https://orcid.org/0000-0002-9914-4471
Figures
Abstract
Background
Frequent enteric infections can damage the small intestine causing inflammation and malabsorption, leading to environmental enteric dysfunction. We aimed to characterize the association between intestinal inflammation and enteric pathogens among children in low- and middle-income countries (LMICs) who presented to care with diarrhea.
Methodology/principle findings
We conducted a cross-sectional analysis within the Enterics for Global Health - Shigella surveillance study at six LMIC sites: Bangladesh, Kenya, Malawi, Pakistan, Peru, and The Gambia. From August 2022 to July 2024, rectal swabs and whole stool samples were collected from 4,903 children with medically attended diarrhea aged 6–35 months (44.4% females, n = 2178/4903; mean age: 15.4 months ± 7.4 months) and were analyzed for Shigella and four fecal inflammatory biomarkers: hemoglobin, lipocalin-2, myeloperoxidase, and calprotectin via Enzyme Linked Immunosorbent Assays. Caregivers and clinicians demonstrated moderate accuracy in identifying blood in stool compared to fecal hemoglobin (area under the curve (AUC)=0.70). Among 10 pathogens evaluated, Shigella-attributable diarrhea had the highest concentrations of calprotectin, hemoglobin, and myeloperoxidase. Shigella culture-/PCR+ episodes had intermediate levels of inflammation between culture-/PCR- and culture+ episodes. In multivariable models restricted to Shigella episodes, dysentery was positively associated and vomiting was negatively associated with biomarker concentrations, with the strongest associations observed for hemoglobin (dysentery geometric mean ratio: 12.82 (95% CI: 7.21, 22.77) and vomiting geometric mean ratio: 0.45 (95% CI: 0.23, 0.86)). Age, sex, and both acute and chronic malnutrition were not associated with inflammatory biomarker concentrations.
Conclusions/significance
Hemoglobin appeared to be a more sensitive marker of blood in stool than visual observation. While Shigella was associated with heightened levels of all inflammatory biomarkers, hemoglobin was most strongly associated with Shigella, especially among attributable and culture positive episodes. The distinct clinical characteristics of Shigella were most closely associated with elevated hemoglobin concentrations.
Author summary
Diarrhea is common among children under five years of age in low- and middle-income countries (LMICs). Repeated diarrheal illnesses can damage the gut, leading to issues with growth and brain development. We examined fecal samples collected from 4,903 children with diarrhea who were enrolled in the Enterics for Global Health - Shigella surveillance study. We tested samples for diarrheal pathogens and measured four inflammatory biomarkers. We found that hemoglobin better identified blood in stool than visual observation of blood. Additionally, certain biomarkers (calprotectin, myeloperoxidase, and most notably hemoglobin) were higher when diarrhea was caused by bacteria such as Shigella than when diarrhea was caused by viruses. Also, we discovered that Shigella episodes identified using molecular diagnostics caused a similar illness as those identified by culture. These results support that Shigella diarrhea episodes identified by molecular diagnostics or culture are more inflammatory than other episodes of diarrhea and may require appropriate antibiotic treatment.
Citation: Horne B, Otieno Onyando B, Badji H, Mujahid W, Bhuiyan TR, Bakali M, et al. (2026) Fecal biomarkers to characterize intestinal inflammation among children with medically attended diarrhea across six low-resource settings: Findings from the Enterics for Global Health (EFGH) Shigella surveillance study. PLoS Negl Trop Dis 20(9): e0014320. https://doi.org/10.1371/journal.pntd.0014320
Editor: Zhe Wang, Shanghai Jiao Tong University, CHINA
Received: April 29, 2026; Accepted: August 31, 2026; Published: September 15, 2026
Copyright: © 2026 Horne et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The EFGH statistical analysis plan (https://clinicaltrials.gov/study/NCT06047821) and study protocol (https://academic.oup.com/ofid/issue/11/Supplement_1) were made publicly available. The de-identified and anonymised dataset is published online (https://search.vivli.org/doiLanding/studies/PR00011860/isLanding) and analytic code is available at https://github.com/sb2yb/efgh-biomarker-substudy.
Funding: This project is supported by the Bill & Melinda Gates Foundation (INV-016650 to PP, INV-031791 to PP, INV-036891 to FQ, INV-036892, INV-028721, INV-041730, INV-044311 to JAP-M, INV-044317 to ETRM). The funders did not play any role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Enteric infections are highly prevalent in low- and middle-income countries (LMICs) where poor sanitary conditions can result in fecally contaminated environments [1,2]. Frequent or persistent enteric infections, especially with Shigella, can lead to intestinal inflammation, increased intestinal permeability, and bacterial translocation of luminal microbes, all of which activate innate and acquired immune responses [1,3]. The combination of these precursor events can result in environmental enteric dysfunction (EED), a largely subclinical condition characterized by villus atrophy, crypt hyperplasia, and accumulation of immune cells in the intestinal mucosa [1,3]. EED often goes undetected as it rarely causes acute gastrointestinal symptoms that prompt medical care, however its consequences are substantial, including impaired nutrient absorption and increased vulnerability to adverse child health outcomes [2].
There is a need to characterize inflammation in children during etiology specific diarrhea episodes to understand how these episodes may contribute to EED. Enteric inflammation has been linked with growth faltering [4], micronutrient deficiencies [5], and impaired neurodevelopment in early childhood in some settings [6]. Importantly, not all enteric pathogens contribute equally to intestinal inflammation [7]. For instance, Shigella is an inflammatory bacterial pathogen [8] that causes epithelial shedding, crypt hyperplasia, and immune cell infiltration [9]. While EED is confirmed histologically via a biopsy of the small intestine, this is not feasible in resource constrained settings or ethically viable for children where biopsy is not medically indicated [10]. Therefore, non-invasive techniques are needed to characterize intestinal damage in children in low resource settings.
Fecal biomarkers are non-invasive and are used to characterize intestinal inflammation (hemoglobin, lipocalin-2, myeloperoxidase, calprotectin), neutrophil damage (myeloperoxidase, calprotectin), and enterocyte damage (lipocalin-2) [7]. Previous studies have demonstrated that fecal calprotectin [11,12], myeloperoxidase [12], and fecal occult blood [12] were higher in bacterial gastroenteritis compared to viral gastroenteritis. However, few studies have quantitatively evaluated whether these inflammatory markers can help distinguish the intestinal impact of specific pathogens or clarify the clinical significance of infections detected by culture versus molecular methods [7]. Prior work has demonstrated that Shigella culture-negative but quantitative Polymerase Chain Reaction (qPCR)-attributable cases represent true Shigella infections and should be managed as culture positive cases [13]. We propose inflammatory biomarkers could be used as independent predictors of etiology to evaluate whether culture- negative, but qPCR-attributable cases cause the same amount of intestinal damage as culture- positive cases.
This study aimed to characterize intestinal inflammation among children who presented to care with diarrhea in LMICs. We sought to 1) characterize the distribution and correlation of four inflammatory biomarkers by day of presentation and etiology, 2) compare detection of hemoglobin with reported blood in stool, 3) evaluate the association between Shigella detection by qPCR and culture with intestinal inflammation; and 4) assess how clinical risk factors for Shigella such as dysentery, fever, and vomiting are associated with biomarker levels. Studying these inflammatory responses could help identify a biomarker that could be used as a point-of-care diagnostic to guide clinicians on antibiotic treatment for shigellosis.
Methods
Ethical statement
This analysis involves data collected as part of the Enterics for Global Health (EFGH) protocol which was reviewed and approved by the Institutional Review Boards (IRB) and the Scientific and Ethical Review Committees of the EFGH collaborating institutions at the study sites. This sub-study was approved by the Emory IRB (Dec 1, 2022) and was deemed exempt from further IRB review and approval. Formal written informed consent was obtained from the parent(s) or primary caregiver(s) of each child who met eligibility criteria before any research activities were performed.
Study design and population
We conducted a cross-sectional nested sub-study on intestinal inflammation among children who presented to care with diarrhea within the Enterics for Global Health (EFGH) - Shigella surveillance study between June 2022 – August 2024. This sub-study was conducted in six countries: Dhaka, Bangladesh (5 enrollment sites), Siaya County, Kenya (6 enrollment sites), Blantyre City, Malawi (1 enrollment site), Karachi, Pakistan (8 enrollment sites), Iquitos, Peru (5 enrollment sites), and Basse, The Gambia (2 enrollment sites) [14]. Primarily, the EFGH study utilized a facility-based enrollment of diarrhea cases, a three month follow-up period, and home stool collection within 24 hours for children who did not produce whole stool at the facility during enrollment [14]. Diarrhea was defined as 3 or more loose or watery stools in the previous 24 hours [14,15] and the onset of illness was within 7 days of study enrollment. Dysentery was defined as caregiver or clinician reported blood in stool at screening, enrollment, discharge, or in the diarrhea diary. Follow-up data was not used for this sub-study.
Clinical data and specimen collection
We collected information on child demographics (e.g., sex, age), anthropometric measurements (mid-upper arm circumference, weight, and length/height), and illness (e.g., dysentery, fever, vomiting) during enrollment. Additionally, we asked caregivers to fill out a diarrhea diary to track illness symptoms (e.g., number of loose stools, dysentery, vomiting, fever) and medication usage for 14 days.
During enrollment, three rectal swabs were collected from each participant by trained study staff using Nylon flocked swabs (Copan Diagnostics). The first swab (FLOQswab) was placed in a dry tube (2-mL screw top vial compatible with a bead beater machine) and stored at −80°C for qPCR testing by TaqMan Array Card (TAC) for identification of enteric pathogens. The remaining two swabs were placed directly in Cary-Blair (CB) transport media (FecalSwab, 4C028S) and modified buffered glycerol saline (mBGS) (FecalSwab, mBGS made locally) respectively for Shigella species identification by culture [16]. During the study period, whole stool samples were collected at enrollment in sterile stool cup containers for the inflammatory biomarker sub-study, shipped to the laboratory, aliquoted, and stored at −80°C. To increase sub-study sample size, home stool collections were conducted within 24 hours of enrollment for participants who did not produce a stool during their enrollment visit.
Inflammatory biomarkers
Whole stool samples were tested for four inflammatory biomarkers: hemoglobin [17], lipocalin-2 [18], myeloperoxidase [19], and calprotectin [20] using commercial kits. Stored stool samples were thawed and analyzed for each biomarker according to the manufacturer’s instructions. Up to 80 samples were tested per plate with duplicates of standards and controls for each plate. The results were uploaded to a custom-built R-based shiny app, to derive a sigmoid 4-parameter curve from the standards to calculate concentrations for the samples. Samples with concentrations higher than the highest standard were further diluted and retested. Samples that remained higher than the highest standard after further dilution were imputed with the value of the highest standard multiplied by the highest dilution factor utilized for that sample. Samples that were lower than the lowest standard were assigned the value of the lowest standard divided by 2.
All biomarker assays were performed using commercially available kits, and calibration was conducted according to the manufacturer’s recommended protocols. Furthermore, analyte concentrations are reported in the units specified by the respective manufacturers. Concentrations cannot be directly compared across biomarkers as each assay captures a distinct component of the immune response.
Culture and qPCR for pathogens
Specimen testing was performed as part of the parent EFGH study. Methods are briefly described here. Rectal swabs were transported in two media: Cary-Blair (CB) and modified Buffer Glycerol Saline (mBGS) for culture on MacConkey and Xylose Lysine Decarboxylase agar. Each swab was cultured on two plates and incubated for 18–24 hours. Suspected Shigella colonies were subjected to biochemical testing on Triple Sugar Iron agar, Motility Indole Ornithine agar, and Urea medium, and incubated for 24 hours. Shigella was confirmed through latex agglutination targeting unique lipopolysaccharides found on the surface (antigens) of bacteria and the isolates were stored in Tryptic Soy Broth+15% glycerol [16]. Total nucleic acid was extracted from the dry swabs or whole stool if dry swabs were unavailable, for the detection of enteric pathogens using qPCR through the TAC platform. Pathogens were included in the analysis if the prevalence was ≥ 5% in the samples at Ct < 35 and could be attributed as the cause of diarrhea. Pathogens were attributed as the cause of diarrhea if detected at quantities higher (Ct lower) than the EFGH pathogen-specific quantity cutoff for attribution of diarrhea [21]. Episodes could be attributed to multiple pathogens across analyses except for the sensitivity analyses on the distribution of the 10 most prevalent enteric pathogens and risk factors (described below). Separate cut offs were used for whole stools and rectal swabs as previously [21]. For each pathogen, the Ct value was transformed into log10 copies (35-Ct value/3.322).
Data analysis
Correlation between biomarkers.
We evaluated the correlation between biomarkers (hemoglobin, lipocalin-2, myeloperoxidase, and calprotectin) using log10 concentrations overall and by individual sites. We also compared biomarker concentration by day of presentation to care relative to illness onset.
Comparison of hemoglobin and report of blood.
We compared the log10 concentrations of hemoglobin in all collected stools with and without reported visible blood (as indicated by the caregiver or clinician at the screening, enrollment or discharge visits, or noted in the diarrhea diary). A receiver operating characteristics (ROC) curve analysis was conducted to evaluate how well reported visible blood in stools corresponded with laboratory detected blood in stools (i.e., log10 concentration of hemoglobin) from the enrollment samples. An optimal cut-point was identified using Youden’s index to quantify the sensitivity and specificity of reported visible blood in stool to identify elevated hemoglobin concentrations.
Distribution of biomarkers by pathogens.
We characterized the distribution of biomarker log10 concentrations across all samples for each of the 10 analytic pathogens (Shigella, typical enteropathogenic Escherichia coli (tEPEC), Campylobacter jejuni/coli, heat-stable enterotoxigenic E. coli (ST-ETEC), Cryptosporidium, adenovirus 40/41, astrovirus, norovirus GII, rotavirus, and sapovirus). In sensitivity analyses, episodes were further restricted to 1) no co-attribution, 2) Shigella-attribution only, Shigella-attribution with co-bacterial, co-Cryptosporidium, and co-viral attribution, and 3) watery diarrhea episodes only. Furthermore, we calculated the mean and standard deviation of the log10 concentrations of the four biomarkers by pathogen. We also performed pairwise correlations among log10 biomarkers by pathogen. To evaluate the relationship between pathogen log10 copies and biomarker concentration, we ran a linear regression with generalized estimating equations (GEE) to account for repeated enrollments and adjusted for country to account for site differences. We also performed a sensitivity analysis using linear regression restricted to one enrollment per child where the first instance of Shigella in a stool sample was included or in the instance that no Shigella was present in any stool sample, the stool sample from the child’s first enrollment was included.
Distribution of biomarkers by Shigella diagnostic results.
We assessed the distribution of log10 biomarker concentrations for Shigella by three categories: Shigella culture-/Shigella qPCR-, Shigella culture-/Shigella qPCR + , and Shigella culture+ regardless of qPCR result. We further evaluated differences in biomarker concentrations by qPCR result where the diarrhea episode was attributed to Shigella (whole stool: Ct < 29.8, rectal swab: Ct < 29.5), where Shigella was detected but not attributed as the etiology (Ct ≥ 29.8 (whole stool) or ≥29.5 (rectal swab) but < 35), and where Shigella was not detected (Ct ≥ 35) [21].
Risk factors for shigellosis.
To determine whether demographic, anthropometric, or illness risk factors were associated with intestinal inflammation among Shigella-positive children, we ran a multivariable linear regression with GEE to account for repeated measures, adjusted for country to account for site differences and co-attributable pathogens. A sensitivity analysis was performed with restriction to Shigella-attribution only.
All analyses were performed via R software version 4.0.2.
Results
From August 2022 to July 2024, a total of 4,903 children aged 6–35 months from the six study sites with whole stools collected within 24 hours of enrollment were included in this study. The mean (SD) biomarker concentrations were 0.04 μg/g (SD: 167.67, n = 4903) for hemoglobin, 14.33 μg/g (SD: 10.33, n = 4903) for lipocalin-2, 1232.54 ng/mL (SD: 5.42, n = 4901) for myeloperoxidase, and 82,035.15 ng/mL (SD: 7.71, n = 4903) for calprotectin.
Correlation between biomarkers
Myeloperoxidase and calprotectin were moderately correlated (R = 0.70), whereas correlations among the other biomarkers were weak, ranging from R = 0.31 to R = 0.44 (Fig 1). Myeloperoxidase and calprotectin were moderately correlated across all sites (R = 0.70-0.74) except Bangladesh (R = 0.55) (Fig A in S1 Appendix). Hemoglobin and myeloperoxidase (R = 0.25-0.36) and hemoglobin and lipocalin-2 (R = 0.23-0.40) had the weakest correlations across sites.
Top row left to right: panel 1 is the correlation between myeloperoxidase and calprotectin, panel 2 is the correlation between hemoglobin and calprotectin, panel 3 is the correlation between lipocalin-2 and calprotectin. Bottom row left to right: panel 4 is the correlation between hemoglobin and myeloperoxidase, panel 5 is the correlation between lipocalin-2 and myeloperoxidase, and panel 6 is the correlation between lipocalin-2 and hemoglobin.
Comparison of hemoglobin and report of blood
Overall, 12.9% (n = 630/4894) of kids had blood reported in their stool prior to or on enrollment. Of the kids who did not have blood in their stool reported prior to or at enrollment, 2.4% developed bloody diarrhea after the enrollment period (n = 104/4264). Among all episodes, hemoglobin concentrations displayed a bimodal distribution when stratified by reported visible blood in stool, suggesting overall agreement between visible and ELISA detected blood, but with some notable discrepancies (Fig 2). There was more concordance among reported visible blood in stool and hemoglobin concentrations in Shigella-attributable diarrhea (Fig B in S1 Appendix) than Shigella detected but not attributed (Fig C in S1 Appendix) and Shigella not detected (Fig D in S1 Appendix). Overall, caregivers and clinicians were fairly accurate in predicting blood in a child’s stool (AUC = 0.70). At the optimal quantitative hemoglobin cut-point, sensitivity of reported blood was low, but specificity was high. Among children with elevated hemoglobin in stool, only 32% of caregivers and clinicians reported blood in stool (Table A in S1 Appendix). Additionally, among children without hemoglobin in the stool, 91% of caregivers and clinicians confirmed there was no blood in the stool.
Concentrations of biomarkers did not vary depending on day of presentation to care (hemoglobin: 4% higher concentration per day (95% CI: -6, 17); lipocalin-2: 5% higher concentration per day (95% CI: 0, 11); myeloperoxidase: 3% lower concentration per day (95% CI: -6, 1); calprotectin: 2% higher concentration per day (95% CI: -2, 6). In contrast, caregivers were 11% more likely to report visible blood in stool for each additional day of the diarrheal episode (RR: 1.11; 95% CI: 1.06–1.17), indicating that blood was more frequently observed later in the course of illness.
Distribution of biomarkers by pathogens
Of the 10 analytic pathogens, Shigella (22.7%) had the highest and astrovirus (1.3%) had the lowest prevalence of attributable episodes (Table B in S1 Appendix). Shigella had the highest concentrations across all biomarkers (hemoglobin mean: -0.33 log10 concentration, SD: 2.07; myeloperoxidase mean: 3.30 log10 concentration, SD: 0.63; and calprotectin mean: 5.21 log10 concentration, SD: 0.72), except lipocalin-2 (mean: 1.33 log10 concentration, SD: 0.91) where C. jejuni/coli was associated with the highest concentration (mean: 1.41 log10 concentration, SD: 0.63) (Fig 3; Table C in S1 Appendix). Results were similar when restricted to no co-attributions (Fig E in S1 Appendix). Shigella-attributable only episodes had slightly higher levels of inflammatory biomarkers when compared to Shigella-attributable episodes with co-bacterial, co-Cryptosporidium, and co-viral infections (Fig F in S1 Appendix). Results remained consistent when restricted to watery diarrhea (Fig G in S1 Appendix). Shigella was the only pathogen to have a positive association with all four biomarkers (Fig H in S1 Appendix). Of note, hemoglobin concentrations were markedly higher among Shigella-attributed episodes than non-attributed episodes (128% higher (95% CI: 102, 157), followed by calprotectin (26% higher; 95% CI: 20, 32), myeloperoxidase (20% higher; 95% CI: 15, 24), and lipocalin-2 (16% higher; 95% CI: 9, 23) (Table 1). The viral pathogens (astrovirus, norovirus GII, rotavirus, and sapovirus) had negative associations across all four biomarkers (Fig H in S1 Appendix, Table 1). A sensitivity analysis restricted to one enrollment per child was consistent with the main findings (Table D in S1 Appendix).
tEPEC = typical enteropathogenic E. coli; ST-ETEC = heat-stable enterotoxigenic E. coli. Co-attributions were permitted, indicating that more than one pathogen could have been the attributable cause of diarrhea. Therefore, a single child’s data could be represented across multiple pathogen bars.
Distribution of biomarkers by Shigella diagnostic results
Shigella culture-/qPCR+ biomarker concentrations were statistically different from Shigella culture-/qPCR- episodes and Shigella culture+ regardless of qPCR result (Fig 4). Shigella-attributed episodes (Fig 5) and Shigella culture-positive episodes (Fig I in S1 Appendix) had higher concentrations of inflammatory biomarkers than non-attributed and culture-negative episodes, respectively.
** = p< 0.01; *** = p< 0.001.
** = p< 0.01; *** = p< 0.001.
Risk factors for shigellosis
Among Shigella-attributed episodes, dysentery was the strongest risk factor for inflammation across biomarkers (significantly so for hemoglobin and calprotectin) but was most notable for hemoglobin where hemoglobin concentrations were 12 times higher among dysentery cases compared to non-dysentery cases (geometric mean ratio: 12.82; 95% CI: 7.21, 22.77; log10 mean difference: 1.11; 95% CI: 0.86, 1.36) (Fig 6; Table E in S1 Appendix). Additionally, having a fever was positively associated across inflammatory markers, significantly so with hemoglobin (geometric mean ratio: 2.25; 95% CI: 1.21, 4.18; log10 mean difference: 0.35; 95% CI: 0.08, 0.62). In contrast, vomiting was negatively associated with all inflammatory biomarkers, significantly so for hemoglobin (geometric mean ratio: 0.45; 95% CI: 0.23, 0.86; log10 mean difference: -0.35; 95% CI: -0.63, -0.07). There was no association between age, sex, and malnutrition with any inflammatory biomarker. Results were largely consistent when subset to Shigella-attribution only (Fig J in S1 Appendix).
MUAC = mid-upper arm circumference; LAZ/HAZ = length for age z-score/height for age z-score; WLZ/WHZ = weight for length z-score/weight for height z-score. Adjusted for country and co-attributable pathogens. Co-attributions were permitted, indicating that more than one pathogen could have been the attributable cause of diarrhea. Therefore, this plot represents data where Shigella is an attributable pathogen but may not be the only attributable pathogen for a diarrheal episode.
Discussion
We conducted a large multisite site study on pediatric diarrhea to characterize intestinal inflammation by etiology. Diarrhea episodes with bacterial etiologies had higher concentrations of all four inflammatory biomarkers (hemoglobin, myeloperoxidase, lipocalin-2, and calprotectin) compared to episodes with viral or parasitic etiologies. These findings are consistent with previous pediatric studies which showed that fecal calprotectin [11,22], myeloperoxidase [11], and fecal occult blood [11] were significantly higher in bacterial than viral gastroenteritis, with good diagnostic discrimination (AUCs ~ 0.74–0.79) [7].
We observed a moderate correlation between myeloperoxidase and calprotectin, while weak correlations were observed among the other pairs, indicating that hemoglobin, myeloperoxidase and lipocalin-2 capture distinct elements of the inflammatory response. Myeloperoxidase and calprotectin may capture a similar response of neutrophil damage [7], which has been previously documented [12].
Shigella‑attributable episodes had the highest concentrations of hemoglobin, myeloperoxidase, and calprotectin, while Campylobacter jejuni/coli had the highest concentrations of lipocalin-2. Among Shigella cases we observed the highest concentrations for each biomarker among the attributable cases, followed by the detected but not attributable cases, then the not detected cases. This would suggest that higher quantities of Shigella correspond to more inflammation. Furthermore, we found that culture-/qPCR+ Shigella episodes had elevated biomarker concentrations compared to culture-/qPCR- episodes, supporting that qPCR-attributed Shigella reflects clinically meaningful Shigella diarrhea rather than incidental carriage of Shigella during diarrhea. This would indicate that qPCR-attributable cases, even if culture negative, cause intestinal inflammation. This finding is consistent with prior work which demonstrated that Shigella culture-/qPCR+ cases represent true Shigella infections and should be managed clinically as culture+ cases would be managed [13].
In analyses restricted to Shigella-attributable episodes, dysentery was positively associated with all biomarkers, but the association was most pronounced for hemoglobin. This is unsurprising as hemoglobin is a marker of red blood cells [7]. Additionally, hemoglobin was also positively associated with fever and inversely associated with vomiting. These associations fit clinical expectations for Shigella diarrhea which is more likely to be associated with fever and bloody stools and less likely to be associated with vomiting [23] unless there is a co-etiology with a viral enteric pathogen [8]. Given hemoglobin’s close alignment with the presentation of Shigella, this could further underscore hemoglobin’s role as a biomarker for identifying diarrheal illnesses requiring antibiotic treatment.
The cross-sectional design of this study limited our ability to assess how biomarker levels changed over time. Children in LMICs tend to have high baseline levels of inflammation [24] so it is possible that we captured underlying enteropathy rather than inflammation due to the acute diarrhea illness for which the child was enrolled in the study. Nonetheless, our cross-sectional design was well suited for our aim of describing the effects of acute diarrheal illness. Future studies could implement longitudinal sampling particularly of fecal hemoglobin to help characterize dynamics of the biomarkers, define an optimal testing window, and identify diagnostic thresholds in this population. Stool samples were collected up to 24 hours after enrollment, at which time biomarker concentrations may have shifted. However, this window was necessary to allow for maximum stool collections across participants.
Elevated fecal hemoglobin was a distinguishing characteristic of shigellosis. Among the other biomarkers, it was most strongly associated with likely Shigella diarrhea, indicating that it could be used to identify treatable diarrhea where etiologic diagnostics are not available. Previous studies have found that point-of-care testing for fecal occult blood show reasonable specificity for identifying invasive bacterial pathogens such as Shigella [7,25], which reinforces the utility of stool blood–based signals as part of an antibiotic decision framework in resource limited settings. This work informed a formal evaluation of the performance of hemoglobin to improve clinical prediction algorithms for shigellosis in EFGH, which demonstrated improvements over both international guidelines and current clinical practice [26].
Supporting information
S1 Appendix. S1 Fig A. Pairwise correlations among log10 biomarker concentrations among children with medically attended diarrhea by the six sites in the EFGH sub-study.
Fig B. Density plot characterizing the distribution of log10 hemoglobin concentrations among Shigella-attributed cases by caregiver or clinician reported visible blood in stool at screening, enrollment or discharge visits or noted in the diarrhea diary among children with medically attended diarrhea across the six sites in the EFGH sub-study. Fig C. Density plot characterizing the distribution of log10 hemoglobin concentrations among Shigella-detected but not attributed cases by caregiver or clinician reported visible blood in stool at screening, enrollment or discharge visits or noted in the diarrhea diary among children with medically attended diarrhea across the six sites in the EFGH sub-study. Fig D. Density plot characterizing the distribution of log10 hemoglobin concentrations among non-Shigella cases by caregiver or clinician reported visible blood in stool at screening, enrollment or discharge visits or noted in the diarrhea diary among children with medically attended diarrhea across the six sites in the EFGH sub-study. Table A. Receiver Operating Characteristics (ROC) curve performance of reported blood in stool and log10 concentrations of hemoglobin by Shigella attribution and detection status among children with medically attended diarrhea across the six sites in the EFGH sub-study. Table B. Pathogen prevalence based on qPCR detection (Ct < 35) and qPCR attribution of the 26 enteric pathogens among children with medically attended diarrhea across the six sites in the EFGH sub-study. Table C. Associations of individual biomarkers with pathogen specific diarrhea etiologies for the 10 most prevalent enteric pathogens considered attributable by qPCR among children with medically attended diarrhea across the six sites in the EFGH sub-study (co-attribution permitted). Fig E. Boxplots characterizing the distribution of log10 biomarker concentrations by the 10 most prevalent enteric pathogens among children with medically attended diarrhea across the six sites in the EFGH sub-study, excluding episodes with more than one attributable pathogen; i.e., co-attribution not permitted. Fig F. Boxplots characterizing the distribution of log10 biomarker concentrations with Shigella attribution alone or with bacterial, Cryptosporidium, or viral co-attribution among children with medically attended diarrhea across the six sites in the EFGH sub-study (co-attribution permitted). Fig G. Boxplots characterizing the distribution of log10 biomarker concentrations by the 10 most prevalent enteric pathogens among children with medically attended watery diarrhea across the six sites in the EFGH sub-study (co-attribution permitted). Fig H. Pairwise correlations among the log10 biomarker concentrations among children with medically attended diarrhea by the six sites in the EFGH sub-study. Table D. Sensitivity analysis of the associations of individual biomarkers with pathogen specific diarrhea etiologies among children with medically attended diarrhea across the six sites in the EFGH sub-study, including only one enrollment per child (multiple enrollments removed). Fig I. Boxplots characterizing the distribution of log10 biomarker concentrations by Shigella culture positivity status among diarrhea cases among children with medically attended diarrhea across the six sites in the EFGH sub-study. Table E. Associations of individual biomarkers with risk factors for shigellosis among children with medically attended diarrhea across the six sites in the EFGH sub-study (co-attribution permitted). Fig J. Associations of individual log10 biomarkers concentrations with risk factors for shigellosis among children with medically attended diarrhea across the six sites in the EFGH sub-study (Shigella-attribution only).
https://doi.org/10.1371/journal.pntd.0014320.s001
(DOCX)
References
- 1. Korpe PS, Petri WA. Environmental enteropathy: critical implications of a poorly understood condition. Trends Mol Med. 2012;18(6):328–36. pmid:22633998
- 2. Watanabe K, Petri WA. Environmental enteropathy: elusive but significant subclinical abnormalities in developing countries. eBioMedicine. 2016;10:25–32.
- 3. Syed S, Ali A, Duggan C. Environmental enteric dysfunction in children. J Pediatr Gastroenterol Nutr. 2016;63(1):6–14. pmid:26974416
- 4. Kosek M, Haque R, Lima A, Babji S, Shrestha S, Qureshi S, et al. Fecal markers of intestinal inflammation and permeability associated with the subsequent acquisition of linear growth deficits in infants. Am J Trop Med Hyg. 2013;88(2):390–6. pmid:23185075
- 5. Kelly P. The contribution of environmental enteropathy to the global problem of micronutrient deficiency. Proc Nutr Soc. 2021;80(3):303–10. pmid:33663621
- 6. Ho GG, Achando BS, Ali S, Hemlock C, Pitchik HO, Stewart CP, et al. Biomarkers of environmental enteric dysfunction and neurodevelopmental outcomes among children in rural Bangladesh and Kenya: A prospective cohort study. Am J Clin Nutr. 2025;122(2):474–87. pmid:40480608
- 7. Babb C, Badji H, Bhuiyan MTR, Cornick J, Qureshi S, Sonye C, et al. Evaluation of fecal inflammatory biomarkers to identify bacterial diarrhea episodes: systematic review and protocol for the Enterics for Global Health Shigella surveillance study. Open Forum Infect Dis. 2024;11(Suppl 1):S65–75. pmid:38532957
- 8. Rogawski McQuade ET, Shaheen F, Kabir F, Rizvi A, Platts-Mills JA, Aziz F, et al. Epidemiology of Shigella infections and diarrhea in the first two years of life using culture-independent diagnostics in 8 low-resource settings. PLoS Negl Trop Dis. 2020;14(8):e0008536. pmid:32804926
- 9.
Hale TL, Keusch GT. Shigella. In: Baron S, editor. Medical Microbiology. 4th ed. Galveston (TX): University of Texas Medical Branch at Galveston. 1996.
- 10. Tickell KD, Atlas HE, Walson JL. Environmental enteric dysfunction: a review of potential mechanisms, consequences and management strategies. BMC Med. 2019;17(1):181. pmid:31760941
- 11. Park Y, Son M, Jekarl DW, Choi HY, Kim SY, Lee S. Clinical significance of inflammatory biomarkers in acute pediatric diarrhea. Pediatr Gastroenterol Hepatol Nutr. 2019;22(4):369–76. pmid:31338312
- 12. Prata M de MG, Havt A, Bolick DT, Pinkerton R, Lima A, Guerrant RL. Comparisons between myeloperoxidase, lactoferrin, calprotectin and lipocalin-2, as fecal biomarkers of intestinal inflammation in malnourished children. J Transl Sci. 2016;2(2):134–9. pmid:27746954
- 13. Pavlinac PB, Platts-Mills JA, Tickell KD, Liu J, Juma J, Kabir F, et al. The clinical presentation of culture-positive and culture-negative, quantitative polymerase chain reaction (qPCR)-Attributable Shigellosis in the global enteric multicenter study and derivation of a Shigella severity score: implications for pediatric shigella vaccine trials. Clin Infect Dis. 2021;73(3):e569–79. pmid:33044509
- 14. Atlas HE, Conteh B, Islam MT, Jere KC, Omore R, Sanogo D, et al. Diarrhea case surveillance in the Enterics for Global Health Shigella surveillance study: epidemiologic methods. Open Forum Infect Dis. 2024;11:S6–16.
- 15.
Diarrhoeal disease. Accessed 2026 July 17. https://www.who.int/news-room/fact-sheets/detail/diarrhoeal-disease
- 16. Horne B, Badji H, Bhuiyan MTR, Romaina Cachique L, Cornick J, Hotwani A, et al. Microbiological methods used in the enterics for global health Shigella surveillance study. Open Forum Infect Dis. 2024;11(Suppl 1):S25–33. pmid:38532949
- 17.
Immundiagnostik AG. Hemoglobin ELISA. [cited 14 July 2023]. Available from: https://www.immundiagnostik.com/en/portfolio/testkits/k-7816d
- 18.
Eagle Biosciences. NGAL (stool) ELISA assay kit: package insert. https://eaglebio.com/wp-content/uploads/2014/06/NGL35-K01_NGAL_Stool_ELISA_Assay_Kit_Package_Insert_v7.1_07.2023-1.pdf. 2023.
- 19.
Immundiagnostik AG. MPO ELISA. [cited 14 July 2023]. Available from: https://www.immundiagnostik.com/en/portfolio/testkits/k-6630
- 20.
Immundiagnostik AG. Calprotectin ELISA. [cited 14 July 2023]. Available from: https://www.immundiagnostik.com/en/portfolio/testkits/kr6927
- 21. Yousafzai MT, Cornick J, Penataro Yori P, Hossain MJ, Keita AM, Atlas HE, et al. The incidence and antimicrobial resistance of Shigella-attributable diarrhoea in young children in low-income and middle-income countries from the multicountry Enterics for Global Health (EFGH) Shigella Surveillance Study: a prospective, facility-based hybrid surveillance study. Lancet Glob Health. 2026;14(5):e749–61. pmid:41831467
- 22. Rehan M, MacBain E, Shajib MS, Mokomane M, Lechiile K, Arscott-Mills T, et al. Correlation of fecal calprotectin levels with the detection of treatable enteric pathogens in children with severe acute diarrheal disease in Botswana. PLoS One. 2025;20(8):e0328764. pmid:40839595
- 23. Platts-Mills JA, Liu J, Rogawski ET, Kabir F, Lertsethtakarn P, Siguas M, et al. Use of quantitative molecular diagnostic methods to assess the aetiology, burden, and clinical characteristics of diarrhoea in children in low-resource settings: a reanalysis of the MAL-ED cohort study. Lancet Glob Health. 2018;6(12):e1309–18. pmid:30287127
- 24. Lima AAM, Leite ÁM, Di Moura A, Lima NL, Soares AM, Abreu CB. Determinant variables, enteric pathogen burden, gut function and immune-related inflammatory biomarkers associated with childhood malnutrition: a prospective case–control study in northeastern Brazil. The Pediatric Infectious Disease Journal. 2017;36:1177.
- 25. Kwasi DA, Adewole PD, Akinlabi OC, Ekpo SE, Okeke IN. Evaluation of fecal occult blood testing for rapid diagnosis of invasive diarrhea in young children. PLOS Glob Public Health. 2023;3(7):e0001629. pmid:37471343
- 26. Ogwel B, Khanam F, Badji H, Charles M, Qureshi S, Horne B, et al. Performance of fecal inflammatory biomarkers to identify watery shigellosis: findings from the Enterics for Global Health (EFGH) Shigella surveillance study. PLoS Negl Trop Dis. 2026;20(6):e0014025. pmid:42224289