Table 1.
Domains of environmental enteric dysfunction and corresponding biomarkers.
Fig 1.
Flow of participants through the SHINE Environmental Enteric Dysfunction (EED) substudy.
(A) Enrolment, treatment allocation, sub-study selection and losses to follow-up. (B) Specimen collection. SOC, standard of care; IYCF, infant and young child feeding; WASH, water, sanitation and hygiene. None means that no serum or stool samples were collected at that visit, but infants remained in the substudy.
Table 2.
Maternal, household and infant baseline characteristics of HIV-negative mothers and their liveborn infants included in analyses, by intervention arm.
Table 3.
Intervention delivery and participant uptake in the SHINE EED substudy by treatment group.
Fig 2.
Geometric means, pointwise 95% confidence intervals and the percent of infants below the limit of detection for biomarkers of intestinal inflammation at 1, 3, 6, 12, and 18 months in the EED substudy by combined treatment arm.
(A) IYCF vs non-IYCF. (B) WASH vs non-WASH. IYCF, infant and young child feeding; WASH, water, sanitation and hygiene; <LOD(%), percent of samples below the limit of detection. Data were smoothed using generalized additive models with cubic splines and 3 knots. Dot (.) indicates a statistically significant difference between treatment arms at that study visit by Tobit regression. Asterisk (*) indicates statistical significance difference after adjustment for multiple testing. Each visit had a window to enable follow-up if the infant was not seen on the target date. These were: month 1 [4–12 weeks], month 3 [12–25 weeks], month 6 [25–51 weeks], month 12 [51–76 weeks], and month 18 [76–104 weeks]. Median[IQR] infant age in months at each visit was 1.3[1.1,1.9], 3.4[3.2,4.0], 6.4[6.2,7.0], 12.3[12.1,12.5], 18.0[17.8,18.3]. KTR was not measured at 18 months. MPO, myeloperoxidase; NEO, neopterin.
Fig 3.
Geometric means, pointwise 95% confidence intervals and the percent of infants below the limit of detection for biomarkers of intestinal damage, biomass and regeneration at 1, 3, 6, 12, and 18 months in the EED substudy by combined treatment arm.
(A) IYCF vs non-IYCF. (B) WASH vs non-WASH. IYCF, infant and young child feeding; WASH, water, sanitation and hygiene; <LOD(%), percent of samples below the limit of detection. Data were smoothed using generalized additive models with cubic splines and 3 knots. Dot (.) indicates a statistically significant difference between treatment arms at that study visit by Tobit regression or by GEE estimated linear regression. Asterisk (*) indicates statistical significance after adjustment for multiple testing. Each visit had a window to enable follow-up if the infant was not seen on the target date. These were: month 1 [4–12 weeks], month 3 [12–25 weeks], month 6 [25–51 weeks], month 12 [51–76 weeks], and month 18 [76–104 weeks]. Median[IQR] infant age in months at each visit was 1.3[1.1,1.9], 3.4[3.2,4.0], 6.4[6.2,7.0], 12.3[12.1,12.5], 18.0[17.8,18.3]. Citrulline was not measured at 18 months. CIT, citrulline; I-FABP, intestinal fatty acid binding protein; REG-1β, regenerating gene 1 beta.
Fig 4.
Geometric means, pointwise 95% confidence intervals and the percent of infants below the limit of detection for EED biomarkers of intestinal permeability and microbial translocation at 1, 3, 6, 12, and 18 months in the EED substudy by combined treatment arm.
(A) IYCF vs non-IYCF. (B) WASH vs non-WASH. IYCF, infant and young child feeding; WASH, water, sanitation and hygiene; <LOD(%), percent of samples below the limit of detection. Data were smoothed using generalized additive models with cubic splines and 3 knots. Dot (.) indicates a statistically significant difference between treatment arms at that study visit by Tobit regression. Asterisk (*) indicates statistical significance after adjustment for multiple testing. Each visit had a window to enable follow-up if the infant was not seen on the target date. These were: month 1 [4–12 weeks], month 3 [12–25 weeks], month 6 [25–51 weeks], month 12 [51–76 weeks], and month 18 [76–104 weeks]. Median[IQR] infant age in months at each visit was 1.3[1.1,1.9], 3.4[3.2,4.0], 6.4[6.2,7.0], 12.3[12.1,12.5], 18.0[17.8,18.3]. A1AT, alpha-1 antitrypsin; LM, lactulose mannitol; sCD14, soluble CD14.
Fig 5.
Geometric means, pointwise 95% confidence intervals and the percent of infants below the limit of detection for EED biomarkers of systemic inflammation and growth hormone activity at 1, 3, 6, 12, and 18 months in the EED substudy by combined treatment arm.
(A) IYCF vs non-IYCF. (B) WASH vs non-WASH. IYCF, infant and young child feeding; WASH, water, sanitation and hygiene; <LOD(%), percent of samples below the limit of detection. Data were smoothed using generalized additive models with cubic splines and 3 knots. Dot (.) indicates a statistically significant difference between treatment arms at that study visit by Tobit regression. Asterisk (*) indicates statistical significance after adjustment for multiple testing. Each visit had a window to enable follow-up if the infant was not seen on the target date. These were: month 1 [4–12 weeks], month 3 [12–25 weeks], month 6 [25–51 weeks], month 12 [51–76 weeks], and month 18 [76–104 weeks]. Median[IQR] infant age in months at each visit was 1.3[1.1,1.9], 3.4[3.2,4.0], 6.4[6.2,7.0], 12.3[12.1,12.5], 18.0[17.8,18.3]. CRP. C-reactive protein; KTR, kynurenine:tryptophan ratio; IGF-1, insulin-like growth factor 1.