Formulation and processing of gruels made from local ingredients, thin enough to flow by gravity in enteral tube feeding

  • Cybèle Maka TagaEmail author
  • Yvette Jiokap Nono
  • Christèle Icard-Vernière
  • Hélène Desmorieux
  • César Kapseu
  • Claire Mouquet-Rivier
Original Article


Designing enteral foods from local ingredients for tube feeding of low-income people who cannot eat orally is needed. Two processing methods, involving the addition of amylase or malt, were used to thin a blenderized tube feeding formula based on sorghum, sesame and soybean seeds. Two composite flours, either with higher carbohydrate (F1D) or higher lipid (F2D) contents were formulated to obtain an enteral food aimed at adults. To thin the formula enough for it to flow inside the feeding tube, increasing concentrations of amylase (0.27–2.17 g/100 g DM) were added to gruels F1D (F1DE) and F2D (F2DE) prepared at 25% DM. Sorghum malt was also added to F1D (F1DM) as an alternative source of amylase. But F1DE and F1DM flow times in a 50 cm feeding tube (10 Fr) remained much longer (up to 14 s) than that of the commercial enteral food (4 s). The F1DE and F1DM osmolalities (485 and 599 mOsmol/Kg water, respectively) were higher compared to that of F1D but remained within the range specified for adult enteral food. F1D, F1DE and F1DM gruels showed pseudoplastic behavior. Their loss ratio (tan\( \delta \)), elastic (G’) and loss (G’’) moduli were similar, but apparent viscosity, flow time in the feeding tube and consistency index (k) showed that F1DE was thinner than F1DM. Adding an incubation step before cooking of F1DM suspension allowed further thinning of the gruel, showing it is possible to formulate an enteral food using local ingredients that flows by gravity in the feeding tube.


Enteral food Formulation α-amylase Malt Flow properties 


Clinical complications and diseases requiring enteral feeding are becoming increasingly common worldwide (Brown et al. 2015; Schneider 2006). Enteral nutrition is a medical indication for people with problems of swallowing, chewing (in the case of Parkinson’s), psychic (depression due to Alzheimer’s disease), for undernutrition, paralysis or inability to use the oral route, replacing or supplementing the normal oral diet to prevent malnutrition or chronic dehydration using tube feeding (Smith and Garcia 2011; Smith et al. 2006). The increase in health care costs generated by patients requiring long-term enteral feeding is not negligible (Brown et al. 2015), but can advantageously be implemented at home to limit the duration of hospitalization and to reduce costs (Ojo 2015). In Cameroon, some gruels have been designed for enteral tube feeding based on local natural sources, thereby allowing a reduction of nearly 90% of the purchase price of these foods compared to imported nutrient solutions with the same energy density. However, these slurries generally cause flow difficulties in tube feedings like clogging or lumps that limit continuous flow by gravity at room temperature. Consequently, gruels are most often bolus-administered using a syringe to force the gruels to flow in the feeding tube (Boullata et al. 2016; Barkhidarian et al. 2011). However, this delivery system is a time-consuming manual method, since it requires the permanent presence of medical staff. Bolus administration also involves several risks including infection, and filling the stomach at a rapid flow rate of about 30 mL/min, may cause tolerance problems with frequent side effects like diarrhea, nausea, vomiting, distension and regurgitation (Smith and Garcia 2011; Schneider 2006; Smith et al. 2006). These effects may be partially due to the composition of commercial formulas that contain many chemical additives which are tolerated by the body to varying extents, which is one of the reasons why natural foods are encouraged in enteral nutrition (Walia et al. 2017; Bobo 2016; Galán and Drago 2014; Kaur 2009). There are few reported studies likely in the same line. Developing enteral formulas of good nutritional quality using local natural sources is possible, drawing in particular on the model of fortified mixtures based on cereals, legumes and/or oilseeds prepared as complementary food for young children (Ndagire et al. 2015; Kana et al. 2013; Van Hoan et al. 2010). However, due to gelatinization of the starch, these cereal products thicken during cooking. To this end, a commonly used technique is incorporating an alpha-amylase (Kampstra et al. 2018; Trèche and Mouquet-Rivier 2008). The amylase may be an industrial purified amylase, or germinated and dried cereal/malt flour, both of which have their advantages and disadvantages. This technique has been implemented in several studies showing interesting results on young children energy intakes (Kampstra et al. 2018; Van Hoan et al. 2010). Despite the high cost of industrial enzymes, the quantities needed to thin infant gruels are small around 0.01% of dry matter (Mouquet et al. 2006), that the cost of their incorporation into infant meals is manageable. But their supply could be difficult, especially in small towns or in rural areas. Malt, which is much less active, requires much higher incorporation ratio but can be produced locally on a small scale at reasonable prices (Traoré et al. 2005). This technique has been successfully implemented to obtain a highly nutritious food for young children (Ndagire et al. 2015). However, this technique requires good hygiene because the humidity and prolonged ambient temperature conditions during germination are conducive to the multiplication of undesirable microorganisms. Like in the case of food for young children, partial hydrolysis of the starch is required in cereal-based blends used as enteral foods to achieve sufficient energy density satisfying the needs of the patient. But the thinning procedure required obtaining a sufficiently fluid gruel for it to flow into the feeding tube is much more important. The use of malt has already been proposed for an enteral nutrition food based on locally available raw materials (Kaur 2009). The objective of the present work was to formulate an enteral food based on sorghum, sesame and soybean seeds whose general composition resembles that of the products sold for enteral nutrition, and to study its fluidity with an amylase source and malt to achieve satisfactory gravity flow in tube feeding.

Materials and methods


Sorghum (Sorghum bicolor (L.) Moench) from the cultivar Safrari and soybean (Glycine max (L.) Merr.) were both supplied by the IRAD (Agricultural Research Institute for Development) in Maroua (Extreme-North Region, Cameroon). White Sesame (Sesamum indicum L.) came from Tcholliré (Northern Region, Cameroon). Locally available iodized salt was also used in the formulations. The α-amylase BAN (800 KNU/g, pH 6, 72 °C, Novo industries) was used to make gruels more fluid. It is a bacterial endo-amylase (Bacillus subtilis) available as microgranules. A commercial food for exclusive tube feeding enteral nutrition (Sondalis® Standard Fibre, Germany) was used as a reference. It is a dietary food intended for people over the age of 3, isocaloric, normal-protein content, lactose-free, gluten-free and with fiber.


Raw material pre-processing

Sorghum, soybean and sesame grains were sorted and winnowed before the experiments. Sorghum grains were washed with demineralized water twice to remove dust and any stones and dried at 45 ± 1 °C in an oven for 24 h. Some of the cleaned sorghum seeds were dehulled using the Tangential Abrasive Dehulling Device (TADD, Venables Machine Works Ltd, Saskatoon, Saskatchewan, Canada) with an average yield of 89%. The seeds were moistened before dehulling to adjust their water content to 11% to favor elimination of the pericarp while limiting breakage and loss of albumen during the process (Hama et al. 2011). The remaining part of the cleaned sorghum seeds destined for malting was soaked in Ondine™ mineral water to avoid the variability of tap water composition, at 25 ± 1 °C for 16 h experimental trial time at which the seeds reached 36% moisture for germination. Germination was carried out at 30 °C for 4 days. The germinated seeds were dried at 45 ± 1 °C in an oven for 24 h to less than 10% moisture, before the radicles were removed by friction on a sieve. The cleaned soybeans were roasted on a hot plate (Ceran, Schott, Type CK112) at 180 °C for 10 min. The surface temperature of the seeds at the end of roasting was 125 °C. The roasted soybeans were dehulled using a Sataké roll dehuller with a yield of 87%. The sesame seeds underwent no further treatment.

Analytical methods

Dry matter contents were determined by oven drying at 105 °C to constant weight. Ash contents were determined by calcination in a furnace at 530 °C for 3 h. Protein contents (N × 6.25) were determined using the Kjehldahl method. Lipid content was determined after extraction in petroleum ether with the semi-automatic 2055 Soxtec system (Foss, Nanterre, France), approved by AOAC according to the 2003.05 and 2003.06 procedures. Fiber contents (neutral fiber content –NDF) were determined according to the method of Van Soest (1963) using a Dosi-fiber extractor (Fibertec, Foss Tecator, no 1020, France). For the determination of iron, zinc, potassium and calcium, 0.4 g of flour was extracted in a 7:1 nitric acid/hydrogen peroxide mixture using a closed vessel microwave digestion system (Ethos 1 Milestone/Thermo Fisher, Saint-Herblain, France). Mineral content was analyzed by atomic absorption spectrometry (AA 800 Perkin Elmer, Les Ulis, France). Total carbohydrates were estimated by difference. Available carbohydrate was obtained by subtracting fiber content from total carbohydrates. Energy values of formulated cereal-based blends were calculated using the Atwater coefficients of 4 kcal/g for protein and available carbohydrate, 9 kcal/g for lipids (Atwater and Benedict 1902) and 2 kcal/g for fibers (FAO 2003), and are expressed on a DM basis. The energy density of the gruel was then calculated by multiplying the energy value by the gruel concentration and expressed in kcal/100 g of gruel.

Formulation of cereal-based blends for enteral feeding

Sorghum, soybean and sesame seeds were mixed in the appropriate proportions to meet nutritional recommendations for macronutrient contents for adults receiving fiber-enriched enteral foods. These proportions were obtained by formulation using ALICOM V.4, a home-made software that was developed for the purpose of calculating formulations with nutritional specifications and previously used for the formulation of complementary foods for young children (Traoré et al. 2005). This software is a comprehensive linear programming tool that enables to formulate the least expensive mixture that complies with a set of nutritional specifications expressed according to the energy intake from an ingredients file including their nutrient composition. This tool is very similar to the one described by Ryan et al. (2014). The formulation specifications were defined based on a synthesis by Smith and Garcia (2011) and the commercial literatures from Laboratoires Abbott Limitée (2010) and NUTRICIA Advanced medical nutrition (2011). Macronutrient content and the compositions of raw and pre-processed materials were determined in the laboratory. The micronutrient contents of the ingredients used for the formulation that were not analyzed in the laboratory were taken from a food composition table (Souci et al. 2008) in order to estimate the final content in the blenderized foods. For the purpose of comparison, the formulas obtained with either raw materials grains or pretreated grains (dehulled sorghum, malted sorghum, dehulled-roasted soybean) were the same. Two formula flours differing by their lipid contents were selected: one with higher carbohydrate content and less lipid (F1D) and the other with lower carbohydrate content and more lipids (F2D) in order to highlight a lubricating effect of lipids that could increase the flowing in feeding tube. The mixtures of the grains in the proportions determined by the formulations were milled and screened using an ultra-centrifugal mill (Retsch GmbH, type ZM 200, Germany) in order to obtain a homogeneous flour consisting of particles \( \le \) 500 μm. The flours were then stored at 4 °C until they were used for the gruel preparation.

Incubation step

This step was only implemented with flours containing malted sorghum. The malt-containing flour was suspended at 25% (W/V) in demineralized water and put in the oven under precise temperature conditions. These temperature conditions favored expression of the hydrolytic enzymes contained in the malt during varying incubation times. The temperatures tested were 35 °C and 55 °C, at pH 5.8 (pH of the flour + water suspension) during various incubation times ranging from 0 to 120 min.

Gruel preparation

Flour mixtures were suspended with deionized water in a cooking pot. The gruels were cooked in standard heating conditions (Ceran, Schott, Type CK112) at different flour concentrations and their final dry matter content was controlled and adjusted by weighing after cooking by adding demineralized water if necessary. A portion of the gruels was lyophilized for further analysis. According to Trèche and Mouquet-Rivier (2008), α-amylase acts optimally on starch at temperatures between the starch starting gelatinization-swelling temperature (from around 65 °C for sorghum, Emmambux and Taylor 2013) and the amylase deactivation temperature (around 80 °C). A slowly rising temperature thus favors a longer action time by the amylase and more enzymatic hydrolysis. The amylase action time was therefore varied as a function of the temperature of the hot plate. A compromise was finally reached concerning gruel cooking incorporating an amylase source taking into account the time constraints, with a plate temperature of 145 °C, corresponding to a total cooking time of 10 min (5 min of rising temperature and 5 min of boiling). Amylase incorporation ratio in the mixture (on a dry matter basis) ranged from 0.27 to 2.12%; total dehulled sorghum was replaced by malted sorghum in the formula with malt. In order to obtain an isoenergetic product with an energy density ranging between 1.0 and 1.2 kcal/g, the gruel concentration when incorporating the amylase was set at 25 g of DM/100 g.

Rheological measurements

The gruel flow time in a nasogastric silicone feeding tube (ENTRAL™) of 10 Fr (\( \emptyset \) 3.33 mm) was measured as described below. The feeding tube was fixed to the tip of the syringe without its plunger rod and a weight was fixed to its other end to stretch the tube. The system was fixed on a rod stem to ensure gravity flow. For the measurement, 10 mL gruel at ambient temperature (23 ± 1 °C) was poured into the body of the syringe, and the flow time was timed over a distance of 50 cm. The same parameter was determined on the enteral nutrition food, Sondalis® Standard Fibre, used as a reference.

The apparent viscosity of the gruels was also measured at 25 °C in a rotational viscometer (VT550, HAAKE, Champlan, France) with the coaxial cylinders MV-DIN, controlled by a PC with Rheowin software 2.67 (HAAKE laboratories, Karlsruhe, Germany). About 40 mL of sample were placed in the stationary cup. The temperature was controlled by water circulating through the jacket surrounding the cup assembly. The shear rate (\( \dot{\gamma } \)) was set at 10 s−1 and the apparent viscosity values were recorded after 1 min of shear. The corresponding shear stress (τ) was recorded and data were fitted with the Oswald–de-Waele or power law model (Steffe 1996):
$$ \varvec{\tau}= \varvec{k} \cdot \varvec{ }{\dot{\mathbf{\gamma }}}^{{\mathbf{n}}} $$
where \( {\tau} \) is shear stress (Pa), \( {\dot{\gamma }} \)  is shear rate (s−1), n is the flow behavior index, k is the consistency index (Pa sn). When n is less than, more than, or equal to 1, the fluid has a shear thinning, dilatant or Newtonian behavior, respectively. The coefficient k is linked to the viscosity and expresses the consistency of the fluid.

The G’ and G’’ moduli were recorded using an oscillatory Rheometer (MCR 301 Anton Paar Physica) with the CP50 cone plate geometry controlled by a PC with Rheoplus/32 V3.40 software. Measurements were performed at a strain sweep ranging from 0.01 to 10,000% and a controlled temperature of 25 °C. Storage modulus G’ measures the elasticity of food, while the loss modulus G’’ measures viscosity. The phase shift δ = tan−1 (G’’/G’) indicates the dominant characteristic: if δ = 0 (tan δ = 0), the material is solid, if δ = 90 (tan δ = ∞), the material is liquid, or somewhere in between for viscoelastic materials. These parameters were also measured on the reference food Sondalis-fibre™.

Osmolality measurement

Osmolality represents the concentration of a solution expressed in osmoles or moles of osmotically active particles per kilogram of water. It was measured using an automatic osmometer (Hermann Roebling MESSTECHNIIK Type 13, D-14129, Berlin), whose principle is to measure the freezing point depression of an aqueous solution in comparison with pure water. One ml of supernatant was obtained by centrifugation of the gruel sample at 20,000 g for 20 min at 4 °C, and introduced in a microtube, fixed on the osmometer and the measurement was made.

Statistical methods

Statistical analyses were performed with the Statgraphics Centurion XVI software version 16.1.18 using analysis of variance (ANOVA) followed, when appropriate by Fischer’s least significant difference tests to compare means at the 5% significance level. A generalized linear model was also used to assess the effect of the variables (temperature, 35 °C or 55 °C; and time 0, 15, 30, 60 or 120 min) of the incubation of flour suspension before cooking on the flow time in the enteral feeding tube.

The coefficient of determination R2 of the rheological analysis was determined using a commercial software [Sigmaplot© version 12.5 (wpcubed, GmbH, Germany)]. All measurements were carried out at least in duplicate.

Results and discussion

  1. a.


Sorghum, soybean and sesame seeds were used to formulate a local enteral food. The targeted fiber-rich enteral food formulation, with fiber content between 2 and 6 g/100 g DM (Smith and Garcia 2011; Laboratoires Abbott Limitée 2010; NUTRICIA Advanced medical nutrition 2011), was not possible to be achieved with the raw materials used, due to their high fiber content (8–9 g/100 g DM; Table 1). Additionally, in an aqueous medium these fibers swell, which may slow down the flow in the probe (Bobo 2016; Schneider 2006, Kaur 2009). To decrease the fiber content of the mixtures, dehulled sorghum and soybean seeds were used. Eliminating the bran reduced the sorghum and soybean fiber contents by 1/4 and 1/2 respectively (Table 1) but led to significant changes in the other nutrient contents. For sorghum, dehulling was followed by a significant decrease in lipid content, attributed to a partial loss of seed germs rich in lipids during the process (Hama et al. 2011). This processing step also resulted in a 61% and a 35% decrease in iron content in dehulled sorghum and roasted-dehulled soybeans, respectively. Two enteral foods were formulated using Alicom software from dehulled sorghum, roasted-dehulled soybeans and sesame seeds, taking their nutritional composition (Table 1) into account and according to specifications for fiber-enriched isoenergetic food (Table 2). The aim of adding various concentrations of sesame or soybean was to obtain a lubricating effect on the flow due to varying lipid contents. The addition of salt, widely available and accessible in Cameroon, was taken into account to supplement sodium deficiency which is very important in maintaining the electrical and chemical equilibrium of the cells for their proper functioning and that of the nervous system. The two formulations selected (Table 2) differed essentially in their higher lipid (F2D) or higher available carbohydrate (F1D) contents. The macronutrient energy contributions of the different formulations were comparable to those of the reference food Sondalis or of other foods reported in the literature (Kaur 2009). The nutritional value of the two formulas, calculated using Alicom software (Table 2), showed that they met the recommended minimum values for lipids, linoleic acid, digestible protein, Na, K, P, Cu, Mg and Mn, as well as for biotin, folic acid and vitamin K1. Other nutrient contents (linolenic acid, Ca, Zn, Se, vitamin A, ascorbic acid, thiamine, riboflavin, nicotinamide, pantothenic acid and vitamin E) were below the recommended minimum values. For Fe, only the F2D formula met the recommendations thanks to the high level of sesame incorporated and its high iron content (Table 1). However, the high iron content could be due to extrinsic iron contamination from dust or soil, whose bioavailability is low. In order to meet all recommendations for micronutrient levels, it would be necessary to add a specifically formulated premix, as it is the case in commercial enteral foods. Due to its higher lipid content, the F2D formula energy value (4.6 kcal/100 g DM) was slightly higher than that of F1D (4.2 kcal/100 g DM). These two values were close and were within the required range (4–4.8 kcal/100 g DM).
Table 1

Proximate and mineral composition of raw and pre-processed ingredients










Dehulled1 and roasted


Dry matter content (%)

89.67 ± 0.11

89.44 ± 0.09

92.00 ± 0.56

94.23 ± 0.01

94.67 ± 0.11

94.82 ± 0.07

Macronutrients (g/100 g DM)


7.95 ± 0.16a

9.98 ± 0.08b

6.80 ± 0.05a

39.76 ± 0.19a

55.01 ± 3.02b

19.40 ± 0.10


2.71 ± 0.08c

1.80 ± 0.11b

1.32 ± 0.11a

18.27 ± 0.47a

20.12 ± 1.78a

54.25 ± 0.59

 NDF2 fibers

8.20 ± 0.56b

6.40 ± 0.01a

9.18 ± 0.34b

8.85 ± 0.13b

4.47 ± 0.08a

9.08 ± 0.29

 Available carbohydrates







 Ash (% DM)

1.22 ± 0.07b

1.33 ± 0.00b

0.98 ± 0.01a

4.61 ± 0.10a

5.22 ± 0.01b

5.31 ± 0.04

Some minerals (mg/100 g DM)


6.0 ± 0.2c

2.3 ± 0.1a

3.2 ± 0.2b

7.6 ± 0.1b

5.0 ± 0.1a

23.4 ± 0.5


1.2 ± 0.0ab

1.4 ± 0.1b

1.0 ± 0.0a

2.6 ± 0.0a

2.9 ± 0.0b

4.4 ± 0.2


9.1 ± 0.6a

8.1 ± 0.2a

55.3 ± 0.7b

198.5 ± 2.7b

186.7 ± 1.5a

586.3 ± 15.4


397 ± 12c

330 ± 7b

76 ± 2a

1644 ± 10a

1628 ± 24a

458 ± 16

 Energy value (kcal/g DM)







Results are mean ± SD of measurements made at least in triplicates. Values in a same line for a given raw material with different superscript letters are significantly different (p < 0.05)

1Dehulling yield = 89% for sorghum and 87% for soybean; 2Neutral Detergent Fibers

Table 2

Nutritional value of blends based on pre-processed ingredients and formulated with ALICOM compared to other enteral foods



Formulation F1D2

Formulation F2D3

Reference food4

Medical food (Kaur 2009)

Energy value (Kcal/100 g DM)






Per 100 kcal

Maximum values

Minimum values

Estimated content

% of max or min values

Estimated content

% of max or min values


Available carbohydrates (g)









Total proteins (g)









Digestible protein (g)








Fibers (g)









Lipids (g)









Linolenic acid (mg)









Linoleic acid (mg)








Sodium (mg)









Sodium (mg)









Potassium (mg)









Potassium (mg)









Calcium (mg)








61.0 + 2.47

Phosphorus (mg)








58.6 + 4.67

Iron (mg)









Copper (µg)









Magnesium (mg)









Zinc (mg)


1. 2







Manganese (µg)









Selenium (µg)









Vitamin A (µg RE)



4.3 6


3.4 6



150 7

Vitamin A (µg RE)









Ascorbic acid (mg)









Thiamin (µg)









Riboflavin (µg)









Nicotinamide (µg)









Vitamin B12 (µg)









Folic acid (µg)









Pantothenic acid (µg)









Vitamin E (mg eq α-tocopherol)









Vitamin K1 (µg)









Biotin (µg)









1For enteral foods, according to Smith and Garcia (2011), Laboratoires Abbott Limitée (2010), NUTRICIA Advanced medical nutrition (2011). 2F1D: Dehulled Sorghum (69.2%); Roasted dehulled soybean (22.1%); Raw sesame (7.7%); Iodized salt (1.0). 3F2D: Dehulled Sorghum (56.4%); Roasted dehulled soybean (18.7%); Raw sesame (23.9%); Iodized salt (1.0). 4Sondalis®: Standard Fiber: dietetic food for exclusive enteral tube feeding for special medical purposes for people over three years. Complete, isoenergetic, normalized-protein, lactose-free and gluten-free food containing fibers. 5Reevaluated including calories available in the fibers. 6Estimated using data from Souci et al. (2008). 7Micronutrient contents adjusted by addition of the premix (Kaur 2009). In bold: maximum values

  1. b.

    Rheological behavior of enteral formulas


Effect of gruel dry matter content on apparent viscosity and flow time in the feeding tube

The flow time in the feeding tube (Fig. 1a) and the apparent viscosity of the gruels (Fig. 1b) prepared from formulas F1D and F2D showed a similar evolution and increase, proportionally to the concentration of the gruel (Fig. 1). This increase was initially low at concentrations ranging between 2.5 and 6 g DM/100 g, values at which the F1D and F2D gruels flowed rapidly into the feeding tube, with a flow time of less than 20 s/50 cm. At concentrations ranging between 6 and 8 g DM/100 g, the increase in the flow time accelerated sharply. At very low concentrations, the behavior was quasi-linear, corresponding to the Newtonian domain linked to very dilute systems. But when the concentration became sufficient to allow intermolecular interactions, the Newtonian domain would be left and apparent viscosity increased strongly. In F1D, this occurred at lower concentrations because of its higher starch content due to the higher proportion of sorghum in its composition. Several authors observed a rapid increase in the viscosity proportional to the gruel concentration, with a similar evolution in both infant nutrition (Ndagire et al. 2015, Trèche and Mouquet-Rivier 2008) and enteral nutrition (Kaur 2009). Thickening of the F1D and F2D gruels was mainly due to their high starch content, which gelatinizes and swells during cooking. At concentrations greater than 5 g DM/100 g, the flow times and apparent viscosities of F1D and F2D gruels were greater than the values obtained for the reference enteral food (4.0 s/50 cm and 16.1 mPa s, respectively). At a concentration of 5 g DM/100 g, the gruel energy densities prepared from the formulas F1D and F2D were 21 and 23 kcal/100 g respectively. These values were well below the target value of 100 kcal/100 g (isoenergetic food), which corresponds to DM contents of around 25 g/100 g to provide enough energy to patients within a reasonable time. At concentrations above 9 g DM/100 g, the 10 Fr feeding tube was observed to be obstructed and the gruels no longer flowed. It was therefore essential to implement processes such as incorporating amylase to make the gruels more fluid with adequate energy density.
Fig. 1

Effect of the concentration of the gruel in the two formulations F1D and F2D on the flow time in a 10 Fr feeding tube (a) and apparent viscosity at 10 s−1 (b). The dashed line represents the flow time and apparent viscosity of the reference enteral food. Results are the means of two measurements with error bars for average deviation

Effect of amylase (BAN) concentration on apparent viscosity and flow time in the feeding tube

The BAN amylase was incorporated in varying proportions in gruels prepared at a concentration of 25% DM (F1DE and F2DE). The resulting flow curves (Fig. 2) revealed different behaviors. At amylase concentrations of less than 0.5 g/100 g DM, the F1DE gruel flow time was longer than that of F2DE. When the amylase concentration exceeded 0.5 g/100 g DM, the trend was reversed. The F1DE mixture, which contained more starch, was more sensitive to thinning by amylase. Beyond 1% amylase, the thinning appeared to reach a maximum, and the flow times stabilized at values of 15 and 28 s/50 cm in F1DE and F2DE, respectively. However, these values were still considerably higher than those of the reference food. Indeed, the incorporated α-amylase only acted on gelatinized starch during cooking, whereas other compounds including fibers, lipids and proteins may also play a role in the consistency of these gruels. When the gelatinized starch was hydrolyzed enough to play a minor role in the gruel consistency, the increase in the α-amylase concentration no longer led to a significant decrease in the flow time. Compared to the 0.01% α-amylase concentration used in complementary food for young children (Mouquet et al. 2006), the concentrations used here were much higher because the enteral foods must be much more fluid than the infant gruels. These results show that starch hydrolysis alone was not sufficient to obtain satisfactory fluidification. Moreover, supplying commercial enzyme in remote areas of Cameroon could be difficult to organize. In this case, replacing the commercial amylase by malted sorghum could be an interesting alternative as it could enable the production of the composite flour for tube feeding by local small or micro-enterprises, or even at household level.
Fig. 2

Effect of the concentration of amylase in the F1DE and F2DE gruels on the flow time in the 10 Fr feeding tube. The gruels were prepared at 25% DM. The dashed line represents the flow time of the reference enteral food. Results are the means of two measurements with error bars for average deviation

Effect of incorporating malted sorghum on the flow properties of blended food gruels

Usually, the incorporation of 5–20% malt, is used in infant flours to increase the gruel energy density while maintaining an acceptable consistency (Trèche and Mouquet 2008). To achieve maximum thinning of the enteral food, the dehulled sorghum was replaced by malted sorghum. Like for dehulling, the content of some nutrients changed during malting (Table 1). Indeed, during the germination process, hydrolytic enzymes are activated in the sorghum seeds to enable the growth of germs and rootlets, which are removed after drying. The germs are rich in proteins and lipids and their removal lead to a decrease in protein and lipid contents, and consequently to a passive increase in total carbohydrates, including available carbohydrate and fibres. Similar results have been reported in sorghum by Traoré et al. (2004). After entering data on the nutrient content of the malted sorghum in the Alicom software, it was possible to check the suitability of the two formulas F1DM and F2DM. With respect to target objectives, only the fibers were in excess: fiber content was estimated to be 1.95 and 1.83 g/100 kcal in F1DM and F2DM, respectively, while the maximum recommended value is 1.5 g/100 kcal. This excess may have adverse effects on patients even at low levels, including compromising energy intake, reducing macronutrient uptake, increasing fecal energy loss and reducing the bioavailability of micronutrients (Tarleton et al. 2013; Green 2001). However, the fiber content in foods should preferably be measured after preparation in the gruel form, since fiber content could decrease due to the effect of the malt enzymes. The energy intakes from macronutrients in the malted F1DM (4.20 kcal/g DM) and F2DM (4.58 kcal/g DM) formulas were identical to those of the formulas F1D and F2D, respectively (Table 2). The F1D formula, which is more sensitive to enzyme activities, was thus chosen for further experiments.

The dynamic rheological properties of the gruels prepared at 25% DM from F1D (without amylase), F1DM and F1DE, were determined and compared with those of the reference food (Table 3). The reference food had much lower modulus than the gruels on a deformation sweep beyond the threshold value, with also a δ tangent ranging from 1 to 1000, which was much higher than that of gruels, indicating more fluid behavior (Table 3). In fact, the most fluid product was that which had the highest δ tangent (ratio G’’: G’). The G’ values were higher than those of G” up to a certain deformation threshold when the tendency was reversed (crossover). The deformation yield of between 19% and 43% for the 4 products was the deformation value at which modules G ‘and G’’ were equal (tanδ = 1), and that is necessary to make the molecules move and cause the flow. Below these threshold values, the products we studied behaved like solid gels. Beyond these threshold values, the product was predominantly viscous and flow (Brummer 2006; Steffe 1996).
Table 3

Dynamic moduli, flow parameters and osmolality of the different enteral foods studied


Viscoelastic moduli (from strain yield to 10,000%)

Power law model parameters


G’ (Pa)

G’’ (Pa)

Tan \( \delta \)


k (mPa sn)


Apparent viscosity (mPa. s) at 10 s−1

Flow time in a 10 Fr feeding tube (s/50 cm)

Osmolality (mOsmol/Kg water)

Reference food



































All measurements were done in duplicate and results are the means

F1D: 69.21% dehulled sorghum, 22.13% roasted-dehulled soybean, 7.69% sesame, 0.97% salt; F1DE: F1D* + 0.5% amylase BAN; F1DM: F1D with dehulled sorghum replaced by malted sorghum; n: flow behavior index; k: consistency index

*Gruels prepared with 25 g DM/100 g

Thus, even when dehulled sorghum was replaced by malted sorghum, the resulting gruels were not sufficiently fluid. The F1D gruels had higher G ‘and G’’ moduli and lower δ tangent (1–18) values than F1DE and F1DM, indicating a much thicker consistency. The F1DE and F1DM gruels had similar dynamic moduli and δ tangent (Table 3) indicating similar viscoelastic behaviors. However, the apparent viscosity and flow time values presented in Table 3 show that the F1DE gruels were somewhat more fluid than F1DM gruels. The consistency index calculated using the power law was also lower for F1DE. The flow index behaviors of the two types of gruels were much lower than 1, indicating their pseudoplastic or thinning character, which appeared to be more marked for gruel with malt. The reference food flow index behavior was equal to 0.98: this value, which is very close to 1, corresponds to that of a Newtonian fluid. Its consistency index was very low. Although gruels obtained with amylase enzyme appeared to be more fluid, it should be noted that the enzyme incorporation ratio of 0.5%, involving industrial amylase, was high and could have a significant impact on the cost of the final product.

For an enteral formula, osmolality is an important factor in tolerance, as it should be close to that of the body fluid (between 275 and 300 mOsmol/Kg water) to be well tolerated by the patients and not cause adverse effects such as cramps, nausea, vomiting, and diarrhea. The results obtained for F1DE and F1DM were in the range of 270–600 mOsmol/Kg water, corresponding to the normal range of enteral values for adults (Smith and Garcia 2011).

Effect of pre-incubation of blends with malt on gruel flow properties

Many enzymes including endo-beta-glucanase, pentosanase and protease are activated during germination, in addition to amylase. Under favorable environmental conditions, they may enable partial hydrolysis of other seed-storage substances that play a role in the consistency of the gruels such as fibers or proteins (Dziedzoave et al. 2010), and thus help thinning them. In the final experiments, immediate cooking of the gruels after flour suspension in the water led to a rapid increase in the temperature, above 60 °C. Above this temperature, enzymes such as proteases (Jones 2005; Gebremariam et al. 2013) were mostly inactivated. We then tested incubating the flour suspension before cooking at temperatures below 60 °C and confirmed that this pretreatment allowed more thorough gruel fluidification (Fig. 3). Fluidification was significantly greater at 55 °C (p = 0.0004) than at 35 °C and increased with incubation time (between 0 and 120 min) (p < 0.0001). However, the measured flow times in the feeding tube were very variable. This could be attributed to the simultaneous action of several enzymes, as well as to the variation in the composition of the medium following enzymatic activities (Barros et al. 2010; Jones 2005). A more detailed study is thus needed to identify the causes of the observed fluidification and the optimum temperature, incubation time and pH to obtain maximum fluidification, while remaining applicable under real conditions.
Fig. 3

Flow time in the 10 Fr feeding tube (s/50 cm) of the different gruels prepared at 25% DM with BAN amylase (0.5%) (F1DE) or malt incorporated in the blended flour (F1DM), according to different temperature–time incubation conditions (pH 5.8) compared to the reference food. Significant differences are shown by different letters (p < 0.05). Measurements were done in triplicate, results are the means with error bars for standard deviation


A blenderized food based on sorghum, soybean and sesame seeds that meets main recommendations for macronutrients to be used, after preparation as a gruel, for tube feeding of adults was formulated. The rheo-fluidification processing conditions of gruels, prepared at the required energy density of 100 kcal/100 g (i.e. 25% DM), that allow them to flow by gravity in the enteral feeding tube, required industrial amylase and sorghum malt addition. Both methods efficiently thinned the gruels and obtained the required flow, but the flow time in the feeding tube was longer than that of the commercial enteral food used as reference. An incubation step of the blends containing malt, (F1DM) was shown to be a promising approach to obtain further gruel thinning. We hypothesized that this could be due to the action of enzymes other than α-amylase in the malt, enzymes that could partially hydrolyze fibers and/or proteins during the incubation step. The results obtained can be directly applied in practice in a first step and under appropriate conditions. The enteral food obtained using local ingredients, much cheaper than current commercial enteral foods, could be sold as a composite flour and the preparation recommendations provided on the packaging. Such process is easy and could be run in remote areas. However, further studies are needed to identify more precisely the effect of the incubation step and the conditions that will allow maximal gruel thinning to better meet practical applications.



We acknowledged the KITE (Knowledge Integration Transparency and Education) scholarship funded by European commission-Erasmus Mundus.


  1. Atwater WO, Benedict FG (1902) Experiments on the metabolism of matter and energy in human body. US Office of Experiment Stations Bulletin No. 109, Government Printing Office, Washington, DC, pp 1898–1900Google Scholar
  2. Barkhidarian B, Seyedhamzeh S, Mousavi N, Norouzy A, Safarian M (2011) PP183-SUN nutritional and physical quality of blenderized enteral diets. Poster presentations—nutritional techniques and formulations I. Clin Nutr Suppl 6(1):92–93CrossRefGoogle Scholar
  3. Barros M, Fleuri LF, Macedo GA (2010) Seed lipases: sources, applications and properties—a review. Braz J Chem Eng 27(01):15–29CrossRefGoogle Scholar
  4. Bobo E (2016) Reemergence of blenderized tube feedings: exploring the evidence. Nutr Clin Pract 31(6):730–735CrossRefGoogle Scholar
  5. Boullata JI, Carrera AL, Harvey L, Escuro AA, Hudson L, Mays A, McGinnis C, Wessel JJ, Bajpai S, Beebe ML, Kinn TJ, Klang MG, Lord L, Martin K, Pompeii-Wolfe C, Sullivan J, Wood A, Malone A, Guenter P (2016) ASPEN safe practices for enteral nutrition therapy, consensus recommendation. J Parenter Enter Nutr 41:15–103CrossRefGoogle Scholar
  6. Brown B, Roehl K, Betz M (2015) Enteral nutrition formula selection: current evidence and implications for practice. Nutr Clin Pract 30(1):72–85CrossRefGoogle Scholar
  7. Brummer R (2006) Rheology essentials of cosmetic and food emulsions. Springer, BerlinGoogle Scholar
  8. Dziedzoave NT, Graffham AJ, Westby A, Komlaga G (2010) Comparative assessment of amylolytic and cellulolytic enzyme activity of malts prepared from tropical cereals. Food Control 21:1349–1353CrossRefGoogle Scholar
  9. Emmambux MN, Taylor JRN (2013) Morphology, physical, chemical, and functional properties of starches from cereals, legumes, and tubers cultivated in Africa: a review. Starch - Stärke 65:715–729CrossRefGoogle Scholar
  10. FAO (2003) Food energy – methods of analysis and conversion factors. FAO Food Nutr Pap 77: ISSN 0254-4725Google Scholar
  11. Galán MG, Drago SR (2014) Food matrix and cooking process affect mineral bioaccessibility of enteral nutrition formulas. J Sci Food Agric 94(3):515–521CrossRefGoogle Scholar
  12. Gebremariam MM, Zarnkow M, Becker T (2013) Thermal stability of starch degrading enzymes of teff (Eragrostis tef) malt during isothermal mashing. Process Biochem 48:1928–1932CrossRefGoogle Scholar
  13. Green CJ (2001) Fibre in enteral nutrition. Clin Nutr 20(1):23–39CrossRefGoogle Scholar
  14. Hama F, Icard-Vernière C, Guyot J-P, Picq C, Diawara B, Mouquet-Rivier C (2011) Changes in micro- and macronutrient composition of pearl millet and white sorghum during in field versus laboratory decortication. J Cereal Sci 54:425–433CrossRefGoogle Scholar
  15. Jones BL (2005) Endoproteases of barley and malt. J Cereal Sci 42:139–156CrossRefGoogle Scholar
  16. Kampstra NA, Nguyen VH, Koenders DJPC, Schoop R, Broersen B, Mouquet-Rivier C, Traoré T, Bruins MJ, de Pee S (2018) Energy and nutrient intake increased by 47–67% when amylase was added to fortified blended foods—a study among 12–35 months old Burkinabe children. Matern Child Nutr 14(1):e12459. CrossRefGoogle Scholar
  17. Kana SMM, Gouado I, Mananga MJ, Ekoule LD, Amvam ZPH, Tetanye E (2013) Evaluation of nutritional status of young children aged 0–2 years in the Douala city (Cameroon), survey of some practices during diversification of complementary foods. Afr J Food Sci Technol 4(2):29–34Google Scholar
  18. Kaur M (2009) Medical foods from natural sources. Springer, New YorkCrossRefGoogle Scholar
  19. Laboratoires Abbott Limitée (2010) Guide de sélection des produits nutritionnels pour adultes et certains produits nutritionnels pour enfants (JEVITY, OSMOLITE). Imprimé au Canada, DIR/161F08-novembre 2010 – 00094,
  20. Mouquet C, Greffeuille V, Treche S (2006) Characterization of the consistency of gruels consumed by infants in developing countries: assessment of the Bostwick consistometer and comparison with viscosity measurements and sensory perception. Int J Food Sci Nutr 57(7):459–469CrossRefGoogle Scholar
  21. Ndagire CT, Muyonga JH, Manju R, Nakimbugwe D (2015) Optimized formulation and processing protocol for a supplementary bean-based composite flour. Food Sci Nutr 3:527–538CrossRefGoogle Scholar
  22. NUTRICIA Advanced medical nutrition (2011) Fiches techniques Alimentation par sonde-Nutrison.
  23. Ojo O (2015) Review—the challenges of home enteral tube feeding: a global perspective. Nutr 7:2524–2538Google Scholar
  24. Ryan KN, Adams KP, Vosti SA, Ordiz MI, Cimo ED, Manary MJ (2014) A comprehensive linear programming tool to optimize formulations of ready-to-use therapeutic foods: an application to Ethiopia. Am J Clin Nutr 100:1551–1558CrossRefGoogle Scholar
  25. Schneider SM (2006) Nutrition entérale: quelle est sa place dans notre arsenal thérapeutique? Gastroenterol Clin Biol 30:988–997CrossRefGoogle Scholar
  26. Smith L, Garcia J (2011) Chapter 89—enteral nutrition. In: Hyams RWS (ed) Pediatric gastrointestinal and liver disease, 4th edn. W.B. Saunders, Saint Louis, pp 978–1001CrossRefGoogle Scholar
  27. Smith L, Casey L, Kennedy-Jones M (2006) Chapter 75—enteral nutrition. In: Hyams RWS, Kay M (eds) Pediatric gastrointestinal and liver disease, 3rd edn. W.B. Saunders, London, pp 1141–1164CrossRefGoogle Scholar
  28. Souci SW, Fachmann W, Kraut H (2008) La composition des aliments, 7th edn. MedPharm Scientific Publishers, StuttgartGoogle Scholar
  29. Steffe JF (1996) Rheological methods in food process engineering, 2nd edn. Freeman Press, East LansingGoogle Scholar
  30. Tarleton SM, Kraft CA, DiBaise JK (2013) Fiber-enriched enteral formulae: advantageous or adding fuel to the fire? Nutrition issues in gastroenterology. Pract Gastroenterol 124:11–22Google Scholar
  31. Traoré T, Mouquet C, Icard-Vernière C, Traoré AS, Trèche S (2004) Changes in nutrient composition, phytate and cyanide contents and [alpha]-amylase activity during cereal malting in small production units in Ouagadougou (Burkina Faso). Food Chem 88(1):105–114CrossRefGoogle Scholar
  32. Traoré T, Vieu MC, Alfred TS, Serge T (2005) Effects of the duration of the habituation period on energy intakes from low and high energy density gruels by Burkinabe infants living in free conditions. Appetite 45:279–286CrossRefGoogle Scholar
  33. Trèche S, Mouquet-Rivier C (2008) Use of amylases in infant food. In: Porta R, Di Pierro P, Mariniello L (eds) Recent research developments in food biotechnology. Enzymes as Additives or Processing Aids, Tirvandrum, pp 231–245Google Scholar
  34. Van Hoan N, Mouquet-Rivier C, Eymard-Duvernay S, Treche S (2010) Effect of extrusion cooking and amylase addition to gruels to increase energy density and nutrient intakes by Vietnamese infants. Asia Pac J Clin Nutr 19:308–315Google Scholar
  35. Van Soest PJ (1963) Use of detergents in the analysis of fibrous feed II—a rapid method for the determination of fiber and lignin. J Assoc Off Anal Chem 46:829–853Google Scholar
  36. Walia C, Van Hoorn M, Edlbeck A, Feuling MB (2017) The registered dietitian nutritionist’s guide to homemade tube feeding. J Acad Nutr Diet 117(1):11–16CrossRefGoogle Scholar

Copyright information

© Association of Food Scientists & Technologists (India) 2019

Authors and Affiliations

  1. 1.Department of Process EngineeringUniversity of NgaoundereNgaoundereCameroon
  2. 2.UMR NutripassIRD, Univ. MontpellierMontpellierFrance
  3. 3.UMR LAGEPPUniv Lyon 1VilleurbanneFrance

Personalised recommendations