Am. J. Trop. Med. Hyg., 62(2), 2000, pp.

225231
Copyright 2000 by The American Society of Tropical Medicine and Hygiene

RANDOM DISTRIBUTION OF MIXED SPECIES MALARIA INFECTIONS IN PAPUA

NEW GUINEA
R. K. MEHLOTRA, K. LORRY, W. KASTENS, S. M. MILLER, M. P. ALPERS, M. BOCKARIE, J. W. KAZURA,
P. A. ZIMMERMAN

AND

Division of Geographic Medicine, Case Western Reserve University School of Medicine, Cleveland, Ohio;
Papua New Guinea Institute of Medical Research, Madang, Papua New Guinea
Abstract. Plasmodium falciparum (Pf), P. vivax (Pv), P. malariae (Pm), and P. ovale (Po) infections are endemic
in coastal areas of Papua New Guinea. Here 2,162 individuals living near Dreikikir, East Sepik Province, have been
analyzed for complexity of malaria infection by blood smear and polymerase chain reaction (PCR) diagnoses. According to blood smear, the overall prevalence of Plasmodium infection was 0.320. Most individuals (0.283) were
infected with a single species only. The prevalence of mixed species infections was low (0.037). Further analysis of
a 173-sample subset by nested PCR of small subunit ribosomal DNA resulted in an overall 3.0-fold increase in
prevalence of infection, with a 17.5-fold increase in the frequency of mixed species infections. Among mixed species
infections detected by PCR, the frequency of double species was 0.364, and that of triple species was 0.237. Nine
individuals (0.052) were infected with all 4 species. To determine if infection status (uninfected, single, and multiple
infections) deviates from an independent random distribution (null hypothesis), observed versus expected frequencies
of all combinations of Plasmodium species infections, or assemblages (Pf-, Pv-, Pm-, Po-, to Pf, Pv, Pm, Po),
were compared using a multiple-kind lottery model. All 4 species were randomly distributed whether diagnosed by
blood smear or PCR in the overall population and when divided into age group categories. These findings suggest
that mixed species malaria infections are common, and that Plasmodium species appear to establish infection independent of one another.
Four species of Plasmodium cause human malaria: P. falciparum (Pf), P. vivax (Pv), P. malariae (Pm), and P. ovale
(Po). They differ greatly with respect to their biology and
clinical manifestations.1 Sympatric combinations of these
species occur in human populations and within infected individuals. Knowledge of their prevalence and transmission
dynamics within a given geographic region is key to the
design of effective control measures. An understanding of
their interactions will be necessary to interpret thoroughly
the outcome of vaccine trials. Previous studies have observed complex relationships among Plasmodium species in
patients treated for syphilis,24 and in surveys of naturally
infected individuals living in malaria-endemic regions
around the world.510 Experimental infections involving nonendemic adults have shown that susceptibility to individual
Plasmodium species may be influenced by the species of
previous or current infections.24 Natural infection relationships among human malaria species appear to vary based
upon geographic differences, which are certain to involve
not only bio-climatic variation but also genetic differences
in human and vector populations.510
Microscopic examination of Giemsa-stained thick and thin
blood smears has been the diagnostic method of choice for
species identification in epidemiologic studies and medical
diagnosis.11 The major limitations of microscopy include
lack of trained personnel and the length of time required for
blood smear examination. Consequently, species identification becomes ambiguous when parasitemia is low and in
cases of mixed species infections, leading to incorrect or
failed species identification.7,9 Several efficient molecularlevel detection methods have been developed to overcome
these limitations, and subsequently validated using field
samples. These include antigen-12,13 or nucleic acid-based detection.1420 With the advent of these methods, significant increases in the reported prevalence of mixed species infections have been observed in various geographic regions.16,2028

Malaria is endemic at altitudes below 1,5001,600 meters

in Papua New Guinea and all 4 malaria parasite species are
transmitted.2933 In the recent study by Genton and others
from the Wosera area of East Sepik Province, the overall
prevalence of Plasmodium species infection was 0.600, the
prevalence of mixed species infections was 0.213, and there
appeared to be no clear-cut seasonal pattern of infection.33
To investigate further the possible interactions between
malaria parasite species, 2,162 Giemsa-stained blood smears
from persons living in 11 villages near Dreikikir in East
Sepik Province have been analyzed. Additionally, a nested
polymerase chain reaction (PCR)based diagnostic protocol
targeting the small subunit ribosomal RNA (ssu rRNA)
genes16 has been adapted to analyze a subset of 173 samples
to compare with blood smearbased diagnosis of mixed species infections. To determine if the prevalence of Plasmodium species infections are independent of species-to-species
effects within the study population, a multiple-kind lottery
model has been used.34 This same model has been used to
illustrate the independent random distribution of various parasite species combinations/assemblages in snails and a variety of vertebrate hosts.34,35 The multiple-kind lottery model
is similar to algorithms developed by Cohen,36 since it relies
upon determining the prevalence of both infected and not
infected individuals to generate expected values for the various parasite species assemblages. All calculations included
in this analysis are easily performed using spreadsheet programs that are widely available for use on personal computers.
MATERIALS AND METHODS

Study population and blood sample collection. The

study was conducted in Dreikikir District, located in the
southern foothills of the Toricelli Mountains in East Sepik

225

226

MEHLOTRA AND OTHERS

Province, Papua New Guinea. Natural vegetation is rain forest. Residents are predominantly subsistence farmers and
have cleared away many areas to make gardens. Samples
were collected from July to September, 1996, which corresponds to the relatively dry season (July to November, rainfall 50150 mm/month; December to June, rainfall 80
290 mm/month).37 The highest human exposure rates to sporozoite-infected mosquitoes occur during this time period37
(Bockarie M, unpublished data). Blood samples from all villagers 5 years old were collected as part of a filariasis
control project.38 More than 85% of eligible subjects agreed
to participate. Ethical approval for this study and the procedure for oral informed consent were obtained from the
Medical Research and Advisory Council of Papua New
Guinea and the Case Western Reserve University/University
Hospitals of Cleveland Human Investigation Committee.
Blood smear examination. Blood films were prepared
and examined as described previously.33 Briefly, thick and
thin smears were stained with 4% Giemsa and examined
under oil-immersion (100) for 100 fields. Parasite species
were identified using both thick and thin film preparations.
Preparation of DNA template. The DNA was extracted
from whole blood (200 l) from study subjects or infected
chimpanzees using a QIAamp 96 or individual spin blood
kits (QIAGEN, Valencia, CA) according to the manufacturers protocol. Plasmodium falciparum (FCB) genomic DNA
was kindly provided by Dr. Xin-zhuan Su (Laboratory of
Malaria Genetics, National Institutes of Health, Bethesda,
MD). Blood samples from chimpanzees infected with P. vivax (type II), P. malariae, or P. ovale were provided by Dr.
W. E. Collins (Centers for Disease Control and Prevention,
Atlanta, GA).
Polymerase chain reaction-based diagnosis. Amplification of ssu rDNA was performed following nested PCR strategies, using genus-specific (nest 1) and species-specific (nest
2) primers (Research Genetics, Huntsville, AL) identical to
those described.16 All reactions (25 l) were performed in
buffer containing 3 pmoles of appropriate upstream and
downstream primers, 67 mM Tris-HCl, pH 8.8, 6.7 mM
MgSO4, 16.6 mM (NH4)2SO4, 10 mM 2-mercaptoethanol,
100 M dATP, dGTP, dCTP, and dTTP, and 2.5 units of
thermostable DNA polymerase. Nest 1 amplification conditions were 95C for 5 min (1); 95C for 1 min, 63C for 2
min, 72C for 2 min (5); 95C for 1 min, 58C for 2 min,
72C for 2 min (35); 95C for 1 min, 55C for 2 min, 72C
for 2 min (5); and 95C for 1 min, 58C for 2 min, 72C
for 2 min (1). A 3-l aliquot of the nest 1 reaction was
then used as a template in nest 2 species-specific reaction.
For control reactions, nest 1 products, amplified from cloned,
species-specific ssu rDNAs, were diluted 1:1,000 or 1:
10,000 in sterile distilled water prior to nest 2 amplification.
Nest 2 amplification conditions were 95C for 5 min (1);
95C for 1 min, 63C for 2 min, 72C for 2 min (5); 95C
for 1 min, 58C for 2 min, 72C for 2 min (30); 95C for
1 min, 55C for 2 min, 72C for 2 min (5); and 72C for
5 min (1). The PCR assays were performed using a Peltier
Thermal Cycler, PTC-225 (MJ Research, Watertown, MA).
The PCR products from nest 1 and nest 2 amplifications
were loaded on 2% Agarose I (Amresco, Solon, OH) gels,
and electrophoresis was performed in 1 TBE buffer (8.9
mM Tris, 8.9 mM boric acid, 2.0 mM EDTA). The gels were

stained for 30 min with SYBR Gold (Molecular Probes, Eugene, OR), diluted 1:10,000 in 1 TBE buffer, and DNA
products were visualized on a Storm 860 using ImageQuant
software (instrumentation and software developed by Molecular Dynamics, Sunnyvale, CA).
To assess the possibility of cross-contamination, a series
of study samples representing the full spectrum of infection
(no infection to all 4 species infection) were subjected to
PCR with negative controls spaced between each sample. No
evidence of PCR amplification of species-specific amplicons
was observed in any of the negative controls.
Polymerase chain reaction-based cloning and DNA sequence analysis of ssu rDNA. Nest 1 amplicons (approximately 1,200 basepairs) from each of the parasites described
above were cloned into the pCR2.1-TOPO plasmid vector
using a TOPO TA cloning kit (Invitrogen, Carlsbad, CA)
according to the manufacturers protocol. The clones were
amplified using extended M13 forward (5-GTT TTC CCA
GTC ACG ACG TTG TAA AAC GAC GGC CAG-3) and
M13 reverse (5-TGA GCG GAT AAC AAT TTC ACA
CAG GAA ACA GCT ATG AC-3) primers. Temperature
cycling conditions were 95C for 30 sec, 45C for 30 sec,
72C for 2 min (5); and 95C for 30 sec, 65C for 30 sec,
72C for 2 min (30). Genus-specific nest 1 amplicons were
purified using a QIAquick PCR purification kit and recommended protocol (QIAGEN). Sequencing of DNA was performed using fluorescence-based sequencing protocols on an
ABI377 automated sequencer (Applied Biosystems, Foster
City, CA). Sequences were analyzed using Sequencher software (Version 3.0.1 Demo; Gene Codes Corporation, Ann
Arbor, MI). Plasmodium falciparum, P. vivax, P. malariae,
and P. ovale sequences were assigned GenBank accession
numbers AF145334, AF145335, AF145336, and AF145337,
respectively. Bacterial clones for plasmids containing each
species-specific ssu rRNA nest 1 amplicon are available
through the Malaria Research and Reference Reagent Resource Center (MR4) at: http://www.malaria.mr4.org.
Statistical analysis. Plasmodium species prevalence was
determined by enumerating infected individuals and dividing
by the total population (2,162) or population subset (173)
size. The multiple-kind lottery model34 was used to calculate
the expected distribution of parasite species assemblages in
the population. Chi-square values were calculated using heterogeneity tests (number of rows number of columns) to
compare observed versus expected values.
RESULTS

Blood smear-based diagnosis of mixed infections. Of

2,162 subjects, 436 individuals (0.202) were infected by Pf,
237 (0.110) by Pv, and 100 (0.046) by Pm. No Po infections
were detected. Table 1 shows that 692 individuals (0.320)
were infected by a single or mixed Plasmodium species,
whereas 1,470 (0.680) were not infected. Furthermore, 612
individuals (0.283) were infected by a single species only,
and 80 individuals (0.037) were infected with a mixture of
species. Three hundred sixty-one individuals (0.167) were
infected with Pf alone, 176 (0.081) with Pv alone, and 75
(0.035) with Pm alone. The 80 mixed species infections included 79 individuals (0.036) infected with 2 species. The

DIAGNOSIS OF MIXED MALARIA INFECTIONS IN PAPUA NEW GUINEA

227

TABLE 1
Summary of not infected and all single, double, triple, and quadruple
infections detected by blood smear diagnosis*
Parasite assemblage

Pf
Pv
Pm
Po
Pf Pv
Pf Pm
Pf Po
Pv Pm
Pv Po
Pm Po
Pf Pv Pm
Pf Pv Po
Pf Pm Po
Pv Pm Po
Pf Pv Pm Po
Not infected

Observed

361
176
75
0
55
19
0
5
0
0
1
0
0
0
0
1,470

Expected

Chi-square

370.25
180.45
71.08
0.00
45.58
17.96
0.00
8.75
0.00
0.00
2.21
0.00
0.00
0.00
0.00
1,465.71
Chi-square (df 15)

0.23
0.11
0.22

1.95
0.06

1.61

0.66

0.01
4.85

* Pf Plasmodium falciparum; Pv P. vivax; Pm P. malariae; Po P. ovale; df

degrees of freedom.

number of individuals infected with various combinations of

2 species were as follows: Pf Pv 55 (0.025), Pf Pm
19 (0.009), and Pv Pm 5 (0.002). One individual had
a positive blood smear for Pf Pv Pm.
To evaluate whether the prevalence of infection within
study subjects (i.e., no infection, single species infections,
or mixed infections) deviates from an independent random
distribution pattern (null hypothesis), the multiple-kind lottery model was applied. Table 1 shows that the observed
and expected values were not significantly different (heterogeneity test [16 rows 2 columns, degrees of freedom [df]
15] 4.85, P not significant).
Specificity of the PCR-based diagnostic assay. Cloned
ssu rDNA templates from the 4 Plasmodium species were
used to confirm the specificity of each primer pair and to
serve as positive controls. Nest 1 products cloned from each
individual Plasmodium species were sequenced to confirm
homology with each of the respective species-specific primer
pairs; no cross-species homology was observed. Under optimized amplification conditions, specific products of the expected sizes (Pf 205 basepairs, Pv 120 basepairs Pm
144 basepairs, Po 786 basepairs)16 were obtained only
when the DNA from the corresponding species was present
(Figure 1).
Comparison of blood smear and PCR-based diagnosis
of mixed Plasmodium species infections. From the 2,162
individuals analyzed by blood smear, a subset of 173 was
further studied by PCR-based diagnosis. One hundred fortyone (0.815) were diagnosed to be infected with Pf, 103
(0.595) with Pv, 64 (0.370) with Pm, and 27 (0.156) with
Po. Table 2 shows that 163 individuals (0.942) were infected
by a single or mixed Plasmodium species, and only 10
(0.060) were not infected. In comparison with the blood
smear analysis, the PCR assay detected a higher prevalence
of infection for Pf (3.0-fold), Pv (3.2-fold), and Pm (2.6fold). When blood smear was considered as the gold-standard diagnostic assay, the sensitivity and specificity of the
PCR-based assay were calculated as follows: for Pf 0.94
(sensitivity) and 0.23 (specificity), for Pv 0.72 and 0.43, and

FIGURE 1. Specificity of the polymerase chain reaction (PCR)

assay using cloned, species-specific, small subunit ribosomal DNA
templates. Nest 1 amplicons cloned into pCR2.1-TOPO from Plasmodium falciparum (Pf), P. vivax (Pv), P. malariae (Pm), and P.
ovale (Po) were used as PCR templates as described in the Materials
and Methods. Each panel (top to bottom) contains nest 2 products
amplified using species-specific primers indicated on the right (Pf,
Pv, Pm, and Po). Organization of samples within each panel (left to
right) includes DNA size marker (100-basepair [bp] ladder), nest 2
products from 2 different plasmid-based clones representing each of
the four human malaria parasite species, and a negative control to
which no DNA was added. Following electrophoresis on a 2% agarose (1 TBE buffer) gel, products were stained using SYBR Gold
and visualized using the Storm 860 fluorescence scanner. The size
of each species-specific product is indicated to the left of each panel
(Pf 205 bp, Pv 120 bp, Pm 144 bp, Po 786 bp).16

for Pm 0.76 and 0.70. Since no individuals were diagnosed

by blood smear to be infected by Po, sensitivity and specificity calculations were not performed.
Further analysis (Table 2) showed that 50 individuals
(0.289) were infected by a single species only, whereas 113
individuals (0.653) were infected with a mixture of parasite
species. Thirty-two individuals (0.185) were infected with Pf
alone, 15 (0.087) with Pv alone, and 3 (0.017) with Pm
alone. No individuals were found to be infected by Po alone.
The 113 mixed species infections included 63 individuals
(0.364) with 2 species infections. The number of individuals
infected with various combinations of 2 species were Pf
Pv 40 (0.231), Pf Pm 16 (0.092), Pf Po 4
(0.023), and Pv Po 3 (0.017). No individuals were infected by combinations of either Pv Po or Pm Po. Fortyone individuals (0.237) were infected with 3 Plasmodium

228

MEHLOTRA AND OTHERS

TABLE 2
Summary of not infected and all single, double, triple, and quadruple
infections detected by polymerase chain reaction diagnosis*
Parasite assemblage

Pf
Pv
Pm
Po
Pf Pv
Pf Pm
Pf Po
Pv Pm
Pv Po
Pm Po
Pf Pv Pm
Pf Pv Po
Pf Pm Po
Pv Pm Po
Pf Pv Pm Po
Not infected

Observed

Expected

32
30.34
15
10.13
3
4.04
0
1.27
40
44.64
16
17.81
4
5.61
3
5.95
0
1.87
0
0.75
27
26.21
8
8.25
5
3.29
1
1.10
9
4.85
10
6.88
Chi-square (df 15)

Chi-square

0.09
2.34
0.27
1.27
0.48
0.18
0.46
1.46
1.87
0.75
0.02
0.01
0.88
0.01
3.56
1.41
15.08

* For definitions of abbreviations, see Table 1.

species as follows: Pf Pv Pm 27 (0.156), Pf Pv

Po 8 (0.046), Pf Pm Po 5 (0.029), and Pv
Pm Po 1 (0.006). Finally, 9 persons (0.052) were infected with all 4 species. Application of the multiple-kind
lottery model showed that observed and expected values

were not significantly different (Table 2, heterogeneity test

[16 rows 2 columns, df 15] 15.08, P not significant).
Given that this may be perceived to be a high prevalence
of infection by the 4 Plasmodium species, Figure 2 is shown
to verify results of PCR diagnoses and provide a comparison
with the blood smear data. Nine individuals were infected
with all 4 species as diagnosed by PCR (lanes 412). One
of these was Pf Pv Pm-infected by blood smear (lane
7). The other 8 were infected singly with either Pf, Pv, or
Pm. Five individuals showed various combinations of 3 species infections (lanes 1317): 2 were infected with Pf Pv
Pm while no infection was observed in their blood smears
(lanes 13 and 15). One was infected with Pv Pm Po
while only Pv infection was seen in the blood smear (lane
14). Two others infected with Pf Pv Pm (lanes 16 and
17) showed Pf Pv (lane 16), and Pf Pm (lane 17) by
blood smear. Of the 2 individuals who were negative by PCR
(lanes 1 and 2), one was Pf-infected and the other was negative by blood smear. One individual (lane 3) was infected
with Pv only by PCR, while no infection was observed in
the blood smear.
From the overall 692 PCR diagnostic assays (4 Plasmodium species 173 study subjects), there were a total of 18
PCR false-negative results. Fourteen individuals were found
to be positive for either a different single or mixed species
infection. The remaining 4 individuals (3 Pf and 1 Pv by

FIGURE 2. Comparison of Giemsa-stained blood smear- and polymerase chain reactionbased diagnosis of mixed Plasmodium species
infections. Each panel (top to bottom) contains nest 2 products amplified using species-specific primers indicated on the right (Pf Plasmodium
falciparum; Pv P. vivax; Pm P. malariae; Po P. ovale). Organization of samples within each panel (left to right) includes DNA size
marker (100-basepair [bp]) bp ladder), a negative control to which no DNA was added, nest 2 products from all 4 species-specific plasmidbased positive controls (Pf, Pv, Pm, and Po), and 17 individual study subjects (lanes 117). Below each panel the Giemsa-stained blood smear
status is indicated for each species infection: infected; not infected.

229

DIAGNOSIS OF MIXED MALARIA INFECTIONS IN PAPUA NEW GUINEA

TABLE 3
Comparison of Plasmodium species infection prevalence diagnosed
by blood smear and small subunit ribosomal RNA DNA-based
polymerase chain reaction (PCR) methods*

TABLE 4
Comparison of observed and expected frequencies of parasite assemblages by age group-blood smear analysis*
510 years (n 37)

Infection status prevalence

Age category

510 years

11 years

Parasite
infection

Blood smear

PCR

Pf
Pv
Pm
Po
Not infected
Pf
Pv
Pm
Po
Not infected

0.432
0.270
0.216
0.000
0.297
0.228
0.162
0.125
0.000
0.551

0.892
0.838
0.297
0.162
0.054
0.794
0.529
0.390
0.154
0.059

* For definitions of abbreviations, see Table 1.

n 37.
n 136.

blood smear) were found to be negative for all 4 species by

PCR. Upon repeat PCR-based analysis, consistent negative
results were obtained for all 4 individuals. In earlier studies,
PCR false negativity has been attributed to a very low copy
number of target sequence, DNA degradation, polymorphism of the target DNA sequence due to mutation, or presence of small amounts of PCR inhibitors.16 To determine if
PCR inhibition was a factor, we checked to determine if
amplification of human target sequences was inhibited in
these 4 samples. The PCR amplification of the human FY
locus39 was not inhibited in any of the samples in question,
The overall false-negative frequency was therefore 0.023.
Differences in Plasmodium infection by age group categories. Further analysis was performed on individuals diagnosed by both blood smear and PCR to determine if the
random distribution of parasite species was altered by age.
For this analysis, study subjects were partitioned into 2 age
group categories, 510 years old (n 37) and 11 years
old (n 136). Table 3 shows that the prevalence of Plasmodium species infection, observed by blood smear analysis,
was decreased by 36% in subjects 11 years old (0.449)
compared with that in subjects 510 years old (0.703). In
contrast, using the PCR assay, no difference in the prevalence of infection was observed (510 years old 0.946;
11 years old 0.941). In the blood smear-based analyses,
a decrease in prevalence of Pf (47%), Pv (40%), and Pm
(42%) infections was observed in subjects 11 years old. In
contrast, PCR-based analyses showed a decrease in prevalence of Pf (11%) and Pv (37%), while Pm increased
(131%), and Po remained unchanged in subjects 11 years
old.
Comparisons of the observed and expected prevalence of
parasite species assemblages in both age groups showed no
significant difference by both blood smear (Table 4) and
PCR (Table 5). The heterogeneity test score for Plasmodium
species infections detected by blood smear was 2.14 in those
510 years old and 4.56 in those 11 years old (respective
P values not significant, df 15) (Table 4). The heterogeneity test score for Plasmodium species infections detected
by PCR was 16.69 in those 510 years old and 11.93 in
those 11 years old (respective P values not significant, df
15) (Table 5). Similar results were obtained when blood

Parasite assemblage

Observed Expected

Chisquare

11 years (n 136)
Observed Expected

Chisquare

Pf
9
9.15 0.00
22
22.74 0.02
Pv
6
4.45 0.54
18
14.86 0.66
Pm
4
3.31 0.14
12
11.00 0.09
Po
0
0.00

0
0.00

Pf Pv
3
3.39 0.04
4
4.39 0.03
Pf Pm
3
2.52 0.09
5
3.25 0.94
Pf Po
0
0.00

0
0.00

Pv Pm
0
1.23 1.23
0
2.12 2.12
Pv Po
0
0.00

0
0.00

Pm Po
0
0.00

0
0.00

Pf Pv Pm
1
0.93 0.00
0
0.63 0.63
Pf Pv Po
0
0.00

0
0.00

Pf Pm Po
0
0.00

0
0.00

Pv Pm Po
0
0.00

0
0.00

Pf Pv Pm Po
0
0.00

0
0.00

Not infected
11
12.01 0.09
75
77.01 0.05
Chi-square (df 15) Chi-square (df 15)
2.14
4.56
* For definitions of abbreviations, see Table 1.

smears for all 2,162 subjects were partitioned into these 2

age group categories and analyzed by the multiple-kind lottery model.
DISCUSSION

The current world status of malaria finds that efforts to

control the disease are largely unsuccessful, especially in
under-developed countries in the tropics. The importance of
this public health problem is underscored by the world-wide
distribution of drug-resistant Plasmodium species parasites
and significant challenges confronting development of effective malaria vaccines. To make progress against this major

TABLE 5
Comparison of observed and expected frequencies of parasite assemblages by age grouppolymerase chain reaction analysis*
510 years (n 37)
Parasite assemblage

Observed Expected

Chisquare

11 years (n 136)
Observed Expected

Chisquare

Pf
4
3.15 0.23
28
26.23 0.12
Pv
2
1.97 0.00
13
7.65 3.74
Pm
0
0.16 0.16
3
4.34 0.41
Po
0
0.07 0.07
0
1.24 1.24
Pf Pv
16
16.28 0.00
24
29.51 1.03
Pf Pm
0
1.33 1.33
16
16.75 0.03
Pf Po
0
0.61 0.61
4
4.79 0.13
Pv Pm
0
0.83 0.83
3
4.88 0.73
Pv Po
0
0.38 0.38
0
1.40 1.40
Pm Po
0
0.03 0.03
0
0.79 0.79
Pf Pv Pm
7
6.89 0.00
20
18.84 0.07
Pf Pv Po
2
3.15 0.42
6
5.39 0.07
Pf Pm Po
0
0.26 0.26
5
3.06 1.23
Pv Pm Po
0
0.16 0.16
1
0.89 0.01
Pf Pv Pm Po
4
1.33 5.34
5
3.44 0.71
Not infected
2
0.38 6.86
8
6.80 0.21
Chi-square (df 15)
Chi-square (df 15)
16.69
11.93
* For definitions of abbreviations, see Table 1.

230

MEHLOTRA AND OTHERS

human pathogen, it is mandatory that the efforts to characterize the magnitude and complexity of Plasmodium species
infections be improved. The importance of characterizing
overall malaria infection accurately has been the focus of a
number of thorough reviews.710
Overall, this study has found that infections in the Dreikikir study population by Pf, Pv, and Pm are approximately
3-fold higher, and that Po is frequently observed when PCRbased diagnosis is compared with conventional blood smear
methods. Furthermore, as a result of the increased sensitivity
of PCR, mixed infections, including those caused by all four
human malaria parasite species, are more common than reported by blood smear diagnosis. Since the observations
from this study are consistent with others based upon molecular diagnostic methodologies, it is important to re-assess
how the sensitivity and specificity of these assays are determined. Recent discussion has suggested that numerous factors related to blood smear diagnosis raise the question as to
whether this methodology should be considered as the true
gold standard.4042 If, in this study, the PCR-based assay is
considered as the gold standard, sensitivity and specificity
of the blood smear-based assay are as follows: 0.31 (sensitivity) and 0.91 (specificity) for Pf, 0.22 and 0.87 for Pv,
and 0.30 and 0.94 for Pm. This evaluation of the two diagnostic assays reflects widely recognized methodologic differences leading to more sensitive detection of malaria parasites by PCR.
Results from this study provide insight beyond the technical advances identified above. Increased sensitivity for detecting infection improves the estimation of the parasite reservoir size and the characteristics of species-to-species interactions within endemic populations. With the PCR-based
prevalence of infection at 0.940 and of mixed infection at
0.653 in this study population, it is likely that all individuals
may be infected by more than one Plasmodium species at
any given time. Furthermore, the prevalence of infection detected by PCR was not observed to decrease with age as was
diagnosed by blood smear. In fact, the PCR prevalence of
P. malariae is observed to increase in the 11-year-old age
group compared with the 510-year-old age group, which is
consistent with a previous report.32 These observations
should be considered in attempts to understand age-acquired
immunity to Plasmodium species parasites more completely.
Since heterologous or cross-species factors have been considered to influence the acquisition of immunity to malaria
parasites,4,8 it is important to monitor species-to-species interactions. In this study, comparisons between observed and
expected prevalence of mixed infection, detected by both
blood smear and PCR, suggest that individual species establish infection independently of the others. Since this finding
was obtained in both age group categories considered, results
suggest that acquisition of immunity does little to influence
the independent distribution of parasites in infected individuals. When these findings were compared with PCR-based
point-prevalence studies specifying the parasites involved in
single and each mixed malaria infection, mixed species infections were randomly distributed in one African-based (Pf-,
Pm-, Po-endemic23) population, two Thai-based (Pf-, Pv-endemic17 and Pf-, Pv-, Pm-, Po-endemic22) populations, and
one South American-based (Pf-, Pv-endemic24) study population. Therefore, findings of an independent distribution of

parasites in infected individuals from this study conducted

in Papua New Guinea do not appear to be unique to this
particular study setting. Two other studies conducted in Malaysia (Pf-, Pv-, Pm-endemic26) and Nigeria (Pf-, Pm-, Poendemic28) have found that observed frequencies of single
species infections are lower than expected, while mixed species infections are higher than expected. Therefore, like the
numerous blood smear-based studies reviewed by Richie9
and McKenzie and Bossert,10 findings from recent PCRbased diagnostic studies provide no consistent trend regarding species-specific facilitation or suppression of infection.
It is important to acknowledge that this study has focused
exclusively on prevalence of infection and not on parasitemia and therefore is unable to assess whether a speciesspecific influence on parasitemia is observed. However,
since blood smear diagnosis is not sufficiently sensitive to
provide a comprehensive assessment of parasites involved in
individual infections, it will be difficult to determine if crossspecies effects influence the intensity of infection. Furthermore, since clinical observations related to malaria were not
made during the study, it is not possible to correlate infection
complexity with malaria pathogenesis. Finally, since pointprevalence surveys are unable to monitor effects of Plasmodium species infection longitudinally, it will be important
to perform follow up studies on this population. From extended studies it may be possible to observe how the ecology
of malaria infection evolves and what factors influence this
evolution.
Acknowledgments: We thank Dr. Georges Snounou, Dr. Wallace Peters, Dr. Christopher Whalen, and David T. McNamara for review
of this manuscript, critical comments, and suggestions; Dr. Godfred
L. Masinde for helpful suggestions in optimizing the PCR assay;
Zachary Dimber for assistance in sample collection; Moses Lagog
and Nandao Torongka for laboratory assistance (Madang); Dr. Hisashi Fujioka for independent confirmation of select mixed species
infection in blood smears; and Cleveland Geonomics for DNA sequence analysis. Furthermore, we thank all of the study volunteers
for their willing participation.
Financial support: This work was supported by the National Institutes of Health (grants AI-142367-01 and AI-36478-04S1) and the
Burroughs-Wellcome Fund (J. W. Kazura).
Authors addresses: R. K. Mehlotra, S. M. Miller, J. W. Kazura, and
P. A. Zimmerman, Division of Geographic Medicine, Case Western
Reserve University School of Medicine, W147D, 2109 Adelbert
Road, Cleveland, OH 44106-4983. K. Lorry, W. Kastens, M. P. Alpers, and M. Bockarie, Papua New Guinea Institute of Medical Research, PO Box 378, Madang, Papua New Guinea.
REFERENCES

1. Garnham PCC, 1966. Malaria Parasites and Other Haemosporidia. Oxford: Blackwell Scientific Publications.
2. Boyd MF, Kitchen SF, 1937. Simultaneous inoculation with
Plasmodium vivax and Plasmodium falciparum. Am J Trop
Med 17: 855861.
3. Mayne B, Young MD, 1938. Antagonism between species of
malaria parasites in induced mixed infections. Public Health
Rep 53: 12891291.
4. Collins WE, Jeffery GM, 1999. A retrospective examination of
sporozoite- and trophozoite-induced infections with Plasmodium falciparum in patients previously infected with heterologous species of Plasmodium: effect on development of parasitologic and clinical immunity. Am J Trop Med Hyg 61
(suppl): 3643.

DIAGNOSIS OF MIXED MALARIA INFECTIONS IN PAPUA NEW GUINEA

5. Molineaux L, Storey J, Cohen JE, Thomas A, 1980. A longitudinal study of human malaria in the West African savanna
in the absence of control measures: relationship between different Plasmodium species, in particular P. falciparum and P.
malariae. Am J Trop Med Hyg 29: 725737.
6. Strickland GT, Zafar-Latif A, Fox E, Khaliq AA, Chowdhry
MA, 1987. Endemic malaria in four villages of the Pakistani
province of Punjab. Trans R Soc Trop Med Hyg 81: 3641.
7. Knowles R, White RS, 1930. Studies in the Parasitology of
Malaria. Calcutta, India: Indian Medical Research Memoirs
No. 18, 436.
8. Cohen JE, 1973. Heterologous immunity in human malaria. Q
Rev Biol 48: 467489.
9. Richie TL, 1988. Interactions between malaria parasites infecting the same vertebrate host. Parasitology 96: 607639.
10. McKenzie FE, Bossert WH, 1997. Mixed-species Plasmodium
infections of humans. J Parasitol 83: 593600.
11. Warhurst DC, Williams JE, 1996. Laboratory diagnosis of malaria. J Clin Pathol 49: 533538.
12. Shiff CJ, Premji Z, Minjas JN, 1993. The rapid manual ParaSight-F test. A new diagnostic tool for Plasmodium falciparum infection. Trans R Soc Trop Med Hyg 87: 646648.
13. Makler MT, Ries JM, Williams JA, Bancroft JE, Piper RC, Gibbins BL, Hinrichs DJ, 1993. Parasite lactate dehydrogenase
as an assay for Plasmodium falciparum drug sensitivity. Am
J Trop Med Hyg 48: 739741.
14. Franzen L, Westin G, Shabo R, Aslund L, Perlmann H, Persson
T, Wigzell H, Petterson U, 1984. Analysis of clinical specimens by hybridization with probe containing repetitive DNA
from Plasmodium falciparum. Lancet i: 525527.
15. Barker Jr RH, Banchongaksorn T, Courval JM, Suwonkerd W,
Rimwungtragoon K, Wirth DF, 1992. A simple method to
detect Plasmodium falciparum directly from blood samples
using the polymerase chain reaction. Am J Trop Med Hyg 46:
416426.
16. Snounou G, Viriyakosol S, Zhu XP, Jarra W, Pinheiro L, do
Rosario VE, Thaithong S, Brown KN, 1993. High sensitivity
of detection of human malaria parasites by the use of nested
polymerase chain reaction. Mol Biochem Parasitol 61: 315
320.
17. Tirasophon W, Rajkulchai P, Ponglikitmongkol M, Wilairat P,
Boonsaeng V, Panyim S, 1994. A highly sensitive, rapid, and
simple polymerase chain reaction-based method to detect human malaria (Plasmodium falciparum and Plasmodium vivax)
in blood samples. Am J Trop Med Hyg 51: 308313.
18. Li J, Wirtz RA, McConkey GA, Sattabongkot J, Waters AP,
Rogers MJ, McCutchan TF, 1995. Plasmodium: genus-conserved primers for species identification and quantitation. Exp
Parasitol 81: 182190.
19. Kimura M, Kaneko O, Liu Q, Zhou M, Kawamoto F, Wataya
Y, Otani S, Yamaguchi Y, Tanabe K, 1997. Identification of
the four species of human malaria parasites by nested PCR
that targets variant sequences in the small subunit rRNA gene.
Parasitol Int 46: 9195.
20. Seesod N, Nopparat P, Hedrum A, Holder A, Thaithong S, Uhlen M, Lundeberg J, 1997. An integrated system using immunomagnetic separation, polymerase chain reaction, and
colorimetric detection for diagnosis of Plasmodium falciparum. Am J Trop Med Hyg 56: 322328.
21. Brown AE, Kain KC, Pipithkul J, Webster HK, 1992. Demonstration by the polymerase chain reaction of mixed Plasmodium falciparum and P. vivax infections undetected by conventional microscopy. Trans R Soc Trop Med Hyg 86: 609
612.
22. Snounou G, Viriyakosol S, Jarra W, Thaithong S, Brown KN,
1993. Identification of the four human malaria parasite species
in field samples by the polymerase chain reaction and detection of a high prevalence of mixed infections. Mol Biochem
Parasitol 58: 283292.
23. Snounou G, Pinheiro L, Goncalves A, Fonseca L, Dias F, Brown

24.
25.

26.

27.
28.

29.
30.
31.

32.
33.

34.

35.
36.
37.

38.

39.
40.

41.
42.

231

KN, do Rosario VE, 1993. The importance of sensitive detection of malaria parasites in the human and insect hosts in
epidemiological studies, as shown by the analysis of field
samples from Guinea Bissau. Trans R Soc Trop Med Hyg 87:
649653.
Postigo M, Mendoza-Leon A, Perez HA, 1998. Malaria diagnosis by the polymerase chain reaction: a field study in southeastern Venezuela. Trans R Soc Trop Med Hyg 92: 509511.
Zhou M, Liu Q, Wongsrichanalai C, Suwonkerd W, Panart K,
Prajakwong S, Pensiri A, Kimura M, Matsuoka H, Ferreira
MU, Isomura S, Kawamoto F, 1998. High prevalence of Plasmodium malariae and Plasmodium ovale in malaria patients
along the Thai-Myanmar border, as revealed by acridine orange staining and PCR-based diagnoses. Trop Med Int Health
3: 304312.
Singh B, Bobogare A, Cox-Singh J, Snounou G, Abdullah MS,
Rahman HA, 1999. A genus- and species-specific nested
polymerase chain reaction malaria detection assay for epidemiologic studies. Am J Trop Med Hyg 60: 687692.
Purnomo, Solihin A, Gomez-Saladin E, Bangs MJ, 1999. Rare
quadruple malaria infection in Irian Jaya Indonesia. J Parasitol 85: 574579.
May J, Mockenhaupt FP, Ademowo OG, Falusi AG, Olumese
PE, Bienzle U, Meyer CG, 1999. High rate of mixed and
subpatent malarial infections in southwest Nigeria. Am J Trop
Med Hyg 61: 339343.
Peters W, Standfast H, 1957. Report on a malaria survey in the
Sepik district. Med J Aust 1: 861868.
Peters W, 1960. Studies on the epidemiology of malaria in New
Guinea, Part I. Holoendemic malariathe clinical picture.
Trans R Soc Trop Med Hyg 54: 242249.
Cattani JA, Tulloch JL, Vrbova H, Jolley D, Gibson FD, Moir
JS, Heywood PF, Alpers MP, Stevenson A, Clancy R, 1986.
The epidemiology of malaria in a population surrounding Madang, Papua New Guinea. Am J Trop Med Hyg 35: 315.
Desowitz RS, Spark RA, 1987. Malaria in Maprik area of the
Sepik region, Papua New Guinea: 19571984. Trans R Soc
Trop Med Hyg 81: 175176.
Genton B, al-Yaman F, Beck HP, Hii J, Mellor S, Narara A,
Gibson N, Smith T, Alpers MP, 1995. The epidemiology of
malaria in the Wosera area, East Sepik Province, Papua New
Guinea, in preparation for vaccine trials. I. Malariometric indices and immunity. Ann Trop Med Parasitol 89: 359376.
Janovy Jr J, Clopton RE, Clopton DA, Snyder SD, Efting A,
Krebs L, 1995. Species density distributions as null models
for ecologically significant interactions of parasite species in
an assemblage. Ecol Model 77: 189196.
Poulin R, 1997. Species richness of parasite assemblages: evolution and patterns. Annu Rev Ecol Syst 28: 341358.
Cohen JE, 1971. Estimation and interaction in a censored 2
2 2 contingency table. Biometrics 27: 379386.
Bockarie M, Kazura J, Alexander N, Dagoro H, Bockarie F,
Perry R, Alpers M, 1996. Transmission dynamics of Wuchereria bancrofti in East Sepik Province, Papua New Guinea.
Am J Trop Med Hyg 54: 577581.
Bockarie MJ, Alexander ND, Hyun P, Dimber Z, Bockarie F,
Ibam E, Alpers MP, Kazura JW, 1998. Randomised community-based trial of annual single-dose diethylcarbamazine with
or without ivermectin against Wuchereria bancrofti infection
in human beings and mosquitoes. Lancet 351: 162168.
Tournamille C, Le Van Kim C, Gane P, Cartron JP, Colin Y,
1995. Molecular basis and PCR-DNA typing of the Fya/Fyb
blood group polymorphism. Hum Genet 95: 407410.
Barker Jr RH, Banchongaksorn T, Courval JM, Suwonkerd W,
Rimwungtragoon K, Wirth DF, 1994. Plasmodium falciparum
and P. vivax: factors affecting sensitivity and specificity of
PCR-based diagnosis of malaria. Exp Parasitol 79: 4149.
Hanscheid T, Valadas E, 1999. Letter to the editor. Am J Trop
Med Hyg 61: 179.
Chiodini Pl, Moody AH, Cooke AH, 1999. Letter to the editor.
Am J Trop Med Hyg 61: 179.