Black soldier fly larvae and cricket fortified crackers are a safe, nutritious, and acceptable snack for school children in Madagascar

Abstract

Edible insects are a nutrient-dense and environmentally friendly alternative to conventional animal-sourced protein. Wild-harvested insects have contributed to diets throughout history, but questions remain regarding the desirability of insect-fortified foods produced by expanding insect farming. We evaluated the safety, nutrition, and acceptability of three locally sourced and produced insect-fortified snacks among primary schoolchildren in Madagascar. Laboratory analyses confirmed the crackers containing crickets and/or black soldier fly larvae were safe for children ages 9 to 13 across all microbiological parameters and heavy metal contaminants, and nutritious, providing significant protein (12.2–14.2 g/100 g), iron (6.6-8.6 mg/100 g), and zinc (3.0-3.4 mg/100 g). Child acceptance (N = 187) was high, with approximately 80% consuming at least 80% of their portions, and 89% selecting insect-fortified crackers over the control when given a choice. Savory recipes were preferred over sweet. In focus groups, caregivers (N = 60) reported positive perceptions of the snacks, with anecdotal improvements in child satiety, energy, and sleep.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-52709-y.

Introduction

The converging pressures of industrial agriculture’s environmental toll, the mounting threat of climate change to food security, widespread nutrition challenges, and accelerating population growth require urgent transformation of our food production systems¹. Despite gains in crop productivity, food waste reductions, and global efforts to encourage plant-focused diets, it is unclear whether these efforts are sufficient to meet the health and environmental challenges faced by developing economies like Madagascar². Low-income communities and children are especially vulnerable to dietary insufficiency, including in contexts where plant-based diets are the norm and food waste is low. Protein and iron deficiency are common in these contexts², thus the development of novel animal-sourced foods that are safe, nutritious, and acceptable would be beneficial.

Increased production and consumption of edible insects and insect-fortified foods is one potential solution^3–5. Insects have been an important part of the diet throughout human history⁶, and more than 2,000 species have been documented as edible⁷. Entomophagy, the practice of eating insects, is most common in the tropics and across the Global South today⁸. Madagascar is no exception. As early as the 1600s, records describe insect eating on the island⁹ and consumption of wild-foraged, seasonally available insects continues¹⁰. Because they have high reproductive rates¹¹, high feed conversion efficiency (arising in part from ectothermic metabolism, meaning they do not use energy to regulate their body temperature)¹², can consume organic waste streams, and require little land, food, and water to grow¹³, farmed insects are more environmentally friendly than conventional livestock and emit fewer greenhouse gases^14,15. Edible insects are widely recognized as nutrient-dense^16,17, containing digestible animal-sourced protein^18,19, essential amino acids^20–24, healthy fatty acids²⁵, micronutrients including iron and zinc²⁶ and, uniquely, prebiotic dietary fiber^27,28. As insects are an animal-sourced protein, their potential to help address protein energy malnutrition is salient; they supply valuable protein and micronutrients typically found in meat that could fill a gap when other animal products are not available^3,4..

Nevertheless, questions regarding the safety (ensuring edible insect-based products are free from significant health risks, including but not limited to microbial, chemical, and physical), acceptability (consumers’ willingness to eat insect-based products), and practicality (feasibility of creating a robust supply chain that could provide continuous insect-based food in proportion to the demand once there is widespread acceptability) of expanding insect use as food remain. Even where local consumption of insects occurs, the acceptability of farmed, processed insect products is not guaranteed; moreover, local insect eating does not equate to a blanket approval, acceptance of, or value placed on all edible insect species^29,30. It is not known whether individuals will eat insect-fortified products on a consistent basis, nor whether they will consume them in amounts that would provide biologically meaningful quantities of nutrients. Further, there remain questions regarding the food safety of insects and insect-fortified products^24,31.

Our paper addresses these knowledge gaps. Working in Madagascar, we co-designed three insect-fortified crackers and one unfortified (control) cracker with a local food technology company, tested these crackers for safety and nutritional content, and assessed their acceptability among school-aged children (9-13y) and their caregivers. We served the four types of crackers as a supplement to a midday school meal in the southeast of the country, a locale characterized by high levels of food insecurity, malnutrition, and poor diets^31,32 where consumption of some insects is culturally acceptable.

Results

Background

We assessed the acceptability of three different insect-fortified crackers: a savory cracker that contained a combination of a local cricket species (Gryllis madagascarensis) and black soldier fly larvae (Hermetia illucens (BSFL)); a sweet cracker that contained both cricket and BSFL; and a savory cracker that contained only BSFL. An unfortified cracker of rice and corn was also included as a control.

Safety

Both crackers were screened for pathogens understood to be associated with insects used as a source of food, namely: Bacterial pathogens most commonly found in insects (Staphylococcus aureus, pathogenic Clostridium spp., and pathogenic species of the Bacillus cereus group)³²; yeasts and molds³², aerobic spores³³, Salmonella and Campylobacter³⁴; and Cronobacter spp³⁵. Because insects and crustaceans are both a subset of the phylum Arthropoda, crackers were tested for Vibrio spp, a pathogen found in crustaceans³⁶. Testing was also done for bacteria that could be introduced during the manufacturing process: Salmonella; Escherichia coli; and Listeria monocytogenes. Across 13 biological parameters tested, there were no instances - either in the prototype or confirmatory tests – where safety limits were exceeded (Table 1).

Table 1.

Food safety data, by cracker type ^#.

Biological parameter	Unit	Safety limit	BSFL and cricket, savory		BSLF and cricket, sweet		BSFL, savory		Control, savory
			Primary test	Confirmation	Primary test	Confirmation	Primary test	Confirmation	Primary test	Confirmation
Staphylococcus aureusA, B	cfu/g	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10
Clostridium perfringensC	cfu/g	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10
Bacillus cereusA, D	cfu/g	≤ 100	20	< 10	< 30	< 10	< 10	< 10	< 10	< 10
MoldsA, B	cfu/g	< 100	< 10		< 10		< 10		< 10
YeastA, B	cfu/g	< 100	< 10		< 10		< 10		< 10
Aerobic Plate Count A, B	cfu/g	< 104	< 10		10 (est)		10 (est)		80 (est)
Microrganism growth	cfu/g	< 1000		< 1000		< 1000		< 1000		< 1000
Escherichia coli A, B,C	cfu/g	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10	< 10
Vibrio spp A		Not detected.	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected
Campyloba-cter Species C		Not detected in 25 g of product	Not detected		Not detected		Not detected		Not detected
Listeria monocytogenes B, D		Not detected.	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected
SalmonellaB, D		Not detected in 25 g of product	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected
Cronobacter sppD		Not detected in 10 g of product	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected	Not detected

Heavy metal parameter	Unit	Food Safety Limit	Primary Test	Confirmation	Primary Test	Confirmation	Primary Test	Confirmation	Primary Test	Confirmation
Arsenic (As) E	mg/kg	0.2	0.0194		0.0149		0.018		0.0336
Cadmium (Cd)A	mg/kg	< 0.06	0.0116		0.0226		0.0157		0.005
Lead (Pb)A	mg/kg	< 0.05	0.0233		0.024		0.0484		0.00597
Mercury (Hg)E	mg/kg	< 0.1	0.005		0.005		0.005		0.005
Manganese (Mn) A	mg/kg	≤ 100	39		38		47		11
Chromium (Cr)	mg/kg	No set limit	0.554		0.334		0.206		1.54
Nickel (Ni)	mg/kg	No set limit	0.215		0.412		0.16		0.244
Tin (Sn)	mg/kg	No set limit	0.0122		0.00749		0.0086		0.0184
Copper (Cu)	mg/kg	No set limit	2.71		3.28		2.39		1.12
Aluminum (Al)	mg/kg	No set limit	15.2		14.7		16.4		10.5
Selenium (Se)	mg/kg	150 µg/day	0.117		0.108		0.149		0.095
Zinc (Zn)	mg/kg	No set limit	32		32		29		13

# Food safety testing results from two rounds of commercial testing. The first Prototype Test was conducted on all cracker recipes in August, 2024 at Eurofins Microbiology Laboratories (Cascade), USA. The second, Confirmation tests, were completed at Institut Pasteur de Madagascar (Laboratoire d’Hygiéne des Aliments et de l’Environnement) in September 2024. These tests were run on the exact batches of crackers used for the study. Superscript letters next to test parameter name indicate source: ^ACELEX (2002)⁵⁹, ^BBNM (2023)⁶¹, ^CHEALTH CANADA (2013)⁶⁰, ^DIPIFF 2024⁵⁸, ^ECODEX (2022)⁵⁷.

Prototype crackers were also tested for 12 heavy metals which can accumulate from the insect diet:³⁷ arsenic; cadmium; lead; mercury; chromium; nickel; tin; copper; aluminum; manganese; selenium; and zinc. There were no instances where heavy metal concentrations exceeded existing safety limits.

Eating insects can elicit allergic reactions caused by cross-reactivity with other taxonomically related food allergens like those found in crustaceans³⁸. Individuals with a shellfish allergy, for example, are at risk of anaphylaxis after the ingestion of cricket products³⁹. During study enrollment, caregivers were asked if their children had an allergy to insects or shellfish. Four children out of 200 were excluded from the study because of shellfish or insect allergies. During the acceptability trial, there were no instances of severe allergic reactions following consumption of the insect crackers and only one instance of a moderate reaction (a minor rash) was reported by caregivers.

Nutrition content

A 50 g serving of crackers contained 6–10 g of cricket and/or BSFL powder and 0.65–1.0 g of insect fiber, resulting in a range of 6.1 to 7.1 g protein (12.2–14.2 g/100 g), 0.9 to 1.5 g of fat (1.8–3.0 g/100 g), and 35.4 to 39.2 of carbohydrates (70.8–78.4 g/100 g), yielding 184 to 189 kcal. A 50 g serving also contained 1.6 to 1.8 g fiber (3.2–3.6 g/100 g), 3.1 to 4.3 mg iron (6.2-8.6 mg/100 g), and 1.5 to 1.7 mg zinc (3.0-3.4 mg/100 g) (Table 2). The savory flavored cracker containing BSFL and cricket, contained the highest amount of protein, fiber, iron, and zinc out of all four cracker types − 20.9% of the recommended intake for protein, 6.9% of the recommendation of fiber for girls, 5.8% of the recommendation⁴⁰ of fiber for boys, 53.8% for iron, and 21.3% for zinc (Table 2).

Table 2.

Content of a 50 g serving of insect-fortified crackers^^.

Energy (kcal)	BSFL and Cricket Savory	BSFL and Cricket Sweet	BSFL Only Savory	Control
	184	193	189	180
Carbohydrate (g) (%RDA; %E)	35.4 (27.2%; 76.96%)	38.1 (29.3%; 78.96%)	39.2 (30.2%; 82.96%)	41.1 (31.6%; 91.33%)
Protein (g) (%RDA; %E)	7.1 (20.9%; 15.43%)	6.5 (19.1%; 13.47%)	6.1 (17.9%; 12.91%)	4.0 (11.8%; 8.89%)
Fat (g) (%RDA; %E)	1.5 (ND; 7.34%)	1.6 (ND; 7.46%)	0.9 (ND; 4.29%)	0.0 (ND; 0.0%)
Dietary Fiber (g) (%AI)	1.8 (5.8, 6.9%)	1.7 (5.5%, 6.5%)	1.6 (5.2%, 6.2%)	1.3 (4.2%, 5.0%)
Iron (mg) (%RDA)	4.3 (53.8%)	3.1 (38.8%)	3.3 (41.3%)	1.4 (17.5%)
Zinc (mg) (%RDA)	1.7 (21.3%)	1.6 (20.0%)	1.5 (18.8%)	0.7 (8.8%)

^^ For carbohydrate (130 g/day), protein (34 g/day), iron (8 mg/day) and zinc (8 mg/day), the nutritional data is presented as grams (%RDA) for children 9 to 13 years. No RDA or adequate intake (AI) value exists for fat in this age group so it remains as not determined (ND). Fiber AI values are gender-specific: 31 g/day for boys, and 26 g/day for girls^39,40. Protein Energy Percentage, listed as %E, is the proportion of total energy contributed from the macronutrient.

Participant demographics

Acceptability was assessed in three schools selected in partnership with Catholic Relief Services (CRS) who were providing school meals as part of the USDA-funded MaMi project in Farafangana and Vondrozo districts. A total of 187 children, ages 9 to 13 years old (mean 10.4y (sd 1.4y)), were enrolled. A majority were male (58.3%) and almost half (46.0%) of the sample attended first or second grade. Each school had access to surface water (pond, river) within 0.25 to 1 km and two latrines (Appendix Table 1).

Acceptability (1): Children’s consumption and perceptions of insect crackers

Figure 1 shows the percentage of crackers consumed, by cracker type, for all days’ consumption and results excluding the first day of the study in each school; this allows us to assess whether consumption was affected by the novelty associated with the start of the acceptability trial and the novelty of the crackers. Across all insect cracker types, 67% of children consumed 80% or more of the insect crackers, this rises to 78.9% when the first day that crackers were provided is excluded. Less than 5% of children consumed less than 10% of the insect crackers provided to them. The savory crackers, BSFL and crickets, and BSFL alone, had higher levels of consumption (81.2 and 81% respectively, excluding the first day) than the sweet BSFL and cricket cracker (73.9%). Boys were more likely to consume 80% or more of the insect crackers (82.9%, excluding first day effect) than girls (73.3%) (Appendix Fig. 1). According to the chi-square test, consumption of 80% or more did not differ by age (Appendix Fig. 2, Appendix Table 7) with chi-square result X² (dof 12) = 16.82, p-value = 0.156, but by sex (Appendix Table 7) with X² (dof 3) = 22.86, p-value = 0.00004. Results from an ordered probit show that girls were 11.8% points less likely to consume > 95% of any type of cracker (p < 0.001). Relative to the control cracker, children were 10.7% points more likely to consume > 95% of the savory BSFL and cricket cracker with p < 0.01 and 11.8% points more likely to consume > 95% of the savory BSFL cracker with p < 0.001 (Appendix Table 2).

Fig. 1.

Cracker consumption, by cracker type, with and without first day of consumption. A stacked schematic describes the percentage in four consumption categories by any insect crackers (BSFL and Cricket (savory), BSFL and Cricket (sweet), or BSFL (savory)), individual insect crackers, and the control cracker. The five pairs of stacked bars are divided based on cracker type, and with and without day-one effects.

On days 1–4, children rated the crackers on a five-point Likert scale (ranging from “liked a lot” to “disliked a lot”) in terms of: color; shape; texture; aroma; and taste. Across all attributes, approximately 80% of children “liked a lot” or “liked a little” both the savory BSFL and cricket cracker, and the savory BSFL cracker. Across these five attributes, between 64.0 and 75.0% of children “liked a lot” or “liked a little” attributes of the sweet BSFL and cricket cracker (Fig. 2). We find children’s evaluation of aroma and taste between different cracker types shows a strong difference, while color, shape, and texture show a difference with a small size effect. However, the R-squared values indicate that cracker type explains only a small fraction (less than 3.1%) of the variance in children’s sensory evaluations (Appendix Table 8).

Fig. 2.

Child’s perception of crackers, by attributes, excluding first day consumption. A stacked schematic describes the percentage in five preference categories by the three insect crackers (BSFL and Cricket (savory), BSFL and Cricket (sweet), or BSFL (savory)), and the control cracker. The four sets of stacked bars are divided based on cracker type, the five organoleptic categories, and without day-one effects.

On days 1–4, children also provided an overall assessment of the crackers they had consumed on the same five-point Likert scale that they used to describe the crackers by attribute. Aggregating across all three insect crackers (Fig. 3), 76% of children “liked a lot” or “liked a little” these crackers. Only 2.4% “Disliked a lot” all insect crackers. The two savory crackers, BSFL and crickets, and BSFL were “liked a lot” or “liked a little” by similar percentages of children, 78.7 and 79.5% respectively. By contrast, 65.8% of children “liked a lot” or “liked a little” the sweet BSFL and cricket cracker. These preferences do not differ by child sex (Appendix Fig. 3) or child age (Appendix Fig. 4). Results from an ordered probit show no association between child age or sex but do show that relative to the control cracker, children were 9.8% points (p < 0.001) less likely to say that they “liked a lot” the sweet BSFL and cricket cracker (Appendix Table 3). Using the average of the five organoleptic categories produces similar results (Appendix Table 4).

Fig. 3.

Child’s perception of crackers, by cracker type. A stacked schematic describes the percentage in five preference categories by any insect crackers (BSFL and Cricket (savory), BSFL and Cricket (sweet), or BSFL (savory)), individual insect crackers, and the control cracker. The five pairs of stacked bars are divided based on cracker type, and with and without day-one effects.

On the fifth day, children chose which cracker to consume, and 89.1% of children chose an insect cracker (Fig. 4). Of children who chose an insect cracker, most (61.9%) chose the savory BSFL and cricket cracker, followed by the savory BSFL cracker (24.5%), and the sweet BSFL and cricket cracker (13.6%). Boys were more likely to choose an insect cracker (92.8%) than girls (83.8%) (Appendix Fig. 5). Children of all ages chose an insect cracker over the control cracker, with the percentage doing so ranging from 95.1% (children aged 10y) to 80% (children aged 13y) (Appendix Fig. 6). Using information on children’s preferred crackers, we re-estimated the ordered probit shown in Appendix Table 2. Conditional on consuming the cracker that children indicated they liked the most, there was no association between child, cracker, or school on the amount of cracker consumed. Consumption rose with increased exposure (as measured by study day, p < 0.001) to the crackers (Appendix Table 5).

Fig. 4.

Child’s cracker choice and consumption of preferred cracker. The three sets of bar graphs describe day-five cracker selection by children. [A] The bar graphs describe the percentage of people who chose any insect cracker versus a control cracker on day five of the study. [B] Of those who choose any insect crackers on day five, the next set is divided into three insect crackers. [C] The final set describes which one of the four categories of consumption the child’s preferred cracker fell in.

Acceptability (2): Caregiver perceptions of insect crackers

A total of 60 parents and caregivers (30 women, 30 men) participated in Focus Group Discussions (FGDs) across the three schools following the taste tests. Thematic analysis of FGD transcripts revealed generally positive perceptions of the insect-fortified crackers across all sites, despite their novelty. Participants responded positively to their children receiving additional food of any kind, and demonstrated an openness to eating processed insect ingredients and a general positive impression of the specific product attributes.

FDGs revealed benefits to children’s wellbeing after consuming the insect fortified crackers, specifically improvements in satiety, energy, and sleep. Multiple participants reported enhanced satiety compared to regular school lunches, with mothers particularly noting this effect. One fourth-grade mother observed, “My daughter was satisfied after her snack at school during the week of the tasting,” while another reported her son “felt full… didn’t finish his meal at the canteen and slept well.” Parents also described improved energy levels that increased throughout the testing period. One mother explained, “We live far from the school and my daughters are often exhausted. After consuming these products, they become lucid,” while a father noted his children “became livelier and more energetic after four days of cracker distribution.” Another father enthusiastically endorsed the benefits: “These crackers are a brilliant idea! My child loves them, and they seem to give him a real energy boost.”

Parents reported overall positive reception by their children to the sensory attributes of the insect-fortified crackers, with widespread interest in continued consumption after the study concluded. However, FGDs did reveal nuanced opinions on some of the products’ characteristics. Though texture and appearance were generally well-received, participants noted concerns about smell and taste. One mother reported her children “didn’t like the smell” although they “appreciated the flavors,” while a father observed “some students were unable to finish their portion due to the taste and aromas, which were unfamiliar to them.” The main practical concern expressed was product dryness, with one father reporting that “some of the crackers caused my child to feel thirsty,” an observation confirmed by staff who provided school meals. Parents reported a clear preference among children for the savory, co-fortified (BSFL and cricket) recipe, and the savory, BSFL-only fortified recipe. Importantly, the incorporation of crickets and BSFL in the crackers—which are not traditionally consumed in this area—generated no significant concerns.

FGDs revealed strong comfort with and enthusiasm for the crackers by parents and caregivers. One father stated that the idea of providing children with insect-fortified crackers is “very good” to “help our children grow up healthy.” Another father was happy to know that his son was “eating something that is both good for him and for the environment.” Parents highlighted the nutritional, environmental, and practical benefits of the product. The idea of eating insects and insect-fortified foods was well received. No objections were noted regarding the inclusion of crickets nor BSFL.

Discussion

Results from our testing of insect-fortified crackers confirmed safety across all microbiological and heavy metal parameters tested. This is a very important consideration for novel food products being introduced to vulnerable populations, including children who may face undernutrition. None of the cracker recipes exceeded safe levels of the 13 biological parameters in original or confirmatory tests, nor heavy metal content. Food safety parameters were run on the exact batch of crackers used for the acceptability study and based on recommendations set by numerous organizations working to establish food safety standards for insect products and insect-fortified foods (see Table 1). As a processed, extruded product including insect powders produced explicitly for human consumption, these results are reassuring. Third party laboratory tests also confirmed the crackers to be nutritious, providing a significant source of protein (18–21% RDA), iron (39–54% RDA), and zinc (19–21% RDA) for children between 9 and 13 years of age^40,41. Insect-fortified crackers were acceptable with nearly 79% of school children consuming at least 80% of their portions, and the vast majority (89%) selecting insect crackers over control options when given a choice. FGDs with caregivers revealed positive perceptions of the crackers and observed notable benefits including improved satiety, energy levels, and sleep in their children.

Despite the traditional focus on whole, wild-harvested insects in the localities where our study was situated, participants were receptive to processed, insect-fortified crackers. This is consistent with previous research demonstrating that the acceptability of processed insect foods tends to be higher in places where insects have traditionally been part of the diet^30,42 and is high among children^43,44. Also of note is our finding on the slight preference for savory over sweet crackers. While children are commonly found to prefer sweet over savory foods, some research has shown that they can accept either, and this has been associated with height and percent body fat⁴⁵. In this instance, it is more likely that the flavor of the insect powders paired better with savory spices than with cocoa and sugar. Additionally, the combination of sugar and heat in the extruder produced a slightly darker, smaller, harder product in contrast to the other three recipes which were more similar in terms of color, shape, and texture.

Our study has several strengths. Our essential food safety parameters are novel and can serve as a guide for future insect food developers in diverse contexts. The tested crackers were sourced from Malagasy-only ingredients (rice, corn, insects). The study design enabled all four products to be tested within the same school, thus avoiding problems associated with a specific taste being preferred in a specific locality. Children were asked to consume a quantity of crackers (50 g per day for a week) that contained meaningful quantities of protein, zinc, and iron. Third, results on acceptability were triangulated through comparative analyses using four methods: (1) by having children rate the sensory characteristics of the crackers, (2) by quantifying daily consumption, (3) by asking children to choose their favorite recipe and (4) by collecting insights and observations from caregivers.

We note the following limitations. Our study design precluded all cracker recipes from an equal number of testing days on day 1 and day 4. On day 1, this design challenge impacted total consumption; thus, we presented percent consumption of each recipe to remove the hesitancy bias. On day 4, this issue may have impacted children’s final selection for preferred recipe, as it was dependent on their memory of eating the snacks in the days prior. We sought to reduce this bias and make it easier for children to remember the crackers by color coding the recipe bags and asking children to remember their favorite snack flavor (by color) each day. We used basic, bright colors (red, yellow, blue, green) and letters for cracker types in our test. Research indicates that color is generally a useful cue for memory, but its effectiveness depends on consistency and task relevance⁴⁶. Considering the simplicity of the study design with just three recipes and a control to taste, a simple color model was appropriate. Children aged 10 are shown to hold three to four items in their visual working memory⁴⁷. Participants saw the colors multiple times as they repeated tasting of each cracker recipe. We do recognize, however, that memory cues work in combination with one another, not in isolation, so pairing color with something congruent and something specific to the product⁴⁸ (perhaps brown for the cocoa sweet crackers) may have been helpful. The one consistent concern raised by students, teachers, and caregivers about these crackers was dryness and increased thirst—an important consideration in contexts that are water insecure.

Additionally, our acceptability trial lasted for one school week, with children receiving eight exposures to the insect fortified crackers over four days and providing them the opportunity to choose consumption of these crackers on the fifth day. The extant literature does not provide guidance on the optimal length of acceptability trials for school aged children. Studies that focus on acceptability of novel foods for infants typically range from two to three days^49–51 to eight to ten exposures over multiple days⁵². The duration of our study is at the upper end of this range; further, by running the study in each school for five days, we are able to account for the potential for reactance by children to our presence⁵³. That said, our study is not designed to assess whether acceptability would continue to be high over a long period of regular consumption.

Our study measured acceptability of the insect-fortified crackers in one region of Madagascar, in one age group, across three schools, where entomophagy was prevalent, driven primarily by historical precedent and the idea of ancestral permission. Results may not translate to different populations.

Another important contribution of this study is new insight into the nutrition and acceptability of farmed insects. Our study adds to a growing knowledge base surrounding edible insects, especially in areas where insects are currently consumed. Today, the vast majority of insects eaten globally are harvested from the wild and are only seasonally available⁷. While this represents a free, locally available food resource, wild harvested insects are rarely available year-round and are not available at a scale large enough to adequately address local dietary deficiencies. Overharvesting of wild insect populations can be ecologically damaging. Moreover, the food safety of wild-harvested insects cannot be guaranteed. Hence, there is a need to assess whether novel food products fortified with commercially farmed edible insects are safe for consumption, nutritious, and acceptable for target populations. Our study fills these knowledge gaps. From a policy perspective, we have shared our results with Catholic Relief Services to inform the MaMi program schools and recipient families of our results. Moreover, our results have also been shared with the cracker producers and the Ministry of Education, such that consideration of integrating insect-fortified crackers into broader school lunch programs will be possible. The current plan is to work with schools and donors to set up a cracker procurement schedule. Local production reduces prices and stimulates the economy, reducing reliance on external imports. Currently, one 50 g serving is estimated to cost $0.14. While this cost is high, it is not representative of what it will be on a larger scale, as packaging accounts for 58% of the price. Larger quantities would allow the producers to reduce packaging costs, raw materials, and BSF powder by 15 to 25%.

We see a future research agenda on this novel food focusing on the following: (1) comparing the cost-effectiveness and environmental footprint of insect-fortified crackers with conventional animal source foods; (2) combining information on consumption of these crackers with information on children’s food consumption at home and at school to assess whether crackers augment or crowd out consumption of food from these other sources; (3) assessing the impact of the consumption of these foods on micronutrient status; and (4) determining whether this product improves gut health, as has been suggested elsewhere^28,54. Insects contain chitin, an insoluble fiber that humans may partially digest through chitinases, potentially converting it to more soluble derivatives that may be fermented by beneficial bacteria in the gut and reduce inflammation^28,55,56. Additionally, work to determine the effectiveness of insect-fortified products in combating general protein-energy malnutrition is warranted. Should results from this future work be positive, insect-fortified products such as the crackers tested in this study may well be an important component of efforts to meet global dietary challenges in a safe, nutritious, acceptable, health-promoting and environmentally sensitive fashion.

Methods

Study design

This study evaluated the acceptability of three different insect-fortified crackers and one control cracker through a series of assessments: (1) a multi-day taste test including sensory evaluations using a Likert Scale with children aged 9 to 13 years attending school in Farafangana district of Madagascar’s Atsimo-Atsinanana region; (2) measurement of the quantity of crackers eaten by students each day, and (3) focus group discussions (FGDs) with parents, caregivers, and school canteen cooks (that are also parents) to explore cultural implications of edible insects, perceptions of insect-fortified crackers, signs of allergies or gastrointestinal distress, and the feasibility of implementing a daily snack time. The study was conducted over four or five school days at each of the three schools in November 2024. Each school day, about 60 children received 25 g of two distinct insect-fortified cracker recipes in sealed bags. The crackers varied both in flavor and insect composition, and they were blinded for the children but not the investigators. Children were given as much time as they needed to eat as much of the crackers as they wanted. Once finished, the research team met with students individually to collect feedback on the crackers. Following snack time, the weight of leftovers was recorded.

Cracker product

A prototype of three commercially produced insect-fortified crackers were obtained from EXA, an insects as food and feed company located in Antananarivo, Madagascar. With input from the study team, EXA developed three cracker recipes and a control cracker, including combinations of the following ingredients: a local cricket species (Gryllis madagascariensis), black soldier fly larvae (Hermetia illucens (BSFL)), rice, corn, sugar, salt, ginger, turmeric, and cocoa. Rice and corn were selected as primary ingredients because they can be locally sourced. Small amounts of sugar and spices were chosen for palatability. Each 50 g daily serving of crackers contained 6–10 g of cricket and/or BSFL flour and was specifically formulated to provide ample nutrients as well as 0.65-1.0 gram of insect fiber. (As a note: insect fiber content is relevant to our broader research objectives to assess the health impacts of insect fiber consumption in children.) To prepare the insect powder, insects were cleaned, boiled at 100 °C for one minute, dry roasted at 110 °C to reach 5% moisture, ground into powder, vacuum sealed in plastic pouches, and stored at 18–24 °C. To reach 5% moisture, crickets were dry roasted for 4 h and BSFL were dry roasted for 45 min. To prepare the insect-fortified crackers, raw materials were weighed and mixed, passed through an extruder (savory snacks: 120 °C; sweet snacks: 150 °C) and die cutter, baked (130 °C) for eight minutes, cooled, packed, and stored (15–30 °C) prior to consumption.

Product Safety and Nutrient Content

The insect powder ingredients and the insect-fortified cracker products both received a formal ‘Certificate of Consumption’ from the government of Madagascar in 2023 and 2024, respectively, indicating approval for commercial sale of this product. Since neither a universally adopted nor U.S.-derived food safety standards list has been set for edible insect products, the nutrition and food safety parameters for the crackers were established based on published standards that do exist for edible insects and other animal-based food products across the globe. In Europe, for example, six Novel Food authorizations have been established for edible insects, making their sale legal across countries⁵⁷. In the United States, based on extensive historical consumption precedent for acceptable levels of insect parts in human food, edible insects farmed explicitly for human consumption can be generally regarded as safe (GRAS) via self-affirmation (CFR 2018). The list of essential food safety parameters, including heaving metals, biologic, and chemical hazards, was developed from standards in the CODEX Alimentarius International Food Standards⁵⁸, the Good Hygiene Practices for European Union Producers of Insects as Food and Feed from IPIFF⁵⁹, the Commission implementing regulation (EU) 2022/188 CELEX⁶⁰, Health Canada (2013)⁶¹, and the Bureau of Madagascar Standards⁶².

Nutritional and heavy-metal analyses of the cracker prototype samples were conducted by Eurofins Microbiology Laboratories (Cascade), USA, using standard AOAC (Association of Official Analytical Chemists) methods, with all tests performed in duplicate. Heavy metals (arsenic, cadmium, mercury, and lead) were quantified using AOAC methods 2011.19, 993.14, and 2015.01, which primarily involve inductively coupled plasma mass spectrometry (ICP-MS) for detecting trace elements. For nutritional composition, calories were determined by using standard Atwater factors in accordance with Code of Federal Regulations (CFR) guidelines. Crude fat was determined by AOAC 920.39, protein by AOAC 990.03 and 992.15, dietary fiber (total, soluble, and insoluble) by AOAC 991.43, and iron by AOAC 984.27. Later, the exact batch samples of crackers used in the study were analyzed using these same methods by the Institut Pasteur de Madagascar (Laboratoire d’Hygiéne des Aliments et de l’Environnement) to ensure the crackers met essential safety standards and were free of pathogens or hazards.

Study sites

This study was conducted in three rural schools in the Farafangana district of the southeastern Atsimo-Atsinanana region of Madagascar. We partnered with Catholic Relief Services (CRS), the implementer of a USDA-funded school feeding program called Mavitrika Mianatra (“active learning” in Malagasy), or MaMi. The MaMi project provides daily rations of fortified rice, beans, and fortified vegetable oil to about 80,000 students across 185 schools in the region. The project also installed storage facilities for the food commodities at all schools, and in some cases latrines and/or other critical infrastructure. The schools were selected for this study among the subset currently receiving rations from the MaMi program that were within a two-hour drive from the town of Farafangana, and large enough to have about ~ 60 students in the targeted age group.

Study population and sample size

Approximately 60 children were enrolled into the study from each school. Children were eligible to participate if they were between the ages of nine and 13, could provide informed assent after a thorough explanation of the study, including what the crackers are made of, were present at school the week of the taste tests, and had a guardian willing to attend an informational session and provide informed consent. Children nine to 13 years of age were chosen for multiple reasons including stomach capacity (ensuring a 50 g snack would not interfere with appetite for the fortified lunch), uniform nutritional needs among this age group (e.g. RDAs for children 1-3yrs, 4-8yrs, 9-13yrs), this age group is understudied, school attendance is more regular than younger students, and they have a better memory of taste. Participant exclusion criteria included: (1) recent or current illnesses including malaria, fever, or severe diarrhea; (2) COVID-19 exposure or symptoms including loss of smell and taste; (3) chronic medical conditions or congenital anomalies requiring frequent medical attention or interfering with nutritional status or dietary consumption; (4) history of food allergies or adverse reactions to eating insects, shellfish, or other ingredients in the snack; (5) inability or unwillingness to comply with study requirements; and (6) absence on the first day of the study when the study doctor was present to treat potential study food allergies.

Study participants were members of rural, agrarian communities in the Atsimo-Atsinanana region. During a scoping trip in May 2023, the research team held conversations with parents at nearby primary schools to characterize the local diet, which is monotonous and primarily composed of rice, cassava, cassava leaves, jackfruit, breadfruit and (sometimes) sweet potato. Wild green leafy vegetables, and fruits (bananas, oranges, lychee, mango) are foraged during a short three-month window (Nov-Jan) by women and children. Eight different types of insects were listed as “snacks” on the seasonal food calendar, including grasshoppers, locusts, crickets, katydids, June beetles, aquatic beetles, wasps, and cicadas. Species consumption is not consistent and varies between families and season. Animal source proteins are rarely consumed due to their economic value, and fish are often sold to buy larger volumes of less nutrient-dense foods.

The minimum sample size for this study was estimated to be 55 children per school. This gave us the ability to detect a minimum effect of a 0.75 point difference on Likert scales with an assumed standard deviation of 1.4 for any two-way comparison of crackers in any one school. In sum, 187 students participated across the schools.

Recruitment and enrollment

A school-wide recruitment, consent, and enrollment event took place one day prior to study commencement, involving parents, school principals, teachers, the MaMi program school monitors, and in some cases the local village chief. The research team presented a study overview to the group and then conducted an individual informed consent process with each child’s parent or legal guardian. Child assent was also obtained from potential study participants (aged 10 to 13 years) meeting the eligibility criteria. Caregivers and child participants were given an opportunity to ask questions about the study in a group setting and privately. In the enrollment log, each caregiver-child dyad was given a participant ID according to school. The enrollment list included the participant ID, child name, gender, age, class, teacher’s name, and parent’s telephone number in case of an allergic reaction.

Intervention & measurements

Each day over a four-day period, children received 25 g of two different insect-fortified or an insect-fortified and (unfortified) control crackers. The cracker packets were labeled (A, B, C, and D), and were color-coded (red, yellow, blue, and green) for easy recall by the children. Throughout the school week, the cracker types were rotated through a series of pairwise comparisons. Each pairwise comparison was provided twice amongst the three schools. The sequence of tests across four days was as follows: school 1 (AD, BC, AB, CD), school 2 (BD, AC, CD, AB), and school 3 (BC, BD, AD, AC). They were monitored throughout product consumption by us and the enumerators. The students came in groups, were allowed enough time to eat, and were surveyed. Next, their leftover crackers were weighed. On the fifth day of each school week, children selected a 50 g packet of their preferred cracker. A summary of the pairwise comparisons is available in Appendix Table 6.

To evaluate acceptability, children were asked to rate each cracker using a 5-point Likert Scale with a visual aid. Children were asked to consider their overall liking and liking via organoleptic variables including appearance (color, shape), texture (finger feel), aroma (smell), and taste (flavor). These data allowed for an absolute rating for each cracker, as well as a pairwise comparison of two crackers. During each day of data collection, the enumerator recorded the quantity (g) of cracker consumed and if the snack bag contents were finished or not finished.

The study crackers were intended to be fed as a morning snack upon arrival (8:30am), but distribution was delayed by at least one hour on four of 12 days. All schools had parents onsite preparing the school lunch of fortified rice, beans, and fortified oil; vegetables (cauliflower, tomatoes, lychee, spinach, cabbage, beets, carrots) were incorporated into the school lunch at school 3. Parents assisted in collecting, boiling, and providing clean water to the study participants at the first two schools and children at the third school were instructed to bring water from home (typically boiled water from cooking rice).

To understand broader perceptions of the insect-fortified crackers, focus group discussions (FGDs) were conducted on the final day of data collection at each school with parents and caregivers of the child participants. School administrators selected and invited 20 parents and caregivers, half women and half men, to participate. Separate FGDs were held for men and women, following a semi-structured FGD guide in classrooms or outdoor areas adjacent to schools. During each one-hour FGD, the method of photo-elicitation4 was also employed to prompt participants to comment on images5 of insects; the images were designed to trigger memories and evoke deeper conversation than verbal questions could alone.6 FGDs explored participant perceptions of insects as food, ancestral practices and taboos, perceptions of the insect-fortified crackers served to the students, and asked for feedback on child experiences and perceptions. Caregivers were also asked to report on their child’s behavior throughout the week, changes they observed in energy or sleep, as well as any signs of allergies or gastrointestinal distress observed among children. All FGDs were audio recorded, summarized and translated to English and hand coded for qualitative thematic analysis.

The consent and assent processes, student interviews, and FGDs were facilitated by the research consulting firm Cabinet CAST, a local team composed of six members (three men and three women) with regional expertise and fluency in the local dialect of Malagasy.

Statistical analysis

Data management: Eligibility and enrollment data was gathered by the school administrators and study enumerators before data collection began. Daily data were entered directly on Android tablets using KoboToolbox (https://eu.kobotoolbox.org/) and saved the data locally in the Kobo Toolbox application. Once completed for each school on study day 5, the survey data was sent to the KoboToolbox survey for aggregation. We also measured the total consumption of snacks based on changes in cracker weight before and after the taste test. This study thus uses merged quantitative data from three different sources: the enrollment data, the snack survey data, and the snack consumption and completion data. The data was cleaned and analyzed using Stata 18.0 MP – Parallel Edition (https://www.stata.com/).

Descriptive analysis The three main types of descriptive statistics the study reports on are demographic characteristics of the study population, such as age, grade, and gender; preference information via a five-point Likert scale on insect crackers and the control cracker by five organoleptic traits; consumption pattern of four crackers under study, and the consumption pattern of the preferred crackers. These statistics are reported with and without first-day effects that impact consumption or preference due to initial hesitancy.

Regression analysis Consistent with the ordinal nature of the data, we follow McKelvey and Zavoina (1975)⁶³, and construct an ordered probit model.

Where, ∼ N (0,1), and is the latent continuous variable that determines the ordered categorical outcomes.

The observed outcome is mapped from using threshold cut points:

Here,  J is the number of cut-off points.

Four different types of o-probit regressions were run to understand the effects of regressors on –four categories of consumption, the Likert scale categories of preference on the five organoleptic traits and overall preference, and the four categories of consumption for preferred crackers. The marginal effects, p-values, and standard errors are reported here.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

MEP, VJS, and JH designed this study and developed the study protocol. MEP, ENR, CDA, JBU, and VJS collected the study data in Farafangana district. JBU and CDA performed the quantitative analysis. VJS, MEP, and ENR conducted the nutrition and safety analysis. VJS conducted the qualitative analysis. All authors contributed to the development of this paper and performed a review prior to submission.

Funding

This study was funded by a gift from King Philanthropies. The funder had no role in study design, data collection, analysis, or the preparation of this manuscript.

Data availability

The data that support the findings of this study are openly available in figshare at https://doi.org/10.6084/m9.figshare.30275950.v2, and the data with replication package for the same are available at https://github.com/chandrakantagme/acceptability_trial_madagascar_2024.git. Citation: Parker, Megan; Agme, Chandrakant (2025). Acceptability of insect-fortified snacks among school children in Madagascar. figshare. Dataset. https://doi.org/10.6084/m9.figshare.30275950.v2.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The study was approved by the Comite Malgache d’ethique pour les Sciences et les Technologies (CMEST) in Antananarivo, Madagascar and the Cornell University Institutional Review Board (Ithaca, NY). Additionally, the study was approved by the Nutrition Director of Farafangana, Chef Fokontany of each community, the National Ministry of Education (Ministere de l’Education Nationale), and the PATH scientific review committee. All research methods were performed in accordance with the relevant guidelines and regulations.