Design, development, and local production of lipid-based nutritional supplements to enhance the complementary feeding diet: A model for collaboration for a feeding trial in Bangladesh

Cultivating collaboration and establishing a communication plan

In order to develop, produce, and test a locally-produced LNS, a range of expertise and routine and transparent communication among the partners with established roles and agreed upon responsibilities was required. The group recognized that convening multiple partners from numerous locations on a regular basis was a critical step towards establishing effective communication and a forum to resolve issues, both of which were invaluable for supporting the overall project objectives. Additionally, the partners felt that developing a co-owned collaborative framework was important for launching and maintaining the effort.

During the first partners’ conference call, occurring 13 months before the expected trial launch date, participants clarified roles and communication strategies and established terms of reference to reflect agreed-upon achievable deadlines in light of project funding and in-kind support, ongoing, non-project related activities, and anticipated challenges (Figure 1, Figure 2). During subsequent meetings, scheduled monthly on average, partners revisited strategies and deadlines, presenting openly their challenges, namely those that would impact achieving the objective of producing the LNS products at the industrial level within the set goal of 1 year. Establishing early in the collaboration a cadence of meetings and anticipated outputs and deliverables ensured open vetting as a group of each partner’s accomplishments, current and projected bottlenecks, strategies to address bottlenecks, and, when necessary, consequent adjustments to the overall timeline.

Figure 2. Timeline of project milestones.

In addition to the communication among partners, it was critical to ensure the Bangladesh government was involved and informed. Thus, partners prioritized presenting the activity to and maintaining routine communication with key public health and nutrition stakeholders and government agencies. Toward this objective, an initial partner task was established focusing on building country-based knowledge about and support for the complementary food supplementation project within the Ministry of Health and Family Welfare (MoHFW), the Bangladesh Institute of Public Health Nutrition (IPHN), and the Bangladesh offices of Alive and Thrive, Nutrition International, UNICEF, and the US Agency for International Development (USAID), among other identified key stakeholders. Additionally, as one of the first actions, the partners introduced the project objectives to the existing Nutrition Working Group, comprised of representatives of the key nutrition partners including UN Agencies, development organizations, civil society, and public sector institutes. This dialogue led to cultivating a venue for answering questions and seeking inputs on the project. A Technical Committee was also formed and led by a leading scientist (TA) and member of a partnering organization, icddr,b, and was composed of child nutrition and health leaders and scientists in the MoHFW, and in academic, medical, and public health fields. This Technical Committee informed and sought guidance from our collaborative project.

Designing a region- and population-specific LNS

Recipe development. An important aspect of developing an LNS as a complementary food supplement for rural Bangladeshi children 6–23 mo of age, guided by existing international products, required creating a recipe that incorporated locally-available, culturally recognized and acceptable ingredients. The partners needed to ensure the products were fortified with multiple micronutrients at agreed upon levels sufficient to bridge the requirement gap in this age group while not adversely affecting the organoleptic or shelf-life qualities of the final product. Detailed consultations among nutritionists and food scientists were needed to finalize the target composition of the LNS including the amount of animal source (dairy) protein in the supplement recipe as well as levels of nutrients included in the premix. Further, the partnership leveraged their broader networks of experts, including a food technologist with WFP based in Pakistan (SM) who was supporting the production and distribution of a chickpea-containing LNS. The food technologist participated in developing the Bangladesh LNS version and provided other technical assistance throughout the LNS production activities.

Product development. Collaborators affiliated with the icddr,b food processing laboratory played a seminal role in the development of the LNS products. Details about how these products were designed, developed, and finalized are described in detail elsewhere²¹. In brief, the icddr,b food technologists created two recipes – one with a rice and lentil base and another with a chickpea base. The formulations were designed as LNS and thus in addition to the grains, the products contained vegetable oil, and dried skim milk powder was the source of animal protein in addition to the plant proteins found in the grain and cereal grains. Producers could procure the ingredients in bulk locally, at a reasonable price, and at a standard required to maintain a consistently high quality LNS.

Once the new formulations were defined and the nutrient composition of the final products measured, the partners determined the desired micronutrient composition of the LNS which would be met by adding a premix to the formulation. The food technologists on the team participated actively in the composition discussions to provide guidance on ensuring the levels and ratios of certain vitamin and minerals and their influence on product stability and organoleptic properties, while the nutritionists ensured fulfilling daily requirement gaps based on a diet of breastmilk and complementary foods. To specify the micronutrient composition, the following aspects were considered: desired portion sizes of the LNS and daily recommended micronutrient intake consumed through one serving by children; the ratio of minerals (such as iron and zinc) to preserve bioavailability; levels of nutrients that could influence organoleptic properties of the product; shelf-life of the product; accounting for micronutrient amounts intrinsic in the LNS ingredients, etc. Once finalized, DSM supported the production and provision of the appropriate micronutrient premix. The chemical form of the micronutrients was chosen considering their application in the specific food matrix, including bioavailability, stability, and organoleptic properties.

The consensus among the partners was that a daily portion of the LNS should provide energy levels found in SQ-LNS (125 kcals) to fill the energy gap in home-based traditional diet. However, because this food would be distributed to a population consuming a diet low in micronutrient content and variety, they devised the premix to provide 75 to 100%, on average, of vitamin and mineral requirements. For a year-long supplementation starting at 6 mo of age, provided by the JiVitA-based RCT, the partners agreed that the LNS would provide children 6–11 mo of age 125 kcals per day through a single sachet of SQ-LNS and children 12–17 mo with 250 kcals per day through two sachets.

Once the food processing lab at icddr,b finalized the recipes, they shared samples with the partners. We then proceeded to conduct an acceptability trial of the two products among the target group of young children in the rural Gaibandha district where the RCT was to be conducted. The results of this study showed high level of acceptability for both products²¹. Concurrent with the acceptability trial, the icddr,b team executed a shelf-life study to determine, among other characteristics, safety (microbes, toxins, heavy metals), water activity, and oil separation tendencies of the products.

Developing a production and distribution plan

Selecting a food manufacturer. Partners had to select a local food manufacturing company interested in and capable of producing the foods in the quantities required for the RCT. This production had to adhere to strict, established food product hygiene standards and requirements²² for children under 2 years of age.

The other success factor for this phase of the project was creating clearly defined tasks and responsibilities for each partner in the manufacturing process. More specifically, a detailed LNS production plan map with a timeline of tasks and individual or groups named responsible for completing tasks guided the effort. Clear reporting from each individual in regular meetings resulted in establishing a good understanding of whether the plan was being executed as anticipated and for course corrections decided collectively.

WFP, a global expert in procurement, logistics, transportation, and storage, played an important role in selecting a suitable and interested manufacturer, leveraging existing collaborations for their fortified biscuit programs. The partners formed a Task Force from the collaborating organizations who developed a shortlist of potential food manufacturers based on their available product range as well as having access to the required equipment, facilities, and quality assurance methods for LNS production.

The selected food manufacturer would have to assume a variety of responsibilities for the study, which required production of a total of 12–15 metric tons each of the two products including creating a separate production line that met the factory specifications for products designed for children less than 2 years old and take this on without knowing if there would be future demand for the product. In addition, before initiating any production, the factory would need to undertake product development efforts to adapt the recipes from the kitchen-based to the manufacturing versions.

The partners approached the five listed companies to request meetings with senior teams and visits to production units to assess the companies’ capacity and willingness to produce food products for young children. Notably, none of the companies, nor any in Bangladesh, were engaged in producing food for children under 2 years of age. Two of the three companies were not interested in initiating the activity due to the resources required to achieve and maintain the strict, high standards, especially without promise of future business and associated revenue for such products. Based on the Task Force assessment, Olympic Industries, Dhaka, one of the largest manufacturers and distributors of biscuits in Bangladesh, met the various criteria including an expressed interest to participate. Olympic was also the most open to the collaboration and was willing to make modifications to their production equipment and lines to meet the needs for production of the complementary foods for children.

The partners established a memorandum of agreement with Olympic Industries with agreed upon payment of the contractual services, largely coordinated by WFP including early investments to set up a separate production room, dedicating equipment, purchase of clean raw ingredients, among others. Following this, the company initiated setting up a production line assembly, modifying existing machines and their location, sourcing raw consumable and package materials, and developing a hygiene and quality assurance plan. The partners supported and monitored these activities. To provide more specialized support, an independent food technologist from Thailand who was an expert in food microbiology and large-scale manufacturing processes especially related to quality control, texture, appearance, and viscosity analyses visited the production site every other week, accompanied by icddr,b and WFP-Bangladesh colleagues, to provide technical inputs during the early phases of the factory production. A main highlight of the project was the successful technology and knowledge transfer from a food processing lab to a local processed food producer for the production of foods for young children requiring markedly high standards. However, one of the challenges faced was that the time required to establish the specialized factory lines was longer than expected.

Olympic Industries first undertook a trial production run, which facilitated addressing numerous issues including better defining the roasting duration of the lentils and chickpeas, adjusting the overall moisture content and size of ground particles, adjusting fluidity to change the products’ stickiness, shifting the recovery rate, and improving extrusion. Simultaneously the partners and company leadership refined hygiene protocols and completed microbiological testing on initial product during the trial.

LNS packaging. Designing single serving sachet packaging for SQ-LNS (i.e. 25–28 g/sachet) required considering the type of material, product stability, labeling, cost, transportation, and storage. After much discussion, the partners finalized sachet packaging material as triple laminated with low-density polyethylene as a compromise between cost and ability to extend shelf life. As required by Codex Alimentarius, a set of food standards established by the World Health Organization (WHO) and the Food and Agriculture Organization of the United Nations (FAO), the partners included labeling of ingredients on each individual serving sachet. Additionally, they decided on the size of each sachet, the number of sachets per box (labelled with ingredients and storage instructions), and the number of boxes in a shipping box. Importantly, the partners ensured that the sachet marked an ‘easy tear’ location after hearing that other regional programs learned during implementation that community members resorted to using razors when unsure how to open the product, which could introduce contamination.

Maintaining quality and hygiene standards. Adhering to Codex Alimentarius requirements was important for LNS designed for young children. Thus, the partners carefully scrutinized manufacturing standards on quality and hygiene and put quality controls in place. The various elements of this component included a) factory line cleaning and hygiene protocols, b) sourcing of affordable but high-quality local ingredients, c) batch-specific microbiological and contaminant testing and approval, and d) shelf-life testing and expiry date labeling.

Olympic Industries management instituted strict protocols for their staff to support proper hygiene. Between batch productions, factory staff rigorously and thoroughly cleaned the machinery to remove any contamination that could have built up. One challenge in Bangladesh was to procure ingredients of consistently high quality. Further, it was not possible to trace the purchased product back to a specific region or farm, as a vestige of the complex, multistep supply chain system in rural, agricultural Bangladesh. The small scale of production by family farms in Bangladesh meant that a bulk purchase of 300 kilograms of lentils was likely composed of product from more than 10, if not a multiple of 10, farms. This aspect of the supply chain highlights the challenges producers may face when trying to procure ingredients of consistent quality at scale. To address this challenge, Olympic Industries routinely procured from a wholesale market three hours away from Dhaka. During each market visit, the procurement lead identified the vendor with the highest quality product. Of note, milk powder and the vitamin and mineral premix were sourced outside of Bangladesh.

Related to the raw ingredients, one of the first challenges the partners faced was that the first few batches had higher than permissible levels of mold contamination. This resulted in increasing rigor in purchasing quality ingredients and in pre-picking, -sorting, and -sifting lentils and chickpeas. These consistent actions helped reduce the risk of contamination.

One of the most persistent challenges the partners faced was achieving and maintaining appropriate microbiological and contaminant standards in the final product. Due to not having local laboratories that could certify the hygiene quality of the products, Olympic Industries and the partners had to ship samples from each lot to the SGS Laboratory in France for testing and certification. At the time of this activity the icddr,b laboratory followed ISO 17025 but had not concluded global accreditation. This testing and certification process in France was time consuming requiring two or three days for sample shipment to SGS and at least 10 days for analyses and report generation and cost over $1000 per sample. Thus, the lack of availability of a credible local laboratory for contaminant testing increased the cost of and time required for testing and ultimately production at scale. This type of delay regularly challenged the supply pipeline and required a dedicated budget.

Another significant hurdle was not having Codex-specified cut-offs for microbiological levels at the time this project was undertaken. Fortunately, we were able to apply WFP’s established cut-offs. The food technologists discussed each report and determined acceptability of the lot. In the event a batch was classified as contaminated, Olympic burned the whole batch, often about three metric tons worth (this occurred for four batches in all) and started over again with production and SGS contaminant testing. Contamination was largely due to elevated total plate count, total coliforms, molds, and Bacillus cereus either unexplainable or possibly resulting from a number of factors ranging from a contaminated batch of ingredients to a protocol breach with staff hygiene practices.