Bile acid dysmetabolism in Bangladeshi infants associated with poor linear growth, enteric inflammation, and small intestine bacterial overgrowth

Abstract

BACKGROUND

Environmental enteric dysfunction (EED) is a subclinical condition caused by fecal-oral contamination leading to enteric inflammation and dysbiosis. Bile acids serve to facilitate lipid digestion and absorption, regulate metabolic pathways associated with childhood growth and inflammation, and may be affected by EED.

AIM

To investigate bile acid metabolism in Bangladeshi children with EED and its association with growth impairment.

METHODS

We conducted a cross-sectional study of 100 Bangladeshi infants (aged 6-9 months) and quantified serum and fecal bile acids using LC-MS/MS. We compared profiles to a control group of 6 American children (6-12 months) and 80 older Bangladeshi children (aged 2 years).

RESULTS

Bangladeshi infants had higher levels of plasma unconjugated primary (65.23% vs 44.25%, P = 0.003) and sulfated primary bile acids (12.98% vs < 0.001%, P = 0.01), with lower primary conjugated bile acids (0.69% vs 2.74%, P ≤ 0.001) compared to American children. Stool unconjugated primary bile acids were inversely associated with weight-for-age [regression coefficient (β) = -0.01, P = 0.01] and height-for-age Z scores (β = -0.01, P = 0.03). Conjugated secondary bile acids were inversely associated with small intestine bacterial overgrowth (β = -1096.68, P = 0.05). Fecal myeloperoxidase was associated with sulfated secondary bile acids (β = -0.40, P = 0.04). Compared to 2-year-old children, the Bangladeshi infant’s serum had higher levels of unconjugated primary bile acids (65.23% vs 9.20%, P ≤ 0.001) and lower levels of primary conjugated bile acids (0.69% vs 80.38%, P ≤ 0.001).

CONCLUSION

Our data suggests an age-dependent defect in conjugation of primary bile acids in Bangladeshi children with compensatory hydrophilic shunting. Additionally, bile acid profiles are associated with intestinal overgrowth.

Key Words: Bile acid metabolism; Malnutrition; Environmental enteric dysfunction; Small intestinal bacterial overgrowth; Bangladesh

Core Tip: This study identifies a previously unrecognized defect in bile acid conjugation among Bangladeshi infants with environmental enteric dysfunction, linking altered bile acid metabolism to poor growth and intestinal inflammation. Elevated unconjugated primary bile acids were strongly associated with anthropometry. Further, the findings highlight a possible age-related delay in maturation of bile acid conjugation pathways in impoverished children. This study may provide the initial insights for exploring novel therapeutic targets through bile acid pathways for treating malnourished children worldwide.

INTRODUCTION

Twenty-one percent of deaths in children under the age of five can be attributed to malnutrition[1]. Pediatric malnutrition in low and middle-income countries is multifactorial with contribution from food insecurity, inadequate access to clean water, and unsanitary living conditions resulting in chronic enteric pathogen exposure. Chronic carriage of enteric pathogens leads to environmental enteric dysfunction (EED), a sub-clinical condition characterized by chronic gut inflammation, small intestinal injury, and small intestinal dysbiosis[2]. EED has been associated with deficient nutrient absorption and systemic inflammation which leads to linear growth shortfalls and poor neurodevelopmental outcomes[3]. Constant exposure to fecal-oral contamination leads to enteric inflammation and subsequent dysbiosis which, in turn, leads to reduced barrier function, villous blunting, and a decreased crypt-to-villus ratio[4]. These histological changes recovered when subjects were removed from areas of poor sanitation and hygiene[4]. While intestinal biopsies are the gold standard for diagnosing EED, endoscopy is invasive and therefore various biomarkers of intestinal and systemic inflammation have been used as proxy measures of EED burden[5]. Childhood EED has been associated with growth stunting, decreased weight, and poor neurodevelopmental outcomes[4].

Among children with EED, small intestine bacterial overgrowth (SIBO) is common. SIBO was present in 11% of 18 weeks old Bangladeshi infants and the prevalence increased to 30%-40% in 1 to 2-year-olds[6]. In this population, SIBO was associated with markers of intestinal inflammation[7]. SIBO in the low-income country setting is also associated with growth stunting and language delay, independently of inflammation[6,8]. Despite the association of SIBO and poor growth being noted in several studies across several continents, the nature of this association and the pathogenesis of SIBO remains unclear. One study did, however, note that the SIBO microbiome led to decreased lipid absorption when transplanted from patients into an in vitro murine intestinal m-IcCl2 cell line[8].

EED associated gut dysbiosis has also been linked to abnormal bile acid patterns[9]. Bile acids act as detergents in the intestine to facilitate lipid digestion and absorption. They also serve as signaling molecules that regulate numerous metabolic pathways associated with childhood growth[10]. Furthermore, bile acids have been shown to play a role in host susceptibility to inflammation and serve as sensitive markers to hepatic and cholestatic function[11]. Bile acid metabolism defects can lead to fat-soluble vitamin deficiency, malnutrition, neurologic impairments, and stunted growth[12]. Lack of bile acid conjugation has even been linked to the action of the bacteria proliferating in the upper small intestine[13]. There is a paucity of literature investigating the role of bile acid metabolism in low- and middle-income country children with EED and malnutrition although preliminary evidence suggests altered bile acid metabolism in this population[9,11,14-16]. A study looking at children in rural Malawi reported that children with EED displayed altered bile acid metabolism from a young age, with a higher proportion of bile acids conjugated with taurine instead of glycine. Total serum bile acids were also approximately 12% lower in children with EED compared with children without EED, which may be reflective of impaired reuptake due to injury in the ileum[9]. The same study also inversely correlated levels of primary unconjugated bile acids with age in children with EED[9]. Another study in undernourished Pakistani children with EED found elevated levels of serum bile acids when compared to well-nourished local American children, with over 70% of the undernourished infants exceeding the upper limit of the normal range[11]. Higher proportions of glycocholic acid (GCA) were also linked to more severe EED changes, with correlations seen in both plasma (r_s = 0.324, P = 0.02) and duodenal aspirates (r_s = 0.307, P = 0.06) from children with refractory wasting who had endoscopies[11].

Based on these previous studies, we hypothesized that stunted children with EED and SIBO have altered bile acid metabolism. To test this hypothesis, we conducted a cross-sectional analysis of Bangladeshi children and compared them to American controls, investigating bile acid patterns in both the serum and stool.

MATERIALS AND METHODS

We conducted a cross-sectional analysis of Bangladeshi toddlers investigating the association between SIBO, EED biomarkers, and bile acid profiles in both serum and stool. We enrolled 100 toddlers from the urban neighborhood of Mirpur in Dhaka, Bangladesh. Samples were obtained as part of a prospective efficacy study in Bangladeshi infants. All specimens were collected at baseline prior to randomization. The present work represents a secondary analysis of these existing samples from a previous study in malnourished Bangladeshi children (ClinicalTrials.gov identifier: NCT03263871). Samples were taken from the pre-intervention study visit. Enrollment was from October 2017 through July 2018. The majority of the homes in Mirpur are mud brick construction. Crowding is common with a mean of 5 people living in 1.5 rooms per dwelling. Uncovered sewers flow throughout the neighborhood with municipal water lines often running through these sewer channels. Due to the location of our study clinic, subjects tended to come from the lowest socioeconomic strata of Mirpur. Enrolled children were between age 6-9 months old with no known chronic medical problems other than mild wasting as defined by a weight-for-age Z (WAZ) score between -1 SD and -3 SD. Exclusion criteria also included milk intolerance, abnormal liver or kidney function, and exclusive breast feeding at time of enrollment.

Blood was collected by a pediatric phlebotomist in our study clinic. The stool was collected in our study clinic by field assistants. Maintaining proper cold chain, samples were transported to our laboratory within 4 hours where they were aliquoted and placed in -80 °C. One aliquot was removed for batched biomarker analysis at the completion of the study. A separate aliquot was later shipped to the United States on dry ice where samples were again stored in -80 °C until they were removed for analysis. The cold chain was monitored throughout the shipping process. We also conducted both a 2-hour dual sugar urinary lactulose-mannitol test by LC-MSMS and glucose-hydrogen breath test on subjects using methods on which we have previously published[7,17]. Pediatric urine collection bags were attached to the patients and were allowed to return to regular diet 30 minutes after LM test solution ingestion. Two mL of urine was collected at the 2-and-5-hour marks. The samples were then analyzed by the HPLC-MSMS system[17]. Glucose-hydrogen breath testing was done by QuinTron BreathTracker SC gas chromatography which collected breath via an age-appropriate anesthesia mask attached at 20-minute intervals for 3 hours. Patients fasted for 3 hours prior to and throughout testing but were allowed water. Children younger than 12 months fasted for 2 hours[7]. SIBO area under the curve (AUC) (i.e., area under the breath hydrogen curve) was calculated by summing the trapezoidal area under the glucose-hydrogen curve using methods previously published[6]. These two tests were conducted within one week of each other but not on the same day. Stool and serum were collected at the same time as glucose-hydrogen breath testing. Biomarkers were tested by commercially available ELISA and included stool regenerating family member 1 beta (Reg 1B) (TechLab, Inc. Blacksburg, VA, United States), stool myeloperoxidase (MPO) (ALPCO. Salem, NH, United States), serum C-reactive protein (CRP) (ALPCO. Salem, NH, United States), and serum soluble CD14 (sCD14) (R&D Systems, Minneapolis, MN, United States). Anthropometry was measured using calibrated infant scales and an infant measuring board by staff trained in the procedure on the day of breath testing. WAZ, height-for-age Z scores (HAZ), weight-for-height Z scores (WHZ), and body mass index-for-age Z scores (BAZ) were calculated using World Health Organization software (WHOAnthro).

We also obtained discarded serum samples collected and stored in EDTA from the clinical laboratories at Virginia Commonwealth University Hospital in Richmond, VA, United States. Samples were collected to be used for comparison, serving as a high-income well-nourished cohort. They were collected for clinical reasons unrelated to this study and screened prior to being discarded from the clinical lab. Samples were selected based on age with only samples from 6-12-month-old screened. Exclusion criteria included acute or chronic malnutrition, any known gastrointestinal diseases, or known metabolic disorders. Only serum was collected as paired stool samples were unavailable. Samples were collected for clinical purposes and sent within minutes of collection to the clinical laboratory where they were processed within 1 hour and stored at 4 °C for 3 days. Samples stored longer than 3 days are generally discarded. Once a sample was identified as marked for discard, having sufficient volume for bile acid testing, and from the appropriate age range, the electronic medical record was reviewed by our team (Donowitz JR) to ensure the children had no known gastrointestinal or metabolic diagnoses associated with altered bile acid metabolism.

Serum from both Bangladeshi and American children, as well as stool from Bangladeshi children underwent bile acid and 7α-hydroxy-4-cholesten-3-one (C4) profiling. C4 is a bile acid precursor, serving as an indicator of overall bile acid production. The serum samples were processed and the composition and levels of individual bile acid metabolism were measured using a Shimadzu liquid chromatography/tandem mass spectrometric 8600 system as described previously[18]. Ten isotope-labeled compounds, each sharing the same structure as the corresponding bile acid analytes, were used as internal standards. The same volume and concentration of internal standards were added to each sample for quality control (QC) and quantitative bile acid analysis. Internal standards help monitor extraction efficiency and correct for instrument variability. The assay was validated using calibration curves and QC samples prepared at various concentrations, with QC samples analyzed in every batch to ensure consistent instrument performance and accurate quantification across all runs.

We first analyzed the composition of bile acids in both the serum and stool samples from both cohorts. Individual bile acid levels were expressed as percentages of total bile acids detected, excluding C4 which was analyzed as a continuous variable. If a bile acid was undetectable, it was assigned a value of 0. If there were no detectable levels, bile acids were valued at a 0. Bile acids were grouped based on the bile acid metabolism pathway as follows: CDCA and CA were categorized as unconjugated primary bile acids, TCA, GCA, TCDCA, and GCDCA were categorized as conjugated primary bile acids, CDCA-3-S and CA-3-S were categorized as sulfated primary bile acids, UDCA, DCA, and LCA were categorized as unconjugated secondary bile acids, TUDCA, GUDCA, MDCA, GDCA, TLCA, and GLCA were categorized as conjugated secondary bile acids, 7keto-DCA, 7keto-LCA, isoDCA, isoLCA, allo-isoLCA, and 3keto-LCA were categorized as secondary bile acid metabolites, LCA-3-S and UDCA-3-S were categorized as sulfated secondary bile acids, and TwMCA, TaMCA, TbMCA, GbMCA, THDCA, GHCA, GHDCA, wMCA, aMCA, bMCA, HCA, and HDCA were categorized as muricholates. Groupings were made based on known metabolic pathways[19,20]. Bile acids within each group were summed to calculate the total percentage detected for each group. Two-sample t-tests were used to determine if there were differences in the percentage of these bile acid groups in the total bile acid pool between the serum samples of the Bangladeshi and American cohorts and between serum and stool samples of the Bangladeshi children. EED biomarkers including Reg 1B, MPO, CRP, sCD14, LM ratio, and SIBO AUC were separately regressed on categories of bile acids that were identified as significantly different between the American and Bangladeshi populations using univariate linear regression. Finally, we created similar linear regression models to predict anthropometric variables including WAZ, BAZ, LAZ, and WHZ scores. As anthropometry was our primary outcome of interest, we next created multivariable regression models with backward stepwise regression and included unconjugated primary bile acids, female, age (in days), Reg 1B, MPO, CRP, sCD14, LM ratio, and SIBO AUC as predictors.

As an exploratory endeavor, we then conducted a secondary analysis of data that was collected as part of two other cohorts in 2-year-old Bangladeshi children from the same neighborhood in Mirpur, Dhaka. SIBO data, collected using identical methodology to that detailed above, was also available on these samples. Eighty paired serum and stool samples collected between May 2011 and March 2016 had bile acid analysis performed by liquid chromatography-mass spectroscopy using the Biocrates AbsoluteIDQ p180 Kit and high-performance liquid chromatography column per manufacturer's protocols (Biocrates, Inc Aliso Viejo, CA, United States). Bile acids were converted into percentages of total bile acids detected. If there were no detectable levels, bile acids were valued at a 0. Then bile acids were grouped based on the bile acid metabolism pathway as follows: CDCA and CA were categorized as unconjugated primary bile acids, TCA, GCA, TCDCA, and GCDCA were categorized as conjugated primary bile acids, UDCA, DCA, and LCA were categorized as unconjugated secondary bile acids, TUDCA, GUDCA, TDCA, GDCA, TLCA, and GLCA were categorized as conjugated secondary bile acids, and TaMCA, TbMCA, aMCA, bMCA, oMCA, and HDCA were categorized as muricholates. Linear regression models were completed between these bile acid categories and SIBO AUC data. Finally, two-sample t-tests were conducted to determine if there were differences between bile acid groups of the 2-year-old cohort and the American children, as well as between the 2-year-old cohort and the original Bangladeshi cohort of 6-9-month-old children.

The collection and analysis of Bangladeshi samples was approved by the Ethics and Research Review Committees at the ICDDR,B and the Institutional Review Board at the University of Virginia. Reliance agreements between the University of Virginia and the ICDDR,B with Virginia Commonwealth University were also established for this study. Informed consent for participation was received by both parents of enrolled children. For the American control group, the study was approved by the Institutional Review Board at Virginia Commonwealth University. A waiver of consent was obtained to use discarded samples.

RESULTS

The cohort of 100 6-9-month-old Bangladeshi consisted of 51 females and 49 males. The average age was 229 days (range: 185-265 days). The average (and SD) of WHZ was -0.75 SD (± 0.69 SD), HAZ was -1.60 SD (± 0.76 SD), WAZ was -1.55 SD (± 0.47 SD), and BAZ was -0.83 SD (± 0.67 SD) (Table 1).

Table 1 Demographic data of Bangladeshi infant cohort.

Characteristics	Value
Female, n	51	51%
Age (days)	229	21
Weight-for-height Z-score	-0.75	0.69
Height-for-age Z-score	-1.60	0.76
Weight-for-age Z-score	-1.55	0.47
BMI-for-age Z-score	-0.83	0.67

The table presents demographic characteristics of the study population of 100 Bangladeshi infants (age 6-9 months), including sex distribution, age, and anthropometric Z-scores. Sex is provided as both count and percentage. Continuous variables are displayed as means with SD. BMI: Body mass index.

In the serum samples, the Bangladeshi children averaged 216540 nmol/L unconjugated primary bile acids, 1537 nmol/L conjugated primary bile acids, 23185 nmol/L primary sulfated bile acids, 4413 nmol/L secondary bile acids, 93 nmol/L secondary conjugated bile acids, 774 nmol/L secondary sulfated bile acids, 23998 nmol/L secondary bile acid derivatives, and 14033 nmol/L muricholates. They had a total average bile acid serum sample concentration of 11956 nmol/L. In comparison, the American group has an average serum distribution of 1097 nmol/L unconjugated primary bile acids, 74 nmol/L conjugated primary bile acids, 0 primary sulfated bile acids, 0 secondary bile acids, 57 nmol/L secondary conjugated bile acids, 0 secondary sulfated bile acids, 20 nmol/L secondary bile acid derivatives, and 695 nmol/L muricholates. They had a total average bile acid serum sample concentration of 2570 nmol/L. For analysis purposes between the two cohorts, the percentage distribution of each category was calculated and then compared. C4 Levels were also obtained in both groups, with Bangladeshi children having an average of 28 nmol/L and American children of 0 nmol/L (Supplementary Table 1).

As compared to American children, Bangladeshi children’s serum had higher levels of primary unconjugated bile acids (65.23% vs 44.25%, P = 0.003), primary sulfated bile acids (12.98% vs < 0.001%, P = 0.01), and secondary bile acids derivatives (11.35% vs 0.97%, P = 0.03). Bangladeshi children had lower percentages of muricholates (6.16% vs 27.02%, P ≤ 0.001), primary conjugated bile acids (0.69% vs 2.74%, P ≤ 0.001), and secondary conjugated bile acids (0.05% vs 3.31%, P ≤ 0.001) compared to American children. There was no significant difference in percentage of unconjugated secondary bile acids, percentage of secondary sulfated bile acids, or total bile acids between the two groups (Figure 1). There was no significant difference between percentage of muricholates, unconjugated primary bile acids, conjugated primary bile acids, sulfated primary bile acids, unconjugated secondary bile acids, conjugated secondary bile acids, sulfated secondary bile acids, or secondary bile acid metabolite levels in the stool and serum of the 6-9-month-old Bangladeshi children (Figure 2).

Figure 1 Percent distributions comparing the younger Bangladeshi infants' serum bile acids and C4 Levels to the American children are displayed with mean and standard deviation. Secondary bile acid derivatives include 7-Keto DCA, 7-Keto LCA, 3-Keto LCA, isoDCA, isoLCA, and allo isoLCA. Bangladeshi infants demonstrate a deficiency in primary bile acid conjugation as compared to American children with compensatory increase in hydrophilic primary sulfated bile acids. C4 is also increased.

Figure 2 Stool and plasma bile acid pools were compared in Bangladeshi 6-9-month-old. Plasma distribution is shown as shaded box-and-whisker plots while stool distribution is illustrated with clear box-and-whisker plots. The central line in each box represents the median, while the box spans the interquartile range (IQR). Whiskers extend to 1.5 times the IQR, with individual points representing outliers. There were no significant differences between the groups in regard to bile acid percent distributions.

In the stool of Bangladeshi children, conjugated secondary bile acids had an inverse relationship with SIBO AUC [regression coefficient (β) = -1096.68, P = 0.05], while MPO had an inverse association with sulfated secondary bile acids (β = -0.40, P = 0.04; Figure 3). No other groups of bile acids were significantly associated with inflammatory biomarkers. Unconjugated primary bile acids were associated with weight-for-age (β = -0.27, P = 0.01) and HAZ (β = -0.01, P = 0.03; Figure 4). There was no significant association against weight-for-height or body mass index (BMI)-for-age. Further, C4 concentrations were associated with HAZ (β = 0.65, P = 0.04; Figure 5). There was no significant association with C4 with weight-for-age, weight-for-height, or BMI-for-age. No other bile acids were significantly associated with weight-for-age, height-for age, weight-for-height, or BMI-for-age.

Figure 3 Compared in the Bangladeshi infants. A: Log transformed secondary conjugated bile acid concentrations were plotted against small intestine bacterial overgrowth area under the curve in the Bangladeshi infants. Small intestine bacterial overgrowth area under the curve is associated with a decrease in secondary conjugated bile acids (β = -1096.68, P = 0.05); B: Log transformed myeloperoxidase were regressed on secondary sulfated bile acid concentrations in the Bangladeshi infants. As myeloperoxidase increases, there is a decrease in the concentration of secondary sulfated bile acids (β = -0.40, P = 0.04). AUC: Area under the curve; SIBO: Small intestine bacterial overgrowth; MPO: Myeloperoxidase.

Figure 4 Compared in the Bangladeshi infants. A: Primary unconjugated bile acid concentrations are plotted against weight-for-age Z scores in the Bangladeshi infants. There is a significant correlation with drop in weight-for-age Z score as primary unconjugated bile acid concentrations increase (β = -0.01, P = 0.01); B: Primary unconjugated bile acid concentrations are plotted against height-for-age Z scores in the Bangladeshi infants. There is a significant correlation with drop in height-for-age Z score as primary unconjugated bile acid concentrations increase (β = -0.01, P = 0.03). HAZ: Height-for-age Z scores; WAZ: Weight-for-height Z scores.

Figure 5 Log transformed C4 concentrations were regressed on height-for-age Z scores in the Bangladeshi infants. Regression shows that there is an increase in height-for-age Z scores with increased C4 Levels (β = 0.65, P = 0.04).

In our multivariable model with HAZ as outcome, the model only retained unconjugated primary bile acids (β = -0.01, P = 0.05), age (β = -0.01, P = 0.08), female (β = 0.28, P = 0.05), and Reg 1B (β = -0.00, P = 0.02) as significant predictors. In our multivariable model with WAZ as outcome, the model only retained unconjugated primary bile acids (β = -0.01, P = 0.01) and CRP (β = -0.00, P = 0.01) as significant predictors.

The plasma samples of the 6-9-month-old Bangladeshi children were compared to the plasma samples of 80 2-year-old Bangladeshi children. As compared to the 2-year-old children, the 6-9-month-old Bangladeshi children’s serum had higher levels of unconjugated primary bile acids (65.23% vs 9.20%, P ≤ 0.001) and muricholates (6.16% vs 0.26%, P ≤ 0.001). They had lower percentages of primary conjugated bile acids (0.69% vs 80.38%, P ≤ 0.001), and conjugated secondary bile acids (0.05% vs 7.24%, P ≤ 0.001) compared to the 2-year-old children. There was no significant difference in percentage of unconjugated secondary bile acids between the two groups (Figure 6). There were no significant associations between the older children’s bile acid group percentages and SIBO AUC.

Figure 6 Box-and-whiskers plot showing percent distribution of bile acids in plasma between the Bangladeshi infants (aged 6-9 months) and the Bangladeshi children (aged 2-years-old). The younger Bangladeshi infants' distributions are shown in blue and the older Bangladeshi children are shown in orange. If P value was not denoted, there was no significant difference between the two groups. The central line in each box represents the median, while the box spans the interquartile range (IQR). Whiskers extend to 1.5 times the IQR, with individual points representing outliers. The younger infants showed higher levels of unconjugated bile acids compared to their older counterparts.

Finally, when compared to the American children, the 2-year-old Bangladeshi children had higher levels of unconjugated secondary bile acids (2.93% vs < 0.001%, P ≤ 0.001) but lower levels of muricholates (0.26% vs 27.02%, P = 0.02). There was no significant difference in percentage of unconjugated primary bile acids, conjugated primary bile acids, or conjugated secondary bile acids between the two groups.

DISCUSSION

Our study demonstrates that dysfunction in bile acid metabolism is associated with stunting and wasting in Bangladeshi children with EED. Overall, when compared to age-matched American children, the Bangladeshi children showed increased levels of unconjugated primary bile acids and sulfated primary bile acids, suggesting arrested metabolism. This dysregulation likely stems from defective hepatic conjugation pathways, as evidenced by significantly reduced levels of conjugated primary bile acids and diminished downstream bile acids. Under physiological conditions, primary bile acids synthesized in the liver are rapidly conjugated, primarily with glycine or taurine, by bile acid-CoA: Amino acid N-acyltransferase (BAAT). Conjugated bile acids, which are more hydrophilic, play a crucial role in nutrient absorption. Conjugation with glycine or taurine enhances solubility and absorption of hydrophobic nutrients such as long chain fatty acids, cholesterol and fat-soluble vitamins by forming mixed micelles in the small intestine[21]. In contrast, hydrophobic bile acids have potent inflammatory properties that can damage the liver, intestine, and other tissues, whereas hydrophilic bile acids exert anti-inflammatory effects[22]. The significant increase in serum C4 in Bangladeshi children may be related to the lack of conjugated bile acids in the enterohepatic circulation, lack of ileal transport, and decreased down-regulation of bile acid synthesis in the liver[23]. The liver responds to increasing rates of bile acid synthesis by increasing hydrophilic muricholate and sulfated bile acids. Therefore, the observed conjugation defect may contribute to long-term stunting and malnutrition by impairing nutrient absorption and promoting both systemic and enteric inflammation. Even after controlling for confounders, it was found that there was still a significant correlation between increased unconjugated primary bile acids and anthropometry. Of note, another study looking at Bangladeshi children found that poor growth was also associated with a higher ratio of conjugated to unconjugated bile acids, however it did not differentiate between primary and secondary bile acids[10]. The study done in Pakistani children also found that secondary bile acids were consistently low in kids with undernutrition and EED[11], a similarity within the Bangladeshi children that may be reflective of an overall conjugation defect in malnourished children.

Younger Bangladeshi children demonstrated a conjugation defect while older children did not, as shown by the elevated levels of primary and secondary conjugated bile acids in the 2-year-olds. This suggests that impoverished Bangladeshi children may have a delay in maturation of their conjugation pathways rather than an intrinsic deficit as seen in patients with BAAT and bile acid-CoA ligase (SLC27A5) gene defects[12]. This finding corroborates a previous study of Malawian children that inversely correlated levels of CA and CDCA with age in children with EED, specifically r = -0.45 for CA (P = 0.02) and r = -0.39 for CDCA (P = 0.05)[9]. These negative correlations also suggested that as the children aged, their serum levels of CA and CDCA decreased, a correlation with the delay we observed in the Bangladeshi cohorts. In a study conducted on microbial challenged mice, protein deficiency led to increased levels of primary bile acids compared to secondary bile acids[24]. The dysfunction noted in our cohort of Bangladeshi children may be attributed to their poor nutritional status from an early age, leading to the delay in conjugation that was noted in this study. Another possibility is a microbiota mediated effect. The gut microbiota plays a crucial role in bile acid metabolism, including processes like deconjugation to produce unconjugated bile acids and 7α-dehydroxylation to yield secondary bile acids[25]. The gut microbiota also matures over the first 2 years of life with children with EED having a more immature microbiome than expected based on chronological age[26]. Another study revealed that when gut microbiota are attenuated, apical sodium-dependent bile acid transporter (ASBT) expression is altered, which is the transporter responsible for bile acid reabsorption in the terminal ileum[27]. It is reasonable to hypothesize that gut microbiota changes from EED could influence ASBT expression as well. The gut microbiota's role in regulating bile acid distribution may explain the differences observed in Bangladeshi children affected by EED.

Muricholic acids are more common in mice but found only in small amounts in humans. While it is normal to have higher levels of muricholates during infancy, it will usually drop to undetectable levels by adulthood[28]. Our findings showed a higher percentage of muricholates in the older Bangladeshi children compared to the younger infants. Notably, the American cohort had a higher percentage of muricholates compared to the Bangladeshi infants, despite being an older cohort on average. However, this finding may be skewed given the overall significant drop in percentage of downstream bile acids due to the sequestration of primary bile acids in these children. Further research is needed to clarify this finding.

Younger Bangladeshi children demonstrated very little difference in the bile acid profile patterns between their stool and serum. In normal physiology children reabsorb their bile acids in their ileum for transport back to the liver via portal blood circulation to initiate negative feedback on subsequent bile acid synthesis, and therefore we expected a difference in bile acid levels between the stool and serum samples[29]. Specifically, we would have expected to see a significantly decreased level of bile acids in the stool in comparison, which we did not. However, it is possible that the intestinal barrier breakdown associated with EED is sufficient to equilibrate the serum and stool bile acid levels.

Currently, there is no known treatment for EED. Our data suggests correction of the early life conjugation defect could present a therapeutic opportunity. A study targeting defective bile acid amidation due to a deficiency in BAAT showed improvement in fat-soluble vitamin absorption after treatment with GCA[12]. Patients in the study demonstrated a lack of conjugated primary bile acids at baseline, similar to the children in our study. Oral GCA (15 mg/kg) was well tolerated and effectively incorporated into bile, with over 60% of biliary bile acids becoming conjugated and most of those as GCA. Treatment improved absorption of fat-soluble vitamins (D₂, E) and promoted growth in prepubertal children with prior delays, demonstrating that GCA supplementation can restore key bile acid functions and support nutritional recovery in amidation defects[12]. While this study enrolled only children with genetically confirmed BAAT deficiency, GCA supplementation may hold promise beyond strictly defined genetic cases. Similarly, another review highlighted the multifaceted therapeutic potential of bile acid supplementation for treating components of metabolic syndrome[30]. Animal models demonstrated that combining probiotics with bile acids produced stronger symptom alleviation than either intervention alone. This likely stems from modulation of the gut microbiota, which in turn influences bile acid signaling pathways, specifically FXR and TGR5[30]. Overall, these works highlight the importance and foundational groundwork needed for future research into bile acid supplementation that may be of benefit for impoverished children.

The 100 6-9-month-old Bangladeshi children also demonstrated an inverse relationship with increased levels of secondary conjugated bile acids correlating with lower levels of SIBO AUC. Small intestinal bacterial overgrowth is already independently linked to growth stunting[6] but this study explores the potential role of SIBO on bile acid metabolism as a mediating factor. Another study linked patients with malabsorption syndrome with SIBO to significantly elevated levels of unconjugated bile acids compared to those without SIBO[31]. This initial correlation warrants further research into how SIBO may affect gut bile acid conjugation and the potential downstream effects on growth and inflammation.

Our study had several notable strengths. First, the study conducted rigorous field collection with concomitant stool and serum samples. Second, we utilized an expanded bile acid profile to understand the accessory pathways being utilized in Bangladeshi children in the absence of effective primary bile acid conjugation. The inclusion of a wide range of bile acids in the analysis, including sulfated forms and muricholates, provides a more detailed understanding of bile acid metabolism and their implications. Finally, we sampled American children of similar age as a high-income country control.

This study also had several notable limitations that should be considered when interpreting its findings. First, as a cross-sectional study, it lacks longitudinal data on the enrolled children. Although we included a separate cohort of older children from the same impoverished neighborhood, these participants were not the same individuals as those in the younger cohort. This discrepancy limits our ability to assess individual-level influences or developmental trajectories in bile acid metabolism over time. Further, the original study did not exclude enrollment based on factors that could have influenced bile acid metabolism. Another key limitation is the relatively small sample size of the American cohort. While the comparison provides valuable insights, the limited size and diversity of the American group may not fully capture the heterogeneity present in a broader population. Also, as a cross-sectional study we are not able to infer causation but only detect association. Finally, results may not be representative of larger populations and diverse settings as the main participants were among rural Bangladeshi children. This highlights a broader gap in the literature, as few studies have attempted to characterize normal variations in bile acid metabolism across genetically and socioeconomically diverse pediatric populations at different developmental stages.

CONCLUSION

This study identified a correlation between conjugation defects and growth deficits in young Bangladeshi children. Further research is needed to better understand the impact of bile acid metabolic dysfunction in impoverished children from low-income countries and to explore potential avenues for novel interventions aimed at mitigating growth impairments.

ACKNOWLEDGEMENTS

Authors would like to acknowledge the children and families of Mirpur Dhaka without whom this research would not have been possible. ICDDR,B acknowledges its core donors, the Government of Bangladesh, and the Government of Canada for providing unrestricted support and commitment to ICDDR,B’s research effort.