Polyparasitism with Schistosoma haematobium and soil-transmitted helminth infections among school children in Loum, Cameroon.

Abstract

Prevalence and abundance of Schistosoma haematobium and soil-transmitted helminths (STH) were assessed among a total of 1600 pupils (urine, n = 1190; faecal samples, n = 1454) attending five schools in Loum, Littoral Province, Cameroon, with the specific aim of assessing the extent of polyparasitism and the extent to which infections were focused in particular subsets of the study group. Prevalence of S. haematobium was 62.8% with an abundance (arithmetic mean of egg counts) of 54 eggs/10 ml urine. For the STH these were 47.7% and 619 eggs per gram of faeces (EPG) for Trichuris trichiura, 65.5% and 3636 EPG for Ascaris lumbricoides, and 1.4% and <0.1 EPG for hookworms. Most children (90.3%) were infected with at least one of these four species, the largest proportion (34.3%) carrying two species; 27.4% carried three and 1.1% carried concurrently all four species of parasites. The average number of species harboured increased with age, as did the prevalences of S. haematobium and T. trichiura but not that of A. lumbricoides. All STH showed marked differences in prevalence between the five schools but only T. trichiura varied significantly between sexes. Mean abundance of infection varied significantly between age classes, among schools and between the sexes, with females showing heavier mean EPGs for A. lumbricoides and T. trichiura and males higher mean eggs/10 ml urine for S. haematobium infections. A highly significant association was detected between A. lumbricoides and T. trichiura, that was not context-dependent. This was confirmed in quantitative analyses after controlling for differences in abundance between schools, sexes and age classes. A weaker context-dependent association (prevalence data) was detected between S. haematobium and A. lumbricoides (sex- and age-dependent) but quantitative associations between these two species, as well as between S. haematobium and T. trichiura, were not convincing.

Full text

Page 1

Polyparasitism with Schistosoma haematobium and
soil-transmitted helminth infections among school children
in Loum, Cameroon
L.-A. Tchuem Tchuente ´1, J. M. Behnke2, F. S. Gilbert2, V. R. Southgate3and J. Vercruysse4
1 Laboratory of General Biology, Faculty of Sciences, University of Yaounde´ , Yaounde´ , Cameroon
2 School of Biology, University of Nottingham, Nottingham, UK
3 Department of Zoology, The Natural History Museum, London, UK
4 Department of Virology, Parasitology and Immunology, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium
Summary
Prevalence and abundance of Schistosoma haematobium and soil-transmitted helminths (STH) were
assessed among a total of 1600 pupils (urine, n ¼ 1190; faecal samples, n ¼ 1454) attending five schools
in Loum, Littoral Province, Cameroon, with the specific aim of assessing the extent of polyparasitism
and the extent to which infections were focused in particular subsets of the study group. Prevalence of
S. haematobium was 62.8% with an abundance (arithmetic mean of egg counts) of 54 eggs/10 ml urine.
For the STH these were 47.7% and 619 eggs per gram of faeces (EPG) for Trichuris trichiura, 65.5%
and 3636 EPG for Ascaris lumbricoides, and 1.4% and <0.1 EPG for hookworms. Most children
(90.3%) were infected with at least one of these four species, the largest proportion (34.3%) carrying
two species; 27.4% carried three and 1.1% carried concurrently all four species of parasites. The average
number of species harboured increased with age, as did the prevalences of S. haematobium and
T. trichiura but not that of A. lumbricoides. All STH showed marked differences in prevalence between
the five schools but only T. trichiura varied significantly between sexes. Mean abundance of infection
varied significantly between age classes, among schools and between the sexes, with females showing
heavier mean EPGs for A. lumbricoides and T. trichiura and males higher mean eggs/10 ml urine for
S. haematobium infections. A highly significant association was detected between A. lumbricoides and
T. trichiura, that was not context-dependent. This was confirmed in quantitative analyses after
controlling for differences in abundance between schools, sexes and age classes. A weaker context-
dependent association (prevalence data) was detected between S. haematobium and A. lumbricoides
(sex- and age-dependent) but quantitative associations between these two species, as well as between
S. haematobium and T. trichiura, were not convincing.
keywords schistosomiasis, Schistosoma haematobium, soil-transmitted helminthiasis, epidemiology,
Cameroon
Introduction
Human communities in the tropical regions of the World,
where soil-transmitted helminth (STH) infections are
endemic, are seldom afflicted by just one species. It is more
usual for several STH to be present in the community
concurrently and for the greatest number of subjects to
carry simultaneously at least two species (Buck et al. 1978;
Keusch & Migasena 1982; Kvalsvig 1988; Ashford et al.
1992; Chunge et al. 1995; Booth et al. 1998a,b), especially
in developing countries where overall rates of parasitism
are high. More disturbingly, recent studies have shown that
for some parasites, egg counts are likely to be higher
among subjects carrying mixed infections than in subjects
carrying single species infections (Booth et al. 1998a;
Needham et al. 1998). This subset of the community is
therefore at high risk of morbidity and, in extreme
circumstances, even mortality. Moreover, Howard et al.
(2001) reviewing the literature on parasitic associations,
observed a significant geographical heterogeneity between
associations although associations within smaller units
such as provinces, districts and ecological zones were
largely homogenous (Howard et al. 2002). They also
recorded increasing strength of associations with increasing
intensity of infection.
In Cameroon, schistosomiasis and STH infections are
important parasitic diseases. Recent estimates indicate that
more than 1.7 million people (of a total of 14 million
Tropical Medicine and International Health
volume 8 no 11 pp 975–986 november 2003
ª 2003 Blackwell Publishing Ltd
975

Page 2

inhabitants) in the country are infected with either
Schistosoma haematobium or S. mansoni, 5.6 million with
Ascaris lumbricoides, 6.5 million with Trichuris trichiura
and 2.6 million with hookworms (Brooker et al. 2000a).
However, infections are unevenly distributed within the
country. The highest transmission levels of schistosomiasis
occur in the northern region, whereas STH infection are
more prevalent in the southern part of the country (Ratard
et al. 1990, 1991; Brooker et al. 2000a; Howard et al.
2002). Although schistosomiasis is primarily predominant
in the north of Cameroon, isolated foci of high prevalence
exist in the south equatorial zone of the country, with
villages endemic for S. haematobium and S. intercalatum
located in both Littoral and Southwest Provinces. One of
the most intense transmission foci is in the town of Loum,
which also represents an interesting and unique example
reflecting the dynamic epidemiological evolution that may
occur in schistosomiasis transmission foci over time.
Indeed, before 1968, only S. intercalatum occurred in
Loum, and subsequently this species was completely
replaced by S. haematobium through introgressive
hybridization (Tchuem Tchuente ´ et al. 1997). However,
STH infections were not assessed in Loum in earlier
surveys.
As part of our study programme on the epidemiology of
schistosomiasis in different ecological zones in Cameroon,
we conducted a detailed survey of infection with
S. haematobium in school children, the subset of the
population at highest risk of infection, attending five
schools in Loum. We also quantified STH infections, as
polyparasitism is known to exist in the region (Howard
et al. 2002). A key objective of this study was to evaluate
the extent to which different helminths co-occur in the
community and particularly in certain subsets of the study
group. These data will provide an essential background for
focussed treatment and control of the parasites among
children in Loum and, in the broad context, a useful
resource for meta-analyses of geographical variation in
associations between helminths in humans.
Materials and methods
Study site
The town of Loum lies at 4 ?43¢N, 9 ?44¢E, 100 km inland
from Douala, at an altitude of about 180 m. To the north
of the town Mount Koupe ´ rises to a height of 2070 m and
the source of the Mbette, a small river which passes
through Loum, is located on the slopes of this mountain
but is reinforced by tributaries from several springs in the
town itself. Loum is situated in the equatorial zone,
Cameroon-type climate, one of the wettest in the world,
with a rainy season, often heavy, of about 9 months
(June–September) followed by a brief dry period.
The town of Loum is subdivided into a number of quarters
(Figure 1) and the main transmission sites of schistosomi-
asis have been identified in those quarters adjacent to the
river Mbette and its tributaries.
Study design
In May 2000, schoolchildren from five schools in Loum
were invited to participate in the study, and were registered
only after the objectives of the study had been explained to
them and to their parents or guardian, and full informed
consent obtained. To avoid any bias, all the available
children attending the schools were examined, almost all
were permanent residents and no mass anthelmintic
treatment had been conducted in the town before this
study. Samples were obtained from 1600 pupils of whom
1454 provided a faecal sample, 1190 a urine sample and
1044 both faecal and urine samples. The composition
of the study population is summarized in Table 1 by
school and sex. There were significant differences in pupil
numbers between the schools as well as among age groups
within schools (the school · sex · age interaction in legend
to Figures 3 and 4), but the sexes were well balanced
within schools.
Collection of stool samples and parasitological
examination
Stool and urine samples were collected from each
participant and transported to a local field laboratory
where all urine and stool samples were examined within
48 h. No preservatives were used. Urine samples were
collected in 50 ml plastic screw-cap vials, between 10.00
and 14.00 hours. Each urine sample was agitated to
ensure adequate dispersal of eggs, 10 ml of urine were
filtered through a Nuclepore?filter, and the filters were
examined by microscopy for the presence of eggs.
Stool samples were collected in similar screw-cap vials,
and examined by a single thick smear technique using a
41.7 mg Kato-Katz template. However, it should be
noted that because of logistic constraints it was not
always possible to examine these within 30 min and
therefore hookworm prevalence and abundance is likely
to be underestimated.
Statistical analysis
The frequency distribution of species richness was tested
for goodness-of-fit to the null model of Janovy et al. (1995)
by chi-square method.
Tropical Medicine and International Health volume 8 no 11 pp 975–986 november 2003
L.-A. Tchuem Tchuente´ et al. Polyparasitism in Cameroon
976
ª 2003 Blackwell Publishing Ltd

Page 3

Confidence limits for prevalence (percentage of subjects
infected) were calculated as described by Rohlf and
Sokal (1995) and prevalence was analysed by maximum
likelihood techniques based on log-linear analysis of
contingency tables, implemented by the software package
Statgraphics Version 7.0. For each species prevalence of
infection (binary factor, infected ¼ 1, not infected ¼ 0),
school (five levels, each corresponding to one school), host
age (12 levels, for analysis subjects were arranged into 12
classes corresponding to age class 1 ¼ 5 years and less, age
class 2 ¼ 6 years etc. until age class 12 ¼ 16 and older)
and sex (two levels) were entered as factors. Beginning with
the most complex model, involving all possible main
effects and interactions, those combinations not contribu-
ting significantly to explaining variation in the data were
eliminated stepwise, beginning with the highest-level
interaction. A minimum sufficient model was then
obtained, for which the likelihood ratio of chi-square was
not significant, indicating that the model was sufficient in
explaining the data.
Quantitative data reflecting parasite abundance within
hosts are shown as arithmetic mean values, although
geometric mean (GM) values are also provided for
comparison because the data were highly overdispersed
Table 1 Composition of the study
population by school and sex of children
in Loum, Cameroon
School
number School name
Stool examination Urine examination
MalesFemales
Total by
schoolMalesFemales
Total by
school
1
2
3
4
5
CBC Gare
CEBEC Loum B
Ec. Pub. Bonkeng
St JM Vianney
St Thomas Bonkeng
30
71
26
71
170
420
29
5628
73
159
313
26
22
78
162
297
32
50
151
321
610
58
142
342
864
50
172
444
21
Total738 7161454 599591 1190
Figure 1 Map of the town of Loum in the Cameroon, showing the location of the five schools that participated in the study. 1, CBC Gare;
2, CEBCE Loum; 3, Ec. Pub. Bonkeng; 4, St J.M. Vianney; 5, St Thomas Bonkeng (¤ school; S, spring; the river is marked in bold).
Tropical Medicine and International Healthvolume 8 no 11 pp 975–986 november 2003
L.-A. Tchuem Tchuente´ et al. Polyparasitism in Cameroon
ª 2003 Blackwell Publishing Ltd
977

Page 4

(Elliott 1977; Dash et al. 1988) and 95% confidence
limits (95% CL) are supplied. However, it should be
noted that while GM provide a better estimate of the
worm burdens of ‘average’ children, the statistical
analysis fits arithmetic mean values and in some cases
GM can be misleading indicators of the direction of
differences between groups (Fulford 1994). Both mean
values reflect the abundance of infection as defined by
Margolis et al. (1982) and include all subjects within the
specified group, infected and not infected, for which
relevant data were available. The degree of aggregation
in quantitative data was calculated by the index of
dispersion (I ¼ the variance to mean ratio) and the index
of discrepancy (D) as described by Poulin (1993) (a value
of 0 indicates an even distribution of counts across all
hosts and a value of 1 indicates all parasites aggregated
in a single host). Frequency distributions of individual
species were also tested for goodness-of-fit to negative
binomial, positive binomial and Poisson models by chi-
square as described by Elliott (1977) and the negative
binomial exponent k is given as appropriate.
Parasite abundance was analysed by GLIM (a statis-
tical system for generalized linear interactive modelling;
GLIM 4, PC version, Royal Statistical Society 1993)
using models with negative binomial errors (Crawley
1993; Wilson & Grenfell 1997). School, and host sex
(see above for levels) were entered as factors. Age was
mostly entered as a factor, although models with age as
a covariate were also explored and those employed were
school · sex + age class. In some cases, where the
iterations in GLIM would not converge satisfactorily,
the models were simplified by pooling combinations of
age classes as described later. The residuals from all
models were checked for negative binomial distribution
and were saved for analysis of interactions between
species.
Quantitative associations between parasites were exam-
ined by correlation analysis (Spearman rank-order corre-
lation test) of raw parasite data from subjects carrying both
species and residuals from minimum sufficient anovas in
order to control for established differences between the
schools, age classes and sex.
Results
Helminth species richness
Four species of helminths were recorded in the survey. The
frequency distribution of species richness across this
subsample of the population was compared with the
distribution predicted by the null model for interactions of
parasite species in an assemblage (Figure 2a; Janovy et al.
1995) and differed significantly. Figure 2a shows that more
than expected subjects carried three and four worm
infections or were not infected.
0
5
10
15
20
25
30
35
40
45
50
01234
No. of species
Percentage of population ±
95% CL
Observed
Expected
1.6
1.7
1.8
1.9
2
2.1
2.2
2.3
2.4
2.5
2.6
12345
School
0.8
1
1.2
1.4
1.6
1.8
2
2.2
<66789 10 11
Age class
21 31415 >15
No. of species ± SEM
No. of species ± SEM
(c)
(b)
(a)
Figure 2 Analysis of species richness. (a) Frequency distribution
of observed and expected (null model of Janovy et al. 1995)
species richness (v2
that provided both urine and stools. (b) Variation in species
richness between schools (GLIM model ¼ school · sex + age
with Poisson errors, main effect of school, v2
P > 0.01). Schools 1–5 comprised 43, 137, 316, 506 and 42 indi-
viduals, respectively. (c) age-dependent variation in species richness
(GLIM model ¼ school · sex + age with Poisson errors, main
effect of age v2
from subjects £5 to ‡16 ages as shown and comprised 13, 73, 65,
113, 98, 131, 168, 143, 94, 71, 40 and 35 individuals, respectively.
4¼ 82.7, P < 0.001) based on 1044 subjects
4¼ 12.3, 0.025 >
1¼ 5.8 (0.025 > P > 0.01). The age classes ranged
Tropical Medicine and International Health volume 8 no 11 pp 975–986 november 2003
L.-A. Tchuem Tchuente´ et al. Polyparasitism in Cameroon
978
ª 2003 Blackwell Publishing Ltd

Page 5

Mean species richness
The overall mean number of species of helminths har-
boured per host was 1.8 ± 0.03, with a variance to mean
ratio of 0.52. There was a significant difference between
the schools (Figure 2b) with children attending schools 1
and 5 (CBC Gare and St Thomas Bonkeng) carrying
significantly more species of helminths than those attending
the other schools. There were no other significant effects
when age was entered as a factor. However, models with
age as a covariate (Figure 2c) indicated a significant
positive increase in the number of species with increasing
age (gradient 0.021 ± 0.009).
Prevalence of infection
The overall prevalence of S. haematobium was
62.9% ± 2.7 (n ¼ 1190). As Figure 3a shows there was a
highly significant effect of age (the age · infection inter-
action), prevalence increasing from 42.9% among the
youngest children (<5) and rising to peak (81.3 ± 11.04%)
among the 15-year-old class. There was also a significant
difference between the schools. The lowest prevalence was
at St JM Vianney G1, the school with the most subjects
(51.3% of study group) and the highest at CBC Gare, the
school with the fewest subjects (4.2% of study group).
Overall there was a significant effect of sample size on
prevalence (rs¼ )0.9, n ¼ 5, P ¼ 0.037). There was
no difference between the sexes (prevalence in males ¼
62.1% ± 3.9; females ¼ 63.6% ± 3.9).
Only 21 subjects carried the hookworm Necator amer-
icanus (1.4%) and so no statistical analysis was possible. In
both A. lumbricoides and T. trichiura there were significant
differences between the schools (Figures 4 and 5), but only
T. trichiura showed varying prevalence between the sexes
(males ¼ 68.6% ± 3.4, females ¼ 62.4% ± 3.5) and age
classes (Figure 5a).
Abundance of infection
As with prevalence there were too few cases of N. amer-
icanus infection to analyse statistically. All three remaining
helminths showed overdispersed distributions that con-
formed to the negative binomial model (Table 2).
The overall mean for S. haematobium egg counts was
54.3 ± 4.14 (GM ¼ 7.1, CL 6.23–8.11) with a range from
0 to 1628 eggs/10 ml. Abundance of infection was highly
dependent on host age (Figure 3a), rising irregularly from
the youngest age classes and peaking among the 15-year
age class. Abundance also differed between the sexes
(males ¼ 63.8 ± 7.0, females ¼ 44.6 ± 4.3) and schools
but interpretation was confounded by a highly significant
interaction between school · sex (Figure 3b). The latter
arose because the mean egg counts were higher among
boys than girls in schools 3, 4 and 5 (accounting for 83.1%
of the study group), similar in school 2 and marginally
lower in school 1.
The egg counts for A. lumbricoides varied from 0 to
101 376 EPG, with a mean of 3636.2 ± 253.5 EPG
(GM ¼ 37.2, CL ¼ 30.06–46.06). Attempts to analyse the
abundance of A. lumbricoides were complicated because
GLIM would not converge in models based on 12 age
classes. However, a model with four age classes (original
12 classes were merged into 3-year ranges) resolved
satisfactorily, and indicated that only the main effects were
0
20
40
60
80
100
120
140
160
180
200
<6 6789 10 11 12 13 14 15 >15
Age class
Mean EPG ± SEM
Mean EPG ± SEM
0
20
40
60
80
100
120
140
160
12345
School
0
10
20
30
40
50
60
70
80
90
100
Prevalence (%) ± 95% CL
10
20
30
40
50
60
70
80
90
100
Prevalence (%) ± 95% CL
Abundance − males
Abundance − females
Prevalence
Abundance
Prevalence
(a)
(b)
Figure 3 PrevalenceandabundanceofSchistosomahaematobium.
(a) Variation in relation to age. For prevalence, age · infection,
v2
v2
to ‡16 ages as shown and comprised 14, 83, 71, 132, 116, 154,
184, 160, 109, 82, 48 and 37 individuals, respectively. (b) Varia-
tion between schools and sex. For prevalence, school · infection,
v2
v2
P < 0.0005; interaction between school and sex, v2
P < 0.0005. Additional significant terms in the minimum sufficient
model for prevalence included a three-way interaction, school
· sex · age, v2
the minimum sufficient model gave v2
for full explanation and Table 1 for n.
11¼ 44.2, P < 0.0001; For abundance, main effect of age,
11¼ 71.07, P < 0.0005. The age classes ranged from subjects £5
4¼ 21.74, P ¼ 0.0002; for abundance, main effect of school,
4¼ 27.43, P < 0.0005; main effect of sex, v2
1¼ 22.16,
4¼ 34.69,
44¼ 67.5, P ¼ 0.013. Test of the goodness-of-fit of
104¼ 115.3, P ¼ 0.2. See text
Tropical Medicine and International Healthvolume 8 no 11 pp 975–986 november 2003
L.-A. Tchuem Tchuente´ et al. Polyparasitism in Cameroon
ª 2003 Blackwell Publishing Ltd
979