Nutritional, Texture, and Sensory Properties of composite biscuits produced from breadfruit and wheat flours enriched with edible fish meal

Abstract This study aimed to develop biscuits with improved nutritional contents using edible fish meal from catfish as the source of macro‐ and micronutrient enrichment while trying to reduce the input of wheat flour in biscuit‐making process. The biscuit was produced using edible fish meal (EFM: 0%–40%) from catfish, improved quality breadfruit (IQBF: 0%–60%), and wheat flours (WF: 0%–40%). Macro (crude protein, fat, fiber, ash, and carbohydrate)‐ and micro (calcium, magnesium, potassium, phosphorus, sodium, and iron)‐nutrient contents of the biscuit were determined. The color (lightness—L*, redness—a*, and yellowness—b*), texture (hardness, springiness, and adhesiveness), and sensory (taste, texture, and overall acceptability) attributes of the biscuits were assessed using standard methods. Model characteristics of the responses were profiled, and numerical optimization technique was used to predict combination/blends that produce biscuits with desired nutritional contents. Moisture, crude protein, fat, fiber, and ash values were in the range of 3.50%–5.57%, 3.06%–15.52%, 13.62%–26.00%, 0.31%–1.40%, and 1.98%–5.32%, respectively. The iron, calcium, and phosphorus contents of the biscuit ranged from 103.85 to 201.30 mg/100 g, 100 to 754 mg/100 g, and 8 mg/100 g to 304 mg/100 g, respectively. Interaction between the models for WF and EFM was significant and this significantly affected the L* (36.37–51.90) and adhesiveness (0.01–0.29) values for color and texture, respectively. Similar observations were also noticed for most of the nutrients. The quadratic models selected for the nutrients were all significant (p < .05) and the adjusted R 2 ranged from 0.61 to 0.84 and 0.59 to 0.97 for the macro‐ and micronutrients, respectively. In conclusion, a biscuit from IQBF, WF, and EFM of 61.33, 0.07, and 38.60 with protein, fat, ash, iron, and calcium contents of 10.41%, 17.59%, 2.05%, 120.52 mg/100 g, and 500.00 mg/100 g, respectively, was produced.


| INTRODUC TI ON
Biscuit (Cookies) are a group of confectionery products made from flour, sugar, egg, and shortening. As convenient food, its inclusion in the diet of Nigerians has increased appreciably, particularly among children and during festivities. Its acceptability as snacks cut across cultural and sociodemographic boundaries. The low-moisture nature of biscuit eases transportation difficulties and reduces potential health hazard from microbial contamination. Wheat flour has been a significant ingredient in conventional biscuit making. It was estimated that wheat imports might grow at 5 percent per annum, and the country could be importing as much as 10mmt per annum by 2030(AEGIC, 2015. Adoption of composite flour technology might have significant saving on foreign exchange, more so when most of the biscuits that appealed to people of various age groups can be made with soft flours. The use of nonwheat adjuncts that can play complementary roles in reducing protein-energy malnutrition and having potentials to alleviate associated social problems might be more promising options. Fish is an essential source of high-quality protein (Ohen & Abang, 2007), providing about 16% of animal protein consumed by the world's population. It provides 40% of the dietary intake of animal protein of the average Nigerian (Federal Department of Fisheries, 1997). Catfish is highly nourishing. It contains lysine as well as vitamin A that is necessary for healthy growth. It contains some quantities of calcium, phosphorus, fat, and other nutrients needed for human growth and health (FAO, 2003). Nigeria is one of the largest importers of fish in the developing world, importing about 600,000 metric ton annually to meet the country's high demand for fish (Olagunju et al., 2007). Updated official production statistics are unavailable, and production is presently estimated at 1.7 million metric tons of fish annually (Abba, 2012). Catfish farming is undertaken by a large number of people, especially the small-scale farmers in Nigeria (Oladejo, 2010). The current motivation of smallscale farmers in Nigeria needs to be encouraged and sustained. Most of these farmers are into mainly catfish production, and these are sold mostly in a raw state to middlemen for sale to domestic consumers. It is, therefore, necessary to develop value addition chain that would encourage increase production and industrial utilization. This can be done by developing an edible fish meal (EFM) from catfish and using it as a form of nutritional supplementation. EFM is the product obtained from the removing of water and oil from whole fish, thereby increasing the concentration of protein and other nutrients (Ibrahim, 2009). EFM provides the opportunity of utilizing other nutrients such as calcium that are available in the fish (Table 1).
Breadfruit (Artocarpus communis Forst) is a staple diet in many tropical countries. The tree fruits early between May and August, producing 50-200 fruits in a year. The mature fruit is round or ovoid, 15-20 cm in diameter, and weighs 2-10 kg on average. The fruit is produced mainly in Malaysia, the South Pacific Island, the Caribbean, and West Africa (Ragone, 2009). Total yearly production in Nigerian is about 10 million metric tons with potential to exceed 100 million metric tons with improved agricultural practice (Adewusi et al., 1995; Amusa et al., 2002). The economic utilization of Breadfruit has been limited by its poor storage properties which are about 1-3 days after harvest (Ragone, 2009). Conversion of breadfruit to the flour would provide a more stable storage form and also enhance its versatility (Morton, 1987;Oladunjoye et al., 2010). Although breadfruit is a perennial crop, it has relatively less competing for domestic uses and has a higher yield per tree per year that can be sustained for many years; an average-sized tree reportedly produced 400-600 fruits per year (NTBG, 2009). Its horticultural features can also be explored at the household level to mitigate the effect of climate change in addition to its economic advantage. The fruit is highly perishable and has an undesirable fruity odor that impacts major organoleptic challenges to the acceptability of products made from it. Breadfruit is enzyme active even during processing to flour. Improvement in flour quality can be achieved by modifying processing procedure and reducing processing time.
Replacement of wheat flour with other adjuncts influences many functional properties of the dough and quality attributes of the resulting biscuits even when it is partial. Biscuits are grouped into various classes such as crackers, sweet biscuits, strongly sweet, semi-sweet, wafers (Manley, 1983), each associated with peculiar quality characteristics, particularly texture. A major concern in applying composite flour technology to the development of bakery product is how to minimize the quality alteration, retain as much as practicable the quality attributes associated with an existing product, and make the product acceptable to the consumers. In cases where these become in impracticable, and products with unique quality have to be developed, it may be necessary to profile the quality attributes of such products. Development of nutritionally enriched biscuits may, therefore, require the profiling of its technological characteristics, particularly, its texture parameters and its relationship with sensorial texture to ascertain its suitability for industrial production and consumer's acceptance.
This study evaluated the quality attributes of biscuit produced from an edible fish meal (EFM), improved quality breadfruit flour (IQBF), and wheat flour (WF). Specifically, it used D-optimal design to model some of the nutritional quality, color, texture, spread ratio, and sensory qualities of the biscuit.
It also assessed the relationships between the sensory texture of the biscuits and some texture parameters of the biscuit. It described the quality attributes of biscuit that should be expected from the modeled responses.

| Preparation of improved quality breadfruit flour (IQBF)
The IQBF was produced within 24 hr of harvest, using a modified method of Bakare et al. (2012). The matured fruits were thoroughly washed to remove adhering latex and dirt, peeled manually, washed, grated, bagged, dewatered, and pulverized. The whole batch of pulverized mash was dried using the flash dryer (Nobex Flash dryer, Nobex Technical Company Limited, Idimu Lagos, Nigeria) with the following conditions. The loading time was 10 min; the temperature in the tube and that of inlet air were 180°C and 200°C, respectively.
The feeding characteristics were moisture content: 45%; mass density: 1,380 kg/m 3 ; and feed rate: 820 kg/hr. The powder was allowed to settle and discharge at every 10 min. The dried breadfruit was milled using locally fabricated hammer mills, sieved (W.S. Tyler, 8570 Blvd, Mentor, OH, United States) through a 250-μm mesh sieve, and sealed in polythene bags and used for analyses.

| Preparation of spice mixture
Formulated instant spice mixture (ISM) was prepared as described by Bakare et al. (2016) and added to confer flavor and functionality properties on the biscuit.

| Preparation of edible fish meal (EFM)
EFM was produced as described by Bakare et al. (2019)

| Experimental design and optimization
The D-optimal design used consisted of sixteen (16)  Fitness of each of the models was analyzed to identify the model that can best be used as a response predictor. The models were assessed for their adequacy for the experimental conditions and significant terms in each of the models were identified, and numeric optimization of the mixture blends was done based on set targets (  Bonfim et al. (2019). The crude fiber content of flour was determined by the trichloroacetic acid method as described by Entwistle and Hunter (1949). The mineral content of the biscuit samples was determined using the method described by Adeniji and Tenkouano (2008). In the analysis, 1 g of sample was weighed into a pyrex glass conical flask, and 10 ml of concentrated nitric acid was introduced into the flask with a straight pipette, and then, 5 ml of perchloric acid was also added.

| Texture profile of biscuit
The texture of the biscuits was determined using the Universal testing machine (model M500-100AT, Testometric, England).
Compressive stress was applied to the samples to determine the behavior of the biscuits under a compressive load. A flat plunger with a 75 mm diameter was attached to the crosshead of the machine. Each biscuit was compressed uniaxially at a depth of 15 mm (30% strain) with a crosshead speed of 102 mm/minutes. Both load cell and strain gauges were connected to a data logging system to record the data.
The stress-strain data were continuously logged into a computer, and the stress at failure was considered as the uniaxial compression strength (UCS) of biscuits. Specifically, the parameter measured were hardness, chewiness, gumminess, cohesiveness, springiness, stringiness, and force at the peak, deformation at the peak, energy to peak, adhesiveness, and energy to break.

| Physical properties
The spread ratio of the biscuit samples was determined, as described by Gaines (1991). Six biscuits edge to edge were used for the evaluation, and the average was noted. Spread ratio was calculated by dividing the diameter by thickness of the biscuits.

| Sensory evaluation
Sensory evaluation was conducted as described by Iwe (2002) by using quantitative acceptance to assess consumers liking for the biscuit. Thirty untrained panelists rated their liking or otherwise for the TA B L E 2 Optimization goals for process variables product from the blends on a nine-point Hedonic scale (1 = disliked extremely as compared to reference sample "R," and 9 = liked extremely as compared to "R").

| Data/statistical analyses
Data were subjected to analysis of variance. Differences between mean values were separated using Duncan multiple range tests (Duncan, 1955). Pearson product-moment correlation coefficients were used to test the relationships between some of the variables.
Independent t test was used for the validation experiment. Statistical analysis package for social science (version 23, IBM, Armonk, NY, USA) was used for these analyses.
Design expert software (version 7.00 Stat, Ease Inc., Minneapolis, MN, USA) was used for the experimental design and subsequent numerical optimization.

| Moisture and macronutrients content
The moisture content of food is one of the most important and widely used indices for determining the quality of dried processed foods. It is a measure of yield and quantity of food solids and can be a direct index of economic value, stability, and quality (WMC, K-State, NCI, and NAEGA, 2008). The moisture content of biscuits samples (  Fat is necessary for steroids and hormones produced in the body and serves as solvents for hormones. It contains essential fatty acids and acts as carriers for fat-soluble vitamins A, D, E, and K and has the highest caloric content compared to protein and carbohydrate. Thus, it provides relatively higher (9 calories per gram of fat) amount of energy (Roth, 2011). Fats are essential for the functioning and structure of body tissues. Extra fat is stored in adipose tissue and is burnt when the body has run out of carbohydrates. Deficiency symptoms of under consumption of fat beneath 10% of the total daily calorie requirement that have been reported include eczema (inflamed and scaly skin condition), retarded growth, and weight loss. On the other hand, excessive fat in the diet can lead to obesity or heart disease (Roth, 2011).
Studies have also pointed to an association between high-fat diets and cancers of the colon, breast, uterus, and prostate (Roth, 2011).
Fat intake is not expected to be more than 30% of total calories (US Department of Agriculture, Center for Nutrition Policy and Promotions, 2005). In this study, the crude fat of biscuits samples ( These values were relatively lower than values reported in the previous study for biscuits produced from fermented breadfruit flour prepared by a conventional process (Bakare et al., 2014). The method of preparation of IQBF used for this study yielded breadfruit flour with reduced crude fiber content (2.49%) compared to the crude fiber content range of 6.32%-9.04% reported for conventional breadfruit flour (Bakare et al., 2012). The implications of dietary fiber have been documented and cited in the previous publication (Bakare et al., 2012). Consumption of dietary fiber is, therefore, necessary for the prevention of constipation, hemorrhoids, and diverticular disease by softening and increasing the size of the stool. The optimal recommendation for dietary fiber intake is 20-35 g/day. However, consumption of too much fiber can induce discomfort, flatulence (abdominal gas), and diarrhea.
It also could obstruct the gastrointestinal tract if intake exceeds 50 grams per day (Roth, 2011). Insoluble fiber contains binders in the form of phytic acid or phytate, which can prevent the absorption of minerals such as calcium, iron, zinc, and magnesium.
Preponderantly, there is little no evidence to assert that the traditional diet of Africans and specifically Nigerians is deficient in dietary fiber despite the dietary transition occasioned by changing lifestyle so excess intake needs to be avoided.
The ash content provides insights into the mineral quality of the biscuit. The ash content of the biscuits ( The total energy content of the biscuit is presented in Table 2.
The estimated energy content of a food substance is a function of the total protein, fat, and carbohydrates present in the biscuit.

Mineral contents
The mineral contents of the biscuit presented in Table 2  (sodium, calcium, magnesium, potassium, and phosphorus) are required in amounts greater than 100 mg a day while the microminerals (iron) are needed in amounts smaller than 100 mg a day. acid-base balance, and regulation of muscle and nerve activities but excess intake of sodium is associated with cardiovascular conditions such as hypertension and congestive heart failure (Roth, 2011). The dietary reference intake (DRI) for sodium has been established at between 1,200 and 1,500 mg/day for adults (Roth, 2011). Therefore, the values observed in this study are within safe limits.
Iron is a component of hemoglobin and myoglobin, and it is needed for the delivery of oxygen to body tissues and cells. It is also needed by enzymes that are involved in the making of amino acids, hormones, and neurotransmitters (Agarwal, 2001;Ward et al., 2014). The recommended dietary reference intake of 10 mg and 15 mg for men and women above 11 years of age, respectively, can be easily met by the consumption of 100 g of this biscuit. The iron content of biscuit ( Magnesium is vital to both hard and soft body tissues. It is essential for metabolism and regulates nerve and muscle function, including the heart, and plays a role in the blood-clotting process (Roth, 2011). Though rare, the deficiency symptoms included nausea and mental, emotional, and muscular disorders. The magnesium content of the biscuits samples ranged from 57 to 72.50 mg/100 g There was significant (p < .05) difference in the level of magnesium of the biscuit samples. The recommended daily intake (RDI) of magnesium ranges from 80 to 130 mg for children of 1 to 8 years and 240 to 420 mg for adults of 9 to 70 years (Roth, 2011). Thus, 100 g of the biscuits with the lowest value of magnesium will supply 43.8%-71.25% of RDI for children.
Potassium is an electrolyte that is predominant in intracellular fluid. Like sodium, it is essential for fluid balance and osmosis. It is also necessary for the transmission of nerve impulses and for muscle contractions. Some of its deficiency symptoms included diarrhea, vomiting, diabetic acidosis, and severe malnutrition. Additional symptoms are nausea, anorexia, fatigue, muscle weakness, and heart abnormalities. The potassium content of the biscuits samples ranged from 377.50 to 582 mg/100 g ( WF, and 0% EFM (experimental runs 11 and 12) while the highest (304 mg/100 g) was in 60% IQBF, 0% WF, and 40% EFM (experimental runs 3 and 7). There were significant (p < .05) differences in the phosphorus content of the biscuit sample. Phosphorus is found in many foods, and its deficiency is therefore rare.

| Color
A combination of the color index, including, L* (degree of lightness), a* (degree of redness) and b* (degree of yellowness) that were determined (

| Texture profiling
The hardness value of biscuits samples ( 6.49% edible EFM) had the least value while experimental runs 15 (66.31% IQBF, 7.05% WF, and 26.64% edible EFM) had the highest value. The hardness values of the biscuits were significantly (p < .05) different from each other. The hardness value is the peak force that occurs during the first compression of the two-cycle imitative tests which attempt to simulate the conditions to which the material is subjected to in the mouth. Hardness, in this case, is an indication of the force required to compress food between the molars (Rosenthal, 1999;Scott-Blair, 1958). The lower the hardness values, the softer the biscuit. The values for hardness recorded in this study were higher than values reported for Maria cookies (Pereira1   et al., 2013). The study was unable to establish a significant relationship between most of the texture (hardness chewiness, gumminess, cohesiveness, springiness, and stringiness) parameters and the composite (IQBF, WF, and EFM) blends. Studies have reported a strong correlation between instrumental and sensory measures of hardness (Campbell et al., 2016). The significant positive relationship (n = 16, r = 0.508 p = .44) between the hardness values and extent of deformation at the peak of the test indicated that the more the force applied to overcome the hardness, the more would be the extent of deformation. This is to be expected since biscuit of this type that is produced from short dough is not expected to be crispy, snappy but crunchy and possesses less resistance to deformation or effort to fracture. The oral processing of dry crunchy biscuits has reduced oral processing and leads to a reduction in the amplitude of jaw movement in both vertical and medial-lateral planes which are the

TA B L E 7 Model properties and regression coefficients for proximate and mineral composition of biscuit
Chewiness value of biscuits samples (  (Rosenthal, 1999). Crunchy biscuit of this type should have reduced muscle activity during chewing and reduction in the number of chews required preparing the biscuit for swallowing. There were significant positive relationships between chewiness and cohesiveness ((n = 16, r = 0.525 p = .037) and also between chewiness and adhesiveness (n = 16, r = 0.629 p = .029).
The gumminess value of biscuits samples (  ((Trinh & Glasgow, 2012). Gumminess may not be mutually exclusive with chewiness since crunchy biscuit; a solid product with chewy textural characteristics upon mastication with saliva may become adhere to the teeth and become gummy. This may explain the significant relationships between gumminess and chewiness (n = 16, r = 0.932, p = .000), cohesiveness (n = 16, r = 0.594 p = .015), and adhesiveness (n = 12, r = 0.692, p = .013), respectively (Table 4).  Springiness is measured several ways, but most typically, by the distance of the detected height (hardness) during the second compression divided by the original compression distance. It organoleptically depicts how well a product physically springs back after it has been deformed during the first compression and has been allowed to wait for the target wait time between strokes. The springiness value of biscuits samples ( Adhesiveness is an indication of the extent of the stickiness of products. Work required overcoming the sticky forces between the sample and the probe (Trinh & Glasgow, 2012). Adhesion is measured as the negative work between the two cycles; however, in many instances, the product has stuck to the probe and does not separate when the highest point between the two cycles is just back to the original product height. There was a significant negative relationship between adhesiveness and IQBF (n = 16, r = −0.724 p = 0 0.008), hardness, chewiness, gumminess, springiness, and stringiness (Table 4) The typical curve for the texture profile is illustrated in Figure 1. The biscuits that are excessively adhesive would require more jaw movement in all three planes of movement during oral processing and more muscle activity even at equal compressive hardness (Çakir et al., 2012).

| Spread ratio
The values of the spread ratio of biscuits (

| Sensory quality
The sensory attributes of the biscuit are presented in Table 5. The values for appearance ranged from 6.36 to 7.12, with blend experimental runs number 3 and 8 having the lowest and highest values, respectively. The blend having 80.00%, 0.00%, and 20.00% IQBF, WF, and EFM were not significantly different (p > .05) in appearance from experimental runs 4 and 10 (without EFM), 14 and 16, respectively.
Sensory texture perception is a dynamic process influenced by senses of touch, sight but most importantly, by the oral processing of food in the mouth. The interrelationships between food structure, structural breakdown during oral processing, and sensory perception of texture are important for understanding the effects on satiation and satiety (Campbell et al., 2016) and also to food process development. The ratings of the texture of the blends (

| Model description
This section deals with the discussion on an appropriate model that best describes the relationship between the response and the factor variables that were selected. Model fitness test that was conducted to identify the model that can best be used as a response predictor. The desire was for the selected model to have insignificant lack of fit in order to demonstrate that the model fits the data (Myers & Montogomery, 2002). The tested models were assessed for their adequacy for the experimental conditions, and the significant terms in each of the models were identified.

| Model fitness for each of the quality attributes
In selecting the appropriate model that best describes the relationship between the responses (nutritional, color, and texture properties) and the factor variables, it was assumed that a second-order relationship would be relatively appropriate and that true function may be approximated by parabolic surface. This assumption was necessary because the actual form of a functional relationship between the response and the factor variables are unknown at this stage. If linear models are used for screening designs or robustness tests, each factor in the model only appears as a linear term. In this case, a linear term means a combination of a coefficient βi and a factor Xi. The interaction has similar uses as the linear but is more complex because of the additional interaction terms. An interaction term is the combination of two factors Xi and Xj with a conjoint coefficients ¯βij. Quadratic models are the most complex of the three basic model types and are used for the optimization processes. The quadratic model for the three factors model is as presented in equation (1) (Johnson & Nachtsheim, 1983). The fitness of each of the model was tested using the sequential model sum of squares (SMSS). The SMSS TA B L E 9 Model properties and regression coefficients for sensory attributes of composite biscuit indicated the contributions of the linear, two-factor interaction (2FI), quadratic and cubic polynomials terms to the totality of the model.

| Model testing
A quadratic model has suggested all the variables (not indicated on any of the Tables because of space constraint) except crude fat, estimation of total energy, calcium, and adhesiveness where the linear model was also suggested. These suggested models had the highest adjusted R 2 except where the model was aliased, cubic, or where there was no suggested model choice.
The residual errors in each of the selected model were compared with the "Pure Error" from replicated design points to ascertain the extent of their lack of fitness. An insignificant lack of fit is indicated by a low probability value ("Prob > F"), low standard deviation, high adjusted R-squared values, and a low predicted residual sum of squares (PRESS). These explained why these and other statistical parameters used in evaluating and selecting the best-fitted model (Zen et al., 2015;Bakare et al., 2019). From the equation constructed for the best-fitted model, positive coefficient presents a positive contribution toward the response and vice versa. Also, a contour plot and three-dimensional response surface graphs (Figures 2-4) for each response were generated for a better explanation.

| Adequacy of the models for the experimental conditions
The models were assessed for their adequacy for the experimental conditions and the significant terms in each of the models were identified (Tables 6-8). These showed the relationships between each of the measured parameters (responses) of each of the quality attributes and the composite ingredients. It is important to reiterate that for most cases, we choose to test the significance or otherwise of the models based on quadratic relationships. This was to avoid the error of selecting cubic models with false assumption of significance and also to avoid under exploring the significance of a selected linear model. For instance, for the proximate and mineral properties (Table 6), the significance or otherwise of the models was tested based on quadratic relationships. The p-value (Prob > F) and model where A, B, and C are as defined in the Tables.
In contrast, most of the textural properties were tested on quadratic models exception being adhesiveness that was tested on linear.  Figure 4). Spread ratio (Table 7) was also not a significant model under texture quality, and this suggested that the shape of the biscuit may need to modify to enable better appreciation of the appearance and measurement of the spread ratio.
Except for appearance, the models for sensory properties ( these variables were AB (IQBF * WF) in addition to the linear components of these models except for color where AC (IQBF * EFM) also featured. The model for sensory texture, in particular, was significant (p < .0013), and its relationship with adhesiveness (n = 16, r = 0.715 p = 0 0.009) was further reinforced by the fact that all the three components (IQBF, WF, EFM) and interactive effects of IQBF and WF were the significant terms in the model.

Attributes of the Biscuit
The goal of the optimization process in food product development is to determine the level of each variable from which a robust product with the desired quality characteristics may be obtained. In this study, nutritional improvement is a core objective without necessarily having to sacrifice the preferred quality attributes of biscuit.
Specific attention was on protein, fat, and ash for the macronutrients while textural and sensory attributes of biscuits are key indices of acceptance by consumers.
The aim of the optimization experiment in this study was to determine the level of each of the flour blends that collectively would result in biscuit with not less than 10% protein contribution from the animal source, reduced fat content preferably not more than 15% and possibly maximum ash content. This aim was accomplished by placing constraints on the responses (Table 1).
A desirability plot was generated by the software (Figure 5). It indicated the region with the optimum combination of flour blends that fall within the constraints that were placed on the responses.  Table 9. Verification experiment conducted on the selected optimization solution showed that there was no significant difference (p > .05) between the predicted and verified values for the quality attributes (degree of freedom = 44, t value = 0.08 and p-value 2-tailed = 0.994) (Table 10).

| CON CLUS ION
Biscuits of acceptable sensory quality were obtained from the sixteen blends of flour. Significant relationships were established between sensory texture and adhesiveness as well as between adhesiveness and the IQBF used for making the biscuits. The model for adhesiveness was the only texture parameter that correlated significantly with any of the composite ingredients (IQBF) and also with sensory texture an indication that specific functional properties might have a significant influence on the adhesive textural property of the biscuit made from the blends. The optimal combination of composite ingredients required to produce biscuits with desired quality attributes were 61.33% IQBF, 38.60% WF, and 0.07 EFM, respectively. The quality attributes include protein, fat, ash, iron, and calcium contents of 10.41%, 17.59%, 2.05%, 120.52 mg/100 g, and 500.00 mg/100 g, respectively.

ACK N OWLED G M ENTS
The authors acknowledge the contribution of the following individuals and body corporate to the success of this study: 4. JGSP Intercon Farms Nigeria for facilitating the procurement of the catfish used for the study.

Material testing laboratory of National Centre for Agricultural
Mechanization (NCAM) Ilorin, Nigeria

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L A PPROVA L
Ethics approval was not required for this research.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.