table 4).
Table 5 shows the final results of the multivariate
analysis of sociodemographic variables and reasons for
water contact and S.mansoni infection. The following
variables were independently associated with infection:
age (10-19 years; OR : 7.1; 95% CI : 4.4-11.1 and &20
years; OR : 3.3; 95% C I : 1.9-5.9), gender (male;
OR : 3.1; 95% CI : 2.3-4.1). Swimming and/or playing
(twice a month or less; OR : 2.2; 95% CI : 1.4-3.2;
more than twice a month; OR : 3.0; 95% CI : 1.5-5.8) and
residence in Belo Horizonte (born in Belo Horizonte;
OR : 2.5; 95% CI : 1.2-5.4).
DISCUSSION
This study revealed the existence of S.mansoni
transmission in the urban area and the occurrence of
infection at 20% prevalence level. Schistosomiasis
dependent morbidity was low: no cases of splenomegaly or splenectomy were found, only 4.7% of infected
subjects reported bloody stools and less than 3% presented hardened enlarged liver. These findings are consistent with low S.mansoni egg counts observed in the
population.24
Studies undertaken in different countries, using
univariate methods of analysis, have found a consistent
association with bloody stools and S.mansoni infection.
Association of diarrhoea or abdominal pain with infection was observed in some endemic areas but not in
others.25"28 In this investigation, an association of bloody
stools with infection was found. The association persisted after adjustments for age, gender, previous treatment and other selected helminths.
Liver impairment is part of schistosomiasis pathogenesis and enlarged liver without specification of its
consistency has been traditionally used as an indication
of schistosomiasis morbidity in areas where malaria or
visceral leishmaniasis are not endemic.29 Recent studies
have shown that palpable liver with normal consistency
has smaller positive predictive value for infection than
hardened liver or bloody stools, indicating that consistency of liver and bloody stools must be considered in
schistosomiasis morbidity studies.27'28 Our results are
consistent with this view.
Sociodemographic variables and reasons for water
contact which were predictive of infection in our study
were age, gender, water contact for swimming and/or
1296
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
TABLE 1 Univariate analysis of signs and symptoms, previous treatment, other selected helminths and S.mansoni infection
Variables
Abdominal pain:
No
Yes
Cases
Controls
310(68.7)
141 (31.3)
Crude odds ratio (95% CI)
299 (64.3)
166(35.7)
1.0
0.8(0.6, 1.1)
404 (86.9)
'***61 (13.1)
1.0
1.0(0.7, 1.4)
/> = 0.155
Diarrhoea:
No
Yes
394 (87.4)
57(12.6)
P = 0.828
Bloody stools:
No
Yes
430 (95.3)
21 (4.7)
463 (99.6)
2 (0.4)
1.0
11.3(2.6,48.5)
424 (92.8)
31 (6.8)
2 (0.4)
1.0
0.4 (0.2, 0.8)
5.0(1.1,23.0)
432 (94.6)
23 (5.0)
2 (0.4)
1.0
0.3(0.1,0.7)
5.5 (1.2,25.1)
409 (88.0)
56(12.0)
1.0
2.4(1.7, 3.4)
287(61.7)
178(38.3)
1.0
1.3 (1.0, 1.7)
457 (98.3)
8(1.7)
2.7 (1.1,6.6)
P = 0.000
Palpable liver at middle clavicular line:
Not palpable
Palpable normal consistency
Palpable hardened
423 (94.6)
13(2.9)
11 (2.5)
P = 0.001
Palpable liver at middle sternal line:
Not palpable
Palpable normal consistency
Palpable hardened
429 (96.0)
7(1.6)
11 (2.4)
P = 0.001
Previous treatment:
No
Yes
340 (75.4)
111 (24.6)
P = 0.000
Ascaris lumbricoides:
No
Yes
248 (55.0)
203 (45.0)
P = 0.039
Ancilostomideos:
No
Yes
431 (95.6)
20 (4.4)
1.0
/> = 0.017
Trichiura trichiuris:
No
Yes
383 (82.4)
82(17.6)
347 (76.9)
104 (23.1)
1.0
1.4 (1.0,2.0)
P = 0.041
Cases: presence of S.mansoni eggs in stools.
Controls: absence of S.mansoni eggs in stools.
P = P value (x2 test).
playing and residence in Belo Horizonte. Age was the
variable that showed the strongest association with
S.mansoni infection, with the highest OR for the second
decade of life. Similar results were found in rural areas
in the State of Minas Gerais where prevalence or incidence of infection was considered.28"30 Regarding the
predominance of infection among males, this is probably a consequence of different water contact patterns
between males and females in the study area.31
A previous study undertaken in Salvador City in
Brazil showed that S.mansoni infection was associated
with low education and low socioeconomic status,
migration and rural origin.7 In our study, infection was
not associated with family monthly income, education
and/or migration. OR for infection presented the highest value for those born in the city.
Swimming and/or playing was the only reason for
water contact that was predictive of infection in this
endemic area. Laundry, washing dishes and bathing
(body hygiene), which are water contacts related to
absence of water supply in the household, were not
associated with S.mansoni infection. These results are
1297
SCHISTOSOM1ASIS IN AN URBAN AREA OF BRAZIL
TABLE 2 Multivariate analysis of selected signs and symptoms
and S.mansoni infection
Signs or symptoms
Adjusted odds ratio (95% CI)
Abdominal pain (yes)
Bloody stools (yes)
Palpable liver at the middle clavicular line:
Normal consistency
Hardened
Palpable liver at the middle sternal line:
Normal consistency
Hardened
0.9(0.7, 1.2)a
8.0(1.8,36.6)'
0.8(0.4, 1.8)b
5.5(l.l,27.7) b
0.6(0.2, 1.6)b
8.0(1.5,42.0)"
* Odds ratio adjusted by multiple logistic regression method for age,
gender, previous treatment, Ascaris lumbricoides, Trichiura trichiuris
and ancilostomideos; no = 0 and yes = 1 (916 individuals participated
in the final analysis).
b
Odds ratio adjusted by multiple logistic regression method for age,
gender, and previous treatment; reference class was liver not palpable
(904 individuals participated in the final analysis).
coherent with the existence of 93% of dwellings with
piped water supply in the study area.
Regarding methodological aspects, all precautions
were taken to avoid inaccuracies in this study: random
procedures, double blind information, quality control of
stool examination, improvement of sensitivity of the
Kato-Katz method by means of two instead of one stool
examination, and extensive training of interviewers and
technicians. Potential sources of bias are those which
can occur in cross-sectional studies when (1) the
characteristic is a consequence of infection and/or (2) it
changes after infection diagnosis and/or (3) it is associated with duration of infection (duration of S.mansoni
infection is estimated at around 3-10 years 31 ). In our
study the first bias is evident in the univariate analysis
(infected cases were more likely to be previously
treated than controls), but this effect disappeared after
adjustment for confounders. The second bias would
occur if individuals have changed their reasons for
TABLE 3 Univariate analysis of selected sociodemographic variables and S.mansoni infection
Variables
Cases
Age group (years):
2-9
10-19
3=20
32(7.1)
291 (64.5)
128 (28.4)
Gender:
Female
Male
146 (32.4)
305 (67.6)
Family monthly income:1
<1
1-2
53
12(2.7)
245 (54.3)
194 (43.0)
Attendance at school (years):b
<1
1-4
55
42 (9.5)
360(81.6)
39 (8,8)
Residence in Belo Horizonte:
< 5 years'
5 5 years'
Born in Belo Horizonte
15(3.3)
201 (44.6)
235(52.1)
Piped water in the household:
Yes
No
399 (88.5)
52(11.5)
Sewerage system in the household:
Yes
No
408 (90.5)
43 (9.5)
* In minimum wages (I minimum wage = USS 50.00).
b
For those ^ 7 years old.
c
For those who were not born in Belo Horizonte.
Cases: presence of S.mansoni eggs in stools.
Controls
P = 0.000
P = 0.000
/* = 0.616
F = 0.465
F = 0.000
P = 0.140
P = 0.970
Crude odds ratio (95% CI)
144(31.0)
163 (35.0)
158(34.0)
1.0
8.0(5.2, 12.2)
3.7 (2.4, 5.8)
278 (59.8)
187(40.2)
1.0
3.1 (2.4,4.0)
8(1.7)
253 (54.4)
204 (43.9)
0.6(0.3, 1.6)
0.6(0.3, 1.6)
1.0
41 (10.7)
312(81.5)
30 (7, 8)
1.0
25 (5.4)
258 (55.5)
182(39.1)
1.0
1.3(0.7,2.5)
2.2(1.1,4.2)
425(91.4)
40 (8.6)
1.0
1.4(0.9,2.1)
421 (90.5)
44 (9.5)
1.0
1.0(0.6, 1.6)
Controls: absence of S.mansoni eggs in stools.
P = P value (x2 test).
F = P value (x 2 for linear trend).
1.1 (0.7, 1.8)
1.3(0.6,2.5)
1 298
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
TABLE 4 Univariate analysis of reasons for water contact and S.mansoni infection
Variables
Controls
(n = 465) %
Cases
(n = 451) '
Laundry:
No
Yes
84.0
16.0
Crude odds ratio (95% Cl)
79.4
20.6
1.0
0.7(0.5, 1.0)
73.5
26.5
1.0
I.I (0.9, 1.5)
75.3
24.7
1.0
1.2(0.9, 1.7)
97.2
1.9
0.9
1.0
2.1 (0.9,5.0)
1.3 (0.4,4.2)
85.6
11.2
3.2
1.0
2.8 (1.9,4.0)
3.8(2.1,6.9)
64.3
35.7
1.0
1.1 (0.8, 1.4)
P = 0.067
Washing dishes:
No
Yes
71.0
29.0
P = 0.380
Bathing:
No
Yes
71.2
28.8
P = 0.162
Fishing:
No
Twice a month or less
More than twice a month
95.1
3.6
1.3
/>' = 0.140
Swimming and/or playing:
No
Twice a month or less
More than twice a month
65.9
24.0
10.1
P' = 0.000
Other reasons:
No
Yes
62.7
37.3
P = 0.626
Cases: presence of S.mansoni eggs in stools.
Controls: absence of S.mansoni eggs in stools.
P = P value (x2 test).
P' = P value (x2 f° r linear trend).
TABLE 5 Final results of the muliivariale analysis of sociodemographic variables, reasons for water contact and S.mansoni
infection
Variables
Age (years):
10-19
5=20
Gender (male)
Swimming and/or playing (yes):
Twice a month or less
More than twice a month
Residence in Belo Horizonte:
3=5 years
Born in Belo Horizonte
Adjusted odds ratio (95% CI) a
7.1 (4.4, 11.4)
3.3(1.9,5.9)
3.1 (2.3,4.1)
2.2(1.4.3.2)
3.0(1.5,5.8)
1.6(0.7.3.3)
2.5 (1.2.5.4)
"Odds ratio adjusted by multiple regression method. Reference classes
were those listed in Table 4 (916 individuals participated in the final
analysis: log likelihood ratio statistic for 8 d.f. = 215.451: P < 0.01).
water contact and/or other characteristics after a previous infection diagnosis.32 In this case, OR for reasons
for water contact could be underestimated in our study;
family monthly income, residence in Belo Horizonte,
water supply and sewerage systems in dwellings are not
changeable by infection diagnosis. In relation to the
third potential bias, there is no evidence that the study
characteristics, except treatment, are related to duration
of S.mansoni infection.
Summarizing, our results lead to the following conclusions: (1) there were foci of S.mansoni transmission
in streams situated in the study area and in the surrounding area; (2) the prevalence of infection was 20%,
the intensity of infection was low, as well as schistosomiasis morbidity (low prevalences of bloody stools,
and liver enlargement, and absence of spleen enlargement); (3) signs and symptoms associated with infection
were bloody stools and hardened palpable liver; (4) age,
gender, residence in Belo Horizonte and water contacts
1299
SCHISTOSOMIAS1S IN AN URBAN AREA OF BRAZIL
for swimming and/or playing were predictive of
infection; these variables might be useful for infection
screening in this endemic area. Ninety per cent of
dwellings had piped water supply and no association
between infection and water supply was found. Our
results emphasize the need for schistosomiasis control
measures focusing on water contacts for leisure in this
industrialized urban area.
ACKNOWLEDGEMENTS
This study was supported by grants from UNDP/World
Bank/WHO Special Programme for Research and
Training in Tropical Diseases and from Financiadora de
Estudos e Projetos (FINEP), Brazil. The authors also
wish to thank Mrs Virginia Schall and Dr Naftale Katz
for their support throughout this project.
REFERENCES
1
Motl K E, Desjeux P, Moncayo A, Ranque P, de Raadt P.
Parasitic diseases and urban development. Bull World
Health Organ 1990; 68: 691-98.
2
Chandiwana S K. Human bilharziasis in a peri-urban area in
Zimbabwe with special reference to its relationship to
malnutrition in school children. Central African J Med
1983; 29: 23-26.
3
Grysells B, Ngimbi N P. Further observations on the urban
Schistosoma mansoni focus in Kinshasa, Zaire. Ann Soc
Beige Med Trop 1983; 63: 341-46.
4
Sarda R K, Simonsen L F, Maitikwquano L F. Urban
transmission of urinary schistosomiasis in Dar es Salaam,
Tanzania. Ada Tropica 1985; 42: 71-78.
5
Crescente J A B, Bichara C N, Cardoso B S, Pinheiro A S,
Abdon N P, Dourado H V. Esquistossomose urbana na
Amazonia: aspectos epidemiologicos. In: Anais do XXVII
Congresso da Sociedade Brasileira de Medicina Tropical.
Uberaba, Brasil, SBMT, 1991, p. 71.
6
Almeida Y M, Souza Filho J S, Bringel J M A, Oliveira E M B,
Moreira A N L, Timbo M J. Siluacao atual da esquislossomose mansonica na cidade de Fortaleza-Ceara (1990).
In: Anais do XXVII Congresso da Sociedade Brasileira de
Medicina Tropical. Uberaba, Brasil, SBMT, 1991, p. 73.
7
Souza S L. Varidveis epidemiol6gicas na esquistossomose
mansoni. (Thesis for Assistant Lecturer). Faculdade de
Medicina da Universidade Federal da Bahia, Salvador,
Brasil, 1973.
8
Silva L J. Crescimento urbano e doen9a: a esquistososmose no
Municfpio de Sao Paulo (Brasil). Rev Saude Publica Sao
Paulo 1985: 19: 1-7.
9
Schall V T, Jurberg P, Willcox H P F, Cavalcante F G, Bagno S.
Esquistososmose mansoni autoctone e outras parasitoses
intestinais em escolares do Bairro Alto da Boa Vista, da
cidade do Rio de Janeiro. Rev Soc Bras Med Trop 1985;
18: 169-74.
10
Katz N, Motta E, Oliveira V B, Carvalho E F. Prevalencia da
esquistososmose em escolares no Estado de Minas Gerais.
In: Anais do XXVII Congresso da Sociedade Brasileira de
Medicina Tropical: Joao Pessoa, Brasil, SBMT, 1978, p. 102.
" Carvalho O S, Guimaraes C T, Massara C L, Bonesio J E R.
Situacao atual da esquistossomose mansoni no lago da
Pampulha, Belo Horizonte, Minas Gerais. Rev Saude
Publica Sao Paulo 1985; 19: 270-77.
12
Guimaraes C T, Souza C P, Carvalho O S, Katz N. Sobre urn
foco de esquistossomose em drea Metropolitans da regiao
sudeste do Brasil. Rev Saude Publica Sao Paulo 1993; 27:
210-13.
I3
IBGE. Anudrio Estatistico do Brasil, Secretaria de Planejamento, Or^amento e Coordenacao, 1992.
14
Pellegrino J, Katz N. Experimental chemotherapy of schislosomiasis mansoni. Adv Parasitol 1968; 6: 233-90.
15
Dean A G, Dean J A, Coulombier D, Brendel K A, Smith D C
et al. Epi Info, version 6. Atlanta, Georgia: Centers for
Disease Control and Prevention, 1994.
16
World Health Organization. Programme for control of diarrhoea! diseases, household sun'ey manual. Geneva: WHO,
1989 (CDD/SER/86.2 Rev.l).
17
Katz N, Chaves A, Pellegrino J P. A simple device for
quantitative stool thick-smear technique in schistosomiasis
mansoni. Rev Inst Med Trop Sao Paulo, 1972; 14: 397-400.
18
Schlesselman J J. Case Control Studies. New York: Oxford
University Press, 1982.
19
Fleiss J L. Statistical Methods for Rates and Proportions. (2nd
edn). New York, John Wiley, 1981.
20
21
Breslow N E, Day N E. Statistical
Methods
in Cancer
Research.
Vol I. The analysis of Case-Control Studies. (IARC
Scientific Publications. No. 32). Lyon: International
Agency for Research on Cancer, 1980.
Greenland S. Modeling and variable selection in epidemiologic
analysis. Am J Public Health 1989; 79: 340-49.
22
E G R E T . Seattle: Statistics and Epidemiology Research Corporation, 1993: 2 1 .
S A S . User's Guide. Version 6. (4th edn) Cary, N C : S A S
Institute, 1990.
24
Lima e Costa M F F, Magalhaes M H A, Rocha R S, Katz N. A
clinico-epidemiological survey of schistosomiasis mansoni
in a hyperendemic area in Minas Gerais State (Comercinho,
Brazil): 1 Differences in t h e manifestations of schistosomiasis in the town centre a n d in the environs. Trans R Soc
Trop Med Hyg 1985; 7 9 : 5 3 9 - 4 5 .
25
Lima e Costa M F F, Rocha R S, Katz N . Morbidade d a
esquistossomose e sua rela?ao c o m a contagem de ovos
de Schistosoma
mansoni e m u m a zona hiperendemica d o
Estado de Minas Gerais. Rev Inst Med Trop Sao Paulo
23
1985; 27:66-75.
26
27
Sukwa T Y, Bulsara M K, Wurapa F K. Evaluation of selected
symptoms in the diagnosis of Schistosoma
mansoni
infection. Trop Geogr Med 1985; 3 7 : 2 4 8 - 5 1 .
Proietti F A, Antunes C M F. Sensitivity, specificity and positive
predictive values of selected clinical signs and symptoms
associated with schislosomiasis mansoni. Int J Epidemiol
1989;18: 680-83.
28
29
30
Lima e Costa M F F, Rocha R S, Colley D, Gazzinelli G, Katz
N. Validity of selected clinical signs and s y m p t o m s in
diagnosis of Schistosoma mansoni infection. Rev Inst Med
Trop Sao Paulo 1991; 3 3 : 1 2 - 1 7 .
Pessoa S B, Barros P R. Notas sobre a epidemiologia da
esquistossomose mansonica no Estado d o Sergipe. Rev
Med Cirurg Sao Paulo 1953; 13: 17-24.
Lima e Costa M F F, Rocha R S, Coura Filho P, Katz N . A 13
year follow-up of treatment and snail control in an endemic
area for Schistosoma
mansoni
in Brazil: incidence of
I 300
31
32
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
infection and reinfection. Bull World Health Organ 1993;
71: 197-205.
Jordan P, Webbe G. Epidemiology. In: Jordan P, Webbe G,
Sturrock RF. Human Schistosomiasis 1993. Wallingford,
UK: CAB International, pp. 87-159.
L i m a e Costa M F F, Rocha R S, Magalhaes M H A, Katz N .
Urn modelo hierarquico d e analise das variaveis socio
economicas e dos padroes de contatos com aguas
associados a forma hepatoesplenica da esquistossomose.
Cad Saude Publica, Rio de Janeiro 1994; 10 (Suppl. 2):
241-53.
(Revised version received May 1995)