Sensory panel results of a dried fish powder supplement among caregivers and young children in Zambia

Objective: To evaluate the acceptability of traditional Zambian dishes fortified with Complementary Food for Africa+Dried Fish Powder (ComFA+Fish), a locally sourced protein/micronutrient blend designed to impact nutrient deficiencies among infants and young children (IYC) and improve pregnancy and birth outcomes among women of reproductive age (WRA). Design: During two sensory panels, caregivers evaluated: (1) the acceptability of four ComFA+Fish dishes for household consumption, including fortified chibwabwa fisashi, savory Kapenta chutney, fortified complementary maize porridge and fortified bean-vegetable soup and (2) whether their IYC found the fortified complementary maize porridge acceptable. Setting: Lake Kariba, Southern Province, Zambia. Participants: Women of reproductive age (n 42) and their IYC aged 6–11 months (n 16) and 12–23 months (n 26) were recruited from fishing villages in Gwembe, Siavonga and Sinazongwe District. Results: A majority of caregivers extremely liked/liked the: (1) fortified chibwabwa fisashi’s sensory attributes (94·7 %), convenience (92·8 %) and overall acceptability (100 %); (2) savory Kapenta chutney’s sensory attributes (81·8 %), convenience (92·8 %) and overall acceptability (100 %); (3) fortified complementary porridge’s sensory attributes (83·5 %), convenience (90·5 %) and overall acceptability (88·1 %) and (4) fortified bean-vegetable soup’s sensory attributes (66·8 %), convenience (87·5 %) and overall acceptability (87·5 %). Further, a majority of caregivers evaluated the fortified complementary porridge as highly acceptable to their IYC. Conclusions: Results suggest that ComFA+Fish is strategically well placed to fill nutritional gaps among IYC and WRA in Zambia and has the potential to be scaled across sub-Saharan Africa.

chronic malnutrition in young children often manifests as stunting (height-for-age<−2), it can also have long-term effects on cognitive development, general health/ability to work and educational attainment that can impact economic productivity and health (e.g.maternal reproductive health) into adulthood (6) .
Although Zambia has made significant progress in reducing child malnutrition in the last two decades (7) , rates of stunting and wasting remain high (particularly in rural areas) and Zambia ranks 140th among 163 countries in meeting the sustainable development goals (8,9) .Among Zambian children under 5 years, 35 % are stunted as a consequence of chronic or recurring malnourishment and 4 % are wasted as a consequence of acute malnourishment (7) .The first 1000 d of lifewhich is defined as the critical period of cognitive and physical development from conception until a child reaches 2 years of ageis identified as a key intervention point for supplementing the diets of infants and young children and of pregnant/lactating women in LMIC, particularly in Africa where stunting prevalence is highest (6) .However, as Eaton et al. (2019)  report in a recent systematic review, a majority of such interventions implemented in LMIC have been cerealbased (v.based on animal-source foods, including aquatic animal-source foods), and their impacts on childhood malnutrition outcomes remain ambiguous (10) .
Dried fish powder is an important but underutilised aquatic animal-source food to address childhood stunting and hidden hunger (i.e.micronutrient deficiencies) linked to extreme poverty, lack of dietary diversity, low consumption of protein and over-dependence on the high-phytate cereal-based diets of many vulnerable households across sub-Saharan Africa (11)(12)(13)(14) .Indeed, leading international organisations (e.g.FAO of the UN, WHO, U.S. Agency for International Development) as well as leading governmental organisations within Zambia (e.g.Ministry of Health, National Food and Nutrition Commission, Ministry of Fisheries and Livestock) are increasingly promoting the use of dried fish powder to meet the nutritional needs of infants and young children and of pregnant/lactating women in low-resource households (15)(16)(17)(18)(19)(20)(21) .Such support is well founded, as a number of current studies have clearly documented the high nutritional value of dried fish powder made from Kapenta (Limnothrissa miodon and Stolothrissa tanganicae) (17)(18)(19) .
For example, Byrd and colleagues found that a 10 g serving of dried Kapenta powder provided 20 % or more of the recommended daily intake for Ca and 37 % or more of the recommended daily intake for DHA for 6-23-month-olds (17) .They also found that the nutrient density of dried Kapenta powder was similar in Fe and Zn and higher in Ca and DHA to that of a small-quantity lipid-based nutrient supplement plus (SQ-LNS-plus) (17) .Likewise, Haug and colleagues found that a small quantity of dried Kapenta powder met the recommended daily intakes of nitrogen, Ca, phosphorus, Mg, Fe, iodine, Se, Zn and fluoride for 12-36-month-olds (18) , while Nölle and colleagues found that 6-7 g (∼1•5 tsp.) of dried Kapenta powder provided the recommended daily intake of vitamin B 12 .Such findings on the nutrient density of dried Kapenta powder are important, given that the gastric capacity of infants and young children is quite limited and their meals, therefore, should be nutrient dense (19) .Such findings also suggest that fortifying traditional dishes with nutrient dense dried fish powder is a formidable resource for enhancing the diets of nutrient-deficient infants, young children and other household members (20) .
The need for a low-cost, locally sourced means by which to increase the nutrient density of traditional dishes across sub-Saharan Africa is clear given that (non-fortified) maize-based foods are highly prevalent in this region, which consumes 21 % of the maize produced worldwide (21,22) .Indeed, considered the national staple food of Zambia (23) , maize is inexpensive enough that many Zambians eat Nshimaa traditional dish of ground white maize meal boiled with water into a stiff smooth pasteat 2-3 of their daily meals.Nshima is also commonly used as a complementary food for infants and young children by thinning the maize paste (usually with water among poorer households) into a porridge of the desired consistency.There is a need to fortify these foods given that the non-fortified maize meal traditionally used to make Nshima and complementary maize porridge is lacking in protein and micronutrients and can reduce absorption of other nutrients.

Complementary Food for AfricaþDried Fish Powder
Although multiple micronutrient powdersinternationally standardised prepackaged sachets of vitamins and minerals produced by a handful of global suppliers and distributed by donors such as UNICEF (24) have helped reduce global rates of stunting (25,26) and anaemia (27)(28)(29) among vulnerable children, they have had less impact on protein malnutrition (27)(28)(29)(30)(31)(32) .In contrast, whole pelagic small fish provide protein, fat, ∼15 essential micronutrients and vitamins A, C, B 12 , E and D even when consumed in small quantities (17)(18)(19)(33)(34)(35)(36) . In orde to help fill both protein and micronutrient gaps among vulnerable children and other household members, we conducted two sensory panels of traditional Zambian foods fortified with a locally sourced, low-cost, high-quality protein/micronutrient blend whose primary ingredient is whole pelagic small fish that are dried and ground into a fine powder.We call this protein/ micronutrient blend, Complementary Food for Africaþ Dried Fish Powder (ComFAþFish).
As this study was conducted at Lake Kariba, which is a major source of Kapenta, we used this fish as the primary ingredient of the ComFAþFish protein/micronutrient blend.One of the most popular freshwater small fish in Zambiaparticularly for low-resource and vulnerable householdstiny Kapenta are a nutritional powerhouse (17)(18)(19)(33)(34)(35)(36) and 'major source of fish protein for all levels of society and make up a significant part of animal protein consumed in the country' (37,2) . Kapent belong to the same family as herring and sardine and because these pelagic fish are small (∼10 cm long), they are dried whole, which further concentrates their nutritional density.An appetising and affordable staple food throughout Zambia and across sub-Saharan Africa for infants, children and adults, pelagic small fish such as Kapenta, Chisense (Potamothrissa acutirostris and Poecilothrissa moeruensis) and Dagaa (Rastrineobola argentea) (38,39) can be consumed fresh, sun dried oras is the case with ComFAþ Fishsun dried, dry roasted and ground into a fine powder that is a rich source of protein and micronutrients.Dried fish powder is produced at the household level in small batches using a mortar and pestle and is also widely available at local markets, where it is commercially produced using a mechanised grinder (e.g. a groundnut grinder) and sold in large batches and as sachets.
As adding a supplement to a food has the potential to change the sensory properties of that food, our objectives in the present study were to: (1) develop four ComFAþFish dishes based on traditional Zambian dishes (including fortified complementary maize porridge); (2) conduct Sensory Panel I among forty-two caregivers to evaluate whether the four ComFAþFish dishes can be produced at the household level and are deemed acceptable for household consumption by the caregivers and (3) conduct Sensory Panel II among the same forty-two caregivers and their children aged 6-23 months in order for caregivers to evaluate their child's global liking of the ComFAþFish complementary maize porridge.

Methods
Kapenta dried fish powder: analysis of nutrient composition Following a four-step sampling protocol, we collected four individual samples of whole dried Kapenta sourced from Lake Kariba from four separate small-scale vendors at open markets in Lusaka (see Fig. 1).Each individual sample weighed a minimum of 0•5 kg, and each sample was inspected for wholesomeness (i.e.no unpleasant odour, visible signs of decay or degradation associated with spoilage).Emulating the traditional process used in Zambian home kitchens, each sample of dried Kapenta was dry roasted in batches in a large pan, and the four samples were combined into one 2-kg sample.This sample was taken to a small-scale miller at an open market for grinding into Kapenta dried fish powder.After grinding, the 2-kg sample of dried fish powder was shipped to the United States, where it was divided and distributed to two accredited commercial labs for analysis, including Mérieux NutriSciences and the Mississippi State Chemical Laboratory.

Sensory panels: study design
The study was conducted in the town of Siavonga, which is ∼195 km from the capital city of Lusaka and abuts the vast artificial water body of Lake Kariba (see Fig. 2).Caregivers were recruited from fishing villages in Gwembe, Siavonga and Sinazongwe Districts in Zambia's Southern Province.The caregiver-child pairs convened in Siavonga and all caregivers were provided with childcare products (e.g.diapers, sanitary wipes and hand sanitiser) and transportation stipends to travel from their villages to the study site.Caregiver-child pairs from Gwembe and Sinazongwe were also provided with lodging at a Siavonga hotel for two nights as Gwembe and Sinazongwe are each more than 250 km from Siavonga.

Eligibility
To be eligible to participate in Sensory Panels I and II, women were required to self-reported that: (1) they were between 18 and 49 years of age; (2) they resided in Gwembe, Siavonga or Sinazongwe Districts; (3) they had no known allergy to dried fish powder, groundnut powder, orange-fleshed sweet potato, peanut/nut oil or vegetable oil; (4) their child was 6-23 months old; (5) their child had no known allergy to dried fish powder, groundnut powder, orange-fleshed sweet potato, peanut/nut oil, or vegetable oil; (6) their child had consumed fish and/or fish-based foods prior to recruitment into the study and (7) their child had consumed complementary maize porridge prior to recruitment into the study.

Preparatory phase
To prepare for Sensory Panels I and II, we conducted group interviews with women in rural fishing villages and other local experts in Zambia's Southern Province to identify: (1) staple traditional dishes that were regularly consumed at the household level (including by infants and (2) locally sourced ingredients that were commonly used in each of these staple traditional dishes and (3) whether dried fish powder was currently a common ingredient in these staple traditional dishes (including when the dishes were prepared as complementary foods for infants and young children).Based on this information, we chose three staple traditional dishes suitable to fortify with ComFAþFish (i.e.chibwabwa fisashi, complementary maize porridge and bean-vegetable soup) and a novel dish (i.e.dried Kapenta chutney) whose primary ingredient is whole dried Kapenta.The four ComFAþFish dishes that we developed use locally sourced Kapenta and other locally sourced staple ingredients (e.g.groundnut powder, orange-fleshed sweet potato powder and pumpkin leaf powder).In order to ensure that each ComFAþFish dish could be produced at the household level, we pre-tested each recipe in Lusaka prior to implementing Sensory Panels I and II at the study site.During the pre-testing process, the ingredients for each dish were purchased at local markets, and a large batch of each dish was prepared in a standard Zambian kitchen using standard cooking tools (mortar and pestle, spoons, stove, etc.) and taste tested by the study team.The person who supervised preparation of the ComFAþFish dishes during the pre-testing process also supervised preparation of the ComFAþFish dishes at the study site.All ingredients used during Sensory Panels I and II were purchased at local markets at the study site.

Sensory panel I assessment
For Sensory Panel I, we developed a technical lexicon for use by caregivers to evaluate five sensory attributes and two additional attributes of the four ComFAþFish dishes.Development of this lexicon and the panel's procedures were informed by Madrelle et al. (2017) and others (40)(41)(42)(43) and the ComFAþFish dishes that were assessed included: (1) fortified chibwabwa fisashi (a savory dish of dried dark green leafy vegetables); (2) savory Kapenta chutney (made from whole dried Kapenta cooked with tomato, onion, green pepper and spices); (3) fortified complementary maize porridge and (4) fortified bean-vegetable soup.Next, we simplified the technical definitions of all seven attributes, translated them from English into Tonga and pre-tested them among Tonga-speakers for clarity of language prior to use by the caregivers.For example, 'distinct aromatic notes associated with the sample' were simplified to 'aroma or smell of the food', and 'distinct flavour notes associated with the sample' were simplified to 'flavour/taste of the food'.
For Sensory Panel I, caregivers were provided with a separate assessment form for each of the four ComFAþFish dishes and were instructed to score the seven attributes of each dish based on the simplified lexicon using a five-point hedonic scale where 1 = extremely disliked and 5 = extremely liked.For this scale, the assessment form included black-and-white 'smiley face' emojis that correlated with each point of the five-point scale.Participants were instructed to mark the appropriate smiley face emoji with an 'x'.All four ComFAþFish dishes evaluated during Sensory Panel I were prepared onsite at separate cooking stations and served to the participants immediately.See Appendix 1 for the technical and simplified lexicons for caregivers.

Sensory panel II assessment
For Sensory Panel II, we developed a three-item set of descriptors for use by caregivers to assess their child's global liking of ComFAþFish complementary maize porridge (see Appendix 2).Development of these descriptors and the panel's procedures were informed by Madrelle et al. (2017) and others (40)(41)(42)(43) .We translated the three items from English into Tonga and pre-tested them among Tonga speakers for clarity of language prior to use by the caregivers.The fortified porridge was prepared onsite and served immediately to each caregivers, who fed the porridge to their child.Caregivers were instructed to feed the fortified porridge to their child during three consecutive tasting intervals (Times 1-3).The serving size of the fortified porridge was 50 g (∼8 oz.), and all children were fed their portion of the porridge using identical infant spoons and identical 8 oz.clear plastic cups.Caregivers were allowed to offer their child sips of bottled water between each of the three tasting intervals.
For item 1 of Sensory Panel II (see Appendix 3), we used the same set of black-and-white smiley face emojis used for Sensory Panel I.For this item, we instructed caregivers to assess their child's global liking of the fortified porridge using a set of five descriptors of positive/negative behaviours exhibited by their child that correlated with a five-point hedonic scale where 1 = child extremely disliked the porridge and 5 = child extremely liked the porridge.For example, caregivers were instructed to indicate that their child extremely disliked the fortified porridge if the child 'spit out the food, pushed the food away, vigorously turned their face away, stopped eating, became fussy, arched their back or pulled their body away when food was offered'.For item 1, caregivers were instructed to offer their child an initial set of three consecutive teaspoonfuls of the fortified porridge at Time 1 (T1), a second set of three teaspoonfuls of fortified porridge at Time 2 (T2) and a final set of three teaspoonfuls of fortified porridge at Time 3 (T3).After each tasting interval (T1, T2 and T3), caregivers were instructed to evaluate how much their child liked the fortified porridge using the five-point hedonic scale.
For item 2, we created black-and-white 'bowl/portion' images and caregivers were instructed to select the bowl/portion image that best correlated with the actual amount of fortified porridge that their child consumed during the entire meal (i.e. the sum of T1 þ T2 þ T3).This was assessed at the end of the meal, and caregivers were instructed to use a five-point hedonic scale where 1 = child consumed less than 1/4 of their portion and 5 = child consumed their entire portion.For item 2, the bowl/portion images correlated with each point of the five-point scale (see Appendix 3).For item 3, we used the same bowl/portion images used in item 2 and caregivers were instructed to select the bowl/portion icon that best correlated with the relative amount of fortified porridge that their child consumed during the entire meal (i.e. the sum of T1 þ T2 þ T3), as compared with their child's regular intake of food during a meal served to the child at that time of day.This was assessed at the end of the meal, and caregivers used the same five-point hedonic scale and for item 3 as was used to answer item 2.
Item 4 was a 'fill-in-the-blank' item where caregivers provided the total number of servings of the fortified porridge that their child consumed.For item 4, caregivers were asked to sum the total number of full portions of fortified porridge that their child consumed during the entire meal (i.e. the sum of T1 þ T2 þ T3 plus any additional portions of the fortified porridge consumed after T3).This was assessed at the end of the meal, and the responses are disaggregated into the following categories in the results section: (1) consumed 3 full portions; (2) consumed 2 full portions; (3) consumed 1•5 full portions; (4) consumed 1 full portion and (5) consumed 1/2 portion.Item 4 also included a 'comments' section to provide caregivers with an opportunity to add any information that they felt might have influenced how their child responded to the fortified porridge during Sensory Panel II (i.e.whether their child was breast-fed right before or during the panel, etc.).

Kapenta dried fish powder: nutrient composition results
For Table 1, ages were categorised into three groups according to dietary reference intakes (DRI), which includes infants aged 7-12 months, children aged 1-3 years and women aged 19-50 years (44,45) .Recommended DRI values for infants, children and women were obtained from the National Academies' Institute of Medicine.DRI values include recommended dietary allowances and adequate intakes.Recommended dietary allowances are the level of intake of essential nutrients that are determined by the Food and Nutrition Board of the National Academies' Institute of Medicine to be adequate to meet the known nutrient needs of practically all healthy persons (44) .Adequate intakes are used when there is not enough data to calculate an average requirement; these are the average nutrient levels consumed daily by a typical healthy population that is assumed to be adequate for the population's needs (ibid.) .
Recommendations regarding highest allowable mercury concentrations in fish were obtained from a joint advisory from the U.S. Food and Drug Administration and the U.S. Environmental Protection Agency (46) .The mercury concentration in the Kapenta dried fish powder was found to be 0•14 ug/100 g.This concentration is below the Food and Drug Administration/Environmental Protection Agency highest allowable average mercury concentration in fish per serving of 15 ug/100 g, when consuming three servings of fish/week (ibid.) .
The total amount of Kapenta dried fish powder used in each ComFAþFish dish varied according to recipe.For example, the recipe for the fortified chibwabwa fisashi included 64 g of dried fish powder, the recipes for both the fortified complementary maize porridge and the fortified bean-vegetable soup included 128 g of dried fish powder and the recipe for the savory Kapenta chutney included Dried fish powder: sensory panels in Zambia 256 g of whole dried fish.The recommended serving sizes for each ComFAþFish dish varied according to age category.For example, recommended serving sizes ranged from 8 to 10 g for infants aged 6-11-month-olds, 16-20 g for children aged 12-23-month-olds, 32-40 g for children aged 24 months and older and 32-40 g for adolescents and adults (44,45) .For Table 1, we calculated the percentage of DRI met for various nutrients, minerals and vitamins based on serving sizes of 10 g for infants aged 7-12 months, 20 g for children aged 1-3 years and 30 g for women aged 19-50 years.
As Table 1 indicates, the sample of Kapenta dried fish powder contained high percentages of the DRI for protein across all age categories and appreciable percentages of the DRI for choline across all age categories.In terms of protein, 10 g of the sample contained 65 % of the DRI for infants, 20 g contained 109 % of the DRI for children and 30 g contained 46 % of the DRI for women.In terms of choline, 10 g of the sample contained 24 % of the DRI for infants, 20 g contained 37 % of the DRI for children and 30 g contained 26 % of the DRI for women.
The sample contained high percentages of the DRI for Ca and Zn across all age categories, appreciable percentages of the DRI for Mg across all age categories and appreciable percentages of the DRI for Fe for children and women.In terms of Ca, 10 g of the sample contained 110 % of the DRI for infants, 20 g contained 82 % of the DRI for children and 30 g contained 86 % of the DRI for women.In terms of Zn, 10 g of the sample contained 43 % of the DRI for infants, 20 g contained 85 % of the DRI for children and 30 g contained 48 % of the DRI for women.In terms of Mg, 10 g of the sample contained 21 % of the DRI for infants, 20 g contained 40 % of the DRI for children and 30 g contained 15 % of the DRI for women.In terms of Fe, 20 g of the sample contained 30 % of the DRI for children and 30 g contained 18 % of the DRI for women.
The sample contained extremely high percentages of the DRI for vitamin B12 across all age categories and appreciable percentages of the DRI for vitamin A and vitamin D across all age categories.In terms of vitamin A, 10 g of the sample contained 25 % of the DRI for infants, 20 g contained 84 % of the DRI for children and 30 g contained 54 % of the DRI for women.In terms of vitamin B 12 , 10 g of the sample contained 220 % of the DRI for infants, 20 g contained 244 % of the DRI for children and 30 g contained 138 % of the DRI for women.In terms of vitamin D, 10 g of the sample contained 43 % of the DRI for infants, 20 g contained 57 % of the DRI for children and 30 g contained 85 % of the DRI for women.

Sensory panel I results: caregivers' evaluation of four ComFAþFish dishes
Comparison of attribute scores Among the four ComFAþFish dishes, the fortified chibwabwa fisashi was evaluated highest among caregivers, followed by the savory Kapenta chutney, the fortified complementary maize porridge and the fortified beanvegetable soup.Below, we present the results for each of the four dishes derived from averaging the results of the five sensory attributes for each dish (i.e.aroma, appearance, mouthfeel/texture, flavour/taste and sweetness).The disaggregated results and the average of these results for each ComFAþFish dish are presented in Tables 2-7.We next combined the averaged results for the two response categories of 'extremely liked' and 'liked' into one category of 'extremely liked/liked,' and this result is presented below for each ComFAþFish dish.Finally, we separately present the results for the two additional attributes of convenience and overall acceptability for each ComFAþ Fish dish.Fortified chibwabwa fisashi: averaged sensory attribute scores.As Table 3 indicates, when caregivers' (n 42) scores for the five sensory attributes of the fortified chibwabwa fisashi were averaged, 94•7 % of caregivers extremely liked/liked the sensory attributes of this dish.A majority of caregivers also extremely liked/liked the convenience (92•8 %) and overall acceptability (100 %) of the fortified chibwabwa fisashi.
Savory Kapenta chutney: averaged sensory attribute scores.As Table 4 indicates, when caregivers' (n 42) scores for the five sensory attributes of the savory Kapenta chutney were averaged, 81•8 % of caregivers extremely liked/liked the sensory attributes of this dish.A majority of caregivers also extremely liked/liked the convenience (92•8 %) and overall acceptability (100 %) of the savory Kapenta chutney.
Fortified complementary maize porridge: averaged sensory attribute scores.As Table 5 indicates, when caregivers' (n 41) scores for the five sensory attributes of the fortified complementary maize porridge were averaged, 83•5 % of caregivers extremely liked/liked the sensory attributes of this dish.A majority of caregivers also extremely liked/liked the convenience (90•5 %) and overall acceptability (88•1 %) of the fortified complementary maize porridge.
Fortified bean-vegetable soup: averaged sensory attribute scores.As Table 6 indicates, when caregivers' (n 24) scores for the five sensory attributes of the fortified bean-vegetable soup were averaged, 66•8 % of caregivers extremely liked/liked the sensory attributes of this dish.A majority of caregivers also extremely liked/liked the convenience (87•5 %) and overall acceptability (87•5 %) of the fortified bean-vegetable soup.

Relative fortified porridge intake among infants aged 6-11 months
In terms of relative porridge intake, a majority of caregivers (68•8 %) reported that their 6-11-month-olds consumed 3/4 or more of their regular amount of food at T1 þ T2 þ T3 as compared with their child's regular intake of food during a meal served to the child at that time of day.In contrast, 31•3 % of caregivers reported that their 6-11-month-olds consumed 1/2 of their regular amount of food at T1 þ T2 þ T3 as compared with their child's regular intake of food during a meal served to the child at that time of day.Finally, a majority of caregivers (87•5 %) reported that their 6-11-month-olds consumed one or more full portions of the fortified porridge (i.e.T1 þ T2 þ T3 plus any additional portions of the fortified porridge consumed after T3).

Relative fortified porridge intake among children aged 12-23 months
In terms of relative porridge intake, a majority of caregivers (96•2 %) reported that their 12-23-month-olds consumed 3/4 or more of their regular amount of food at T1 þ T2 þ T3.In contrast, 3•8 % of caregivers reported that their 12-23-month-olds consumed less than 1/4 of their regular amount of food at T1 þ T2 þ T3.Finally, all caregivers (100 %) reported that their 12-23-month-olds consumed one or more full portions of the fortified When the disaggregated scores are examined, the appearance of the fortified bean-vegetable soup was assessed as less desirable than the appearance of the other three ComFAþFish dishes.For example, nearly 43 % of caregivers extremely liked/liked the soup's appearance, while nearly 21 % of caregivers were neutral regarding the soup's appearance.In contrast, an overwhelming majority of caregivers extremely liked/liked the appearance of the fortified chibwabwa fisashi (∼98 %), savory Kapenta chutney (100 %) and fortified complementary maize porridge (∼93 %), and no caregivers were neutral regarding the appearance of these ComFAþFish dishes.These results suggest that the recipe for the fortified bean-vegetable soup should be adjusted to improve appearance.

Sensory panel I: non-sensory attribute results among caregivers
The results suggest high acceptability among caregivers of the attributes of convenience and of overall acceptability for all four ComFAþFish dishes.A majority of caregivers extremely liked/liked the convenience of all four of the ComFAþFish dishes, and a majority of caregivers scored the overall acceptability as high for all four of the ComFAþFish dishes.Given that the extreme 'time poverty' of resource-limited women in LMIC is well documented (47)(48)(49) , the attribute of convenience is key to scaling a protein/micronutrient blend such as ComFAþFish and ensuring that resource-limited women: (1) adopt the product long-term and (2) use it on a regular basis (i.e. in at least one meal/d for children aged 6-23 months).A nutrient-dense supplement that does not account for the extreme time poverty of female caregivers (i.e.does not ensure that the product is easy-to-use on a daily basis) will likely encounter barriers to widespread adoption regardless of its efficacy.Therefore, ComFAþFish is designed with convenience as a central attribute.By utilising dried fish powder as its primary ingredient, ComFAþFish is a nutrient-dense, locally accessible product that can easily be added to traditional dishes that family members are already used to consuming, including infants and young children who are in the complementary feeding stage.It is expected that this approach also helped ensure that the ComFAþFish dishes received high scores for overall acceptability, as this is an attribute that Puri et al. (2022) argue is more important 'from the consumer point of view' than are individual sensory attributes (41,5) .

Sensory panel II: evaluation by caregivers of global liking among infants and young children
The Sensory Panel II results indicated that the ComFAþFish complementary maize porridge was evaluated among the majority of caregivers as highly acceptable to both 6-11month-olds and 12-23-month-olds.These results are promising given that traditional (non-fortified) maize porridge is arguably the most commonly consumed complementary food among infants and young children in Zambia and many other LMIC across sub-Saharan Africa (37,38) .That a dried fish powder-based protein/ micronutrient blend such as ComFAþFish might not be appetising to infants and young children when added to complementary maize porridge was of critical concern in terms of scalability.These results suggest that fortifying complementary maize porridge with ComFAþFish did not have a negative impact on infants' and young children's food intake.This is a promising result, given that infants' and young children's gastric capacity can only accommodate small meals and, therefore, every meal should be nutrient dense.
The present study's results should be interpreted within the limitations of a pilot study that included a relatively small sample recruited using a convenience sampling strategy.Since ComFAþFish is an innovative protein/ micronutrient blend and the study is the first to evaluate its sensory and other attributes, there is currently no reference to which to compare our results.Social desirability bias might have affected caregivers' responses.For example, it is important to note that only twenty-four of forty-two caregivers completed their evaluation of the fortified bean-vegetable soup, and their responses for the soup's sensory attributes were not as positive as were caregivers' responses for the sensory attributes of the other three ComFAþFish dishes.These results suggest that the recipe for the fortified bean-vegetable soup will need to be modified in order to increase its sensory attribute scores.Further, the level of non-responses for the sensory attribute of sweetness may have skewed the averaged results.For example, while thirty-five of forty-two caregivers scored the attribute of sweetness for the fortified complementary maize porridge, twenty-eight of forty-two caregivers scored this attribute for the fortified chibwabwa fisashi, twenty-six of forty-two caregivers scored this attribute for the savory Kapenta chutney and fifteen of twenty-four caregivers scored this attribute for the fortified bean-vegetable soup.These results suggest that this attribute's definition should be clarified and revised for future sensory panels.

Conclusion
The present study contributes to the growing body of evidence that the consumption of aquatic animal-source foods including locally sourced dried fish powder are promising as ways to reduce stunting and anemia (15,(17)(18)(19)50) . The esults suggest that ComFAþFish is strategically well placed to fill nutritional gaps to address micronutrient deficiencies among vulnerable 6-23-month-olds, pregnant/lactating women and other household members (e.g.adolescent girls) in Zambia and other LMIC.That all four ComFAþFish dishes received high scores for the attribute of convenience suggests that this protein/micronutrient blend has the potential for wide scale adoption for daily fortification of complementary maize porridge and other traditional dishes consumed in Zambia and across sub-Saharan Africa.
A second advantage of ComFAþFishwhose primary ingredient is whole pelagic small fish that are dried and ground into a fine powderis that cold-storage methods of preserving fresh fish are rarely available to small-scale men and women fishers and processors in LMIC and, as a consequence, fresh fish can be lost through spoilage before they can be sold or processed.Pelagic small fish are easily dried whole, which not only concentrates their nutritional density but prevents spoilage.A third advantage of ComFAþFish is that it utilises widely available local ingredients (e.g.dried fish powder, dried pumpkin leaves, etc.) and, therefore, can be produced at the householdlevel using common kitchen tools and is also commercially produced with a mechanised grinder for sale in large batches or as affordable sachets in local markets (15,(17)(18)(19)39) . Becase the key ingredient of ComFAþFish is locally sourced dried fish powder, it has the potential to allow wider access to affordable and palatable sources of protein and micronutrients for urban and rural populations living in extreme poverty in Zambia and across sub-Saharan Africa for whom the lack of dietary diversity and reliance on highphytate, low-protein maize-based diets increases their vulnerability to hidden hunger.
What makes the results promising is thatrather than trying to introduce a new nutrient-dense product to market that would likely not be accessible or affordable for the urban and rural poor -ComFAþFish is a nutrient-dense product whose primary ingredient is an aquatic animalsourced food (whole pelagic small fish ground into a fine powder) that is widely accessible, affordable and is both commonly produced at-home and purchased as a pre-ground powder in urban and rural local markets.Because ComFAþFish uses locally sourced ingredients that already have high acceptability among our target population of vulnerable urban and rural households, scaling ComFAþFish across different regions and countries will leverage the familiarity of local foods.
Having determined the acceptability of four ComFAþ Fish dishes, our next steps are to: (1) complete a 6-month shelf-life study of Kapenta dried fish powder (underway); (2) as needed, adjust the amount of dried fish powder per recipe to ensure that the recommended serving size per age group (e.g. 10 g for infants aged 7-12 months) meets the recommended DRI values for infants, children and women without compromising convenience, overall acceptability and palatability and (3) collaborate with tiered in-country scaling partners to scale ComFAþFish at the national (e.g.school feeding programs), regional (mid-level entrepreneurs) and village levels (micro-enterprises and households) across Zambia and sub-Saharan Africa.

Fig. 2
Fig. 2 Map of Zambia's Lake Kariba showing the village recruitment sites of Gwembe, Siavonga and Sinazongwe Districts

Table 1
Dietary reference intakes (DRI) and percentages of DRI met for infants, children and women per serving of Kapenta dried fish powder Note: *This table presents recommended dietary allowances (RDA) in bold type and adequate intakes (AI) in ordinary type followed by an asterisk.

Table 7
Sensory panel II: evaluation by caregivers of global liking, actual intake and relative intake of ComFAþFish complementary maize porridge among 6-11-month-olds (n 16)

Table 8
Sensory panel II: evaluation by caregivers of global liking, actual intake and relative intake of ComFAþFish complementary maize porridge among 12-23-month-olds (n 26)