Influence of maturity, smoking, and drying of fresh maize on sensory acceptability and nutritional content of the developed porridges

Abstract This study investigated the potential of using the underutilized fresh maize in the preparation of porridge to contribute toward complementary feeding of children, and reductions in pre‐harvest losses. Fresh maize was harvested at different stages of maturity, blanched, smoked, and sun dried before milling into flours that were blended with soy flours for preparation of test porridges. The test flours were analyzed using the Association of the Official Analytical Chemists (AOAC) methods to determine their nutrient composition before preparation of the porridges. A trained consumer panel of 12 people, mothers and nursery school children tasted the porridges to rank acceptability and preference. Analysis of nutritional data showed that the test flours contained similar amounts of proteins, fats, and carbohydrates as the commonly used dried maize–soy flour blends. All the test porridges were generally accepted by the mothers and children due to the unique smoky and roasted aroma, brown color and the sweeter flavor even without the addition of sugar. Grainy texture and the presence of residues were the only unacceptable attributes in some of the test porridges. In conclusion, fresh maize–soy floor blends can be potentially used in complementary feeding of children at home and school as an alternative to other traditional maize flours. Optimization and fortification can help make the flours nutrient‐dense and most appropriate for child feeding at scale.


| INTRODUC TI ON
Maize is consumed widely in the world. In Malawi and other countries in sub-Saharan Africa, maize is the staple food eaten in different forms as porridge during breakfast or stiff paste for lunch and supper. Different groupings and tribes use different names for the stiff paste in the region (e.g., nsima, sadza, pap, ugali, and poshto).
Fresh and immature maize is one of the maize products that is highly traded and utilized in almost all households in Malawi. While maize remains an important staple, utilization of fresh maize is limited to boiling and roasting. A lot of fresh maize is harvested for consumption, but due to its perishability and lack of processing techniques, leftovers get wasted.
To extend the shelf life, women preserve fresh maize through blanching and drying by open sun and using smoke from firewood or charcoal. Fresh maize preserved this way is traditionally known as Viselera in some parts of Malawi. Williamson (1975). It is shelf stable, widely accepted and mainly eaten as a boiled snack by all family members in times of scarcity. Despite its popularity particularly in rural Malawi, there is no known industrial application of the product and, as a result, it remains un-investigated and not well understood. Since fresh maize has a wide acceptance, there is need to consider its use in a wider range; for example, by processing it into flour and developing new products in a bid to reduce pre-harvest losses and diversify diets. Currently, diets intended for feeding of children less than five are less diversified and predominated by native dried maize. Although highly consumed, maize has low protein (9.42%) and is deficient in essential amino acids, such as lysine and tryptophan, but has fair amounts of sulfur-containing amino acids (methionine and cysteine) according to FAO (1992). To improve the protein quality, maize is consumed together with other protein sources such as legumes, milk, soybeans, and amaranth seeds and leaves, which are relatively rich sources of lysine and tryptophan, but are low in sulfur amino acids contained in maize (Food and Agriculture Organization of the United Nations, 1992). In Malawi and other countries in sub-Saharan Africa, maize-soy blends are produced commercially and locally promoted by governments and their development partners for feeding young children and adults in schools, hospitals, orphanage centers, and refugee camps (Ng'ong'ola-Manani, Mwangwela, Sch€uller, Østlie, & Wicklund, 2014;Lawless & Heymann, 2010).
The aim of this study was to assess the nutritional content of smoked and sun-dried fresh maize flours and evaluate consumer acceptability of porridges prepared to contribute toward the development of foods for complementary feeding of young children in the country. Being the first of its kind, the present study will contribute valuable data to the food composition database currently under development in the country in addition to informing the public about other possible uses of fresh maize besides boiling and roasting.
Given that child under nutrition is high in Malawi with 37 in every 100 children under 5 years suffering from stunting and 11.7 percent underweight (National Statistical office 2015-2016) the study will F I G U R E 1 Process flow chart for the preparation of fresh maize preserves and flour

| Preparation of fresh maize preserves and maize preserve flours
Fresh maize preserves were prepared with slight modifications of method proposed by Dauda (2014). Fresh green maize harvested at different stages of maturity (20DAP, 27 DAP, 40 DAP) was blanched in boiling water for 15 min and cooled in cold water immediately. The sheaths were then removed before tying the cobs to poles in a wooden smoke kitchen and open sun dried ( Figure 1).
After drying, the blanched and dried maize was milled into flours that were blended with soy and used to prepare the test porridges.

| Preparation of the control porridge
Maize dried in the field after full maturity was milled and combine with soy to prepare control porridges for the study as traditionally done by Malawian women.

| Preparation of flour blends for test porridges
A total of ten flour blends were formulated by blending 80 g of either dried fresh maize flour or native maize flour with 20 g boiled or roasted, dried, and milled soy. Flours formulated were as follows: traditional maize flour + roasted soy flour(MRS); maize + Boiled soy (MBS); smoked milk stage maize + roasted soy(SMRS); smoked physiological maturity fresh maize flour roasted + soy(SMPRS); smoked milk + boiled soy (SMBS); smoked dent + boiled soy (SMDBS); smoked physiological maturity + boiled soy (SMPMS); smoked dent + roasted soy (SMDRS); sun-dried physiological maturity + boiled soy (SNPMBS); and sun-dried physiological maturity roasted soy (SNPMR). The test porridges were prepared by adding the flour to warm water until a soft paste was formed using high heat for 10 min and simmering at low heat for 50 min to avoid burning. Homemade maize-soy flour blends (locally known as Likuni phala) comprising of flours from dried fully matured maize and roasted soy were prepared and cooked in a similar manner as the control.

| Descriptive sensory analysis
The researcher recruited a panel of 10 trained judges consisting of five women and five men chosen purposively from the University staff and students. The judges undertook a rigorous descriptive analysis of six representative fresh and normally dried maize-soy blends (MRS, MBS, SNPMRS, SNPMBS, SMPMRS, and SMPMBS) before other sensory evaluation tests were conducted. Only judges with the ability to differentiate four basic tests (salty, sour, sweet, and bitter) in the triangle test used were recruited among the 20 tasters who expressed interest. Training of the sensory panel was done as described by Lawless and Heymann (2010) and Ng'ong'ola- Manani et al. (2014). The exercise helped to generate consensus regarding terms to be used to describe porridge samples of this study.

| Sample preparation and presentation
The six prepared warm porridges were presented to the panelists in the identical coded containers at random to avoid bias and obtain valid responses for analysis and interpretation.

| Descriptive sensory analysis
Descriptive sensory evaluation was conducted in triplicate using randomized blocked design in a well-ventilated, aerated, and lighted class room for 3 days after the seven-day training. The panelists Description Code evaluated the samples, rinsing their mouth with clean tap water were necessary in the process. Scoring was done using a nine-point hedonic scale ranging from 1 "dislike extremely" to 9 "like extremely." Attributes with statistically significant scores were taken for further analysis as described by Tomic et al. (2009 (Lawless & Heymann, 2010), with a score of one meaning that the child disliked the porridge, while a score of five meant that she liked it extremely. The mothers were instructed to present any remaining porridge to the research assistants who weighed and indicated the amount of plate waste in grams as part of the acceptability results.

| Consumer acceptability tests by the mothers
Mothers are the ones who cook porridge at home for children. After feeding young children as part of sensory evaluation for this study, mothers were asked to choose the most acceptable porridge based on color, taste, and texture using a nine Likert type hedonic scale.
The seven test porridges evaluated by the children were the ones given to the mothers of this study. On each tasting day, the mothers were asked to taste the porridge they were feeding their children and rate it using the aforesaid scale that ranged from 1 "dislike extremely" to 9 "like extremely."

| Preference ranking tests by the mothers
Besides the consumer acceptability test, the mothers were also asked to taste and rank their preference of the test porridges using a seven-point Food Action Rating Scale (FACT) sheet (Table 1) as described by (Lawless & Heymann, 2010).

| Statistical analysis
Sensory and nutrient analysis data were entered and analyzed in SPSS version 20 (Statistical Package for Social Sciences, Inc., Chicago, IL, USA). Descriptive statistics, such as means, frequencies, and percentages, were generated and used to describe the findings.
Tests for robustness of means were done using Games-Howell tests.
Principal component analysis (PCA), PSLR was used to determine variations and bring out strong patterns in the dataset as described by (Ng'ong'ola- Manani et al., 2014).

| Socio-demographic characteristics of the study subjects
The study subjects were 82 mothers and 113 nursery school children 3-5 years of age. Porridge is the most common food for young children while mothers are the main decision makers on what kind of porridge to prepare for their children. Table 2 presents more details about the socio-demographic characteristics of the study subjects. Table 3 shows the results for nutritional composition of fresh and traditional dried maize flours. In general, the flours prepared from dried and smoked fresh maize have shown similar amounts of proteins, fats, zinc, and iron with traditional dried maize, but higher amounts of ash and iron compared to maize flour from traditionally dried maize. The amounts of carbohydrates were almost similar to those of normally dried maize, but with lower levels of potassium.

| Descriptive sensory analysis
Six test porridges were evaluated by the trained panel of 12 people before assessments by the mothers and children. Attributes generated by the sensory panel to describe different test porridges and their scores are given in Table 4. Figure 3 presents a bi-plot to show sensory loadings and variations for the porridges generated from principal components analysis (PCA). Table 5 presents results from acceptability tests of the seven porridges. There were significant differences (p < 0.05) in the general acceptability of the prepared maize-soy blends.

| Observed consumption and left overs by children
Similarly, the study found no significant differences in the acceptability and preference of the different maize-soy blends presented.
Observations of plate waste showed that all the test porridges were well-accepted by the children, with exception of dent fresh maize-soy blend (SMDBS) porridge that had some leftovers. No significant differences (p < 0.05) were also recorded on the amounts of leftovers.

| Sensory results from the mothers and children
To understand the differences in acceptability of the developed porridges between mothers and children, Pareto charts were developed presenting samples from higher liking to the lowest liking ( Figure 4a and b).

| Drivers of acceptability and preference
To understand drivers of liking of different porridge blends, partial least square regression analysis (PLSR) and external preference mapping were performed on the data from consumer acceptability study. Figure 5a and b summarizes the results obtained.
TA B L E 3 Proximate nutrient composition of fresh maize flours dried by smoke and sun drying, and the control (normally dried maize flour)

| D ISCUSS I ON
Porridge is the most common food for young children. Nevertheless, its energy and nutrient content is often too low to meet their needs fully due to the high starch content of staple foods, such as maize, millet, sorghum, and cassava. Food analysis data from this study (Table 3)  In this study, nearly all the mothers were aware of the traditional practice of preserving fresh maize in the villages for reconstitution in future, but none of them had ever done so before. The majority TA B L E 4 Attributes generated to describe different maize-soy porridges of them claimed preparing children's porridge using dried maize-soy flour blends (locally known as Likuni Phala). As already explained, when asked to evaluate the test porridges whose maize was harvested while fresh and smoked, the mothers of this study found the porridges acceptable and preferred most the unique smoky and sweet flavors even without the addition of sugar. Cluster analysis and external preference mapping (Figure 5a et al., 2014). This study is the first of its kind to formulate and test the acceptability and preference of porridges prepared from smoked and dried fresh maize-soy blends. Given that every country is different with respect to diet and lifestyles by virtue of cultural, socioeconomic and contextual factors, there is still a strong need for more research in a wider variety of countries in sub-Saharan Africa to generate country-specific data for guiding nutrition programming and cookery of complementary foods.
The study was not without limitations. The green maize season falls in rainy months which generally presented problems in drying, with some of the maize cobs developing moulds.
Considering the need to get valid responses and since children have small stomachs, the study designed to carry out the sensory evaluation of the seven test porridges within two weeks, relying on the same cohort of subjects to allow easy comparisons of the results. In real life situation, this period was longer than anticipated; some mothers dropped out in the process.
Besides this, the cessation of the study led to frustrations of many children who got used to eating the porridges in the evaluation exercise. For example, when the study began the children were very excited taking it as one of the school feeding programmes and the numbers kept on increasing, for example, from 27 kids at the beginning to 111 at the end, only to realize that it ended sooner than they expected. No analysis of polyaromatic hydrocarbons(PAH) in the smoked maize products was conducted in this study. However, literature reviewed from other studies ( Table 6) reveals the wide nature of PAH in foods and the lack of a specific legislation on allowable PAH levels in maize. Although an allowable limit has been set by the European Union for baby products, none of such legislation exists for PAH in foods in Malawi.

| CON CLUS ION
Although fresh maize is consumed widely in developing countries     Gomaa et al. (1993) Smoked meat and meat products(Thirty-eight samples) smoked fish (39 samples)

Sweden
Nine samples of smoked meat (produced by traditional "sauna" smoking, where the food is directly exposed to hot smoke from a burning log fire) showed high BaP levels ranging from 6.6 to 36.9 μg/kg, exceeding the 5.0 μg/kg maximum level for smoked meat and fish established by the European Commission (Regulation [EC] No 208/2005).Six samples of smoked fish had BaP levels exceeding 5.0 μg/kg, the concentrations ranging from 8.4 to 14.4 μg/kg. Samples of meat and fish smoked by indirect technique, using smoke from an external smoke generator, all had BaP levels below the limit of quantification, that is, 0.3 μg/kg Wretling, Eriksson, Eskhult, and Larsson (2010) Toasted bread (direct toasting (flame-toasting, coal-grilling or gas oven-toasting) or indirect toasting (electric oven-toasting) None of electric oven and toaster bread were polluted; samples toasted by charcoal and flame grilling contained up to 350 μg/kg of total PAHs. Very low levels PAH levels were reported in several commercial toasted samples of bread. Benzo[a] pyrene ranged from no detectable to 0.23 μg/kg. Salgueiro, Falcón, Carballo, and Gándara (2008).

ACK N OWLED G M ENTS
We are grateful to our research assistants Gift Chisapo, Gift Monyadira, Madalitso Chilembo for their valuable contribution during product formulation and sensory evaluation sessions. Sincere thanks should also go to the Health Surveillance Assistants (HSAs) for their role in informing the chiefs and sensitizing women about the study during child growth monitoring sessions, the headmen of the villages where mothers came from and university staff and students for taking part in the descriptive sensory analysis, Lastly, we would like to thank the management of the University for funding the study as part of the MSc. program by the researcher.

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

E TH I C A L S TATEM ENTS
This study was approved by the Ethics Committee at Lilongwe University of Agriculture and Natural Resources (LUANAR). Actual participation in the study was based on full consent from the mothers and nursery school administrators. The study participants were informed by the researcher about the voluntary nature of the study, highlighting on their rights to refuse participation, the right to skip a particular sensory session they did not want to participate, and the right to discontinue the exercise at any time. Each respondent was also assured of the confidentiality of the sensory responses given.
The researcher and her assistants abided by their professional ethical conduct, such as neutrality, respect for tester's dignity, sensory perceptions, and data verification, throughout the period of data collection. Completed forms were not shared with anyone outside the study team.