Food Chemistry 173 (2015) 250–256

Contents lists available at ScienceDirect

Food Chemistry

journal homepage: www.elsevier .com/locate / foodchem

Can bread processing conditions alter glycaemic response?

http://dx.doi.org/10.1016/j.foodchem.2014.10.0400308-8146/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Singapore Institute for Clinical Sciences, ClinicalNutrition Research Centre, 14 Medical Drive, #07-02, Singapore 117599, Singapore.Tel.: +65 6407 0793; fax: +65 6776 6840.

E-mail address: jeya_henry@sics.a-star.edu.sg (J. Henry).

Evelyn Lau a, Yean Yean Soong a, Weibiao Zhou b, Jeyakumar Henry a,c,⇑a Clinical Nutrition Research Centre, Singapore Institute for Clinical Sciences, 14 Medical Drive, #07-02, Singapore 117599, Singaporeb Food Science and Technology Programme, Department of Chemistry, National University of Singapore, S14 Level 5, Science Drive 2, Singapore 117543, Singaporec Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, 14 Medical Drive, #14-01, Singapore 117599, Singapore

a r t i c l e i n f o a b s t r a c t

Article history:Received 7 July 2014Received in revised form 12 September2014Accepted 7 October 2014Available online 19 October 2014

Keywords:BreadGlycaemic responseProcessing

Bread is a staple food that is traditionally made from wheat flour. This study aimed to compare the starchdigestibility of western baked bread and oriental steamed bread. Four types of bread were prepared:western baked bread (WBB) and oriental steamed bread (OSB), modified baked bread (MBB) made withthe OSB recipe and WBB processing, and modified steamed bread (MSB) made with the WBB recipe andOSB processing. MBB showed the highest starch digestibility in vitro, followed by WBB, OSB and MSB.A similar trend was observed for glycaemic response in vivo. MBB, WBB, OSB and MSB had a glycaemicindex of 75 ± 4, 71 ± 5, 68 ± 5 and 65 ± 4, respectively. Processing differences had a more pronouncedeffect on starch digestibility in bread, and steamed bread was healthier in terms of glycaemic response.The manipulation of processing conditions could be an innovative route to alter the glycaemic responseof carbohydrate-rich foods.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Carbohydrates are quantitatively the most important dietaryenergy source for humans, typically accounting for 45–70% of totalenergy intake (Lafiandra, Riccardi, & Shewry, 2014). They play animportant role in energy metabolism and glucose homeostasis.Carbohydrate foods that increase blood glucose rapidly are knownas high glycaemic index (GI) foods, and those that increase bloodglucose gradually are known as low GI foods (Jenkins, Wolever, &Taylor, 1981). They may be divided into low GI (GI 6 55), mediumGI (56 6 GI 6 69) or high GI (GI P 70). Diet is known to play a cri-tical role in the aetiology and management of obesity and diabetes(Thomas & Pfeiffer, 2012). A large number of studies, includingobservational prospective cohort studies as well as randomizedcontrolled trials, show a positive association between consumptionof low GI food in the prevention of obesity, diabetes and cardiovas-cular diseases (Brand-Miller, McMillan-Price, Steinbeck, &Caterson, 2009). The glycaemic response is also influenced by thequantity of carbohydrates consumed. Glycaemic load (GL) is ameasure to quantify the overall glycaemic effect of a portion offood, and takes into account the amount of carbohydrates presentin a serving (Foster-Powell, Holt, & Brand-Miller, 2002).

Bread is a staple food that is traditionally made from wheatflour. It is consumed in different parts of the world, albeit in differ-ent forms due to variations in the choice of ingredients used, andprocessing techniques employed. The GI value of bread wasreported to range from 40 to 97 (Fardet, Leenhardt, Lioger,Scalbert, & Remesy, 2006). The wide-ranging GI values may bedue to: (a) differences in molecular configuration of starch present,(b) variations in cooking and processing methods that resulted indiffering degrees of starch gelatinization, (c) differences instructure of bread in terms of compactness and viscosity, and(d) possible interactions with other food components, such asstarch–protein and starch–lipid interactions, that could impedestarch digestibility (Thondre, 2013).

In Asia, notably China and its surrounding regions, steamedbread is a popular staple. In contrast to baked bread, orientalsteamed bread is typically made using low to medium protein con-tent wheat flour that undergoes fermentation and is further cookedby steaming, rather than baking. A distinction should be madebetween steamed bread (‘‘mantou’’) and steamed bun (‘‘baozi’’).Mantou is plain steamed bread without any filling, whereas baoziis steamed bread containing sweet or savoury fillings made of beanpaste or minced meat. Thus, western oven-baked bread and orien-tal steamed bread vary in both the ingredients used, preparationmethods, and heating methods applied. Although glycaemicresponse of western bread has been researched intensively inrecent years, there is a dearth of information related to the glycae-mic response and glycaemic index of oriental steamed bread. This

E. Lau et al. / Food Chemistry 173 (2015) 250–256 251

study aimed to compare the in vitro starch digestibility and in vivoglycaemic response of western baked bread and oriental steamedbread. More importantly, the study focused on investigating howdifferences in macronutrient composition and processing condi-tions (namely, mixing time, mixing intensity, proofing period andmethod of cooking) influenced glycaemic response.

2. Materials and methods

2.1. Chemicals, reagents and bread ingredients

Sodium hydroxide pellets and concentrated hydrochloric acid(5 M) were purchased from Merck chemicals (Darmstadt,Germany). Pancreatin (P7545, 8X USP specifications), pepsin(800–2500 units/mg) and amyloglucosidase (P300 U/ml) used inthe in vitro digestion protocol were purchased from Sigma–AldrichCompany Ltd. (St. Louis, USA). Amyloglucosidase (E-AMGDF,3260 U/ml) used for the secondary digestion in the reducing sugarassay was obtained from Megazyme International (Wicklow,Ireland). Absolute ethanol was obtained from Fisher ScientificCompany (Fairfield, USA). Milli-Q ultrapure water was usedthroughout the experiments (Billerica, USA). The maleate (0.2 M/pH 6) and acetate (0.1 M/pH 5.2) buffers were prepared accordingto previously described methods (Mishra & Monro, 2009). Lancetswere purchased from Abbott (Abbott, UK) for in vivo GI testing.Glucotrol (Eurotrol, Sweden) was used for daily quality checkingof glucose metres to ensure reliability of results.

High protein wheat flour (Prima bread flour, Singapore), med-ium protein wheat flour (Hong Kong Bake King Flour, Singapore),vegetable shortening (Bake King, Singapore), yeast (SAF instant,France), salt (Fairprice, Singapore) and sugar (Fairprice, Singapore)were purchased from the local supermarket. Potable water wasused for preparation of bread.

2.2. Bread making process

Four types of bread were prepared for this study (Table 1).Western baked bread (WBB) was prepared using standard recipeingredients (with the use of high protein flour) and processingsteps, including baking at 210 �C, and oriental steamed bread(OSB) was prepared using standard recipe ingredients (with theuse of medium protein flour) and processing steps, includingsteaming at 100 �C. The standard recipes for WBB and OSB wereadapted from previously described methods (Ananingsih, Gao, &Zhou, 2013; Wang & Zhou, 2004). Modified baked bread (MBB)was prepared using oriental steamed bread recipe ingredientsand baked bread processing steps (baking at 210 �C). Modifiedsteamed bread (MSB) was prepared using western baked breadrecipe ingredients and steamed bread processing steps (steamingat 100 �C). Bread was freshly prepared on the morning of the study.

2.3. Analytical methods

Protein, fat and moisture contents were determined accordingto AACC methods 46-11.02, 30-25.01 and 44-01.01 respectively(AACC International, 2000). Protein content was analysed withFOSS Kjeltec Systems (FOSS, Denmark). Fat content was analysedwith FOSS Soxtec 2055 (FOSS, Denmark). Dough, proofed doughand bread volume were determined with a VSP 600 Volscan Profi-ler (Stable Micro System Ltd., UK). The sample was mounted on astand, and a laser sensor was used to scan the rotating sample tomeasure the contours of the sample at regular intervals for calcu-lation of volume using the installed computer software. Specificvolume was determined by calculating the ratio between thevolume of the dough or bread and its weight. These measurements

were carried out in triplicates, with two measurements per analy-sis. The total available carbohydrate content of each type of breadwas determined using Megazyme assay kit (Megazyme, Ireland).

2.4. In vitro analysis of starch digestibility

In vitro starch hydrolysis and quantification of sugars releasedduring digestion were carried out according to previouslydescribed methodology (Mishra & Monro, 2009; Ranawana &Henry, 2013). About 2.5 g of each type of bread was weighed andanalysed. Rapidly digestible starch (RDS) and slowly digestiblestarch (SDS) were quantified. RDS was defined as starch that wasrapidly digested within 20 min into pancreatic digestion phase,whereas SDS defined as starch that was digested within20–120 min of pancreatic digestion phase. Triplicates were carriedout for the in vitro analysis of starch digestibility.

2.5. In vivo analysis of glycaemic response

2.5.1. SubjectsThirteen healthy subjects (seven male and eight female; age

24 ± 5 years old; BMI 21.2 ± 1.8 kg/m2; values expressed asmeans ± SD) participated in the study. The inclusion criteria forhealthy subjects were: age between 21 and 50 years old, BMIvalues ranging between 18.0 and 24.9 kg/m2, blood pressure values6120/80 mmHg, and fasting blood glucose levels 66.0 mmol/l.Subjects who smoked, had metabolic diseases or took part in sportsat competitive levels were excluded from the study. Ethicsapproval was given by the National Healthcare Group DomainSpecific Review Board. Written informed consent was obtainedfrom subjects prior to participation in the study.

2.5.2. Study designThe study protocol used was in accordance with procedures

recommended by the FAO/WHO/ISO (FAO, 1998). The subjectswere instructed to avoid vigorous exercise and excessive alcoholthe day before the test, and to consume dinner in standard portionsthe evening before in order to avoid the second meal effect(Wolever, Jenkins, Ocana, Rao, & Collier, 1988). Subjects wererequested to fast for at least 10 h prior to the test, and to reportto the testing site between 0800 and 0900 h on the day of the testsession. Bread was presented to subjects in a randomized order onfour separate test sessions. Each portion of bread served wasequivalent to 50 g of total available carbohydrate content toaccount for differences in recipe formulation. The reference foodwas 50 g of anhydrous glucose dissolved in 250 ml of potablewater. The mean glycaemic response of reference food, calculatedas the average glycaemic response from three separate test ses-sions of consumption of 50 g glucose, was used for the calculationof GI of test bread. This was done to account for inter-day variabil-ity of subjects. There was at least a 1 day gap between GI measure-ments to minimise carry-over effects. The test and reference foodwere served with plain drinking water. Subjects were instructedto finish the test food within 15 min, and physical activity was keptto a minimum during the test.

Fasting blood samples were taken at �5 min and 0 min. Fingerswere gently massaged prior to finger pricking. Baseline valueswere calculated as a mean of the two values with coefficient of var-iation <4%. The test food was consumed after taking baseline mea-surements, and timing started with the first bite of the test food.Further blood samples were taken at 15, 30, 45, 60, 90 and120 min. Blood glucose was measured with a HemoCue Glucose201+ RT analyser (Hemocue� Ltd., Sweden). The glucose metreswere checked daily using Glucotrol to ensure reliability of the mea-surements. Incremental area under blood glucose curves (IAUC)was calculated geometrically (Brouns et al., 2005). GI value of

Table 1Ingredients and processing parameters for WBB, MSB, OSB and MBB.

WBB MSB OSB MBB

Western baked bread Modified steamed bread made with conventionalbaked bread recipe

Oriental steamed bread Modified baked bread made with steamed breadrecipe

IngredientsFlour (g) 1000 1000 1000 1000Water (g) 590 590 550 550Sugar (g) 40 40 10 10Salt (g) 10 10 10 10Yeast (g) 10 10 10 10Shortening

(g)30 30 0 0

ProcessingMixing – Mix dry ingredients at level 1 intensity for

1 min– Add water, mix at level 1 intensity for 1 min,

followed by level 3 intensity for 7 min

– Mix dry ingredients at level 1 intensity for1 min

– Add water, mix at level 1 intensity for 1 min,followed by level 2 intensity for 4 min

– Mix dry ingredients at level 1 intensity for1 min

– Add water, mix at level 1 intensity for 1 min,followed by level 2 intensity for 4 min

– Mix dry ingredients at level 1 intensity for1 min

– Add water, mix at level 1 intensity for 1 min,followed by level 3 intensity for 7 min

Restinga 15 min 15 min 15 min 15 minProofingb 70 min 40 min 40 min 70 minCooking Baking at 210 �C, 11 min Steaming at 100 �C, 10 min Steaming at 100 �C, 10 min Baking at 210 �C, 11 min

a Resting was carried out at room temperature.b Proofing was carried out at 40 �C, 85% relative humidity.

252E.Lau

etal./Food

Chemistry

173(2015)

250–256

E. Lau et al. / Food Chemistry 173 (2015) 250–256 253

bread for each subject was calculated by expressing the IAUC ofbread as a percentage of IAUC of the mean of the three glucosetests. The GI of each type of bread was calculated as the mean GIfor all the subjects. The GL of each type of bread was calculatedby: (GI � 50 g total available carbohydrate)/100.

2.6. Statistical analysis

Data processing was carried out using SPSS software (version16, USA). Data was presented as means ± standard deviation (SD)or means ± standard error of mean (SEM), as indicated. The specificvolume results were analysed using Kruskal–Wallis test, and posthoc comparisons were carried out using Mann–Whitney U testwith Bonferroni’s correction. In vitro starch digestibility resultswere evaluated using one-way ANOVA, and post hoc comparisonswere carried out using Bonferroni’s correction. For in vivo GI analy-sis, differences in postprandial blood glucose concentrations, IAUCand GI values were evaluated using repeated-measures ANOVAwith Bonferroni’s correction. The glycaemic response of the differ-ent types of bread was evaluated using repeated measures ANOVA,with Bonferroni’s correction. The differences in IAUC and GIbetween male and female subjects were compared using indepen-dent samples t-test. Statistical significance was set at p < 0.05 forall cases.

3. Results and discussion

3.1. Specific volume of bread

The specific volume of dough, proofed dough and bread weremeasured to compare the development of bread structure (Table 2).MBB dough had the highest specific volume, and specific volumefor other types of dough did not differ significantly (p > 0.05).Increased protein content of flour has previously been found tohave a positive effect on bread specific volume (Barak, Mudgil, &Khatkar, 2013). WBB had the highest specific volume for bothproofed dough and baked bread, followed by MBB, OSB and MSB.The higher protein content in WBB (Table 3) may result in the for-mation of a more extensive gluten network during dough develop-ment which enabled the more effective retention of gas duringproofing, thereby resulting in a higher specific volume as comparedto MBB. On the other hand, the specific volumes of MSB proofeddough and bread were found to be significantly lower than OSB,

Table 3Nutrient composition of wheat flours (g/100 g), WBB, MSB, OSB and MBB (per 50 g availa

Component Wheat flour

High protein Medium pro

Portion size (g) – –Total available carbohydrate (g) 72.2 77.3Moisture (g) 12.2 11.8Protein (g) 12.8 7.4Fat (g) 1.8 1.3

Table 2Specific volume (ml/g) of dough, proofed dough and bread of WBB, MSB, OSB andMBB.

WBB MSB OSB MBB

Dough 0.87 ± 0.01a 0.85 ± 0.01a 0.88 ± 0.01a 0.96 ± 0.02b

Proofed Dough 2.74 ± 0.22a 1.77 ± 0.09b 2.06 ± 0.12c 2.36 ± 0.12d

Bread 3.95 ± 0.34a 2.59 ± 0.10b 2.95 ± 0.11c 3.43 ± 0.10d

Values are represented as means ± SD. Values within a row with different super-script letters are significantly different (p < 0.05).

despite having higher protein content. This suggested that specificvolume of bread was not only dependent on protein content, butwas also influenced by processing parameters.

During optimal dough development, gluten proteins arehydrated and undergo dis-aggregation, in which glutenins alignand form cross links with glutenins, and cross-linked proteinsheets are layered one over another to form the gluten network(Kontogiorgos, 2011). Energy input during dough developmentincreased with mixing intensity and duration. The gluten networkhas been shown to be weakened and ruptured with excessivelyhigh energy input, resulting in reduced dough visco-elasticity(Peighambardoust, Fallah, Hamer, & van der Goot, 2010). Increas-ing the length of fermentation time also had a positive effect onbread volume for optimally developed dough, in which the expan-sion of gas cells brought about an increase in porosity of breadcrumb. The use of higher mixing speed, longer mixing times andlonger proofing period during the WBB and MBB bread makingprocedure resulted in WBB and MBB with significantly higher spe-cific volume for both proofed dough and bread, as compared to OSBand MSB. In addition, bread typically experiences oven-springupon heating, due to further expansion of gas volumes at elevatedtemperatures (Mondal & Datta, 2008). The use of higher tempera-tures during baking (210 �C), as compared to steaming (100 �C)may have resulted in a faster rate of temperature increase of doughfor baked bread as compared to steamed bread, resulting in higherspecific volumes for WBB and MBB, as compared to OSB and MSB.MSB had the lowest specific volume, possibly due to under-development of dough. High protein flour was used for preparationof MSB, but underwent a short mixing period with low intensity,and this could have resulted in insufficient development of the glu-ten network. The combination of low energy input during doughdevelopment, as well as a shortened fermentation period, resultedin MSB having the least porous structure as compared to othertypes of bread.

3.2. Starch digestibility

3.2.1. In vitro starch digestibilityBread was subjected to in vitro enzymatic digestion under con-

trolled conditions to quantify the amount of rapidly digestiblestarch (RDS) and slowly digestible starch (SDS) (Fig. 1). In a pre-vious study, RDS had been found to show good correlation within vivo glycaemic response, and it could therefore be a proxy indi-cator of GI value (Englyst, Vinoy, Englyst, & Lang, 2003). RDS wasfound to be the predominant starch fraction in all four types ofbread, as most of the starch was fully gelatinized, and bread typi-cally has an open structure, rendering starch highly accessible tohydrolysis by amylase (Ranawana & Henry, 2013; Ronda, Rivero,Caballero, & Quilez, 2012). Comparing MBB with WBB (differentrecipe ingredients and same processing procedures), it was foundthat MBB had significantly higher RDS content than WBB. Simi-larly, comparing OSB with MSB (different recipe ingredients andsame processing procedures), OSB had significantly higher RDScontent than MSB. Although more sugar was used in the recipe

ble carbohydrate portion).

Bread

tein WBB MSB OSB MBB

102 120 105 8750 50 50 5035 52.6 47 29.59.1 9.0 4.9 5.05.6 6.6 1.9 1.6

Fig. 1. Rapidly digestible starch and slowly digestible starch contents of WBB, MSB, OSB and MBB.

254 E. Lau et al. / Food Chemistry 173 (2015) 250–256

for preparing WBB and MSB, the medium protein content wheatflour used for preparing OSB and MBB had a higher available carbo-hydrate content as compared to the high protein content wheatflour used for preparing WBB and MSB (Table 3). Even though lesssugar was used in the recipe for OSB and MBB, it resulted in higherRDS amounts being detected during in vitro studies.

Processing conditions had a marked effect on in vitro digestibil-ity. Baked breads (WBB and MBB) were found to have higher RDScontent. This could be attributed to the higher specific volumeand porosity, resulting in increased accessibility of amylases tostarch granules, rendering starch more susceptible to hydrolysis.A previous study has shown that an increase in degree of mixingin dough results in higher amounts of RDS content, possibly dueto a weakened gluten matrix that renders starch granules moreaccessible to enzymatic digestion (Parada & Aguilera, 2011). WBBand MBB were found to have significantly higher SDS content thanMSB and OSB (p < 0.05). Rapid evaporation of water from the out-ermost region of dough in the presence of dry heat resulted inincomplete gelatinization of starch granules (Primo-Martin, vanNieuwenhuijzen, Hamer, & van Vliet, 2007). Digestibility has beenshown to reduce with decreased gelatinization, as limited swellingand hydration decreases the chemical reactivity of starch granulestowards amylolytic enzymes (Parada & Aguilera, 2009). The use ofmoist heat during steaming did not result in crust formation,accounting for the significantly lesser SDS content found in OSBand MSB.

Fig. 2. Blood glucose responses to 50 g equivalent

3.2.2. In vivo glycaemic responseThe postprandial blood glucose responses to steamed and baked

bread showed different temporal profiles as shown in Fig. 2. Inagreement with a previous study (Brouns et al., 2005), there wereno significant differences in IAUC and GI between male and femalesubjects. Peak glycaemic response for OSB and MSB was observedat 30 min, whereas WBB and MBB showed a delayed peak glycae-mic response at 45 min. WBB and MBB gave rise to higher peakglucose concentrations than OSB and MSB, and the higher bloodglucose concentrations were sustained until 120 min. This resultedin higher IAUC for WBB and MBB, but the differences were notsignificant amongst the four types of bread. Physical structurewas found to be an important factor in determining glycaemicresponse of bread, as reported in previous studies (Burton &Lightowler, 2006; Lioger et al., 2009). The manipulation of physicalstructure, and in turn, starch digestibility, was brought about bydifferences in processing procedures. A more compact bread struc-ture could have hindered the accessibility of amylase to starchgranules, resulting in a slower rate of glucose release, and reducedglycaemic response in OSB and MSB.

The total IAUC of MBB was slightly higher than WBB, with GIvalues of 75 and 71 respectively (Table 4). MBB and WBB were pre-pared using different recipe ingredients, but underwent the sameprocessing procedure. A similar trend was observed for OSB andMSB, which had GI values of 68 and 65, respectively. The modestdifferences in glycaemic response in this study suggested that

portions of bread in healthy subjects (n = 13).

Table 4Postprandial blood glucose characteristics of WBB, MSB, OSB and MBB.

WBB MSB OSB MBB

Max incremental peak rise (mmol/L) 3.2 ± 0.2a 2.7 ± 0.2a 2.9 ± 0.3a 3.2 ± 0.2a

Time of peak rise (min) 45a 30b 30b 45a

Incremental area under curve (IAUC) 196 ± 17a 179 ± 13a 184 ± 13a 211 ± 18a

GI 71 ± 5a 65 ± 4a 68 ± 5a 75 ± 4a

GL 36a 33a 34a 38a

Values are represented as means ± SEM. Values within a row with same superscript letters are not significantly different (p > 0.05).

E. Lau et al. / Food Chemistry 173 (2015) 250–256 255

macronutrient composition played a minor role in digestibility ofstarch. Digestion of starch is known to be hindered in the presenceof proteins and lipids, as proteins may encapsulate starch granulesto form a protective barrier against enzymatic hydrolysis, andstarch, particularly amylose, may form complexes with lipids toresist breakdown. The lower protein and fat content in MBB andOSB, as compared to WBB and MSB (Table 3), did not change post-prandial glycaemic response markedly. MBB showed a slightlyhigher glycaemic response than WBB in vivo, and had significantlyhigher in vitro starch digestibility than WBB. Despite having alower specific volume, the greater extent of starch digestibility ofMBB, both in vitro and in vivo may be partly attributed to the lowerprotein content in the flour used for the preparation of MBB. MBBdough was subjected to intense energy input during dough devel-opment, and the weakened gluten network in MBB, as seen fromthe lower specific volume of bread, as compared to WBB, couldhave resulted in reduced resistance during enzymatic digestion,and correspondingly a higher glycaemic response.

In vitro starch digestibility and in vivo glycaemic responseresults were in agreement and showed the same ranking for thefour types of bread. MBB was found to have the highest glycaemicresponse, followed by WBB, OSB and MBB. This indicated thatdigestibility of starch and the release of glucose were consistentin both studies. Bread was predominantly made up of RDS, andthe amount of SDS was comparatively too low to significantlyimpact on the glycaemic response. Hence, the higher SDS contentsdid not contribute to reduced glycaemic response in this case. Thephysical structure of bread could be manipulated by processingconditions (namely mixing time and duration, fermentation periodand method of cooking), resulting in pronounced changes in gly-caemic response. When ascertaining whether macronutrient com-position or processing parameters had a greater impact onglycaemic response, the latter was found to play a more pivotalrole in this study, as both types of steamed bread, OSB and MSB,demonstrated lower starch digestibility in vitro and reduced gly-caemic response in vivo. It has been reported that the risks ofdeveloping cardiovascular diseases is associated with an elevatedblood glucose levels in healthy individuals (Braun, Bopp, & Faeh,2013). Modest increases in postprandial blood glucose levels couldalso reduce the long-term risk of developing Type 2 diabetes(Buyken, Mitchell, Ceriello, & Brand-Miller, 2010). In this study,steamed bread emerged as a ‘‘healthier’’ alternative to bakedbread.

Glucose tolerance is known to decrease with age due toimpaired insulin secretion and action. Higher postprandial glucoseconcentrations after consumption of a mixed meal has beenobserved in elderly adults as compared to younger individuals(Basu et al., 2006). Subjects within the age band of 19–50 yearsare typically recruited for GI studies (Wolever et al., 2008), how-ever, the mean age of the subjects recruited for this study was24 years. It is important to recognise that the results in elderlyadults, compared to the relatively young adults in our study, couldbe different. Future studies could be carried out to assess the gly-caemic response and insulinemic index in elderly adults.

4. Conclusion

This study demonstrated for the first time that even with theuse of identical bread recipe ingredients, the application of variedprocessing conditions, namely mixing time, mixing intensity,proofing period and method of cooking, resulted in lower starchdigestibility in vitro and reduced glycaemic response in vivo ascompared to baked bread. Processing played a major role in affect-ing the physical structure of bread. It has been customary to usefood ingredients such as b-glucan, galactomannan, non-starchpolysaccharides and polyols to reduce the glycaemic index of highGI foods (Thondre, 2013). The observation that processing para-meters could impact on glycaemic response of wheat-based foodsprovides us with a new approach to manipulate the glycaemicindex of carbohydrate-rich foods. Work is in process to elucidatethe differences in the microstructure of bread to understand thepotential links between food microstructure and glycaemicresponse.

Conflict of interest

None of the authors declare any conflict of interest.

Acknowledgements

This research was funded by the A⁄STAR Health and LifestyleGrant (Grant No. 112 177 0033). We would also like to thankDr. Tan Sze Wee, Deputy Director of Biomedical Research Council,for his continued support.

References

AACC International. (2000). Approved methods (11th ed.). St Paul, Minnesota:American Association of Cereal Chemists International.

Ananingsih, V., Gao, J., & Zhou, W. (2013). Impact of green tea extract and fungalalpha-amylase on dough proofing and steaming. Food and Bioprocess Technology,6(12), 3400–3411.

Barak, S., Mudgil, D., & Khatkar, B. S. (2013). Relationship of gliadin and gluteninproteins with dough rheology, flour pasting and bread making performance ofwheat varieties. LWT – Food Science and Technology, 51(1), 211–217.

Basu, R., Man, C. D., Campioni, M., Basu, A., Klee, G., Toffolo, G., et al. (2006). Effectsof age and sex on postprandial glucose metabolism differences in glucoseturnover, insulin secretion, insulin action, and hepatic insulin extraction.Diabetes, 55(7), 2001–2014.

Brand-Miller, J., McMillan-Price, J., Steinbeck, K., & Caterson, I. (2009). Dietaryglycemic index: health implications. Journal of the American College of Nutrition,28(Suppl. 4), 446S–449S.

Braun, J., Bopp, M., & Faeh, D. (2013). Blood glucose may be an alternative tocholesterol in CVD risk prediction charts. Cardiovascular Diabetology, 12(1), 24.

Brouns, F., Bjorck, I., Frayn, K. N., Gibbs, A. L., Lang, V., Slama, G., et al. (2005).Glycaemic index methodology. Nutrition Research Reviews, 18(1), 145–171.

Burton, P., & Lightowler, H. J. (2006). Influence of bread volume on glycaemicresponse and satiety. British Journal of Nutrition, 96(5), 877–882.

Buyken, A. E., Mitchell, P., Ceriello, A., & Brand-Miller, J. (2010). Optimal dietaryapproaches for prevention of type 2 diabetes: A life-course perspective.Diabetologia, 53(3), 406–418.

Englyst, K. N., Vinoy, S., Englyst, H. N., & Lang, V. (2003). Glycaemic index of cerealproducts explained by their content of rapidly and slowly available glucose.British Journal of Nutrition, 89(3), 329–339.

256 E. Lau et al. / Food Chemistry 173 (2015) 250–256

FAO (1998). Carbohydrates in human nutrition. Rome: Report of a Joint FAO/WHOExpert Consultation.

Fardet, A., Leenhardt, F., Lioger, D., Scalbert, A., & Remesy, C. (2006). Parameterscontrolling the glycaemic response to breads. Nutrition Research Reviews, 19(01),18–25.

Foster-Powell, K., Holt, S. H. A., & Brand-Miller, J. C. (2002). International table ofglycemic index and glycemic load values: 2002. American Journal of ClinicalNutrition, 76(1), 5–56.

Jenkins, D. J. A., Wolever, T. M. S., & Taylor, R. H. (1981). Glycemic index of foods: Aphysiological basis for carbohydrate exchange. American Journal of ClinicalNutrition, 34(3), 362–366.

Kontogiorgos, V. (2011). Microstructure of hydrated gluten network. Food ResearchInternational, 44(9), 2582–2586.

Lafiandra, D., Riccardi, G., & Shewry, P. R. (2014). Improving cereal graincarbohydrates for diet and health. Journal of Cereal Science, 59(3),312–326.

Lioger, D., Fardet, A., Foassert, P., Davicco, M. J., Mardon, J., Gaillard-Martinie, B.,et al. (2009). Influence of sourdough prefermentation, of steam cookingsuppression and of decreased sucrose content during wheat flakes processingon the plasma glucose and insulin responses and satiety of healthy subjects.Journal of the American College of Nutrition, 28(1), 30–36.

Mishra, S., & Monro, J. A. (2009). Digestibility of starch fractions in wholegrain rolledoats. Journal of Cereal Science, 50(1), 61–66.

Mondal, A., & Datta, A. K. (2008). Bread baking – A review. Journal of FoodEngineering, 86(4), 465–474.

Parada, J., & Aguilera, J. M. (2009). In vitro digestibility and glycemic response ofpotato starch is related to granule size and degree of gelatinization. Journal ofFood Science, 74(1), E34–E38.

Parada, J., & Aguilera, J. M. (2011). Microstructure, mechanical properties, andstarch digestibility of a cooked dough made with potato starch and wheatgluten. LWT – Food Science and Technology, 44(8), 1739–1744.

Peighambardoust, S. H., Fallah, E., Hamer, R. J., & van der Goot, A. J. (2010). Aerationof bread dough influenced by different way of processing. Journal of CerealScience, 51(1), 89–95.

Primo-Martin, C., van Nieuwenhuijzen, N. H., Hamer, R. J., & van Vliet, T. (2007).Crystallinity changes in wheat starch during the bread-making process: Starchcrystallinity in the bread crust. Journal of Cereal Science, 45(2), 219–226.

Ranawana, V., & Henry, C. J. K. (2013). A comparative evaluation of the glycaemicpotential of commercial breads consumed in South East Asia. InternationalJournal of Food Sciences and Nutrition, 64(2), 223–229.

Ronda, F., Rivero, P., Caballero, P. A., & Quilez, J. (2012). High insoluble fibre contentincreases in vitro starch digestibility in partially baked breads. InternationalJournal of Food Sciences and Nutrition, 63(8), 971–977.

Thomas, T., & Pfeiffer, A. F. H. (2012). Foods for the prevention of diabetes: How dothey work? Diabetes/Metabolism Research and Reviews, 28(1), 25–49.

Thondre, P. S. (2013). Food-based ingredients to modulate blood glucose. Advancesin Food and Nutrition Research, 70, 181–227.

Wang, R., & Zhou, W. (2004). Stability of tea catechins in the breadmaking process.Journal of Agricultural and Food Chemistry, 52(26), 8224–8229.

Wolever, T. M. S., Jenkins, D. J. A., Ocana, A. M., Rao, V. A., & Collier, G. R. (1988).Second-meal effect: Low-glycemic-index foods eaten at dinner improvesubsequent breakfast glycemic response. American Journal of Clinical Nutrition,48(4), 1041–1047.

Wolever, T. M. S., Brand-Miller, J. C., Abernethy, J., Astrup, A., Atkinson, F., Axelsen,M., et al. (2008). Measuring the glycemic index of foods: Interlaboratory study.American Journal of Clinical Nutrition, 87(1), 247S–257S.