Situation analysis of procurement and production of multiple micronutrient supplements in 12 lower and upper middle‐income countries

Abstract Globally, there are few vitamin and mineral ingredient manufacturers. To support local, in‐country or regional procurement and production of multiple micronutrient supplements (MMS), the following production scenarios are possible: (a) straight ingredients of vitamins and minerals forms imported or locally produced that are mixed, tableted, or encapsulated and packaged by a local manufacturer; (b) import or local production of a vitamin and minerals premix that is tableted or encapsulated and packaged locally; (c) import of a bulk, finished product (tablets or capsules) that is packaged and branded; and (d) or import of a branded packaged product. This paper is a situation analysis of the market, manufacturing, and policy factors that are driving the production of MMS in 12 lower and upper middle‐income countries. Key informants completed a self‐administered structured questionnaire, which examined the local context of products available in the market and their cost, regulations and policies, in Brazil, Colombia, Guatemala, Mexico, Peru, Bangladesh, India, Vietnam, Ghana, Kenya, Nigeria, and South Africa. Our study found that although most countries have the capacity to produce locally MMS, the major barriers observed for sustainable and affordable production include (a) poor technical capacity and policies for ensuring quality along the value chain and (b) lack of policy coherence to incentivize local production and lower the manufacture and retail price of MMS. Also, better guidelines and government oversight will be required because not one country had an MMS formulation that matched the globally recommended formulation of the United Nations Multiple Micronutrient Preparation (UNIMMAP).

for World Health Organization (WHO) to issue a global recommendation, WHO did recommend in their antenatal guidelines that "policymakers in populations with a high prevalence of nutritional deficiencies might consider the positive effects of multiple micronutrient supplement on maternal health to outweigh the disadvantages, and might choose to give multiple micronutrient supplement which include IFA" (WHO, 2016).
As policy leaders weigh the health benefits of MMS during pregnancy, programming for large-scale micronutrient interventions, such as MMS, must address several upstream processes for delivery and scale up (Menon et al., 2014); considerations related to the local procurement of MMS are crucial to guarantee a stable, affordable, and high-quality supply of MMS (Bhutta, Imdad, Ramakrishnan, & Martorell, 2012;Shrimpton, Huffman, Zehner, Darnton-Hill, & Dalmiya, 2009). The supply of MMS in country will be influenced not only by local manufacturing capacities but also by the regulatory policies that enable manufacturing scenarios, as well as the capacity of countries to monitor the quality of MMS along the value chain.
Globally, there are few vitamin and mineral ingredient manufacturers. For in-country or regional procurement of MMS, there are four possible scenarios (Table 1). Scenarios A and B require local manufacturing capacities for blending and mixing; scenario C is likely to be found in countries with lower manufacturing capabilities; and scenario D in countries with no capacity or with small consumer markets.
This paper explores these four scenarios by examining the manufacturing capacities and policies for MMS production that exist in a sample of 12 lower and upper middle-income countries (hereafter middle-income countries, MIC).

| METHODS
We selected 12 MIC and 3 high-income countries (HIC) that were included for each region as a comparison. MIC were selected if they met two criteria: (a) presence of DSM staff to facilitate data gathering, and (b) countries with moderate to poor maternal health indicators during pregnancy as indicated by low birthweight or maternal anaemia in pregnancy (Table 2). We selected the following countries: Brazil, Colombia, Guatemala, Mexico, Peru, (for Latin America [LATAM]), Bangladesh, India, Vietnam (for Asia), Ghana, Kenya, Nigeria, and South Africa (for sub-Saharan Africa). Because of the strong manufacturing and quality control (QC) policies for MMS in the United States, Japan, and Germany, and their potential as regionally suppliers for MMS, these countries served as comparison for LATAM, Asia, and Africa, respectively (Wallace, 2015). A 40-question, self-administered questionnaire was designed to collect data on four key areas: product availability in local markets, production capacity, taxes and commercial policies, and regulatory landscape ( Figure 1). The questionnaire was sent to key informants (employees or key stakeholders) of DSM with knowledge about the local consumer market and regulatory policies. Data on policies and regulations were augmented by key informants in the regulatory affairs unit of DSM with knowledge in these markets. Information from this questionnaire was verified with each key informant and, where possible, cross-checked via web search or with information available to the United Nations Children Fund (UNICEF).
The data collection was restricted to supplements containing vitamins and minerals in tablets or capsules for adults (either targeted to male or female needs) or for women who were pregnant or lactating. We excluded supplements for children, products sold as syrups and powders, due to poor stability and palatability of these products, and those that contained herbal ingredients for lack of scientific evidence in using these products for improved pregnancy outcomes. For countries with high manufacturing capabilities and numerous MMS brands on the market, we restricted the analysis to the top 5 selling brands/companies. For these countries, we assumed that local production or availability of MMS could be achieved via four scenarios as described in Table 1. In situations where these scenarios did not apply, we inquired about MMS production locally, and policy and market factors.
Pictures of products and product labels were collected in each of the 12 MIC. Only product label information was analysed. The

Key messages
• Despite the availability of MMS in the marketplace, not one survey country has a product that matches the internationally recommended UNIMMAP formulation.
• In some countries where imported brands are preferred by consumers, local companies will need to invest in marketing to compete with international brands.
• Regional trade agreements or technology transfer should be leveraged to ensure affordable local supply of high-quality MMS.
• Monitoring and surveillance systems need to be strengthened, and local food and drug authorities need better oversight on the MMS quality and potency.
• To achieve policy coherence, national governments may need to consider consumer-facing and manufacturingfriendly policies to ensure affordable supply of MMS. The term straight ingredients is preferred as it includes both active pharmaceutical ingredient (vitamin) and its forms for dry tableting.
UNICEF/WHO/United Nations University (UNU) international multiple micronutrient preparation (UNIMMAP) formulation was considered the benchmark for ingredients and recommended daily dose due to the scientific evidence of its efficacy on pregnancy outcomes (UNICEF, WHO, and UNU, 1999;West et al., 2015).  (Tables 3a and 3b). Seventy-one 2012. Low birth weight is a weight at birth of less than 2,500 g irrespective of gestational age. N/A: data not available.
FIGURE 1 Conceptual framework identifies four domains describing the conditions for local procurement and production of multiple micronutrient supplement within a given country products were collected from 12 MIC and 12 products from 3 HIC.
These products belong to either Category I or II, as defined in Section 2. There were no products belonging to categories III or IV.  Mean price (in USD) ± SD (unweighted). Category I-less than 10 ingredients and dosage (80% or more of the recommended dosage) consistent with UNIMMAP; Category II-at least 10 ingredients and dosage consistent with UNIMMAP.
b Although the total number of products sampled in the Africa region was 33, for one product we were unable to verify if it was local or imported.  Note. MMS = multiple micronutrient supplement. a Mean price (in USD) ± SD (unweighted). Category I-less than 10 ingredients and dosage consistent (80% or more of the recommended dosage) with UNIMMAP; Category II-at least 10 ingredients and dosage consistent with UNIMMAP.
b Although the total number of products sampled in the Africa region was 33, for one product we were unable to verify if it was local or imported. similar products sold in the five Latin American countries (USD 15.20 ± 6.82 for a pack of 30 capsules or tablets). However, the price of MMS from the United States varies considerably depending on the retail location. For example, in drugstores, wholesale stores and ecommerce, the price per capsule or tablet is significantly lower (nearly 75%) than the range mentioned above. Also, these supplements, in both Categories I and II, in the United States lack significant amounts of one or more of vitamins A, B1, B2, and B3 (0% to 60% of the recommended UNIMMAP dosage).

| Africa
Both local and imported MMS in blister or bottle packaging are found in the four African countries surveyed-Ghana, Kenya, Nigeria, and South Africa. MMS can be purchased through several channels-pharmacy, supermarket, and "mom and pop" stores, and directly sold or home delivered to consumers in these four countries. Mean retail prices for a pack of 30 supplements are reported in Table 3a. The four MMS from Japan belonged to Category I and retailed at USD 6.11 ± 6.23 for 30 tablets or capsules.
We also observed some general trends for pricing of MMS across regions (Tables 3a and 3b). In Africa and Asia, imported products tended to be more expensive than local products. This was not observed in LATAM, where imported products seemed to be competitively priced. However, the regression analyses show that in general, we can expect that imported products will be significantly more expensive by USD 4.25 per package of 30 tablets or capsules than local brands (Table 4). In LATAM and Africa, products were more expensive if they had more ingredients and higher dosage, though this difference was not statistically significant (

| Product registration
We ascertained whether MMS for adults or pregnancy were registered as food or drug in each country. Drugs undergo stringent controls for quality, and the cost of registration is high. Foods, on the other hand, are usually easier and cheaper to register in countries. The preference for drug over food registration might signal greater degree of government oversight/monitoring from production to market place.
With regard to cost of registration, there were major differences between high-income countries (HIC) and MIC, with Germany and Japan registration costing tens of thousands of dollars more than their regional analogues (Table 5).

| Import taxes and exemptions
The import taxes reported by informants varied widely, with African countries having some of the highest taxes for each of the four scenarios (Table 6). Ghana, for example, reportedly charges a 38-40% tax for all nutritional products, whereas Peru imposes zero tax on straights or premix blends and a lower tax on imported, finished products. In some countries, the import tax varies by nutritional product. India applies a 7% tax to vitamins only, whereas minerals, vitamin-mineral blends, and omega fatty acids are taxed at 30% of the import value. A similar, but less drastic scenario, is found in Guatemala, Mexico, and Colombia (Table 6). 3.4 | Manufacturing scenarios and regional trade

| Preferred scenario
We asked key informants to discuss which scenario ( In Kenya, Ghana, and Nigeria, although some local manufacturing from straights and premix does take place, the preferred scenario is imports of bulk products (Scenarios C) or branded supplements (Scenario D) for several reasons identified by our key informants: (a) manufacturing capacity is low in these countries, (b) consumers prefer the branded supplements, and (c) there is a small consumer market for these products. In contrast, South Africa imports straights, premix, or bulk products, as the country has a higher manufacturing capacity,  and takes advantage of free trade agreements with the EU. In the HIC Germany, the preferred scenario is to import or produce straights.
Asia, Japan, and Vietnam prefer to import straights or premix because of lower import taxes and high manufacturing capacity for blending. Bangladesh prefers to import straights.
India locally manufactures most vitamins (i.e., vitamin A, C, D, E, B2, B3, and B12, biotin, and folic acid) and manufactures straights for all minerals. When other vitamins are needed, the straight scenario would be used.

| Regional trade
We also explored whether trade agreements could facilitate regional production of MMS, if local production was not possible. Of the countries with lower manufacturing capacities, only Guatemala and Peru had regional trade agreements that would facilitate acquiring products produced regionally. In East Africa, regional MMS production may be facilitated with the common trade tariff among these countries.

| Taxes and commercial policies-Incentives and disincentives
We also sought to understand how other economic policies might encourage (i.e., price subsidies and tax exemptions) or discourage (i.e., price ceilings) local manufacturing or may potentially limit consumer demand (i.e., commercial taxes). These are described below.

| Tax exemptions
Very few countries offer tax exemptions. Colombia offers exemptions for straights and premix blends of vitamins and on bulk products or branded supplements that are imported as drugs. In India, there are excise duty and sales tax exemptions for products manufactured in duty free zones.
While in the United States, it is the state authorities that may grant sales tax exemptions to MMS products provided through social programmes.

| Commercial taxes
We also asked about commercial taxes because these are indirect taxes ultimately borne by the final consumer at point of purchase.
Countries reported three types of commercial taxes: sales tax, value added tax, and excise tax. These commercial taxes are part of the retail price and are regressive policies because they place a greater economic burden on low-income consumers compared to higher income groups.
Excises are inland taxes typically imposed on producers and manufacturers and ultimately passed onto the consumer. The commercial taxes varied widely across the regions and within regions ( Table 7) with some of the highest taxes paid on the imported, branded products. No data were available for Ghana or Japan.

| Price ceilings
There was little use of price ceilings for MMS across the countries, except for Brazil and India. In Brazil, MMS is registered as drug so it We found that all countries require at least the manufacture and expiry date and list of ingredients, as drug facts or nutrient facts, depending on how the product is registered in each country (see Table 5). Ingredient claims are voluntary across all countries and regions.
Government endorsement on label is not allowed in Asia, Africa, or Peru. Health claims are allowed to appear on the labels of MMS sold in Guatemala, Peru, Bangladesh, India, and Vietnam. In Mexico and Colombia, health claims were not permitted on food supplements. In Ghana, Kenya, Nigeria, and South Africa, no disease reducing claims could be made, but ingredient and content claims are mandatory.

| Local manufacturing capacity
Each of the countries surveyed indicated presence of local manufacturing capabilities of MMS. Local manufacturing capacity is defined as capacity of production facilities established in the country either by a multinational (mostly companies from the United States or Europe), regional (on the same continent), or a native company. These production facilities either purchase straights (Scenario A) or premix blends (Scenario B) to manufacture MMS tablets or capsules. Factory production cost for MMS by countries or regions are specified in Table 9.

| Latin America
Overall in the LATAM region, retail sales of MMS grew by a cumulative annual growth rate of 14% from 2008 to 2013. In Brazil, Colombia, Guatemala, Mexico, and Peru, 197 million MMS were sold in 2014 by 13 local manufacturers (Strobel, 2014). Except for Mexico, each of the native companies had at least one leading brand. In Mexico, imported brands seem to be preferred. Most of the brands sold in LATAM are native companies except for those in Guatemala and one company in Colombia, where the production facilities were established by either a multinational or regional company. Ten companies import straights, with capabilities for blending vitamins and minerals, tableting or capsuling, and packing. All of the brands were produced in facilities that are WHO-GMP certified. Total cost of production of MMS is highest in Brazil due to high import taxes on supplements. Mexico and Colombia are seen to be agile and competitive in manufacturing MMS. Except for Peru, all local manufacturers indicated an underutilized capacity and could increase production should there be a rise in demand for MMS.
In comparison, the Unites States has sales, on an average, of 200 million capsules or tablets of MMS each in the year 2014. It has several local manufacturers, and the cost of production in the United States is comparable to that in LATAM at USD 0.01 or less for a tablet or approximately USD 0.30 for 30 tablets or capsules (Table 9). However, the cost of production should be evaluated along with the quality of production because it affects the potency and performance of the MMS. We did not find any objective data on whether the quality of production in LATAM is comparable to that in the United States.

| Africa
There is local manufacturing capacity in Ghana for MMS in tablet form.
All MMS brands are imported from the United Kingdom. Information for local manufacturing capacity in Nigeria was not available, despite being able to find one local brand. In Kenya, five WHO-GMP certified manufacturers produce MMS. Four of them import ingredients indicating the existence of premixing and tableting or capsuling capabilities in Kenya. These five manufacturing facilities are stated to be underutilized and have the capabilities to expand. For very large scale mass production, capital goods (e.g., machine or equipment) would need to be imported. South Africa also has local manufacturers with premixing and tableting or capsuling capabilities that are underutilized.
The largest production facility in South Africa is a multinational, which produces 3 million tablets of a leading brand of MMS for pregnancy.

| Asia
Overall in the Asia Pacific region, retail sales of MMS grew by a cumulative annual growth rate of 12% from 2008 to 2013. In Bangladesh, India, and Vietnam, there are 15 local manufacturers of MMS that produce 14 leading brands. Among these 15 local manufacturers, only two were multinational companies, one in India and one in Vietnam, which The shelf life varies from product to product, but the minimum and most common is 2 years but may be up to 3 years. South Africa

50%
As close to 100% as possible As close to 100% as possible   (Table 9).
However, we did not find data that compared the quality of finished products among countries in the region.
Indian and Bangladeshi manufacturers procure straights and those

| QA and QC
Although MIC expressed several challenges along the MMS or IFA value chains, no major barriers were expressed in HIC (i.e., Japan, United States and Germany) as long as actors along the value chain invested in high-quality ingredients, infrastructure to maintain requirements for temperature and humidity, and a skilled workforce.

| Latin America
In Brazil, Colombia, Guatemala, Mexico, and Peru, high quality and consistent supply of ingredients (e.g., folic acid) was cited as a major barrier for sourcing. Ensuring homogeneity during the premixing or blending process was considered to be a challenge and hence requires investment in skilled workforce to ensure good manufacturing practices are being followed. There was not enough in-country information on QC during transportation and in warehouses. Key informants reported that lack of capacity or knowledge for conducting quality checks on finished MMS is a major challenge for Colombia and Guatemala.  (Galloway & McGuire, 1994), weak supply chain management systems (Victora et al., 2012), and consumer purchase capacity will continue to limit uptake (UNICEF, WHO and UNU, 1999).

| Africa
Upstream policies to improve affordability include setting competitive and sustainable price ceilings coupled to import subsidies, import tax exemptions, commercial tax exemptions, company tax subsidies, or a combination of these. Taxes borne by the manufacturer create an incentive structure that makes MMS a premium product. In Colombia and India, for example, our data show that some nutritional ingredients, such as omega-3 fatty acids, which have been shown to be beneficial during pregnancy for reducing risk on maternal depression and poor fetal neurodevelopment (Su et al., 2008), are subject to much higher import taxes compared to vitamins. Even in countries with strong local manufacturing capabilities, an affordable retail or wholesale selling price may depend on how governments structure import tax for straights, premix, and bulk products, as the manufacture is more likely to pass on the tax cost to the consumer (Food and Agriculture Organization, 2010).
Although this study provides a comprehensive overview of the manufacturing and policy factors, further in-country examination will be required to improve the robustness of the data because we did not (a) assess the quality of products on sold in the market (USDA, 2015), (b) verify our key informant responses with national regulatory bodies, or (c) use representative sample of products. Furthermore, despite numerous follow-ups with key informants, we still received incomplete data for some countries, namely, Germany and the United States. Missing data were common for market information, installed capacity, and QA/QC along the value chain. We mitigated these missing data points by triangulating with other data sources, such as IMD database and other key informants within DSM, and additional crosscheck by UNICEF on the regulatory data. Despite the limitations, we believe this to be an adequate representation of the key market and policy factors that are shaping the MMS availability in the 12 MIC countries that we sampled.
In contrast to MMS, WHO does recommend daily IFA supplement during pregnancy for reducing the risk of low birth weight and maternal anaemia or iron deficiency. There is nonetheless a significant lack of information on supply chains for IFA (Sununtnasuk, D 0 Agostino, & Fiedler, 2016). Although we did not specifically collect information on the production and procurement of IFA, a number of our informants reported problems for sourcing (raw material quality and blending), production (good manufacturing practices and QC), distribution (adequate storage facilities), and monitoring (shelf life). Therefore a number of these challenges would apply to both IFA (WHO 2016, UNIDO 2014) and MMS, although a situation analysis of IFA supply chain would be needed to estimate the factors that might impair the implementation of IFA. In one study that looked at the bottlenecks in the IFA supply chain in Senegal, stock outs at local level were common due to poor ordering processes and lack of funds for timely orders (Gueye, Pendame, Ndiaye, Diop, & Daff, 2015). Taken together, our data and the literature seem to suggest that procuring MMS in low-income countries is likely to be more challenging than what is reported here for lower-middle income countries.
In conclusion, this situation analysis offers a framework on the manufacturing and policy enablers and barriers for MMS local production. Importantly, a potential scale up from iron-folate to MMS will need to be viewed holistically considering manufacturing, policy, and human capacity/skill factors, as these together will influence the implementation, sustainability, and affordability of MMS in MIC.
paper; they do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated or the decisions, policies, or views of the World Health Organization. The opinions expressed in this publication are those of the authors and are not attributable to the conveners.

CONTRIBUTIONS
ECM, KB, and KK conceptualized the study design and questionnaire.
ECM and KB organized the data collection and led the analysis and interpretation of findings and drafting of the manuscript. KGvZ assisted in data organization and drafting of manuscript. KK, GS, RK, and AF helped in interpretation of findings and contributed to the critical revision of the manuscript for final submission. All authors approved the draft.