Use of a novel supplementary food and measures to control inflammation in malnourished pregnant women in Sierra Leone to improve birth outcomes: Study protocol for a prospective, randomized, controlled clinical effectiveness trial

Background: The negative synergy between poor nutritional status and infectious diseases is doubly detrimental in pregnancy. In Sierra Leone, maternal malnutrition is amongst the highest in the world, while maternal mortality is high at 1320/100,000 live births and stunting in under-five is 37.9%, ranked 110/132 worldwide. Maternal malnutrition has been associated with preterm birth, small-for-gestational age infants, and poor maternal outcomes. Infants born prematurely or small-for-gestational age experience higher mortality and are at risk for stunting and decreased cognitive performance. Nutritional interventions alone during pregnancy may not be as effective in the setting of increased inflammation from repeated infections. Interventions are needed to improve maternal outcomes and reduce stunting in this population. Methods/design: This will be a prospective, randomized, controlled clinical effectiveness trial of an improved supplementary food plus anti-infective therapies compared to standard therapy in malnourished pregnant women. Pregnant women will be randomized to receive a low water activity, ready-to-use supplementary food plus five anti-infective interventions or the standard of care which is 3.5 kg corn/ soy blended flour with 350 mL vegetable oil every two weeks. The five anti-infective interventions are 1) insecticide-treated mosquito net at the time of enrollment into the study, 2) sulfadoxine-pyrimethamine given every 4 weeks, beginning at enrollment or at 13 weeks’ gestation, whichever is later, 3)azithromycin at a dose of 1 g given once at enrollment (after first trimester)and again during 28-34 weeks of gestation, 4)single dose 400 mg albendazole given in second trimester, and 5) testing and treatment for bacterial vaginosis at enrollment and again at 28-34 weeks of gestation. Treatment will be provided for the duration of the pregnancy. The primary outcome measure will be birth length. Secondary outcomes in the mothers will include rates of maternal weight gain and increase in mid-upper arm circumference, and time to maternal anthropometric recovery. Secondary outcomes in the infants will include birth weight, birth head circumference, and linear and ponderal growth. Discussion: Malnutrition remains a major problem in the developing world with lasting maternal and infant consequences. Maternal malnutrition has been associated with intrauterine growth retardation, low birth weight (LBW), pre-term delivery and poor cognitive development. Nutritional interventions alone have not been successful in reducing stunting. By bundling nutritional and anti-infective interventions, we aim to reduce intrauterine growth restriction and low birth weight in moderately malnourished pregnant women in Sierra Leone. If successful, this bundle can easily be implemented by governments or non-governmental organizations. Trial registration: Clinicaltrials.gov NCT03079388; Date: March 5, 2017. In this study, we aim to determine whether a combined intervention of a specialized nutritious food for pregnant women and a package of anti-inflammatory interventions results in improved birth anthropometry compared with the standard of care. This will be a prospective, randomized, controlled clinical effectiveness trial in pregnant women with malnutrition comparing the impact of combined nutrition and infection control interventions with the standard of care in Sierra Leone. The study will enroll pregnant women at less than 35 weeks of gestation with moderate or severe malnutrition. A probable patient diagram is shown in Fig. 1. Patient flow diagram for clinical trial. *Women 16 years of age or older will be enrolled and followed until 6 months post-partum. **Women less than 35 weeks of gestation will be enrolled The study will enroll patients at twenty antenatal clinics in rural Pujehun District in southern Sierra Leone and 10 antenatal clinics in the Rural Western Area District. In 2013, Pujehun District had the highest rate of under 5 stunting at 46.4% which is 8.4% above the national rate. Mothers who are thin (BMI < 18 kg/m2) are more likely to have children that are stunted, identifying Pujehun as an ideal district for a maternal nutrition intervention. The population predominantly belongs to the Mende ethnic group. Plantation farming is a major economic activity in the region with agricultural production of palm oil, cassava, and sweet potatoes [34]. Pujehun is one of the country’s poorest and least developed districts. We aim to complete the study using the highest ethical standards for clinical trials research in accordance with the Declaration of Helsinki [35]. Ethical approval has been obtained from the Sierra Leone Ethics and Scientific Review Committee (SLESRC) and from the Human Research Protection Office at Washington University in St. Louis. Support has also been obtained from Pujehun District Health Officer. Each subject will be briefed on the study procedures and interventions. Women older than 16 years of age will be eligible to consent for themselves. If a woman younger than the age of 16 desires to participate, consent will be obtained from the parent or guardian. Both verbal and written informed consent will be obtained from all enrollees. Those who are unable to sign or write their names will be asked to document consent using thumbprints. Pregnant women presenting to any of the thirty selected antenatal clinics will be screened for enrolment eligibility. Women who have a fundal height no greater than 35 cm, MUAC ≤23 cm, and planning to reside in the study area and attend the antenatal clinic will be considered eligible. MUAC is a reproducible, reliable measure in the field and easily understandable by both health workers and participants. Fundal height will be used as a measure of gestational age. Exclusion criteria will include pregnant women less than 16 years of age who do not have an adult community member willing to consent for them or known complications of pregnancy such as gestational diabetes, pre-eclampsia, hypertension, or severe anemia. The primary outcome measure of interest will be birth length in centimeters. The secondary outcome measures of interests include the total and average weekly maternal weight gain, recovery from maternal malnutrition identified as a MUAC > 23, newborn head circumference, birth weight, infant linear and ponderal growth, and infant survival at 3 months and 6 months. The sample size will be 1514 pregnant women with malnutrition, divided equally between RUSF and control groups. This sample size allows for a 20% drop out rate and/or exclusion rate due to fundal height, leaving a final sample size of 1200 (600 per group) with a two tailed significance of 0.05, power of 80%. This will allow for detection of a difference of 0.22 SD or 0.5 cm in birth length and a difference of 0.19 SD or 80 g in birth weight. Subjects will be randomized to receive the RUSF or the standard of care (supercereal, oil, and iron and folic acid tablets), using a random number generator which prospectively assigns participants to a study group. Since RUSF is visually distinct from flour, neither the study subjects nor the research study team members working directly with participants will be blinded. Study managers will be blinded to treatment during data analysis. A study team member will attend each participating antenatal clinic for subject recruitment. All women attending the antenatal clinic will be briefly informed of the study and entry criteria in a group setting. Women willing to be measured and potentially participate will be offered measurement of MUAC and screened for study eligibility. All women meeting eligibility criteria will be offered enrollment and agree via informed consent. Informed consent will be obtained by the on-site study coordinator. All participants who are 16 years of age or older will be able to consent for themselves as adults. If a participant who is younger than 16 wants to participate in the study, they will be required to have consent from a parent or guardian. Upon enrollment, the participant will be interviewed. Demographic information will be recorded, as well as, time of last menses, estimated date of delivery, and fundal height. Anthropometry will include current weight, height, MUAC, and weight history if available. Available medical records of the women will be reviewed with attention to any current or previous pregnancy complications and medications. Subjects will answer a standard clinical symptoms questionnaire, a food frequency questionnaire, and a Household Food Insecurity Access Scale (HFIAS). Participants will return for follow-up every 2 weeks for measurements and provision of food until delivery. Since birth measurements are primary outcomes, a birth weight team will be identified and trained in order to obtain accurate birth measurements within 48 h of delivery. Participation in the study will continue throughout pregnancy, and follow-up with the mother and infant will continue until 6 months postpartum. Mothers and infants will be followed at usual infant clinic follow-up at 6 weeks, 3 months, and 6 months post-partum contact point clinic. Infant weight, linear growth, and morbidity will be monitored. Maternal weight, MUAC and morbidity will be assessed at these visits, as well. The primary investigator will be responsible for the overall management of the trial. The on-site investigator will be responsible for day to day management of clinic sites. Adverse events have been defined as any reactions to RUSF, emesis, diarrhea, or rashes suggestive of food allergy. To monitor the safety of participants, a data safety monitoring meeting will be held every other month. The meeting will be convened by the Director of Nutrition for the Ministry of Health and Sanitation for the government of Sierra Leone. Present at this meeting will be the director, a study nurse, the district health officer, and one of the investigators. At the meeting routine morbidity data, enrollment data, and participant satisfaction will be reviewed. Minutes of these meetings will be recorded and forwarded to the Sierra Leone Ethics and Scientific Review Committee (SLESRC). A special inquiry will be convened whenever a serious adverse event is discovered. At any time a study staff member is informed about an unexpected event of a participant by a health surveillance assistant, national clinic staff, or community member, this will be reported to the Director of Nutrition for the Ministry of Health and Sanitation for the government of Sierra Leone within 24 h. Information will be gathered, and a summary of the event will be given to the data safety monitoring board within 24 h of being informed. The event will be reviewed in detail by the principle investigator and if the information supports the concern that this could be study related, an independent physician will be asked to review the case and render an opinion. The findings and conclusions will be recorded and forward to SLESRC and the Washington University IRB. Linear programming (LP) technology, a formulaic computer database program listing all potential ingredients, nutritional composition, prices, and country specific availability for Sierra Leone has been used to develop and optimize the RUSF. The RUSF will be produced by Project Peanut Butter (PPB), Sierra Leone, a local non-profit organization that produces and distributes ready to use therapeutic foods. PPB operates a factory in Freetown, Sierra Leone that has been internationally certified to supply RUSF to UN agencies. The spreadsheet-based tool identified a low-cost formula for RUSF for pregnant women. The tool allows users to consider essentially any candidate ingredient, as long as data on nutrient content, technical parameters, and prices are available. A preliminary step in the RUSF development process was conducting a price survey of all potential local ingredients that could be used in a formulation. An initial list of 89 possible locally-available candidate ingredients was identified through the following multi-step, systematic approach: After this draft list was developed, additional ingredients were excluded based on feedback from local producers, nutritionists, food technologists, Sierra Leone government ministries, and other local experts. A subsequent list of 27 locally available candidate ingredients remained following these additional exclusions. The final candidate ingredient list for the Sierra Leone model consisted of 53 ingredients (Table 1). Candidate ingredient list of locally produced and imported foods Using the LP tool designed for Sierra Leone, preliminary formulations were generated in the Food and Nutrition Food Lab at Washington University School of Medicine (WUSTL). Four potential RUSF formulas were created at WUSTL that contained peanuts, milk powder, vegetable oil, sugar, and a vitamin mineral complex. The major difference between the four was the indigenous ingredient present: millet, split pea, cowpea, bambara bean. These four potential RUSF formulas were then tested informally by pregnant women at an antenatal clinic in Pujehun, Sierra Leone. Informal acceptability showed the formula with millet was the overwhelming favorite. The millet formula was further optimized for nutrient content, feasibility of production and organoleptic acceptability. Rudimentary blending and heating techniques were used to create a homogenous lipid matrix containing a variety of legumes, cereals, milk powders, vegetable oils, sugar, vitamin mineral complex and an emulsifying agent until the final formula was determined. Table 2 specifies the ingredient composition of the study RUSF, named Mama Dutasi. The RUSF will provide a total of 520 kcal, 18 g protein, 200% of RDA for most micronutrients during pregnancy (Table 3). The supplement has also been optimized to provide excellent protein quality. The formulation contains an optimal polyunsaturated fatty acid composition by reducing the percent of energy from linoleic acid (omega-6) and increasing the percent of energy from α-linolenic acid (omega-3) to 4.5% and 1.5%, respectively [36–38]. RUSF Ingredient Composition Nutrient profile of study foods a Amount recommended for supplementation, with 200 mcg to come from dietary sources for total RDA of 600 mg/d b Daily ration: 250 g supercereal (CSB+) with sugar, 20 ml oil, 60 mg Fe supplement with 400 mcg folic acid c Calculated based on a WFP supercereal product containing ~ 64% corn, ~ 24% soybean,~ 20 ml vegetable oil Large scale production of this product requires the millet be in pre-processed form. The most effective pre-processing methods are roasting or extrusion cooking followed by milling. In this instance, the process used was sorting, drying, roasting, and milling. Pre-processing was performed at FINIC Industries in Kissy, Sierra Leone. The following outlines the pre-processing of millet in more detail: Proper sorting and cleaning is a necessary step to all raw commodities regardless of the harvest process. Sorting/cleaning the millet required removing dirt, debris, and stones. Visible debris like twigs and grass were removed by hand while dirt and stones were removed with a de-stoner. Once cleaned, the millet was allowed to drain. Roasting is a dry heat method which bakes/toasts the ingredients. The whole grains will be placed in a rotary oven heated by charcoal and dried for 12+ hours. This process is used to destroy microorganisms, lower the water activity, inactivate enzymes and catalyzes flavors via browning reactions. Milling or size reduction of whole grains was necessary prior at incorporating into the process line as to not damage the existing equipment. Size reduction of ingredients is also important for edibility. Although grinding of the final product occurs during the production process for particle reduction and homogenization, it is important that the millet have a small enough particle size to not damage the grinder plates. The roasted, cooled, dried millet was ground to a particle size similar to baking flour and packaged in clean lined peanut buckets. The millet flour was cleared by quality control and food safety testing for microbial contamination at Eurofins Scientific Inc., Des Moines, Iowa, USA and was stored until production of the RUSF product. The final product was packaged in 100 g foil sachets and identified with a custom Mama Dutasi label. Analytical testing was performed for food safety and quality assurance for aflatoxin and microbial contamination at Eurofins Scientific Inc., Des Moines, Iowa, USA. The control group will receive the standard of care for malnourished pregnant women in Sierra Leone which is 3.5 kg super cereal with 350 mL vegetable oil every two weeks. This provides 250 mg portion per day of the super cereal and 25 g oil per day for the mother. Iron and folic acid supplement goal is 90 pills through pregnancy. The study will assure that women have an uninterrupted access to the ready to use local supplementary food and super cereal if they are in the control group. Table ​Table33 provides a comparison of the nutritional composition of the study RUSF, standard of care food, and recommended daily intake for pregnancy. Women will receive the food for the duration of their pregnancy. The intervention group will receive a package of five interventions aimed at decreasing the burden of infections in pregnant women in order to decrease the risk of stillbirth, fetal growth restriction, and premature delivery. This package will include an insecticide-treated mosquito net at the time of enrollment, sufadoxine-pyrimethamine given every 4 weeks beginning at 13 weeks gestation, azithromycin given at the time of enrollment and at 28–34 weeks of gestation, single dose of albendazole at enrollment, and testing and treatment for bacterial vaginosis at enrollment and again in the third trimester. Table 4 details the anti-inflammatory interventions. Anti-Inflammatory Interventions, Dosing, and Timing of Administration • Enrollment or 13 weeks gestation, whichever is later • Every 4 weeks until delivery • Enrollment or second trimester, whichever is later • Weeks 28–34 • Enrollment • Weeks 28–34 The control group will receive the current recommendations of the government of Sierra Leone, which includes 60 mg iron and 400 μg folic acid supplementation starting at enrollment, an insecticide-treated mosquito net at the time of enrollment, three doses of sulfadoxine-pyrimethamine during second and third trimester, and a single dose of albendazole for deworming at enrollment. Table 5 provides a summary of the schedule of activities and interventions for subjects by week of study participation. Schedule of activities/ interventions for subjects by week of study participation aHFIAS household food insecurity assessment scale Food Safety specifications: RUSF specifications will be in accordance with the CODEX guidelines for supplementary food products [39]. Clinical data including demographics, anthropometry, morbidity, and mortality data will be collected by field workers using standardized forms. Field workers will be trained in the questionnaires and measurements prior to collecting any data. Completed data forms will be stored in a secure locked central location. All data will be double entered, compared, and sealed in a password-protected electronic database before the randomization code is broken. All data discrepancies will be resolved by examination of the original data cards and discussion with the relevant field workers. The data set will be locked after all discrepancies have been resolved. Descriptive statistics will be used to characterize the population. Student’s t-test will be used to assess whether primary outcomes of infant birth weight and length, and secondary outcome of maternal weight gain improved significantly in the intervention group compared with the control. Fisher’s exact test will be used to determine if the proportion of mothers recovering from malnutrition and of infants with LBW are improved in the intervention group. Regression modeling will be used to investigate factors that influence maternal and infant outcomes. Independent variables to be considered as covariates for inclusion in the models include: maternal age, weight, number of previous pregnancies, anemia, illness and number of weeks of intervention. For longitudinal measures, the groups will be compared using repeated measures mixed model analysis of variance. Covariates will include age, number of previous pregnancies, years of education and clinic location. Findings of this study will be disseminated through publication in peer reviewed journals and presentations at national and international conferences. The findings will also be shared with the Director of Nutrition for the Ministry of Health and Sanitation for the government of Sierra Leone. Study results will be presented to the communities and participants involved through community meetings and information sessions.