What Is Known
- Diarrhea is a leading cause of child morbidity and mortality worldwide.
- Traditional methods to identify enteric pathogens in low-resource settings (including bacterial culture, stool microscopy, and antigen-based assays) lack specificity and sensitivity.
- Childhood diarrhea is associated with suboptimal child growth.
What Is New
- Quantitative polymerase chain reaction stool analysis methods allow for the rapid, sensitive screening of stool for multiple causative enteropathogens, and may help identify those children with diarrhea who are most at risk for suboptimal growth.
- Serum antibodies to the bacterial products lipopolysaccharide and flagellin hold promise as emerging intestinal biomarkers.
Diarrhea continues to be a leading cause of morbidity and mortality for children worldwide. In 2011, an estimated 700,000 child deaths were caused by diarrhea, with 72% of those deaths occurring in children younger than 2 years (1). Diarrhea and nutritional status are closely interrelated, as undernutrition increases susceptibility to diarrhea-causing infections (2), and childhood diarrhea predisposes to growth faltering. It is estimated that each episode of diarrhea before age 2 years increases a child's risk of stunting by 5% (3). The pathologic mechanisms linking diarrhea and undernutrition are complex, but likely involve alterations in immune function, increased intestinal permeability, impaired nutrient absorption, and repeated exposure to multiple enteropathogens (4).
An essential step in reducing diarrheal morbidity is the identification of causative infectious pathogens, but traditional methods of diagnosis including bacterial culture, stool microscopy, and antigen-based assays have poor sensitivity (5,6). Newer polymerase chain reaction (PCR)-based methods to detect diarrhea-causing enteropathogens have recently been developed (7), but may detect pathogens that do not cause clinical symptoms of diarrhea. For example, in a longitudinal study that compared diarrheal samples to prediarrheal-matched controls, PCR stool analysis was shown to correlate with diarrhea caused by rotavirus, astrovirus, and Shigella/enteroinvasive Escherichia coli only when the data were analyzed quantitatively (8). Similarly, a study of enteropathogens in diarrhea cases and nondiarrheal controls found high rates of detection even in asymptomatic controls; the specificity improved when pathogen quantity was assessed (9). In these ways, the calculation and interpretation of pathogen burden is of critical importance to the utility of these PCR methods.
In resource-limited settings, diarrhea-causing enteropathogens have most commonly been described in hospitalized children with acute/persistent severe diarrhea (10,11). In hospitalized children younger than 5 years in Dar es Salaam, Tanzania, diarrheagenic E coli was found by standard bacteriologic culture (12) in 22.9% of patients (13). The etiology of more common, ambulatory diarrhea in Tanzania, however, remains poorly reported. We sought to determine the etiology of community-acquired diarrhea in Tanzanian infants and young children, and hypothesized that identifiable diarrheal infections would correlate with a high pathogen burden and place children at risk for suboptimal growth. In addition, we evaluated the relation of novel biomarkers of gastrointestinal function with outcomes in this cohort.
A convenience sample of infants was selected from a prospective randomized trial of daily zinc and/or multivitamin supplementation in Dar es Salaam, Tanzania (14,15) (clinicaltrials.gov identifier NCT00421668). Consenting mothers who were confirmed to be HIV-negative were enrolled into the study and their infants were randomized between 5 and 7 weeks of age. Infants of multiple births and infants with congenital anomalies or other conditions that would interfere with the study procedures were excluded. Infants were enrolled at 6 weeks of age and studied for 18 months. Demographic data were collected at enrollment, and morbidity and anthropometric data were collected at monthly outpatient follow-up visits. All of the children received a periodic dose of vitamin A every 6 months as per standard of care in Tanzania.
Mothers were asked to return to the study clinic with their infants every 4 weeks for data collection and standard clinical care, including growth monitoring, immunizations, and routine medical treatment for illnesses. During these visits, nurses interviewed caregivers about the child's morbidity history, and measured child's weight on a digital infant balance with 10-g precision (Tanita, Inc. Arlington Heights, IL) and length with 1-mm precision using a rigid length board with a movable foot piece.
Stool samples were collected between January 2009 and January 2011 from children with acute diarrhea, defined as 3 or more watery stools in a 24-hour period. Stool from children with diarrhea was collected at the study site via fecal swab from either the diaper (napkin) or collection container (pot) and placed into vials containing transport media. Vials were placed into a cooler and transported to a central laboratory daily, where they were promptly frozen at −80°C. Blood was drawn from infants at 6 weeks and 6 months of age, and frozen at −80°C.
A PCR-based TaqMan array method was used to screen stool samples for 15 enteropathogens. Previously described pathogen-specific quantification cycle (Cq) cutoffs were used to define positive stool PCR results (7). The primary outcome was the prevalence of any identifiable stool enteropathogen. Secondary outcomes included weight-for-age z score, length-for-age z score (LAZ), and weight-for-length z score (WLZ; as measured by standard anthropometric measurements) and intestinal biomarkers (including antibodies to lipopolysaccharide [LPS] and flagellin and the amino acid citrulline [CIT]).
Flagellin- and LPS-specific IgA and immunoglobulin G (igG) levels were quantitated by enzyme-linked immunosorbent assay, as has been previously reported (16). Microtiter plates were coated overnight with purified E coli flagellin (100 ng/well), or purified E coli LPS (2 μg/well). Serum samples from study subjects diluted 1:200 were applied to wells coated with flagellin or LPS. After incubation and washing, the wells were incubated with either anti-human IgA (KPL, Inc., Gaithersburg, MD) or IgG (GE Healthcare, Inc., Pittsburgh, PA) coupled to horseradish peroxidase. Quantitation of total immunoglobulins was performed using the colorimetric peroxidase substrate tetramethylbenzidine, and optical density was read at 450 nm with an enzyme-linked immunosorbent assay plate reader. Data are reported as optical density corrected by subtracting background (determined by readings in samples lacking serum).
Underivatized CIT concentrations in serum specimens were evaluated in the Clinical and Epidemiological Research Laboratory in the Department of Laboratory Medicine, Boston Children's Hospital. Samples were ultracentrifuged to remove high molecular weight proteins. The separation selectivity of CIT was obtained using a C18 BEH column and the ion-pairing agent pentadecafluorooctanoic acid on a Waters Acquity UPLC System (Waters Corporation, Milford, MA). Positive electrospray ionization and a Waters Quattro premier triple quadrupole mass spectrometer were used to detect CIT by its characteristic 176 to 170 and 176 to 113 m/z transitions.
Data were double-entered using Microsoft Access software at the central study site, then converted to SAS datasets and uploaded to a secured UNIX-based server for analysis. t Tests were used to compare mean anthropometric measures, CIT, flagellin, and LPS concentrations at 6 weeks, 6 months, and 12 months of age among children who tested positive for any pathogen vs. those who tested negative for all pathogens. t Tests used the pooled method when variances were equal and Satterthwaite method when variances were unequal. A scatter plot was constructed to assess the relation between enteropathogen burden (as measured by PCR Cq) and anthropometric outcomes. All analyses were performed using SAS Software (version 9.2; SAS Institute, Inc. Cary, North Carolina). Institutional approval was granted by the Harvard School of Public Health Human Subjects Committee, the Muhimbili University of Health and Allied Science Committee of Research and Publications, and the Tanzanian National Institute of Medical Research.
One hundred twenty-three subjects with diarrhea were enrolled. Table 1 lists the baseline characteristics of the substudy population, including both maternal and infant demographic and nutritional status. Eighty-seven (70.7%) of the mothers had been pregnant at least once previously. One hundred twenty (97.6%) of infants had a birth weight >2500 g, and 8 (6.5%) were born before 37 weeks of gestational age.
The mean ± SD age of diarrheal illness that resulted in a stool sample was 12.4 ± 3.9 months (range 2.9–19.9 months). One stool sample was collected from each of the n = 123 subjects. Table 2 lists the enteric pathogens detected. Thirty-five pathogens were identified in 34 (27.6%) subjects: 11 rotavirus, 9 Cryptosporidium spp, 7 Shigella spp, 3 Campylobacter jejuni/coli, 3 ST-enterotoxigenic E coli, and 2 enteropathogenic E coli (EPEC). The Cq thresholds used to determine a positive sample were <30 for rotavirus, <25 for heat stable toxin producing enterotoxigenic Escherichia coli (ST-ETEC), <20 for EPEC eae, < 25 for EPEC bfpa, <30 for shigella/enteroinvasive e. coli (EIEC), <30 for campylobacter, and <30 for cryptosporidium. These are quantities previously associated with diarrhea in similar settings (7). There was no difference in pathogen detection in children receiving zinc (n = 14 vs n = 47 in the group with any pathogen vs no pathogen detected, respectively, P = 0.25) or multivitamin supplementation (n = 17 vs n = 51, P = 0.47). One subject had 2 pathogens detected simultaneously (EPEC and Campylobacter). Compared with children without an identified pathogen, subjects with any identified enteropathogen had significantly lower WLZ (−0.55 ± 1.10 vs 0.03 ± 1.30, P = 0.03) at the final clinic visit. There were no significant differences in weight-for-age z score or LAZ between the 2 groups.
The mean ± SD age at the final clinic visit was 18.2 ± 2.5 months. The mean ± SD time of follow-up after stool collection was 5.9 ± 4.0 months. Figure 1 shows the relation between PCR Cq and LAZ at the final clinic visit (R = 0.329, P = 0.062) in those subjects with any positive enteropathogen, suggesting a possible trend between lower Cq (indicating higher pathogen burden) and linear growth deficits.
Fifty of the 123 subjects (40.7%) underwent serum analysis for antibodies to the bacterial products LPS and flagellin, which are emerging intestinal biomarkers that may correlate with intestinal permeability (16,17). Table 3 reports the antibody levels to LPS and flagellin at 6 months of age, in those with any identified enteropathogen versus no detected enteropathogen. Subjects with any identified diarrheal enteropathogen had lower IgA antibodies to LPS (0.75 ± 0.27 vs 1.13 ± 0.77, P = 0.01) and flagellin (0.52 ± 0.16 vs 0.73 ± 0.47, P = 0.02) at age 6 months than those without an identified pathogen. These differences were not present at age 6 weeks. There was no significant difference between IgG antibodies to LPS and flagellin between the 2 groups.
CIT, a nonprotein amino acid produced by enterocytes, is a biomarker of intestinal mass and function (18–20). Subjects with any identified diarrheal enteropathogen had a nonsignificantly lower CIT concentration (17.2 ± 1.6 vs 21.0 ± 8.8, P = 0.06) at age 6 weeks than those without an identified pathogen. This trend did not persist at age 6 months of age (20.8 ± 5.1 vs 20.7 ± 6.5, P = 0.99).
The identification of diarrhea-causing enteropathogens worldwide is an important factor in reducing global child morbidity and mortality, and our understanding of enteropathogens in various geographic settings is growing. The Global Enteric Multicenter Study (GEMS) (21) identified 4 causative enteropathogens in nonhospitalized children in Africa and Asia with moderate-to-severe diarrhea, including rotavirus, Cryptosporidium, enterotoxigenic E coli and Shigella spp. This study used traditional stool assays (22) (conventional bacterial culture, enzyme-linked immunosorbent assays and RT-PCR for viruses). In contrast to conventional methods, multitarget PCR stool analysis allows for a single stool sample to be screened for multiple bacterial, viral, and parasitic enteropathogens. Previous studies using qualitative PCR stool analysis have reported high levels of pathogen detection in control, nondiarrheal stool samples (79% in the study by Kabayiza et al (23), (84)% in the study by Elfving et al (9)). Platts-Mills et al (8) reported a median of 3 positive enteropathogens detected in each prediarrheal stool sample. In this way, the high sensitivity of PCR makes it difficult to differentiate between intestinal microbial colonization versus pathogenic infection. One solution to this problem is the use of the PCR Cq, which allows for quantification of pathogen burden and helps differentiate between colonizing pathogens and those causing diarrheal disease. Lower Cq (indicating higher pathogen burden) has been shown to correlate with an increased diarrheal risk (7,9) and in the case of rotavirus, ETEC and Shigella, more severe diarrhea. (23). Figure 1 shows the relation between lower Cq and LAZ at the final clinic visit, and suggests that higher pathogen burden may be associated with the development of linear growth faltering.
Rotavirus was the most commonly identified enteropathogen in our cohort. Rotavirus is a known global cause of moderate-to-severe diarrhea in young children (21), and oral rotavirus vaccination has proven to be safe and effective in African infants (24,25). In our cohort, after applying a PCR Cq threshold of <30 cycles, 8.9% of subjects with diarrhea were positive for rotavirus. In a previous study of Tanzanian children age 0 to 60 months hospitalized for diarrhea, 18% were found positive for rotavirus by enzyme-linked immunosorbent assay (26). One likely explanation for this difference is the expected higher incidence of severe diarrhea caused by rotavirus in the hospital setting as compared to the community. The Interactions of Malnutrition & Enteric Infections: Consequences for Child Health and Development (MAL-ED) study, which used ELIZA methodology to test for rotavirus (27), reported an adjusted attributable fraction of diarrhea due to rotavirus of 14.3% in Tanzania during the second year of life (28). Of note, our sample collection was completed before Tanzania initiated routine oral rotavirus vaccination in 2012, and therefore would not reflect widespread oral rotavirus vaccination within the cohort.
We found that any identified diarrheal pathogen was associated with lower subsequent WLZ, indicating acute or subacute undernutrition. Many enteropathogens have been correlated with suboptimal child growth. Cryptosporidium, which was found in 7.3% of subjects in our cohort, has been associated with persistent diarrhea (29) and childhood malnutrition (30). Nearly 5% of subjects in our cohort had diarrhea due to ETEC or EPEC, which have been associated with severe diarrhea and child malnutrition (31). In a study of infants with diarrhea in Bangladesh, Qadri et al (32) reported that infants with ETEC-positive diarrhea were more likely to become stunted and wasted at 12 months than those with ETEC-negative diarrhea, and that growth faltering persisted through age 24 months. Shigella/EIEC was detected in 5.7% of our cohort, and has previously been associated with prolonged diarrhea and growth faltering (29). Stool protein loss followed by inadequate postdiarrhea protein consumption has been proposed as a mechanism for growth faltering in the setting of Shigella infection (33).
The extent to which intestinal permeability influences diarrheal disease and child growth is controversial. Environmental enteric dysfunction has been hypothesized to be associated with small intestinal exposure to pathogenic organisms and has been proposed as more proximal cause of suboptimal childhood growth than diarrheal illness (34,35). To investigate this possibility, we examined levels of serum antibodies to LPS and flagellin. Based on the original descriptions of such antibodies as indirect markers of gut permeability, one may have postulated that stunting would correlate with increased levels of these bacterially directed antibodies. In fact, we found that lower IgA antibody levels to LPS and flagellin at age 6 months correlated with the identification of at least 1 stool pathogen. When considering more recent studies that mucosal antibodies, particularly IgA, play a critical role in regulating flagellated bacteria both by altering microbiota composition and flagellin gene expression (36,37), a possible explanation for our observations herein is that those subjects with elevated anti-flagellin and anti-LPS antibodies could have more robust immune function. This increased immunity, whether endogenous or maternal, may help to contain the quantity and/or activity of potentially diarrheagenic microbes. Currently, this assay cannot reliably distinguish between passively transferred maternal antibodies and infant antibodies, which is of particular importance given that the measurements were taken at age 6 months. Therefore, as these intestinal biomarkers become further refined, it will be important to understand whether and/or how maternal immunity affects anti-LPS and anti-flagellin antibodies. In addition to this, testing at later timepoints, including after passively transferred maternal immunity wanes, will be important. Further study in this area is needed to elucidate the complex relation between intestinal permeability, bacterial translocation, infant diet, diarrhea, and growth.
There were several limitations to our study. First, our sample size was relatively small and may have limited our ability to detect significant relations between diarrhea etiology and linear growth. Second, despite our efforts to do so, we were unable to collect prediarrheal control stools, which likely limited our ability to better differentiate colonized pathogens from disease-causing pathogens. Fortunately, enteropathogen detection in prediarrheal control samples has been well described previously (7,19). Finally, only a small proportion our cohort was able to undergo intestinal biomarker testing. Addressing these limitations in future studies will be important.
In conclusion, quantitative PCR stool analysis methods allow for the rapid screening of stool for multiple causative enteropathogens and may help identify those children with diarrhea who are most at risk for subsequent suboptimal growth. Applying PCR Cq cutoffs allows for assessment of diarrheal pathogen burden and will likely become more clinically important as the pathogen-specific Cqs become further refined. Novel intestinal biomarkers may hold promise in identifying those infants most at risk for diarrheal illness.
The authors thank the research staff who made this study possible, including the late Esther Kibona, and all of the families who participated.
1. Walker CL, Rudan I, Liu L, et al Global burden of childhood pneumonia and diarrhoea. Lancet
2. Wierzba TF, El-Yazeed RA, Savarino SJ, et al The interrelationship of malnutrition and diarrhea
in a periurban area outside Alexandria, Egypt. J Pediatr Gastroenterol Nutr
3. Black RE, Allen LH, Bhutta ZA, et al Maternal and child undernutrition
: global and regional exposures and health consequences. Lancet
4. Guerrant RL, Oria RB, Moore SR, et al Malnutrition as an enteric infectious disease with long-term effects on child development. Nutr Rev
5. Koplan JP, Fineberg HV, Ferraro MJ, et al Value of stool cultures. Lancet
6. Costantini V, Grenz L, Fritzinger A, et al Diagnostic accuracy and analytical sensitivity of IDEIA Norovirus assay for routine screening of human norovirus. J Clin Microbiol
7. Liu J, Kabir F, Manneh J, et al Development and assessment of molecular diagnostic tests for 15 enteropathogens causing childhood diarrhoea: a multicentre study. Lancet Infect Dis
8. Platts-Mills JA, Gratz J, Mduma E, et al Association between stool enteropathogen
quantity and disease in Tanzanian children using TaqMan array cards: a nested case-control study. Am J Trop Med Hyg
9. Elfving K, Andersson M, Msellem MI, et al Real-time PCR threshold cycle cutoffs help to identify agents causing acute childhood diarrhea
in Zanzibar. J Clin Microbiol
10. Tellevik MG, Moyo SJ, Blomberg B, et al Prevalence of Cryptosporidium parvum
, Entamoeba histolytica
and Giardia lamblia
among Young Children with and without Diarrhea
in Dar es Salaam, Tanzania
. PLoS Negl Trop Dis
11. Moyo SJ, Blomberg B, Hanevik K, et al Genetic diversity of circulating rotavirus
strains in Tanzania
prior to the introduction of vaccination. PLoS One
12. Moyo SJ, Maselle SY, Matee MI, et al Identification of diarrheagenic Escherichia coli
isolated from infants and children in Dar es Salaam, Tanzania
. BMC Infect Dis
13. Moyo SJ, Gro N, Matee MI, et al Age specific aetiological agents of diarrhoea in hospitalized children aged less than five years in Dar es Salaam, Tanzania
. BMC Pediatr
14. McDonald CM, Manji KP, Kisenge R, et al Daily zinc but not multivitamin supplementation reduces diarrhea
and upper respiratory infections in Tanzanian Infants: a randomized, double-blind, placebo-controlled clinical trial. J Nutr
15. Locks LM, Manji KP, McDonald CM, et al Effect of zinc and multivitamin supplementation on the growth of Tanzanian children aged 6–84 wk: a randomized, placebo-controlled, double-blind trial. Am J Clin Nutr
16. Ziegler TR, Luo M, Estivariz CF, et al Detectable serum flagellin and lipopolysaccharide and upregulated anti-flagellin and lipopolysaccharide immunoglobulins in human short bowel syndrome. Am J Physiol Regul Integr Comp Physiol
17. Sitaraman SV, Klapproth JM, Moore DA 3rd, et al Elevated flagellin-specific immunoglobulins in Crohn's disease. Am J Physiol Gastrointest Liver Physiol
18. Crenn P, Coudray-Lucas C, Thuillier F, et al Postabsorptive plasma citrulline concentration is a marker of absorptive enterocyte mass and intestinal failure in humans. Gastroenterology
19. Crenn P, De Truchis P, Neveux N, et al Plasma citrulline is a biomarker of enterocyte mass and an indicator of parenteral nutrition in HIV-infected patients. Am J Clin Nutr
20. Gosselin KB, Feldman HA, Sonis AL, et al Serum citrulline as a biomarker of gastrointestinal function during hematopoietic cell transplantation in children. J Pediatr Gastroenterol Nutr
21. Kotloff KL, Nataro JP, Blackwelder WC, et al Burden and aetiology of diarrhoeal disease in infants and young children in developing countries (the Global Enteric Multicenter Study, GEMS): a prospective, case-control study. Lancet
22. Panchalingam S, Antonio M, Hossain A, et al Diagnostic microbiologic methods in the GEMS-1 case/control study. Clin Infect Dis
2012; 55 (suppl 4):S294–S302.
23. Kabayiza JC, Andersson ME, Nilsson S, et al Diarrhoeagenic microbes by real-time PCR in Rwandan children under 5 years of age with acute gastroenteritis. Clin Microbiol Infect
24. Madhi SA, Cunliffe NA, Steele D, et al Effect of human rotavirus
vaccine on severe diarrhea
in African infants. N Engl J Med
25. Armah GE, Sow SO, Breiman RF, et al Efficacy of pentavalent rotavirus
vaccine against severe rotavirus
gastroenteritis in infants in developing countries in sub-Saharan Africa: a randomised, double-blind, placebo-controlled trial. Lancet
26. Moyo SJ, Gro N, Kirsti V, et al Prevalence of enteropathogenic viruses and molecular characterization of group A rotavirus
among children with diarrhea
in Dar es Salaam Tanzania
. BMC Public Health
27. Houpt E, Gratz J, Kosek M, et al Microbiologic methods utilized in the MAL-ED cohort study. Clin Infect Dis
2014; 59 (suppl 4):S225–S232.
28. Platts-Mills JA, Babji S, Bodhidatta L, et al Pathogen-specific burdens of community diarrhoea in developing countries: a multisite birth cohort study (MAL-ED). Lancet Glob Health
29. Moore SR, Lima NL, Soares AM, et al Prolonged episodes of acute diarrhea
reduce growth and increase risk of persistent diarrhea
in children. Gastroenterology
30. Opintan JA, Newman MJ, Ayeh-Kumi PF, et al Pediatric diarrhea
in southern Ghana: etiology and association with intestinal inflammation and malnutrition. Am J Trop Med Hyg
31. Ferdous F, Das SK, Ahmed S, et al Severity of diarrhea
and malnutrition among under five-year-old children in rural Bangladesh. Am J Trop Med Hyg
32. Qadri F, Saha A, Ahmed T, et al Disease burden due to enterotoxigenic Escherichia coli
in the first 2 years of life in an urban community in Bangladesh. Infect Immun
33. Black RE, Brown KH. Becker S Effects of diarrhea
associated with specific enteropathogens on the growth of children in rural Bangladesh. Pediatrics
34. Humphrey JH. Child undernutrition
, tropical enteropathy, toilets, and handwashing. Lancet
35. Syed S, Ali A, Duggan C. Environmental enteric dysfunction in children: a review. J Pediatr Gastroenterol Nutr
36. Chassaing B, Koren O, Carvalho FA, et al AIEC pathobiont instigates chronic colitis in susceptible hosts by altering microbiota composition. Gut
37. Cullender TC, Chassaing B, Janzon A, et al Innate and adaptive immunity interact to quench microbiome flagellar motility in the gut. Cell Host Microbe