Milling interventions for the production of atta for Indian flat bread with low carbohydrate digestibility

Abstract

Study was aimed to produce atta for chapati, an Indian flat bread with low carbohydrate digestibility through different milling interventions; processing and formulating a functional ingredient mix (FM). Granulation, physico-chemical, rheological and chapati making characteristics of chakki atta, CA (control), roller mill atta (RA); RA replaced with 5, 10 and 15% FM (5, 10 and 15% RAFM) were evaluated. RA and RAFM samples showed lower water absorption, higher dough stability, pasting temperature and peak viscosity than CA. Evaluation of carbohydrate digestive profile showed differences in the pattern of carbohydrate digestibility and glucose release between the chapatis prepared from CA, RA and 10% RAFM. Rapidly available glucose (RAG), an indicator of glycemic response in vivo, was found to be lower in the 10% RAFM than CA. It can be concluded that milling interventions and compositional differences together determine the carbohydrate digestibility of the atta.

Introduction

Comminution or ‘size reduction' is a unit operation in which the average size of the solid food is reduced. This improves the palatability of food and enhances its suitability in a wide range of products. The grinding equipments mostly used for whole wheat flour production are chakki or stone mill, roller mill, pin and hammer mill or pulverizer. Stone mill uses the force of compression, abrasion and shear; roller mill impart compression and shear in various degrees in fluted and polished roll system. The different type of forces imparted on the material to be ground by the grinding equipment influences the particle size, granulation range, damaged starch, bran size and colour of the flour (Loncin and Merson 1979). These differences will impact on the techno-functional properties and the final product quality (Inamdar et al. 2015). Finely milled flours generally have a high glycemic index (GI) while coarse stone-ground flours have larger-sized particles and lower GI (Brand-Miller 2003). This has been corroborated by Heaton et al. (1988) who found that there was a gradual stepwise increase in glucose and plasma insulin responses in subjects consuming whole or cracked grains or wholegrain flour.

In addition to particle size, food structure is found to have a strong influence on postprandial blood glucose and insulin response. Any process that disrupts the integrity of the food structure will adversely reflect on the plasma glucose and insulin responses. Whole grains tend to slow down the glycaemic response owing to its fibre content while processing tends to increase it (Jenkins et al. 1988). Marangoni and Poli (2008), has demonstrated that use of mixture of fibers reduced markedly the GI of bread and biscuits. Incorporation of psyllium into food products is more effective at reducing blood glucose response than the usage of soluble fiber supplement that is separate from the food (Hanssen et al. 1992). Proteins were also reported to modulate the carbohydrate digestibility of foods, Bengal gram being a rich source of protein and low in carbohydrate unlike the cereals has shown reducing post-feast climb in glucose in people with diabetes (Srivastava and Vasishtha 2012).

Owing to the health scenario of our population with increasing chronic degenerative diseases such as diabetes and obesity, the need for exploration of the food formulations or modifying the staple foods for sustained glucose release has increased. Since whole wheat flour termed as atta is a staple for the majority of the population and most of the atta is consumed in the form of chapati, the Indian flat bread it was felt essential to identify the process that has a lower carbohydrate digestibility. Hence the investigation was undertaken to study the role of milling interventions for production of atta, formulation of functional ingredients mix (FM) using selected non-wheat grains and partial replacement of roller flour mill atta with FM (RAFM) and to study the physico-chemical, rheological, carbohydrate digestive profile and chapati making characteristics of differently produced atta.

Materials and methods

Materials

Commercial medium hard wheat (Lok 1 variety), chickpea (Cicer arietinum), barley seed (Hordeum vulgare), defatted soya flour, oats (Avina Sativa), fenugreek fiber (Trigonellafoenum-graecum), psyllium husk (Plantago ovata) and salt (sodium chloride) were purchased from the domestic market.

The glucose oxidase enzyme, Aspergillus niger (128,200 units/g solid, Cat.no G7141-50KU, Lot SLBN8314V), amyloglucosidase (source Aspergillus niger, ≥ 300U/ml, Cat.no A7095, Lot SLBJ2506V), α-amylase (source Bacillus amyloliquefaciens (≥ 250 units/g, Cat.no A7595, Lot SLBK3763V), pancreatin (source porcine pancreas (8 × USP, Cat.no P7545, Lot SLBH7449V), pepsin (source porcine gastric mucosa (≥ 250 units/mg, Cat.no P7000, Lot BCBF9832V), and peroxidase (source horseradish (163 units/mg, Cat.no P8250, Lot SLBQ1119V), invertase (source bakers yeast, 355 units/mg solid, Cat.no I4504, Lot 111K7480), pepsin (sourcs gastric mucosa (1000 NF U/mg, Cat.no P7000, Lot BCBF9832V) were purchased from Sigma Aldrich Ltd (St. Louis, MO, USA). Analytical grade chemicals, organic solvents and triple distilled water were used for all the study.

Methods

Wheat milling

Whole wheat flour was produced by the application of two different techniques; stone mill and roller milling.

Stone milling

Cleaned wheat in a batch of 2 kgs was passed through stone mill (Make Indica, Mixer Jumbo model) having a 1.5 HP motor and stone diameter of 175 mm. The rpm of the stone chakki was 960 and the flowrate was 2 kg per hour. The stone gap was maintained at 0.3 mm. The flour obtained was coded as chakki atta (CA).

Roller milling

The recombination process by remixing the bran and all the flour streams to obtain whole wheat flour was applied using laboratory roller mill (Buhler, MLU 202). A batch of 2 kg of cleaned wheat without conditioning was passed through the mill having three breaks and three reduction system. The flour samples from all the break and reduction passages were collected together. The coarse and fine bran from the last break and reduction passages respectively were passed through the bran duster to retain the fine endosperm and was mixed with the flour. The dusted bran was ground in the hammer mill (Natraj, Premium model, 1 Hp motor) with 0.75 mm sieve opening. The ground bran was mixed with the flour fractions in a mixer for 10 min to obtain roller milled whole wheat flour coded as Roller flour mill atta (RA).

Functional ingredients milling

Chick pea, barley, oat, and psyllium husk were ground separately in the hammer mill using the fine sieve. Defatted soya flour was purchased from local market in Mysore and directly used for the study. Fenugreek fibre flour was produced as per the method described by Sakhare and Prabhasankar (2017) using roller flour mill.

Preparation of blends

The ground flours from wheat namely CA and RA where the base atta. To improve the nutritional quality of chapati from RA, functional ingredients mix (FM) was prepared by combining chickpea flour, barley flour, defatted soya flour, oat flour, fenugreek fiber and psyllium husk in the proportion of 30:30:20:15:2.5:2.5 respectively. The RA was replaced using the above FM at the level of 5, 10 and 15%. These were labelled as 5, 10 and 15% functional ingredients mixed roller flour mill atta (RAFM) respectively. Selection and fixing of functional ingredients level were based on the preliminary trials.

Particle size index

The percentage proportion of particles in the different size classes (< 100 µm, < 250 µm and < 500 µm) present in different atta samples were measured by the Laser 132 Particle Size Analyzer (S3500, Microtrac Inc., USA), that utilises three red laser diodes placed precisely to analyse particles accurately. The data were analysed using the software Flex in Microtrac S3500.

Chemical and rheological characteristic

Determination of moisture, ash, protein, SDS—sedimentation value, Hagberg’s falling number, damaged starch, farinograph, extensograph, amylograph was according to AACC methods (AACC 2010) and total dietary fibre (AOAC 2010).

Chapati making characteristics

The chapati was prepared using the following formulation: CA (100 g)/RA (100 g)/FMRA (95:5; 90:10; 85:15) and water (farinograph water absorption). The atta and water were mixed for 3 min at I speed (58 rpm) in a dough mixer (Model N-50, Hobart, GmbH, Offenburg, Germany). The dough was rested, divided (40 g each), rounded, sheeted to 3 mm thickness, cut into circular shape of 15 cm diameter, baked on a thermostatically controlled hot plate on one side (side 1) for 20 s, 15 s on the other side (side 2) at 180 °C. The baked chapati was allowed to puff by baking it in a gas tandoor oven with side 1 facing the lid for 15 s. The baked chapatis were cooled for 15 min and packed using poly propylene pouches.

Objective and sensory evaluation

The physical parameters of chapati such as weight, diameter and thickness were measured. The shear value which is the force required to shear a rectangular piece of chapati (2 cm × 6 cm) was measured using Warner–Bratzler shear attachment with a plunger speed of 100 mm/min of texture analyser (Model TAHdi, Stable Microsystems, Surrey, UK).

The quality parameters of chapati such as colour, appearance, pliability, tearing strength, taste and aroma and mouthfeel were evaluated using a 10-point scale by a panel of 24 panellists. The combined score (60) of these six quality parameters was considered as the overall quality score. An excellent chapati should offer slight resistance to tear (tearing strength); have pleasant brown colour with a smooth surface (appearance), greater folding ability (pliability), soft texture with little chewiness (mouthfeel) and little sweetish taste with typical wheatish aroma (taste and aroma). The chapatis having these characteristics were given maximum scores by the judges. On the other hand, the chapatis offering higher resistance to tear, with dull brown colour and rough surface, little pliability, hard or leathery texture, excessive chewiness (mouthfeel), bland taste were given least scores.

Scanning electron microscopy (SEM) studies

SEM studies were carried out according to the procedure followed by Bhargava et al. (2013) using a Leo scanning electron microscope (Model 435 VP, Leo Electronic Systems, Cambridge, UK).

Nutritional characteristics

From the chapati making trials, it was noted that incorporation of FM to RA above 10% caused adverse effect on the quality characteristics of chapati; hence FM at 10% level was fixed and used for all the nutritional analysis in comparison with CA. Determination of moisture, ash, total protein, fat (AACC 2010) and dietary fiber (AOAC 2010) in chapati samples was carried out (AACC 2010; AOAC 2010) by following the methods described earlier.

Carbohydrate digestive profile

Chapati prepared from differently milled wheat flours such as CA, RA and 10% RAFM were analyzed for carbohydrate digestive fractions by a simulated gastrointestinal digestion method (Englyst et al. 1992). Here the pattern of carbohydrate digestion and pattern of glucose released at definite periods of stimulated digestion, i.e., RAG is the glucose while SAG is the glucose released after subsequent 100 min of digestion. Total glucose aliquots were obtained after subsequent hydrolysis with potassium hydroxide and complete enzymatic digestion. Resistant starch was the starch that was left undigested. Free glucose was determined separately. Glucose oxidase–peroxidase method was used to estimate glucose concentration in all the fractions.

Statistical analysis

Statistical analysis of data was carried out using analysis of variance in completely randomized design for the determination of the level of significance. The means were further compared using Duncan's new multiple range tests. The significant level was set to P ≤ 0.05.

Results and discussion

Particle size index

The results of the particle size analysis of the samples showed wide variation in the fineness of the atta samples produced by different milling methods and for the different RAFM samples (Table 1). These differences in the particle size can be attributed to the differently imparted forces during grinding. The stone milled atta (CA) was finest among different samples because the percentage of particles passing through 100 μm size was highest (73.3%). The percentage of the other samples for the same size was found to be 37.5, 40.6, 42.9 and 43.1% respectively for RA, 5, 10 and 15% RAFM respectively. These results indicate that RA was coarsest when compared to 5–15% RAFM because RA was ground in roller flour mill by the technique of first separating the flour streams from bran and then grinding the bran in a hammer mill and recombining it back with the flours. The bigger bran size is responsible here for the granulation of atta being coarser compared to CA. It was observed that relatively finer granulation of 5, 10 and 15% RAFM samples has an increase in the percentage of the number of particles passing through 100 μm size from 40.6 to 43.1%. This is because of the finer particle size of the non-wheat grains which were ground in the hammer mill using the fine sieve. The percentages of particles passing through 250 and 500 μm also confirmed that among different atta samples CA was finest, RA was coarsest and 5–15% RAFM was medium fine.

Table 1.

Proportion of particles (%) in the different size classes

Samples	< 100 µm	< 250 µm	< 500 µm
CA	73.3d ± 0.15	94.5d ± 0.22	99.7d ± 0.30
RA	37.5a ± 0.22	64.6a ± 0.25	88.7a ± 0.35
5% RAFM	40.6b ± 0.15	69.3b ± 0.12	96.5b ± 0.20
10% RAFM	42.9c ± 0.22	72.3c ± 0.18	97.2c ± 0.19
15% RAFM	43.1c ± 0.35	72.8c ± 0.45	97.4c ± 0.25

Values are mean ± standard deviation of three separate determinations (n = 3). Mean SD values followed by different superscripts within the same column are significantly different (P ≤ 0.05)

µm: micro meter; CA: chakki atta; RA: roller flour mill atta; RAFM: roller flour mill atta with functional ingredients mix

Physico-chemical characteristics

The moisture content of the different flours was influenced by the variations in the method of grinding applied for the production of atta. It was observed that moisture content was lowest with 5.81% for CA wherein the grinding was done to get the finer flour compared to other methods, and the values for the moisture content for RA was observed to be 7.32% and it gradually decreased to 6.58% with increased addition of 5–15% FM to RA (Table 2). McCabe et al. (2001) observed that the surface area generated increases greatly as the particle size decreases and the particles lose more moisture in the presence of higher temperature. The higher moisture from RA was because much cooler grinding as only a surface of the rolls touches each other allowing the flour to lose lesser moisture. In 5–15% RAFM the moisture content is being influenced by the initial moistures of the different grains. Milling did not influence the ash or the mineral content of the atta but increased with the addition of 5 to 15% FM to RA which is contributed by the different non-wheat raw materials present in the mix. Damaged starch was found to be 17.56% for the CA, RA (9.22%), 5% RAFM (9.20%), 10% RAFM (9.01%) and 15% RAFM (8.95%).

Table 2.

Physico-chemical characteristics of atta

Atta	Moisture (%)	ash (%)	Damaged Starch (%)	Falling number	Protein (%)	Dietary fibre (%)
CA	5.81a ± 0.11	1.22a ± 0.02	17.56c ± 0.18	427a ± 10	10.01a ± 0.02	9.98a ± 0.02
RA	7.32c ± 0.13	1.28a ± 0.02	9.22b ± 0.12	459c ± 15	10.14a ± 0.02	10.02a ± 0.02
RAFM (%)
5	7.21c ± 0.15	1.41b ± 0.01	9.20b ± 0.11	446b ± 10	11.71b ± 0.02	11.25b ± 0.02
10	6.79b ± 0.15	1.46b ± 0.01	8.51a ± 0.11	440b ± 10	14.1c ± 0.02	12.50c ± 0.02
15	6.58b ± 0.15	1.47b ± 0.01	8.05a ± 0.11	430a ± 10	15.2d ± 0.02	13.50d ± 0.02

Values are mean ± standard deviation of three separate determinations (n = 3). Mean SD values followed by different superscripts within the same column are significantly different (P ≤ 0.05)

Refer Table 1 for abbreviations

These results show that among differently milled atta samples namely CA and RA, CA had the highest damaged starch (17.56%) and RA had the lowest (9.22%) indicating severity of grinding in stone mill. The damaged starch gradually decreased from 9.22 to 8.95% with increased addition of FM from 0 to 15% may be due to increase in protein and dietary fiber and decrease in starch contents. Damaged starch is those starch granules that have been physically changed when compared to their native granular form during wheat milling process. During the preparation of dough, the damaged starch particles take up water more easily compared to intact starch granules (Miralbés 2004). The forces acting on the grains during grinding influence the amount of damaged starch produced. It is observed that the quantity of the damaged starch content produced is directly proportional to the fineness of flour. The two important factors that are responsible for the production of damaged starch are the external factor conforming to the scratching effect by the grooved surface of the mill and the internal factor materializing during the reduction phase when granules are flattened or broken (Ghodke et al. 2009). Depending upon the type of mill employed to produce chapati flour, a varied degree of starch damage is produced. Several authors also observed that extent of size reduction and starch damage are complementary to each other (Solanki et al. 2005; Sharma et al. 2008) and Niu et al. 2014). The falling number value indicates the α-amylase activity in flour and the test is performed to detect degradation of gelatinized starch paste by α-amylase. Generally the falling number value is inversely proportional to the value of damaged starch. Damaged starch is that which has lost its crystallinity, which renders starch granules susceptible to enzymatic degradation in comparison to native starch granules. Every et al. (2002) in their studies on the mill streams have observed a relation between the damaged starch and falling number and reasoned the lower values of damaged starch responsible for higher falling number values (lower α-amylase activity). The lowest falling number value of 427 was observed for CA when compared to all other samples which might be due to the presence of the highest starch damage than other atta samples. Falling Number value decreased from 459 to 430 with an increase in percentage of FM from 0 to 15%.

Protein content increased with increase in the addition of 5, 10 and 15% FM containing protein-rich raw materials like defatted soya flour and Bengal gram flour to RA. The protein content was 10.01% for CA and RA (10.14%) showing not much difference. However, with an increase in the addition of 5, 10 and 15% FM containing protein rich raw materials like defatted soya flour and Bengal gram flour to RA the protein content increased to 11.71, 14.1 and 15.2%. A similar effect was observed in the dietary fibre content, where in the fibre content increased from 10.02 to 13.50% with the incorporation of 15% FM containing fibre sources viz., psyllium husk, oats, barley, and fenugreek fibre to RA as against CA (9.98%).

Rheological characteristics

Farinograph

The farinograph water absorption was 86 and 71.8% for samples CA and RA respectively. It increased with the incorporation of the non-wheat raw materials to RA and ranged from 72.4 to 73.9% (Table 3). Water absorption capacity of the flour increased with lower moisture, higher protein, bran and damaged starch contents in the flour (Hallén et al. 2004). The increase in the water absorption was linear with the fineness of the atta samples. Damaged starch has more water absorption capacity than native or intact starch that is desirable to a certain extent in wheat based products (Mulla et al. 2010). Reduced wheat bran particle size could rise farinograph water absorption by initiating more water interaction through hydrogen bonding in fiber structure (Penella et al. 2008). These results are in line with the amount of damaged starch, protein and moisture content in the respective atta. There was slightly lower dough development values, and higher stability values for RA and 5–15% RAFM samples when compared to CA. It is also reported that there is no correlation with damaged starch in flour samples and dough stability (Haridas Rao et al. 1989).

Table 3.

Rheological characteristics of atta

Parameters	CA	RA	RAFM (%)
			5	10	15
Farinograph
Water absorption (%)	86.0e ± 0.21	71.8a ± 0.42	72.4b ± 0.35	73.1c ± 0.25	73.9d ± 0.35
Dough development time (min)	3.18b ± 0.12	2.31a ± 0.32	2.32a ± 0.15	2.35a ± 0.25	2.42a ± 0.15
Dough stability (min)	1.28a ± 0.02	1.52c ± 0.04	1.45c ± 0.05	1.41b ± 0.04	1.4 b ± 0.05
Amylograph
Pasting temperature °C	60.8a ± 0.25	61.8c ± 0.28	61.4b ± 0.35	61.2b ± 0.15	61.0b ± 0.18
Peak viscosity (BU)	435a ± 5.54	568b ± 6.50	576c ± 4.53	582c ± 6.05	604d ± 8.52
Hot paste viscosity (BU)	311a ± 5.58	394b ± 5.53	400b ± 4.54	411c ± 3.55	424c ± 2.55
Cold paste viscosity (BU)	632a ± 3.55	771c ± 2.55	765c ± 5.55	757b ± 4.55	750b ± 2.85
Breakdown (BU)	124a ± 3.55	174b ± 2.54	176c ± 1.55	177c ± 3.05	180c ± 2.12
Setback (BU)	321a ± 1.54	377d ± 2.44	365c ± 3.25	346b ± 1.52	326a ± 2.25

Values are mean ± standard deviation of three separate determinations (n = 3). Mean SD values followed by different superscripts within the same row are significantly different (P ≤ 0.05)

Refer Table 1 for abbreviations

Amylograph

Pasting temperature provides an indication of the minimum temperature required to cook a given sample. The pasting temperature of CA and RA were 60.8 and 61.8 °C respectively. The lower pasting temperature of CA indicating early onset of initial viscosity could be due to faster water binding capacity, more surface area available and higher damaged starch content present in the finer flour particles in CA. Higher pasting temperature is an indication of delayed swelling due to high resistance of starch granules. The results are in agreement with Inamdar and Prabhasankar (2017). It is also reported that the gelatinisation temperature is less for the flour with finer particle size when compared to coarser fractions (Ahmed et al. 2015). Addition of FM decreased the pasting temperature of RA. Peak viscosity indicates the water-holding capacity of the starch or mixture. The peak viscosity values representing maximum viscosity during heating from 30 to 95 ⁰C describe the swelling capacity of the starch granules prior to their physical breakdown (Tipples et al. 1990). Results of the amylograph studies are shown in Table 3; Among different atta samples, sample CA had the lowest viscosity during heating, cooking, cooling; break down and set back values when compared to RA and RAFM. These results indicate that the starch present in CA has shown early gelatinisation and cooking; lower water absorption capacity and swelling ability prior to physical break down during heating; lesser stability during cooking; lesser viscosity of the paste after cooling to 50 ⁰C. The lower break down and set back values of CA indicate higher rigidity of starch granules during cooking and lower synerisis of starch upon the cooling of the cooked starch paste. Addition of different levels of 5–15% FM to RA decreased the pasting temperature, cold paste viscosity, set back, increased the peak viscosity and hot paste viscosity values.

Overall it can be concluded that the mixture of different starch granules present in 5, 10 and 15% RAFM showed high resistance to swelling, had increased ability to swell freely before breakdown, showed increased resistance to rupture after cooking and decreased retrogradation tendency when compared to CA. These differences in the behavior of starch of RAFM during heating, cooking and cooling are due to lower fineness and damaged starch content present in the samples.

Quality of chapati

Physico-sensory

The moisture content of the chapatis made from RA, 5, 10 and 15% RAFM was found to be in the range of 45.43 to 36.53% when compared to CA (46.51%). A brown color, even surface with brownish baked spots scattered on the surface is desirable characteristics of a good quality chapati. It should be pliable, have a soft texture with little chewiness and possesses typical wheat aroma. The chapati made from sample CA recorded highest sensory score for colour (9.5), appearance (9.5), pliability (9.5), tearing strength (9.5), taste and aroma (9.0) and mouthfeel (9.0) with an overall quality score of 56.0 for the maximum score of 60 (Table 4). CA containing finer particle size, an ideal proportion of damaged starch produced chapatis with better pliability, tearing strength and mouthfeel. The results are in agreement with Ghodke et al. (2009). The chapatti from RA showed dull brown colour, lesser folding ability, lesser softness with more chewiness, slight bland taste and aroma without any sweetish taste. Hence the chapatti from RA had lower sensory scores for these characteristics when compared to chapatti from CA. This is reflected in the overall quality score of 45 as against 56 for CA.

Table 5.

Carbohydrate digestive profile of chapati – g/100 g on as is basis

Chapati	CA	RA	10% RAFM
Rapidly available glucose	40.74c ± 1.20	37.1b ± 1.20	33.60a ± 1.00
Slowly available glucose	3.24a ± 0.10	6.07 b ± 0.10	8.46 c ± 0.10
Free sugar glucose	2.46a ± 0.80	2.67b ± 0.70	2.70c ± 0.10b
Rapidly digestible starch	34.45c ± 0.90	31.02b ± 10.80	27.81a ± 0.80
Slowly digestible starch	2.92a ± 0.10	5.46b ± 0.10	7.61c ± 0.10
Total glucose	46.12b ± 1.60	47.20 c ± 1.30	45.50a ± 1.50
Total starch	39.29b ± 1.20	40.08b ± 0.80	38.82a ± 0.90
Resistant starch	1.93a ± 0.10	2.55b ± 0.10	3.10 c ± 0.10

Values are mean ± standard deviation of three separate determinations (n = 3). Mean SD values followed by different superscripts within the same row are significantly different (P ≤ 0.05)

Refer Table 1 for abbreviations

Table 4.

Chapati making characteristics of atta

Parameters	CA	RA	RAFM (%)
			5	10	15
Physical
Shear force (g)	1350e ± 3.55	1300d ± 8.52	1250c ± 7.50	1180b ± 5.52	1125a ± 3.55
Moisture (%)	46.51d ± 0.05	45.43c ± 0.03	42.36b ± 0.10	40.39b ± 0.06	36.53a ± 0.11
Sensory
Colour (10)	9.5d ± 0.05	8.0b ± 0.01	8.5c ± 0.03	8.5c ± 0.12	7.5a ± 0.15
Appearance (10)	9.5d ± 0.03	8.0b ± 0.02	8.5c ± 0.02	8.5c ± 0.15	7.5a ± 0.05
Pliability (10)	9.5d ± 0.02	8.0a ± 0.03	8.5b ± 0.03	9.0c ± 0.12	8.0a ± 0.03
Tearing strength (10)	9.5d ± 0.05	8.0a ± 0.05	8.5b ± 0.05	9.0c ± 0.08	8.0a ± 0.02
Taste and aroma (10)	9.0d ± 0.06	6.5a ± 0.06	7.5b ± 0.06	8.5c ± 0.09	8.0c ± 0.06
Mouthfeel (10)	9.0e ± 0.04	6.5a ± 0.10	7.5c ± 0.05	8.5d ± 0.12	7.0b ± 0.10
OQS (60)	56.0e ± 0.30	45.0a ± 0.42	49.0c ± 0.53	52.0 d ± 0.54	46.0b ± 0.52

Values are mean ± standard deviation of three separate determinations (n = 3). Mean SD values followed by different superscripts within the same row are significantly different (P ≤ 0.05)

Refer Table 1 for abbreviations

OQS overall quality score

The sensory scores for appearance, pliability, tearing strength, aroma, eating quality increased with increase in the addition of FM up to 10%. At 15% level of FM, the chapaties were slightly rough, less pliable, chewy and showed foreign taste. The overall quality score of RA chapati was 45, 5% RAFM (49), 10% RAFM (52) and 15% RAFM (46) indicating the significant adverse effect at above 15% level of FM. Thus, use of FM at 10% level is recommended.

SEM of atta

Microscopic examination showed that the degree of distortion and irregularity of starch granules was more in CA (Fig. 1a). Here the structural integrity of starch granules changed from smooth to that of a flaky, rough surface (Tester & Morrison 1990). Also, there is an increased presence of clumps formed by distorted granules along with blebs and pits on the surface of these starch granules reflecting the extensive damage caused due to chakki/stone mill. The observations were similar to the granular damage observed by Baldwin et al. (1995) in case of ball milled potato starch. In Fig. 1b, c, the granules observed were more regularly shaped, smooth surface with lesser evidence of cracks indicating that the extent of damage in roller mill was very less as compared to that of chakki/stone mill. In Fig. 1c, the increased presence of both small and large starch granules was observed indicating that the heterogeneity is due to compositional differences in the atta. There is also an increase in starch granules embedded in the gum coating of proteins and this may be attributed to increased protein content in 10% RAFM (Ray et al. 2018). Several authors have reported protein bodies attached to starch granules in soya, chick pea and fenugreek seeds (Indrani et al. 2010; Chin et al. 2000).

Fig. 1.

Scanning electron micrographs (SEM) of atta. a: CA; b RA; c 10% RAFM. Refer Table 1 for abbreviations

Thus, it is evident that the stone mill causes increased surface damage to starch granules and corroborates with our earlier findings that roller mill has lesser damaged starch content. Thus the increase in damaged starch granules obtained through stone milling render them more susceptible to hydration and enzymatic attack. It is consequently hypothesized that this surface damage renders the granules to increased susceptibility to hydration which in turn affects the rate of carbohydrate digestibility.

Nutritional characteristics

The chapatis from CA and RA almost had similar protein and dietary fiber contents. Addition of 10% FM to RA resulted in an increase of these contents by 1.3, 1.2 times respectively and it is due to the presence of protein rich raw materials like chick pea and soya,

Carbohydrate digestive profile

The carbohydrate digestibility of the foods is a reflection of changes in the blood sugar over some time. Food formulations with higher SAG, RS, and lower RAG are considered to be beneficial to health. The rapidly available glucose (RAG) of the chapati made from RA was found to be 3.6 g/100 g lower than that of CA. With the incorporation of 10% FM to RA, a further decrease in RAG of about 3–4 g/100 g was observed. This accounted for a 9% decrease in RAG of RA compared to CA and 9.4% decrease in RAG from RA to 10% RAFM. The differences in the RAG between differently milled samples (P ≤ 0.05) and between RA and RAFM (P ≤ 0.05) were found to be statistically significant. The reduction in RAG from CA to 10% RAFM was 17.5%, and the extent of reduction was of much higher level (P ≤ 0.001). Slowly available glucose (SAG) in 10% RAFM increased to 8.46 g/100 g as against CA (3.24 g/100 g). The extent of increase in SAG in 10% RAFM was found to be statistically significant (P ≤ 0.001). An inverse relationship is found between RAG and SAG along with RS. The decrease in RAG in 10% RAFM resulted in a corresponding increase in both SAG (i.e. 8.46 g/100 g as against 3. 24 g/100 g of CA), resistant starch content (i.e. 3.10 g/100 g as against 1.93 g/100 g of CA) and the differences in both the cases were found to be statistically significant (P ≤ 0.001). Milling or grinding methods studied here resulted in physico-chemical and structural differences in the flours. It is to be noted that damaged starch influenced water absorption capacity and in turn the carbohydrate digestibility. RAG is dependent on multitude factors such as particle size, damaged starch, compositional differences of the native and formulated matrix among which dietary fibre and protein content being the major ones.

Conclusion

The present study shows that the milling or grinding methods adopted in the production of atta has an effect on physico-chemical and structural differences in the flours which influences the particle size and starch intactness which in turn affects carbohydrate digestibility. The effect of added dietary fiber and protein based ingredients on the modulation of carbohydrate digestibility suggest that processing treatments and added ingredients work synergistically. Since atta is a staple food and the demand for the packaged atta is on the rise, the application of the present study provides an option for the flour miller to launch a new atta in the health and wellness segment.

Compliance with ethical standards

Conflict of interest

The authors have declared no conflict of interest.