Bacterial, viral and parasitic enteric pathogens associated with acute diarrhea in hospitalized children from northern Jordan

Authors


*Corresponding author. Tel.: +1 (404) 639-3334; Fax: +1 (404) 639-3333, E-mail address: mcy6@cdc.gov

Abstract

To determine the etiology of acute diarrhea in Jordanian children under 5 years of age, we examined stool samples from 265 children admitted to the pediatric ward at Princess Rahma Hospital for Children, Irbid, Jordan, for parasites, rotavirus and enteric bacteria. Using both traditional and molecular diagnostic techniques, we detected enteropathogens in 66.4% of patients with diarrhea. A single enteric pathogen was detected in 50.9% of the children, and multiple pathogens were detected in 15.5%. The prevalence of enteropathogens identified was as follows: rotavirus (32.5%), enteropathogenic Escherichia coli (12.8%), enteroaggregative E. coli (10.2), enterotoxigenic E. coli (5.7%), Shigella spp. (4.9%), Entamoeba histolytica (4.9%), Salmonella spp. (4.5%), Campylobacter jejuni/coli (1.5%), Cryptosporidium spp. (1.5%), enteroinvasive E. coli (1.5%), eae-, Ehly-positive E. coli (0.8%), Giardia lamblia (0.8%) and Yersinia enterocolitica (0.4%). No Vibrio cholerae, Shiga toxin-producing E. coli, microsporidia, adenovirus or small round virus were detected. Findings from this study demonstrate that rotavirus and several types of diarrheagenic E. coli, which are not screened for during routine examinations of stool samples in public health laboratories, were the most frequently detected enteropathogens in these children. Our findings highlight the value of using a combination of traditional and molecular techniques in the diagnosis of diarrheal disease in this population.

1Introduction

Diarrheal disease continues to be an important cause of morbidity and mortality among young children in developing countries [1]. The primary health care programs in Jordan have played an important role in reducing the morbidity and mortality associated with these diseases; however, the amount of diarrheal illness in the population, particularly among young children, remains a concern, and knowledge of its causes is limited. Methods currently used in public health laboratories allow for the identification of Salmonella spp., Shigella spp., Entamoeba histolytica and Giardia, but they would not detect the following recognized enteropathogens: rotavirus, Cryptosporidium, Campylobacter spp. and diarrheagenic strains of Escherichia coli. Because an etiologic agent is not detected for a large portion of patients with diarrhea, the possibility exists that a portion of the undiagnosed illness may be attributable to one or more of the latter enteropathogens. Several studies have examined the role of specific enteropathogens in childhood diarrheas in Jordan [2–4]; however, no comprehensive studies describing the prevalence of viral, bacterial and parasitic enteropathogens, especially newly recognized ones, and the clinical features of their infections in Jordanian children have been reported. Because knowledge of the etiologies of diarrhea in a population is the first step in developing effective control and prevention programs, we used a combination of traditional and molecular diagnostic techniques to survey the stools of Jordanian children with acute diarrhea for the above enteropathogens. The present study is the first comprehensive survey to assess the prevalence of these organisms in this population and evaluate the clinical features of their infections.

2Materials and methods

2.1Study population

During the peak diarrheal season (May–August) of 1993 and 1994, all children up to 5 years of age who were admitted with acute diarrheal diseases to Princes Rahma Hospital for Children, Irbid, Jordan, were studied. Diarrhea was defined as the passage of three or more loose or watery stools in the preceeding 24 h. Clinical examination of each patient by pediatricians was followed by treatment and monitoring until discharge. The pediatrician completed a questionnaire for each patient containing the following information: age, sex, clinical symptoms (fever, vomiting, dehydration status), type of diarrhea and medical care sought before hospitalization.

2.2Stool and blood samples

Fresh stool samples were collected from diarrheal patients and transferred to the Microbiology Laboratory at Yarmouk University, Irbid, Jordan, on ice packs and processed within 4 h of collection. Aliquots were also stored at −70°C for subsequent rotavirus detection. Blood was collected from children enrolled in the study for routine analysis of hematocrit, white blood cell counts and serum electrolytes.

2.3Parasite detection

A smear of feces in 0.9% saline was examined microscopically for the presence of leukocytes, red blood cells and E. histolytica and Giardia lamblia cysts and trophozoites. In addition, stool samples were concentrated using a formyl-ether technique for identification of cysts of E. histolytica and G. lamblia. Thin smears were prepared from unconcentrated suspensions of stool, fixed with methanol for 5 min, air-dried and examined for microsporidia and Cryptosporidium spp. For detection of Cryptosporidium oocysts, smears were prepared with Kinyoun and auramine–rhodamine stains (Difco, Detroit, MI, USA) according to the manufacturer's instructions. Slides were scanned at 400× magnification, and suspected oocysts were confirmed under oil immersion. Microsporidia were detected by light microscopic examination and by a rapid fluorescence technique. For the detection of microsporidial spores by light microscopy, smears were stained by Weber's modified trichome method as described [5]. Slides were examined by the scanning of 100 oil immersion fields (1000× magnification). With this stain, the microsporidial spore walls stain bright pinkish color. For the fluorescence technique, fixed smears were stained by using the fluochrome stain Uvitex 2B as described by van Gool et al. [6]. Stained smears were examined under a Leitz fluorescence microscope equipped with an excitation filter with a transmission range from 355 to 425 nm [6]. The stained microsporidial spores appeared white fluorescence or reddish brown.

2.4Virus detection

Enzyme-linked immunosorbent assay. Stool samples were tested for rotavirus specific antigens using a commercial kit according to the manufacturer's instructions (Amico Laboratories, Inc., IL, USA).

Direct electron microscopy. A 10% stool suspension in phosphate-buffered saline (PBS) was centrifuged at 8497×g at 4°C for 10 min (Sigma 2K15) to remove debris. The supernatant was collected and centrifuged at 48 840×g at 4°C for 2 h using a MSE centrifuge (Euoropa, 24M). The pellet was resuspended in 200 μl of PBS. A 100-μl aliquot of the suspension was placed on a wax sheet, and a 200-mesh Formvar carbon-coated copper grid was floated on the surface for 3 min. The excess fluid was absorbed by a filter paper. The grid was then floated on a drop of distilled water for 1 min. After removal of the excess fluid, the grid was floated on 100 μl of 2% aqueous phosphotungstic acid (pH 6.8) for 3 min. The grid was then blotted, air-dried and examined for virus particles with a Zeiss 10CR electron microscope at a magnification of 50 000. At least 10 grid squares were examined for viral particles for a period of 15–20 min per grid.

2.5Bacterial detection

All specimens were tested for Salmonella spp., Shigella spp., Vibrio cholerae, Campylobacter jejuni/coli, Yersinia enterocolitica and diarrheagenic E. coli. Fresh stool specimens were inoculated onto Drigalski agar medium (Institute Pasteur Production, Paris, France), thiosulfate–citrate–bile salts–sucrose medium (Difco) and SalmonellaShigella agar medium (Institute Pasteur Production), and were incubated at 37°C for 24 h. C. jejuni/coli was isolated on Butzler selective medium (Oxoid) at 42°C for 48 h under microaerophilic conditions (10% CO2, 5% O2, 85% N2) and was identified by confirming the darting motility by phase contrast microscopy and the characteristic morphology by Gram staining. Y. enterocolitica was detected on Drigalski agar plates incubated at room temperature for 1 week. Bacterial isolates were identified by standard microbiological methods [7].

To test for diarrheagenic E. coli, 10 colonies that morphologically resembled E. coli were picked from Drigalski agar plates for each stool sample studied, pooled in brain–heart infusion broth (Difco) containing 10% glycerol, and stored at −70°C for further use. At the time of the study, individual isolates were subcultured onto blood agar plates (BAP) for DNA template preparation for the polymerase chain reaction (PCR) assay. A 1-cm long sweep of bacterial growth from the quadrant of the BAP with confluent or the heaviest growth was suspended in 300 μl sterile distilled water and boiled for 10 min to prepare the DNA template. A 2-μl aliquot of this suspension was added to 22 μl of PCR mixture (50 mM KCl, 10 mM Tris–HCl (pH 8.3), 1.5 mM MgCl2, 0.2 mM each deoxynucleotide triphosphate and 0.6 U of Taq polymerase (Boehringer Mannheim Biochemicals, Indianapolis, IN, USA)) and 1 μl of primer mix 1 or 2 containing the primer at the concentrations shown in Table 1.

Table 1.  Oligonucleotide primers used in the detection of diarrheagenic E. coli
Target gene/pathogenPrimerPrimer sequence (5′–3′)Concentration (μM)Product (bp)Reference
Mix 1
LT/ETECLTIb-1TCT CTA TGT GCA TAC GGA GC2.0320[8]
 LTI-2CCA TAC TGA TTG CCG CAA T2.0  
ST/ETECSTI-1TTA ATA GCA CCC GGT ACA AGC AGG4.0147[8]
 STI-2CTT GAC TCT TCA AAA GAG AAA ATT AC4.0  
stx1/STEC (EHEC)SLTI-1CAG TTA ATG TGG TGG CGA AG0.4475[8]
 SLTINS2CAC AGA CTG CGT CAG TGA GG0.4  
stx2/STEC (EHEC)SLTII-1CTT CGG TAT CCT ATT CCC GG0.4863[8]
 SLTIINS7CGC TGC AGC TGT ATT ACT TTC0.4  
ipaH/EIECIPAH15BGCC GGT CAG CCA CCC TCT GAG0.4620[9]
 IPAH8BGTT CCT TGA CCG CCT TTC CGA0.4  
Mix 2
EAF plasmid/EPECEAF1CAG GGT AAA AGA AAG ATG ATA A1.0397[11]
 EAF25TAT GGG GAC CAT GTA TTA TCA1.0  
eae/EPEC, EHECC1TCG TCA CAG TTG CAG GCC TGG T0.41110[10]
 C2CGA AGT CTT ATC CGC CGT AAA GT0.4  
uidA/O157:H7PT-2GCG AAA ACT GTG GAA TTG GG0.4252[12]
 PT-3TGA TGC TCC ATA ACT TCC TG0.4  
pCVD432S/EAggECpCVD432/STARTCTG GCG AAA GAC TGT ATC AT0.4630[14]
 pCVD432/STOPCAA TGT ATA GAA ATC CGC TGT T0.4  
Ehly/EHECMFS1FACG ATG TGG TTT ATT CTG GA4.0166[13]
 MFS1RCTT CAC GTC ACC ATA CAT AT4.0  

Published primers specific for genes encoding thermolabile (LT) and thermostable (ST) enterotoxins [8], Shiga toxin 1 (Stx1) and Shiga toxin 2 (Stx2) [8], and invasion plasmid antigen H (ipaH) [9] were incorporated in reaction 1. Primers specific for sequences encoding intimin [10], the enteropathogenic E. coli (EPEC) adherence factor (EAF) plasmid [11], the O157 allele of β-glucuronidase [12], the sequences encoding enterohemolysin [13], and the 60-MDa plasmid that plays a role in mediating the enteroaggregate pattern of adherence in E. coli[14] were incorporated in reaction 2.

Primers were combined if they had similar annealing temperatures and produced amplicons of sizes that could be separated and differentiated by agarose gel electrophoresis. The final reaction mixture was overlaid with 50 μl sterile mineral oil.

All samples were amplified in a programmable Thermocycler (Biometra, Göttingen, Germany) for the following cycling parameters: 95°C for 30 s to denature DNA, at 55°C for 1 min to anneal the primers, and then at 72°C for 1 min to extend the annealed primers. The cycle was repeated 35 times, and final extension was performed at 72°C for 10 min. A control containing water instead of template was included in each experiment to exclude the possibility of reagent contamination. E. coli strains used as controls in the PCR tests included TX1 (O78:H12) containing the gene for ST; H10407 (O78:H11) containing the genes for ST and LT; B170 (O111:NM) containing the genes for eae and the EAF plasmid; C4193-1 (O157:H7) containing the genes for stx1, stx2, eae, Ehly and the O157 allele of uidA; EDL1284 (O124:NM) containing the genes for ipaH; and 3591–78 (O75:NM) containing the plasmid involved in mediating the enteroaggregate type adherence to HEP-2 cells in vitro. The sequences of the oligonucleotide primers used to detect the different categories of diarrheagenic E. coli are shown in Table 1. PCR products were electrophoresed in 1.5% SeaKem agarose gels (FMC Bioproducts, Rockland, ME, USA) for 3 h at 150 V using a horizontal electrophoresis apparatus, stained with ethidium bromide, and visualized under UV light. A 123-bp DNA ladder and 1-kb ladder (Gibco BRL, Gaithersburg, MD, USA) were used as molecular mass markers. PCR was also performed on 10–20 isolated colonies to confirm results obtained from sweeps.

2.6Statistical analysis

Statistical analysis was performed using SAS, version 7.0 (SAS Institute, Inc., Cary, NC, USA), S-Plus, version 4.5 (MathSoft, Inc., Seattle, WA, USA) and EpiInfo, version 6.0 (CDC, Atlanta, GA, USA).

3Results

3.1Children

During the peak diarrheal seasons of 1993 and 1994, fecal samples from 265 children admitted with acute diarrhea to the pediatric ward at Princess Rahma Hospital for Children were examined for enteric pathogens. Table 2 shows the clinical symptoms and other features of children included in the survey. The male:female ratio was 1.3:1, and 183 (69.1%) were under 1 year of age. Before admission to the hospital, 151 (57.0%) of the children studied had received medical care at one or more health care facilities. Ninety-five (35.9%) sought medical care at general practitioners, 87 (32.8%) at pediatricians, 52 (19.6%) at health centers and 29 (10.9%) were previously hospitalized. Of the 151 patients who sought medical care, 92 (60.9%) received oral rehydration solution (ORS) and 36 (23.8%) received antibiotic therapy. The characteristics and clinical symptoms of patients whose specimens yielded a pathogen (n=176) were not significantly different from those of patients whose specimens did not yield a pathogen (n=89). There were no deaths among the 265 children included in the study.

Table 2.  Clinical symptoms and other characteristics of 265 children hospitalized with diarrhea
  1. aMean±S.D. number of loose or watery stools per 24 h.

Clinical and other characteristicsNo. (%)
Age (months)
   0–226 (9.8)
   3–564 (24.2)
   6–1193 (35.1)
   12–2354 (20.4)
   24–3516 (6.0)
   36–5912 (4.5)
Sex ratio (M/F)1.3/1
Sought medical care before hospitalization151 (57.0)
Diarrhea
   Frequencya6.4±2.4
   Watery243 (91.7)
   Mucus148 (55.8)
   Bloody stools28 (10.6)
Vomiting206 (77.7)
Temperature ≥38.5°C139 (52.5)
Severe dehydration26 (9.8)
Therapy received during hospitalization
   IV fluid166 (62.6)
   ORS154 (58.1)
   Antibiotics104 (39.2)

3.2Prevalence of enteric pathogens

The various enteric pathogens and their frequency of recovery are shown in Table 3. A single enteric pathogen was identified in stools of 135 (50.9%) children admitted with acute diarrhea. Multiple pathogens were identified in 41 (15.5%) stools, and none were identified in 89 (33.6%). Among the 36 patients who had received antibiotic therapy before hospitalization, bacterial enteropathogens were isolated from 11 (30.6%). Rotavirus was the most common enteropathogen, detected in 86 (32.5%) patients, followed by EPEC in 34 (12.8%) patients, enteroaggregative E. coli (EAggEC) in 27 (10.2%) patients and enterotoxigenic E. coli (ETEC) in 15 (5.7%) patients. Among the 15 patients from whom ETEC strains were isolated, LT-producing E. coli strains were detected in nine (60%), ST-producing E. coli strains in four (26.7%) and LT/ST-producing E. coli strains in two (13.3%). The most frequent patterns of multiple infections were the combination of EAggEC with rotavirus (13 patients), followed by EPEC with rotavirus (10 patients) and ETEC with EAggEC (five patients).

Table 3.  Prevalence of viral, bacterial and parasitic enteropathogens detected in children hospitalized with acute diarrhea
  1. aTotal exceeds 176 because of mixed infections.

  2. bThe % of children from whom the designated pathogen was isolated or detected.

EnteropathogenNumber of isolatesa (%)b
Rotavirus86 (32.5)
EPEC34 (12.8
EAggEC27 (10.2)
ETEC 
   LT-producing9 (3.4)
   ST-producing4 (1.5)
   LT/ST-producing2 (0.8)
Shigella spp.13 (4.9)
E. histolytica13 (4.9)
Salmonella spp.12 (4.5)
C. jejuni/coli4 (1.5)
Cryptosporidium spp.4 (1.5)
Enteroinvasive E. coli4 (1.5)
eae-, Ehly-positive E. coli2 (0.8)
G. lamblia2 (0.8)
Y. enterocolitica1 (0.4)
V. cholerae0 (0.0)
STEC0 (0.0)
Microsporidia0 (0.0)
Adenovirus0 (0.0)
Small round virus0 (0.0)

3.3Clinical and other features

Table 4 summarizes the clinical features of hospitalized patients in relation to the microorganisms isolated. An enteropathogen was listed if it infected at least eight individuals. The number of stools in the 24 h prior to admission was 6–9 for all patient groups. Watery stools occurred in 62.5% of patients infected with EAggEC, in 70% of patients infected with Salmonella, and in 90–100% of patients infected with EPEC, rotavirus, Shigella, ETEC and E. histolytica. Blood was observed in stools from about 60% of patients infected with Shigella, from 25% infected with E. histolytica and from 20% of patients infected with Salmonella. There were no significant differences in clinical symptoms of patients from whom a single enteric pathogen was isolated, and those with multiple enteric pathogens.

Table 4.  Clinical features of hospitalized patients infected with single etiologic agentsa
  1. aData shown only for etiologic agents isolated from eight or more patients.

Etiologic agentNo. of patientsMean age (months)±S.D.Watery stool (%)Mucus in stool (%)Visible blood in stool (%)Vomiting (%)Temperature ≥38.5°C (%)
Rotavirus537.2±5.296.254.79.477.449.1
EPEC197.0±5.09057.95.384.247.4
Shigella spp.1017.4±15.810080607060
Salmonella spp.108.8±6.67050207080
ETEC1015.2±10.21006006060
EAggEC88.1±3.562.537.5075.537.5
E. histolytica810.8±6.710062.52510087.5
S.D.: standard deviations.

The laboratory data for all groups of patients on admission showed normal electrolytes, urea and packed cell volume, but slightly elevated white blood counts.

4Discussion

Diarrheal diseases are a major public health problem for children in Jordan, as in other developing countries. Knowledge of the enteropathogens responsible for diarrheal illnesses is essential for implementation of appropriate public health measures to control these diseases. This study, which covered the peak diarrheal seasons of two consecutive years (1993–1994), is the first study conducted in Jordan to determine the prevalence of viral, bacterial and parasitic enteropathogens associated with acute diarrhea in hospitalized children under 5 years of age. Using a combination of traditional and molecular diagnostic techniques, we detected one or more enteropathogens in 66% of patients with diarrhea. With the methods routinely used by public health laboratories, enteropathogens would have been detected in only 15% of patients. The prevalence of the various pathogens observed during the course of the study did not vary significantly between the first and second year of the study.

Rotavirus was the most prevalent etiologic agent and was detected in 33% of hospitalized children. Our findings for rotavirus infection were similar to those reported by Meqdam et al. [2], who reported that approximately 40% of infants and young children under 3 years of age with diarrhea from northern Jordan were infected with rotavirus. The prevalence of rotavirus in Jordan as estimated by these two studies is comparable to that reported for other developing and developed countries, which ranges from 30% to 50%[15].

Diarrheagenic E. coli strains were the second most common group of enteropathogens detected. We identified these strains by molecular techniques and classified them according to the combination of virulence, or virulence-associated, markers they possess. We considered strains to be EPEC if they contain the gene for intimin (eae) and the EAF plasmid, which play an essential role in mediating attaching and effacing lesions in vivo. Strains containing a plasmid that plays a role in mediating an aggregative type of adherence pattern to eukaryotic cells in vitro we considered to be EAggEC, and those that contain the gene for IpaH (ipaH), which is carried on the plasmid that plays an essential role in mediating the invasion of intestinal epithelial cells, we identified as EIEC. We classified strains as ETEC or Shiga toxin-producing E. coli (STEC) if they contained the genes for the toxins identified with their respective categories of diarrheagenic E. coli.

Of the diarrheagenic E. coli strains detected, EPEC strains were the most prevalent (12.8%), followed by EAggEC (10.2%) and ETEC (5.7%). The relative prevalence of these categories of diarrheagenic E. coli was similar to that observed earlier among malnourished children [16] and children with acute diarrhea in northern Jordan [4], and their presence in children with diarrhea in other developing countries has been documented [17–22].

No STEC strains, of which E. coli serotypes O157:H7, O26:H11 and O111:H8/NM are the most prevalent types, were detected in the present study. The traditional Jordanian practice of consuming thoroughly cooked foods and boiled milk may, in part, explain this finding because STEC infections are often associated with the consumption of undercooked, contaminated foods, such as ground beef and unpasteurized milk from cows [23]. Although no STECs were detected, potential diarrheagenic E. coli strains harboring eae and a marker (Ehly) for a 60-MDa plasmid (EHEC plasmid) that plays a role in regulating adherence in E. coli O157:H7 were detected in two patients with non-bloody diarrhea. These markers and the genes for Shiga toxins were initially proposed to define the enterohemorrhagic E. coli category of diarrheagenic E. coli[24]. Because no other recognized enteropathogens were detected among these patients, further studies are needed to assess the role of bacteria carrying eae and the EHEC plasmid in diarrheal illness.

The prevalence of C. jejuni/coli, Y. enterocolitica and Cryptosporidium spp. was similar to that previously reported in separate studies from northern Jordan [3,4], and V. cholerae, which is not endemic in Jordan, and microsporidia, which are usually associated with diarrhea in immunocompromised patients, were not detected.

The clinical symptoms associated with infection by the most prevalent enteropathogens are shown in Table 4. We examined the collective predictive capacity of temperature, vomiting and the clinical characteristics of diarrhea using multi-way contingency tables, logistic regression models, tree classification models and cluster analysis. We found no relationships for predicting the presence of individual pathogens or a category (viral, bacterial and parasitic) of diarrheal pathogens. The diarrhea for the majority of children resolved by the time of discharge from the hospital; however, 13 children were discharged with failure to thrive and three with chronic diarrhea. There was no association of a particular enteropathogen with children discharged with these conditions.

The information about the prevalence of a wide range of enteropathogens should facilitate the control and management of diarrheal diseases among infants and children in the country. Although this study was not a case-control study, the large number of rotavirus and EPEC infections suggests that these organisms are important causes of diarrhea in this population. The use of vaccines to control rotavirus infections is now recommended in developing countries [25], and its use in Jordan merits consideration. More information is needed about the sources, modes of transmission and risk factors of EPEC infections in Jordanian children to develop methods to control these infections.

Acknowledgements

The authors would like to thank Robert M. Hoekestra at the Division of Bacterial and Mycotic Diseases, CDC, Atlanta, GA, USA, for conducting statistical analysis.

Ancillary