Faecal regenerating 1B protein concentration is not associated with child growth in rural Malawi

Aim: This study was designed to determine whether faecal regenerating 1B protein (REG1B) concentration is associated with physical growth among 6–30-month-old children in rural Malawi. Methods: This was a secondary analysis from a randomised controlled trial in rural Malawi in which we followed-up 790 live-born infants from birth to 30 months of age. We collected anthropometric data at the age of 6, 12, 18, 24 and 30 months. We measured faecal REG1B concentration by enzyme-linked immunosorbent assay (ELISA) technique using stool samples collected at 6, 18 and 30 months of age. We assessed the association between faecal REG1B concentration and children’s physical growth using linear regression and longitudinal data analysis. Results: Of 790 live-born infants enrolled, 694 (87%) with at least one faecal REG1B concentration measurement were included in the analysis. Faecal REG1B concentration was not associated with the children’s concurrent length-for-age z-score (LAZ), weight-for-age z-score (WAZ), weight-for-length z-score (WLZ) and mid-upper arm circumference-for-age z-score (MUACZ) at any time point (P > 0.05), nor with a change in their anthropometric indices in the subsequent 6-month period (P > 0.05). Conclusions: Faecal REG1B concentration is not associated with LAZ, WAZ, WLZ and MUACZ among 6–30-month-old infants and children in rural Malawi. This is a secondary analysis of data that were collected in a randomised controlled trial conducted in two hospitals (Mangochi, Malindi) and two health centres (Lungwena, Namwera) in Mangochi District, rural Malawi, South‐East Africa between February 2011 and April 2015. Details of this trial have been described elsewhere. 13 The total population in the study area was about 190 000 and most of them spoke Chiyao and subsisted on farming and fishing. In brief, pregnant women with less than 20 completed gestation weeks were enrolled and randomly allocated into three groups, receiving daily 60 mg iron +400 μg folic acid (IFA) in IFA group, a tablet of multiple micronutrients (MMN) in MMN group or 20 g of lipid‐based nutrient supplements (LNS) in LNS group as interventions. After delivery, 790 live‐born infants were followed up until age of 30 months. Clinic and home visits were conducted to collect both data using questionnaires and biological samples. Further details of trial design and its main outcomes have been published earlier. 14 The study was approved by ethics committees in Malawi (College of Medicine) and Finland (Pirkanmaa Hospital District) and performed in accordance with the principles of Helsinki declaration and regulatory guidelines in Malawi. Written informed consents were obtained from caregivers. Research Assistants collected stool samples which had been placed in collection containers by mothers on the same day during home visits at 6, 18 and 30 months. The samples were on receipt immediately put in cooler bags. If the child had diarrhoea, sample collection was postponed by 2 weeks. The Research Assistants transported the samples in cooler bags to the site laboratory and laboratory technicians aliquoted the samples to cryovial tubes and stored them first in a −20°C freezer. Within 48 h, the samples were transported to a central laboratory where they were frozen at −80°C until being shipped on dry ice to Tampere University of Finland for analysis. An enzyme‐linked immunosorbent assay (ELISA) technique (TECHLAB, Inc., Blacksburg, VA, USA) was used to quantify REG1B concentration in stool samples. Samples were diluted at 1:10 000 before adding 100 μL of standards, controls and stool samples in duplicates to plates with pre‐coated immobilised polyclonal antibody against REG1B. The plates were incubated at 37°C for 20 min followed by shaking and five times of washing before adding conjugate solution into each well. The incubation and washing were then repeated before adding substrate solution, followed by incubating for 15 min at room temperature. After adding stop solution, plates were read using optical density (OD) 450/620 nm (Multiskan FC Microplate Photometer, Thermo Fisher Scientific Inc., Waltham, MA, USA). The linear standard curve was made by plotting standard absorbance against standard concentration to calculate REG1B concentration. Concentration was expressed as μg/g. Anthropometric measurements of children were taken by trained study staff at clinic visits at 6, 12, 18, 24 and 30 months. The staff measured length or height to 1 mm using a length board (Harpenden Infantometer, Holtain Limited, Crosswell, UK) and weight with reading increments of 10 g using an electronic infant weighing scale (SECA 735) and a digital adult weighing scale (SECA 874). Also, they measured mid upper arm circumference (MUAC) and head circumference to 1 mm with the use of non‐stretchable plastic insertion tapes. We calculated length‐for‐age z‐score (LAZ), weight‐for‐age z‐score (WAZ), weight‐for‐length z‐score (WLZ), head circumference‐for‐age z‐score (HCZ) and mid‐upper arm circumference‐for‐age z‐score (MUACZ) using World Health Organisation Child Growth Standards. 15 Change in z‐score was calculated by subtracting the anthropometric z‐score at the end of the interval of interest from that at the beginning of the interval. Baseline information of mothers and infants was obtained at both home and clinical visits. Maternal body mass index (BMI) and HIV infection was assessed at enrolment. Maternal malaria was diagnosed by the Rapid Diagnosis Test using Clearview Malaria Combo (British Biocell International Ltd., Dundee, UK). Research nurses recorded duration of pregnancy, infant sex and birthweight. Trained study staff collected breastfeeding information using questionnaires. Information on household food insecurity, expressed as household food insecurity access scores, 16 was also collected to assess the situation of household food intake in the past month. Statistical analyses were performed with STATA version 15.0 (StataCorp, College Station, TX, USA). The definition of age for 6, 12, 18, 24 and 30 months was 20–32 weeks, 46–58 weeks, 72–84 weeks, 98–110 weeks and 124–136 weeks, respectively. Linear regression models were used to analyse the association between REG1B concentration at 6, 18 and 30 months and anthropometric data at the same time point respectively, and the association between REG1B concentration at 6 or 18 months and change in anthropometric z‐score in subsequent 6 months. Random effects model was used to estimate the association between repeated anthropometric indices at 6, 18 and 30 months and repeated faecal REG1B concentration also from the same multiple time points. Models were adjusted for child age, birthweight, breastfeeding after delivery (yes/no), maternal HIV infection (positive/negative), child sex, duration of pregnancy, maternal malaria (positive/negative) and household food insecurity access scores. These variables were selected as potential confounders in advance of the analysis, as recommended in a recent textbook on statistical analysis of child growth. 17 To better understand the impact of possible confounders, we present as also unadjusted analyses as Supplementary tables. For the repeated measurement analysis, adjustment for child age was included also in the sensitivity analysis because of the strong association between it and the children’s intestinal biomarker concentration. The numbers of participants included were different in different models because of missing values in REG1B data at different visit times.