Content uploaded by Jayashree Arcot
Author content
All content in this area was uploaded by Jayashree Arcot
Content may be subject to copyright.
A Preliminary
Assessment of
the Glycemic
Index of Honey
A report for the Rural Industries Research
and Development Corporation
by Dr Jayashree Arcot and
Prof Jennie Brand-Miller
March 2005
RIRDC Publication No 05/027
RIRDC Project No UNS-17A
ii
© 2005 Rural Industries Research and Development Corporation.
All rights reserved.
ISBN 1 74151 126 7
ISSN 1440-6845
A preliminary assessment of the Glycemic Index of honey
Publication No. 05/027
Project No. UNS-17A
The information contained in this publication is intended for general use to assist public knowledge and discussion
and to help improve the development of sustainable industries. The information should not be relied upon for the
purpose of a particular matter. Specialist and/or appropriate legal advice should be obtained before any action or
decision is taken on the basis of any material in this document. The Commonwealth of Australia, Rural Industries
Research and Development Corporation, the authors or contributors do not assume liability of any kind
whatsoever resulting from any person's use or reliance upon the content of this document.
This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the
Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the Publications
Manager on phone 02 6272 3186.
Researcher Contact Details
Dr. Jayashree Arcot
Department of Food Science and Technology
University of New South Wales
Phone: 02 9385 5360
Fax: 02 9385 5931
Email: j.arcot@unsw.edu.au
Prof. Jennie Brand Miller
Human Nutrition Unit
University of Sydney
Phone: 02 9351 3759
Fax: 02 9351 6022
Email: j.brandmiller@biochem.usyd.edu.au
In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form.
RIRDC Contact Details
Rural Industries Research and Development Corporation
Level 1, AMA House
42 Macquarie Street
BARTON ACT 2600
PO Box 4776
KINGSTON ACT 2604
Phone: 02 6272 4819
Fax: 02 6272 5877
Email: rirdc@rirdc.gov.au.
Webs ite: ht tp:/ /www.rir dc.gov .au
Published in March 2005
Printed on environmentally friendly paper by Canprint
iii
Foreword
In recent years scientists have been investigating the physiological responses (effect on blood sugar
levels) to food, particularly the effects of different carbohydrate containing foods. Honey has been
classified to be a food containing simple sugars and this has several implications on the choice of
foods for diabetics.
Glycemic Index factor is a ranking of foods based on their overall effects on blood sugar levels. The
source of honey decided the sugar and acid composition of honey which can show differences in the
GI factor. Little or no information exists on the GI of honey and in particular no information exists on
the differences in the GI of different honey varieties.
The quantitative measurement of organic acid and carbohydrate composition of different floral
varieties would therefore enable the study of GI of honey, and lead the way to understanding whether
or not all types of honey should be classified as one type of food for people with Diabetes.
RIRDC has been able to facilitate this study by providing the funding for this project. This report
discusses the sugar and acid composition of six floral varieties of honey, namely Red Gum, Salvation
Jane, Ironbark, Yellow Box, Stringybark and Yapunyah and two commercial blends obtained in 2001,
and their effects on the blood glucose response in humans.
This report is presented in a convenient format ready to publish in industry journals thus ensuring that
the beekeepers benefit from the findings of the study.
This project was funded from industry revenue which is matched by funds provided by the Federal
Government.
This report, a new addition to RIRDC’s diverse range of over 1200 research publications, forms part
of our Honeybee R&D program, which aims to improve the productivity and profitability of the
Australian Beekeeping Industry.
Most of our publications are available for viewing, downloading or purchasing online through our
website:
• downloads at www.rirdc.gov.au/fullreports/index.html
• purchases at www.rirdc.gov.au/eshop
Peter O’Brien
Managing Director
Rural Industries Research and Development Corporation
iv
Acknowledgments
There have been a number of beekeepers and suppliers who have supplied us with the required floral
sources of honey. I wish to acknowledge their contribution here.
Honey suppliers
Bill Winner (Capilano Honey, QLD)
Rob Manning (Department of Agriculture, WA)
Eduard Plankton (Wescobee Ltd, WA)
Trevor Lehmann (Leabrooks Honey, SA)
Discussion on sample collection
Bruce White (NSW Agriculture)
Rosie Stern (Consultant Dietitian)
Literature Collection
Rosie Stern (Consultant Dietitian)
Chemical Analyses
Shahnaz Shahtehmasebi (Honorary Research Associate, Department of Food Science and Technology,
UNSW, 2001)
GI testing
Susanna Holt (University of Sydney)
Abbreviations
GI Glycemic Index
II Insulin Index
v
Contents
Foreword...............................................................................................................................................iii
Acknowledgments................................................................................................................................. iv
Abbreviations........................................................................................................................................ iv
Executive Summary ............................................................................................................................. vi
1. Introduction ................................................................................................................................... 1
2. Objectives ....................................................................................................................................... 2
3. Review of Literature...................................................................................................................... 3
3.1 Honey....................................................................................................................................... 3
3.2 Nutritional Value ..................................................................................................................... 6
3.3 Colour ...................................................................................................................................... 7
3.4 Honey Quality.......................................................................................................................... 7
4. Honey Types and Geographical Distribution.............................................................................. 9
4.1 The Glycemic Index................................................................................................................. 9
5. Methodology................................................................................................................................. 12
5.1 Honey Collection................................................................................................................... 12
5.2 Chemical Analyses................................................................................................................. 12
5.3 Glycemic Index Testing......................................................................................................... 12
6. Results and Discussion ................................................................................................................ 16
6.1 Chemical analyses.................................................................................................................. 16
6.2 Glycemic Index Testing......................................................................................................... 17
7. Implications.................................................................................................................................. 23
8. Recommendations........................................................................................................................ 23
9. References..................................................................................................................................... 24
vi
Executive Summary
Objective
To obtain a clear understanding of the differences between the blood glucose responses of the different
sources of honey based on sugar and organic acid contents and identify those varieties with low GI
factor to use as a major marketing strategy to increase consumption, especially in Diabetics.
Method
Initial discussions with the Department of Agriculture, NSW were held to identify the common floral
varieties of honey that were available in 2001 depending on the season. The relevant suppliers in
Queensland, NSW, South Australia and Western Australia were approached for the supply of honey
with floral authentication. Six floral varieties namely, Red Gum (E. camaldulensis), Salvation Jane
(Echium lycopsis), Ironbark (E. nubilis), Yellow Box (E. melliodora), Stringybark (E. macrorhyncha)
and Yapunyah (E. ochrophloia) and two commercial blends were obtained from the above suppliers.
When the samples were received in the laboratory, they were stored at -180C until further analyses.
Samples were analysed for their sugar contents, namely, fructose, glucose, maltose and sucrose and
organic acids using standard HPLC techniques; and pH. The samples were tested for Glycemic Index
and Insulin Index through a human study comprising at least 10 healthy individuals. Glycemic Index
is a method developed in order to rank equal portions of different foods according to the extent to
which they increase blood glucose levels after being eaten. On the basis of the available carbohydrate
(sugars) content of the honeys, an amount of honey containing 25 grams of carbohydrate was given to
each volunteer to eat after an overnight fast. Over the next two hours, finger prick capillary blood
samples were collected and compared similarly with a reference food namely 25g bread. The Insulin
Index was also studied using the same procedure in the same subjects except that the concentration of
insulin in the plasma component was analysed instead.
Results and Discussion
The major results from the study are that:
• There were a lot of differences in the physical form of the honeys. Some were more solid and
crystallised and others were more fluid.
• The fructose contents varied from 27.5 – 52.4 g/100g of the honey amongst the varieties studied
with a commercial blend (1) having the lowest and Stringybark having the highest.
• The glucose contents varied from 20.3 – 32.9 g/100g with the same commercial blend (1) having
the least but the Red gum variety having the highest.
• Glucose and fructose were the predominant sugars found in all the honeys tested.
• Malic and succinic acids were the predominant organic acids found in all the varieties. Total acid
content was lowest in the commercial blend (1) and highest in Stringybark
• The pH of the honeys revealed that the range was between 5.2 (Salvation Jane) and 6.4 (Iron
Bark).
• Yellow Box, Stringybark , Red Gum, Iron Bark and Yapunyah honeys were considered to be of
low GI. Hence these are more suitable for consumption in controlled amounts by people with
diabetes and other health problems associated with poor blood glucose control (eg. pancreatic
disease, polycystic ovarian syndrome). The commercial blend (1) and Salvation Jane honeys are
of moderate GI and Commercial blend (2) was considered to be a high GI food. The insulin
responses were not exaggerated in relation to their corresponding glycemic responses. Therefore
the eight honeys tested do not appear to contain any insulinogenic components, other than sugar.
vii
• Only the honey’s fructose content was significantly associated with the average GI values and
average II values. The other individual sugars were not significantly associated with either the GI
or the II values.
Outcomes
The results of this study showed that different honeys could have significantly different effects on
blood glucose and insulin levels, due to differences in their sugar content and physical form, and
should not all be classified as one type of food for people with diabetes.
Now armed with the knowledge that there are differences in the GI between the floral varieties of
honey, and the fact that Yellow Box, Stringybark , Red Gum, Iron Bark and Yapunyah honeys were
considered to be of low GI and Salvation Jane and a commercial blend (1) were of moderate GI, it
should now be possible to better market these honeys as suitable for consumption in controlled
amounts for the diabetics. Commercial blends may vary in their composition depending on the
availability of honeys in that particular season and hence should be treated with caution if the varieties
that have gone into the blend are unknown as GI will be variable too. From the consumers’ point of
view, the floral varieties identified above with low to moderate GI should ideally be produced more
and marketed better by the Honey Industry. There may be other floral varieties that may be available
which should be studied in future for their GI.
viii
1
1. Introduction
Honey is the natural sweet substance produced by Apis mellifera bees from the nectar of plants or from
secretions of living parts of plants or excretions of plant sucking insects on the living parts of plants,
which the bees collect, transform by combining with specific substances of their own, deposit,
dehydrate, store and leave in the honey comb to ripen and mature (Codex standard, 1987).
Honey has a health attribute of being a readily available energy source. The overall health effect of
honey on individuals is often equated by health professionals with table sugar since the total energy
levels of white sugar and honey are quite similar. In addition total carbohydrates for honey and table
sugar are high. Over 82% of the solids in honey are composed of sugars. Health professionals are
often unaware of the types of carbohydrates in honey as compared to table sugar. Honey’s major
carbohydrates are the monosaccharides fructose and glucose while table sugar’s major carbohydrate is
sucrose, which is a disaccharide made up of glucose and fructose. The daily consumption of honey
that might cause problems in people with diabetes would be very similar to the amount of sugar
consumption that would supposedly cause a problem. Many diabetes associations in the world still put
a cap on sugar consumption which is around 30g/day or 5% of the total energy intake. Since honey has
high water content, an amount of 35-40g/day should be considered an upper limit of consumption.
Consequently there are differing physiological effects on blood sugar levels for honey as compared to
table sugar. The blood glucose response is lower for honey compared to table sugar. The physiological
effect of a carbohydrate on blood sugar levels is termed the Glycemic Index of the carbohydrate. The
Glycemic Index of a food is the ranking of a food based on the glycemic effect compared with a
standard food. The standard is usually bread or glucose.
The Glycemic Index of a food varies depending on factors such as processing methods and levels of
organic acids. It has been used to classify carbohydrate foods for various applications, including health
effects relating to diabetes, sports nutrition and weight management (Brand-Miller and Foster-Powell,
K. 1999).
The Glycemic Index of Australian honey has been shown to be 58 (Brand-Miller, 1995). Only a
limited number of studies worldwide have concentrated on the Glycemic Index of honey and these
studies have looked at the Glycemic Index of blended honeys rather than individual honeys from
different floral sources.
The aim of this research was to determine the Glycemic Index of six different Australian honey
varieties and two commercial blends as available in the Australian market. The honey types were
Stringybark, Iron bark, Red gum, Salvation Jane, Yellow Box, Yapunyah and two blended honeys.
The honey types varied in composition depending on the floral sources from which the nectar and
pollens were sourced. Honey was sourced from most Australian states. In addition to measuring the
Glycemic Index, total sugars, types of sugars, total acids and types of acids were analysed for each
honey type.
As a consequence of the study, an insight into the health effects, in relation to diabetes and sports
nutrition, of the different types of honey was determined. It is anticipated that the image of honey to
health professionals and sports persons will be improved. The Glycemic Index of the individual honey
types can be further used as a marketing strategy to improve the saleability of honey.
2
2. Objectives
The objectives of this study were to:
• Analyse the total available carbohydrate and organic acid contents of six individual varieties and
two blends of honey
• Assess the Glycemic index of the honeys in humans
3
3. Review of Literature
3.1 Honey
Honey has been described as a sweet viscous fluid made by honeybees from the nectar that they obtain
from plants, mainly flowers. It is ready to be consumed as produced and is essentially a fructose
solution supersaturated with glucose (White and Underwood, 1974).
In Australia honey is produced in most regions and approximately 75% of the Australian honey
originates from natural Eucalypt forests.
Honey contains more than 180 identified substances but honey consists mainly of sugars with the
remainder consisting of flavouring materials, minerals, acids, enzymes and pigments. The total amount
of sugars and the relative amounts of the different sugars (sucrose, fructose and glucose) vary in
nectars from different Eucalypts, and ultimately contribute to the different flavours and colours of
honey (Rostaim Faraji-Haremi, 1976). The general composition and properties of honey are
summarised in Tables 1 and 2.
Table 1. General composition of Australian honey
Composition
Moisture 15 –18 (% w/w basis)
Fructose 36 – 50%
Dextrose (Glucose) 28 – 36%
Sucrose 0.8 – 5.0%
Maltose 1.7 – 11.8%
Nitrogen 0.05 – 0.38%
Ash 0.04 – 0.93%
pH 3.3 – 5.6
Enzymes Invertase, diastase, glucose oxidase
Acid 0.5% (mainly Gluconic acid)
Free Acid 12 – 40 m-equiv./kg
Vitamins Minimal, less than 10% of Australian RDI
Minerals Minimal, less than 10% of Australian RDI
Source: Winner, 2001 (personal comm.)
Table 2. Physical properties of honey
Characteristic Value
Specific gravity (17% moist 200C) 1.423
Viscosity (17.1% moist 250C) 150 poise
Specific heat (17.4% moist 200C) 2.26 kJ / kg / K
Thermal conductivity
(17% moist 210C)
(17% moist 710C)
5.36 x 10-5 W/MK
5.95 x 10-5 W/MK
Freezing point (15%soln.) -1.42 - -1.530C
Water activity (aw) 0.5 – 0.6
Source: (D’arcy et. al., 1999)
Chandler et. al. (1974) performed chemical analysis of over 100 honeys from authenticated floral
sources. Most samples were from commercial Australian honeys. The Australian sourced honeys came
from all major honey-producing districts and represented over 60 different floral sources. Table 3
displays some representative floral sources from the different states and their corresponding
composition. A more detailed composition of both pure single floral species honeys and blended
honeys are available in Chandler et. al, (1974).
4
Table 3. Representative floral sources from Australian States (Chandler et. al. 1974)
Sample
No. Botanical Name Local Name Moisture
(%) pH Total Acid
(m-equiv/kg) Total Sugars
(%) Fructose
(%) Glucose
(%) Sucrose
(%)
New South Wales
1 E. camaldulensis River (red) gum 15.4 4.77 12.5 75.1 46.6 28.5 0.5
2 E. macrorhyncha Red stringybark 16.5 4.44 17.7 74.0 44.0 30.0 2.0
3 E. maculata Spotted gum 16.8 4.24 26.6 77.1 45.9 31.2 0.3
5 E. melliodora Yellow box 15.8 4.05 20.3 76.2 49.2 27.0 2.0
8 E. ochrophloia Napunyah 16.3 4.43 14.8 77.9 39.6 38.3 1.0
9 E scabra White stringybark 14.9 4.58 6.8 66.1 45.9 20.2 11.6
12 E. sideroxylon Mugga 15.7 4.17 16.1 75.5 45.2 30.3 2.0
15 E. viridis Green mallee 14.3 4.54 9.0 75.9 45.7 30.2 2.5
85 Echium lycopsis Paterson’s curse 14.2 3.81 23.8 73.4 43.0 30.4 4.8
86 Echium lycopsis Salvation Jane 15.5 3.80 23.1 76.4 43.1 33.3 2.2
Queensland
17 E. nubilis Dusky-leaved ironbark 17.4 4.48 10.0 74.8 44.4 30.4 1.4
20 E. melliodora & E dealbata Yellow box & hill gum 13.6 4.50 17.4 72.0 44.6 27.4 0.1
South Australia
22 E. camaldulensis River (red) gum 15.3 4.02 19.3 74.3 40.5 33.8 0.9
26 E. leucoxylon S.A. blue gum 15.5 3.99 21.3 76.6 43.6 33.0 1.5
Victoria
34 E. leucoxylon Yellow gum 15.3 4.10 20.7 70.3 40.2 30.1 5.7
Western Australia
41 E. marginata Jarrah 14.7 6.32 9.1 74.3 51.9 22.4 0.3
45 E. calophylla & E. diversicolor Marri & Karri 15.0 5.18 10.7 73.7 43.7 30.0 1.2
Tasmania
68 Eucryphia lucida Leatherwood 15.9 4.90 11.6 73.5 44.5 29.0 3.7
70 Eucryphia lucida Leatherwood 15.2 4.65 18.8 73.9 42.5 31.4 1.1
5
Honeys from the principal Australian floral source, the Eucalypts, show general uniformity in
chemical composition with a light amber colour, low moisture content, low acid and high pH values,
high glucose-to-moisture ratios and variable (low to high) granulation tendencies. Honeys from the
floral sources, white stringybark, yellow gum and yellow box have higher sucrose content. Messmate
honey has a high fructose and low glucose level, while Yapunyah has a high glucose and low fructose
content. Honeys from non-eucalypt Australian species, mainly tea tree flora were darker in colour,
higher in acidity (but not pH), higher in sucrose, and had a greater proportion of strong granulating
tendencies. The acidity-pH-ash relationships for these honeys were abnormal and suggest the
involvement of other factors besides acid and ash contents in determining the pH. Honeys from exotic
floral sources such as ground flora with the exception of orange blossom, showed high granulation
tendencies and low moisture contents. The honeys were lighter in colour (extra light amber), lower in
ash and pH, and higher in acidity than eucalypt honeys (Chandler et. al, 1974).
3.1.1 Sugars of Honey
Honey is a carbohydrate with the sugars accounting for 82 – 85% of the solids content of honey. Since
the sugars are the most important component of honey, the physical attributes of honey are largely
determined by the kinds and concentrations of the carbohydrates present (Crane, 1976).
In most honeys, the monosaccharide fructose predominates but exceptions occur such as in rapeseed
(canola) honey that contains greater amounts of glucose than fructose. There are at least twelve
disaccharides in honey in addition to fructose and glucose. These are sucrose, maltose, isomaltose,
nigerose, turanose, maltulose, leucrose, kojibiose, neotrehalase, gentiobiose, laminaribiose and
isomaltulose (D’arcy et. al., 1999).
3.1.2 Acids in Honey
The acids present in honey make up 0.5% of the total honey solids. The acids contribute to the
flavours of honey. The organic acids reported to be present in honey include: gluconic, formic, acetic,
butyric, lactic, oxalic, citric, succinic, tartaric, maleic, malic, pyroglutamic, pyruvic, α-ketoglutamic,
glycolic, α or β glycerophosphate and glucose-6-phosphate (Crane, 1976). Not only do acid levels
contribute to honey flavour, but the level of acidity of honey contributes to its’ stability towards
micro-organisms. Gluconic acid is present in honey in a higher amount than all other acids. It is
produced by the action of an enzyme in honey on the glucose in it.
Except for gluconic acid, the sources of the various honey acids are not known. Many of the acids are
intermediates in the Krebs cycle of biological oxidation, are of widespread occurrence and may be
present already in the nectar.
The identification of gluconic acid in honey provides an explanation of a difficulty long encountered
by analysts seeking to measure the total amount of the various acids in honey. This was done by
titration with alkali, and an indistinct or fading endpoint is often encountered, which lead to
uncertainty or error in the measurement. Gluconic acid exists in solution in equilibrium with its’
lactone, or internal ester, which does not have an acid function.
3.1.3 Ash, Acidity and Ph
Standards for ash content are designed to reject honeys that have become contaminated by metal
pickup from containers. There is a direct relationship between ash contents and pH, with Eucalypts
generally having higher ash contents and higher acidities (ie lower pH values). Lowest pHs have been
recorded for South Australian bluegum, spotted gum, mugga and bloodwood, with highest pHs for
white stringybark, jarrah, kurri/mauri, greybox and stoney mallee (Chandler, 1974).
6
3.2 Nutritional Value
A 100g serve of honey supplies 1320 kilojoules of energy compared to 100g of table sugar (sucrose)
which contains 1600 kilojoules of energy. Total carbohydrates vary with 82.1g/100g for honey and
100g/100g for table sugar (sucrose) (English and Lewis, 1992).
3.2.1 Proteins and Amino Acids
The nitrogen content of honey is quite low, on average 0.4%, though it may range to 1% of the total
solids. Only 40-65% of the total nitrogen in honey is protein in nature. The remainder of the nitrogen
is derived from the free amino acids found only in trace amounts. The most predominant of these are:
proline, glutamic acid, alanine, phenylalanine, tyrosine, leucine and isoleucine (D’arcy et. al., 1999)
3.2.2 Minerals, Vitamins and Enzymes
Honey contains small amounts of minerals and vitamins (see Table 4). Many minerals have been
identified, including potassium, sodium, calcium, magnesium, iron, copper, chlorine, phosphorous and
sulphur. These are of little significance due to their small quantities. All minerals and vitamins in
honey are less than 10% of the RDI (Recommended Dietary Intakes) for these micronutrients.
Invertase is the most significant enzyme in honey since invertase added by the honeybee splits the
sucrose into constituent sugars and produces other more complex sugars in small percentages during
the process. The substrate for invertase is sucrose which is hydrolysed to give glucose and fructose.
Diastase (α- and β- amylases) is another predominant enzyme and is frequently used to measure honey
quality. It is used as a predictor to determine if honey has undergone any heat treatment. Additionally,
glucose oxidase is found in honey and is responsible for the conversion of glucose to gluconolactone,
which in turn forms gluconic acid which is the dominant acid in honey (D’arcy et. al., 1999).
Table 4. Minerals and Vitamins in Honey
Vitamins Amount in 100g of Honey
Thiamine <0.006mg
Riboflavin <0.06mg
Niacin <0.36mg
Pyrodoxine (B6) <0.32mg
Ascorbic Acid (C) 2.2 – 2.4mg
Minerals
Calcium 4.4 – 9.20mg
Iron 0.06 – 1.50mg
Magnesium 1.2 – 3.5mg
Phosphorus 1.9 – 6.3mg
Potassium 13.2 – 16.8mg
Sodium 0.0 – 7.6mg
Zinc 0.03 – 0.4mg
Source: Stern, 1999
3.2.3 Moisture Content of Honey
The moisture content of honey can vary from as low as 12% to as high as 27% w/w basis with
Australian honeys usually 16-18%. The low moisture content together with a high osmotic pressure of
honey prevents the growth of bacteria. The water activity of honey is low, 0.5 – 0.6 which is at a level
where most bacteria and fungi do not grow.
7
3.3 Colour
Colour (measured using the industry standard pfund scale) is used as a measure of quality in Australia,
with certain colour grades gaining premium prices compared to poorer ones. The colour grades and
corresponding pfund values are listed in Table 5.
Colour varies greatly from nearly colourless to yellow, yellow green, gold, amber, dark brown or red
brown to almost black.
Table 5. Colour grades of honey and their corresponding pfund values
Colour Grade Pfund value Examples
White Less than 34mm White Clover
Extra light amber (ELA) 35 – 48mm Brush Box, Iron Bar,
Light amber (LA) 49 – 65mm Stringybark, supermarket
blend
Pale amber (PA) 66 – 82mm Blue Gum
Medium amber (MA) 83 – 100mm Tea Tree, Sugar Cane
Amber 100 – 114mm Candied
Dark Amber More than 114mm Rainforest
Source: D’arcy et. al., 1999 and Lower Clarence Skills Centre, 1996
Eucalypt honeys are generally darker than other honeys (Chandler et. al., 1974). Most consumers
prefer lighter coloured honey compared to darker coloured honeys (Lower Clarence Skills Centre,
1996).
3.4 Honey Quality
The Australian Honey Industry general specifications for honey of high quality are presented in Table
6.
Table 6. Australian Honey Quality Specification
Moisture Not more than 20%
Apparent Reducing Sugar Not less than 65%
Apparent Sucrose Range from 5 to not more than 15% depending on
floral source
Water insoluble solids Not more than 0.1%
Mineral content (ash) Not more than 0.6%
Acidity Not more than 40 milliequivalents acid per 1000
grams
Diastase Activity Not less than 3
Hydroxymenthylfurfural (HMF) Not more than 80 mg/kg
Colour Shall be graded using the Pfund grading standard
Source: Australian Honey Quality Specifications, 2001
The large honey packers of Australia including Capilano (Qld and NSW) and Westcobee (WA)
operate a program of quality control keeping safety as a priority. In addition, honey producers who
supply the large honey packers are trained in HACCP (food safety), and are regularly audited for food
safety and quality.
8
3.4.1 Granulation and Crystallinity
Glucose monohydrate spontaneously crystallises from honeys that are a supersaturated solution under
ordinary storage conditions. Therefore granulation is the result of the crystallisation of glucose caused
by a change in the supersaturated state and, in theory, whether a honey will granulate or not will
depend on the proportion of glucose to other components of the mixture. Several formulae using the
glucose, water and fructose contents of a honey have been suggested for predicting its’ susceptibility
to crystallisation. None of these formulae are reliable indicators of crystallisation (Chandler et. al.,
1974).
Rapid crystallisation is expected in honeys from mallee, yapunyah, river red gum, while problems of
crystallisation should not occur in honeys from coastal blackbutt, grey box, jarrah, messmate, pink
gum, white stringybark and yellow box floral sources (Chandler et. al., 1974).
9
4. Honey Types and Geographical
Distribution
4.1 The Glycemic Index
The Glycemic Index (GI) is a physiologically based method used to classify carbohydrate foods
according to their blood glucose-raising potential. The concept has been widely adopted in diabetes
management in Australia, New Zealand, Canada, the United Kingdom and France. The GI compares
rise in blood sugar levels after equal carbohydrate portions of foods are ingested and ranks them
relative to a standard which is usually glucose or white bread (Brand-Miller et. al., 2000).
The Glycemic Index measures the area under the glycemic response curve during a 2-hour period after
consumption of a 50g carbohydrate serve from a test food, with values being expressed relative to the
effect of either white bread or glucose. As a result, the Glycemic Index is considered a specific
property of foods. As shown in Table 12, high Glycemic Index foods are those that have the highest
peak circulating glucose in the 2 hour period following food ingestion and the highest area under the
curve for the increase in blood glucose above fasting baseline. Conversely, low Glycemic Index foods
are those that cause lower peak glucose, demonstrating a smaller area under the curve for the change in
blood glucose in the 2-hour period and have a lower risk of causing hypoglycaemia (Roberts, 2000).
Over the last 2 decades more than 600 individual foods have been tested for their Glycemic Index
Table 12 summarises some of the foods tested.
Contrary to popular belief, low Glycemic Index foods are not the same as foods based on high
complex carbohydrate and fibre, nor are high Glycemic Index foods those based on simple sugars. The
foods that produce the highest glycemic responses include many of the starchy foods consumed by
people in industrialised countries, including bread, breakfast cereals, and potatoes, whether high or
low in fibre. This is because the starch is fully gelatinised and can be rapidly digested and absorbed.
The foods with the lowest Glycemic Index values include pasta, relatively unprocessed cereal foods,
baked beans, dairy products, and many types of fruit and vegetables. Sugary foods often cause lower
levels of glycaemia per gram of carbohydrate than the common starchy staples of western diets. This is
because up to half of the weight of carbohydrate is fructose (as is the case with honey), a sugar that
has little effect on glycaemia. In fact, the overall Glycemic Index of the diet has been shown to have
an inverse correlation with total sugars (refined plus naturally occurring) expressed as a proportion of
total carbohydrate (Brand-Miller et. al., 2000).
In general, high Glycemic Index foods are those with high carbohydrate content and foods that are
rapidly digested. Specific factors that favour increased Glycemic Index include: high-refined
carbohydrate content (because fat and protein have minimal effect on blood glucose compared with
carbohydrate); high glucose and/or starch content relative to lactose, sucrose and fructose contents
(because these sugars yield less glucose, none in the case of fructose); low soluble fibre content
because soluble fibre forms a gel in the stomach and reduces the rate of gastric emptying and hence
the rate of digestion; and finally, soft, overcooked, highly processed, or over ripened food textures
because they are digested more rapidly than foods with greater structural integrity such as firm raw
foods, intact grains, and discrete harder pieces of food (Roberts, 2000).
10
Table 12. The Glycemic Index (GI) of foods
Low Glycemic Index (<55) GI Moderate Glycemic Index (56-69) GI High Glycemic Index (>70) GI
Breads
Pumpernickel 41
Heavy mixed grain 30-45
Breakfast cereals
All Bran 42
Toasted muesli 43
Psyllium-
b
ased processed cereal
42
Dairy foods
Milk, full fat 27
Milk, skim 32
Yoghurt, low fat, fruit 33
Confectionery
Chocolate (Dove) 45
M&Ms 33
Snickers Bar 41
Fruits
Apple 36
Orange 43
Peach 28
Legumes
Lentils 28
Soybeans 18
Baked Beans 48
Breads
Sourdough 57
Barley bread 65
Rye bread 65
Breakfast cereals
Cream of wheat 66
Muesli 66
Dairy foods
Ice cream, full fat 61
Confectionery
Mars Bar 65
Fruits
Pineapple 52
Pawpaw 58
Rices 50-60
(high amylose varieties, eg basmati)
Honey (blended Australian) 58
Breads
White bread 70
Wholemeal bread 72
French bread 95
Breakfast cereals
Cornflakes 84
Rice Bubbles 82
Potatoes 80-100
Confectionery
Jelly beans 80
Life Savers 70
Fruits
Watermelon 72
Rices 70-90
(low amylose, white or brown)
Honey (blended not Australian)
87
Reference food is Glucose = 100
Source: Brand-Miller, J and Foster-Powell, K. (1999)
The International Tables of Glycemic Index lists honey as having a Glycemic Index of either 58 or 87
(Powel et. al., 1995). The Glycemic Index of 58 is an Australian blended honey (Brand-Miller, 1995)
while the honey with a Glycemic Index of 87 has not documented the source (Jenkins et. al., 1981).
Since the composition of sugars in honey varies depending upon the floral source, it can be assumed
that the Glycemic Index of honey will vary depending upon the floral source of the honey. There are
important differences between Glycemic Indices of the monosaccharides in honey, notably glucose
and fructose levels. Fructose has a Glycemic Index of only about 23. The Glycemic Index of a sugar
can be predicted on the basis of the molar ratio of glucose to other monosaccharides in the sugar
molecule. This explains why maltose (a disaccharide with two glucose units) has a score close to
glucose at 100, whereas sucrose (a disaccharide of glucose and fructose) has a Glycemic Index of only
61. Honey, which contains mixtures of glucose and fructose, may therefore have index values with
various ranges (Gurr 1997).
The Australian honey industry has shown a keen interest in finding out the Glycemic Index of
individual honeys since the industry was made aware of the Glycemic Index of Australian honey at an
annual NSW Apiarists Association conference and follow up paper (Stern, 1999).
11
4.1.1 Health and the Glycemic Index
Most research relating to the Glycemic Index and health indicates the clinical usefulness in the
treatment of diabetes and hyperlipidaemia. Short term studies in lean healthy people, obese
individuals, and people with diabetes show consistently higher day long insulin levels with diets based
on high Glycemic index foods in comparison with low Glycemic index diets of similar nutrient
composition. In people with diabetes, the consumption of high Glycemic index foods results in a far
more exaggerated glycemic and insulin response, which may lead to worsening insulin resistance and
eventually the need for drug or insulin therapy. Furthermore, higher day long insulin levels promote
carbohydrate oxidation at the expense of fatty acid oxidation, thereby encouraging synthesis of very
low density lipoprotein cholesterol (VLDL) in the liver and fat storage in adipose tissue. A
combination of high Glycemic Index carbohydrate and high fat (of any type) in a meal therefore may
be synergistic in promoting weight gain.
Long term studies in animal models show that high Glycemic Index starch increases fasting insulin
levels and promotes insulin resistance, in comparison with identical diets based on low Glycemic
Index starch.
Recent epidemiological studies indicate that the Glycemic Index of the diet may be the most important
dietary factor in preventing type 2 diabetes. Two large scale prospective studies, one in female nurses
and one in male health professionals, showed that diets with a high glycemic load (GI x carbohydrate
content) increases the risk of developing type 2 diabetes after controlling for known risk factors such
as age and body mass index. A similar picture emerged with acute coronary heart disease in the
Nurses’ study. The underlying mechanism postulated by these authors is the demand for insulin
generated by high Glycemic Index foods. Because hyperinsulinaemia is linked with all of the facets of
the “metabolic syndrome” (insulin resistance, hyperlipidaemia, hypertension and visceral obesity), the
Glycemic Index of foods eventually may be linked with all so-called diseases of affluence.
In healthy people as well as those with type 2 diabetes, high-carbohydrate diets (>50% energy) have
been shown to worsen aspects of the blood lipid profile, including the TG, VLDL, HDL and
lipoprotein. Individuals with insulin resistance are more susceptible to these adverse effects.
The Glycemic Index has implications for weight control in people with diabetes because slowly
digested carbohydrate is associated with higher satiety. The prolonged presence of food in the gut may
stimulate chemical and pressure receptors that signal satiety. Low insulinaemic diets have been shown
to increase the rate of weight loss on energy restricted diets through the mechanism of lower insulin
levels. Thus, low Glycemic Index diets may promote weight control by both enhancing satiety and
reducing insulinaemia.
There is also some evidence that the Glycemic Index is relevant to sports nutrition. Low Glycemic
Index foods eaten before prolonged strenuous exercise increases endurance time and provides higher
concentrations of plasma fuel towards the end of exercise, while high Glycemic Index foods lead to
faster replenishment of muscle glycogen after exercise (FAO/WHO, 1997).
12
5. Methodology
5.1 Honey Collection
Six floral varieties and two commercial blends were chosen for the study after discussions with the
RIRDC, R&D Advisory Committee. They were, Red Gum, Salvation Jane, Ironbark, Yellow Box,
Stringybark and Yapunyah and two commercial blends obtained in 2001. Duplicate collection of all
honeys at source was ensured. The main Industry contributor was Capilano Honey Ltd from
Queensland. The other producers who supplied the honeys were, Wescobee Ltd from WA, Leabrooks
Honey Ltd from SA and Department of Agriculture, WA. Red Gum was particularly obtained from SA
and WA and the rest were obtained from Capilano Ltd. The purity of the honeys was to a large extent
assured by the suppliers. The two commercial blends were obtained from Capilano Honey Ltd and
Wescobee Ltd. The composition of the floral varieties or the types that went into the blends was not
known. Once the honey samples were received in the laboratory at the Department of Food Science
and Technology, The University of New South Wales, they were stored at -180C until further analyses.
5.2 Chemical Analyses
The composition of sugars in all the honeys were analysed by a High Pressure Liquid
Chromatographic (HPLC) Technique as suggested by Wills et al, (1980). Glucose, fructose, sucrose
and maltose were analysed. The pH of the samples was tested using a pH meter. Organic acids (oxalic,
malic, succinic, lactic, acetic, propionic, citric and butyric acids) were analysed using a standard
HPLC technique (AOAC, 2000).
5.3 Glycemic Index Testing
This study was conducted using internationally recognised GI methodology, which has been
validated by small experimental studies and large multi-centre research trials. The experimental
procedures used in this study were in accordance with international standards for conducting ethical
research with humans and were approved by the Medical Ethics Review Committee of Sydney
University.
5.3.1 Study Participants (Subjects)
For both parts of the study, a group of 9-10 healthy, non-smoking people aged between 18-45 years
was recruited from the staff and student population of the University of Sydney. People volunteering
to participate in the study were excluded if they were overweight, were dieting, had a family history of
diabetes, were suffering from any illness or food allergy, or were regularly taking prescription
medication.
In the first part of the study, a group of seven females and three males was recruited. The average age
of the subjects was 27.5 years (range: 19.8 – 44.9 years) and the average body mass index (BMI) score
was 22.5 kg/m2 (range: 20.8 – 25.0 kg/m2). The BMI score is a measure of a person’s weight in
relation to their height. BMI values between 19 – 25 kg/m2 are within the healthy weight range. In
the second part of the study, a group of nine females were recruited, five of which had also
participated in the first part of the study. Therefore, the two groups of subjects were relatively similar
in terms of their age and BMI ranges, and both groups predominantly consisted of females. The
average age of the subjects in the second part of the study was 27.9 years (range: 19.7 – 44.9 years)
and the average body mass index (BMI) score was 22.3 kg/m2 (range: 20.2 – 25.0 kg/m2).
Sample number: With 10 subjects in each group a power of 80% is seen to test the difference of one
SD at the 0.05 level. Differences of less than one SD would not be considered a clinically significant
difference. This is the usual acceptable size in all GI investigations.
13
As there is no difference in the glycemic response/glycemic index between male and female, or
between lighter and heavier people, each honey was tested in 9 or 10 different individuals. The
reference food was tested twice and the average area under the curve for each individual was used.
5.3.2 Test Foods
In both parts of the study, pure glucose sugar (dextrose (D-glucose), Sigma-Aldrich chemical
company, Castle Hill, NSW, Australia) dissolved in water was used as the standard reference food and
was consumed by each study participant on two separate occasions. Each of the eight types of honey
was consumed by each study participant on one occasion only. The reference food and the eight
honeys were all served in amounts containing 25 grams of available (digestible) carbohydrate. The
weight and sugar content of the test portions of the honeys are listed in Table 13.
Table 13. The weight (g) and sugar content (g) of the test portion of the reference food and
honeys
Food
Portion
size
Av.
Carbohydrate
Fructose
Glucose
Sucrose
Maltose
Reference
food
25 g
glucose 250
mL water
25.0
0.0
25.0
0.0
0.0
Commercial
Blend 1
49.6
25.0
13.6
15.5
0.8
0.7
Commercial
Blend 2
35.6
25.0
13.6
10.5
0.3
0.6
Iron Bark
41.7
25.0
14.1
9.8
0.5
0.6
Red Gum
33.9
25.0
11.7
11.2
0.8
1.3
Salvation
Jane
40.6
25.0
12.9
11.2
0.4
0.4
Stringy bark
30.4
25.0
15.9
8.5
0.3
0.3
Yapunya
36.9
25.0
15.5
8.8
0.3
0.3
For each study participant, the reference food was prepared the day before required by dissolving 25
grams of pure glucose sugar in a glass of 250 mL of warm water, which was then covered and stored
overnight in a fridge. The solution was taken from the fridge shortly before serving to the study
participants with a glass of 250 mL of plain water. The test portions of the honeys were weighed into
standard glass bowls the day before required and were covered with airtight plastic wrap and stored
overnight in a fridge. The next morning, the test portions of the honeys were taken from the fridge
14
shortly before being served to the study participants with a spoon and a glass of 250 mL of plain
water. The study participants were required to consume all of the honey or glucose solution and water
served to them. The study participants were instructed to consume all the honey out of the serving
bowls.
5.3.3 Experimental Procedures
Using standard GI methodology to determine a food’s GI value, a portion of the food containing 25
grams of available carbohydrate is fed to 8-10 healthy people in the morning after they have fasted for
10-12 hours overnight. (The amount of carbohydrate chosen depends on the energy density of the test
foods and the size of the test portions. A smaller dose of carbohydrate (25 g) was chosen as subjects
could not consume all the honey given to them as the portion size for 50 g carbohydrate content was
equivalent to between 30-100g of honey by weight which was too large to be consumed comfortably
within 12 minutes). A fasting blood sample is obtained and then the food is consumed, after which
additional blood samples are obtained at regular intervals during the next two hours. In this way, it’s
possible to measure the total increase in blood sugar (glucose) and insulin levels produced by that
food. The two-hour blood glucose (glycemic) response for this test food is then compared to the two-
hour blood glucose response produced by the same amount of carbohydrate in the form of pure
glucose sugar (the reference food: GI value of glucose = 100%). Therefore, GI values for foods are
relative measures (ie. they indicate how high blood sugar levels rise after eating a particular food
compared to the very high blood sugar response produced by glucose sugar). Insulin index (II) values
are calculated in the same way as GI values, substituting the blood glucose response values in the GI
equation (see page 8) with the corresponding blood insulin values.
In both parts of this study, the study participants consumed the reference food on two separate
occasions, while the honeys were each consumed on one occasion only. The reference food was
consumed on both the first and last test sessions, and the honeys were consumed in random order in
between.
The day before each test session, the study participants were required to refrain from unusual amounts
of eating and exercise, and were required to consume at least 300 grams of carbohydrate for the whole
day. In addition, they were required to refrain from consuming alcohol for the whole day and refrain
from consuming a legume-based meal during the evening. The night before a test session, the study
participants ate a regular evening meal and then fasted for 10-12 hours overnight. During the fasting
period, they were allowed to drink only water.
The next morning they reported to the research centre in a fasting condition. The study participants
first warmed a hand in a bucket of hot water for two minutes, after which a fasting finger-prick blood
sample was obtained from a finger (approximately 0.9-1.2 mL of blood) using an automatic lancet
device (Safe-T-Pro®, Boehringer Mannheim Gmbh, Mannheim, Germany). After the fasting blood
sample was obtained, study participants were given a fixed portion of a reference food or a honey,
which they consumed with 250 mL of plain water at a comfortable pace within 12 minutes. The study
participants were required to consume all of the honey or reference food and water served to them.
The participants were then required to remain seated at the research centre and refrain from eating and
drinking during the next two hours. Additional blood samples were taken 15, 30, 45, 60, 90 and 120
minutes after eating had commenced. Therefore, a total of seven blood samples were collected from
each subject during each two-hour test session.
5.3.4 Measurement of Blood Glucose Responses
For each study participant, the concentration of glucose in the plasma component in each of their
seven blood samples was analysed in duplicate using the glucose hexokinase enzymatic method
(Roche Diagnostic Systems, Sydney, Australia) and an automatic centrifugal spectrophotometric
analyser (Roche/Hitachi 912®, Boehringer Mannheim Gmbh, Mannheim, Germany) using internal
15
controls. The glucose concentrations in the seven blood samples were then used to graph a two-hour
blood glucose response curve, which represents the total two-hour glycemic response to that food (ie.
the total rise in blood sugar induced by the digested food). The area under this two-hour blood plasma
glucose response curve (AUC) was calculated using the trapezoidal rule (1), in order to obtain a single
number, which indicates the magnitude of the total blood glucose response during the two-hour period.
A glycemic index (GI) value for the test food was then calculated by dividing the two-hour blood
glucose AUC value for this test food by the subject’s average two-hour blood glucose AUC value for
the reference food and multiplying by 100 to obtain a percentage score.
GI value for test food (%) = Blood glucose AUC value for the test food x 100
AUC value for the same carbohydrate portion of the reference food
Due to differences in body weight and metabolism, blood glucose responses to the same food can vary
between different people. The use of the reference food to calculate GI values reduces the variation
between the subjects’ blood glucose results to the same food arising from these natural differences.
Therefore, the GI value for the same food varies less between the subjects than their glucose AUC
values for this food. In this study, the final GI value for each honey is the average of the 9-10
subjects’ GI values for that honey.
5.3.5 Measurement of Blood Insulin Responses
For each study participant, the concentration of insulin in the plasma component in each of their seven
blood samples collected during each test session was analysed using a solid-phase antibody-coated
tube radioimmunoassay kit (Diagnostic Products Corporation, Los Angeles, CA, USA) with internal
controls. The plasma insulin concentrations in the seven blood samples were then used to graph a
two-hour blood insulin response curve, which represents the study participant’s total two-hour
insulinaemic response to that food. The area under this two-hour blood plasma insulin response curve
(AUC) was calculated and an insulin index (II) value for the honey was then calculated using the GI
formula shown above, substituting the insulin AUC results for the glucose AUC results.
16
6. Results and Discussion
6.1 Chemical analyses
6.1.1 Carbohydrate composition
The mean sugar content of the six floral varieties and two commercial blends are given in Table 14.
Table 14. Mean available sugar content (g/100g) of honeys
Honey varieties Glucose Fructose Sucrose Maltose Total
Commercial
Blend 1(NSW) 20.3 + 0.3 27.5 + 1.0 1.1 + 0.2 1.5 + 0.2 50.4 + 0.4
Commercial
Blend 2 (WA) 29.6 + 0.4 38.1 + 1.1 0.9 + 0.1 1.6 + 0.3 70.2 + 0.5
Iron Bark 23.6 + 0.5 33.8 + 0.8 1.1 + 0.2 1.4 + 0.2 59.9 + 0.5
Red gum 32.9 + 0.9 34.6 + 0.7 2.5 + 0.4 3.7 + 0.4 73.7 + 0.6
Salvation Jane 27.7 + 0.8 31.9 + 1.0 0.9 + 0.1 1.1 + 0.1 61.6 + 0.5
Stringybark 27.9 + 1.0 52.4 + 1.3 1.0 + 0.1 1.0 + 0.1 78.3 + 0.6
Yapunya 23.9 + 0.5 42.1 + 1.9 0.8 + 0.1 0.9 + 0.2 67.7 + 0.6
Yellow Box 26.8 + 0.7 45.5 + 2.1 0.9 + 0.1 1.1 + 0.2 74.3 +0.8
The above values are means + S.D of duplicate determinations
The glucose content ranged from 20.3- 32.9 g/100g with the commercial blend (1) (NSW) having the
least and Red gum having the highest. Fructose levels varied from 27.5-52.4 g/100g with the
commercial blend (1) having the least and Stringy bark having the highest. Sucrose content was low in
all samples (0.9-1.1g/100g) excepting in Red gum. The maltose levels ranged from 0.9-3.7 g/100g
with Yapunyah having the least and Red gum having the highest. The total sugar content varied
between 50.4 and 78.3 g/100g with the commercial blend (1) having the least and Stringybark having
the highest. Sugar contents are in line with literature values.
6.1.2 Organic acids
The organic acid contents measured using a HPLC technique revealed Malic and succinic acids to be
the predominant ones. Oxalic, tartaric, malic, succinic, lactic, acetic, propionic, citric and butyric acids
were identified. Table 15 indicates the contents of the acids.
17
Table 15. Mean organic acids (mg/100g) content of honeys
Honey Oxalic Tartar Malic Succin Lactic Acetic Propio Citric Butyri
Com.1 0 0 0.97 0.12 0 0.04 0 0.03 0
Com.2 0.03 0 1.07 0.13 0 0.04 0.04 0 0
Red Gum 0.13 0.1 0.02 1.1 0.02 0.02 0.02 0 0
S. Jane 0 0 0.92 0.12 0 0 0 0.01 0
Iron Bark 0 0.13 1.41 0.13 0 0 0 0.05 0
Yellow Box 0.01 0.06 1.33 0.07 0 0 0 0.01 0
Stringy
Bark 0.11 0.1 1.35 0.26 0 0 0 0.03 0
Yapunyah 0 0.01 1.4 0.22 0 0 0 0.04 0.6
The above values are means of duplicate analyses.
6.1.3 pH of the Honeys
The pH of the honeys ranged from 5.2-6.4. Some honeys ranged from 5.2-5.8 (Salvation Jane,
Commercial blend (1) and Yellow Box. The others ranged from 6.0- 6.4 (Yapunyah, Stringybark,
Commercial blend (2), Red Gum and Iron Bark).
6.1.4 Osmolality of Honeys
Using an osmometer, the honeys were tested for their osmolality in replicates. They ranged from 4804-
4884 for Salvation Jane, Iron Bark, Commercial blend (2), and Red gum. The rest had a range of
5676-5708 for Yellow Box, Stringybark and Commercial blend (1).
6.2 Glycemic Index Testing
The study was divided into two parts with 3 honeys tested in the first lot and the rest in the second lot.
The three honeys that were tested first were Yellow Box, Iron Bark and Salvation Jane. The average
two-hour blood glucose response curves for the reference food and the three honeys tested in the first
part of the study are shown in Figure 1. The reference food produced the highest overall glycemic
response curve producing a large rise and fall in the level of blood glucose. The peak blood glucose
level and pattern of the two-hour glycemic response curves varied among the honeys. On average, the
Salvation Jane honey produced the largest response curve and the Yellow Box honey produced the
lowest response curve.
18
-0.05
0
0.05
0.1
0.15
0 50 100 150
Time (min)
Change in plasma
glucose concentration
(mmol/L)
Glucose
Salv.Jane
Iron Bark
Yellow Box
Figure 1. The average plasma glucose response curves for the reference food and the three honeys
tested in the first part of the study, depicted as the change in glucose concentration from the fasting
baseline level.
The average blood insulin response curves for the three honeys tested in part 1. The average two-hour
blood glucose response curves for the reference food and the three honeys tested in the first part of the
study are shown in Figure 1. The patterns of the blood insulin response curves were not exactly the
same as their corresponding glycemic response curves. The reference food produced the highest
overall insulin response curve and the insulin response curves for the honeys varied to a greater extent
than their glycemic response curves. The reference food produced the highest insulin response curve
followed by Salvation Jane honey, Iron Bark honey, and, lastly, the Yellow Box honey.
The average plasma insulin response curves for the
reference food and the four honeys tested depicted as
the change in blood insulin concentration from the fasting
baseline level
-50
0
50
100
150
0 50 100 150
Time (min)
Change in plasma
insulin concentration
(pmol/L)
Glucose
Salvation Jane
Iron Bark
Yellow Box
Figure 2. The average plasma insulin response curves for the reference food and the three honeys
tested in the first part of the study, depicted as the change in blood insulin concentration from the
fasting baseline level.
The average two-hour blood glucose response curves for the reference food and the five honeys tested
in the second part of the study are shown in Figure 3. The reference food produced the highest overall
19
glycemic response curve and the response curves for the five honeys varied markedly with the
Commercial Blend #1 honey producing the largest overall response curve and the Stringybark honey
producing the smallest response curve.
Figure 3. The average plasma glucose response curves for the reference food and the five honeys
tested in the second part of the study, depicted as the change in glucose concentration from the fasting
baseline level.
6.2.1 The average blood insulin response curves for the four honeys tested in
part 2
The average two-hour blood glucose response curves for the reference food and the five honeys tested
in the second part of the study are shown in Figure 4. The reference food produced the highest insulin
response curve and the five honeys produced a range of insulin responses, varying in the peak insulin
concentration, and the rate of rise and fall in blood insulin levels. Among the honeys, the Commercial
blend # 1 honey produced the highest integrated two-hour insulin response and the Stringybark honey
produced the lowest.
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0 50 100 150
Time (min)
Cahnge in plasma
glucose concentration
(mmol/L)
Gluc ose
Stringybark
Yapunya
Commercial
blend 1
Commercial 2
Red gum
20
The average plasma insulin response curves for the
reference food and the six honeys depicted as the change in
blood insulin concentration from the fasting baseline level
-20
0
20
40
60
80
100
120
140
160
0510
Time (min)
Change in plasma insulin
concentration (pmol/L)
Gluc os e
Comm.b len d 1
Yapuny a
Comm.b len d 2
Stringybark
Figure 4. The average plasma insulin response curves for the reference food and the four honeys
tested in the first part of the study, depicted as the change in blood insulin concentration from the
fasting baseline level.
6.2.2 Glycemic and Insulin Index Values
The GI and II values for each of the honeys tested varied among the 10 subjects who participated in
the study. This variation in GI and II values for the same food between people is normal and is due to
a number of factors, such as the different rates at which the subjects ingested the foods, and genetic
factors affecting the metabolism of carbohydrate. The average (mean) GI and II values (mean ±
standard error of the mean) for the nine honeys are listed in Table 16.
Table 16. The mean ± SEM GI and II values for the eight honeys, using glucose sugar as the
reference food (ie GI and II value for glucose = 100)
Test Food GI value (%) II value (%)
Low
Yellow Box honey 35 ± 4 40 ± 5
Stringybark honey 44 ± 4 47 ± 3
Red Gum honey 46 ± 3 51 ± 3
Iron Bark honey 48 ± 3 42 ± 4
Yapunya honey 52 ± 5 49 ± 3
Moderate
Commercial blend # 2 honey 62 ± 3 62 ± 4
Salvation Jane honey 64 ± 5 52 ± 3
High
Commercial blend # 1 honey 72 ± 6 67 ± 6
Reference food (glucose) 100 ± 0 100 ± 0
21
6.2.3 Significant Differences Among the Honeys’ Average GI and II Values
Standard statistical tests (Analysis of Variance and the Fisher PLSD test for multiple comparisons)
were used to determine whether the average GI and II values for the honeys were significantly lower
than the GI and II value of the reference food and whether significant differences existed among the
honeys’. The smaller the p value, the more significant the difference, with p<0.001 (99.9%) being the
most significant difference followed by p<0.01 and lastly p<0.05.
Significant Differences Among GI Values
Due to the different number of subjects in each part of the study, the results from the two parts of the
study were analysed separately. For the first part of the study, the GI value for the reference food
(glucose sugar) was significantly greater than the average GI values for all three honeys tested, with
the difference being highly significant (p<0.001). The average GI value of the Salvation Jane honey
was significantly greater than the average GI values of the Yellow Box honey (p<0.001) and the Iron
Bark honey (p<0.01). The average GI value of the Iron Bark honey was significantly greater than that
of the Yellow Box honey (p<0.01).
For the second part of the study, the GI value for the reference food (glucose sugar) was significantly
greater than the average GI values for all five honeys tested, with the difference being highly
significant (p<0.001). The average GI value of the Commercial blend honey # 1 was significantly
greater than the average GI values of the Stringybark, Yapunya, Red Gum (p<0.001), and Commercial
blend # 2 (p<0.05) honeys. The average GI value of the Commercial blend honey # 2 was
significantly greater than the average GI values of the Stringybark (p<0.001), Red Gum (p<0.01), and
Yapunya (p<0.05) honeys.
Significant Differences Among II Values
Due to the different number of subjects in each part of the study, the results from the two parts of the
study were analysed separately. For the first part of the study, the II value for the reference food
(glucose sugar) was significantly greater than the average II values for all three honeys tested, with the
difference being highly significant (p<0.001). The average II value of the Salvation Jane honey was
significantly greater than the average II values of the Iron Bark and Yellow Box honeys (p<0.05).
For the second part of the study, the II value for the reference food (glucose sugar) was significantly
greater than the average II values for all five honeys tested, with the difference being highly
significant (p<0.001). The average II value of the Commercial blend honey # 1 was significantly
greater than the average II values of the Stringybark, Yapunya, and Red Gum (p<0.001) honeys. The
average II value of the Commercial blend honey # 2 was significantly greater than the average II
values of the Stringybark, Yapunya (p<0.01) and Red Gum (p<0.05) honeys.
Relationship Between the Honeys’ Average GI and II Values
Linear correlation analysis showed that the average GI and II values for the honeys were significantly
associated (r = 0.875, n = 9, p<0.001) (Figure 5). Plasma glucose and insulin responses typically show
a highly significant association for low-fat, high-carbohydrate foods. The insulin responses were not
exaggerated in relation to their corresponding glycemic responses. Therefore, the nine honeys tested
do not appear to contain any insulinogenic components, other than sugar.
22
Figure 5. The relationship between the eight honeys’ average GI and II values.
Relationship Between the Honeys’ Sugar Contents and GI and II Values
Linear correlation analysis was used to examine the association between the honeys’ content of single
sugars (fructose, glucose, sucrose and maltose (g/100 g)) and the average GI and II values. Only the
honeys’ fructose content was significantly associated with the average GI values (r = - 0.76, n = 9,
p<0.05) and average II values (r = - 0.67, n = 9, p<0.05). The other individual sugars were not
significantly associated with either the GI or II values.
80 70605040 30
30
40
50
60
70
Average GI value (%)
Average II value (%)
y = 16.451 + 0.65705x; r = 0.87, n = 9, p<0.001
23
7. Implications
Using glucose as the reference food (GI = 100), foods with a GI value of 55 or less are currently
considered to be low-GI foods. Foods with a GI value between 56-69 have an intermediate or
moderate GI rating, and foods with a GI value of 70 or more are high-GI foods. Therefore, the Yellow
Box, Stringybark, Red Gum, Iron Bark and Yapunya honeys are low GI foods and are more suitable for
consumption, in controlled amounts, by people with diabetes and other health problems associated
with poor blood glucose control (eg. pancreatic disease, polycystic ovarian syndrome, Diabetes), in
line with their dietary requirements. Commercial blend # 2 (SA) and Salvation Jane honeys are
moderate GI foods and the Commercial blend # 1 (NSW) honey is a high-GI food. There is no cut-off
value for insulin index. At present, we do not know the clinical significance of a food which has a low
GI but high insulin index.
The results of this study show that different honeys can have significantly different effects on blood
glucose and insulin levels, due to differences in their sugar content and physical form, and should not
all be classified as one type of food for people with diabetes.
8. Recommendations
• The low GI honeys such as Yellow Box, Stringybark, Red Gum, Iron Bark, Yapunyah and the
moderate GI honeys such as Commercial blend #2 and Salvation Jane can be marketed by stating
in their promotional materials that the GI values of the honeys were measured using valid scientific
methodology through this project.
• The values should be published in relevant GI publications particularly in the future editions of
Brand-Miller’s books about the GI (The GI Factor series) which will be appropriately referenced.
• Finally there may be more floral varieties of honey that need to be tested. One question that needs
further research – is any pure floral honey low GI? For example, is any Yellow Box honey low GI?
24
9. References
1. AOAC (2000) Official Methods of Analysis of the Association of Analytical Chemists. 17th
edition.
2. Australian Honey Quality Specifications 2001. Wescobee Honey – Quality Data
www.wescobe.com. 14/2/2001
3. Brand-Miller JC (1994) The importance of glycemic index in diabetes. American Journal of Clinical
Nutrition 59 (Suppl), 747 S – 752 S.
4. Brand-Miller J (1995) International tables of glycemic index. American Journal of Clinical Nutrition
62 (Suppl), 871 S – 890 S.
5. Brand-Miller J, Foster-Powell K, Colagiuri S & Leeds T (2000) The GI Factor (revised edition).
Sydney: Hodder & Stoughton
6. Brand-Miller, J. and Foster-Powell, K. (1999). Diets with a low Glycaemic Index: From Theory to
Practice. Nutrition Today Vol 34 No 2
7. Chandler, B.V., Fenwick, D., O’Hara, T. and Reynolds, T. (1974). Composition of Australian
Honeys. CSIRO Aust. Div. Fd. Res. Tech. Pap No. 38, 1-39.
8. CODEX STAN 12-1981, Rev 1 (1987). Agriculture, Fisheries and Forestry Australia
9. Crane, E. 1976. Honey, A comprehensive survey. Heinemann. London
10. D’Arcy, B., Caffin, N., Bhandari, B., Squires, N., Fedorow, P. and Mackay, D. (1999). Australian
liquid honey in commercial bakery products. RIRDC publication No 99/145.
11. English, R. and Lewis, J. (1992). Nutritional Values of Australian Foods Australian Govt
Publishing Service, Vic
12. FAO/WHO Expert Consultation Report (1997). Rome 14-18 April Carbohydrates in human
nutrition. FAO Food and Nutrition Paper 66.
13. Gurr, M. 1997. Nutritional and Health Aspects of Sugars. Evaluation of New Findings. ILSI series
1997.
14. Jenkins DJA, Wolever TMS, Taylor RH, et al. (1981) Glycemic index of foods: a physiological
basis for carbohydrate exchange. American Journal of Clinical Nutrition 34, 362 - 366.
15. Joint FAO/WHO Report (1998) Carbohydrates in Human Nutrition. FAO Food and Nutrition, Paper
66. Rome: FAO.
16. Lower Clarence Skills Centre (1996). The Study into Potential for the expansion of Clarence
Valley Honey Industry.
17. Roberts, S.B. 2000. High Glycaemic Index Foods, Hunger and Obesity: Is there a connection?
Nutrition Reviews vol 58, No. 6.
18. Rostaim Faraji – Harem, 1976 Colour and Chemical Composition of Honeys from Known Floral
Sources. PhD Thesis UNSW.
19. Stern, R. (1999) What is the Glycaemic Index? The Australian Beekeeper
20. Warhurst, P. DPI Note, Beekeeping honey grades. Department of Primary Industries Queensland.
21. White, J.W. And Underwood, J.C. (1974). Maple syrup and honey. In Symposium Sweeteners GE
Intlett et, al. eds AVI Publishing Co., Westpoint U.S.A.
22. Wills, RBH, Balmer, N & Greenfield, HG. (1980) Composition of Australian Foods.2.Methods of
Analysis. Food Tech. Aust. 32(4):198-204.
23. Winner,B. Capilano Product Specification.
www.capilano.com.au/honey/products/prod_spec2.html 22/2/2001
