Tortilla added with unripe banana and cassava flours: chemical composition and starch digestibility

Abstract

Tortilla is the main staple food in Mexico. The principal fractions of tortillas are carbohydrates, where starch is the major constituent. Tortillas containing corn dough (600 g/kg) plus unripe banana flour (400 g/kg) (Musa paradisiaca L.) (UBF), and corn dough (500 g/kg) plus cassava flour (500 g/kg) (Mahinot esculenta Crantz) (CF) were analyzed for chemical composition and in vitro starch digestibility, and the glycemic index was also predicted. Protein and dietary fiber content decreased in tortillas added with UBF and CF. Fresh tortilla with UBF had the highest resistant starch (RS) content and with CF the lowest one. An increase in RS content with the storage time was determined in control tortilla and with CF, but tortilla with UBF did not. The predicted glycemic index decreased in stored tortillas. The change in digestible carbohydrates in tortilla with the addition of non-traditional ingredients can be an alternative for people with special nutrition requirements.

La tortilla representa un alimento principal en la alimentación de la población de México. El principal componente de las tortillas son los carbohidratos, de éstos el almidón es la molécula más abundante. Se prepararon tortillas con masa de maíz (600 g/kg, base seca) y harina de plátano (Musa paradisiaca L.) (400 g/kg, base seca) y tortillas con masa de maíz (500 g/kg, base seca) y harina de yuca (Mahinot esculenta Crantz) (500 g/kg, base seca). Se almacenaron y analizó la composición química, digestibilidad del almidón y predicción del índice glucémico. El contenido de proteína y fibra dietaria disminuyó con la adición de las harinas. El contenido almidón resistente aumentó con el almacenamiento en la tortilla control y en la preparada con harina de yuca. La predicción del índice glucémico disminuyó con el almacenamiento. Por los cambios en la composición química y la digestibilidad de almidón, las tortillas elaboradas con harinas de fuentes no convencionales podrían ser una alternativa para la población con requerimientos de alimentación especiales.

Keywords:

• tortilla
• banana
• cassava
• starch digestibility
• resistant starch
• dietary fiber

Palabras clave:

• tortilla
• plátano
• yuca
• digestibilidad del almidón
• almidón resistente
• fibra dietaria

Introduction

For many Mexicans, corn tortillas represent about half of their daily calories. Maize tortilla is the main staple in the rural zones of Mexico, where people blend the masa (dough) with other ingredients such as bean, chickpea, faba bean, cheese, etc., in order to diversify the taste, odor, texture and nutritional characteristics of the tortillas. In the city zones, parents send their children to the neighborhood tortilla store (tortilleria) for fresh tortillas for every meal. On the other hand, dry instant corn-dough flours have been developed for industrial production and introduced into the marketplace. Nixtamalized corn flour (NCF) has a much longer shelf-life than the traditional wet corn-dough, and contains other added ingredients, such as vitamins and minerals, hydrocolloids, etc., to increase the functionality of these flours. NCF was blended with bean flour to produce a product named “enfrijolada” (MASECA, Tlanepantla, Estado de México, México) to produce a tortilla with improved nutritional characteristics and starch digestibility given by the legume (Hernández-Salazar, Agama-Acevedo, Sáyago-Ayerdi, Tovar, & Bello-Pérez, 2006). Government surveys have shown that a large percentage of Mexican children are deficient in iron, zinc and folic acid, resulting in stunted growth and other public health issues. However, the tortillas are lacking in key vitamins and minerals. Therefore, projects have been developed to improve the nutritional quality of tortillas, for example tortillas with added iron and other micronutrients (Dunn, Serna-Saldivar, Sanchez-Hernandez, & Griffin, 2008; Richins, Burton, Pahulu, Jefferies, & Dunn, 2008). Because corn tortilla is deficient in essential amino acids such as lysine, transgenic maize (genetically modified maize with the cDNA of amarantin to increase essential aminoacids) was nixtamalized and some technological and nutritional characteristics of tortilla were evaluated (Ayala-Rodriguez et al., 2009). Recently, due to concerns regarding obesity problems there is interest in blending nutraceutical ingredients with “masa” to obtain “healthy” tortillas. These ingredients would provide a greater volume per calories, which would help people to feel satisfied longer on fewer total calories, thus permitting weight loss and preventing weight gain (Rolls, 2009). Our research group reported starch digestibility and predicted glycemic index in tortilla with cooked bean (“taco”) (Sayago-Ayerdi, Tovar, Osorio-Diaz, Paredes-Lopez, & Bello-Perez, 2005), tortilla prepared with amaranth flour (Islas-Hernández, Rendón-Villalobos, Agama-Acevedo, Tovar, & Bello-Pérez, 2007) and tortilla with added flaxseed flour (Rendon-Villalobos, Agama-Acevedo, Osorio-Diaz, Tovar, & Bello-Pérez, 2009), and a blend of bean and corn dough of quality protein maize (Grajales-García et al., 2012). Resistant starch (RS), in part, is a food component that contributes to decreased glycemic index; due to this reason it is included in the dietary fiber (DF). Unripe banana is a rich source of RS (Faisant, Gallant, Bouchet, & Champ, 1995), and recently it was reported that cassava has RS content between 50 and 196 g/kg (Mejía-Agüero, Galeno, Hernández-Hernández, Matheus, & Tovar, 2012). Maize, unripe banana and cassava (in a lower amount) contain plant cell wall materials that include DF. Dietary fiber is another food component that interferes with starch digestion. Starch is the principal component of cereals, legumes, tubers and unripe fruits and is the main source of energy in human diets. Starch can be inaccessible to digestive enzymes because other components present in the food (as DF) produce a physical barrier or gelatinized starch during cooling and can produce a retrograded structure (Hendrich, 2010). Unripe banana and cassava are starchy raw materials that are traditionally blended with corn dough to produce tortilla in the small towns in the south of Mexico. Banana and cassava are food crops that are underused in México because both are harvested in specific regions (tropical and subtropical) (Bello-Perez, Agama-Acevedo, Gilbert, & Dufour, 2012), and their human consumption is restricted. However, there are no scientific studies that report on the starch digestibility of those tortillas. The objective of this study was to evaluate the addition of banana and cassava flours on the chemical composition and RS content of maize tortilla.

Materials and methods

Flour preparation

Commercial hard green (unripe) preclimateric banana (Musa paradisiaca L.) and cassava tubers (Mahinot esculenta Crantz) were purchased from a local market in Tuxtepec, Oaxaca, Mexico. The banana and cassava flours were prepared as previously reported (Ovando-Martinez, Sayago-Ayerdi, Agama-Acevedo, Goñi, & Bello-Perez, 2009).

Preparation of tortillas

Commercial “masa” (corn dough) was purchased in a local “tortilleria” of Yautepec, Morelos, Mexico. Corn dough was blended with banana flour to produce a 60:40 ratio (weight/weight, dry basis) and with cassava flour to produce a 50:50 ratio (weight/weight, dry basis), followed by addition of water to form “masa” with the consistency to make tortillas. Previous experiments at our laboratory showed that these were the maximum substitution levels allowed to prepare tortillas with adequate functional characteristics. Dough was molded by pressure and extruded (Tortilladoras González, Naucalpan, México) into thin circles to obtain 2-mm-thick tortillas. Tortillas were cooked on a hot griddle for 1 min per side at an approximate temperature of 250 ± 5°C. After cooling, the samples were frozen using liquid nitrogen and freeze-dried. In the case of stored tortillas (24, 48 and 72 h), the samples were frozen using liquid nitrogen and freeze-dried. Commercial tortillas from the same “tortilleria” were used as a control.

Chemical composition

Moisture content was determined gravimetrically (130 ± 2°C for 2 h) using 2–3 g of ground sample. Ash, protein, and fat were analyzed according to AACC (2000) methods 08-01, 46-13, and 30-25, respectively. Dietary fiber was tested using 985.29 AOAC (1999) method.

In vitro digestibility tests

Total starch content was determined by the method of Goñi, Garcia, and Saura-Calixto (1997). In brief, a 50 mg sample was dispersed in 2 mol/L KOH to hydrolyze all the starch (30 min) and subsequently incubated (60°C, 45 min, pH 4.75) with amyloglucosidase (Roche No. 102 857, Roche Diagnostics, Indianapolis, IN, USA); the glucose content was then determined using the glucose oxidase/peroxidase (GOD/PAD) assay (SERA-PAK® Plus, Bayer de México, SA de CV). Total starch content was calculated as glucose (mg) × 0.9; potato starch was used as a reference.

Total RS was measured by the method proposed by Goñi, Garcia, Mañas, and Saura-Calixto (1996). This method reports the sum of native indigestible fractions, retrograded fractions and a substantial part of physically inaccessible starch. In brief, the protocol comprises removal of protein with pepsin P-7012 (Sigma Chemical Co., St Louis, MO, USA) at 40°C and pH 1.5 for 1 h, incubation with amylase A-3176 (Sigma Chemical Co., St Louis, MO, USA) at 37°C for 16 h to hydrolyze digestible starch, treatment of the precipitate with 2 mol/L KOH, and incubation with amyloglucosidase A-7255 (Sigma Chemical Co., St Louis, MO, USA) at 60°C and pH 4.75 for 45 min and determination of glucose using the GOD/PAD assay (SERA-PAK® Plus, Bayer de México, SA de CV). The in vitro rate of hydrolysis was measured using hog pancreatic a-amylase according to Holm, Bjorck, Drews, and Asp (1986). Each assay was run with 500 mg of available starch. The predicted glycemic index was calculated from the α-amylolysis curves, using the empiric formula proposed by Goñi et al. (1997).

Determination of rapidly digestible starch (RDS), slowly digestible starch (SDS), and resistant starch

RDS, SDS, and RS starch fractions were measured on the basis of Englyst, Kingman, and Cummings (1992). Samples of food (containing < 0.8 g carbohydrates) were weighed into 250 mL Erlenmeyer flasks. Fifty milligrams of guar gum powder were added. Then, 10 mL of pepsin solution (50 g/L of pepsin in 0.05 mol/L of HCl) was added. The flasks were placed in a water bath at 37°C for 30 min. Five milliliters of 0.5 mol sodium acetate/L (equilibrated to 37°C) were added to each tube to form a buffer at pH 5.2. After adding 5 mL of enzyme mixture, the shaking action of the water bath was started, which was taken as time zero for the incubation and was not interrupted until all the G₁₂₀ portions were collected (see below). After 20 min, an aliquot (0.5 mL, G₂₀) from each flask was taken, and placed in a centrifuge tube containing 20 mL 66% ethanol. After a further 100 min a second, 0.5 mL sample (G₁₂₀) was taken in the same way. The G20 and G120 portions were centrifuged at a low speed (1500×g) for 5 min.

When all the G₁₂₀ portions had been collected, the flasks were placed together into a boiling water bath and left for 30 min. The tubes were cooled in ice water for 15 min. Then, 10 mL of 7 mol/L KOH was added. The flasks were placed in a shaking water bath containing ice water and shaken for 30 min. Tubes were removed singly from the ice water and 1 mL of the content was transferred into tubes containing 10 mL of acetic acid (0.5 mol/L). Amyloglucosidase solution was added to these tubes, and the tubes were placed into a 70°C water bath for 30 min followed by an additional 10 min in a boiling water bath. The tubes were cooled to room temperature before the addition of 40 mL water and centrifuged at 1500 × g for 5 min. This was the total glucose (TG) portion. The glucose was determined from G₂₀, G₁₂₀ and TG, using GOD/PAD assay.

Statistical analysis

The results were expressed as the mean ± standard error of three separate determinations. Comparison of means was performed by one-way analysis of variance (ANOVA) followed by Tukey's test. Statistical analyses were run using SPSS Version 6.0 software (SPSS Institute Inc., Cary, NC, USA).

Results and discussion

Chemical composition

The ash content of tortillas with the addition of unripe banana and cassava flour was lower than its respective flour, and no difference was found among the three tortillas (Table 1). This pattern can be due to the reason that the ash content in maize (between 11.0 and 17.0 g/kg, Méndez-Montealvo et al., 2005) is lower than in unripe banana and cassava flours, producing a dilution effect in the tortillas elaborated with those non-conventional flours.

Table 1. Chemical composition of tortilla added with banana and cassava flour.
Tabla 1. Composición química de tortilla adicionada con harina de plátano y harina de casava.

Sample	Ash (g/kg)	Fat (g/kg)	Protein^A (g/kg)	Total starch (g/kg)	TDF (g/kg)
Cassava flour	21.0 ± 1.0^a	53.0 ± 1.4^a	12.0 ± 0.7^a	808.0 ± 8.0^a	98.0 ± 3.0^a
Banana flour	25.0 ± 1.0^b	4.4 ± 0.2^b	40.0 ± 0.5^b	726.0 ± 3.0^b	107.0 ± 2.0^b
Control tortilla	14.0 ± 1.0^c	23.0 ± 2.0^c	77.0 ± 3.0^c	780.0 ± 25.9^a	108.0 ± 2.0^b
Cassava tortilla (50:50)	15.0 ± 0.7^c	15.0 ± 0.7^d	52.0 ± 0.5^d	757.0 ± 9.0^a,b	52.0 ± 3.0^c
Banana tortilla (40:60)	14.0 ± 0.6^c	25.0 ± 1.0^c	58.0 ± 1.0^d	786.0 ± 7.6^a	49.0 ± 3.0^c
Note: Results are means of three replicates ± SEM. Means with the same superscript letter with in a column are not significantly different at p < 0.05 level by Tukey's multiple range tests. Composition expressed on a dry weight basis. SEM = standard error of the means; TDF = total dietary fiber; ^AN × 6.25.
Nota: Los valores son la media de tres repeticiones ± EE. Medias con la misma letra dentro de la misma columna no son significativamente diferentes, p < 0,05, con la prueba de Tukey. Los valores están expresados en base seca. SEM = EE = Error estándar. TDF = Fibra dietaria total.

Tortilla prepared with unripe banana flour (UBF) shows a slightly higher protein content than tortilla prepared with cassava flour, although an important difference was found in the protein content between both flours. This effect can be explained by the protein content of maize since values between 83.0 and 113.0 g/kg were reported in different Mexican hybrids and varieties (Méndez-Montealvo et al., 2005), producing a dilution effect in this component when maize and these flours are blended to elaborate tortillas. However, other protein-rich flours, such as amaranth and flaxseed, increased protein content of tortilla when they were blended in a ratio 80:20 (nixtamalized maize flour:flour), with reported values of 106.0 g/kg (Islas-Hernández et al., 2007) and 129.0 g/kg (Rendon-Villalobos et al., 2009), respectively.

Tortilla added with UBF displayed the lowest fat content; this value agrees with the levels of this component determined in its flour. Tortilla added with cassava flour and control tortilla had similar fat content, pattern related with fat content of cassava flour (53.0 g/kg), and the fat content determined in different Mexican maize varieties (between 41.0 and 70.0 g/kg, MéndezMontealvo et al. 2005).

Cassava flour presented higher total starch content than UBF (Table 1), values that are lower than those determined in hybrids H515 (860 g/kg) and H516 (805 g/kg) used for tortilla production in the south of Mexico (Méndez-Montealvo et al., 2005). Cassava also presents high total starch content as was reported in 25 varieties, ranging between 742 and 814 g/kg (dry basis) (Mejía-Agüero et al., 2012). However, the total starch content in the three tortillas studied was not different. Tortilla added with 20% of flaxseed flour showed 602.0 g/kg of total starch content (Rendon-Villalobos et al., 2009), and with amaranth flour at the same level did not present an important difference in total starch with its control (747.0 g/kg) (Islas-Hernández et al., 2007).

A lower value of DF was determined in tortillas added with cassava and UBF than the control tortilla (Table 1). Cassava and UBFs had higher DF content than tortillas added with these flours. Wide variation in DF content was determined in maize varieties, ranged between 71.4 and 130.5 g/kg (Méndez-Montealvo et al. 2005). A dilution effect can explain this pattern. A non-starch polysaccharides content of 132.0 g/kg was reported in tortilla prepared with 20% flaxseed flour, and the control tortilla analyzed in this study (98.0 g/kg, Rendon-Villalobos et al., 2009) presented similar value to that determined in the control tortilla here.

Resistant starch content of stored tortilla

“Fresh” tortilla (0 h) added with UBF showed the highest RS value followed by control tortillas and tortilla added with cassava flour had the lowest RS level (Table 2). This is in agreement with the high RS content of unripe banana reported using two methods (473.0 g/kg and 572.0 g/kg) (Faisant et al., 1995), and that determined in UBF (304.0 g/kg) (Rodriguez-Ambriz, Islas-Hernandez, Agama-Acevedo, Tovar, & Bello-Perez, 2008). However, after heat treatment, a decrease in RS content was recorded in UBF (175.0 g/kg). This confirms the effect of the heat treatment on the sample, destroying the granular crystalline structure of starch with the concomitant decrease in RS amount. Tortilla prepared with cassava flour had the lowest RS value, indicating that starch retrogradation was minor in this sample. This pattern could be related with crystalline arrange of cassava starch that present low resistance to enzymatic hydrolysis. It was reported that amylopectin with high amount of short chains has higher retrogradation rate (Yuan, Thompson, & Boyer, 1993). Additionally, higher amylose content was reported in banana starch (370.0 g/kg), and can be responsible for its higher retrogradation rate (Aparicio-Saguilán et al., 2005).

Table 2. Resistant starch in tortilla of tortilla added with banana and cassava flour stored at different times.
Tabla 2. Contenido de almidón resistente en tortilla adicionada con harina de plátano y harina de casava, a diferente tiempo de almacenamiento.

	Resistant starch (g/kg)
Sample	0 h	24 h	48 h	72 h
Control	18.0 ± 5.0^a	25.0 ± 1.0^a	27.0 ± 2.0^a	28.8 ± 2.0^a
Cassava tortilla (50:50)	6.0 ± 1.0^b	10.0 ± 1.0^b	11.0 ± 0.1^b	15.0 ± 2.0^b
Banana tortilla (40:60)	26.0 ± 3.0^c	29.0 ± 1.0^c	28.0 ± 1.0^a	29.0 ± 2.0^a
Note: Values followed by the same superscript within the same column do not differ significantly at p < 0.05.
Nota: Valores con la misma letra dentro de la misma columna no son significativamente diferentes, p < 0,05.

An increase in RS content was recorded in the sample stored for 24 h, depending on the tortilla type, and a plateau was obtained after 48 h of storage (Table 2). A similar pattern was found for tortilla added with amaranth flour; “fresh tortilla” showed 14.0 g/kg of RS and 20.0 g/kg was determined in tortilla stored for 72 h (Islas-Hernández et al., 2007). However, “fresh tortilla” with 20% of flaxseed flour presented an RS content of 51.0 g/kg (Rendon-Villalobos et al., 2009), but stored samples were not analyzed. The amylopectin structure plays an important role in the starch retrogradation, and more studies are necessary in this sense.

Rate of enzymatic starch hydrolysis

Fresh tortillas showed a slight difference in the hydrolysis rate (Figure 1a); in general, they were not significant (see error bars). Control tortilla showed lower hydrolysis values in the first 15 minutes of the test; then, the values were not significantly different (p < 0.05) in tortillas added with unripe banana and cassava flours. The addition of unripe banana and cassava flours apparently does not prevent starch reorganization in maize to decrease hydrolysis rate. It was reported in textural (Román-Brito, Agama-Acevedo, Méndez-Montealvo, & Bello-Pérez, 2007) and starch digestibility (Bello-Perez, Rendon-Villalobos, Agama-Acevedo, & Islas-Hernandez, 2006) studies that starch reorganization in tortillas can be produced immediately after cooling. However, a dilution effect cannot be discarded because the molecular and structural characteristics of starch in unripe banana and cassava flours could not be responsible for producing a reorganized structure resistant to enzymatic hydrolysis. At the longest storage time (72 h) the difference between control tortilla and those added with unripe banana and cassava flours was more evident (Figure 1b). Control tortilla displayed lower hydrolysis rate during the course of the in vitro test than tortillas prepared with unripe banana and cassava flours. This confirmed that the composition of maize and unripe banana or cassava flour produces tortilla with a higher in vitro hydrolysis rate, and this pattern can be explained by the low starch reorganization (retrogradation). Tortillas at different storage times were compared (Figure 2). Two groups can be observed: fresh tortilla and those stored for 24 h showed similar hydrolysis pattern, as did those stored for 48 and 72 h; however, the difference was minimum. The difference in the hydrolysis susceptibility was in accordance with the increase in RS content in tortilla added with amaranth flour (Islas-Hernández et al., 2007).

Figure 1. In vitro starch hydrolysis of tortilla added with banana and cassava flours. (a) 0 h; (b) 72 h of storage time. Δ, Control; •, Cassava tortilla; □, Banana tortilla.

Figura 1. Hidrólisis in vitro del almidón de tortilla adicionada con harina de plátano inmaduro y harina de casava. a) 0 horas; b) 72 horas de almacenamiento. Δ, Control; •, Tortilla con harina de casava; □, Tortilla con harina de plátano.

Figure 2. In vitro starch hydrolysis of tortilla added with banana flour stored at times. Δ, 0 h; •, 24 h; □, 48 h; ♦, 72 h.

Figura 2. Hidrólisis in vitro del almidón de tortilla adicionada con harina de plátano, almacenada a diferente tiempo. Δ, 0 h; •, 24 h; □, 48 h; ♦, 72 h.

When the glycemic index cannot be tested in vivo, it can be estimated in vitro. The starch hydrolysis rate data for tortillas after different storage times allowed the calculation of predicted glycemic index (pGI) according to Goñi et al. (1997). Fresh tortilla prepared with UBF showed the highest pGI value, and no difference was found for control tortillas and tortillas added with cassava flour (Table 3). Tortillas stored for 24 h had similar pGI value, tortillas stored for longer times displayed lower pGI, and the effect was higher for control tortilla and tortilla prepared with UBF. This pattern agrees with the RS content determined in both samples. The molecular structure and the arrangement of the starch components play an important role in this pattern (Casarrubias-Castillo, Hamaker, Rodriguez-Ambriz, & Bello-Perez, 2012).

Table 3. Effect of storage time on predicted glycemic index^a of tortilla added with unripe banana and cassava flours.
Tabla 3. Efecto del tiempo de almacenamiento en la predicción del índice glucémico de tortillas adicionadas con harina de plátano y harina de casava.

	Glycemic index
Sample	0 h	24 h	48 h	72 h
Control	102.34^a	103.13^a	90.53^a	91.61^a
Cassava tortilla (50:50)	102.47^a	103.31^a	98.63^b	99.87^c
Banana tortilla (40:60)	106.31^b	104.48^b	98.02^b	96.46^b
Note: Calculated from the equation proposed by Goñi et al. (1997). Values followed by the same superscript within the same column do not differ significantly at p < 0.05.
Nota: Calculado con la ecuación propuesta por Goñi et al. 1997. Valores con la misma letra dentro de la misma columna no son significativamente diferentes, p < 0,05.

Digestion properties of tortillas

Fresh tortillas (control and those added with unripe banana and cassava flours) had similar RDS. RDS is the main fraction present in fresh tortilla followed by SDS and RS (Table 4). Native starch showed high resistance to enzymatic hydrolysis. When the starchy products are cooked, the semicrystalline structure of native starch is completely disorganized with the increase in the RDS (Zhang, Ao, & Hamaker, 2006). In this sense, the addition of both flours to tortillas did not modify the RDS content. This pattern may be due to the fact that fresh samples were analyzed where starch is more easily digested, and perhaps higher effect in the three starch fractions (RDS, SDS and RS) could be seen in stored tortillas. Diverse cereal products were tested for rapidly available glucose with the Englyst test that is comparable with RDS. Special K® (645.0 g/kg), claimed as a product with high fiber content, had a similar RDS value to the tortillas. The slowly digestible starch content of control tortilla and that with cassava flour was similar, and the highest value was determined in tortilla with added UBF (Table 3). This pattern shows the specific characteristics of starch digestion of unripe banana due to the molecular and structural characteristics of its starch (Aparicio-Saguilan et al., 2005; Espinosa-Solis, Jane, & Bello-Perez, 2009).

Table 4. In vitro starch digestibility values of tortillas added with unripe banana and cassava flours.
Tabla 4. Digestibilidad in vitro del almidón de tortilla adicionadas con harina de plátanos y harina de casava.

	Starch fraction (g/kg)
Sample	RDS	SDS	RS
Control	665.8 ± 4.6^a	65.2 ± 1.2^a	19.1 ± 2.9^a
Cassava tortilla (50:50)	658.8 ± 7.8^a	70.0 ± 9.4^a	28.3 ± 8.6^a
Banana tortilla (40:60)	667.1 ± 3.2^a	88.2 ± 8.6^b	30.7 ± 5.9^b
Note: Results are means of three replicates ± SEM. Means with the same superscript letter within a column are not significantly different at p < 0.05 level by Turkey's multiple range tests. SEM = standard error of the means; RDS, rapidly digested starch; SDS, slowly digested starch; RS, resistant starch; TS, total starch. RS was calculated as TS – (RDS + SDS).
Nota: Los valores son la media de tres repeticiones ± EE. Medias con la misma letra dentro de la misma columna no son significativamente diferentes, p < 0,05, con la prueba de Tukey. SEM = EE = Error estándar; RDS, almidón rápidamente digerible; SDS, almidón lentamente digerible; RS, almidón resitente; TS, almidón total. RS se calculó como TS – (RDS + SDS).

SDS of cooked normal maize starch was 83.0 g/kg (Zhang et al., 2006), which is similar to that determined in tortillas. However, the process to make tortillas is more severe, the cooking of maize with calcium hydroxide, the grinding of nixtamal and the cooking of the tortillas produces extensive starch gelatinization. Still, tortillas retain important SDS content. Cornflakes (31.0 g/kg) and Special K® (30.0 g/kg) (Englyst, Vinoy, Englyst, & Lang, 2003) had low slowly available glucose, and these values were lower than those determined in tortillas, but crackers had slowly available glucose value of 68.0 and 96.0 g/kg (Englyst et al., 2003) that are comparable to those determined in tortilla. These results show the slow digestion properties of tortillas, and currently there are no studies reported on these starch fractions in tortillas. Products with high level of SDS provide the nutritional benefits of starch (as supply of glucose), because they do not produce postprandial hyperglycemic and hyperinsulinemic spikes associated with the RDS (Han & BeMiller, 2007). RS determined by Englyst's method (Englyst et al., 1992) for control tortilla and tortillas with added UBF was similar to those determined by Goñi's method (Goñi et al., 1996). Lower RS content with Goñi's method was assessed for tortilla prepared with cassava flour than Englyst's method.

Conclusions

Tortillas with added UBF and cassava flour showed lower protein content than control tortillas. The DF content of tortillas with added unripe banana and cassava flours was similar, and control tortillas presented a DF value up to 100% that obtained in the composites. Fresh tortillas with added UBF had the highest RS content, and a plateau was obtained after 48 h of storage. Tortillas with unripe banana and cassava flours showed higher in vitro hydrolysis rates than control tortillas. Tortillas prepared with UBF displayed the highest SDS content. Tortillas elaborated with the blends UBF-“masa” and cassava flour-“masa” may be an alternative to diversify the use of both food crops due to their chemical and starch digestibility characteristics in maize tortilla.

Acknowledgements

We appreciate the financial support from SIP-IPN, COFAA-IPN and EDI-IPN.