Digestibility of Selected Resistant Starches in Humans
by
Asmaa Alraefaei
A thesis submitted in conformity with the requirements for the degree of Master of Science
Department of Nutritional Sciences
University of Toronto
© Copyright by Asmaa Alraefaei, 2015
ii
Digestibility of Selected Resistant Starches in Humans
Asmaa Alraefaei
Master of Science
Department of Nutritional Sciences University of Toronto
2015
Abstract
Resistant starches (RSs) are food ingredient sources of fiber. Their fiber content is measured in
vitro, which may not accurately reflect in vivo digestibility. Carbohydrate content of commercial
starches (Hi-Maize® 260 [HM], Hylon® VII [HY] and Amioca [AM]) was determined in vivo.
Fiber values for HM, HY and AM were: 32.8 g, 11.5 g and 0 g in vitro, 12.5 g, 40.8 g and 6.2 g
from blood glucose study (n=10) and 18.7 g, 21.6 g and 1.7 g from ileostomy study (n=3).
13.4% and 43.5% of RS in HM and HY was recovered in effluent. Total bacterial count (TBC)
was 109 CFU/ g of effluent and positive correlation (r=0.4955) existed between TBC and
carbohydrates by difference for HM and HY. The in vitro AOAC 991.43 method underestimates
fiber content in HY and AM and overestimates fiber content in HM and does not accurately
reflect fiber content of RS products.
iii
Acknowledgements
It feels like yesterday when I received an e-mail from the Department of Nutritional
Sciences at the University of Toronto to inform me that I was admitted to the M.Sc. program.
It was one of the most powerful moments in my life and ultimately, one of the saddest.
If I was going to accept the offer of admission, I knew I had to leave my 2 year old son
and 8 month daughter for a while in Kuwait until things have resolved.
I still remember the night I left vividly, I picked up my daughter from her crib, took her
upstairs to my mother’s room, kissed her and lay her down to sleep next to my mother.
With a shattered heart, I walked away to pursue an education.
To my brilliant supervisor, Dr. Thomas Wolever, I want to thank you sincerely for
offering me a position in your lab and I hope I have lived up to your expectations.
To my advisory committee, Dr. Young-In Kim, Dr. Deborah O’Connor and Dr. Elena
Comelli, thank you for your patience and invaluable guidance.
To Dr. David Jenkins, it is a profound honor to have you as my external examiner.
To the sponsor of my project, Ingredion, thank you for allowing this project to take place.
To Glycemic Index Laboratories, thank you for being wonderful people to work with.
To Dr. Wolever’s lab members, Ahmad, Judlyn, Shannan, Sari, Tracy, Evan, Halah and
Kervan, thank you all for your wonderful feedback and training.
A very special thank you to Amel Taibi for training me religiously during my time in
Dr. Comelli’s lab.
iv
To my husband, Ammar, for your endless support and children, Abdulmalek and Dalya.
I love you all more and more each day and I hope my actions have spoken volumes.
To my sisters, my best friends, Sara, Noha, Buthayna and Hadeel.
To Buthayna especially, for always knowing how to put me back together
when I was a million little pieces.
To my mother, Mona, there is nothing I can every say and nothing I can ever do to express my
utmost love and respect for you, you are simply my universe.
To my father, Majid, for teaching me persistence and determination.
To my grandmothers, Moyassar and Sara Margie, may you rest in eternal peace.
To R.E., you will forever hold a very special place in my heart, M.N.M.K.
I dedicate this degree to every Muslim female seeking the right to an education.
This degree is whole-heartedly for you.
Above all, I thank God for all the blessings in my life and pray that may He continue blessing me
with health, wealth, knowledge, peace and happiness in my life.
May this degree be the start to boundless devotion and contribution to academia.
v
Table of Contents
Contents Page No.
Abstract ii
Acknowledgements iii
List of Tables ix
List of Figures x
List of Abbreviations xi
List of Appendices xii
Chapter 1: Introduction 1
Chapter 2: Literature Review 4
2.1: Resistant Starch 4
2.1.1: Coining and Defining of the Term ‘Resistant Starch’ 4
2.1.2: Formation of Resistant Starch in High-Amylose Maize Starch 5
2.1.3: Classification of Resistant Starch 6
2.1.4: Health Benefits of Resistant Starch 7
2.2: Measurement of Resistant Starch 10
2.2.1: In Vitro Methods 10
2.2.1.1: Background 10
2.2.1.2: Initial Methods Developed for the Determination of Resistant Starch 11
2.2.1.3: Development of AOAC Official Method 2002.02 15
2.2.1.4: AOAC 991.43 Method vs. AOAC 2002.02 Method 17
2.2.1.5: Incorporation of Resistant Starch into Dietary Fiber Measurements and the Development of AOAC 2009.01 18
vi
2.2.2: In Vivo Methods 21
2.2.2.1: Breath Hydrogen Test 22
2.2.2.2: Ileostomy Model 23
2.2.2.2.1: Potential Problems with the Ileostomy Model 26
2.3: Resistant Starch as a Food Ingredient 26
2.4: Resistant Starch and Dietary Fiber Labeling 28
Chapter 3: Rationale, Objectives and Hypotheses 31
3.1: Rationale 31
3.2: Objectives 34
3.3: Hypotheses 34
Chapter 4: Study 1 – Digestibility of Selected Resistant Starches Estimated from the Glycemic Responses they Elicit 35
4.1: Background 35
4.2: Methods 36
4.2.1: Subjects 36
4.2.2: Protocol 37
4.2.3: Test Meals 38
4.2.4: Palatability 40
4.2.5: Blood Samples 41
4.2.6: Available Carbohydrates Calculations 41
4.2.7: Statistical Analysis 43
4.3: Results 44
4.3.1: Baseline Characteristics 44
4.3.2: Blood Glucose Analysis 44
4.3.3: Glycemic Response Elicited by Reference Food 45
vii
4.3.4: Palatability 45
4.3.5: Blood Glucose Responses 47
4.3.6: Incremental Areas Under the Blood Glucose Response Curves 48
4.3.7: Relative Glycemic Responses 49
4.3.8: Available Carbohydrates Calculated from Blood Glucose Responses 50
4.3.9: Estimates of Available Carbohydrates in the Test Starches 51
4.4: Discussion 52
Chapter 5: Study 2 – Digestibility of Selected Resistant Starches in Ileostomates 57
5.1: Background 57
5.2: Methods 57
5.2.1: Subjects 57
5.2.2: Protocol 58
5.2.3: Test Meals 60
5.2.4: Freeze-Drying Ileal Effluent 60
5.2.5: Ileal Effluent Analysis 61
5.2.5.1: Proximate Analysis 61
5.2.5.2: AOAC Official Method 2002.02 62
5.2.5.3: Total Bacterial Count Determined by Real-Time Quantitative Polymerase Chain Reaction 62
5.2.6: Statistical Analysis 63
5.3: Results 63
5.3.1: Baseline Characteristics 64
5.3.2: Control Diets 65
5.3.2.1: Summary of the Control Diets 65
viii
5.3.3: Amount of Ileal Effluent Collected 65
5.3.3.1: Amount of Dry Matter Produced 66
5.3.3.2: Mean Percent Dry Matter 67
5.3.4: Proximate Analysis of Ileal Effluent 67
5.3.5: Comparison of Resistant Starch Fed vs. Recovered 69
5.3.6: Comparison of Carbohydrates by Difference Fed and Recovered 70 5.3.7: Amount of Fiber and Available Carbohydrates by Different Methods 71
5.3.8: Total Bacterial Count Determined by Real-Time Quantitative Polymerase Chain Reaction 72
5.3.8.1: Comparison of Log Transformed Values of Total Bacterial Count 73
5.3.8.2: Correlations Between Total Bacterial Count and Carbohydrate Measure 74
5.3.8.2.1: Correlation Between Total Bacterial Count and Resistant Starch 74
5.3.8.2.2: Correlation Between Total Bacterial Count and Carbohydrates by Difference 75
5.4: Discussion 77
Chapter 6: Conclusion and Implications 82
Chapter 7: Limitations and Future Directions 85
References 87
Appendices 101
ix
List of Tables
Tables Page No.
Table 1 – Comparison of Main Initial In Vitro Methods Used to Determine Resistant Starch Content 14
Table 2 – Comparison of Resistant Starch Values Obtained Using Several In Vitro
Methods and In Vivo Results 16 Table 3 – Total Dietary Fiber and Resistant Starch Content of a Range of Samples 18 Table 4 – Comparison of Resistant Starch Values Obtained Using Several In Vitro
Methods (Including AOAC 2009.02 Method) and In Vivo Results 20 Table 5 – Composition of the Test Meals 40 Table 6 – Baseline Characteristics of Study 1 44 Table 7 – Palatability, Incremental Areas Under the Blood Glucose Response Curves
and Relative Glycemic Responses of the Test Meals 46
Table 8 – Available Carbohydrate Calculated from Blood Glucose Responses 50 Table 9 – Estimates of Available Carbohydrate in Test Starches 52 Table 10 – Baseline Characteristics of Study 2 64 Table 11 – Summary of the Control Diets 65 Table 12 – Amount of Ileal Effluent Collected in 10 Hours 66 Table 13 – Amount of Dry Matter Produced in 10 Hours 66 Table 14 – Mean Percent Dry Matter Produced in 10 Hours 67 Table 15 – Proximate Analysis of Ileal Effluent 68 Table 16 – Comparison of Resistant Starch Fed vs. Recovered 69 Table 17 – Comparison of Carbohydrates by Difference Fed and Recovered 70 Table 18 – Amount of Fiber and Available Carbohydrates by Different Methods 71 Table 19 – Total Bacterial Count of Ileal Effluent 73 Table 20 – Comparison of Log Transformed Values of Total Bacterial Count 74
x
List of Figures
Figures Page No.
Figure 1 – Dietary fiber components measured by AOAC Method 991.43 21 Figure 2 – Dietary fiber components measured by AOAC Method 2009.01 21 Figure 3 – Illustration of the Protocol for the Blood Glucose Response Study 38 Figure 4 – Palatability of the Test Meals 46 Figure 5 – Blood Glucose Responses Elicited by the Test Meals 47 Figure 6 – Incremental Areas Under the Blood Glucose Response Curves 48 Figure 7 – Relative Glycemic Responses 49 Figure 8 – Available Carbohydrate Calculated from Blood Glucose Responses 51 Figure 9 – Illustration of the Protocol for the Ileostomy Study 60 Figure 10 – Amount of Fiber and Available Carbohydrates by Different Methods 72 Figure 11 – Correlation Between Total Bacterial Count and Resistant Starch 75 Figure 12 – Correlation Between Total Bacterial Count and Carbohydrates by 76
Difference
xi
List of Abbreviations
Abbreviation Full Term
AM Amioca
ANOVA Analysis of Variance
AOAC Association of Official Analytical Chemists
avCHO Available Carbohydrates
BGR Blood Glucose Response
BMI Body Mass Index
CFU Colony Forming Units
CHOD Carbohydrates by Difference
CV Coefficient of Variation
GI Glycemic Index
HM Hi-maize® 260
HY Hylon® VII
iAUC Incremental Area Under the Blood Glucose Response Curve
ITDF Integrated Total Dietary Fiber
RGR Relative Glycemic Response
RM Repeated Measures
RS Resistant Starch
RT-qPCR Real-Time Quantitative Polymerase Chain Reaction
SIBO Small Intestinal Bacterial Overgrowth
tCHO Total Carbohydrates
TDF Total Dietary Fiber
xii
List of Appendices
Appendices Page No.
A) Appendix 1 – Ileostomy Study Menu 101
B) Appendix 2 – Control Diet of Subject 1 (AR001) 105
C) Appendix 3 – Control Diet of Subject 2 (AR002) 106
D) Appendix 4 – Control Diet of Subject 3 (AR003) 107
1
Chapter 1 Introduction
1.0: Introduction
Starch is a complex carbohydrate produced by most green plants for energy storage and consists
of a large number of glucose units joined together by covalent, glycosidic bonds (Ahuja et al.,
2013). Globally, it is the most common carbohydrate in the human diet, with more than 50% of
the caloric requirement being fulfilled by starch-based staple foods (Ahuja et al., 2013) such as
wheat, maize/corn, rice, potatoes and cassava (Food and Agriculture Organization, 1995). Corn
contains 72% of its dry matter as starch and is considered the third most valuable crop in Canada
with 10.7 million metric tons produced in 2011. The consumption of fresh corn more than
doubled since the 1970s and reached 0.79 kg per capita in 2011 (Statistics Canada, 2014).
Starch is present in amyloplasts as semi-crystalline intracellular water-insoluble granules, with
alternating crystalline and amorphous layers. Starch is composed of two main types of
molecules: linear amylose (25%) and branched amylopectin (75%) along with traces of lipids
(0.1-1.0%) and proteins (0.05-0.5%). Amylose is a linear glucan polymer, composed of α-1,4
linked glucose residues with a degree of polymerization ranging from 800 in maize to 4,500 in
potato (Morrison & Karkalas, 1990; Alexander, 1995), and is capable of forming single or
double helices (Ahuja et al., 2013). Amylopectin is a highly branched glucan polymer with α-
1,6 glycosidic linkages interspersed along the linear polymer and has a degree of polymerization
ranging from 105-107 glucose units (Myers et al., 2000). Amylopectin is highly digestible
because its branched structure makes it readily able to gelatinize, a process that breaks down the
intermolecular bonds of starch molecules in the presence of water and heat, allowing the sites to
engage more water. Conversely, amylose is less digestible because its linear structure allows
2
adjacent molecules to join by hydrogen bonding which reduces their ability to gelatinize (Ahuja
et al., 2013).
Starch is classified as readily digestible starch, slowly digestible starch or resistant starch (RS)
based on in vitro enzymatic hydrolysis which is assumed to represent the rate of glucose release
and its absorption in the gastrointestinal tract (Englyst et al., 1992a). Readily digestible starch
and slowly digestible starch represent the amount of starch digested in vitro within 20 minutes
and between 20-100 minutes respectively while RS is not digested in vitro within 120 minutes
and is assumed to pass into the large intestine undigested (Englyst et al., 1992a). RS in
particular has garnered much interest from leading research scientists due to its diverse
properties, numerous food and beverage applications, commercial significance and perceived
health benefits (Shi & Maningat, 2013).
The following paragraphs provide a brief rationale for this project as well as the research
questions, research gap and implications. In terms of the rationale, a large body of evidence
suggests that dietary fiber consumption is essential for optimal health and well-being (Jenkins et
al., 2012; Murphy et al. 2012; Sun et al., 2010; Yao et al., 2014). As a result, there is great
interest in the food industry to produce products enriched with dietary fiber to promote health
(Agriculture and Agri-Food Canada, 2014). Since whole grains are a good source of dietary
fiber (Health Canada, 2008; Office of Disease Prevention and Health Promotion, 2011),
everyday processed foods were enriched with whole grains to increase dietary fiber content
(Agriculture and Agri-Food Canada, 2014). The problem with this was that the addition of
whole grains altered the sensory properties of processed foods (Arvola et al., 2007; Bakke &
Vickers, 2007) and the food industry diverted its efforts into using other high-fiber food
ingredients with minimal effect on their sensory properties. An example of such an ingredient is
RS (Fuentes-Zaragoza et al., 2010; Korus et al., 2009). RSs are a good source of dietary fiber
3
with neutral characteristics and have no effect on the sensory properties of processed foods.
Therefore, RSs are now widely used as ingredients to increase the fiber content of products
(Food Innovation, 2014). Having said that, accurate dietary fiber labeling of RS products is
essential. The dietary fiber content of RS products is usually determined in vitro (McCleary &
Rossiter, 2004); however, this may not reflect true in vivo digestibility of RS (Shukri et al., 2013)
and might affect accurate dietary fiber labeling of these products.
This project is composed of two studies aimed to determine the available carbohydrate and
dietary fiber content of commercial RS products currently used in the food industry as novel
food ingredient sources of dietary fiber. This project aimed to answer two main questions:
1) Do in vitro methods accurately reflect true in vivo digestibility of RS?
2) How much dietary fiber, in the form of RS, do commercial RS products contain?
In vivo digestibility of RS is mainly assessed by the breath hydrogen method (Jenkins et al.,
1998) or the ileostomy model (Englyst et al., 1992a), the former of which possesses a number of
limitations and the latter of which is becoming increasingly difficult to implement. The literature
lacks in the description of a new in vivo method that can be used for determining starch
digestibility; hence, this project aimed at developing a new in vivo method used to estimate the
available carbohydrate content and calculate the dietary fiber content of novel RS products from
blood glucose responses (Chapter 4). There is currently no evidence in the literature to suggest
that the available carbohydrate content of RSs has been estimated using blood glucose responses;
thus, this project is the first to describe such a method.
The implications of this project lie mainly in the importance of accurate dietary fiber labeling of
novel RS products. This is not only important for RS as food ingredient sources, but also
important for the products that incorporate such ingredients. It is crucial to assure that the
4
requirements for products made with a “high in fiber” claim are satisfied and that what is
purchased from these products delivers the promised claims to consumers.
Chapter 2 Literature Review
2.0: Literature Review
2.1: Resistant Starch
2.1.1: Coining and Defining of the Term ‘Resistant Starch’
Dietary fiber is divided into three types: soluble, insoluble and RS (About Resistant Starch,
2012). RS is the portion of starch which is not hydrolyzed in vitro until after 120 minutes have
elapsed and is assumed to pass into the large intestine undigested where it acts as a prebiotic
(Englyst et al.,1992a).
Englyst et al. (1982) observed that the treatment that a food receives prior to analysis alters the
properties of its polysaccharides. After boiling potatoes for 20 minutes, a sample was kept for 2
hours at 0˚C then frozen in solid carbon dioxide – methanol and kept for an additional 72 hours
at -25˚C while another sample was kept for 72 hours at -25˚C. In both samples, non-cellulose
polysaccharide glucose and cellulose increased in comparison to the control. Hence, cooking
and freezing allowed starch to gelatinize then retrograde to form a material resistant to α-
amylase hydrolysis which was responsible for the increase in non-cellulose polysaccharide
glucose and cellulose values; and the researchers used the term ‘resistant starch’ to describe this
resistant fraction. Using ileostomy patients as a model for starch digestibility, Englyst and
Cummings (1985) were able to confirm that the resistant fraction of starch in oats, cornflakes and
white bread was not only resistant to digestive enzymes in vitro, but also in vivo in the human
5
small intestine where it was recovered, undigested, in ileal effluent. For oats, a trace amount of
RS was fed (determined in vitro by Englyst et al. (1982) method) and 0.04 g was recovered in
ileal effluent. For cornflakes, 2.99 g of RS was fed and 1.78±0.32 g was recovered and for white
bread, 1.07 g of RS was fed and 0.89±0.19 g was recovered. Therefore, the in vitro and in vivo
values were somewhat similar but not equal.
Similar studies on ileostomy patients were conducted by Englyst and Cummings (1986; 1987)
with bananas and potatoes respectively that lead to a new observation which redefined RS. The
researchers observed that starch that is trapped within whole plant cells or within the food matrix
and some starch granules that have not been fully gelatinized were hydrolyzed very slowly by α-
amylase; and therefore are not digested in the small intestine. The term RS was thus redefined to
“starch and starch-degradation products that, on average, reach the human large intestine”
(Englyst & Cummings, 1987).
RS was officially defined by Asp (1992) as “the sum of starch and products of starch degradation
that are not absorbed in the small intestines of healthy individuals.”
2.1.2: Formation of Resistant Starch in High-Amylose Maize Starch
An increasing interest in the RS of high-amylose maize starch (HAMS) has been observed in the
past two decades due to consumer awareness of the benefits of healthy food components such as
fiber (Sajilata et al., 2006). The RS residues of HAMS consist of mainly semi-crystalline
structures with the B-type polymorph and possess long-chain double helices of amylose which
are present in blocks; hence, preventing the starch granule from swelling at 95-100ºC and
protecting the bulk of amylose and a small portion of amylopectin from enzymatic hydrolysis.
RS residues consist of large molecules with a degree of polymerization of 840-951 and small
molecules with a degree of polymerization of 59-74 (Jiang et al., 2010b). Although HAMS
6
contain minimal lipids (0.2-0.7%), the lipids present in the starch granules also reduce the
enzyme digestibility of the starch at 95-100ºC. After the lipids are removed from the starch
granules, the RS content of HAMS is reduced from 10.6-43.4% to 9.0-28.9% (Jiang et al.,
2010b).
2.1.3: Classification of Resistant Starch
RS is classified into 5 different types based on botanical source and processing (Maningat &
Seib, 2013). This classification is significant because it classifies RSs in light of the source of
enzymatic resistance (Evans, 2013). For RS1, the source of enzymatic resistance is the location
of the starch granules and for RS2, RS3 and RS5, enzymatic resistance is due to the physical
structure of the starch molecules. As for RS4, enzymatic resistance is due to chemical
modification of the starch granules (Evans, 2013). A description of RS classification is given
below (Brown et al., 1995; Englyst et al., 1992a; Nugent, 2005):
1) RS1: is physically-entrapped starch present in whole or coarsely-ground grains and
pulses where the starch granules are encapsulated within a cell wall so that amylase
enzymes are prevented or delayed from having access to the glycosidic bonds
2) RS2: is raw granular starch found in uncooked potatoes, green bananas and high-
amylose corn with intrinsic resistance to digestion due to its crystalline structure
3) RS3: Retrograded or recrystallized starch such as in cooked and cooled potatoes
4) RS4: Chemically-modified starch typically through esterification, cross-linking or
transglycosylation (Robertson, 2012) where the D-glucan chains hinder enzymatic
hydrolysis and make neighboring glycosidic bonds resistant to degradation
5) RS5: amylose-lipid complexes commonly found in native starch granules and
processed starch and which have shown to impede granule swelling during heating in
7
excess water; hence, reducing enzyme accessibility to hydrolyze the starch granules
(Hasjim et al., 2013)
2.1.4: Health Benefits of Resistant Starch
Incorporation of RS into foods has been shown to reduce postprandial glycemic and insulinemic
responses, to enhance insulin sensitivity, to increase satiety and reduce food intake, to produce
short-chain fatty acids (SCFA) through colonic fermentation and to impact gut microbiota
composition. However, the role of RS in colon cancer risk is studied to a limited extent and is
not very well understood.
In terms of glycemic responses, RS4 potato starches when consumed as a drink elicited a
significantly decreased glycemic response compared to dextrose (control) (Haub et al., 2012);
however no differences in glycemic responses and no satiety effect was present when the same
doses of RS4 potato starches were added to dextrose then compared to dextrose alone.
Furthermore, nutritional bars formulated with 34% of phosphorylated cross-linked RS4 wheat
starch displayed attenuated postprandial blood glucose and insulin levels when compared to a
control glucose drink and to another nutritional bar in which the same RS4 wheat starch was
replaced by an equivalent amount of puffed wheat. These results were noteworthy as the bars
also contained high levels of glycemic carbohydrates (11% brown sugar and 20% corn syrup)
(Al-Tamimi et al., 2010).
A human feeding study conducted by Hasjim et al. (2010) compared control white bread and
bread containing 60% RS5 on a dry weight basis and the results showed that the total
incremental area under the blood glucose response curve (iAUC) of the postprandial glycemic
response was reduced to 55% after ingestion of the RS5 bread compared with the control
(100%). Similarly, Behall et al. (2006) reported that the iAUC was reduced when the RS2
8
content of test muffins was increased to 38% and the greatest iAUC reduction (33%) occurred
after meals containing both high ß-glucan and high RS2. Bodinham et al., 2010, Raben et al.
1994 and Robertson et al., 2005 reported significantly lower postprandial insulin responses
following the RS supplement and feeding RS to human subjects has been shown to enhance
insulin sensitivity when compared with other treatments (Behall et al., 2006; Johnston et al.,
2010; Maki et al., 2012; Robertson et al., 2003; Robertson et al., 2005).
As for satiety, Willis et al. (2009) reported that RS and corn bran had the most impact on satiety,
whereas polydextrose had little effect and behaved like the low-fiber treatment and these results
confirmed the earlier work of Nilsson et al. (2008) in healthy human subjects. Bodinham et al.,
2010 and Anderson et al., 2010 observed a significantly lower food intake in human subjects
after a RS meal and Bodinham et al., 2010 suggested that the consumption of 48g RS daily may
be useful in the management of metabolic syndrome and in controlling appetite.
Many of the effects of RS on colonic function are exerted through the action of SCFAs, namely
butyrate, acetate and propionate, produced by bacterial fermentation of RS in the colon. Studies
by Phillips et al. (1995), Cummings et al. (1996) and Muir et al. (2004) have shown an increase
in the fecal concentrations of SCFA after the ingestion of a RS supplement. SCFA are the
preferred fuel of the cells lining the colon (colonocytes) and they have been shown to increase
colonic blood flow, lower luminal pH and help prevent the development of abnormal colonic cell
populations such as adenomas (Topping & Clifton, 2001).
Martinez et al. (2010) studied the impact of RS2 high-amylose corn starch and RS4 wheat starch
on gastrointestinal microbiota composition in human volunteers. The volunteers were fed 100 g
crackers with either 33 g RS2 or RS4 starch. Pyrosequencing analysis of fecal DNA
demonstrated that RS4 crackers significantly decreased Firmicutes and increased Bacteroidetes
9
and Actinobacteria. These changes were due to a decrease in the family Ruminococcaceae and
increases in the genera Parabacteroidetes and Bifidobacterium. The total cell numbers of
bifidobacteria increased more than three-folds with RS4 crackers while RS2 crackers doubled the
cell numbers. This is further confirmed by the fact that RS is a prebiotic and assists in promoting
the colonization of beneficial microbiota in the colon (Topping et al., 2003).
There are a limited number of reviews and meta-analyses that discuss the role of RS in the
reduction of colon cancer risk. However, a number of randomized controlled trials have been
conducted in the past years with inconsistent results regarding the protective role of RS on colon
cancer. Dronamraju et al. (2009) fed 13 g of RS per day for 4 weeks to 24 pre-operative
colorectal cancer patients and reported a lesser proportion of mitotic cells in the upper half of the
colon crypts than 10 similar subjects who were given amylopectin as a placebo. The reduction in
this premalignancy marker signifies a potential role of RS in protecting against colon cancer.
Similarly, Burn et al. (2008) fed 13g of the same RS per day for 29 months to subjects
previously treated for hereditary nonpolyposis colon cancer and reported no difference between
incidence of colon neoplasms between this group and similar subjects fed amylopectin as a
placebo.
Grubben et al. (2001) fed 28 g of RS or 45 g of highly digestible maltodextrin per day for 4
weeks to 23 patients after colon adenomectomy. They reported a significant decrease in fecal
bile acid concentrations in subjects given RS but SCFAs and colorectal cell proliferation did not
differ between the treatments. This suggests that a decrease in fecal bile acid concentrations is a
possible antineoplastic effect of RS since it may be considered as a colon tumor promoter
(Nagengast et al., 1995). Conversely, van Gorkom et al. (2002) fed 19 g of RS per day for 2
months to 28 subjects with sporadic adenoma and reported that RS did not alter epithelial cell
proliferation in colorectal biopsies compared with 28 controls given ordinary starch as a placebo.
10
While it is true that inconsistencies appear regarding the protective role of RS in colon cancer,
the suppression of colon cancer proliferation (Dronamraju et al., 2009) and a decrease in fecal
bile acid concentrations (Grubben et al., 2001) indicate that RS may have an antineoplastic effect
in the colon. However, direct evidence for the role of RS in the protection against colon cancer
is still not available and should warrant further investigation.
2.2: Measurement of Resistant Starch
2.2.1: In Vitro Methods
2.2.1.1: Background
As early as the 1960s, Leach & Schoch (1961) and Sandstedt et al., (1962) observed that raw
high-amylose corn starch was resistant to in vitro digestion with amylases and Fuwa (1977)
witnessed the same phenomenon during his work on raw banana starch and amylase. Englyst et
al., (1982) coined the term ‘resistant starch’ for retrograded starch resistant to α-amylase
digestion in cooked and cooled potatoes during his work on the determination of non-starch
polysaccharides by gas-liquid chromatography.
The in vitro methods for the determination of RS discussed below give inconsistent results when
reporting RS content or percentage in various samples. Some in vitro methods either marginally
or greatly overestimate RS content while others either marginally or greatly underestimate RS
content when compared to in vivo results from ileostomy patients (Tables 2 & 4). This supports
the rationale (section 3.1) of this project which describes that in vitro methods do not accurately
reflect in vivo digestibility of RS.
11
2.2.1.2: Initial Methods Developed for the Determination of Resistant Starch
In the late 1980s, two methods by Berry (1986) and Björck et al. (1986) were developed to
determine RS content. Berry (1986) modified the in vitro method of Englyst et al. (1982) to
represent physiological conditions of digestion more closely. These modifications included
omitting the initial heating step at 100°C, performing incubations at 37°C, and reemploying the
use of pancreatic α-amylase and pullulanase enzymes. Using this new method, the measured RS
contents of samples were much higher than those previously obtained suggesting that Englyst et
al.’s (1982) initial method underestimated the amount of RS content in the samples. The method
developed by Björck et al. (1986) was similar to that of Berry (1986) except that there was an
initial heating step in which the starch samples were boiled for 15 minutes after the addition of α-
amylase.
During the 1992 European Research Program, EURESTA, several new or modified methods for
the measurement of RS were developed. The Champ (1992) method was based on modifications
to the method of Berry (1986) where RS determinations were performed on a pellet and gave a
direct measurement of RS. The Champ method increased the sample size from 10 mg to 100 mg
and the sample was digested only with pancreatic α-amylase. Further, the pH used by the
Champ method (pH 6.9) was higher than that used by the methods of Englyst et al. (1982;
1992a) and Berry (1986) (pH 5.2). The author concluded that the modified Berry (1986) method
yielded 48% RS on an “as is” basis while the Björck et al. (1986) method failed to detect any RS
because of the initial boiling step.
The Englyst et al. (1992a) method was one of the most widely used methods for the
determination of RS because Englyst et al. (1982) coined the term “resistant starch” during their
work on the determination of non-starch polysaccharides in cooked and cooled potatoes. This
12
method was a modification of the Englyst et al. (1982) method in which various types of starch
were determined by controlled enzymatic hydrolysis with measurement of the released glucose
using glucose oxidase-peroxidase-4-aminoantipyrine (GOD-PAP). This method chemically
defined RS as the starch that is not hydrolyzed after incubation with pancreatin and
amyloglucosidase after 120 minutes at 37ºC (Englyst et al., 1992a).
Muir & O’Dea (1992; 1993) developed a procedure for the determination of RS in which
samples were chewed, treated with pepsin and then with a mixture of pancreatic α-amylase and
amyloglucosidase in a shaking water bath at pH 5.0, 37ºC for 15 hours. The RS containing pellet
was recovered by centrifugation and washed with acetate buffer after which the RS was digested
by a combination of heat, dimethyl sulfoxide and thermostable α-amylase treatments. The
results indicated that different food-processing techniques produced different amounts of RS and
the amount of RS in food decreased with increased chewing; thus, the authors concluded: “This
assay may be useful in predicting which foods and processing techniques result in high amounts
of starch escaping digestion in the small intestine” (Muir & O’Dea, 1992).
In the mid-1990s, two methods were developed for the determination of RS content, the Faisant
et al. (1995b) method and Goñi et al. (1996) method. The Faisant et al. (1995b) method is a
modification of the Berry (1986) method in which RS is quantified directly in the residue which
is isolated by 80% ethanolic precipitation. As for the Goñi et al. (1996) method, the sample is
incubated with pepsin and then hydrolyzed by pancreatic α-amylase with no alcoholic
precipitation. The water-insoluble residue is collected by centrifugation, as opposed to alcoholic
precipitation, and this residue contains the RS.
The Åkerberg et al. (1998) and Champ et al. (1999a) were two methods that were developed in
the late 1990s for the determination of RS content. The Åkerberg et al. (1998) method included
13
chewing for the sample pre-treatment and RS is recovered by ethanol precipitation and filtration.
The Champ et al. (1999a) method is a modification of the Berry (1986) method except that it did
not include a protein hydrolysis step which may be a problem with substrates containing a high
percentage of protein; although such a problem has not been observed by the authors.
Of the initial in vitro methods discussed in this section, the Björck et al. (1986), Englyst et al.
(1992), Muir & O’Dea (1992; 1993), Goñi et al. (1996), Åkerberg et al. (1998), and Champ et al.
(1999a) methods were considered the main in vitro methods widely used for the determination of
RS content before the development of the Association of Official Analytical Chemists (AOAC)
Official Methods for RS determination. Table 1 below compares these methods based on the
following parameters: sample size, sample pre-treatment, dispersion of RS, RS hydrolysis,
glucose determination and validation (Champ et al., 2003).
In terms of validation, Englyst et al. (1992), Muir & O’Dea (1992; 1993), Goñi et al. (1996),
Åkerberg et al. (1998) and Champ et al. (1999a) validated their in vitro methods based on in vivo
data from ileostomy patients while Björck et al. (1986) validated their in vitro method based on
in vivo data obtained from antibiotic-treated rats.
14
Table 1 – Comparison of Main Initial In Vitro Methods Used to Determine Resistant Starch Content:
Björck et al. (1986)
Englyst et al. (1992a)
Muir & O’Dea (1992; 1993)
Goñi et al. (1996)
Åkerberg et al. (1998)
Champ et al. (1999a)
Sample Size 100 mg of fiber residue
0.8-4.0 g depending on
water and starch content of the
sample
≈0.1 g carbohydrate basis
100 mg of dry sample
1 g of total starch basis
50 mg of total starch basis
Sample Pre-treatment
- Minced (9mm Øa holes)
Chewing Dry samples milled (Ø a≤1 mm) and fresh samples homogenized
Chewing (15 times in 15 seconds)
Minced (9mm Øa holes)
Dispersion of Resistant Starch
Boiling (20 minutes) +
2 M KOHb (RTc for 30 minutes)
No Boiling (20 minutes) +
2 M KOHb (0°C for 30 minutes)
2 M KOHb (RTc for 30
minutes)
2 M KOHb (for 30 minutes)
Boiling (20 minutes) +
2 M KOHb (0°C for 30 minutes)
Resistant Starch Hydrolysis
Amyloglucosidase (pH 4.75 at 60°C for 30 minutes)
No Amyloglucosidase Amyloglucosidase (pH 4.75 at 60°C for 30 minutes)
Termamyl + Amyloglucosidase
Amyloglucosidase (14 units/ml at
70°C for 30 minutes + 100°C for 10 minutes)
Glucose Determination
Enzymatic, GOD-PODd
Enzymatic, GOD-PAPe
Enzymatic, GOD-PAPe
Enzymatic, GOD-PAPe
Enzymatic, GOD-PODd
Enzymatic, GOD-PAPe
Validation In vivo data obtained with
antibiotic-treated rats
In vivo ileostomy data
In vivo ileostomy data
In vivo ileostomy data
(from literature)
In vivo ileostomy data
(from literature)
In vivo ileostomy and intubation data
Table adapted from Champ et al. (2003) c RT – Room Temperature a Ø – Diameter d GOD-POD – glucose oxidase peroxidase b 2 M KOH – 2 Molar potassium hydroxide e GOD-PAP – glucose oxidase-peroxidase-4-aminoantipyrine
15
2.2.1.3: Development of AOAC Official Method 2002.02
While significant steps were made in the development of in vitro methods for the measurement
of RS during the 1990s, none of these methods were successfully subjected to interlaboratory
evaluation (McCleary, 2013). This prompted the development of a new method that gave values
in line with those obtained from ileostomy patients and that could survive the rigors of
interlaboratory evaluation.
A survey of scientists initiated in 1993 by Lee & Prosky (1995) showed that 80% favored the
inclusion of RS in the definition of dietary fiber, which led to the development of AOAC Official
Method 2002.02 for the measurement of RS (McCleary & Monaghan, 2002). The incubation
conditions, level of pancreatic α-amylase and amyloglucosidase used, protease pretreatment and
procedure for recovery of RS, among others, were adjusted to optimize an assay for the direct
measurement of RS. This method was subjected to evaluation under the auspices of AOAC
International and American Association of Cereal Chemists International to determine its
interlaboratory performance. Thirty-seven laboratories tested eight pairs of blind duplicate
starch or plant material samples with RS values between 0.6-64% fresh weight basis (McCleary
et al., 2002). Two statistical measurements of precision, the relative standard deviation
repeatability (RSDr) and relative standard deviation reproducibility (RSDR) were used in order to
validate the method. The RSDr expresses the measurement results under repeatability conditions
where independent results are obtained with the same method on identical test items using the
same laboratory, operator and equipment within short intervals of time (Horie et al., 2008). The
ideal RSDr value is 4% and the reported RSDr values for this method ranged from 1.97-4.2%
(McCleary & Monaghan, 2002). The RSDR expresses the measurement results under
reproducibility conditions where results are obtained with the same method on identical test
items using different laboratories, operators and equipment (Horie et al., 2008). The ideal RSDR
16
value is 6% and the reported RSDR values for this method ranged from 4.58-10.9% (McCleary et
al., 2002). The method was not suitable for samples with less than 1% RS because for such
samples, the RSDr and RSDR values were very high and considered unacceptable; therefore, this
method was applicable to samples containing 2-64% RS content. On the basis of the previous
evaluation, the method was accepted as AOAC Official Method 2002.02 and American
Association of Cereal Chemists Recommended Method 32-40.01 (McCleary et al., 2002).
Table 2 – Comparison of Resistant Starch Values Obtained Using Several In Vitro Methods and In Vivo Results: Resistant Starch (in vitro results, %)
Source of Starch
Englysta Faisantb Champc McCleary & Monaghan
(AOAC 2002.02)d
Goñie Resistant Starch (in vivo
results, %)f
Potato starch (native)
66.5 83.0 77.7 77.0 – 78.8g
Amylomaize starch
(native)
71.4 72.2 52.8 51.7 – 50.3g
Amylomaize starch
(retrograded)
30.5 36.4 29.6 42.0 37.8 30.1g
Bean flakes 10.6 12.4 11.2 14.3 15.3 9-10.9h Corn flakes 3.9 4.9 4.3 4.0 4.7 3.1-5.0i
ActiStar 63j – 57j 58.0 57j 59j Table adapted from McCleary & Monaghan (2002) a Englyst et al., 1992a b Faisant et al., 1995a c Champ et al., 1999a d McCleary & Monaghan, 2002 e Goñi et al., 1996 (assuming a starch content of 40% for bean flakes and 70% for corn flakes) f In vivo results from ileostomy patients g Champ et al., 2003 (personal communication between Champ et al., 2003 and Langkilde, A.M., Andersson, H. and Bouns, F.) h Schweizer et al., 1988 i Muir & O’Dea (1993) and Englyst et al., 1992a j Results provided by Bernd Kettlitz, Cerestar (Vilvoorde, Belgium). Englyst et al., 1992a value produced by Englyst Carbohydrate Services, Champ value at INRA (Nantes, France) and Goñi value at Cerestar (Vilvoorde, Belgium) In Table 2, McCleary & Monaghan (2002) reported the values for 4 different in vitro methods
for the determination of RS, as a percentage of total starch, and compared them to RS
17
percentages obtained in vivo from ileostomy patients. For amylomaize starch (native), the RS
percentage of the sample obtained in vivo was 50.3% as opposed to 72.2% from the Faisant et al.
(1995b) method, 71.4% from the Englyst et al. (1992a) method, 52.8% from the Champ (1992)
method and 51.7% from the McCleary & Monaghan (2002) AOAC 2002.02 method. Thus, the
Faisant et al. (1995b) and Englyst et al. (1992a) methods greatly overestimated RS percentage in
amylomaize starch (native) while the Champ (1992) and McCleary & Monaghan (2002) AOAC
2002.02 methods marginally overestimated RS content. In addition, the inconsistency is not only
apparent between the in vitro methods but also within the same in vitro method when
determining RS content from different sources of starch. For example: when comparing the
Englyst et al., (1992a) RS percentages with in vivo RS percentages, this method underestimated
RS by 12.3% for potato starch (native), overestimated RS by 21.1% for amylomaize starch
(native) and closely resembled the in vivo percentage for amylomaize starch (retrograded),
30.5% in vitro and 30.1% in vivo. Overall, AOAC 2002.02 method developed by McCleary &
Monaghan (2002) seems to be the in vitro method whose values closely resemble RS recoveries
in vivo.
2.2.1.4: AOAC 991.43 Method vs. AOAC 2002.02 Method
Lee et al. (1992) developed the AOAC Official Method 991.43 for the determination of total,
soluble, and insoluble dietary fiber in foods, an enzymatic-gravimetric based method. This
method underestimated the total dietary fiber (TDF) content of the majority of food materials
because it incompletely analyzed RS content (McCleary & Rossiter, 2004). This was
demonstrated when McCleary & Rossiter (2004) compared the TDF percentage determined by
AOAC 991.43 and a modified version of AOAC 991.43 where dimethyl sulfoxide was added to
solubilize RS. RS was determined by AOAC 2002.02 method (Table 3) and expressed as a
percentage of total starch. For Hylon VII, RS percentage determined by AOAC 2002.02 was
18
53.7% but was not reflected in the TDF percentages because AOAC 991.43 method determined
25.9% TDF and the modified AOAC 991.43 method determined 1.0% TDF. This comparison
was also true for ActiStar, native potato starch and CrystaLean. As for Hi-maize 1043, Hi-maize
bread, rye crispbread and kidney beans, the results were inconsistent. For Hi-maize 1043 and
Hi-maize bread, AOAC 991.43 percentages were higher than the AOAC 2002.02 percentages;
however, the modified AOAC 991.43 percentages were lower than the AOAC 2002.02
percentages. For rye crispbread and kidney beans, both AOAC 991.43 and modified AOAC
991.43 had TDF percentages higher than AOAC 2002.02 method. Hence, the results of AOAC
991.43 and modified AOAC 991.43 methods in Table 3 suggest that these methods are
unsuitable for the determination of RS content in RS containing food materials.
Table 3 – Total Dietary Fiber and Resistant Starch Content of a Range of Samples:
Total Dietary Fiber (%) Resistant Starch (%) Sample AOAC 991.43 AOAC 991.43a AOAC 2002.02
Hylon VII 25.9 1.0 53.7 ActiStar <0.1 < 0.1 58.0
Native potato starch <0.1 < 0.1 78.1 CrystaLean 34.0 0.3 40.9
Hi-maize 1043 54.5 0.5 45.7 Hi-maize bread 9.2 3.5 5.1 Rye crispbread 15.0 13.6 1.2 Kidney beans 21.5 5.4 5.3
Table adapted from McCleary & Rossiter (2004) a Modification to AOAC 991.43 was made in which the sample was first cooked with dimethyl sulfoxide to solubilize resistant starch 2.2.1.5: Incorporation of Resistant Starch into Dietary Fiber
Measurements and the Development of AOAC 2009.01
Hipsley (1953) coined the term ‘dietary fiber’ to describe the non-digestible constituents of
plants that make up the plant cell wall such as cellulose, hemicellulose and lignin. Twenty-three
years later, Trowell et al. (1976), broadened the definition into a physiological one based on
edibility and resistance to digestion in the human small intestine. This new definition included
indigestible polysaccharides such as gums, modified celluloses, mucilages and pectin and non-
19
digestible oligosaccharides (NDO). In 1990, the British Nutrition Foundation Task Force
proposed that complex carbohydrates be defined as the sum of starches and non-starch
polysacchardies (British Nutrition Foundation, 1990). In this proposal, starches are considered
avCHO and all other fractions are considered unavailable carbohydrates. This proposal was
dismissed because this classification underestimated the importance of the physiological function
of RS; simply, “the contribution of RS to the body’s physiological functions is much too
significant to be buried under the general term available carbohydrates” (Schweizer et al., 1990).
The first official method for the measurement of dietary fiber was the AOAC Official Method
985.29 – ‘Total dietary fibre in foods; enzymatic-gravimetric method’ developed by Prosky et al.
(1985). This method was later extended by Lee et al. (1992) to allow measurements of TDF,
soluble dietary fiber and insoluble dietary fiber in foods and was known as AOAC Official
Method 991.43 (Figure 1). The recurrent problem with these methods is the partial measurement
of RS content in some food materials; subsequently, underestimating RS content in foods (Tables
3 & 4).
The fact that a single method to measure all dietary fiber components (including RS) is needed,
has been known for some time. While it is possible to measure many individual fiber
components with specific and non-specific methods, TDF cannot simply be calculated by adding
the values for these specific components to the determined value of high molecular weight
dietary fiber as measured by AOAC 991.43. Since AOAC 991.43 also measures some RS in
food materials, summation leads to “double counting” of this material and an eventual
overestimation of RS content (McCleary, 2013). To resolve this issue, an integrated method for
the measurement of total dietary fiber (ITDF), AOAC 2009.01 (Figure 2), was published by
McCleary (2007) and it allowed the accurate measurement of total high molecular weight dietary
fiber, which includes insoluble dietary fiber, including RS, and higher molecular weight soluble
20
dietary fiber. For most food and ingredient samples analyzed, the RS values obtained with
AOAC 2009.01 were higher than those obtained from AOAC 991.43. This method was
successfully subjected to interlaboratory evaluation and accepted as AOAC Method 2009.01
(McCleary et al., 2009) and has been recommended for use in dietary fiber labeling as it takes
into account all dietary fiber fractions including RS (Figure 2).
Table 4 – Comparison of Resistant Starch Values Obtained Using Several In Vitro Methods (Including AOAC 2009.02 Method) and In Vivo Results: Resistant Starch (in vitro results, %)
Source of Starch
Englyst AOAC 991.43
AOAC 2002.02
AOAC 2009.01
Resistant Starch (in vivo
results, %)c Raw potato
starch 66.5a 0.9b 64.9b 56.8b 67.9d
High-amylose corn
starch
71.4a 25.6b 50.0b 49.3b 43.7d
Corn flakes 3.9a 3.3e 2.2b 2.4b 3.1-5.0e Raw green
banana 54.2b 7.5b 51b 38b 55.3d
Table adapted from Shukri et al. (2013) a Englyst et al., 1992a b McCleary (2007) c In vivo results from ileostomy patients d Langkilde & Andersson (1995) e Muir & O’Dea (1992) In Table 4, Shukri et al. (2013) reported the values for 4 different in vitro methods for the
determination of RS, as a percentage of total starch, and compared them to RS percentages
obtained in vivo from ileostomy patients. When comparing RS percentage determined in vitro
by AOAC 2009.01 method with in vivo RS percentages, this method underestimated RS content
for raw potato starch and raw green banana and overestimated RS content in high-amylose corn
starch. Out of the four in vitro methods, AOAC 991.43 was the method that greatly
underestimated RS content when compared to in vivo results. Overall, Table 4 shows that no one
in vitro method accurately reflects in vivo digestibility of RS and results remain inconsistent
within the same method when comparing different sources of RS. However, AOAC 2009.01 as
21
a general method for the determination of TDF and AOAC 2002.02 as a specific method for the
determination of RS content seem to be the methods that closely reflect RS content in vivo when
compared to other in vitro methods.
2.2.2: In Vivo Methods
The two main in vivo methods for determining starch digestibility include the breath hydrogen
test and the ileostomy model. The breath hydrogen test exhibits a number of limitations and the
ileostomy model is generally accepted as the “golden standard” for the measurement of the
undigested fraction of carbohydrates.
Figure 2 – Dietary fiber components measured by AOAC Official Method 2009.01: This method takes into account the complete fraction of RS as a component of dietary fiber. * FOS – Fructo-oligosaccharides Source: Medallion Labs. (2014). Dietary Fiber. Retrieved December 3, 2014 from http://www.medallionlabs.com/TestOfferings/Fiber.aspx Used with permission from Medallion Labs.
Figure 1 – Dietary fiber components measured by AOAC Official Method 991.43: This method underestimates RS content because it only measures a fraction of it. * FOS – Fructo-oligosaccharides Source: Megazyme. (2014). Measurement of Dietary Fiber. Retrieved December 3, 2014 from http://www.megazyme.com/resources/dietary-fiber/measurement-of-dietary-fiber Used with permission from Megazyme.
22
2.2.2.1: Breath Hydrogen Test
Undigested carbohydrates in the small intestine are fermented by the gut microbiota in the colon
and the products of carbohydrate fermentation are methane, hydrogen gas, carbon dioxide and
SCFAs (Hijova & Chmelarova, 2007). Since hydrogen gas is one of the four products of
carbohydrate fermentation, a large number of studies have used the breath hydrogen test to
assess the undigested fraction of carbohydrates. Breath hydrogen tests are simple, convenient,
inexpensive, non-invasive and do not disturb the normal physiological functions of the intestine
(Rumessen, 1992).
In the 1990s, several studies were conducted on RS and the breath hydrogen test. Hylla et al.
(1998) and Muir et al. (1995b) reported that the AUC of breath hydrogen was significantly
higher during the high-RS diet compared to the low-RS diet and van Munster et al. (1994)
reported that the mean rise in breath hydrogen relative to placebo was 35% (p=0.03) for all
subjects and 60% for the eight non-methane producing subjects (p=0.02) in the study. Wolever
et al. (1986) reported that direct measurement of the available carbohydrate in ileal effluent after
the consumption of three test meals by three ileostomates gave values similar to those
determined by the breath hydrogen method (white bread 10% from effluent and 11% from breath
hydrogen, wholemeal bread 8% from both effluent and breath hydrogen and red lentils 22% from
effluent and 18% from breath hydrogen). However, a study by Jenkins et al. (1998) reported that
commercial RS2 and RS3 supplements from high-amylose corn starch did not alter breath
hydrogen and methane when compared to the control treatment (low-fiber supplement). Since
this project involves commercial RS products, the results from the Jenkins et al. (1998) study
directed the methodology of this project to use the ileostomy model, instead of breath hydrogen,
as an in vivo model of determining starch digestibility.
23
In addition to the inconsistency in the breath hydrogen results mentioned above, other limitations
are associated with this in vivo method including:
1) 20% of subjects are non-producers of hydrogen gas (Gilat et al., 1978; Saltzberg et al.,
1988) and 5% are low-producers (Bond & Levitt, 1977; Vogelsang et al., 1988) and this
largely depends on the type of microbiota dominating the gut (i.e.: methanogens) (Gilat et
al., 1978);
2) Acidic colon microclimate (Perman et al., 1981; Vogelsang et al., 1988) antibiotic use
(Gilat et al., 1978; Murphy & Calloway, 1972), enemas and colonoscopy preparation
(Gilat et al., 1978) have all shown to reduce hydrogen gas production;
3) The nature of the indigestible carbohydrate such as decreased fiber particle size (Ehle et
al., 1982; Heller et al., 1980; Robertson, 1988), increased hydration (Millard & Chesson,
1984), increased fiber porosity and solubility and decreased viscosity (Tomlin et al.,
1986) all increase fermentability; thus, increasing hydrogen gas production; and
4) Prolonged small intestinal transit decreases hydrogen gas production (Read et al., 1985).
In addition, high inter- and intra-individual variation among subjects has been reported due to
great differences in their dietary habits and hydrogen gas production capacity (Rumessen, 1992;
Simrén & Stotzer, 2006).
2.2.2.2: Ileostomy Model
An ileostomy is a surgical opening constructed by diverting the end of the small intestine (ileum)
out onto the abdominal wall. Diseases of the large intestine such as Crohn’s disease, ulcerative
colitis and colorectal cancer may require the colon to be partially or wholly removed and the
ileostomy becomes a necessary measure to eliminate digestive waste. Digestive waste passes out
24
of the ileostomy and is collected in an external pouch that is adhered to the skin. Ileostomies are
usually located above the groin on the right hand side of the abdomen (Ileostomy, 2014).
Sandberg et al. (1981) were the first to use the ileostomy model as an experimental model for the
in vivo determination of dietary fiber. Nine ileostomates were fed wheat bran fiber and their
effluent was collected over a 24-hour period after which it was analyzed for different
polysaccharides using gas-liquid chromatography. Of the added bran, 80-100% of hemicellulose
and 75-100% of cellulose were recovered in ileostomy contents and the authors concluded that:
“…it seems probable that determination of dietary fibre by in vivo digestion in ileostomy patients
comes very close to the theoretical definition of dietary fibre” (Sandberg et al., 1981).
Hans N. Englyst was one of the first scientists to extensively study the digestibility of RS using
the ileostomy model (Englyst & Cummings, 1985; 1986; 1987; Englyst et al., 1996; Silvester et
al., 1995). The procedure entails feeding healthy ileostomy patients RS-containing test meals
and collecting their effluent for a minimum of 10 hours to compare the amount of RS fed vs. the
amount recovered. The Englyst method was first developed in 1982 (Englyst et al., 1982) to
determine non-starch polysaccharides in plant foods. The method was validated on the basis of
ileostomy studies (Englyst & Cummings, 1985; 1986; 1987) and claimed to give results in line
with those from ileostomy patients. The method was modified in 1992 (Englyst et al., 1992a)
with optimization in enzymatic hydrolysis and adaptation to a larger variety of substrates
(Champ et al., 2003). When Englyst redefined RS to include starch-degradation products
(Englyst & Cummings, 1987), the authors mentioned that “in vivo recoveries and in vitro
analysis yielded an average of 0.4% difference in the amount of RS recovered in ileal effluent.”
This is not necessarily true when RS percentage from potato starch, determined by Englyst et al.
(1992a) method, was compared with in vivo RS recoveries from ileostomy patients (Tables 2 &
4) in which the methods underestimated RS content from 1.4-12.3%.
25
A number of studies using the ileostomy model reported the mean fiber or RS recoveries after
the ingestion of starchy foods such as maize, wheat, potatoes, bananas or beans. Englyst et al.
(1996) reported that the mean starch recoveries in ileal effluent ranged from 91-106% of the RS
intakes measured in vitro after the ingestion of a number of starchy foods and over 90% after the
ingestion of bananas (Englyst & Cummings, 1986) and potatoes (Englyst & Cummings, 1987).
Jenkins et al. (1987) reported a similar percentage of a mean fiber recovery value of 91.5±11.3%
from a number of starchy foods and higher values of 105±6% RS recovery from potatoes and
100±3% RS recovery from beans were reported by Schweizer et al. (1990).
A study on oats, cornflakes and white bread by Englyst & Cummings (1985) compared the
amount of RS fed vs. RS recovered from ileostomy patients. For the oats, cornflakes and white
bread respectively, the values were: trace vs 0.04 g, 2.99 g vs. 1.78±0.32 g and 1.07 g vs.
0.89±0.19 g. A similar study by Englyst et al. (1996) reported values for the amount of RS fed
vs. RS recovered in ileal effluent. For the maize-RS and wheat-RS, the RS fed was 8.5 g and the
RS recovered was 8.4 g and 8.8 g respectively. Furthermore, a review by Champ et al. (2003)
recalculated some values produced from studies conducted by Muir & O’Dea (1993) and Muir et
al. (1995a) in order to compare the percentage of RS recovered in vivo with that recovered in
vitro. For the high-RS meal, pearl barley, cornflakes and ground rice, the values were: 37.8% vs.
35.1%, 5.5% vs. 5.5%, 3.1% vs. 2.9% and 0.7% vs. 0.8% respectively (Champ et al., 2003).
Although these studies have reported similar results in vitro vs. in vivo, Table 2 and Table 4
show that this is not always the case when comparing in vitro vs. in vivo results from different
studies for the same source of starch. Nonetheless, the ileostomy model appears to be the
superior form of studying starch digestibility in vivo.
26
2.2.2.2.1: Potential Problems with the Ileostomy Model
The ileostomy model is to-date a superior model for studying starch digestibility; however, due
to improvements in surgical techniques, fewer conventional ileostomies are being created and
ileostomates are becoming harder to find. Thus, other methods for measuring RS in vivo are
needed such as the method described in Chapter 4. In addition, the model is not perfect and
exhibits potential problems in need of further evaluation as outlined by the questions below
(Rumessen, 1992):
1) How much carbohydrate is degraded due to asymptomatic small intestinal
bacterial overgrowth (SIBO) in the terminal part of the ileum?
2) How reliable are determinations of different carbohydrates recovered from the
ileostomy bag where further breakdown is possible due to the presence of amylase and
fermenting bacteria?
3) Is small intestinal absorption or motility normal in patients with Crohn’s
disease or other gastrointestinal tract diseases?
Further breakdown in the ileostomy bag, prolongation of GI-transit, hyperabsorption, SIBO and
consequent fermentation may lead to an underestimation of the malabsorbed fraction such as RS
in ileostomy patients (Rumessen, 1992).
2.3: Resistant Starch as a Food Ingredient
RS is used as a food ingredient to increase fiber content of foods and to deliver perceived health
benefits. RSs types 1-5 from different botanical sources are traditionally white in color, possess
fine particle size and are neutral in flavor. However, they have distinct differences in water-
holding capacity, which affect processing parameters of bakery foods such as biscuits, muffins,
cakes and scones (Woo et al., 2009). Product developers and designers use RS as an ingredient
27
to boost fiber content of food and beverage products for the fulfillment of nutrient and caloric
labeling claims without jeopardizing the taste and texture of their products (Erickson, 2005;
Topping, 2007). Food manufacturers essentially add RS to products in order to increase fiber
content, reduce available carbohydrate content and reduce the caloric count of their products
(Maningat et al., 2005; 2008).
Hi-maize™ was the first RS ingredient to hit the market in 1993 (Brown et al., 1995), only 11
years after the term ‘resistant starch’ was coined and the growth of RS as a food ingredient has
been phenomenal with at least 30 products now being sold and marketed on a worldwide basis
(Maningat & Seib, 2013). Hi-maize™ is rich in RS and is a natural source of dietary fiber that
can be added to various foods without adversely affecting their organoleptic (sense-related)
properties. Hi-maize™ is also able to contribute functionally to foods by increasing moisture
retention, improving expansion after extrusion and providing a barrier film that blocks the uptake
of fats in snack foods (Food Innovation, 2014). Another RS ingredient, PROMITER™ Resistant
Starch, reduced the oil pick-up in fried snacks by 15-25% allowing for a ‘reduced calories’ and a
‘reduced fat’ claim (PROMITOR™ Dietary Fiber, 2011). These results confirmed the earlier
work of Han et al. (2007) in which the RS4 starch produced fried batters with improved
crispiness and hardness and reduced oil absorption during frying.
In order to declare a ‘good source of fiber’ claim, Erickson (2005) formulated an oatmeal cookie
by complete replacement of the flour portion with RS4 starch. The 12.8% RS4 used delivered a
‘good source of fiber’ claim which translated to 3 g of dietary fiber per 30 g serving size of the
cookie. Bustos et al. (2011) added 2.5-10.0% RS2 and RS4 high-amylose corn starch in pasta
and observed a significant decrease in water absorption and swelling index compared to the
control pasta. The authors also reported that as the level of RS4 starch increased in the pasta, the
28
hardness tended to increase but the springiness, cohesiveness and chewiness tended to decrease
(Bustos et al., 2011).
Chemically modified food starches (RS4) are used as additional ingredients to enhance the
processing performance, physical attributes, sensory properties and storage stability of packaged
food products. For example, pregelatinized RS4 is added to a variety of reduced-fat foods,
bakery products, instant mashed potato, sausages, salad dressings, desserts, ice cream, yogurt and
cream fillings for product enhancement (Maningat & Seib, 2013).
2.4: Resistant Starch and Dietary Fiber Labeling
The aim of the nutrition label is to provide the information that a consumer needs in order to
make healthy choices about the foods they eat and it is the responsibility of the food
manufacturer to declare values that fairly represent the true content of the product (Champ et al.,
2003). Health authorities in North America have set the Dietary Reference Intake for dietary
fiber as 38 g/day for males and 25 g/day for females with no specific mention of the amount of
RS that should be consumed on a daily basis (Dietary Reference Intakes, 2005).
In February 2012, the Health Canada Bureau of Nutritional Science Food Directorate of the
Health Products and Food Branch issued a report titled: Policy for Labeling and Advertising of
Dietary Fibre-Containing Food Products (Health Canada, 2012), and the aim of this report was to
make Canada up-to-date with international standards regarding the definition of dietary fiber.
The report discussed a revised definition for dietary fiber and set a new caloric value of 2 kcal (8
kJ)/g to represent the amount of energy yielded from dietary fiber. Within the report, Health
Canada, in consultation with the Canadian Food Inspection Agency (CFIA), proposed a list of
appropriate analytical methods for the determination of TDF and its many components. AOAC
985.29 and 991.43 (among others) were accepted for the partial determination of RS, AOAC
29
2009.01 was accepted for the determination of TDF (including RS content) and AOAC 2002.02
was accepted as a specific method that can accurately determine RS content in its entirety.
However, the CFIA now requires using AOAC 2009.01 for nutrition labeling purposes because
they believe it is effective for measuring the TDF content of a food, regardless of the fiber’s
chemical structure. The CFIA believes that AOAC 2009.01 method eliminates the issues of
double counting when certain potential fiber fractions such as RS, polydextrose and inulin are
partially and completely measured by a combination of general and specific methods (Health
Canada, 2012). Shukri et al. (2013) demonstrated that AOAC 2009.01 was the general method
which best reflected RS recoveries in vivo from ileostomy patients (Table 4) providing further
explanation as to why this method is now preferred for labeling purposes by the CFIA (as
opposed to AOAC 991.43 general method).
Moreover, the issue of which RS content should be mentioned on the food product label remains
a challenging one and this is mainly due to the fact that RS content changes due to processing
and preparation of foods; hence, should the label mention the RS content of the food as sold or as
consumed? (Champ et al., 2003). The RS content of food products are influenced by two
factors: 1) product ageing and 2) culinary preparation of the food. Product ageing leads to either
an increase or decrease of RS content in the food product that is stored for several months after
its production. For example, most canned or frozen foods that contain starch will probably
generate RS due to retrogradation while RS2 content of banana decreases as ripening progresses.
This is not the case for dry foods such as biscuits and rusks as they do not seem to be susceptible
to storage-induced RS content changes and this poses yet another challenge for regulatory
agencies because companies producing different RS containing products will have to adhere to
different product labeling regulations (Champ et al., 2003)
30
The second factor that influences the RS content of food includes how it is prepared in the home.
Heating a product containing starch at a temperature higher than the gelatinization temperature
(>70°C) then cooling it will result in retrograded amylose and amylopectin and thus an increase
in the amount of RS3 present in the food item. This is why two-column labeling for the addition
of information on the product seems to be required for products in which significant amounts of
RS may appear during preparation. This way, the consumer knows how much RS is present in
the product as sold and as consumed (Champ et al., 2003). With this in mind, Lee & Prosky
(1992) state that: “The nutrition column on the label of food packages is not designed to display
research progress of nutrient chemistry; rather, it is designed to present accurate, concise
information about a food’s nutrient content and the physiological significance of that content for
the consumer’s benefit.” For this reason, Health Canada (2003) only requires the nutrition label
to declare the nutrient content of the food product ‘as sold’ meaning that changes in RS content
due to aging and culinary preparation practices is not taken into account (Health Canada, 2003).
The literature review above focuses on the presence of inconsistencies between in vitro methods
for the determination of RS content and whether they truly reflect RS digestibility in vivo or not;
therefore, a new method that reflects true RS digestibility in vivo is needed and a proposed
method is discussed in Chapter 4.
31
Chapter 3 Rationale, Objectives and Hypotheses
3.0: Rationale, Objectives and Hypotheses
3.1: Rationale
In the past two decades, the role of dietary fiber in health and disease has been extensively
studied by the scientific community, especially in the area of chronic disease risk reduction. The
role of dietary fiber has been well documented for risk reduction of type 2 diabetes (Hu et al.,
2001; Meyer et al., 2000; Sun et al., 2010; Yao et al., 2014), cardiovascular disease (Chandalia
et al., 2000; Jenkins et al., 2002; Jenkins et al., 2012) and colorectal cancer (Aune et al., 2011;
Bingham et al., 2003; Howe et al. 1992; Murphy et al. 2012).
An increasing body of evidence supporting the health benefits of dietary fiber as well as dietary
guidelines recommendation to increase whole grain consumption (Health Canada, 2008; Office
of Disease Prevention and Health Promotion, 2011) has lead to an increase in public awareness
(Whole Grains Council, 2013) and consumer demand (Agriculture and Agri-Food Canada, 2014)
for high-fiber products. Based on the 2014 Food & Health Survey (International Food
Information Council Foundation, 2014), 72% of the respondents (n=1,005) made an effort to eat
more foods with whole grains and 53% tried to get a certain amount or as much as possible of
fiber and whole grains when purchasing food products. In addition, whole grains scored the
highest percentage of health benefits associated with their consumption compared to soy and
flavonoids. Eighty, seventy-eight and sixty percent of the respondents were aware of the health
benefits of whole grains on the maintenance of a healthy digestive system, promotion of heart
health and promotion of healthy blood glucose levels respectively (International Food
Information Council Foundation, 2014).
32
Due to increased public awareness and consumer demands for high-fiber products, great efforts
have been put forth by the food industry to produce products enriched with dietary fiber
(Agriculture and Agri-Food Canada, 2014). One way for the food industry to achieve this is by
using more whole grains such as wheat, rice, oats and barley as fiber ingredients in bread and
bread products, cold cereals, snacks, energy bars and a wide range of food products (Agriculture
and Agri-Food Canada, 2014). However, the use of whole grains in food products has altered
their organoleptic properties such as taste and texture (Arvola et al., 2007; Bakke & Vickers,
2007). The 2014 Food & Health Survey (International Food Information Council Foundation,
2014) reported that “Taste” remains the most important factor that impacts the decision of
purchasing food products in which it scored 90% while “Healthfulness” scored 71%.
As a result of the effect of whole grains on the taste and texture of food products, the food
industry has directed its efforts into using other high-fiber food ingredients with less effect on
their taste and texture. An example of such an ingredient that can provide the health benefits of
dietary fiber without altering the product’s taste and texture is RS (Fuentes-Zaragoza et al., 2010;
Korus et al., 2009). RS is naturally white in color, finely ground and easily incorporated
into foods such as bread, biscuits, cereals and pasta by partially replacing refined flour (Food
Innovation, 2014). In food products, the taste and texture of RS resembles that of refined flour
which is a primary low-fiber ingredient in processed carbohydrate-rich foods. Hence, addition of
RS as a food ingredient in everyday processed foods increases their dietary fiber content without
alterations in organoleptic properties such as taste and texture (Natural Hi-maize Resistant
Starch, 2012).
Since novel RS food ingredients are added to products to increase dietary fiber content (Natural
Hi-maize Resistant Starch, 2012), accurate fiber labeling of these products is of great
importance. In terms of determining dietary fiber content, AOAC 991.43 method (Lee et al.,
33
1992) remains the most common in vitro method used in the food industry for the determination
of the TDF content in food products (Turowski et al., 2007; Möller, 2011). This is also true for
the commercial RS products used in this project in which their dietary fiber content was
determined by AOAC 991.43 method (Ingredion – AMIOCA TF (04400108), 2012; Ingredion –
HI-MAIZE® (22000B00), 2012; Ingredion – HYLON® VII (04400B01), 2012). As discussed in
the literature review above, AOAC 991.43 method has shown to underestimate RS content of
food products and does not accurately reflect true in vivo digestibility of RS (McCleary &
Rossiter, 2004; Shukri et al., 2013). Hence, the use of AOAC 991.43 as a method to determine
the TDF content of RS food ingredients may lead to inaccurate fiber labeling and possible
mislabeling of these products.
To-date, the golden standard for determining true digestibility of RS is the in vivo ileostomy
model (Englyst & Cummings, 1987; Englyst et al., 1996; Muir & O’Dea, 1993; Muir et al.,
1995a; Schweizer et al., 1990). A current limitation of using the ileostomy model is the need to
recruit patients with conventional ileostomies. Conventional ileostomies are being replaced by
newer types of ileostomies and are no longer routinely performed (United Ostomy Associations
of America, Inc., 2011); therefore, finding ileostomates that satisfy the inclusion criteria is
becoming increasing difficult. Having said that, there is a need to develop a new in vivo method
for the determination of the digestibility of RS using health subjects from the general public.
Study 1 of this project proposed a new in vivo method aimed at determining the digestibility of
RSs from blood glucose responses (BGRs). Subsequently, the new in vivo method proposed in
study 1 will need to be validated by the golden standard of studying starch digestibility, i.e.: the
ileostomy model. The ileostomy model used in study 2 assessed the digestibility of RSs in
subjects with a conventional ileostomy and was implemented to confirm the results from study 1.
34
The background for these 2 studies will be discussed further in the beginning of Chapter 4 and
Chapter 5.
3.2: Objectives
A) Study 1 – Digestibility of Selected Resistant Starches Estimated from the Glycemic Responses they Elicit:
1) To estimate the available carbohydrate content of three commercially available starches (Hi-
maize® 260, Hylon® VII and Amioca) from the glycemic responses they elicit.
B) Study 2 – Digestibility of Selected Resistant Starches in Ileostomates:
1) To determine the amount of carbohydrates in three commercially available starches (Hi-
maize® 260, Hylon® VII and Amioca) which escape digestion in the human small intestine and
to compare these results to the results from study 1.
2) To quantify the total bacteria present in ileal effluent and compare it with results from the
literature about healthy subjects.
3) To correlate the total bacterial count with the carbohydrates by difference and RS recovered in
ileal effluent.
3.3: Hypotheses
1) AOAC Official Method 991.43 initially used in vitro to determine the total dietary fiber
content of the three investigational starch products will underestimate the total dietary fiber
content of the products by underestimating RS content. The results from the BGR study will be
accurately reflected in the ileostomy study.
2) Ileostomates will exhibit a higher quantity of total bacteria in their ileum compared to data
from healthy subjects in the literature.
35
3) A strong negative association between the total bacterial count and the carbohydrates by
difference and RS recovered in ileal effluent will explain an underestimation of these
components.
Chapter 4 Study 1
Digestibility of Selected Resistant Starches Estimated from the Glycemic Responses they Elicit
4.0: Study 1 – Digestibility of Selected Resistant Starches Estimated from the Glycemic Responses they Elicit
4.1: Background
Study 1 was a blood glucose response study conducted from March to June 2013. This study
was the first to estimate the bioavailability of RSs by measuring glycemic responses. The only
way blood glucose can rise immediately after the ingestion of starch is if it is digested into
glucose and absorbed into the blood stream. An example of such starch is readily digestible
starch and slowly digestible starch. They represent the amount of starch digested in vitro within
20 minutes and between 20-100 minutes respectively. Readily digestible starch and slowly
digestible starch each have a different effect on glycemia because the rate of glucose release and
absorption from readily digestible starch is higher than slowly digestible starch (Englyst et al.,
1992a).
Wolever et al. (2006) have previously shown that there is a non-linear relationship between the
amount of available carbohydrates ingested and the glycemic response. This was based on two
studies which attempted to determine how GI and the amount of carbohydrates consumed (in
grams) influenced blood glucose responses. One study looked at the effect of different doses of
4 starchy foods (instant potato, bread, spaghetti and barley) (Wolever & Bolognesi, 1996a) and
36
the other studied different doses of bread and 3 sugars (glucose, sucrose and fructose) (Lee &
Wolever, 1998). These studies showed that the iAUCs over 2 hours increased non-linearly
across the range of intakes from 0-100 g of carbohydrates. Hence, these dose-response studies
can be used to estimate the effect on glycemic responses of replacing part of the available
carbohydrate in a food with an unavailable carbohydrate which has no effect on digestion and
absorption of the available carbohydrate. The non-linear regression equation includes the iAUC,
product GI and available carbohydrates (in grams). The equation can be rearranged to find the
amount of ingested available carbohydrates by substituting the iAUC and GI of the product.
Therefore, we propose to use this novel method to estimate the amount of available
carbohydrates from 3 commercially available starches (Hi-maize® 260 [HM], Hylon® VII [HY]
and Amioca [AM]). This will in turn allow for the calculation of the amount of dietary fiber in
the investigational starch products.
4.2: Methods
4.2.1: Subjects
A total of 10 healthy subjects were recruited from a subject database at Glycemic Index
Laboratories (GI Labs on 3rd floor, 20 Victoria Street, Toronto). The subjects were randomly
chosen and contacted via telephone by the Clinic Manager who briefly described the study.
Interested subjects were then briefly screened for eligibility and those who met the inclusion
criteria were given a time and date for their first visit. Monetary compensation was provided for
the subjects for their time spent at GI Labs.
For the power analysis, using the t-distribution and assuming an average coefficient of variation
(CV) of within-individual variation of iAUC values of 25%, n=10 subjects has 80% power to
detect a 33% difference in iAUC with two-tailed p<0.05.
37
The inclusion criteria included males and non-pregnant females aged 18-75 years and in good
health. The exclusion criteria was as follows: 1) age less than 18 years or more than 75 years, 2)
known history of heart disease, 3) subjects using medications or with any condition which might
make participation dangerous to themselves or to others or which may affect the results in any
way, 4) subjects who cannot or will not comply with the experimental procedures or do not
follow GI testing safety guidelines and 5) severe food allergies of any kind. All subjects
provided written informed consent before participation and the protocol was approved by the
Human Subjects Review Committee at the University of Toronto.
4.2.2: Protocol
The study was open-label with a randomized, cross-over design with each subject studied on 9
separate non-consecutive days (maximum 3 days per week). On each test day, subjects came to
GI Labs in the morning after a 10-14 hour overnight fast. On each test occasion the subject was
weighed, and two fasting blood samples were obtained by finger-prick at 5-minute intervals.
After the second fasting sample, the subject consumed their test meal. At the first bite, a timer
was started and additional blood samples were taken at 15, 30, 45, 60, 90 and 120 minutes post
test meal ingestion. Before and during the visit, a Blood Glucose Test Record Sheet (BGTRS)
was filled out with the subject's initials, ID number, date, body weight, test meal, beverage, time
of starting to eat, time it took to eat, palatability, if alcohol was consumed the previous evening,
changes to medication, time and composition of last meal and any unusual activities. During the
2 hours of the test, subjects remained seated quietly. The protocol is illustrated in Figure 3
below.
38
Figure 3 – Illustration of the Protocol for the Blood Glucose Response Study:
4.2.3: Test Meals
The composition of the test meals is shown in Table 5. The 3 investigational starch products
were provided by the sponsor of this study, Ingredion (10 Finderne Ave., Bridgewater, N.J.,
U.S.A). Fifty grams total carbohydrates (tCHO) from the investigational starch products, with
varying amounts of avCHO and dietary fiber, were mixed in 100 ml of water and fed to each
subject.
The first product, Amioca, is produced from waxy maize (100% amylopectin endosperm)
providing 350 kcal, 87 g tCHO and <1.0 g of dietary fiber per 100 g. Approximately 56.4 g of
Amioca was fed to each subject providing 50 g avCHO and 0 g dietary fiber. The sponsor
provided all composition values and dietary fiber was determined in vitro by AOAC 991.43
method (Ingredion – AMIOCA TF (04400108), 2012). Amioca is typically used as a natural
BGTRS – Blood Glucose Test Record Sheet
Visits
Fasting 10-‐14 hours
Time (mins)
15
30
45
60
90
120
1 2 3 4 5 6 7 8 9
-‐5
0
Weight
Test Meal
& Beverage
BGTRS
Blood sample
39
thickener and texturizing agent in food products and used in cooked snacks to yield an open,
expanded texture with a long shelf life (Ingredion – AMIOCA TF (04400108), 2012).
The second product, Hylon® VII, is produced from unmodified high amylose corn (70%
amylose) providing 277 kcal, 87 g tCHO and approximately 23 g of dietary fiber as RS per 100
g. Approximately 56.9 g of Hylon® VII was fed to each subject providing 38.5 g avCHO and
11.5 g dietary fiber. The sponsor provided all composition values and dietary fiber was
determined in vitro by AOAC 991.43 method (Ingredion – HYLON® VII (04400B01), 2012).
Hylon® VII is used in the confectionery industry as a gelling agent in the manufacture of high
jelly gum candies. It can also be used as an ingredient in batter coatings or extruded and fried
snack products. Its film-forming properties help produce a crisp, evenly browned product while
preventing excessive penetration of cooking oils (Ingredion – HYLON® VII (04400B01), 2012).
The third product, Hi-maize® 260 is a modified (heat-treated) high amylose corn starch providing
130 kcal, 87 g tCHO and approximately 60 g dietary fiber as RS per 100 g. Approximately
55.3 g of Hi-maize® 260 was fed to each subject providing 17.2 g avCHO and 32.8 g dietary
fiber. The sponsor provided all composition values and dietary fiber was determined in vitro by
AOAC 991.43 method (Ingredion – HI-MAIZE® 260 (22000B00), 2012). Hi-maize® 260 is
ideal for use in baked goods (breads, muffins, cakes and cookies) pasta, snacks, ready-to-eat
cereals, nutrition bars and other low-moisture foods. It can be used as a substitute for flour when
formulating healthy products due to its low water-holding capacity and physical characteristics.
It is promoted as providing the most health benefits between the 3 investigational starch products
due to its high RS content. Thus, HM is advertised as ideal for increasing fiber content, lowering
the glycemic impact of processed carbohydrates or formulating foods targeting diabetics or
weight management (Ingredion – HI-MAIZE® 260 (22000B00), 2012).
40
Because the glycemic responses of all the other test meals were expressed relative to the
response elicited by 50 g dextrose, the 50 g dextrose test meal was consumed 3 times by each
subject (beginning, middle and end of the visits) in order to obtain a more representative value of
the subject’s glycemic response. The 10g dextrose and white bread test meals were included as
positive controls to confirm the accuracy and precision of the method.
Each test meal was consumed with a standard drink. Subjects chose to have one or 2 cups of
water, tea or coffee with or without 2% milk and artificial sweetener, and the drink the subject
chose was the same for all tests done by that subject. Test meals were consumed within 10
minutes.
Table 5 – Composition of the Test Meals:
Test Meal Abbr. Weight (g)
Fiber (g)a
avCHO (g)
Protein (g)
Fat (g)
Hi-maize® 260 HM 55.3 32.8 17.2 0.4 0.4 Hylon® VII HY 56.9 11.5 38.5 <0.4 <0.6
Amioca AM 56.4 0.0 50.0 <0.3 <0.08 Dextrose Dex50 54.6 0 50 0 0
White Bread WB 55 1 25 5.4b 0.6b Dextrose Dex10 10.5 0 10 0 0
Starch composition values based on in vitro analysis provided by the sponsor a Fiber reported “as is” and analyzed by AOAC Official Method 991.43 b Proximate analysis results from Gelda Scientific for March 2013 (provided by Glycemic Index Labs)
4.2.4: Palatability
After consuming the test meal, subjects rated the palatability of the test meal using a visual
analogue scale consisting of a 100 mm line anchored at the left end by “very unpalatable” and
the right end by “very palatable”. Subjects made a vertical mark along the line to indicate their
perceived palatability. The distance from the left end of the line to the mark made by the subject
is the palatability rating. The longer the line, the higher the perceived palatability.
41
4.2.5: Blood Samples
Blood samples (2-3 drops for each time point) were collected into 5ml tubes containing a small
amount of sodium fluoride/potassium oxalate as an anticoagulant and preservative. The samples
were mixed by rotating the tube vigorously and refrigerated during the testing session. After
completion of the test session, the samples were stored at -20°C prior to glucose analysis. Blood
glucose analysis, using a YSI Model 2300 Glucose Analyzer (Yellow Spring Instruments, OH,
U.S.A) took place within 3 days of sample collection.
4.2.6: Available Carbohydrate Calculations
The trapezoid rule was used to calculate the iAUC, ignoring the area beneath baseline (Wolever,
2006). For the purpose of the iAUC calculation, fasting glucose was taken to be the mean of the
glucose concentrations at time -5 and 0 minutes.
In order to calculate the avCHOs from glycemic responses, the GI of each product needs to be
known. The dose of AM fed was assumed to contain 50 g avCHO (value provided by the
sponsor); thus, the GI of AM was determined in this experiment and was assumed to reflect the
GI of HM and HY. The amount of digestible carbohydrates absorbed from the test meals was
determined based on the rationale outlined below.
Wolever et al. (2006) showed that the iAUC elicited by various doses of carbohydrates from
white bread could be expressed as:
[1] iAUC = Z x (1 – e-0.0233 g) + 7.8
where ‘g’ is the amount of avCHO in the test meal. The value ‘Z’ depends upon the GI of the
food fed and the carbohydrate tolerance of the subject, ‘S’. Hence, equation [1] can be re-written
as follows:
[2] iAUC = GI x S x (1 – e-0.0233 g) + 7.8
42
If the test carbohydrate is glucose (GI=100), the value for ‘Z’ can be derived for each subject by
substituting ‘R’ (mean iAUC after 50 g dextrose) for iAUC and 50 for ‘g’ in equation [1] and
rearranging the equation as follows:
R = Z x (1 – e-1.165 g) + 7.8
Therefore, expressing the equation in terms of ‘Z’ yields:
[3] Z = (R – 7.8) 0.688
Since ‘Z = GI x S’; thus, ‘S = Z/GI’. Since the GI of glucose = 100, from equation [3], we can
calculate the carbohydrate tolerance ‘S’ as:
[4] S = (R – 7.8) 68.8
Thus, the amount of avCHO contained in the test carbohydrate was calculated by rearranging
equation [2] to solve for ‘g’ as follows:
[5] avCHO = {ln(1-[(F-7.8)/(GI x S)])}/-0.0233
where ‘F’ is the iAUC elicited by the test carbohydrate and ‘S’ is derived from equation [4]. The
accuracy of this equation was tested by calculating the avCHO for 10 g dextrose (GI=100) and
55 g white bread (GI=71) from their iAUC and GI values in which a value of 10 was expected
for 10 g dextrose and 24.8 for 55 g white bread.
When the experiment was completed, we found that equation [5] did not work for 7 of the 90
glycemic responses because the value (F-7.8)/(GIxS) was greater than 1 so that the value of 1-
[(F-7.8)/(GIxS)] was less than 0 and it is not mathematically possible to take the natural log of a
negative number. The problem reflects the large day-to-day variation of glycemic responses and
was caused by high iAUC values relative to GI and/or the carbohydrate tolerance ‘S’. Two
solutions to this problem were implemented: 1) the mean iAUC was used in equation [5] instead
of each subject’s iAUC and 2) another equation [5a] derived from iAUCs after the consumption
43
of various doses of different sources of carbohydrates was developed by Wolever (2006); thus,
equation [5a] was derived as follows:
[1a] iAUC = Z x (1-e-0.0222 g)
[2a] iAUC = GI x S x (1-e-0.0222 g)
R = Z x (1-e-1.11)
[3a] Z = (R) 0.670
[4a] S = (R) 67.0
[5a] avCHO = {ln(1-[(F)/(GI x S)])}/-0.0222
Use of equation [5a] resulted in somewhat different estimates of avCHO, but (F)/(GIxS)
remained greater than 1 for the 7 problematic glycemic responses described above. Because
mean iAUC was used in equation [5], mean iAUC was also used in equation [5a] and the mean
of these 2 equations was taken as the final mean BGR avCHO content of the test meals. Use of
the mean iAUC generated 2 avCHO values, 1 for each equation; hence, a measure of variability
for the avCHO content was not possible and affected how this data was statistically analyzed in
section 4.3.8 below.
4.2.7: Statistical Analysis
The results were assessed using repeated measures analysis of variance (RM-ANOVA) and if a
statistically significant difference was found, this was followed by pairwise comparisons using
Tukey’s test. Outliers were excluded for values >2xSD above the mean.
44
4.3: Results
4.3.1: Baseline Characteristics
A sample of 10 healthy, normoglycemic subjects (4 males and 6 females) aged 39±13 years with
a body mass index (BMI) of 26.5±4.2 kg/m2 participated in this study. One subject dropped out
at the beginning of the study due to new employment responsibilities. Three subjects were
taking medication or supplements on a daily basis. Subject ID 452 was taking one-a-day
multivitamin and 1,500 IU vitamin D3/ day. Subject ID 559 was taking 400 mg Seroquel, an
atypical antipsychotic drug, once a day. Subject ID 564 was taking 50 mg Tradozone as an
antidepressant, 75 mg Arthrotec for her arthritis and 40 mg Tecta to reduce gastric acid once a
day. The individual details are shown in Table 6 below.
Table 6 – Baseline Characteristics of Study 1:
Subject ID
Gender Ethnicity Age (yrs)
Height (m)
Weight (kg)
BMI (kg/m2)
336 M Latin American 48 1.77 70.0 22.3 341 M South Asian 47 1.88 110.5 31.3 354 F Caucasian 43 1.66 71.2 25.8 452 M Caucasian 63 1.65 75.8 27.8 538 M Caucasian 21 1.83 81.8 24.4 550 F Arab/West Asian 34 1.51 51.5 22.6 559 F Caucasian 37 1.64 92.7 34.5 560 F Arab/West Asian 32 1.62 66.3 25.3 563 F Caucasian 22 1.73 65.9 22.0 564 F Caucasian 47 1.58 73.2 29.3
Mean±SD 39±13 1.69±0.11 75.9±16.2 26.5±4.2
4.3.2: Blood Glucose Analysis
Of the 720 blood samples, only 1 (0.1%) BGR value was missing (subject ID 563 at 45 minute
blood collection on the first 50 g dextrose test) and was replaced by the mean of the 30 and 60
minute BGR values. Due to a low volume of blood, duplicate analysis could not be done on 10
of the 90 0 minute samples. For the remaining 80 0 minute BGRs, the mean±SD was 4.57±0.33
45
mmol/L yielding a coefficient of variation (CV%) of 7.27%, well over 3% as a result of
analytical variation. The mean±SD for the 90 -5 and 0 minute BGRs was 4.62±0.32 mmol/L
yielding a CV% of 6.86%, reflecting both analytical variation and minute-to-minute variation in
BGRs.
4.3.3: Glycemic Response Elicited by Reference Food
The Dex50 test meal was consumed 3 times by each subject. The mean±SEM of the iAUC
values were 221.2±29.0, 212.4±31.4 and 199.1±28.7 mmolxmin/L for the 1st, 2nd and 3rd tests of
Dex50 respectively and no significant differences were observed among them (F=0.19, p=0.77).
The mean±SEM of the within-subject CV% of the iAUC values after the Dex50 test meals was
25.6±3.1% and this is satisfactory since the mean within-subject CV for the repeated reference
food tests should be less than 30%.
4.3.4: Palatability
Palatability, iAUCs and relative glycemic responses (RGRs) for every treatment are shown in
Table 7 below. A one-way RM-ANOVA was performed on the palatability results (F=8.3,
p=0.002) with significant differences observed among the 6 test meals. WB was significantly
different from AM (p≤0.01) and HM (p≤0.001) and Dex50 was significantly different from HY
and HM (p≤0.05). The starch products scored fairly low on the palatability scale with HM
scoring the lowest, 20.2±5.7 mm and HY scoring the highest, 30.7±9.6 mm among the products
(Figure 4).
46
Table 7 – Palatability, Incremental Areas Under the Blood Glucose Response Curves and
Relative Glycemic Responses of the Test Meals:
Values are expressed as means±SEM for each treatment
Figure 4 – Palatability of the Test Meals:
Test Meal Abbr. Palatability (mm)
iAUC (mmolxmin/L)
Relative Glycemic
Response (%) Hi-maize® 260 HM 20.2±5.7 70.5±20.4 32.4±8.5
Hylon® VII HY 30.7±9.6 25.8±13.0 9.7±3.8 Amioca AM 29.0±4.5 77.0±14.2 37.7±5.5 Dextrose Dex50 53.6±8.9 210.9±25.6 100.0±0.0
White Bread WB 61.6±6.1 105.8±26.4 51.3±10.5 Dextrose Dex10 49.0±8.8 61.4±11.0 31.4±4.8
A) Bars represent mean±SEM of the palatability values for each treatment for n=10 subjects B) Abbreviations: WB – White Bread, Dex50 – Dextrose 50, Dex10 – Dextrose 10, HY – Hylon® VII, AM – Amioca, HM – Hi-maize® 260 C) Bars with different letters differ significantly
1 1 1 1 1 1
0
10
20
30
40
50
60
70
80
90
100
Treatment
Pala
tabi
lity
(mm
)
WB Dex50 Dex10 HY AM HM
abad
ae
be
cde
ce
47
4.3.5: Blood Glucose Responses
A two-way RM-ANOVA was performed on the BGRs with a significant time (F=56.6,
p<0.0001), treatment (F=7.8, p<0.0001), and time x treatment interaction (F=11.1, p<0.0001).
Significant differences in the BGRs were observed among the 6 test meals at 15, 30, 45 and 60
minutes. AM and HM elicited nearly identical BGRs at 30 minutes, 5.62±0.1 mmol/L and
5.63±0.25 mmol/L respectively, and did not differ significantly at any time point (p>0.05).
Among the 3 investigational starch products, AM elicited the highest BGR, followed closely by
HM and HY elicited the lowest BGR.
Figure 5 – Blood Glucose Responses Elicited by the Test Meals:
A) Blood glucose response curves represent the mean±SEM for each treatment at each time point (0, 15, 30, 45, 60, 90 and 120 minutes) for n=10 subjects B) Abbreviations: Dex50 – Dextrose 50, WB – White Bread, Dex10 – Dextrose 10, HM – Hi-maize® 260, AM – Amioca, HY – Hylon® VII
0 30 60 90 1204
5
6
7
8HM
HYAM
WBDex10
Dex50
Time (minutes)
Blo
od
Glu
cose
Res
po
nse
(mm
ol/L
)
48
4.3.6: Incremental Areas Under the Blood Glucose Response Curves
iAUC values for the treatments are shown in Table 7 above. A one-way RM-ANOVA was
performed on the iAUC values (F=19.2, p<0.0001) and Dex50 had a significantly higher iAUC,
210.9±25.6 mmolxmin/L, than all the other treatments (Figure 6). AM’s iAUC was significantly
higher than HY (p≤0.01) and the iAUC of HY was 34% that of AM and 37% that of HM. The
GI of AM was calculated to be 38 (low) and assumed to reflect the GI of HY and HM.
Figure 6 – Incremental Areas Under the Blood Glucose Response Curves:
A) Bars represent the mean±SEM of the iAUC values for each treatment for n=10 subjects B) Abbreviations: Dex50 – Dextrose 50, WB – White Bread, AM – Amioca, HM – Hi-maize® 260, Dex10 – Dextrose 10, HY – Hylon® VII C) Bars with different letters differ significantly
0
25
50
75
100
125
150
175
200
225
250
Treatment
Blo
od
Glu
cose
iAU
C (m
mo
lxm
in/L
)
Dex50 WB AM HM Dex10 HY
a
bc
b bc
b
c
49
4.3.7: Relative Glycemic Responses
RGR values are shown in Table 7 above. A one-way RM-ANOVA was performed on the RGRs
(F=28.2, p<0.0001) with Dex50 (100%) significantly different from all the other treatments. The
RGR of AM was significantly different from HY (p≤0.01) and remained significant after the
outlier was excluded from HY (one outlier only for subject 341). The highest RGR of the
starches was from AM (37.7%) followed by HM (32.4%) and HY (9.7% and excluding the
outlier: 6.3%).
Figure 7 – Relative Glycemic Responses:
A) Bars represent the relative glycemic response percentages for each treatment for n=10 subjects B) Abbreviations: Dex50 – Dextrose 50, WB – White Bread, AM – Amioca, HM – Hi-maize® 260, Dex10 – Dextrose 10, HY – Hylon® VII C) Bars with different letters differ significantly
1 1 1 1 1 1
0
20
40
60
80
100
Treatment
Rel
ativ
e R
esp
on
se (%
)
Dex50 WB AM HM Dex10 HY
a
b
b bcb
c
50
4.3.8: Available Carbohydrate Calculated from Blood Glucose Responses
avCHO values are shown in Table 8 and Figure 8 below and values are expressed as means only
(explained in section 4.2.6 above). Hence, a two-way ANOVA and multiple comparisons were
not relevant to perform on these data. The mean BGR values from equations [5] and [5a] in
Table 8, column 5 are the final BGR avCHO values and will be referred to as ‘mean BGR’
values in this and the following sections. The mean BGR avCHO values for the starches are as
follows: HM was 37.5 g (2.2 times higher than the in vitro value), AM was 43.8 g (0.9 times or
10% less than the in vitro value) and HY was 9.2 g (0.2 times or 80% less than the in vitro
value).
Table 8 – Available Carbohydrate Calculated from Blood Glucose Responses:
Available Carbohydrates (g) Test Meal In vitroa BGR
Equation [5]b BGR
Equation [5a]b Mean BGR Equations [5] & [5a]
HM 17.2 35.1 40.0 37.5 AM 50.0 41.2 46.4 43.8 HY 38.5 7.5 10.9 9.2
Dex50 50.0 50.1 49.9 49.9 WB 25.0 27.1 28.9 27.9
Dex10 10.0 8.6 9.8 9.2 a In vitro values provided by the sponsor b BGR values expressed as means for n=10 subjects Abbreviations: HM - Hi-maize® 260, AM – Amioca, HY - Hylon® VII, Dex50 – Dextrose 50, WB – White Bread, Dex10 – Dextrose 10
51
Figure 8 – Available Carbohydrate Calculated from Blood Glucose Responses:
4.3.9: Estimates of Available Carbohydrate in Test Starches
Estimates of avCHOs in the test starches are shown in Table 9 below. The amount of avCHOs
per 100 g of the test starches are as follows (in vitro vs. mean BGR): HM was 31.1 g vs 67.9 g,
AM was 88.7 g vs 77.7 g and HY was 67.7 g vs. 16.2 g. The largest difference observed in the
in vitro vs. mean BGR values was for HY.
A) Bars represent the mean avCHO content of each treatment by each method for n=10 subjects B) Abbreviations: Dex50 – Dextrose 50, WB – White Bread, Dex10 – Dextrose10, AM – Amioca, HY - Hylon® VII, HM - Hi-maize® 260
1 1 1 1 1 1
0
5
10
15
20
25
30
35
40
45
50
55
60
Treatment
Ava
ilab
le C
arb
oh
ydra
te (g
/test
mea
l)
Dex50 WB Dex10 AM HY HM
In-vitroMean BGR [5]Mean BGR [5a]Mean BGR [5&5a]
52
Table 9 – Estimates of Available Carbohydrate in Test Starches:
Available Carbohydrates (g/portion)
Available Carbohydrates (g/100 g)
Test Meal Weight (g) In vitroa Mean BGRb In vitroa Mean BGRb HM 55.3 17.2 37.5 31.1 67.9 AM 56.4 50.0 43.8 88.7 77.7 HY 56.9 38.5 9.2 67.7 16.2
a In vitro values provided by the sponsor b Mean BGR values represent mean of equations [5] and [5a] and are expressed as means for n=10 subjects Abbreviations: HM - Hi-maize® 260, AM – Amioca, HY - Hylon® VII 4.4: Discussion
The research question for this study was: Do in vitro methods accurately reflect true in vivo
digestibility of RS? To answer this question, our results showed that the iAUCs and RGRs of the
3 test starches were significantly less than that of Dex50 and suggest that the in vitro AOAC
991.43 method greatly underestimated the fiber content in HY, marginally underestimated the
fiber content in AM and greatly overestimated the fiber content in HM. No previous studies
have been done on estimating avCHO content of RS products from glycemic responses; hence no
direct comparison of our data with data from the literature was possible. Nonetheless, our results
for HM were confirmed in Table 4 when AOAC 991.43 method overestimated RS content of
high-amylose corn starch in comparison to in vivo data from ileostomy patients.
The first parameter of interest in this study was the perceived taste of the raw starch products
measured by the palatability scale. The test starches scored fairly low on the palatability scale
with HM scoring the lowest (20%), HY scoring the highest (31%) and AM scoring in between
(29%). The taste of the raw starch mixed in water was predominantly described as “wallpaper
paste.” It is important to note that these starch products are not manufactured in order to be
consumed raw; however, are food ingredients meant to be incorporated into cakes, cookies,
biscuits etc. to increase fiber content without effect on taste and texture. Hence, these
53
palatability results are not representative of the effect that these starch products would have on
taste and texture had they been added to cake for example, instead of being consumed raw.
The iAUC of Dex50, 210.9±25.6 mmolxmin/L, and RGR, 100%, were significantly higher than
all the other treatments. WB had half the amount of avCHO of Dex50 (50 g vs. 25 g) and this
clearly shows from the iAUC values in which the iAUC of WB, 105.8 mmolxmin/L, was half
that of Dex50. Dex10 had a similar BGR to WB at 30 minutes but reduced below fasting levels
after 45 minutes up until 120 minutes. When calculating the iAUC, the area below fasting levels
is disregarded, possibly explaining why the iAUC of Dex10 (61.4±11.0 mmolxmin/L) is similar
to HM (70.4±20.4 mmolxmin/L). Surprisingly, the BGRs of HM (iAUC=70.4±20.4
mmolxmin/L) and AM (iAUC=77.0±14.2 mmolxmin/L) were very similar and nearly identical at
30 minutes although in vitro analysis has indicated that HM contains 60% RS as dietary fiber and
Amioca has 0% dietary fiber. HY elicited the lowest BGR (iAUC=25.8±13.0 mmolxmin/L)
although in vitro analysis indicated an avCHO content of 38.5 g per 50 g tCHO content,
theoretically placing its BGR closer to AM.
Of interest was the comparison of the RGR and mean BGR avCHO content of HM, Dex10 and
HY. The RGR of HM and Dex10 were very similar, 32% vs. 31% respectively, although the
estimated mean BGR avCHO content was 37.5 g per 50 g tCHO portion of HM and 10 g for
Dex10. On the other hand, the estimated mean BGR avCHO content of HY was 9.2 g per 50 g
tCHO portion, identical to the value estimated for Dex10, 9.2 g. Taking a look back at the RGR
of Dex10 and HY, they were 31% vs. 10%. This confirms the very concept of GI in which the
same amount of ingested avCHO from different carbohydrate sources, i.e.: dextrose and starch,
elicit completely varying BGR, iAUC and ultimately RGR values.
54
The similarity in the BGR, iAUC and RGR values of HM and AM was alarming, due to the fact
that, based on in vitro analysis, HM contained 17.2 g avCHO per 50 g tCHO portion while AM
contained 50 g avCHO per 50 g tCHO portion. This suggests that there are more avCHO in HM
and less avCHO in AM than accounted for in vitro and this was confirmed by the mean BGR
avCHO calculations in which HM contained 37.5 g of avCHO and AM contained 43.8 g of
avCHO per 50 g tCHO, a mere 6.3 g difference between them. This also suggests that a large
fraction of the starch in HM is slowly digestible starch, as opposed to RS, that is slowly but
completely digested and absorbed into the blood stream. Also, AM might not only contain
readily digestible starch but also slowly digestible starch which might explain why HM and AM
elicited similar glycemic responses. As for HY, the in vitro method indicated that it contained
38.5 g avCHO per 50 g tCHO portion; however, mean BGR avCHO calculations showed that
HY only contained 9.2 g avCHO, explaining why it elicited the lowest BGR, iAUC and RGR
values.
The avCHOs calculated for Dex10 (9.2 g) was a better representation of the accuracy of the
method than WB because it was within 8% of its in vitro value as opposed to WB (27.9 g) that
was 12% higher than its in vitro value. A possible reason for this is that subjects vary in their
ability to process avCHOs from different carbohydrate sources, i.e.: dextrose and flour. If the
individual iAUC values, as opposed to the mean values, for Dex10 and WB were used to
determine the avCHO content of these test meals, more accurate values could have been
observed, possibly within 2-3% of their in vitro values. For example: using individual iAUCs
for Dex10 for equation [5] yielded an average of 10 g avCHOs and equation [5a] yielded an
average of 11 g avCHOs and the mean of the equations was 10.5 g avCHO content. Hence, a
better representation of the avCHO content of Dex10 could have been achieved had individual
iAUCs, rather than mean iAUC, been used. This was not possible to do with our data because of
55
the 7 problematic glycemic responses described in section 4.2.6; thus, mean iAUCs were used to
estimate the avCHO content of all the test meals.
When it comes to determining the digestibility of RS, two main in vivo methods are used, the
breath hydrogen method (Jenkins et al., 1998) and the ileostomy model (Englyst et al, 1992a).
The breath hydrogen method assesses digestibility based on how much hydrogen gas is produced
from the RS that escapes digestion in the small intestine. The ileostomy model assesses
digestibility of RS by direct collection of ileal effluent and further analysis to determine amount
of RS recovered. Both of these methods do not take into account the affect that RS has on
glycemic responses when it replaces a portion of avCHOs. This novel in vivo method has
several advantages including the use of healthy, normoglycemic subjects, sample collection takes
place in 2 hours as opposed to 10 hours for the ileostomy model and it avoids the issue of the
type of gut microbiota inhabiting the colon of subjects, i.e.: methanogens (Gilat et al., 1978),
from the breath hydrogen method.
The results of this novel in vivo approach further emphasize the importance of using the relevant
in vitro methods for the dietary fiber fraction of interest. This ties directly with the implications
of this project in which accurate dietary fiber labeling of novel RS products is essential to ensure
accurate fiber labeling of products which incorporate RS.
There was one main limitation of the BGR method used in this study, i.e.: accurate knowledge of
the product GI was necessary to give accurate estimates of the avCHO content of the starch
products. For example: the GI of Amioca was calculated to be 38 and the mean BGR avCHO
content was 43.8 g/ 50 g tCHO content of the product. Let’s assume that the GI of Amioca was
calculated to be 36 or 40 (±2). If this was the case, the mean BGR avCHO content for AM for a
56
GI of 36 would have been 48 g and 37 g if the GI was 42. Therefore, a -2 in GI resulted in a +5
g in the avCHO content of AM and a +2 in GI resulted in a -5 g in the avCHO content of AM.
It remains unclear whether the novel in vivo approach to estimating avCHO content of RS
products used in this study produces accurate results of the true avCHO content of commercial
RS containing products. Having said that, this novel approach needs to be validated by
comparing its results with the results from the ileostomy model since it is considered the golden
standard of studying starch digestibility.
57
Chapter 5 Study 2
Digestibility of Selected Resistant Starches in Ileostomates
5.0: Digestibility of Selected Resistant Starches in Ileostomates
5.1: Background
Study 2 was the ileostomy study conducted from October 2013 – April 2014. This study was
conducted in order to validate the results from the novel in vivo BGR method described in
Chapter 4. The best way to conclude whether the method described in Chapter 4 truly reflects
the bioavailability of RSs is by comparing its results to the results from the golden standard of
studying starch digestibility, i.e.: the ileostomy model. This study will help us answer the
following question: Does the novel BGR method accurately reflect the bioavailability of novel
RS products? This study was registered in ClinicalTrials.gov (NCT01939600).
5.2: Methods
5.2.1: Subjects
We aimed to have a total of 10 subjects to complete this study and the subjects were recruited
from the Division of Gastroenterology at St. Michael’s Hospital. The gastroenterologists
referred patients with a conventional ileostomy to the study coordinator who briefly assessed
them for eligibility via telephone. Patients interested in the study and who met the inclusion
criteria were invited to the Risk Factor Modification Center of St. Michael’s Hospital to
complete the consent form, subject screening form as well as the meal plan. All the study visits
took place at GI Labs and monetary compensation was provided for the subjects for their time
and effort.
58
The inclusion criteria included males or non-pregnant females with a conventional ileostomy and
who are clinically stable with no clinical evidence of malabsorption. The exclusion criteria was
as follows: 1) short bowel syndrome (defined as having less than 200 cm of intact small
intestines left), 2) prone to high ileostomy output with change in diet, 3) ileostomy created less
than 6 months from the first study visit unless output is stable, 4) subjects using medications
which influence gastrointestinal motility or absorption, 5) any condition which might make
participation dangerous to the subjects themselves or to others or which may affect the results in
any way, 6) subjects who cannot or will not comply with the experimental procedures and 7)
severe food allergies of any kind. All subjects provided written informed consent before
participation and the protocol was approved by St. Michael’s Hospital Research Ethics board as
well as the Human Subjects Review Committee at the University of Toronto.
5.2.2: Protocol
The study was open-label with a randomized, cross-over design with each subject studied on 4
occasions (control and 3 investigational starch products) over a minimum of 4 weeks (one test
per week) and a maximum of 7 weeks. On the evening before each test day, subjects consumed
a starch- and fiber-free dinner. The meal was planned by the study coordinator and the subject
and selected from a menu (Appendix 1) suitable for subjects who consume animal products as
well as vegetarians. The dinner meal was picked up by the subject from Glycemic Index
Laboratories (3rd floor, 20 Victoria Street, Toronto) or was delivered to them at home or work 1
or 2 days before each test day. Subjects were provided with a Test Meal Record Sheet (TMRS)
to fill out indicating the composition of all the study meals consumed, the time they were
consumed, any changes to medication and any unusual activities.
Each test day started at approximately 8:00 am after a 10-14 hour overnight fast where the
subject first changed their ileostomy pouch, started a timer and consumed a breakfast test meal
59
consisting of a polysaccharide-free breakfast meal alone (control) or a breakfast test meal with
one of the 3 investigational starch products in randomized order (Table 5). A starch- and fiber-
free lunch meal (similar to the dinner meal) was consumed 4 hours after the breakfast meal and
the subjects were allowed a mid-morning snack before lunch consisting of either cheese, yogurt
or pudding with fruit juice or the meal-replacement. When the ileostomy pouch was a third full,
effluent was collected and placed immediately in the -80ºC deep freezer and the subjects
continued to collect ileostomy effluent for 10 hours.
Subjects had the choice to complete their visits at GI Labs or another location such as their home
or workplace. If the latter, a study staff delivered the test breakfast and lunch meals and a box of
dry ice to the subject’s location, supervised the breakfast meal consumption and showed the
subject how to handle and store the box of dry ice safely. Ileostomy effluent collected at home
or work was delivered to GI Labs within 24 hours of collection either by the subject or by taxi or
was picked up from the subject's home or work by the study staff.
The dinner, breakfast and lunch meals chosen remained constant over the 4 test occasions and
were recorded in the TMRS. Subjects were allowed drinks of water, coffee or tea freely during
the day. The protocol is illustrated in Figure 9 below.
60
Figure 9 – Illustration of the Protocol for the Ileostomy Study:
5.2.3: Test Meals
The 3 investigational starch products as well as the amount fed in this study were equivalent to
the ones used in study 1 (Table 5). Subjects chose breakfast from a menu (Appendix 1) of items
including: milk-based cheese or cheese-alternative, milk- or soy-based yogurt, pudding, fruit
salad, pastries, fruit juice, meal-replacement (Ensure Nutrition Plus), milk and coffee or tea. The
test starches were mixed into milk, yogurt, pudding, fruit juice, water or were taken in the meal-
replacement. The selected breakfast meal was the same for all 4 test occasions.
5.2.4: Freeze-Drying Effluent
Ileal effluent samples were removed from the deep-freezer and placed in a shaking water bath at
room temperature for 15-20 minutes to thaw. After thawing, the samples were weighed and the
weight was recorded. Each sample consisted of 2-3 collection bags which were emptied into a
blender and homogenized for 1 minute at lowest speed. The freeze-drier flask was weighed and
350 g of homogenized effluent was poured into the flask. The sample was rotated in the flask to
Fasting 10-‐14 hours
Night Before Test Day 1
Night Before Test Day 2
1
Night Before Test Day 3
1
Night Before Test Day 4
Dinner
Test Meal Record Sheet (TMRS)
Control Breakfast or Breakfast with Added Resistant Starch
Snack
Lunch
TMRS
Ileal Effluent Collection
10 hours
Control Breakfast or Breakfast with Added Resistant Starch
Snack
Lunch
TMRS
Ileal Effluent Collection
10 hours
Control Breakfast or Breakfast with Added Resistant Starch
Snack
Lunch
TMRS
Ileal Effluent Collection
10 hours
Control Breakfast or Breakfast with Added Resistant Starch
Snack
Lunch
TMRS
Ileal Effluent Collection
10 hours
Dinner
Test Meal Record Sheet
Dinner
Test Meal Record Sheet
Dinner
Test Meal Record Sheet
Visits
Fasting 10-‐14 hours
Fasting 10-‐14 hours
Fasting 10-‐14 hours
61
reduce thickness and increase surface area then placed at a 45º angle in the -80ºC freezer
overnight. The next day, the freeze-dryer was turned on and the frozen samples were loaded
onto the freeze-dryer. Samples from the control diet took approximately 4 days to dry and
samples from the starch test days took approximately 7 days to dry. The samples were dry when
the external side of the flask was warm and dry and when big cracks appeared through the
effluent. After the samples had dried they were removed from the freeze-dryer and weighed.
The dry material was transferred into a plastic bag and lightly pounded into powder which was
later transferred into hinged-lid plastic containers. The freeze-dried samples were sent to Dr.
Suzanne Hendrich’s lab, Professor of Food Science and Human Nutrition at Iowa State
University for proximate and RS analysis.
5.2.5: Ileal Effluent Analysis
5.2.5.1: Proximate Analysis
The moisture content of the freeze-dried effluent samples was determined using the Oven-Drying
Method (Horwitz, 2007) where the percent moisture was calculated as follows:
% Moisture = (weight (g) of sample before drying – weight (g) of sample after drying)/
weight (g) of sample before drying x 100%.
The fat content was determined using the Goldfisch Method (Horwitz, 2007) in which:
% Oil = % oil (wet basis)/ [1-(% moisture/100)]
The protein content was determined by the DUMAS Nitrogen Combustion Method (Simonne et
al., 1997) in which:
% Protein = weight of nitrogen (N) (g) × 6.25 × 100%/ weight (g) of sample
The ash content (Horwitz, 2007) was determined by placing 5g of the sample in a crucible in a
550ºC furnace overnight and the:
% Ash = weight of ash (g)/ weight of sample (g) x 100%
62
The carbohydrate by difference (Horwitz, 2007) was calculated as:
100 - (weight in grams [protein + fat + moisture + ash]) in 100g of the effluent
Analysis was performed on duplicate samples and the mean was taken as the final value for
further calculations.
5.2.5.2: AOAC Official Method 2002.02
The RS content of the ileal effluent samples was determined by AOAC Official Method 2002.02
(McCleary et al., 2002) and this method is applicable to samples containing more than 2%
weight/weight (w/w) of RS. The samples were first prepared with the assumption that they
contained <10% RS. If the absorbance read over 1 absorbance unit, it was indicative that the
sample contained >10% RS and was diluted to 100 ml with a second reading of the absorbance.
For samples >10% RS:
A) RS (g/100 g of the sample) = ∆E x F/W x 90
where ‘E’ is the absorbance read against the reagent blank, ‘F’ is the conversion from absorbance
to micrograms of D-glucose and ‘W’ is the dry weight of the sample analyzed.
For samples <10% RS:
B) RS (g/100 g of the sample) = ∆E x F/W x 9.27
The difference in the multiplication factor of 90 in equation A and 9.27 in equation B is due to
the volume correction of the solution. For equation A, 0.1 ml was taken from 100 ml of the
solution and for equation B, 0.1 ml was taken from 10.3 ml of the solution.
5.2.5.3: Total Bacterial Count Determined by Real-Time Quantitative Polymerase Chain Reaction
Bacterial DNA was extracted from 500 µg of undiluted ileal effluent using E.Z.N.A.™ Stool
DNA Isolation Kit (Omega Bio-Tek, Georgia, U.S.A) according to the manufacturer’s
instructions, but modified to include a lysozyme digestion step (20 mg/ml at 37ºC for 30
63
minutes). The lysozyme digestion step was included to facilitate maximal bacterial cell wall
lysis in order to improve efficiency of DNA extraction (Singh et al., 2013). DNA concentration
and purity were measured using Thermo Scientific NanoDrop 2000 spectrophotometer (Thermo
Fisher Scientific, Massachusetts, U.S.A) at 230, 260 and 280 nm. Purity ratios above 1.7 were
considered of good quality and used for the real-time quantitative polymerase chain reaction (R-
T qPCR).
R-T qPCR was employed using 10 ng of total DNA, total bacteria TaqMan assay (Furet et al.,
2009; Suzuki et al., 2000) and TaqMan Gene Expression Master Mix (Applied Biosystems,
U.S.A.) in 10 µl PCR reaction. All reactions were run in triplicates in a 384-well optical plate in
Applied Biosystems 7900 HT Real-Time PCR machine with default thermocycling conditions as
per the manufacturer’s instructions.
5.2.6: Statistical Analysis
The results were assessed using repeated measures analysis of variance (RM-ANOVA) and if a
statistically significant difference was found, this was followed by pairwise comparisons using
Tukey’s test. Pearson correlations were performed on the total bacterial count, carbohydrates by
difference and RS.
5.3: Results
We aimed to recruit 10 subjects for this study but only 3 ileostomates were enrolled. The
recruitment period extended for 5 months where 7 ileostomates were approached and 3 agreed to
participate. It was difficult to find patients with conventional ileostomies because this surgical
procedure is becoming less commonly performed and is being replaced by newer procedures
aimed at enhancing the quality of life of these patients. Of the 4 ileostomates that declined
participation, 2 had difficulties with transportation, one was on a strict dietary regimen and the
64
last ileostomate did not have stable output due to her ileostomy being created less than 2 months
from the time of contact. Therefore, only 3 ileostomates completed the study.
5.3.1: Baseline Characteristics
A sample of 3 Caucasian, female patients with conventional ileostomies completed the study.
The mean±SD age of the ileostomates was 44.3±26.8 years with a normal BMI of 23.5±6.2
kg/m2. Subject 1 (AR001) had her ileostomy for 30 years due to Crohn’s disease with 76 cm of
her small bowel removed and occasionally takes domperidone to relieve nausea. Subject 2
(AR002) has had her ileostomy for 16 years due to Crohn’s disease with 15 cm of her small
bowel removed and takes multivitamins, vitamin D and fish oil supplements. Subject 3 (AR003)
has had her ileostomy for 4 months due to dysmotility of the large intestine as well as
gastroparesis without the resection of any of her small bowel and takes a combination of
medication to relieve nausea and pain. The individual details are shown in Table 10 below.
Table 10 – Baseline Characteristics of Study 2:
Subject ID
Gender Ethnicity Age (yrs)
Height (m)
Weight (kg)
BMI (kg/m2)
Amount of Small Bowel
Resected (cm)
AR001 F Caucasian 73 1.5 50.0 21.9 76 AR002 F Caucasian 40 1.6 46.7 18.2 15 AR003 F Caucasian 20 1.7 85.4 30.3 0
Mean±SD 44.3±26.8 1.6±0.1 60.7±21.5 23.5±6.2 30.3±40.3
65
5.3.2: Control Diets
Due to each subjects’ dietary restrictions, a customized control meal plan was implemented
based on each of their personal preferences. Subject 1 (AR001)’s control meal plan is described
in Appendix 2, Subject 2 (AR002)’s control meal plan is described in Appendix 3 and Subject 3
(AR003)’s control meal plan is described in Appendix 4.
5.3.2.1: Summary of the Control Diets
A summary of the 3 control diets is outlined in Table 11 below. Subject 3 (AR003) had the
highest amount of energy, carbohydrates, fiber and fat among the meal plans and subject 2
(AR002) had the highest amount of protein among the meal plans.
Table 11 – Summary of the Control Diets: Energy
(kcal) CHO (g) Fiber (g) Protein (g) Fat (g)
AR001 1,501 200 5 66 46 AR002 1,682 196 8 95 54 AR003 1,876 247 11 71 69
Mean±SD 1,686±188 214±28.4 8±3 77±15.5 56±11.7
5.3.3: Amount of Ileal Effluent Collected
At the beginning of every visit, the subjects were asked to change their ileostomy pouch to a new
one and ileal effluent was collected for 10 hours from 8:30am to 6:30 pm. Effluent was collected
approximately every 3 hours and immediately placed in the -80°C deep-freezer until freeze-
drying took place. The amount of effluent collected within subjects was less variable than the
amount of effluent collected from different subjects among treatments. This is clearly illustrated
by comparing the %CV for each subject (3.3%, 29.1% and 16.8%) in Table 12 below with the
%CV for each treatment (69.8%, 74.0%, 60.7% and 90.4%) suggesting a difference in the
digestive pattern of each individual subject in response to the starches. Subject 1 (AR001)
66
produced the highest amount of effluent (715.3±23.7 g) followed by subject 3 (AR003)
(295.3±49.7 g) and subject 2 (AR002) produced the lowest amount of effluent (168.3±49.0 g).
The individual details are shown in Table 12 below.
Table 12 – Amount of Ileal Effluent Collected in 10 Hours: Treatment
Subject Control (g) HM (g) HY (g) AM (g) Mean±SD %CV AR001 715 747 709 690 715.3±23.7 3.3 AR002 164 187 219 103 168.3±49.0 29.1 AR003 333 281 336 231 295.3±49.7 16.8
Mean±SD 404.0±282.3 405.0±299.9 421.3±255.9 341.3±308.7 %CV 69.8 74.0 60.7 90.4
Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca 5.3.3.1: Amount of Dry Matter Produced
Dry matter produced from effluent after each treatment was determined by freeze-drying it to
remove the liquid component and weighing the remaining solids (Table 13). HM produced the
highest amount of dry matter (58.0±33.8 g) followed by HY (58.0±20.9 g), AM (40.7±41.1 g)
and control (35.3±24.8 g). The individual details are shown in Table 13 below.
Table 13 – Amount of Dry Matter Produced in 10 Hours: Treatment
Subject Control (g) HM (g) HY (g) AM (g) Mean±SD AR001 64 97 82 88 82.8±13.9 AR002 20 37 48 14 29.8±15.6 AR003 22 40 44 20 31.5±12.3
Mean±SD 35.3±24.8 58.0±33.8 58.0±20.9 40.7±41.1 Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
67
5.3.3.2: Mean Percent Dry Matter
The mean percent dry matter produced in 10 hours for the control treatment was 9.2% followed
by 11.8% from AM, 15.5% from HY and 15.7% from HM. Since in vitro analysis showed that
AM contained 0g dietary fiber, the mean percent dry matter should have been closer to the
control value, 9.2%; however, it was 2.6% higher indicating that AM may contain more fiber
than accounted for in vitro. The similarity in the mean percent dry matter of HM and HY
signifies that these starches may contain similar fiber values. The individual details are shown in
Table 14 below.
Table 14 – Mean Percent Dry Matter Produced in 10 Hours:
Abbreviations: HM –Hi-maize® 260, HY – Hylon® VII, AM – Amioca 5.3.4: Proximate Analysis of Ileal Effluent
The results of the proximate analysis of ileal effluent are given in Table 15 below. The table
presents the unadjusted values of moisture, fat, protein, ash and carbohydrates by difference
(CHOD) and shows their mean±SEM for each treatment. The dietary fiber content of HM, HY
and AM as provided by the sponsor were 32.8 g, 11.5 g and 0 g/ 50g tCHO portion respectively
and the adjusted CHOD values from proximate analysis indicated that HM, HY and AM
contained 18.7 g, 21.6 g and 2.9 g as dietary fiber.
Treatment Subject Control (g) HM (g) HY (g) AM (g) AR001 8.9 13.0 11.5 12.8 AR002 12.1 19.9 21.9 13.9 AR003 6.5 14.2 13.1 8.6 Mean 9.2 15.7 15.5 11.8
68
Table 15 – Proximate Analysis of Ileal Effluent:
Proximate Analysis (g) per 10 hours Subject Treatment Moisture (g) Fat (g) Protein (g) Ash (g) CHOD (g) AR001 Control 5.6 16.1 17.6 11.8 11.8 AR002 Control 1.1 1.1 4.6 4.1 8.7 AR003 Control 1.8 1.5 4.4 3.6 10.0
Mean±SEM 2.8±1.4 6.3±4.9 8.9±4.4 6.5±2.7 10.2±0.9
AR001 HM 4.7 16.4 17.9 13.4 36.0 AR002 HM 0.9 1.4 4.6 4.4 24.0 AR003 HM 1.3 1.1 3.9 2.8 26.8
Mean±SEM 2.3±1.2 6.3±5.1 8.8±4.6 6.9±3.3 28.9±3.6
AR001 HY 2.2 13.4 14.3 7.9 34.0 AR002 HY 1.3 1.1 5.5 4.7 33.2 AR003 HY 0.9 1.78 5.3 3.6 28.0
Mean±SEM 1.5±0.4 5.4±4.0 8.4±3.0 5.4±1.3 31.7±1.9
AR001 AM 6.7 20.5 24.5 14.6 20.8 AR002 AM 0.7 1.8 3.3 2.5 5.9 AR003 AM 1.0 2.1 4.9 2.5 8.9
Mean±SEM 2.8±2.0 8.2±6.2 10.9±6.8 6.5±4.0 11.9±4.6 Proximate analysis values are unadjusted Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca, CHOD – Carbohydrates by Difference
69
5.3.5: Comparison of Resistant Starch Fed vs. Recovered Amounts of RS fed (in vitro) vs. recovered (in vivo) are shown in Table 16 below. A one-way
RM-ANOVA was performed on the results (F=169.2, p=0.006) and there were significant
differences in the means of the control treatment and HM (p≤0.01), HM and HY (p≤0.0001) and
HM and AM (p≤0.01). The largest difference between RS fed vs. recovered values for the 3
starches was for HM (-28.39 g) followed by HY (-6.52 g) and AM (0.14 g). Only 13.4% of the
RS fed in HM was recovered in ileal effluent and the percentage increased to 43.4% in HY.
Table 16 – Comparison of Resistant Starch Fed vs. Recovered: Treatment Control (g) HM (g) HY (g) AM (g)
Subject F Rb Fª Rbc Fª Rbc Fª Rbc AR001 0 0.9 32.8 7.78 11.5 9.02 0 0.37 AR002 0.28 1.68 1.84 -0.08 AR003 0.32 3.76 4.09 0.14 Mean 0 0.50 32.8 4.41 11.5 4.99 0 0.14
Mean Difference (R-F)* 0.50d -28.39e -6.52d 0.14d
SD for (R-F) 0.35 3.10 3.67 0.23 95% CI for (R-F) -0.36,1.36 -36.10, -20.69 -15.64, 2.61 -0.42, 0.70
ª ‘F’ are the fed fiber values based on in vitro analysis (AOAC Official Method 991.43) provided by the sponsor (g/50 g total carbohydrates) b ‘R’ are the recovered RS values in ileal effluent and analyzed based on AOAC Official Method 2002.02 c ‘R’ values are adjusted (R = RS recovered with starch – RS recovered on control day) Values with different letters differ significantly Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
70
5.3.6: Comparison of Carbohydrates by Difference Fed and Recovered
Amounts of CHOD fed (in vitro) vs. recovered (in vivo) are shown in Table 17 below. A one-
way RM-ANOVA was performed on the results (F=34.14, p=0.01) and there were significant
differences in the means of the control treatment and HM (p≤0.05), HM and HY (p≤0.05) and
HM and AM (p≤0.01). The largest difference between CHOD fed vs. recovered values for the 3
starches was for HM (-14.07 g) followed by HY (10.05 g) and AM (1.72 g). Approximately
57% of the CHOD fed in HM was recovered in ileal effluent and the percentage increased to
187% in HY.
Table 17 – Comparison of Carbohydrates by Difference Fed and Recovered: Treatment Control (g) HM (g) HY (g) AM (g)
Subject F Rb Fª Rbc Fª Rbc Fª Rbc AR001 0 11.83 32.8 24.13 11.5 22.14 0 9.01 AR002 8.69 15.29 24.52 -2.76 AR003 10.02 16.75 18.00 -1.08 Mean 0 10.18 32.8 18.73 11.5 21.55 0 1.72
Mean Difference (R-F)* 10.18d -14.07e 10.05d 1.72d
SD for (R-F) 1.58 4.74 3.30 6.37 95% CI for (R-F) 6.27, 14.10 -25.85, -2.31 1.86, 18.25 -14.09, 17.54
ª ‘F’ are the fed fiber values based on in vitro analysis (AOAC Official Method 991.43) provided by the sponsor (g/50 g total carbohydrates) b ‘R’ are the recovered CHOD values in ileal effluent and analyzed based on proximate analysis c ‘R’ values are adjusted (R = CHOD recovered with starch – CHOD recovered on control day) Values with different letters differ significantly Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
71
5.3.7: Amount of Fiber and Available Carbohydrates by Different Methods
The amount of fiber and avCHOs by different methods (in vitro, in vivo BGR study and in vivo
ileostomy study) are shown in Table 18 and Figure 10 below. The values for each method were
expressed as means only since a measure of variability was not provided for the in vitro values
and not possible from the in vivo BGR study as discussed in 4.2.6 above. Hence, a two-way
ANOVA and multiple comparisons were not relevant to perform on these type of data. The fiber
values for HY and AM from the ileostomy study are less than those from the BGR study and the
fiber value for HM from the ileostomy study was more than the fiber value from the BGR study.
Table 18 – Amount of Fiber and Available Carbohydrates by Different Methods:
Fiber Available Carbohydrates Method HM (g) HY (g) AM (g) Method HM (g) HY (g) AM (g) In vitroa 32.8 11.5 0 In vitroa 17.2 38.5 50 In vivo BGR
Studyb
12.5 40.8 6.2 In vivo BGR
Studyc
37.5 9.2 43.8
In vivo Ileostomy Studyd,e
18.7 21.6 1.7 In vivo Ileostomy
Studyf
31.3 28.4 48.3
ª Values provided by the sponsor (g/50 g total carbohydrates) and fiber analyzed based on AOAC Official Method 991.43 b Fiber (g) = 50 g total carbohydrates – avCHO (g) c Mean BGR values of equations 5 and 5a d Fiber values are the CHOD values from ileal effluent calculated from proximate analysis e Fiber values are adjusted (Fiber (g) = Calculated CHOD value (g) with starch – calculated CHOD value (g) on control day) f avCHO (g) = 50 g total carbohydrates – fiber (g) (CHOD value from ileal effluent calculated from proximate analysis)
Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
72
Figure 10 – Amount of Fiber and Available Carbohydrates by Different Methods:
5.3.8: Total Bacterial Count Determined by Real-Time Quantitative
Polymerase Chain Reaction
The total bacterial count (TBC) of the ileal effluent after each treatment is shown in Table 19
below. The TBC after each treatment indicated a mean value of 109 colony forming units
(CFU)/g of ileal effluent, substantially higher than the amount of total bacteria in the small
intestines of healthy subjects (<104 (CFU)/ml of intestinal fluid). A two way RM-ANOVA was
performed on the values and a significant main effect of subjects (F=49.5, p=0.02), treatment
(F=37, p=0.007) and subject x treatment interaction (F=37.4, p=0.0002) were observed.
AHMinvitro
HMBGR
HMileo
HYinvitro
HYBGRHYile
o
AMinvitro
AMBGR
AMileo
FHMinvitro
HMBGR
HMileo
HYinvitro
HYBGRHYile
o
AMinvitro
AMBGR
AMileo
0
5
10
15
20
25
30
35
40
45
50
Car
bohy
drat
es (g
) per
50g
tCH
O p
ortio
n of
Pro
duct
In vitroIn vivo BGRIn vivo Ileostomy
HM HY AM HM HY AMTreatment
Fiber Available Carbohydrates
A) Bars represent the mean fiber and available carbohydrate content of each treatment by each method B) Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
73
For AR001, there were significant differences between control and HM, HM and HY and HM
and AM (p≤0.0001) and control and AM (p≤0.05). No significant differences were observed for
AR002 and AR003 among the treatments and no significant differences were observed for all 3
subjects after each treatment.
Table 19 – Total Bacterial Count of Ileal Effluent:
Treatment Subject Control HM HY AM Mean±SD AR001 1.8x1010a 7.7x109b 2.0x1010ad 2.05x1010cd 1.64x1010±6.0x109
AR002 2.7x108 3.3x108 1.1x108 1.0x108 2.02x108±1.1x108
AR003 9.3x108 7.5x108 1.8x109 2.1x109 1.38x109±6.5x108
Mean±SD 6.2x109±9.8x109 2.9x109±4.1x109 7.3x109±1.1x1010 7.6x109±1.1x1010 Values are expressed as CFU/g of ileal effluent Values with different letters differ significantly Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca 5.3.8.1: Comparison of Log Transformed Values of Total Bacterial Count
The log transformed values of TBC for each subject after each treatment are shown in Table 20
below. A two-way RM ANOVA was performed on the data with subjects being the largest
source of variation (93.1%) (p<0.0001). There was a significant difference between log
transformed values of control and HM for Subject 1 (AR001) (p≤0.01), control and HY for
Subject 2 (AR002) (p≤0.01) and Subject 3 (AR003) (p≤0.05) and control and AM for Subject 2
(AR002) and Subject 3 (AR003) (p≤0.01). No significant differences were found among the log
transformed values after each treatment (F=0.84, p= 0.54).
74
Table 20 – Comparison of Log Transformed Values of Total Bacterial Count:
Values expressed as log count/g of ileal effluent Values with different letters differ significantly Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
5.3.8.2: Correlations Between Total Bacterial Count and Carbohydrate Measure
5.3.8.2.1: Correlation Between Total Bacterial Count and Resistant Starch
The correlation between TBC and RS is shown in Figure 11 below. Two correlations were
performed; one for the treatments that contained a higher content of RS (HM and HY) and the
other for the treatments that contained minimal or no RS (AM and control). For HM and HY,
the Pearson correlation coefficient r was 0.9646 (95% CI = 0.7049 to 0.9963 and R2 = 0.9305)
indicating a strong positive association between TBC, HM and HY. The association was
significant at two-tailed (p=0.002) and TBC accounted for 93% of the variation in the RS
measure for HM and HY ((0.9646)2 x 100% = 93%). The linear regression equation was
y=3.829x – 29.81 and the slope had a significant deviation from zero (F=53.56, p=0.002).
For AM and control, the Pearson correlation coefficient r was 0.9216 (95% CI = 0.4363 to
0.9915 and R2 = 0.8493) indicating a strong positive association between TBC, control and AM.
The association was significant at two-tailed (p=0.009) and TBC accounted for 85% of the
variation in the RS measure for AM and control ((0.9216)2 x 100% = 85%). The linear
Treatment Control HM HY AM
Subject Rep 1 Rep 2 Rep 1 Rep 2 Rep 1 Rep 2 Rep 1 Rep 2 AR001 10.18 10.30 9.81 9.94 10.26 10.34 10.23 10.38 Mean 10.25a 9.85b 10.30a 10.30a
AR002 8.32 8.51 8.48 8.56 7.99 8.08 7.92 8.08 Mean 8.40a 8.55a 8.05b 8.00b
AR003 8.95 8.99 8.88 8.87 9.20 9.28 9.28 9.36 Mean 8.95a 8.90a 9.25b 9.35b
Mean±SEM 9.21±0.48c 9.09±0.37c 9.19±0.59c 9.21±0.60c
75
regression equation was y=0.4134x – 3.235 and the slope had a significant deviation from zero
(F=22.54, p=0.009).
Figure 11 – Correlation Between Total Bacterial Count and Resistant Starch:
5.3.8.2.2: Correlation Between Total Bacterial Count and Carbohydrates by Difference
The correlation between TBC and CHOD is shown in Figure 12 below. Two correlations were
performed; one for the treatments that contained a higher content of dietary fiber (HM and HY)
and the other for the treatments that contained minimal or no dietary fiber (AM and control). For
HM and HY, the Pearson correlation coefficient r was 0.4955 (95% CI = -0.5286 to 0.9322 and
R2 = 0.2455) indicating a moderate positive association between TBC, HM and HY. The
association was not significant at two-tailed (p=0.32) and TBC accounted for 25% of the
variation in the CHOD measure for HM and HY ((0.4955)2 x 100% = 25%). The linear
regression equation was y=2.768x + 5.019 and the slope was not significantly deviated from zero
(F=1.3, p=0.32).
RS values are unadjusted Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.00
1
2
3
4
5
6
7
8
9
10
Total Bacteria (log count/g of ileal effluent)
Res
ista
nt S
tarc
h (g
)
1
1
1
12
22
2
33
33
Subject
r=0.9216
1 - AR0012 - AR0023 - AR003
Treatment
HMHYAM
r=0.9646
Control
76
For AM and control, the Pearson correlation coefficient r was 0.8068 (95% CI = -0.0139 to
0.9780 and R2 = 0.6509) indicating a strong positive association between TBC, control and AM.
The association was not significant at two-tailed (p=0.052) and TBC accounted for 65% of the
variation in the CHOD measure for AM and control ((0.8068)2 x 100% = 65%). The linear
regression equation was y=4.426x – 29.71 and the slope was not significantly deviated from zero
zero (F=7.5, p=0.052).
Figure 12 – Correlation Between Total Bacterial Count and Carbohydrates by Difference:
CHOD values are unadjusted Abbreviations: HM – Hi-maize® 260, HY – Hylon® VII, AM – Amioca
7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.00
5
10
15
20
25
30
35
40
Total Bacteria (log count/g of ileal effluent)
Car
bohy
drat
es b
y D
iffer
ence
(g)
r=0.8068
1
11
1
2
2
2
2
3 3
3
r=0.4955
3
Treatment
Subject1 - AR0012 - AR0023 - AR003
HMHYAM
Control
77
5.4: Discussion
Our results showed that dietary fiber content determined in vitro did not reflect the dietary fiber
content in vivo based on the results from the BGR and ileostomy studies. Even though results
from the ileostomy study were different from the BGR study, the conclusion remained consistent
that in vitro analysis underestimated fiber in HY and AM and overestimated fiber in HM.
The amount of ileal effluent collected from each subject was quite variable but a consistent
pattern was observed (Table 12). Subject 1 (AR001) had the highest output for both the control
and starch test days, followed by Subject 3 (AR003) and Subject 2 (AR002) had the lowest
output. Generally, all the subjects experienced lower output than usual due to prolonged GI
transit time but one subject experienced this more than the others. Subject 2 (AR002) initially
described her daily output as “very active”; however, repeatedly mentioned that her output
during the visits was very low due to the fact that her usual diet had three times more fiber than
what was provided in her meal plan. This likely caused a disturbance in her small intestinal
motility and subsequently, a great increase in her GI transit time. To clarify, Subject 2’s
(AR002) first visit consisted of AM as the treatment that she consumed at 8:30am. Due to low
output, her first collection was at 5:00 pm and we managed to collect 103 g of effluent in a 10
hour period compared with 690 g from AM for Subject 1 (AR001) and 231 g for Subject 3
(AR003). The mean±SD amount of ileal effluent collected for the control test days was
404.0±282.3 g, well within the range for healthy ileostomates reported previously by Englyst &
Cummings (1985), Englyst et al. (1996) and McNeil et al. (1982).
The mean percent dry matter produced in 10 hours (Table 14) for the control test days was
previously reported to be between 7.5-9% (Englyst & Cummings 1986; Englyst et al., 1996;
Silvester et al., 1995) and for the high-RS treatment, it was reported to be between 10-12%
78
(Englyst et al., 1996; Langkilde et al., 2002; Silvester et al., 1995). The mean percent dry matter
produced in 10 hours for the control test days was 9.2% and 15.7% for the high-RS starch (HM),
values higher than those previously reported. One explanation for the higher mean percent dry
matter values is because the control diets in this study had a mean of 8 g of fiber compared to 0 g
of fiber reported for the control diets in previous ileostomy studies (Englyst & Cummings, 1985;
1986; 1987). Hence, at baseline, the fiber values for the control diet were higher, leading to a
higher percentage of dry matter for all treatments. Although the control diets were required to be
polysaccharide-free, the source of fiber was from the prepackaged food items available on the
menu for the subjects’ meal plans. Great efforts were made by the study coordinator to provide
menu items with 0 g fiber; however, this proved to be difficult as prepackaged food items were
necessary for this study to reduce food handling.
As for proximate analysis (Table 15), some of the results of the protein and fat during the starch
test days were less than that from the control day whereby an adjustment results in negative
values. This can be largely due to the subjects’ prolongation in GI transit time caused by the fine
particle size of RS as previously reported by Cummings et al. (1996); hence, suggesting
macronutrient hyperabsorption in the small intestines.
Englyst et al. (1996), Schweizer et al. (1990) and Silvester et al. (1995) reported that the mean
RS recoveries in ileal effluent were 95-100% of what was fed to the subjects. When comparing
the amount of RS fed (in vitro value determined by AOAC 991.43) to the adjusted amount of RS
recovered (in vivo value determined by AOAC 2002.02), 13.4% and 43.3% of the RS in HM and
HY respectively were recovered in ileal effluent and AM contained more RS (0.14 g) than
accounted for in vitro (0 g) (Table 16). Our percentages were much less than what was
previously reported indicative that the in vitro method AOAC 991.43 did not accurately reflect
the true RS content of the investigational products. The comparison would have been much
79
more meaningful had the initial in vitro method been specific for RS content, i.e.: comparing in
vitro and in vivo values of RS both determined by AOAC 2002.02 method. Similarly, when
comparing the amount of CHOD fed to the adjusted amount of CHOD recovered, 57% and 187%
of the CHOD in HM and HY respectively were recovered in ileal effluent and AM contained
more CHOD (1.72 g) than accounted for in vitro (0 g) (Table 17). As claimed by the sponsor,
the dietary fiber component of the RS products is entirely from an RS source. If this claim is
true, the CHOD values and RS values for each treatment should be equal, i.e.: RS entirely
constitutes the dietary fiber component of the investigational products. Comparing CHOD and
RS values showed that HM and HY contained 4.2 and 4.3 times more CHOD than RS in the
products. A possible explanation for this phenomenon is that the RS in the products was partly
digested in the small intestine so was no longer analyzed as RS but was still considered CHOD.
This sheds light on the importance of measuring carbohydrates in novel starch products using a
combination of general and specific methods to detect any differences in total carbohydrates vs.
RS fraction and thus, reducing the risk of inaccurate fiber labeling.
Table 18 gives the final fiber and avCHO values for each treatment for each method (in vitro, in
vivo BGR study and in vivo ileostomy study). No one single method produced the highest fiber
or avCHO values for the treatments. The fiber values from the BGR study were not in line with
the fiber values from the ileostomy study. For example, the fiber value for HM from the
ileostomy study was 51% higher than the value from the BGR study. In terms of variation, the
%CV of the fiber values from the BGR and ileostomy studies were as follows: 29% for HM,
44% for HY and 81% for AM. As described in the discussion of Chapter 4 (section 4.4),
accurate knowledge of the product’s GI was necessary to produce accurate avCHO values which
in terms produce accurate fiber values. A slight increase or decrease in the product GI
calculation can result in big differences in the avCHO content of the products, affecting fiber
80
values. This is one possible scenario that can explain the variation in the fiber values of the BGR
and ileostomy studies.
There was an interest in the ileostomy study to determine the quantity of total bacteria in the
ileum of the subjects. It was reasoned that a strong negative correlation between TBC, CHOD
and RS may provide an explanation to lower amounts of recovered CHOD and RS in effluent
due to bacterial fermentation. The mean TBC after each treatment was 109 CFU/g of ileal
effluent and for the subjects, Subject 1 (AR001) had the highest TBC, 1010 CFU/g of ileal
effluent followed by Subject 3 (AR003), 109 CFU/g of ileal effluent and last, Subject 2 (AR002)
with 108 CFU/g of ileal effluent. TBC calculated for each treatment and for each subject
individually all warrant a diagnosis of small intestinal bacterial overgrowth (SIBO), defined as
>105 CFU/ml of intestinal fluid (DiBaise, 2008) and the subjects’ TBC were all substantially
higher than those of healthy patients (<104 CFU/ml of intestinal) (Quigley & Quera, 2006).
The TBC was transformed into log values and there was no consistent pattern observed with the
log transformed values, i.e.: they did not increase or decrease in the same manner in response to
a single starch treatment. All correlations between TBC, CHOD and RS were positive
confirming that SIBO was not a confounding factor in this study, i.e.: TBC had no effect on the
recovered CHOD and RS values.
The end of this discussion raises an important question: How reliable are in vitro methods for
determining the carbohydrate content of manufactured starch products? The answer is not simply
‘reliable’ or ‘not reliable.’ The answer needs to take into account which type of in vitro method
is employed to measure the carbohydrate fraction of interest. If the in vitro method takes into
account all dietary fiber components, ex: AOAC 2009.01 ITDF method, and is complemented
with a secondary method that measures the specific carbohydrate fraction of interest, ex: AOAC
81
2002.02 RS method, then the methods can be considered reliable and the food manufacturer need
not pursue further in vivo investigations. However, if the in vitro method has been proven to be
incompetent for the measurement of the specific carbohydrate fraction like RS, relevant in vitro
methods or in vivo methods can be used in order to produce an accurate measure of the fraction
of interest. In this case, AOAC 991.43 was not the relevant in vitro method to measure the TDF
content of the 3 investigational starch products.
82
Chapter 6.0 Conclusion and Implications
6.0: Conclusion and Implications
The first hypothesis of this project was that the AOAC Official Method 991.43 initially used in
vitro to determine the total dietary fiber content of the three investigational starch products will
underestimate the total dietary fiber content of the products by underestimating RS content. It is
concluded that the in vitro AOAC 991.43 method greatly underestimated the fiber content in HY,
marginally underestimated the fiber content in AM; however, overestimated the fiber content in
HM. Our first hypothesis was true for HY and AM but not for HM. The first hypothesis also
stated that the results from the BGR study will be accurately reflected in the ileostomy study, but
the results from the BGR study were not in line with the results from the ileostomy study due to
the sensitivity of the BGR method to slight changes in GI.
The second hypothesis stated that ileostomates will exhibit a higher quantity of total bacteria in
their ileum and it is concluded that ileostomates exhibited a higher than normal quantity of
bacteria in their ileum compared to data from healthy subjects in the literature.
The third hypothesis stated that a strong negative association between TBC, CHOD and RS will
provide an explanation for an underestimation in these components and it is concluded that since
the associations were positive, high TBC in the ileum of the ileostomates did not have an effect
on these components.
The main implication of this project reflects the importance of accurate dietary fiber labeling of
novel RS products. For manufacturers, it is crucial to identify the relevant general (ex: AOAC
2009.01 ITDF method) and specific (ex: AOAC 2002.02 RS method) in vitro methods used to
accurately determine the amount of RS in novel RS products. Accurate dietary fiber labeling for
83
RS products is not only important for the food ingredient itself, but for accurate labeling of the
food products which incorporate the RS products. For example, in order to obtain a significant
reduction in calories or an increase in dietary fiber per serving of cookies, 25-50% of refined
flour must be replaced by an RS ingredient (Haynes et al., 2013) such as HM. If one cookie
accounts for one serving and the average size of the cookie is 12g, 4g of HM (33%) is required to
satisfy a “high source of fiber” claim (Food and Drug Regulations (C.R.C., c.870, page 52),
2014). If the same HM product used in this study was used as an ingredient in these cookies, 4g
of HM would not have provided 4g of fiber in the form of RS because data from the ileostomy
study showed that only 13.4% (Table 17) of the RS fed was recovered. This would mean that
only 0.5g of the 4g of HM used could actually be considered as RS. Hence, since these cookies
do not satisfy the requirements for a “high source of fiber” claim, they are considered
mislabeled, i.e.: if a label is false or misleading in any particular way such as in product
composition (Consumer Packaging and Labelling Act (R.S.C.. 1985, c. C-38, page 3), 2014).
Accurate dietary fiber labeling is also essential for the health of people who suffer from chronic
diseases such as type 1 diabetes. If an individual with type 1 diabetes were to purchase a high-
RS containing product that really contains more avCHO than fiber, it may affect the amount of
insulin they need to inject or the amount of carbohydrates they need to consume along with the
RS product, i.e.: disturbing the delicate balance between carbohydrate counting and type and
amount of injected insulin units (National Diabetes Information Clearinghouse, 2014).
There is also a need for tighter regulations on a global basis on the methods used to determine
the total amount of dietary fiber in RS containing products for labeling purposes. Fortunately in
Canada, Health Canada in consultation with the Canadian Food Inspection Agency (CFIA) has
made it mandatory that the TDF content on nutrition labels needs to be determined by AOAC
84
2009.01 method because it takes into account not only the RS fraction in its entirety, but all the
other dietary fiber fractions as well (Health Canada, 2012).
The overall conclusions and implications of this project can be summed up in the following
points:
1) The new in vivo BGR method seems promising but might be carbohydrate specific, i.e.:
produces accurate results with some sources of carbohydrates and not others
2) Difficulty in finding conventional ileostomates for this project further emphasizes the
need for a new in vivo method for the determination of the avCHO and fiber content of
RS products
3) Manufacturers need to identify the relevant general and specific methods used to measure
the dietary fiber fraction of interest in their products
4) Wider recognition from health authorities that RS is one type of dietary fiber will call for
stricter regulations on the methodology used for determining the fiber content for
products high in RS content; thus, benefitting all consumers
85
Chapter 7.0 Limitations and Future Directions
7.0: Limitations and Future Directions
One of the limitations of study 1 was the sensitivity of the BGR method to slight changes in GI.
For the future, the method needs development in such a way where slight changes in GI won’t
produce great changes in avCHO content of the products of interest.
As for study 2, it was true that the sample size was small but it was sufficient to detect
statistically significant differences among the treatments as described in the results section 5.3.
Another limitation of study 2 was providing prepackaged food items as part of the study meal
plan in order to minimize food handling. As a result, finding prepackaged food items with 0g
fiber to satisfy the ‘polysaccharide-free control diet’ requirement was difficult. This limitation
could have been avoided had the study been conducted in a metabolic kitchen where
macronutrient content can be strictly regulated. The last limitation of study 2 was the time at
which the effluent was collected. We initially planned to collect effluent every 2 hours;
however, as a result of pooling the samples in the end, we decided to collect effluent when the
pouch was one third full. A change in diet as well as the addition of starch to the ileostomates
diet greatly increased their GI transit time and the first effluent collection was usually around 1
pm (starch was consumed at 8:30am). The ileostomates mentioned that by 1 pm on regular days,
they would have had to empty their pouches at least once if not more. This essentially means
that the collection of ileal effluent every 2 hours to prevent further enzymatic breakdown and
fermentation of the organic matter in the pouch wouldn’t have even been possible due to low
output.
86
The starch products had their own limitations too. It would have been best if a combination of
the relevant general (AOAC 2009.01) and specific (AOAC 2002.02) methods were used to
measure the dietary fiber and resistant starch fraction, respectively, of the products. Since we
used the AOAC 2002.02 method to measure the RS fraction in ileal effluent, it would have been
interesting to compare this value with the value obtained in vitro had this method been initially
used to determine RS content of the starch products. For study 2, we compared the RS value
obtained in vivo, and measured by AOAC 2002.02 method, with the fiber value obtained in vitro,
and measured by AOAC 991.43 method. Thus, we were comparing a specific method for a
dietary fiber fraction with a general method for the total measurement of dietary fiber and
whether this comparison is valid or not remains questionable.
As for future directions, it would be interesting to compare RS values when the chemical
analysis remains constant, i.e.: AOAC 2002.02, and the method of starch digestibility is variable,
i.e. in vitro vs. in vivo. For example, AOAC 2002.02 is used to measure the RS fraction in a
food product in vitro and this value is compared to the RS value recovered in vivo from ileal
effluent, also measured by AOAC 2002.02. Moreover, feedback from the ileostomates focused
on an interest in the role of RS, by its ability to prolong GI transit time, in the enhancement of
the athletic performance of athletes with an ileostomy and this can be extended to ileostomates in
the general public who are active outdoors (ex: camping, hiking, swimming and horseback
riding). This can also be applied to ileostomates who enjoy traveling and are required to be on
long plane rides where they can sleep comfortably without stressing about their pouches getting
too full and leaking. Increasing the confidence and quality of life of ileostomates by adding RS
to their daily diets is a great example of knowledge translation and customer benefits.
87
References About Resistant Starch. (2012). Retrieved August 5, 2014 from http://www.resistantstarch.com/ResistantStarch/About+RS/ Agriculture and Agri-Food Canada. (2014). Whole Grain Claims in the Marketplace. Retrieved on December 4, 2014 from http://www.agr.gc.ca/eng/industry-markets-and-trade/food-regulations/food-policy-and-regulatory-issues/reports-and-reviews/whole-grain-claims-in-the-marketplace/?id=1329944262611 Ahuja, G., Jaiswal, G., Chibbar, R.N. (2013). Starch Biosynthesis in Relation to Resistant Starch In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 1-22. West Sussex, UK: John Wiley & Sons, Ltd. Åkerberg, K.E., Liljberg, G.M., Granfeldt, Y.E., Drews, A.W., Bjorck, M.E. (1998). An in vitro method, based on chewing, to predict resistant starch content in foods allows parallel determination of potentially available starch and dietary fiber. Journal of Nutrition, 128, 651-660. Alexander, R.J. (1995). Starch in Plastics. Cereal Foods World, 41, 426-427. Anderson, G.H., Cho, C.E., Akhavan, T., Mollard, R.C., Luhovyy, B.L., Finocchiaro, E.T. (2010). Relation between estimates of cornstarch digestibility by the Englyst in vitro method and glycemic response, subjective appetite, and short-term food intake in young men. American Journal of Clinical Nutrition, 91, 932-939. Al-Tamimi, E., Seib, P.A., Snyder, B.S., Haub, M.D. (2010). Consumption of cross-linked resistant starch (RS4XL) on glucose and insulin responses in humans. Journal of Nutrition and Metabolism, doi: 10.1155/2010/651063. Arvola, A., Lähteenmäki, L., Dean, M., Vassallo, M., Winkelmann, M., Claupein, E., Saba, A., Shepherd, R. (2007). Consumers’ beliefs about whole and refined grain products in the UK, Italy and Finland. Journal of Cereal Science, 46(3), 197-206. Asp, N.G. (1992). Resistant Starch. Proceedings from the Second Plenary Meeting of EURESTA: European FLAIR Concerted Action No. 11 on Physiological Implications of the Consumption of Resistant Starch in Man. Preface. European Journal of Clinical Nutrition, 46(Suppl 2), S1-148. Aune, D., Chan, D.S., Lau, R. et al. (2011). Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies. British Medical Journal. 343: d6617. Behall, K.M., Scholfield, D.J., Hallfrisch, J.G., Liljeberg-Elmståhl, H.G.M. (2006). Consumption of Both Resistant Starch and ß-Glucan Improves Postprandial Plasma Glucose and Insulin in Women. Diabetes Care, 29(5), 976-981.
88
Berry, C.S. (1986). Resistant Starch: formation and measurement of starch that survives exhaustive digestion with amylolytic enzymes during the determination of dietary fibre. Journal of Cereal Science, 4, 301-314. Bingham, S.A., Day, N.E., Luben, R., Ferrari, P., Slimani, N., Norat, T., Clavel-Chapelon, F., Kesse, E., Nieters, A., Boeing, H., Tjϕnneland, A., Overvad, K., Martinez, C., Dorronsoro, M., Gonzalez, C., Key, T.J., Trichopoulou, A., Naska, A., Vineis, P., Tumino, R., Krogh, V., Bueno-de-Mesquita, H.B., Peeters, P.H.M. (2003). Dietary fibre in food and protection against colorectal cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC): an observational study. The Lancet, 361(9368), 1496-1501. Bakke, A., Vickers, Z. (2007). Consumer Liking of Refined and Whole Wheat Breads. Journal of Food Science, 72(7), S473-S480. Björck, I., Nyman, M., Pedersen, B., Siljeström, M., Asp, N.-G., Eggum, B.O. (1986). On the digestibility of starch in wheat bread studies in vitro and in vivo. Journal of Cereal Science, 4, 1-11. Bodinham, C.L., Frost, G.S., Robertson, M.D. (2010). Acute Ingestion of resistant starch reduces food intake in healthy adults. British Journal of Nutrition, 103, 917-922. Bond, J.H., Levitt, M.D. (1977). Use of breath hydrogen (H2) in the study of carbohydrate absorption. American Journal of Digestive Diseases, 22(4), 379-382. British Nutrition Foundation. (1990). Complex Carbohydrates in Foods. London, UK: Chapman & Hall. Brown, I.L., McNight, K.J., Moloney, E. (1995). Hi-maize™: new directions in starch technology and nutrition. Food Australia, 47, 272-275. Burn, J.1., Bishop, D.T., Mecklin, J.P., Macrae, F., Möslein, G., Olschwang, S., Bisgaard, M.L., Ramesar, R., Eccles, D., Maher, E.R., Bertario, L., Jarvinen, H.J., Lindblom, A., Evans, D.G., Lubinski, J., Morrison, P.J., Ho, J.W., Vasen, H.F., Side, L., Thomas, H.J., Scott, R.J., Dunlop, M., Barker, G., Elliott, F., Jass, J.R., Fodde, R., Lynch, H.T., Mathers, J.C. (2008). Effect of aspirin or resistant starch on colorectal neoplasia in the Lynch syndrome. The New England Journal of Medicine, 359(24), 2567-2578. Bustos, M.C., Perez, G.T., Leon, A.E. (2011). Effect of four types of dietary fiber on the technological quality of pasta. Food Science and Technology International, 17, 213-221. Champ, M. (1992). Determination of resistant starch in foods and food products: interlaboratory study. European Journal of Clinical Nutrition, 46(Suppl. 2), S51-S62.
89
Champ, M., Langkilde, A.-M., Brouns, F., Kettlitz, B., Le Bail-Collet, Y. (2003). Advances in dietary fibre characterization. 2. Consumption, chemistry, physiology and measurement of resistant starch; implications for health and food labeling, Nutrition Research Reviews, 16, 143-161. Champ, M., Martin, L., Noah, L., Gratas, M. (1999a). Analytical methods for resistant starch. In: Cho, S.S., Prosky, L., Dreher, M. (eds). Complex Carbohydrates in Foods, 169-187. Marcel Dekker, Inc., New York, U.S.A. Chandalia, M., Garg, A., Lutjohann, D., von Bergmann, K., Grundy, S.M., Crinkley, L.J. (2000). Beneficial Effects of High Dietary Fiber Intake in Patients with Type 2 Diabetes Mellitus. The New England Journal of Medicine, 342, 1392-1398. Consumer Packaging and Labelling Act (R.S.C.. 1985, c. C-38, page 3). (2014). Government of Canada – Justice Laws Website. Retrieved on December 10, 2014 from http://laws-lois.justice.gc.ca/eng/acts/C-38/page-3.html#h-5 Cummings, J.H., Beatty, E.R., Kingman, S.M., Bingham, S.A., Englyst, H.N. (1996). Digestion and physiological properties of resistant starch in the human large bowel. British Journal of Nutrition, 75(5), 733-747. DiBaise, J.K. (2008). Nutritional Consequences of Small Intestinal Bacterial Overgrowth. Practical Gastroenterology, 69, 15-28. Dietary Reference Intakes for Energy, Carbohydrates, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. (2005). Food and Nutrition Board of the Institute of Medicine of the National Academies. Retrieved December 3, 2014 from http://www.nap.edu/openbook.php?record_id=10490&page=339 Dronamraju, S.S., Coxhead, J.M., Kelly, S.B., Burn, J., Mathers, J.C. (2009). Cell Kinetics and gene expression changes in colorectal cancer patients given resistant starch: a randomized controlled trial. Gut, 58(3), 413-420. Ehle, F.R., Murphy, M.R., Clark, J.H. (1982). In situ particle size reduction and the effect of particle size on degradation of crude protein and dry matter in the rumen of dairy steers. Journal of Dairy Science, 65, 963-971. Englyst, H. N., Cummings, J.H. (1985). Digestion of the polysaccharides of some cereal foods in the human small intestine. American Journal of Clinical Nutrition, 42(5), 778-787. Englyst, H. N., Cummings, J.H. (1986). Digestion of the carbohydrates of banana (Musa paradisiaca sapientum) in the human small intestine. American Journal of Clinical Nutrition, 44, 42-50. Englyst, H. N., Cummings, J.H. (1987). Digestion of the polysaccharides of potato in the small intestine of man. American Journal of Clinical Nutrition, 45, 423-431.
90
Englyst, H.N., Kingman, S.M., Hudson, G.J., Cummings, J.H. (1996). Measurement of resistant starch in vitro and in vivo. British Journal of Nutrition, 75, 749-755. Englyst, H. N., Kingman, S.M., Cummings, J.H. (1992a). Classification and measurement of nutritionally important starch fractions. European Journal of Clinical Nutrition, 46, S33-S50. Englyst, H.N., Wiggins, H.S., Cummings, J.H. (1982). Determination of the Non-starch Polysaccharides in Plant Foods by Gas-Liquid Chromatography of Constituent Sugars as Alditol Acetates. Analyst, 107, 307-318. Erickson, W. (2005). The baker’s toolbox for developing health-promoting products. Cereal Foods World, 50, 6-8. Evans, A. (2013). Digestion Resistant Carbohydrates In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 95-109. West Sussex, UK: John Wiley & Sons, Ltd. Faisant, N., Buléon, A., Colonna, P., Molis, C., Lartigue, S., Galmiche, J.P., Champ, M. (1995a). Digestion of raw banana starch in the small intestine of healthy humans: structural features of resistant starch. British Journal of Nutrition, 73, 111-123. Faisant, N., Planchot, V., Kozlowski, F., Pacouret, P., Colonna, P., Champ, M. (1995b). Resistant starch determination adapted to products containing high levels of resistant starch. Sciences des Aliments, 15, 83-89. Food and Agriculture Organization. (1995). Dimensions of need - Staple foods: What do people eat? Retrieved December 1, 2014 from http://www.fao.org/docrep/u8480e/u8480e07.htm Food and Drug Regulations (C.R.C., c.870, page 52). (2014). Government of Canada – Justice Laws Website. Retrieved on December 10, 2014 from http://laws-lois.justice.gc.ca/eng/regulations/C.R.C.%2C%5Fc.%5F870/page-52.html#h-48 Food Innovation. (2014). Why Hi-maize® resistant starch. Retrieved August 26, 2014 from http://www.foodinnovation.com/foodinnovation/en-us/Hi-maize/Pages/Why-Hi-maize.aspx Fuentes-Zaragoza, E., Riquelme-Navarrete, M.J., Sánchez-Zapata, E., Pérez-Álvarez, J.A. (2010). Resistant starch as functional ingredient: A review. Food Research International, 43(4), 931-942. Furet, J.-P., Firmesse, O., Gourmelon, M., Bridonneau, C., Tap, J., Mondot, S., Doré, J., Corthier, G. (2009). Comparative assessment of human and farm animal faecal microbiota using real-time quantitative PCR. FEMS Microbiology Ecology, 68(3), 351-362. Fuwa, E. (1977). Digestion of various starch granules by amylase. Journal of the Japanese Society of Starch Science, 24, 128-140. Gilat, T., Ben Hur, H., Gelman-Malachi, E., Terdiman, R., Peled, Y. (1978). Alterations of the colonic flora and their effect on the hydrogen breath test. Gut, 19, 602-605.
91
Goñi, I., Manas, E., Garcia-Diz, F., Saura-Calixto, F. (1996). Analysis of resistant starch: a method for food and food products. Food Chemistry, 56, 445-449. Grubben, M.J.1., van den Braak, C.C., Essenberg, M., Olthof, M., Tangerman, A., Katan, M.B., Nagengast, F.M. (2001). Effect of resistant starch on potential biomarkers for colonic cancer risk in patients with colonic adenomas: a controlled trial. Digestive Diseases and Sciences, 46(4), 750-756. Han, J.-A., Lee, M.-J., Lim, S.-T. (2007). Utilization of oxidized and cross-linked corn starches in wheat flour batter. Cereal Chemistry, 84, 582-586. Hasjim, J., Ai, Y. Jane, J-l. (2013). Novel Applications of Amylose-Lipid Complex as Resistant Starch Type 5 In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 79-94. West Sussex, UK: John Wiley & Sons, Ltd. Hasjim, J., Lee, S.-O., Hendrich, S., Setiawan, S., Ai, Y., Jane, J. (2010). Characterization of a novel resistant-starch and its effects on postprandial plasma-glucose and insulin responses. Cereal Chemistry, 87(4), 257-262. Haub, M., Louk, J.A., Lopez, T.C. (2012). Novel resistant potato starches on glycemia and satiety in humans. Journal of Nutrition and Metabolism, doi: 10.1155/2012/478043. Haynes, L., Zimeri, J., Arora, V. (2013). Biscuit Baking and Extruded Snack Applications of Type III Resistant Starch In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 167-189. West Sussex, UK: John Wiley & Sons, Ltd. Health Canada. (2003). ARCHIVED – Nutrition Labelling and Claims Regulations – Frequently Asked Questions. Retrieved on August 26, 2014 from http://www.hc-sc.gc.ca/fn-an/label-etiquet/nutrition/reg/regulations-reglements-faqs-eng.php Health Canada. (2008). Grain Products – Canada’s Food Guide. Retrieved on December 4, 2014 from http://www.hc-sc.gc.ca/fn-an/food-guide-aliment/choose-choix/grain-cereal/index-eng.php Health Canada. (2012). Policy for Labelling and Advertising of Dietary Fibre-Containing Food Products. Retrieved August 26, 2014 from http://www.hc-sc.gc.ca/fn-an/alt_formats/pdf/legislation/pol/fibre-label-etiquetage-eng.pdf Heller, S.N., Hackler, L.R., Rivers, J.M., Soest, P.J.V., Roe, D.A., Lewis, B.A., Robertson, J. (1980). Dietary fiber: the effect of particle size of wheat bran on colonic function in young adult men. American Journal of Clinical Nutrition, 33, 1734-1744. Hijova, E. Chmelarova, A. (2007). Short chain fatty acids and colonic health. British Medical Journal, 108(8), 354-358. Hipsley, E.H. (1953). Dietary fibre and pregnancy toxaemia. British Medical Journal, 2, 420-422.
92
Horie, K., Tsutsumi, Y., Takao, Y., Suzukui, T. (2008). Calculation of repeatability and reproducibility for qualitative data. Retrieved October 26, 2014 from bm.nsysu.edu.tw/tutorial/iylu/ANQ%202008/11.../G5-01.doc Horwitz, William. (2007). Official Methods of Analysis of AOAC International. Gaithersburg, Maryland (USA): AOAC International. Howe, G.R, Benito, E., Castelleto, R., Cornée, J., Estève, J., Gallagher, R.P., Iscovich, J.M., Deng-ao, J., Kaaks, R., Kune, G.A., Kune, S., L'Abbé, K.A., Lee, H.P., Lee, M., Miller, A.B., Peters, R.K., Potter, J.D., Riboli, E., Slattery, M.L., Trichopoulos, D., Tuyns, A., Tzonou, A., Whittemore, A.S., Wu-Williams, A.H., Shu, Z. (1992). Dietary Intake of Fiber and Decreased Risk of Cancers of the Colon and Rectum: Evidence From the Combined Analysis of 13 Case-Control Studies. Journal of the National Cancer Institute, 84(24), 1887-1896. Hu, F.B., Manson, J.E., Stampfer, M.J., Colditz, G., Liu, S., Solomon, C.G., Willett, W.C. (2001). Diet, Lifestyle, and the Risk of Type 2 Diabetes Mellitus in Women. The New England Journal of Medicine, 345, 790-797. Hylla, S., Gostner, A., Dusel, G., Anger, H., Bartram, H.-P., Christl, S.U., Kasper, H., Scheppach, W. (1998). Effects of resistant starch on the colon in healthy volunteers: possible implications for cancer prevention. American Journal of Clinical Nutrition, 67, 136-142. Ileostomy. (2014). In Wikipedia, The Free Encyclopedia. Retrieved August 11, 2014 from http://en.wikipedia.org/wiki/Ileostomy Ingredion – AMIOCA TF (04400108). (2012). Product Data Sheet Confidential, Nutritional Information and Technical Service Bulletin. (Confidential documents provided by sponsor). Ingredion – HI-MAIZE® (22000B00). (2012). Product Data Sheet Confidential, Nutritional Information and Technical Service Bulletin. (Confidential documents provided by sponsor). Ingredion – HYLON® VII (04400B01). (2012). Product Data Sheet Confidential, Nutritional Information and Technical Service Bulletin. (Confidential documents provided by sponsor). International Food Information Council Foundation. (2014). 2014 Food & Health Survey – The Pulse of America’s Diet: From Beliefs to Behaviors. Retrieved on December 5, 2014 from http://www.foodinsight.org/sites/default/files/2014%20Food%20and%20Health%20Survey%20Full%20Report.pdf Jenkins, D.J.A, Cuff, D., Wolever, T.M.S., Knowland, D., Thompson, L., Cohen, Z., Prokipchuk, E. (1987). Digestibility of Carbohydrate Foods in an Ileostomate: Relationship to Dietary Fiber, in Vitro Digestibility, and Glycemic Response. American Journal of Gastroenterology, 82(8), 709-717.
93
Jenkins, D.J.A., Kendall, C.W.C., Vuksan, V., Vidgen, E., Parker, T., Faulker, T., Mehling, C.C., Garsetti, M., Testolin, G, Cunnane, S.C., Ryan, M.A., Corey, P.N. (2002). Soluble fiber intake at a dose approved by the US Food and Drug Administration for a claim of health benefits: serum lipid risk factors for cardiovascular disease assessed in a randomized controlled crossover trial. American Journal of Clinical Nutrition, 75(5), 834-839. Jenkins, D.J.A., Kendall, C.W.C., Augustin, L.S.A., Mitchell, S., Sahye-Pudaruth, S., Mejia, S.B., Chiavaroli, L., Mirrahimi, A., Ireland, C., Bashyam, B., Vidgen, E., de Souza, R.J., Sievenpiper, J.L., Coveney, J., Leiter, L.A., Josse, R.G. (2012). Effect of Legumes as Part of a Low Glycemic Index Diet on Glycemic Control and Cardiovascular Risk Factors in Type 2 Diabetes Mellitus – A Randomized Controlled Trial. Journal of the American Medical Association, 172(21), 1653-1660. Jenkins, D.J.A., Vuksan, V., Kendall, C.W.C., Würsch, P., Jeffcoat, R., Waring, S., Mehling, C.C., Vidgen, E., Augustin, L.S.A., Wong, Ev. (1998). Physiological Effects of Resistant Starches on Fecal Bulk, Short Chain Fatty Acids, Blood Lipids and Glycemic Index. Journal of the American College of Nutrition, 17(6), 609-616. Jiang, H., Campbell, M., Blanco, M., Jane, J. (2010b). Characterization of maize amylose-extender (ae) mutant starches. Part II: Structures and properties of starch residues remaining after enzymatic hydrolysis at boiling-water temperature. Carbohydrate Polymers, 80(1), 1-12. Johnston, K.L., Thomas E.L., Bell, J.D., Frost, G.S., Robertson, M.D. (2010). Resistant starch improves insulin sensitivity in metabolic syndrome. Diabetic Medicine, 27, 391-397. Korus, J., Witczak, M., Ziobro, R., Juszczak, L. (2009). The impact of resistant starch on characteristics of gluten-free dough and bread. Food Hydrocolloids, 23, 988-995. Langkilde, A.M., Andersson, H. (1995). In vivo quantification of resistant starch in EURESTA reference materials using the ileostomy model. In: Asp, N.G., van Amelsvoot, J.M.M., Hantvast, J.G.A.J. (eds). Proceedings of the concluding plenary meeting of EURESTA, 31-32. European Flair Concerted Action on Resistant Starch, No. 11 (COST 911). Langkilde, A.M., Champ, M., Andersson, H. (2002). Effects of high-resistant-starch banana flour (RS2) on in vitro fermentation and the small-bowel excretion of energy, nutrients, and sterols: an ileostomy study. American Journal of Clinical Nutrition, 75, 104-111. Leach, H.W., Schoch, T.J. (1961). Structure of the starch granule II. Action of various amylases on granular starches. Cereal Chemistry, 38, 34-46. Lee, B.M., Wolever, T.M.S. (1998). Effect of glucose, sucrose and fructose on plasma glucose and insulin responses in normal humans: comparison with white bread. European Journal of Clinical Nutrition, 52, 924-928. Lee, S.C., Prosky, L., DeVries, J.W. (1992). Determination of total, soluble, and insoluble dietary fiber in foods – Enzymatic-gravimetric method, MES-TRIS buffer: Collaborative study. Journal of the AOAC International, 75, 395-416.
94
Lee, S.C., Prosky, L. (1995). International survey on dietary fiber definition, analysis and reference materials. Journal of the AOAC International, 78, 22-36. Maki, K.C., Pelkman, C.L., Finocchiaro, E.T., Kelley, K.M., Lawless, A.L., Schild, A.L., Rains, T.M. (2012). Resistant starch from high-amylose maize increases insulin sensitivity in overweight and obese men. Journal of Nutrition, 142(4), 717-723. Maningat, C., Bassi, S.D., Ranjan, S., Gaul, J.A., Stempien, G.J., Dohl, C.T., Woo, K. (2008). High-protein, reduced-carbohydrate bakery and other food products. Australian Patent No. 2004253169. Maningat, C., Bassi, S., Woo, K., Dohl, C., Gaul, J., Stempien, G., Moore, T. (2005). Formulation of high-protein, high-fiber (low-carbohydrate), reduced calorie breads. AIB Research Department Technical Bulletin, 27(4), 1-16. Maningat, C.C., Seib, P.A. (2013). RS4-Type Resistant Starch: Chemistry, Functionality and Health Benefits In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 43-77. West Sussex, UK: John Wiley & Sons, Ltd. Martinez, I., Kim, J., Duffy, P.R., Schlegel, V.L., Walter, J. (2010). Resistant starches types 2 and 4 have differential effects on the composition of the fecal microbiota in human subjects. PLoS One, 5(11), e15046, doi: 10.1371/journal.pone.0015046. McCleary, B.V. (2007). An integrated procedure for the measurement of total dietary fibre (including resistant starch), non-digestible oligosaccharides and available carbohydrates. Analytical Bioanalytical Chemistry, 389, 291-308. McCleary, B.V. (2013). Measurement of Resistant Starch and Incorporation of Resistant Starch into Dietary Fibre Measurements In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 131-144. West Sussex, UK: John Wiley & Sons, Ltd. McCleary, B.V., McNally, M., Rossiter, O. (2002). Measurement of resistant starch by enzymatic digestion in starch samples and selected plant materials: Collaborative Study. Journal of the Association of Official Analytical Chemists, 85, 1103-1111. McCleary, B.V., Mills, C., Draga, A. (2009). Development and evaluation of an integrated method for the measurement of total dietary fibre. Quality Assurance and Safety of Crops and Foods, 1, 213-224. McCleary, B.V., Monaghan, D.A. (2002). Measurement of resistant starch. Journal of the Association of Official Analytical Chemists, 85, 665-675. McCleary, B.V., Rossiter, P. (2004). Measurement of novel dietary fibres. Journal of the Association of Official Analytical Chemists, 87, 707-717. McNeil, N.I., Bingham, S., Cole, T.J., Grant, A.M., Cummings, J.H. (1982). Diet and health of people with an ileostomy. II. Ileostomy function and nutritional state. British Journal of Nutrition, 47, 407-415.
95
Medallion Labs. (2014). Dietary Fiber. Retrieved December 3, 2014 from http://www.medallionlabs.com/TestOfferings/Fiber.aspx Used with permission from Medallion Labs. Megazyme. (2014). Measurement of Dietary Fiber. Retrieved December 3, 2014 from http://www.megazyme.com/resources/dietary-fiber/measurement-of-dietary-fiber Used with permission from Megazyme. Meyer, K.A., Kushi, L.H., Jacobs Jr., D.R., Slavin, J., Sellers, T.A., Folsom, A.R. (2000). Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. American Journal of Clinical Nutrition, 71(4), 921-930. Millard, P., Chesson, A. (1984). Modifications to swede (Brassica napus L.) anterior to the terminal ileum of pigs: some implications for analysis of dietary fibre. British Journal of Nutrition, 52, 583-594. Möller, J. (2011). Dietary Fibre – Codex Definitions and method of analysis. FOSS, P/N 1026514(1). Retrieved on December 5, 2014 from http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CB0QFjAA&url=http%3A%2F%2Fwww.foss.dk%2F-%2Fmedia%2F3ae835abb7c6472aace7e4fbb37e9cdb.ashx&ei=8CiCVIzuCMKyyATcjYLABg&usg=AFQjCNErg6dzs6uk7mquSjKsbEKoOrMd5Q&bvm=bv.80642063,d.aWw Morrison, W.R., Karkalas, J. (1990). Starch In P.M. Dey (ed)., Methods in Plant Biochemistry, 323-352. London, UK: Academic Press. Muir, J.G., Birkett, A., Brown, I., Jones, G., O’Dea, K. (1995a). Food processing and maize variety affects amounts of starch escaping digestion in the small intestine. American Journal of Nutrition, 61, 82-89. Muir, J.G., Lu, Z.X., Young, G.P., Cameron-Smith, D., Collier, G.R., O’Dea, K. (1995b). Resistant starch in the diet increases breath hydrogen and serum acetate in human subjects. American Journal of Clinical Nutrition, 61, 792-799. Muir, J.G., O’Dea, K. (1992). Measurement of resistant starch: factors affecting the amount of starch escaping digestion in vitro. American Journal of Clinical Nutrition, 56(1), 123-127. Muir, J.G., O’Dea, K. (1993). Validation of an in vitro assay for predicting the amount of starch that escapes digestion in the small intestine of humans. American Journal of Clinical Nutrition, 1993, 57, 540-546. Muir, J.G., Yeow, E.G., Keogh, J., Pizzey, C., Bird, A.R., Sharpe, K., O’Dea, K., Macrae, F.A. (2004). Combining wheat bran with resistant starch has more beneficial effects on fecal indexes than does wheat bran alone. American Journal of Clinical Nutrition, 79(6), 1020-1028.
96
Murphy, E.L., Calloway, D.H. (1972). The effect of antibiotic drugs on the volume and composition of intestinal gas from beans. Digestive Diseases, 17(7), 639-642. Murphy, N., Norat, T., Ferrari, P., Jenab, M., Bueno-de-Mesquita, B., Skeie, G., Dahm, C.C., Overvad, K., Olsen, A., Tjonneland, A., Clavel-Chapelon, F., Boutron-Ruault, M.C., Racine, A., Kaaks, R., Teucher, B., Boeing, H., Bergmann, M.M., Trichopoulou, A., Trichopoulos, D., Lagiou, P., Palli, D., Pala, V., Panico, S., Tumino, R., Vineis, P., Siersema, P., van Duijnhoven, F., Peeters, P.H., Hjartaker, A., Engeset, D., Gonzalez, C.A., Sanchez, M.J., Dorronsoro, M., Navarro, C., Ardanaz, E., Quiros, J.R., Sonestedt, E., Ericson, U., Nilsson, L., Palmqvist, R., Khaw, K.T., Wareham, N., Key, T.J., Crowe, F.L., Fedirko, V., Wark, P.A., Chuang, S.C., Riboli, E. (2012). Dietary fibre intake and risks of cancers of the colon and rectum in the European prospective investigation into cancer and nutrition (EPIC). PLoS ONE, 7(6):e39361. Myers, A.M., Morell, M.K., James, M.G., Ball, S.G. (2000). Recent progress toward understanding biosynthesis of the amylopectin crystal. Plant Physiology, 122, 989-997. Nagengast, F.M., Grubben, M.J., van Munster, I.P. (1995). Role of bile acids in colorectal carcinogenesis. European Journal of Cancer, 31A(7-8), 1067-1070. National Diabetes Information Clearinghouse. (2014). National Institute of Diabetes and Digestive and Kidney Disease – National Institutes of Health. Retrieved on December 10 from http://diabetes.niddk.nih.gov/dm/pubs/carbohydrate_ez/index.aspx Natural Hi-maize Resistant Starch. (2012). How can Hi-maize® Resistant Starch help me? Retrieved on December 5, 2014 from http://www.hi-maize.com/HiMaize/USA/Consumers/How+can+Hi-maize+help+me/ Nilsson, A.C., Ostman, E.M., Holst, J.J., Björck, I.M. (2008). Including indigestible carbohydrates in the evening meal of healthy subjects improves glucose tolerance, lowers inflammatory markers, and increases satiety after a subsequent standardized breakfast. Journal of Nutrition, 138(4), 732-739. Nugent, A.P. (2005). Health properties of resistant starch. Nutrition Bulletin, 30(1), 27-54. Office of Disease Prevention and Health Promotion. (2011). Dietary Guidelines for Americans, 2010. Retrieved December 4, 2014 from http://www.health.gov/dietaryguidelines/2010.asp Perman, J.A., Modler, S., Olson, A.C. (1981). Role of pH in production of hydrogen from carbohydrates by colonic bacterial flora. Journal of Clinical Investigation, 67, 643-650. Phillips, J, Muir, H.G., Birket, J., Lu, Z.X., Jones, G.P., O’Dea, K., Young, G.P. (1995). Effect of resistant starch on fecal bulk and fermentation-dependent events in humans. American Journal of Clinical Nutrition, 62(1), 121-130. PROMITOR™ Dietary Fiber. (2011). PROMITOR™ Resistant Starch. Retrieved August 26, 2014 from http://www.promitorfiber.com/promitor_dietary_fibers/pages/promitor_resistant_ starch.aspx
97
Prosky, L., Asp, N.-G., Furda, I., DeVries, J.W., Schweizer, T.F., Harland, B.F. (1985). Determination of total dietary fiber in food and food products: collaborative study. Journal of the AOAC International, 68, 677-679. Quigley, E., Quera, R. (2006). Small intestinal bacterial overgrowth: roles of antibiotics, prebiotics, and probiotics. Gastroenterology, 130 (2 Suppl 1): S78-90. Raben, A., Tagliabue, A., Christensen, N.J., Madsen, J., Holst, J.J. and Astrup, A. (1994). Resistant starch: the effect on postprandial glycemia, hormonal response, and satiety. American Journal of Clinical Nutrition, 60, 544-551. Read, N.W., Al-Janabi, M.N., Bates, T.E., Holgate, A.M., Cann, P.A., Kinsman, R.I., McFarlane, A., Brown, C. (1985). Interpretation of the breath hydrogen profile obtained after ingesting a solid meal containing unabsorbable carbohydrates. Gut, 26, 834-842. Robertson, J.A. (1988). Physicochemical characteristics of food and the digestion of starch and dietary fibre during gut transit. Proceedings of the Nutrition Society, 47, 143-52. Robertson, M.D. (2012). Dietary-resistant starch and glucose metabolism. Current Opinion in Clinical Nutrition and Metabolic Care, 15(4), 362-367. Robertson, M.D., Bickerton, A.S., Dennis, A.L., Vidal, H., Frayn, K.N. (2005). Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. American Journal of Clinical Nutrition, 82(3), 559-567. Robertson, M.D., Currie, J.M., Morgan, L.M., Jewell, D.P., Frayn, K.N. (2003). Prior short-term consumption of resistant starch enhances postprandial insulin sensitivity in healthy subjects. Diabetologia, 46(5), 659-665. Rumessen, J.J. (1992). Hydrogen and methane breath tests for evaluation of resistant carbohydrates. European Journal of Clinical Nutrition, 46(Suppl. 2), S77-S90. Sajilata, M.G., Singhal, R.S., Kulkarni, P.R. (2006). Resistant Starch – a review. Comprehensive Reviews in Food Science and Food Safety, 5(1), 1-17. Saltzberg, D.M., Levine, G.M., Lubar, C. (1988). Impact of Age, Sex, Race, and Functional Complaints on Hydrogen (H2) Production. Digestive Diseases and Sciences, 33(3), 308-313. Sandberg, A.S., Andersson, H., Hallgren, B., Hasselblad, K., Isaksson, B., Hultén L. (1981). Experimental model for in vivo determination of dietary fibre and its effect on the absorption of nutrients in the small intestine. British Journal of Nutrition, 45(2), 283-294. Sandstedt, R.M., Strahan, D., Ueda, S., Abbott, R.C. (1962). The digestibility of high-amylose corn starches compared to that of other starches. The apparent effect of the ae gene on susceptibility to amylase action. Cereal Chemistry, 39, 123-131.
98
Schweizer, T.F., Andersson, H., Langkilde, A.M., Reimann, S., Torsdottir, I. (1990). Nutrients excreted in ileostomy effluents after consumption of mixed diets with beans or potatoes. II. Starch, dietary fibre and sugars. European Journal of Clinical Nutrition, 44, 567-575. Schweizer, T.F., Walter, E., Venetz, P. (1988). Collaborative study for the enzymatic, gravimetric determination of total dietary fibre in foods. Mitteilungen aus der Gebiete der Lebensmitteluntersuchung und Hygiene, 79, 57-68. Shi, Y-C., Maningat, C.C. (2013). Resistant Starch: Sources, Applications and Health Benefits. West Sussex, UK: John Wiley & Sons, Ltd. Shukri, R., Seib, P.A., Maningat, C.C., Shi, Y.-C. (2013). In Vitro Enzymatic Testing Method and Digestion Mechanism of Cross-Linked Wheat Starch In Y.-C. Shi & C.C. Maningat (eds)., Resistant Starch: Sources, Applications and Health Benefits, 145-166. West Sussex, UK: John Wiley & Sons, Ltd. Silvester, K.R., Englyst, H.N., Cummings, J.H. (1995). Ileal recovery of starch from whole diets containing resistant starch measured in vitro and fermentation of ileal effluent. American Journal of Clinical Nutrition, 62, 403-411. Simonne, A.H., Simonne, E.H., Eitenmiller, R.R., Mills, H.A., and Cresman, C.P. (1997). Could the Dumas Method Replace the Kjeldahl Digestion for Nitrogen and Crude Protein Determination in Foods? Journal of the Science of Food and Agriculture, 73(1), 39-45. Simrén, M., Stotzer, P.O. (2006). Use and Abuse of hydrogen breath tests. Gut, 55, 297-303. Singh, N., Arioli, S., Wang, A., Villa, C., Jahani, R., Song, Y.S., Mora, D., Guglielmetti, S., Comelli, E.M. (2013). Impact of Bifiobacterium bifidum MIMBb75 on mouse intestinal microorganisms. FEMS Microbiology Ecology, 85(2), 369-375. Statistics Canada. (2014). Corn: Canada’s third most valuable crop. Retrieved November 4, 2014 from http://www.statcan.gc.ca/pub/96-325-x/2014001/article/11913-eng.htm#wb-tphp Sun, Q., Spiegelman, D., van Dam, R.M., Holmes, M.D., Malik, V.S., Willett, W.C., Hu, F.B. (2010). White rice, brown rice, and risk of type 2 diabetes in US men and women. Archives of Internal Medicine, 170(11), 961-969. Suzuki, M.T., Taylor, L.T., DeLong, E.F. (2000). Quantitative analysis of small-subunit rRNA genes in mixed microbial populations via 5’- nuclease assays. Applied Environmental Microbiology, 11, 4605-4614. Tomlin, J., Read, N.W., Edwards, C.A., Duerden, B.I. (1986). The degradation of guar gum by a faecal incubation system. British Journal of Nutrition, 55, 481-486. Topping, D. (2007). Cereal complex carbohydrates and their contribution to human health. Journal of Cereal Science, 46, 220-229.
99
Topping, D.L. & Clifton, P.M. (2001). Short-chain fatty acids and human colonic function: roles of resistant starch and nonstarch polysaccharides. Physiological Reviews, 81(3), 1031-1064. Topping, D.L., Fukushima, M., Bird, A.R. (2003). Resistant starch as a prebiotic and synbiotic: state of the art. Proceedings of the Nutrition Society, 62(1), 171-176. Trowell, H.C., Southgate, D.A.T., Wolever, T.M.S., Leeds, A.R., Gassull, M.A., Jenkins, D.J.A. (1976). Dietary fiber redefined. Lancet, 1, 967. Turowski, M., Deshmukh, B., Harfmann, R., Conklin, J., Lynch, S. (2007). A method for determination of soluble dietary fiber in methylcellulose and hydroxypropyl methylcellulose food gums. Journal of Food Composition and Analysis, 20, 420-429. United Ostomy Associations of America, Inc. (2011). Ileostomy Guide. Retrieved on December 6, 2014 from http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0CCwQFjAC&url=http%3A%2F%2Fwww.ostomy.org%2Fuploaded%2Ffiles%2Fostomy_info%2FIleostomyGuide.pdf&ei=0y2DVPDbO9KPyATlxoGQCw&usg=AFQjCNFYfyCVgpSyRcikL1uY2VMSEwSQWg&bvm=bv.80642063,d.aWw van Gorkom, B.A.P., Karrenbeld, A., van der Sluis, T., Zwart, N., van der Meer, R., de Vries, E.G.E., Kleibeuker, J.H. (2002). Calcium or resistant starch does not affect colonic epithelial cell proliferation throughout the colon in adenoma patients: A randomized controlled trial. Nutrition and Cancer – an International Journal, 43(1), 31-38. van Munster, I.P., de Boer, H.M., Jansen, M.C., de Haan, A.F., Katan, M.B., van Amelsvoort, J.M., Nagengast, F.M. (1994). Effect of resistant starch on breath-hydrogen and methane excretion in healthy volunteers. American Journal of Clinical Nutrition, 59, 626-630. Vogelsang, H., Ferenci, P., Frotz, S., Meryn, S., Gangl, A. (1988). Acidic colonic microclimate – possible reason for false negative hydrogen breath tests. Gut, 29, 21-26. Whole Grains Council. (2013). Whole Grain Statistics. Retrieved on December 4, 2014 from http://wholegrainscouncil.org/newsroom/whole-grain-statistics Wolever, T.M.S. (2006). The Glycaemic Index: A Physiological Classification of Dietary Carbohydrates. Wallingford, UK: CABI. pp: 18, 66. Wolever, T.M.S., Bolognesi, C. (1996a). Source and amount of carbohydrate affect postprandial glucose and insulin in normal subjects. Journal of Nutrition, 126, 2798-2806. Wolever, T.M., Cohen, Z., Thompson, L.U., Thorne, M.J., Jenkins, M.J., Prokipchuk, E.J., Jenkins, D.J. (1986). Ileal loss of available carbohydrate in man: comparison of a breath hydrogen method with direct measurement using a human ileostomy model. American Journal of Gastroenterology, 81(2), 115-122. Wolever, T.M.S., Gibbs, A.L., Spolar, M., Hitchner, E.V., Heimowitz, C. (2006). Equivalent glycemic load (EGL): a method for quantifying the glycemic responses elicited by low carbohydrate foods. Nutrition and Metabolism, 3:33.
100
Woo, K.S., Maningat, C.C., Seib, P.A. (2009). Increasing dietary fiber in foods: The case for phosphorylated cross-linked resistant starch, a highly concentrated form of dietary fiber. Cereal Foods World, 54(5), 217-223. Willis, H.J., Eldridge, A.L., Beiseigel, J., Thomas, W., Slavin, J.L. (2009). Greater satiety response with resistant starch and corn bran in humans subjects. Nutrition Research, 29(2), 100-105. Yao, B., Fang, H., Xu, W., Yan, Y. et al. (2014). Dietary fiber intake and risk of type 2 diabetes: a dose-response analysis of prospective studies. European Journal of Epidemiology. 29(2):79-88.
101
Appendix 1
Ileostomy Study Menu
Subjects can choose from a variety of food items on this menu. Subjects can choose from 1, 2 or all categories of each meal as desired.
Food items chosen and the amount desired by the subjects will remain constant for all 4 test days. Please feel free to ask the study staff questions about any of the food items listed below.
Meal Category Food Item* Amount Energy (cal)
Carbs (g)
Fiber (g)
Starch (g)
Sugars (g)
Protein (g)
Fat (g)
Dinnera Frozen Entréesb
Meat Options
Vegetarian
Options
Healthy Choice Gourmet Steamers: Grilled Balsamic
Chicken
Per entrée 283g
320 41 4 N/Ac 10 21 8
Blue Menu Italian Lasagna
Per entrée 320g
340 46 4 N/A 9 20 8
Blue Menu Roasted Vegetable
Lasagna
Per entrée 300g
310 46 5 N/A 10 15 7
Blue Menu Macaroni & 3
Cheeses
Per entrée 300g
360 50 5 N/A 4 20 9
Dessertb President’s Choice Banana Cake with Chocolate Chips
Per 1 cake 30g
130 16 1 N/A 10 2 6
President’s Choice Brownies with Dark
Chocolate Chips
Per 1 brownie 35g
150 23 1 N/A 16 2 5
Fruit Juiceb Minute Maid 100% Apple Juice
from Concentrate
1 package 200ml
90 23 N/A N/A 21 0.3 0
Minute Maid 100% Orange Juice
from Concentrate
1 package 200ml
100 23 N/A N/A 20 1 0
102
Meal Category Food Item* Amount Energy (cal)
Carbs (g)
Fiber (g)
Starch (g)
Sugars (g)
Protein (g)
Fat (g)
Breakfast Test meal is consumed during breakfast by mixing it with waterd, milke, yogurt, pudding, fruit juice or the meal-replacement
Cheeseb
Milk-based
Cheese-
alternative
Blue Menu Swiss Cheese (8 slices)
1 slice 19g
60 0 0 N/A 0 7 3
Old Fort Cheddar Cheese (8 slices)
1 slice 20g
90 1 0 N/A 0 5 7
Galaxy Nutritional Foods
Canadian Flavor Slices (8 slices)
1 slice 21g
50 1 0 N/A 0 4 3
Galaxy Nutritional Foods
Italian Flavor Slices (8 slices)
1 slice 21g
50 1 0 N/A 0 4 3
Yogurtb
Milk-based
Soy-based
Pudding
Liberté Greek Yogourt Vanilla
1 container 100g
90 12 0 N/A 11 8 2
Liberté Greek Yogourt
Strawberry & Banana
1 container 100g
90 12 1 N/A 11 7 1.5
SoyGo Vanilla (240g)
Per 125g 103 12 2 N/A 6 5 4
SoyGo Blueberry (240g)
Per 125g 100 12 2 N/A 6 5 4
Snack Pack Vanilla Pudding
Per 1 cup 99g
120 20 0 N/A 15 1 3.5
Snack Pack Chocolate Pudding
Per 1 cup 99g
120 21 1 N/A 16 2 3.5
Fruitsf Fresh Fruit Saladg 320g 160 70 3.5 N/A 33 2.5 0.7 Pastryb Strawberry Cheese
Fruit Bites 3 pastries
60g 200 21 1 N/A 4 3 12
Mini Braided Apple Strudel
1 strudel 71g
250 28 1 N/A 5 3 14
(Continued)
103
Meal Category Food Item* Amount Energy
(cal) Carbs
(g) Fiber
(g) Starch
(g) Sugars
(g) Protein
(g) Fat (g)
Breakfast Fruit Juiceb Minute Maid 100% Apple Juice
from Concentrate
1 package 200ml
90 23 N/A N/A 21 0.3 0
Minute Maid 100% Orange Juice
from Concentrate
1 package 200ml
100 23 N/A N/A 20 1 0
Meal-Replacementb
Ensure Nutrition Shake
Milk Chocolate
1 bottle 240ml
250 40 1 N/A 22 9 6
Ensure Nutrition Shake
Vanilla
1 bottle 240ml
250 40 0 N/A 23 9 6
Mid-morning Snack: Subject can choose to snack from any of the following categories: cheese, yogurt or pudding along with fruit juice. They also have the option of choosing a meal-replacement.
104
Meal Category Food Item* Amount Energy (cal)
Carbs (g)
Fiber (g)
Starch (g)
Sugars (g)
Protein (g)
Fat (g)
Lunch Sandwichh
Deli Meat Optionsb
Wonder Classic Thin Sandwich Soft White
Bread
2 slices 140 27 1 N/A 5 4 1.5
Blue Menu Oven Roasted Turkey
Breast Fully Cooked
(175g)
3 slices 66g
70 2 0 N/A 0 14 1
Blue Menu Extra Lean Stone
Roasted Ham (175g)
3 slices 53g
60 2 0 N/A 1 11 1
Cheese included in this sandwich is the same cheese or cheese-alternative the subject has chosen for breakfast (8-slice package).
Fruit Juiceb Minute Maid 100% Apple Juice from
Concentrate
1 package 200ml
90 23 N/A N/A 21 0.3 0
Minute Maid 100% Orange Juice from
Concentrate
1 package 200ml
100 23 N/A N/A 20 1 0
* All food items are from Loblaws (60 Carlton Street, Toronto, ON M5B 1L1). a Dinner meal consumed the night before the test day. b Subject will choose 1 option only. c N/A – Information not available. d Water can be consumed as much as needed throughout the day. e Subjects are responsible for providing their own milk, tea, coffee, sugar or sweetener. These items can be consumed as desired throughout the day. f Fruits may vary depending on the season. g Product did not contain a nutrition label and was not found in the Canadian Nutrient File, version 2010. Nutritional information was found in DailyBurn Tracker Food Database (Source: No Author. (2013). Loblaws Fruit Salad. DailyBurn Tracker. Retrieved May 26, 2013, from http://tracker.dailyburn.com/recipes/loblaws_fruit_salad). h Subjects can choose to have one or two sandwiches.
105
Appendix 2 Control Diet of Subject 1 (AR001)
Subject 1 (AR001) preferred the inclusion of animal products (i.e.: meat and cow’s
milk) in her meal plan because of the satiating effect of animal proteins. Her meal
plan is outlined in the table below.
Meal Food Itema Amount Energy (cal)
tCHOb (g)
Fiber (g)
Protein (g)
Fat (g)
Dinner Italian Lasagna 320 g 340 46 4 20 8 Apple Juice 200 ml 90 23 N/Ac 0.3 0
Diet Gingerale 355 ml 0 0 0 0 0 Breakfast Cheddar Cheese 78.5 g 353.3 3.9 0 19.6 27.5
Vanilla Pudding 90 g 109.1 18.2 0 0.9 3.2 Coffeed 340 ml 2 0 0 0.4 0.06 2% milk 160 ml 83.2 7.6 0 5.8 3.2
Sugar 15 g 60 15 0 0 0 Snack Orange Juice 200 ml 100 23 N/Ac 1 0 Lunch Thin White
Bread 54 g 140 27 1 4 1.5
Oven Roasted Turkey Breast
48 g 50.9 1.5 0 10.2 0.7
Orange Juice 200 ml 100 23 N/Ac 1 0 Other Coffeed 170 ml 1 0 0 0.2 0.03
2% milk 80 ml 41.6 3.8 0 2.9 1.6 Sugar 7.5 g 30 7.5 0 0 0
TOTAL 1,501 200 5 66 46 a Food item’s nutrient composition taken from nutrition label unless otherwise stated b Total Carbohydrates
c N/A: Not Available d Coffee, brewed / Café infuse (Food Code: 2873). Canadian Nutrient File, 2010 <http://webprod3.hc-sc.gc.ca/cnf-fce/report-rapport.do?lang=eng>.
106
Appendix 3 Control Diet of Subject 2 (AR002)
Due to Subject 2’s (AR002) religious beliefs, all food items in her meal plan were
carefully checked and selected by her. Her meal plan is outlined in the table below.
Meal Food Itema Amount Energy (cal)
tCHOb (g)
Fiber (g)
Protein (g)
Fat (g)
Dinner Roasted Vegetable Lasagna
300 g 310 46 5 15 7
Apple Juice 200 ml 90 23 N/Ac 0.3 0 Breakfast Swiss Cheese 38.5 g 121.6 0 0 14.2 6.1
Raspberry Vanilla Yoghurt
90 g 72 10.8 0 4.5 1.8
Apple Strudel 70 g 246.5 27.6 0.99 2.96 13.8 Coffeed 170 ml 1 0 0 0.2 0.03 2% milk 80 ml 41.6 3.8 0 2.9 1.6
Snack Swiss Cheese 52.5 g 165.8 0 0 19.3 8.3 Coffeed 170 ml 1 0 0 0.2 0.03 2% milk 80 ml 41.6 3.8 0 2.9 1.6
Lunch Thin White Bread
54 g 140 27 1 4 1.5
Swiss Cheese 38.5 g 121.6 0 0 14.2 6.1 Apple Juice 200 ml 90 23 N/Ac 0.3 0
Other Coffeed 170 ml 1 0 0 0.2 0.03 2% milk 80 ml 41.6 3.8 0 2.9 1.6
Thin White Bread
54 g 140 27 1 4 1.5
Swiss Cheese 18 g 56.8 0 0 6.6 2.8 TOTAL 1,682 196 8 95 54
a Food item’s nutrient composition taken from nutrition label unless otherwise stated b Total Carbohydrates
c N/A: Not Available d Coffee, brewed / Café infuse (Food Code: 2873). Canadian Nutrient File, 2010 <http://webprod3.hc-sc.gc.ca/cnf-fce/report-rapport.do?lang=eng>.
107
Appendix 4 Control Diet of Subject 3 (AR003)
Subject 3 (AR003) suffers from eosinophilic esophagitis and restricts soy, mushroom,
red peppers among other food items in her diet. All the ingredients in her meal plan
were meticulously checked and approved by her before the commencement of her
visits. Her meal plan is outlined in the table below.
Meal Food Itema Amount Energy (cal)
tCHOb (g)
Fiber (g)
Protein (g)
Fat (g)
Dinner Macaroni & 3 Cheese
300 g 360 50 5 20 9
Apple Juice 200 ml 90 23 N/Ac 0.3 0 Breakfast Cheddar Cheese 38.5 g 173.3 1.9 0 9.6 13.5
Chocolate Pudding
90 g 109.1 19.1 0.91 1.8 3.2
Apple Strudel 70 g 246.5 27.6 0.99 2.96 13.8 Blueberriesd 50 g 28.5 7.3 1.3 0.37 0.17 Red Grapese 50 g 34.5 9 0.5 0.5 0 Apple Juice 200 ml 90 23 N/Ac 0.3 0
Snack Cheddar Cheese 38.5 g 173.3 1.9 0 9.6 13.5 Blueberriesd 25 g 14.3 3.7 0.7 0.19 0.09 Red Grapese 25 g 17.3 4.5 0.25 0.25 0 Apple Juice 200 ml 90 23 N/Ac 0.3 0
Lunch Thick White Bread
54 g 135 26.3 1.5 4.5 1.2
Oven Roasted Turkey Breast
48 g 50.9 1.5 0 10.2 0.7
Cheddar Cheese 38.5 g 173.3 1.9 0 9.6 13.5 Apple Juice 200 ml 90 23 N/Ac 0.3 0
TOTAL 1,876 247 11 71 69 a Food item’s nutrient composition taken from nutrition label unless otherwise stated b Total Carbohydrates
c N/A: Not Available d Blueberry, raw / Bleuet, cru (Food Code: 1705). Canadian Nutrient File, 2010 <http://webprod3.hc-sc.gc.ca/cnf-fce/report-rapport.do?lang=eng>. e Grape, red or green (European type, such as Thompson seedless), adherent skin, raw / Raisin, rouge ou vert (Type Européen, comme Thompson sans grains), peau adhérente, cru (Food Code: 1718). Canadian Nutrient File, 2010 <http://webprod3.hc-sc.gc.ca/cnf-fce/report-rapport.do?lang=eng>.