Lactoferrin and lysozyme to reduce environmental enteric dysfunction and stunting in Malawian children: study protocol for a randomized controlled trial

Environmental enteric dysfunction (EED) is a subclinical chronic inflammatory condition of the gut. Known in the past as tropical sprue, tropical enteropathy, or environmental enteropathy, it is condition that primarily affects rural, poor, but seemingly healthy children, most prominently below 2–3 years of age [3‐5]. EED is histologically characterized by diffuse small bowel villous atrophy and T-cell infiltration of the intestinal mucosa that causes chronic immunostimulation [6], leading to increased intestinal permeability and impaired mucosal barrier function, which further leads to translocation of gut microbes and endotoxins, and nutrient malabsorption [7]. Although EED itself is a subclinical condition, affected children are subjected to increased rates of infection, poorer response to oral vaccines, and growth stunting.

Interventions aimed at combating EED have had mixed results. Trials involving antibiotics [8] and probiotics [9] to modulate the gut microbiota and decrease microbial-stimulated inflammatory responses have had limited success. Supplementation with multiple vitamin A [10], alanyl-glutamine [11], long-chain polyunsaturated fatty acids [12], albendazole [13], zinc [13], multiple micronutrient supplements [14], and mesalazine [15] have all been shown to at least transiently benefit small intestine function. However, these benefits have not always been consistent—even in combination—among children in the periods of highest food insecurity [16], and any demonstrated improvements in linear growth and development remain lacking. Given that malnutrition to this day still accounts for a significant amount of morbidity and mortality in children less than 5 years of age, interventions aimed at reducing EED are integral to improving child survival worldwide [17].

Breastfeeding provides innumerable benefits to infants and children, both in the short and long term. Infants in low- and middle-income countries who have never breastfed are 14 times likely to die from all causes compared to exclusively breastfed infants in the first 5 months of life [18]; most of this benefit stems from a lower risk of acute infections.

Similar to children with minimal dietary diversity, breast milk has also been shown to be protective against increased intestinal permeability [19]. Lactoferrin and lysozyme are two proteins present in substantial concentrations within various human mucosal secretions, particularly breast milk [20]. Lactoferrin is an 80-kDa iron-binding glycoprotein which accounts for 25% of the protein in breast milk (1–3 g/mL) and functions to sequester iron, thereby inhibiting bacterial growth [19, 21]. Lysozyme is a 15-kDa protein present in breast milk at 0.1–0.3 mg/mL that acts as an antimicrobial by lysing a specific connection within the peptidoglycan layer of bacterial cell walls [22].

Previous research has demonstrated that both lactoferrin and lysozyme supplementation are safe and efficacious in children as young as 5 months of age [23]. Lactoferrin has been shown to improve hematocrit levels and decrease the rate of lower respiratory tract infections in neonates without adverse effects [24]. Lysozyme levels in breast milk have been positively correlated with neonatal weight gain [25].

Lysozyme has also been shown to modulate inflammatory response and gut structure, as well as alter microbiota in animal models. One study showed that pigs which were fed human lysozyme-enriched milk expressed significantly elevated blood levels of the anti-inflammatory cytokine transforming growth factor (TGF)-β1 and tended to have longer villi with thinner lamina propia, which are linked to increased intestinal surface area and increased nutritional absorptive capacity [26]. In this animal model, lysozyme also led to altered gut microbiota, relatively enriched in Bifidobacteriacea and Lactobacillacea, two populations associated with a healthy microbiome, and decreased Mycobacteriaceae and Campylobacterales populations, which are associated with disease [27, 28]. Another study of supplementation with transgenic milk containing human lactoferrin in malnourished pigs showed benefits to jejunal architecture, reduced intestinal permeability, and promotion of weight gain [29].

In combination, lactoferrin and lysozyme have been shown to work synergistically as bactericidal agents against both Gram-negative and Gram-positive bacteria. Lactoferrin destabilizes the outer membrane likely through a mechanism involving iron sequestration and binding to membrane lipopolysaccharide, while lysozyme acts on the inner peptidoglycan membrane [22, 30, 31]. Importantly, both are resistant to proteolytic degradation [32, 33], which is essential for any therapeutic use as an oral agent in humans.

In a study of Peruvian children between the ages of 5 and 33 months, supplementing oral rehydration solutions with lactoferrin and lysozyme in the setting of acute diarrhea with dehydration decreased the severity and duration of diarrheal episodes, without any adverse effects related to treatment [34]. These therapeutic properties make lactoferrin and lysozyme attractive candidates for the prevention and treatment of EED, and more broadly may be useful supplements for infants who cannot breastfeed or who are weaning from breast milk but may still benefit from their therapeutic properties.

There are numerous factors that contribute to stunting among impoverished children worldwide, poor gut health being among the most important. We plan to investigate the effect of nutritional supplementation with lactoferrin and lysozyme on the growth and gut health of children in rural Malawi, a setting where over 80% of children have EED [8, 9, 13, 14, 16, 35, 36].

This will be a prospective, randomized, double-masked controlled clinical trial to assess the effect of lactoferrin and lysozyme supplementation on growth and gut health in rural Malawian children. Children between the ages of 12 and 23 months will be eligible for enrollment, and those enrolled will receive the intervention daily for 16 weeks. A 1:1 allocation ratio will be used to assign children to either the intervention or placebo groups. The anticipated trial flow is depicted in Fig. 1. The Standard Protocol Items, Recommendations for Interventional Trials (SPIRIT) checklist is available as Additional file 1.

This study will enroll children in the villages surrounding the community of Mmwenye, Machinga District, Malawi. Most villagers work as subsistence farmers with maize as the primary crop. Water is primarily acquired from boreholes or carried by hand from nearby streams. Homes are generally small, with foundations constructed of mud brick and roofs of thatch or sheet metal. It is not uncommon for multiple generations to live within the same house. Defecation and urination are primarily done in the open [37] or for some families in pit latrines dug into the ground. Children interact frequently with farm animals including chickens, goat, and cattle, some of which are permitted within houses. Geophagy is common [38, 39]. Electricity, running water, and other utilities are generally absent; most cooking is accomplished on open wood or charcoal fires and most cleaning is performed at streams or with water obtained from those streams. Most household responsibilities and child-rearing are performed by women. It is common for women to marry and assume household responsibilities relatively early in life, and thus most women are illiterate and have not completed primary school. Most hospitals are situated in the larger cities and are therefore prohibitively far away from villagers; more commonly, local health clinics have only modest resources and staffing. Health promotion and disease prevention is primarily accomplished in the community by Health Surveillance Assistants (HSAs) [40].

The study has been structured in accordance with the Helsinki Declaration [41] and approved by the University of Malawi College of Medicine Research and Ethics Committee and Washington University School of Medicine Human Research Protection Office. Consent for conducting the study was obtained from district health officials and community leaders. Prior to enrollment, large-group community meetings will be held to inform potential participants, their caregivers, and any interested members of the community about the study and answer their concerns. Consent from all caregivers will be obtained in written and verbal forms by native Chichewa-speaking Malawian research nurses, and those who cannot sign their names will submit thumbprints instead (Additional file 2).

Given that the intervention proteins are a normal component of breast milk, it is extremely unlikely that any adverse effects will be observed. No specific data safety monitoring committee will be used, but rather daily monitoring for any adverse events will be conducted by the field staff and reviewed and investigated by the study principal investigators. Additionally, at each follow-up visit caretakers will be specifically queried about the presence of any ill symptoms or unusual clinical events in their children in an attempt to proactively identify any untoward effects.

The primary outcomes of interest will be improvement of a major biomarker of EED, the percent of ingested lactulose excreted in the urine (Δ%L), from baseline to 8 weeks and to 16 weeks. Linear growth has been shown to best correlate with this component of the traditional dual-sugar absorption test [42]. The enrollment goal will be at least 86 subjects in each of two equally sized intervention or control groups. This sample size was chosen to detect a decrease of 0.05% in Δ%L between the two groups (considered to be a clinically significant improvement in gut barrier function) with an alpha value of 5% and 90% power. This assumes that urinary lactulose measurements will be distributed similarly to prior studies that showed that more than 80% of apparently healthy rural Malawian children had EED. Additional children will be enrolled to account for children who drop out or are excluded during the course of the study, as well as for some children who will not have a satisfactory data collection due to technical issues with the urine collection. The differences in lactulose recovery before and after the intervention will be calculated and compared between the intervention and control groups using t tests if the data are normally distributed or the non-parametric Wilcoxon Rank Sum Test if the data are not normally distributed.

Secondary outcomes of interest include linear and ponderal growth, adverse gastrointestinal symptoms, and any side effects due to the interventions. Linear growth will be calculated as the change in length over time, expressed as mm/day. Weight change will be calculated as the change in weight over time, normalized to the enrollment weight, and expressed as g/kg/day. Exploratory analyses of mRNA markers of intestinal inflammation found in the stool will also be conducted as part of our ongoing investigations into identifying novel biomarkers for EED [43, 44].

Following randomization, the experimental group will receive 1.5 g of bovine lactoferrin powder and 40 g of unpurified rice powder that contains 0.2 g of recombinant human lysozyme daily. The control group will receive 41.5 g of locally sourced rice powder daily. The amount of enzymes to be consumed per child per day is a conservative estimate based on the amount a typical breastfeeding 12-month old would receive daily and is consistent with previous studies [34, 45, 46]. Each group will receive their respective interventions for 16 weeks.

Bovine lactoferrin has been granted Generally Regarded as Safe (GRAS) status by the US Food and Drug Administration, and is widely used in studies of lactoferrin efficacy [23, 47, 48]. It will be acquired from FrieslandCampina (Amersfoort, Netherlands).

Bovine milk contains only trace amounts of lysozyme and is therefore not a viable source for this enzyme, but genetic modification has successfully integrated recombinant human lysozyme into rice. The transgenic rice has a normal plant phenotype; after harvest, the recombinant proteins are extracted and purified from milled rice powder as partially iron-saturated lysozyme [34]. The amino acid profile, iron content, peptide matching, chemical characterization, and functionality is identical to human lysozyme present in breast milk [49]. Ventria (Junction City, Kansas, USA) will be the supplier of the rice-based lysozyme product used in this study [34].

For the control group, a local variety of rice called faya will be used. It is an aromatic rice locally grown in Malawi with a nutritional profile similar to that used in the experimental formulation. The only processing this rice will undergo will be mechanical grinding; there will be no cooking, addition of chemicals, or other processes.

Participants will receive their interventions packaged in clear plastic jars with adequate room for shaking and mixing, and will be instructed to store their interventions in a cool location with the caps screwed on tightly. Caregivers will be instructed to add the interventions to the child’s “phala”, which is the traditional morning dish of maize porridge that is commonly fed as the preferred complementary food. Phala is made by adding maize flour to boiling water and stirring over heat for about 15 min until thickened to the desired consistency. To avoid any denaturation or alteration of the interventions that may come with boiling, caregivers will be instructed to add the interventions after cooking is complete. Each participant will be given a small scoop with which to portion the correct amount of intervention, and they will be taught to mix the specified number of these scoops into the cooked phala. Acceptability tests in local children have shown that both the experimental and control interventions are palatable when mixed into this phala, with no adverse reactions noted.

Each number from 1 to 250 will be randomly pr-assigned to either the control or experimental group by a random number generator, with 125 numbers assigned to each group. These numbers will be assigned in sequential order to participants as they are enrolled. The staff that generate the randomization list, and prepare and label each intervention/control jar with study numbers will do so prior to delivery of the interventions to the field and those staff will have no further involvement in the study. A designated field team member not involved in sample collection or anthropometric measurements will distribute the jars to the study participants. These field researchers will be trained in anthropometry, study protocols, the assessment of acute malnutrition, and good clinical research practices by the senior physicians and investigators. Researchers involved in the collection of samples and analysis of data will not directly handle, distribute, or view the intervention flours, nor will any of the paperwork they handle include information about the assigned intervention group. Care will be taken to ensure that the experimental and control powders look as similar as possible, but there are very subtle color, particulate, and texture differences between the two. Although caregivers will not be told to which group their children are randomized, it will be possible for them to differentiate the two different powders being used in the study should they undertake a very close comparison with another subject’s flour.

On the enrollment date, children will be screened for eligibility via clinical anthropometric measurements and a short questionnaire regarding recent illnesses. Children without any exclusion criteria will be given a study number, and will complete the lactulose absorption test, stool collection, and blood draw after consent is provided. HSAs will work with caregivers to complete a questionnaire that includes basic sociodemographic, economic, sanitary, and dietary data specific to lactoferrin- and lysozyme-containing foods, along with an account of recent illnesses in the previous 7 days. After completion of sample collection, each participant will receive a 4-week supply of the intervention corresponding to their assigned study number. Local staff will then educate caregivers on how to prepare and feed the interventions to the children. Recruitment of trial participants began in October 2016 and the trial is ongoing.

At subsequent monthly visits, study participants will return for anthropometric measurements, with or without sample collections. Questionnaires that primarily detail recent illnesses, medical visits, diet, and compliance by asking the caretaker how many days of the intervention were missed will be completed. The complete schedule of study visits, testing, and distribution of interventions is provided in the SPIRIT format in Fig. 2.

Anthropometric data will be obtained at every visit. Weight will be assessed to the nearest 5 g using standard digital scales (Seca 334, Chino, California, USA) and height will be measured to the nearest 0.1 cm using standard rigid height boards (Seca 417, Chino, California, USA). MUAC will be measured to the nearest 0.1 cm using standard flexible tape inserts (TALC, St. Albans, UK). Each child will be assessed for edema indicative of kwashiorkor by pressing on the dorsum of the left foot. The field research staff supervising anthropometry have been working in similar field sites and study conditions and trials for 5–15 years each, and will ensure that any junior research staff are fully trained and vetted prior to initiating the study. A random sampling of measurements each day will also be conducted by at least two field staff in order to verify that consistent and accurate measurements are being taken and corrective measures initiated as soon as any inaccuracies are identified.

Anthropometric Z scores will be calculated based on the 2006 World Health Organization (WHO) Multicentre Growth Reference Survey [50]. During both enrollment and subsequent visits, any child with a weight-for-height Z score less than –2, MUAC less than 12.5 cm, or presence of pitting edema will be classified as acutely malnourished and excluded from the study. Children excluded due to acute malnutrition will then be enrolled in supplementary or therapeutic feeding programs and receive usual medical care according to standard protocols [51‐53].

The lactulose absorption test is a modification of the dual-sugar absorption test [54] and will be used as a measure of gut integrity. This test has previously been correlated with histological changes in the small bowel [55], severity of diarrhea [56], micronutrient status [57, 58], metabolic abnormalities [59], and anthropometric growth [60]. This test involves the oral ingestion of a sugar solution with subsequent urine collection to determine the degree to which the sugars are absorbed within the intestine and enter the systemic circulation.

Lactulose is a large disaccharide that cannot cross these tight junctions, and can only traverse the intestinal epithelium in large quantities if there are significant disruptions [61]. The percentage of ingested lactulose excreted in the urine is a measure of intestinal mucosal permeability, with a higher percentage signifying increased intestinal permeability.

Children will be asked to take no food (water and breast milk are permitted) after 10 pm on the evening before each sample collection date (weeks 0, 8, and 16). Each child will meet at the community meeting point in the village at 6 am with the research team. Each child then will drink 20 mL of a solution containing 5 g lactulose. Children who spit up or vomit the solution will return on a subsequent day to try again. Adhesive urine bags will then be placed on the child’s perineum to capture each urination. Once a child has voided any quantity of urine, the bag will be removed and the urine will be transferred to a clean container specific to the child, and then a new urine bag will be placed on the child. Thimerosol (10 mg) dissolved in 20 μL of water will be added to each child’s urine container to prevent bacterial degradation of the sugars. Caretakers and children are told to avoid breastfeeding, and will remain at the community meeting point for the next 4 h. Four hours after drinking the sugar solution, children will be encouraged to void and complete the urine collection, after which the total volume of urine will be quantified and aliquots portioned into cryovials. These samples will be stored in liquid nitrogen tanks in the field and then stored in –80 °C freezers until they can be transported back to the United States. The urinary concentration of lactulose will be measured by high-pressure liquid chromatography (HPLC) as described previously [62]. The total excretion of lactulose will be determined by multiplying this concentration by the total volume of urine.

Clinical data focusing on demographic and anthropometric values will be collected on standardized forms by field workers. Data will be double-entered into a password-protected Microsoft Access database by research assistants blinded to the intervention assignment. All discrepancies will be resolved by examination of the original forms and discussion with the relevant field workers. After entering all data and resolving inconsistencies, the data set will be locked. For anthropometric values, comparisons will be made between the lactoferrin-lysozyme mixture and the control rice flour groups. Fisher’s exact test will generally be used to compare discrete parameters, and Student’s t test will be used for continuous parameters. A difference with P < 0.05 will be considered statistically significant. These statistical methods will also be used for all clinical data and EED measures.

All trial data, including personal information, will be kept in password-protected encrypted computer databases on password-protected computers. Paper records will be kept in locked file cabinets and the research team’s headquarters in locked buildings. Given the relatively short duration of the study and the need to analyze all urine concentrations simultaneously at the end of the study, no interim analyses are planned and no stopping guidelines have been established. The full protocol, participant-level dataset, and details of all statistical analyses will be made available to any interested parties after the conclusion of the trial by contacting the principal investigators.