Nutrition Research Reviews (2003), 16, 163191 DOI: 101079NRR200371 [610260]

Nutrition Research Reviews (2003), 16, 163–191 DOI: 10·1079/NRR200371
© The Author 2003
Health potential of polyols as sugar replacers, with emphasis on
low glycaemic properties
Geoffrey Livesey
Independent Nutrition Logic, Pealerswell House, Wymondham, Norfolk NR18 OQX, UK
Polyols are hydrogenated carbohydrates used as sugar replacers. Interest now arises because of
their multiple potential health benefits. They are non-cariogenic (sugar-free tooth-friendly), low-glycaemic (potentially helpful in diabetes and cardiovascular disease), low-energy and low-insulinaemic (potentially helpful in obesity), low-digestible (potentially helpful in the colon),osmotic (colon-hydrating, laxative and purifying) carbohydrates. Such potential health benefitsare reviewed. A major focus here is the glycaemic index (GI) of polyols as regards the healthimplications of low-GI foods. The literature on glycaemia and insulinaemia after polyol inges-tion was analysed and expressed in the GI and insulinaemic index (II) modes, which yielded thevalues: erythritol 0, 2; xylitol 13, 11; sorbitol 9, 11; mannitol 0, 0; maltitol 35, 27; isomalt 9, 6;lactitol 6, 4; polyglycitol 39, 23. These values are all much lower than sucrose 65, 43 or glucose100, 100. GI values on replacing sucrose were independent of both intake (up to 50 g) and thestate of carbohydrate metabolism (normal, type 1 with artificial pancreas and type 2 diabetesmellitus). The assignment of foods and polyols to GI bands is considered, these being: high (>70), intermediate (> 55–70), low (> 40–55), and very low (< 40) including non-glycaemic; thelast aims to target particularly low-GI-carbohydrate-based foods. Polyols ranged from low tovery low GI. An examination was made of the dietary factors affecting the GI of polyols andfoods. Polyol and other food GI values could be used to estimate the GI of food mixtures con-taining polyols without underestimation. Among foods and polyols a departure of II from GIwas observed due to fat elevating II and reducing GI. Fat exerted an additional negative influ-ence on GI, presumed due to reduced rates of gastric emptying. Among the foods examined, theinteraction was prominent with snack foods; this potentially damaging insulinaemia could bereduced using polyols. Improved glycated haemoglobin as a marker of glycaemic control wasfound in a 12-week study of type 2 diabetes mellitus patients consuming polyol, adding to otherstudies showing improved glucose control on ingestion of low-GI carbohydrate. In general someimprovement in long-term glycaemic control was discernible on reducing the glycaemic loadvia GI by as little as 15–20 g daily. Similar amounts of polyols are normally acceptable.Although polyols are not essential nutrients, they contribute to clinically recognised mainte-nance of a healthy colonic environment and function. A role for polyols and polyol foods tohydrate the colonic contents and aid laxation is now recognised by physicians. Polyols favoursaccharolytic anaerobes and aciduric organisms in the colon, purifying the colon of endotoxic,putrefying and pathological organisms, which has clinical relevance. Polyols also contributetowards short-chain organic acid formation for a healthy colonic epithelium. Polyol tooth-friendliness and reduced energy values are affirmed and add to the potential benefits. In regardto gastrointestinal tolerance, food scientists and nutritionists, physicians, and dentists have intheir independent professional capacities each now described sensible approaches to the use andconsumption of polyols.
Diabetes; Coronary heart disease; Caries; Laxation; Digestive health; Glycaemic index;
Insulinaemic index; Food energy; Sucrose; Sucrose replacer; Polyol; Erythritol; Xylitol;
Sorbitol; Mannitol; Maltitol; Isomalt; Lactitol; Polyglycitol
Abbreviations: DM, diabetes mellitus; FPG, fasting plasma glucose; GI, glycaemic index; GL, glycaemic load; GT, glucose tolerance;
HbA1c, glycated or glycosylated haemoglobin; II, insulinaemic index; RIR, relative insulin response.
Corresponding author: Dr G. Livesey, fax +44 1953 600218, email glivesey@inlogic.co.uk

Introduction
Although polyols (hydrogenated carbohydrates) have been
reviewed from various perspectives (Wang & van Eys,1981; Ziesenitz & Siebert, 1987; Dills, 1989; Livesey,1992, 2001; Zumbé et al . 2001) no reviews have consid-
ered their glycaemic indices (GI) and there has been littleconsideration of their prospects in respect of the health ofthe digestive tract other than their role in caries prevention.GI ranks foods and carbohydrates according to their abilityto raise the concentration of glucose in the blood (Jenkinset al. 1981). Also the overall glycaemic load (GL) from the
diet has implications for the development and managementof metabolic syndrome, diabetes, and CHD and the controlof metabolic markers such as glycated proteins (glycatedhaemoglobin (HbA
1c), fructosamine), plasma triacylglyc-
erols, HDL and sensitivity to insulin (Brand et al . 1991;
European Association for the Study of Diabetes, 1995,2000; Salmerón et al . 1997 a,b; Food and Agriculture
Organization, 1998; Frost et al . 1998, 1999; Bär, 2000;
Bastyr et al . 2000; Buyken et al . 2000, 2001; Canadian
Diabetes Association, 2000; Diabetes UK, 2000, 2002; Liuet al. 2000 a,b; Stratton et al. 2000; Bellisle, 2001; Ford &
Liu, 2001; Gilbertson et al . 2001; Kapur & Kapur, 2001;
Khaw et al. 2001; International Diabetes Institute Australia,
2002; Jenkins et al. 2002; Livesey, 2002 a).
Widespread knowledge of the GI concept largely post-
dates many relevant studies on polyols and those studieshaving reported the GI of polyols have sometimes used cal-culation methods that are no longer acceptable. These stud-ies are revisited to place information available in a moderncontext. The multiple potential health benefits from usingpolyols as replacers of sugars, maltodextrins and glucosesyrups or in laxation are examined under the concepts ofglycaemia and insulinaemia, reduced energy, caries reduc-tion, and digestive health. The review begins with back-ground on the definition, description and metabolism ofpolyols.
Polyols
Definition of ‘polyol’
‘Sugar replacer’, ‘sugar alcohol’, ‘hydrogenated carbohy-
drate’, and ‘polyol’ are synonyms for a sub-class of carbo-hydrates present in foods. The defining characteristic is theoccurrence of an alcohol group (>CH-OH) in place of thecarbonyl group (>C=O) in the aldose and ketose moietiesof mono-, di-, oligo- and polysaccharides; hence polyolsare not sugars, and generally carry the suffix ‘-itol’ in placeof the suffix ‘-ose’ according to modern carbohydratenomenclature (McNaught, 1996). The name ‘polyol’ is anabridgement of ‘polyalcohol’ or ‘polyhydric alcohol’.Preferred names are ‘polyol’ or ‘hydrogenated carbohy-drate’; the latter makes explicit that these substances arecarbohydrate. Individual polyols are described in Table 1and in more detail later (p. 164).
Classification amongst other carbohydrates
Because polyols are not sugars they are permitted in sugar-
free and tooth-friendly products (European Communities,1994). The distinction between sugars and polyols is
important yet frequently overlooked, the consultation bythe Food and Agriculture Organization (1998) being themost significant recent example. Sugars are legally definedfor nutrition labelling purposes as mono- and disaccharideonly. In contrast, polyols may be hydrogenated mono-, di-,but also oligo- and polysaccharide (Table 2). Polyols alsocontribute unavailable carbohydrate to fermentation analo-gous to dietary fibre to which it may contribute (AmericanAssociation of Cereal Chemists, 2001). Examples of namedcarbohydrates in each subclass of food carbohydrates aregiven in Table 2 to help show their difference from polyols.The overall order in which the carbohydrates are listed hereis governed by molecular weight or degree of polymerisa-tion, as suggested by the Food and AgricultureOrganization (1998). However, no real physiological mean-ing can be attached to this order, nor does this order helpinterpretation of carbohydrate terminology in regulatoryfood codes. To be usefully informative the nutrition infor-mation panel will in future require other information; inthis context the GI, GL (GI ×amount of carbohydrate) and
other possible expressions of glycaemic potential are candi-dates for possible inclusion in future food labelling andfood tables.
Individual polyols: description, absorption and metabolism
The physiological attributes of polyols, i.e. low cariogenic-
ity, low glycaemia, low insulinaemia, low energy value,source of substrate for a healthy colon and intestinal toler-ance are linked through the common property of polyolsbeing difficult to digest or slow to metabolise yet relativelyeasy to ferment in the colon. This property results from thehindrance to digestion and absorption by the alcohol groupthat replaces the carbonyl group and the occurrence of sac-charide linkages other than the α1–4 and α1–6 present in
starches and sucrose. Thus, a low digestibility and/or slowhepatic glucose release is the determinant of their low gly-caemic and insulinaemic response properties.
During the time polyols are resident in the mouth, they
resist fermentation and acidogenesis by the micro-organ-isms of dental plaque (Willibald-Ettle & Schiweck, 1996;Kandelman, 1997) and are not absorbed via the stomach toany significant degree. Absorption that does occur is bypassive diffusion of monosaccharide polyol along a con-centration gradient (Herman, 1974). Disaccharide andhigher polyols are too large to diffuse from the gut into thecirculation in amounts more than 2 % of oral intake(Livesey, 1992). Some di-, oligo- and polysaccharide poly-ols may liberate glucose, but as their digestion is slow andincomplete this does not result in a substantial rise in bloodglucose, as will be shown in later sections (p. 168). Thesmall intestine is probably less permeable distally so thatco-released monosaccharide polyol may be less readilyabsorbed than the same monosaccharide polyol takenorally. Once absorbed, monosaccharide polyols areexcreted via the kidneys, oxidised directly or converted toglycogen or glucose in the liver; the route of metabolismand excretion depends on their structure. Unabsorbed car-bohydrate from polyols is generally fermented completelyby the colonic microflora (Livesey, 1992).
164 G. Livesey

Health potential of polyols 165
Table 1. Polyol specifications
Saccharide Molecular
Polyol Formula type Generic form weight (Da) Synonyms* Further details†
Erythritol C4H10O4Mono- Tetritol 122 .12 Hydrogenated erythrose FNP 52/7
meso -Erythritol
Erthritetetra-Hydroxybutane
1,2,3,4-ButanetetrolErythrolPhysitol
Xylitol C
5H12O5Mono- Pentitol 152 .15 Hydrogenated xylose FNP 52/4
Xylite
Mannitol C6H14O6Mono- Hexitol 182 .17 Hydrogenated mannose FMP 52/4
D-Mannitol
Mannite
Sorbitol C6H14O6Mono- Hexitol 182 .17 Hydrogenated glucose FNP 52/4
D-Sorbitol
GlucitolSorbolSorbit
Sorbitol syrup Mixed mono- and smaller amounts of other hydrogenated Hydrogenated glucose syrup FNP 52/4
saccharides
‡D-Glucitol syrup
Lactitol C12H24O11Di- Hexopyranosyl- 344 .3 Hydrogenated lactose FNP 52/4
hexitol β-D-Galactopyranosyl-1-4- D-sorbitol
β-D-Galactopyranosyl-1-4- D-glucitol
LactositolLactitLactosbiosit
Isomalt C
12H24O11Mixed di- Hexopyranosyl- 344 .3 Hydrogenated isomaltulose FNP 52/4
hexitol Hydrogenated palatinose
Mixture of α-D-glucopyranosyl-1-6- D-sorbitol
and α-D-glucopyranosyl-1-1- D-mannitol
Maltitol C12H24O11Di- Hexopyranosyl- 344 .3 Hydrogenated maltose FNP 52/4
hexitol α-D-Glucopyranosyl-1-4- D-sorbitol
α-D-Glucopyranosyl-1-4- D-glucitol
Maltitol syrups Mixed, ≥50 % di-, and lesser amounts of mono- and Hydrogenated high-maltose glucose syrup FNP 52/5
higher saccharides‡ Hydrogenated starch hydrolysate
Dried maltitol syrupMaltitol syrup powderSeveral forms are available:
Regular, about 53 % maltitolIntermediate, about 73 % maltitolHigh, about 98 % maltitolHigh polymer, about 50 % hydrogenated
polymer
Polyglycitol Mixed, < 50 % di- and of other especially oligo- and Polyglucitol FNP 52/6
polysaccharides‡ Hydrogenated starch hydrolysate
FNP 52, Food and Nutrition paper 52.
* Excluding proprietary names.† Food and Agriculture Organization (1996–1999), addenda 4–7.‡ For details, see p. 167.
Representative values for the absorption, fermentation
and urinary excretion of polyols are shown in Table 3. Thedata are drawn from information collected by Livesey(1992), the Life Sciences Research Office (1994, 1999),and other material described later (pp. 166–168). For thepresent, no distinction is made between the results ofdigestibility studies assessing absorption from liquids andsolids on the ground that such distinctions at moderatepolyol intake are based on invasive methodology in whichsolids may increase the non-recovery of polyols at theileum by increasing retention in the stomach and upper
gastrointestinal tract rather than increasing absorption.Differences in the tolerance of polyols when consumed inliquid and solid meals are ascribable to different rates ofstomach emptying rather than differences in the extent ofdigestion and absorption (Livesey, 1990 a, 2001).
Excessive intake might cause absorption to be lowered andpotentially would affect the glycaemic response to polyols,though, as will be seen in subsequent dose–response data(p. 169), this appears not to happen.


Erythritol. This small (four-carbon, tetritol) molecule is
absorbed readily by diffusion, with approximately 10 %escaping to the large intestine in man (Oku & Noda, 1990;Noda et al. 1994; Bornet et al. 1996 a). Absorbed erythritol
distributes widely through the tissues but its metabolism isminimal and being poorly reabsorbed via the kidneys it isessentially excreted unused in urine (Bernt et al. 1996).
Xylitol. Absorption of xylitol from the small intestine
occurs less readily than the smaller molecule erythritol,causing more to be fermented in the large bowel. Estimatesof the extent of fermentation range from 50 to 75 %(Livesey, 1992; Life Sciences Research Office, 1994) withthe lower value being more consistent with the size of thismolecule. Thus, based on
D-arabitol as a non-metabolisable
marker of pentitol absorption, a similar absorption of oralxylitol in man would suggest it to be 53 % (Bär, 1990). Thisis corroborated by the present author who has predicted itsabsorption based on molecular weight for a series of polyols(glycerol, erythritol, mannitol and lactitol) to be 48 % (seeLivesey, 1992). On the basis of energy values for xylitolproposed by several experts and authorities, absorbability byconsensus is 49 %; this being the average of values esti-mated by the Dutch Nutrition Council (1987), Bär (1990),
Bernier & Pascal (1990), Livesey (1992); Life SciencesResearch Office (1994), and Brooks (1995). The liver read-ily sequesters absorbed xylitol where it is dehydrogenatedby a non-specific cytoplasmic NAD-dependent dehydroge-nase (synonyms iditol dehydrogenase; polyol dehydroge-nase). The xylulose so produced is phosphorylated via aspecific xylulokinase to xylulose-5-phosphate, an intermedi-ate of the pentose-phosphate pathway before conversion toglucose, which is only slowly released into the bloodstreamor stored as glycogen (Keller & Froesch, 1972).
Mannitol. Various forms of evidence indicate that approx-
imately 25 % of oral mannitol in solution is absorbed(reviewed in Livesey, 1992). Absorbed mannitol is excretedin urine because it is virtually non-metabolisable in the tis-sues (Nasrallah & Iber, 1969) and the remainder or unab-sorbed mannitol is slowly fermented.
Sorbitol. Estimates of absorption from oral solutions
range from 25 to 80 % of the ingested dose (Beaugerie et al . 1990; Livesey, 1992), with the lower value being166 G. Livesey
Table 2. Classification of the major carbohydrates in foods (modified from Food and Agriculture Organization, 1998)
Class DP Sub-class Examples
Monosaccharides 1 Sugars Glucose, fructose, galactose
Hydrogenated monosaccharides Erythritol, xylitol, mannitol, sorbitol
Disaccharides 2 Sugars Sucrose, maltose, lactose, trehalose
Hydrogenated disaccharides Maltitol, isomalt, lactitol
Oligosaccharides 3–9 Malto-oligosaccharides Maltodextrins.
Other oligosaccharides Raffinose, stachyose, fructo-oligosaccharides,
galacto-oligosaccharides.
Hydrogenated oligosaccharides Hydrogenated starch hydrolysate
Polysaccharides >9 Starch Amylose, amylopectin, modified starches
NSP Cellulose, hemicelluloses, pectins, etc
Hydrogenated polysaccharides Polyglycitol, hydrogenated polydextrose
DP, degree of polymerisation.
Table 3. Approximate absorption, fermentation and urinary excretion of polyols*
Absorption (g/100 g) Fermentation (g/100 g) Urinary excretion (g/100 g)
Erythritol 90 10 90
Xylitol 50 50 <2Sorbitol 25 75 <2Mannitol 25 75 25Isomalt 10 90 <2Lactitol 2 98 <2Maltitol 40 60 <2Maltitol syrup
Regular, intermediate, high about 50† about 50† <2High-polymer about 40‡ about 60‡ –
Polyglycitol about 40† about 60† <2
* Data are given to the nearest 5 %, except when close to zero, when data are to the nearest 2 % or for urinary excretion where an
upper limit of 2 % appears. For references, see pp. 166–168.
† Data are based solely on glycaemic and insulinaemic responses, which may give a lower limit.‡ Based on in vitro digestion.

more consistent with the size of this molecule (Livesey,
1992) and the higher value possibly due to the use of inva-sive methodology and non-recovery. Slow and late
14CO2
excretion from labelled sorbitol (compared with glucose) innon-invasive studies in human subjects suggests lowerabsorption (Tsuji et al. 1990) though this could also be due
to the temporal storage of [
14C]carbon as glycogen.
Absorbed sorbitol is practically metabolised fully as only atrace is excreted (Adcock & Grey, 1957). Dehydrogenationin the liver is via the non-specific cytoplasmic NAD-depen-dent dehydrogenase, as for xylitol, with the production offructose then glycogen or glucose that may be slowlyreleased into the bloodstream. Unabsorbed sorbitol isextensively fermented to short-chain organic acids andgases (Hyams, 1983), with a considerable yield of butyricacid in vitro (Mortensen et al. 1988; Clausen et al. 1998).
Sorbitol syrup. The biological response to sorbitol syrup
is based on the combined individual responses to its con-stituents, which are mainly sorbitol and mannitol (Table 1,see earlier; p. 166).
Maltitol. This is a disaccharide polyol (moieties of glucose
and sorbitol; Table 1) for which hydrolysis is required beforeabsorption. Absorption in human subjects is reported to rangefrom 5 to 80 % (Beaugerie et al. 1990; Life SciencesResearch Office, 1999); the wide range is partly due to theuse of invasive methods and partly due to the incorrect evalu-ation of results from non-invasive methods. Account needs tobe taken of three non-invasive study approaches in humansubjects. First, comparison of the time course of
14CO2pro-
duction from [U14C]maltitol, [U14C]glucose (fully available)
and [U14C]fructo-oligosaccharides (fully unavailable) (see
data in Livesey, 1993) indicates by the simplest of computa-tional models a lower limit to absorption of 35 % for maltitol(10 g) in solution. Second, glycaemia and insulinaemia (seepp. 167–168) indicate a lower limit to absorption of 35 to27 % respectively formaltitol (25–50 g) in solution. Third,based on indirect calorimetry following the ingestion of ahigh-polymer maltitol syrup containing 50 % maltitol and 50% polymer and separate study of the polymer fraction(Sinaud et al. 2002) the energy value of maltitol can be esti-mated. This estimated energy value corresponds to maltitolabsorption of approximately 32 % when consumed in threemixed solid meals interspersed by three maltitol drinks(totalling 50 g maltitol in 50 g polymer daily). On the basis ofenergy values for maltitol proposed by several authorities,absorbability by consensus is 45 % (Dutch Nutrition Council,1987; Bär, 1990; Bernier & Pascal, 1990; Life SciencesResearch Office, 1994, 1999; Brooks, 1995; Australia NewZealand Food Authority, 2001; American DiabetesAssociation, 2002; Food and Agriculture Organization,unpublished results). The products of hydrolysis by intestinalbrush-border disaccharidases are glucose and sorbitol, themetabolism of which has been described earlier (p. 166).Maltitol syrup(s). These are hydrogenated starch
hydrolysates and consist of a mixture of sorbitol, maltitol,and hydrogenated oligo- and polysaccharides (Table 1).The terminology ‘regular-, intermediate- and higher-malti-tol syrups and high-polymer maltitol syrup’ is applied hereto conveniently identify four distinctly different products,all of which bear the same general name ‘maltitol syrup’.Information on the availability of carbohydrate from hydro-genated oligo- and polysaccharide fractions of regular,intermediate- and high-maltitol syrups is not evidentlyavailable. However, based on glycaemic and insulinaemicresponse data (derived later, see p. 169), it is probably closeto 50 %.
A maltitol syrup comprising 50 % maltitol and 50 %
hydrogenated polymer has recently been introduced(Sinaud et al . 2002), which here is referred to as ‘high-
polymer maltitol syrup’. The high-polymer fraction isobtained by heating starch at high temperature and lowmoisture in the presence of an acid catalyst, which yieldsafter separation a product with an average degree of poly-merisation of about 17, the introduction of 1–2 and 1–3glucosidic linkages and so a proportion of branched link-ages. Digestibility of the high-polymer maltitol syrup in
vitro is about 40 % based on hydrolysis with α-amylase and
amyloglucosidase and the release of sorbitol and glucose(Sinaud et al. 2002). This value is consistent with the gly-
caemia and insulinaemia described in the present review.
Polyglycitol syrup. Similar to the maltitol syrups this is a
hydrogenated starch hydrolysate, though it has more sorbitol(< 20 v.< 8 %) and less maltitol (< 50 v.≥ 50 %). The
absorption of carbohydrate from polyglycitol syrup is uncer-tain in extent. However, with a GI and insulinaemic index(II) similar to those for maltitol (see pp. 169–171) it proba-bly has a similar small-intestinal digestibility, at about 40 %.
Isomalt. This is a mixed disaccharide polyol (Table 1).
The products of hydrolysis are glucose, sorbitol and manni-tol, the metabolism of which is described earlier (p. 166).However, a variety of studies including non-invasive meth-ods in human subjects and methods in animals (Livesey,1990 a,b, 2000 a) together with the present studies on gly-
caemia and insulinaemia suggest 0 to 14 % of isomalt isavailable as carbohydrate in man. On the basis of theenergy values of isomalt suggested by various authoritiesand experts (Dutch Nutrition Council, 1987; Livesey, 1992;Life Sciences Research Office, 1994; Brooks, 1995) a con-sensus of approximately 90 % is fermented in the colon,with a stoichiometry in vivo and in vitro indicating rela-
tively little H
2gas production (Livesey et al. 1993).
Lactitol. Very little of this disaccharide polyol is absorbed,
perhaps 2 % as lactitol and its hydrolysis products galactoseand sorbitol. This is due to a very low activity of Health potential of polyols 167

/H9252-galactosidase in the human intestine (Nilsson & Jägerstad,
1987; Grimble et al. 1988). The liver readily uses absorbed
galactose and sorbitol in either hepatic glycogen storage orhepatic glucose production. Unabsorbed lactitol is com-pletely fermented with a stoichiometry giving a generousyield of H
2gas in vivo and in vitro (Livesey et al. 1993) and
butyric acid in vitro (Clausen et al. 1998).
Glycaemia and insulinaemia
Definitions
Glycaemic index. GI is a measure of a specific property of
carbohydrate in a food or meal or diet (Jenkins et al. 1981;
Wolever et al . 1991; Food and Agriculture Organization,
1998). It is defined as ‘the incremental area under the bloodglucose response curve of a 50 g carbohydrate portion of atest food expressed as a percentage of the response to thesame amount of carbohydrate from a standard food taken bythe same subject’ (Food and Agriculture Organization,1998). In this definition carbohydrate usually means avail-able carbohydrate, though has included for comparative pur-poses any carbohydrate that might replace availablecarbohydrate in a foodstuff (Pelletier et al. 1994; Bär, 2000;
Zumbé et al . 2001; Foster-Powell et al . 2002; Sydney
University’s Glycaemic Index Research Service, 2002). The50 g carbohydrate portion mentioned in the definition is notalways practical and smaller portions (down to 25 g) can beused when this is more realistic of the conditions of con-sumption. The standard food mentioned in the definition isusually glucose in water or white bread. To avoid confusion,it is useful to express the GI relative to glucose (for exam-ple, G = 100 GI units), and to state the standard food (beingwell defined, glucose is preferred) and its GI. Measurementsare usually made on eight to ten adults consuming each testfood on one occasion and the standard food ideally on threeoccasions. The calculation of GI has been standardised(Food and Agriculture Organization, 1998) and applied herewith the following additional instructions: calculations wereperformed on group mean plasma glucose responses to car-bohydrate ingestion; this minimises a bias caused by dis-counting below-baseline areas that occur due to randomeffects. Baseline values were taken at zero time rather thanaveraged across time before zero time; this minimises a biasdue to the fall in basal glucose concentrations with time inthe basal state. In studies reporting GI values these calcula-tions were still necessary to ensure a standard and consistentapproach was used.
GI values obtained in normal individuals usually apply to
those with abnormal carbohydrate metabolism (Wolever et
al. 1987; Foster-Powell et al. 2002); namely patients with
type 1 diabetes mellitus (DM) (previously called juvenile orinsulin-dependent DM, which results from inadequateinsulin secretion) and more commonly type 2 DM patients(previously called adult or late onset diabetes, which isassociated with the resistance of tissues to insulin). A provi-sional WHO classification of diabetes is available (Alberti& Zimmet, 1998), together with useful desktop guides on
type 1 and 2 DM (European Diabetes Policy Group,1999 a,b), and criteria for impaired glucose tolerance (GT)
and impaired fasting glycaemia (Unwin et al. 2002).
Glycaemic load. GL is formally the product of the carbo-
hydrate content and GI of a food and so is primarily a mea-sure of the quantity and apparent quality of thecarbohydrate in the food item and has units of weight (g).Foods with the same GL have practically the same impacton the integrated blood-glucose response, which in diabetesmanagement is the main target.
Insulinaemic index. II is obtained under identical condi-
tions to those for GI, simply replacing the measure of glu-cose with a measure of insulin. The index was introduced asa result of possible concern that blood-glucose responsesmight not adequately reflect the responses of the major ana-bolic hormone insulin, which is central to abnormal carbohy-drate metabolism in DM (Holt et al. 1997; Wolever, 2000).
Insulin load. IL is calculated in the same way as GL, but
replacing glucose measurements with insulin measurements.
Composite foods, meals and diets. The composite GL is
the sum of GL from each food or ingredient item. Dividingthis sum by the sum weight of the carbohydrate eaten givesthe composite GI. Substitution of measures of glycaemiawith measures of insulinaemia gives the composite insulinload and composite II.
Statistics. Studies from which GI and II values were calcu-
lated were generally of similar size and for simplicity wereconsidered of equal weight when deriving overall means.
Time course of acute glycaemic responses to polyols
Glycaemic responses to sugars and polyols in fasted normal
individuals of both sexes were summarised from the literature(Fig. 1). The curves are representative of 25 g doses taken inwater or tea without milk or other nutrients (80 to 500 ml).The sugars (glucose and sucrose) result in higher responses30–60 min after ingestion and lower glucose concentrationsafter 90 min than any of the polyols (erythritol, xylitol, sor-bitol, mannitol, maltitol, isomalt, and lactitol). The responsesto all polyols are lower or much lower than for sucrose.Glycaemic and insulinaemic responses (incremental areas) forsucrose and polyols, relative to equivalent intakes of glucose,were calculated for the studies represented in Fig. 1 and otherstudies. The responses were calculated using a wider range ofintakes (10 to 70 g), type 1 and type 2 DM patients in addi-tion to normal subjects, and maltitol syrups and polyglycitolin addition to the other polyols mentioned (Table 4). In the168 G. Livesey

majority of publications such calculations had either not been
undertaken or had been undertaken incorrectly due to the lit-erature pre-dating knowledge of the current GI calculationmethod. In type 1 DM patients supported by an artificial pan-creas the rate of insulin delivery was in some cases used as asurrogate for insulinaemia.
Glucose and insulin measurements were invariably made
on venous plasma or capillary blood. Glucose was the mostcommon reference carbohydrate used. In a small number ofcases sucrose was the reference carbohydrate, in whichcase responses were still expressed relative to glucose hav-ing 100 GI units. Statistical presentations (means, standarderrors and differences) are omitted from Table 4 due to thepossible heterogeneous nature of the data with respect tothe level of polyol intake and the condition of subjects’ car-bohydrate metabolism, which are now examined.
Glycaemic responses in normal, type 1 and type 2 diabetes
mellitus subjects
Information on glycaemic responses was available for sor-
bitol, isomalt and hydrogenated starch hydrolysates (inter-mediate- and high-maltitol syrups combined) in normal,type 1 and type 2 DM subjects and for maltitol in normaland type 1 DM subjects. Diabetics had HbA
1cvalues of less
than 12 % indicating a degree of glucose control, thoughtless than aimed for nowadays. For each polyol, glycaemicresponses expressed relative to glucose in both types of dia-betes were similar to those in normal subjects (Fig. 2).
Relationship of glycaemic response to intake of polyols
Information was available for sucrose, maltitol, high-malti-
tol syrup, isomalt, lactitol and sorbitol to assess the rela-tionship between intake and glycaemic response relative tothe most commonly used reference, glucose (Fig. 3).
Sorbitol, isomalt and lactitol had very low to littleresponses at all intakes and there was no association withdose. Responses tended to fall either significantly ornumerically with increasing dose for sucrose ( P< 0·02),
maltitol ( P= 0·06) and high-maltitol syrup ( P= 0·16). Such
possible dose dependence is not limited to soluble carbohy-drates as it is also observed for bread v.glucose (Jenkins et
al. 1981; Wolever & Bolognesi, 1996; Lee & Wolever,
1998).
Sucrose is a carbohydrate often replaced by polyols in
foodstuffs. When incremental glucose-response areas forpolyols were re-expressed relative to a sucrose standard setat 65 GI units at all intakes (Fig. 3 (b)), the relative gly-caemic response to all polyols was clearly independent ofdose.
Glycaemic and insulinaemic indices of polyols
The achievement of low postprandial glycaemia is an
important goal and has greater significance when accompa-nied by low insulinaemia. All polyols had lower GI and IIvalues than either glucose or sucrose (Table 5).
Among these carbohydrates GI and II were related prac-
tically linearly (Fig. 4) with a slope of association of 0·75(
SE0·05) (dimensionless); this slope is significantly less
than might be expected ( P< 0·0001); for glucose the value
would by definition fall on a line passing through the originof slope 1·00.
Variations about mean GI and II values possibly
increased with increasing value; homogeneity was achievedby transformation to the square root (Fig. 4 (b)).Observations falling below the line of identity (Fig. 4 (a)and (b)) are consistent with causing demand on the pan-creas for insulin that is lower than that due to glucose, andHealth potential of polyols 169Incremental glucose (mmol/l)
0 30 60 90 120 150 180
Time (min)–0.50.00.51.01.52.02.53.0
Incremental glucose (mmol/l)
0 30 60 90 120 150 180
Time (min)–0.50.00.51.01.52.02.53.0
Fig. 1. (a), Glycaemic curves for glucose (––), sucrose ( ––) and polyols maltitol ( –-), isomalt (– – –) and lactitol ( ––) in normal individuals;
(b) glycaemic curves for glucose (––), sucrose ( ––) and polyols xylitol ( –-), sorbitol (– – –), erythritol ( ····) and mannitol ( ––) in normal
individuals. Data from several publications were pooled to yield curves representative of 25 g doses (20–64 g for erythritol) i n water or tea (80
to 500 ml) without other nutrients. In practice, individual studies used various doses, and dose was used as a covariate at eac h time point to
obtain curves representing 25 g intake. Based on data in Table 4 for normal subjects.

170 G. Livesey
Table 4. Estimates of the relative glucose response (RGR) and relative insulin response (RIR) to sucrose and polyols (glucose = 100)
Composition (%)
Reference (RGR, RIR)* Intake (g) RGR RIR Subjects n S Ma H Source
Sucrose
Glucose (100, 100)† 20 89 33 Normal 9 M MacDonald et al. (1978)
Glucose (100, –) 20 87 na Normal 12 Samata et al. (1985)
Glucose (100, –) 20 89 na Type 2 DM 8 Samata et al. (1985)
Glucose (100, –) 20 79 na Type 1 DM 6 Samata et al. (1985)
Glucose (100, 100) 25 58 58 Normal 4 M + 4 F Lee & Wolever (1998)Glucose (100, 100) 25 80 43 Normal 8 M Pelletier et al. (1994)
Glucose (100, 100)† 35 58 23 Normal 9 M MacDonald et al. (1978)
Glucose (100, 100) 50 58 45 Normal 4 M + 4 F Lee & Wolever (1998)Glucose (100, 100)† 50 64 43 Normal 9 M MacDonald et al. (1978)
Glucose (100, –) 50 65 na Normal – Brand-Miller et al. (1999)
Glucose (100, 100)† 70 75 45 Normal 9 M MacDonald et al. (1978)
Glucose (100, 100) 100 58 67 Normal 4 M + 4 F Lee & Wolever (1998)
Erythritol
Glucose (100, 100) 17 21 3 Normal 5 M Noda et al. (1994)
Glucose (100, 100)‡ 64 0 (–5) 1 Normal 3 M + 3 F Bornet et al. (1996 a)
Glucose (100, 100)‡ 20 0 (–20) 3 Type 2 DM 3 M + 8 F Ishikawa et al. (1996)
Glucose (100, –)‡ 40 3 – Normal 6 PD Cock, Cerestar
(unpublished results)
Xylitol
Glucose (100, 100) 20 13 4 Normal 5 M + 5 F Nguyen et al. (1993)
Glucose (100, 100) 25 9 31 Normal 8 M Natah et al. (1997)
Glucose (100, 100) 30 14 12 Normal 3 M + 3 F Salminen et al. (1982)
Glucose (100, –) 30 15 na Normal 5 M + 5 F Müller-Hess et al. (1975)
Glucose (100, 100) 50 7 14 Normal 30 Tong et al. (1987)
Glucose (100, 100) 50 18 14 Normal 5 M + 5 F Müller-Hess et al. (1975)
Mannitol
Glucose (100, 100) 25 0 0 Normal 5 Ellis & Krantz (1941)
Sorbitol
Glucose (100, 100) 20 13 4 Normal 8 M Nguyen et al. (1993)
Glucose (100, 100)† 20 7 36 Normal 9 M MacDonald et al. (1978)
Sucrose (81, –) 20 14 7 Type 1 DM§ 18 M + 6 F Kaspar & Spengler (1984)Glucose (100, –) 25 10 na Normal 7 Ellis & Krantz (1941)Glucose (100, 100)† 35 3 12 Normal 9 M MacDonald et al. (1978)
Sucrose (68, 40) 40|| 11 19 Type 2 DM 10 M + 8 F Petzoldt et al. (1982 b)
Glucose (100, –) 50 14 na Normal 2 Ellis & Krantz (1941)Glucose (100, 100)† 50 6 15 Normal 9 M MacDonald et al. (1978)
Glucose (100, 100) 50 8 6 Normal 9 (M + F) Mimura et al. (1972)
Glucose (100, –) 50 4 na Type 2 DM 13 Ellis & Krantz (1943)
Maltitol
Glucose (100, 100) 20 44 23 Normal 5 M + 5 F – 98 – Nguyen et al. (1993)
Glucose (100, 100) 25 49 30 Normal 8 M – 99 – Pelletier et al. (1994)
Glucose (100, –) 50 25 na Normal 9 (M + F) 98¶ Mimura et al. (1972)
Glucose (100, 100) 50 37 21 Type 2 DM 11 (M + F) 98¶ Mimura et al. (1972)
Glucose (100, 100) 50 39 29 Normal 12 M – 99 – Kamoi (1974)Glucose (100, –) 50 31 na Normal 14 M + 5 F – 99 – Kamoi (1974)Glucose (100, –) 50 39 na ‘Diabetic’ 14 M + 7 F – 99 – Kamoi (1974)Glucose (100, 100) 50 25 27 Type 2 DM 6 – 98 – Slama (1989)Glucose (100, 100) 50 29 33 Normal 6 – 98 – Slama (1989)
Maltitol syrups
High-maltitol syrup (about 89 % maltitol)
Glucose (100, 100)‡ 10 65 na Normal 6 – 89 – Secchi et al. (1986)
Glucose (100, 100)‡ 25 48 na Normal 6 – 89 – Secchi et al. (1986)
Glucose (100, 100) 25 37 48 Normal 8 M 5 88 7 Pelletier et al. (1994)
Sucrose (71, 34) 30 55 28 Normal 8 2 88 10 Felber et al. (1987)
Glucose (100, 100) 35 47 28 Normal 8 M + 8 F 5 89 6 Kearsley et al. (1982)
Glucose (100, 100) 50 33 36 Normal 6 – 89 – Secchi et al. (1986)
Glucose (100, 100)‡ 50 50 na Normal 6 – 89 – Secchi et al. (1986)
Intermediate-maltitol syrup (about 70 % maltitol)
Glucose (100, 100) 25 54 29 Normal 8 M 2 72 26 Pelletier et al. (1994)
Glucose (100, 100) 50 52 27 Normal 3 M + 3 F 7 69 33 Wheeler et al. (1990)
Glucose (100, 100) 50 52 na Type 1 DM 3 M + 3 F 8 69 33 Wheeler et al. (1990)
Glucose (100, 100) 50 56 66 Type 2 DM 3 M + 3 F 9 69 33 Wheeler et al. (1990)

more than simply due to the GI of polyols and sucrose
being lower than for glucose.
Composite glycaemic index, glycaemic load, insulinaemic
index and insulin load: potential interactions
Interactions between polyols and sugar, and between poly-
ols and foods. Seven studies involving polyols provided
the possibility to assess whether the sum of the GL for mealcomponents fed separately from one another would equalthe GL of the entire meal (Table 6).
A meal composed of glucose and sorbitol (monosac-
charide mixture) yielded a GL less than predicted fromthe sum of the loads for glucose and sorbitol separately(Table 6; cases 1 and 2). Incomplete hydrolysis cannotexplain this result; possibly sorbitol slows stomachemptying or hurries the glucose to a site where absorp-tion distally is less rapid (Livesey et al . 1998) or signifi-cantly dilutes luminal glucose concentration through its
osmotic effect. A similar observation is made for a mealof a disaccharide mixture, sucrose and lactitol (Table 6;case 3). Likewise a similar result is observed for sorbitoltaken in comparatively complex meals; a breakfast com-prising mainly bread and butter (Table 6; cases 4 and 5)and a protein-and-carbohydrate-based breakfast, mainlyscrambled eggs and farina cereal (1088 kJ (260 kcal);Akgün & Ertel, 1980) to which was added either sucroseor fructose or sorbitol (35 g) (Table 6; case 6). The laststudy was repeated in type 2 DM patients with similarresults (Table 6; case 7).
Similar results were obtained when GI and load were
replaced by II and load (Table 7), suggesting that the inter-action affecting glycaemia was not the result of interactionto elevated insulin secretion.
The general case is evidently that a mixture involving a
polyol yields a value less than the sum of its individualHealth potential of polyols 171
Table 4. Continued
Composition (%)
Reference (RGR, RIR)* Intake (g) RGR RIR Subjects n S Ma H Source
Regular-maltitol syrup (about 53 % maltitol)
Glucose (100, 100) 20 43 19 Type 2 DM 5 M + 5 F – >50 – Nguyen et al. (1993)
Glucose (100, 100) 25 53 47 Normal 8 M 3 53 40 Pelletier et al. (1994)
Glucose (100, 100) 35 67 36 Normal 8 M + 8 F 7 53 40 Kearsley et al. (1982)
Glucose (100, 100) 66 54 52 Normal 6 5 55 40 Slama (1989)Glucose (100, 100) 66 41 64 Type 2 DM 6 5 55 40 Slama (1989)
High-polymer maltitol syrup
Glucose (100, 100) 50 47 23 Normal 6 – 50 50 Rizkalla et al. (2002)
Glucose (100, 100) 50 25 39 Type 2 DM 6 – 50 50 Rizkalla et al. (2002)
Polyglycitol syrup
Glucose (100, 100) 25 32 16 Normal 3 M + 3 F 14 8 78 Wheeler et al. (1990)
Glucose (100, –) 30 45 na Type 1 DM 3 M + 3 F 14 8 78 Wheeler et al. (1990)
Glucose (100, 100) 35 38 30 Type 2 DM 3 M + 3 F 14 8 78 Wheeler et al. (1990)
Isomalt
Glucose (100, 100) 20 11 7 Type 1 DM§ 18 M + 6 F Kaspar & Spengler (1984)Glucose (100, 100) 25 2 8 Normal 10 Sydney University Glycaemic
Index Research Service (2002)
Sucrose (71, 34) 30 11 4 Normal 10 M Thiébaud et al. (1984)
Glucose (100, 100) 50 7 5 Type 2 DM 8 F + 16 M Petzoldt et al. (1982 a)
Sucrose (65, 45) 50 12 3 Type 2 DM 24 Drost et al. (1980)
Sucrose (65, 45) 50 6 3 Type 2 DM 3 M + 9 F Bachmann et al. (1984)
Glucose (100, 100) 50 12 15 Type 2 DM 6 Slama (1989)Glucose (100, 100) 50 8 5 Normal 6 Slama (1989)Sucrose (62, 47) 70 11 4 Normal 6 Keup & Püttner (1974)
Lactitol
Glucose (100, 100) 25 3 6 Normal 8 M Natah et al. (1997)
Glucose (100, 100) 25 7 1 Normal 7 Doorenbos (1977)
Sucrose (65, –) 50 7 na Normal 8 Zaal & Ottenhof (1977)**
S, sorbitol; Ma, maltitol; H, hydrogenated saccharides with degree of polymerisation > 2; M, male; na, information not availabl e; DM, diabetes mellitus; F, female.
* Values of RGR and RIR at the intakes of reference substrate used are shown in parentheses. These data are used to adjust to a glucose reference of 100 when
the reference substrate in the study was other than glucose.
† Unpaired reference: data were adjusted for glucose responsiveness according to the treatment group’s fasting glucose concentr ations.
‡ Unpaired reference: data taken from a separate publication with adjustment for fasting glucose concentration. RGR data in par entheses are acutal, outside
parentheses are conventional.
§ Subjects with artificial pancreas.|| Intake was 4 x 10 g doses at hourly intervals over 240 min.¶ 98 % assumed based on production by hydrogenation of maltose.** Partially reported by van Velthuijsen (1990).

172 G. Livesey
0102030405060708090100GI (glucose = 100)44 2
44 1 5 21
05
3
Normal Type 2 DM Type 1 DM
020406080100
0 1 02 03 04 05 06 0
Dose (g)Glycaemic response (glucose = 100)020406080100
02 0 4 0 6 0(b)
Sucrose
89 % Maltitol
98 % Maltitol
Sorbitol,
isomalt and
lactitolSucrose
89 % Maltitol
98 % Maltitol
Sorbitol, isomalt
and lactitol(a)Fig. 2. The glycaemic response for four polyols relative to glucose ( □) in normal, type 2 diabetes mellitus (DM) and type 1 DM subjects.
Values are the means of the responses for individual studies shown in Table 4, which cites the sources of information for the c alculations
made. The numbers of studies represented are shown above each column and the vertical bars represent either standard deviation (n≥3
studies) or range ( n2) or are absent ( n1). Hydrogenated starch hydrolysate ( ) is equally weighted information combined from polyglycitol
and regular maltitol syrup. ( ), Maltitol; ( ), isomalt; ( ), sorbitol.
Fig. 3. Glycaemic response area for sucrose and polyols relative to glucose (glycaemic index (GI) = 100) (a) or sucrose (GI = 65) (b).
Carbohydrates were taken in water or tea without other nutrients (80 to 500 ml). Data are inclusive of normal, type 2 diabetes m ellitus (DM)
and type 1 DM subjects and are from Table 4, which cites the sources of information used in the calculations. ( /H17003), Sucrose; ( /H17009._), high-maltitol
syrup; ( /H17005), maltitol; ( /H17034), sorbitol; ( /H17033), isomalt; ( /H17040), lactitol. Regression curves for glucose = 100 at each intake were:
Sucrose relative glucose response (RGR) = 95 ( SE8) + intake ×(–0·70 ( SE0·24)), P= 0·02; High-maltitol syrup RGR = 62 ( SE9) + intake ×
(–0·43 ( SE0·27)), P= 0·16; Maltitol RGR = 56 ( SE10) + intake ×(–0·42 ( SE0·23)), P= 0·10; Sorbitol RGR = 12 ( SE4) + intake ×(–0·09 ( SE
0·10)), P= 0·39; Isomalt RGR = 6 ( SE4) + intake ×(0·06 ( SE0·06)), P= 0·53; Lactitol RGR = 3 ( SE5) + intake ×(0·07 ( SE0·14)), P= 0·72.

parts. A single instance departed from the general case and
occurred in type 2 DM patients (Table 7; case 7). Here theinteraction is as expected for the GL, but not for the insuli-naemic load, and this could be due to a marked impairmentof insulin secretion in the patients studied.In conclusion, the GI and II and loads of the polyols
apply approximately in the context of simple meals of sug-ars (glucose, sucrose), starches (bread) and protein (scram-bled egg and farina cereal) and without overestimation(Tables 6 and 7). A similar conclusion was drawn forHealth potential of polyols 173
Table 5. Glycaemic and insulinaemic indices of polyols*
Glycaemic index Insulinaemic index
(glucose = 100) (glucose = 100)
Polyol Mean SD n† Mean SD n†
Erythritol 0 17 4 2 1 3
Xylitol 13 4 6 11 5 4Sorbitol 9 4 10 11 6 6Mannitol 0 – 1 0 – 1Isomalt 9 3 9 6 4 9Lactitol 6 2 3 4 3‡ 2Maltitol 35 9 9 27 5 6Maltitol syrups
High-maltitol syrup 48 11 7 35 10 4Intermediate-maltitol syrup 53 2 4 41 22 3Regular-maltitol syrup 52 10 5 44 17 5High-polymer maltitol syrup 36 11‡ 2 31 8‡ 2
Polyglycitol 39 7 3 23 7‡ 2
* Data are the means of study values for relative glucose responses and relative insulin responses in Table 4, ignoring intake as a cause of variance
when glucose is the reference carbohydrate. Observations obtained with > 50 g intake were excluded from the analysis. For the i nsulinaemic
index, one observation on xylitol and one observation on sorbitol were excluded as outliers from the analysis due to their bein g > 6 standardised
residuals from the results shown.
† No. of studies.‡ Plus and minus half range of the two values.
010203040506070
0 1 02 03 04 05 06 07 08 0
Glycaemic index (glucose = 100)Insulinaemic index
(glucose = 100)02468
05(b)
(a)
Fig. 4. Relationship (—–; Slope = 0·75 ( SE0·05)) of the insulinaemic index to the glycaemic index for polyols and sucrose for untransformed
data (a) and square root transformations (b). Data are from Table 5 and are means, with standard errors represented by vertical and horizontal
bars (among studies) for sucrose ( /H17034), regular-maltitol syrup ( /H17040), intermediate-maltitol syrup ( /H17039), high-maltitol syrup ( /H17009._), polyglycitol ( /H17005),
maltitol ( /H17004), sorbitol ( /H17034), xylitol ( /H17033), isomalt ( /H17003), lactitol ( /H17039), erythritol ( /H17040), and mannitol ( /H17004). (- – -), Unity.

individual foods in the context of foods and more complex
mixed meals (Collier et al. 1986; Wolever & Jenkins, 1986;
Bornet et al. 1987). In both circumstances the composite GI
(and GL) and II (and IL) were slightly less than predictedfrom the GI and II of individual components or foods.Importantly, the present results indicate that the potentialbenefits of low GI and II would not be diminished due tothe co-ingestion of very low-GI polyols with protein andavailable carbohydrate in a meal context.
Interaction between polyols and fat. Chocolate is a source
of both carbohydrate and fat. The glycaemic response tosucrose (GI 65 (
SD9)) is lower when in chocolate (GI 30
(SD9)) (values recalculated from Pelletier et al . 1994).Similarly, the glycaemic response to maltitol (GI 35 ( SD7))
may be lower in chocolate (GI 29 ( SD7)) (Pelletier et al .
1994). These responses may be attributed to slower stom-ach emptying, but also to a higher insulin response in thepresence of fat. Indeed, interactions between carbohydrateand fat are known to elevate insulinaemia and reduce gly-caemia (Collier et al. 1988; Morgan et al. 1988). Thus, the
II of sucrose (43 (
SD14)) is higher when in chocolate (76
(SD24)); likewise the II of maltitol (27 ( SD10)) v.maltitol
in chocolate (82 ( SD25)) (Pelletier et al. 1994). With other
polyols (isomalt, erythritol) no such interactions were evi-dent when comparing the present results for pure polyols(Table 5) with those from elsewhere for polyols eaten withfat, in chocolate (Gee et al . 1991; Bornet et al . 1996 b).174 G. Livesey
Table 6. Interaction between polyols and other dietary components affecting glycaemic index*
Glycaemic
Intake (g) Index Loadl† (g)
Case 1: normal subjects, n16, 8 M + 8 F (Kearsley et al. 1982)
Sorbitol 17 .51 0 1 .75
Glucose 17 .5 100 17 .5
Predicted sum for mixture 19.25
Observed for mixture 15.05
Observed/predicted value 0.78
Case 2: normal subjects, n16, 8 M + 8 F (Kearsley et al. 1982)
Sorbitol 15 .11 0 1 .51
Glucose 20 .0 100 19 .95
Predicted sum for mixture 21.46
Observed for mixture 17.50
Observed/predicted value 0.82
Case 3: normal subjects, n8 (Zaal & Ottenhof,1977)
Lactitol 50 6 3Sucrose 50 65 32 .5
Predicted sum for mixture 35.5
Observed for mixture 25.8
Observed/predicted value 0.73
Case 4: type 2 DM, n12 (Drost et al. 1985)
Bread (and butter) 36 70 25 .2
Sorbitol 22 10 2 .2
Predicted sum for mixture 27.4
Observed for mixture 16.7
Observed/predicted value 0.61
Case 5: type 1 DM‡, n9, 3 M + 6 F (Vaaler et al. 1987)
Bread (and butter) 75 70 52 .5
Sorbitol 21 10 2 .1
Predicted sum for mixture 54.6
Observed for mixture 49.8
Observed/predicted value 0.91
Cases 6 and 7: normal, n10; type 2 DM, n6 respectively (Akgün & Ertel, 1980)
Protein and carb meal + sucrose § 56 .13 7 .0
Protein and carb meal + fructose § 41 .42 2 .3
‘Protein and carb meal’ predicted § 33 .41 4 .3
Sorbitol 35 10 3 .53 .5
Predicted sum for mixture 36.91 7 .8
Observed for mixture 11.61 5 .4
Observed/predicted value 0.31 0 .87
M, male; F, female; DM, diabetes mellitus; carb, carbohydrate; GI, glycaemic index.
* GI values are from Table 5 or calculated references cited.† Glycaemic load = intake /H11003GI/100.
‡ Supported with continuous subcutaneous insulin infusion.§ Glycaemic loads calculated assuming the difference in glycaemic response between the sucrose and fructose meal was equal to t he difference in
glycaemic loads from sucrose (35 g /H11003GI 65/100) and fructose (35 g /H11003GI 23/100).

Importantly, the potential benefit of a low glycaemic
response per se to polyols is not lost when co-ingested with
fat. The data would suggest, however, that to achieve lowinsulin responses in products that can be made only withappreciable amounts of fats then carbohydrate of particu-larly low glycaemic response would be needed. In view ofa current understanding that high insulinogenic foods anddiets may be adverse for health reasons it may be just asimportant (or possibly more important) to reduce the GI ofcarbohydrate in high-fat foods as it is to lower the amountof fat in the foods.
Polyol-based snack foods
Healthy individuals and individuals with disorders of car-
bohydrate metabolism alike can desire the sweet taste offoods requiring bulk sweeteners; sugars and polyols(Mehnert, 1971). A number of snack foods (which may alsobe eaten at mealtimes), sugars and polyols are listed inTable 8, ranked by II. The carbohydrate content of a rea-soned portion, as used in these studies, is also shown and isabout 25 g, much of which might be replaceable with poly-
ols in manufactured goods.
Polyols rank very low on the II scale; however, this is
not the case for all polyol products, thus (as discussed ear-lier; pp. 174–175) maltitol-based chocolate has an II com-parable with sucrose-based chocolate, and much above theII for isomalt- and erythritol-based chocolate products. Thelatter two polyol products have II and GI values less thansome fruits (oranges, apples, banana, grapes) and yoghurt.Both fruits and polyol products have II values that are lessthan for many other products. Some polyols may thereforebe used to generate snack foods lower in II and GI than reg-ular snack foods.
The lowering of insulinaemia between meals is well
demonstrated for a polyol-based product by Bornet et al .
(1996 b). They fed sucrose- and erythritol-based chocolate
between breakfast and lunch to type 2 DM patients, show-ing considerable savings on the demand for insulin (Fig. 5).Such responses are not limited to snacks since they are alsoobserved after mixed meals as noted later (pp. 176–178).Thus also Hassinger et al . (1981) established that in dia-
betics requiring insulin, 30 g xylitol behaves as a low-Health potential of polyols 175
Table 7. Interaction between polyols and other components affecting insulinaemic index*
Insulinaemic
Intake (g) Index Load† (g equivalent)
Case 1: normal subjects, n16, 8 M + 8 F
Sorbitol 17 .51 1 1 .9
Glucose 17 .5 100 17 .5
Predicted sum for mixture 19.4
Observed for mixture 13.7
Observed/predicted value 0.71
Case 2: normal subjects, n16, 8 M + 8 F
Sorbitol 15 .11 1 1 .7
Glucose 20 .0 100 20 .0
Predicted sum for mixture 21.6
Observed for mixture 13.1
Observed/predicted value 0.61
Case 3: no insulin dataCase 4: type 2 DM, n12
Bread (and butter) 36 90 32 .4
Sorbitol 22 11 2 .4
Predicted sum for mixture 34.8
Observed for mixture 29.7
Observed/predicted value 0.85
Case 5: no insulin dataCases 6 and 7: normal, n10; type 2 DM, n6 respectively
Protein and carb meal + sucrose ‡ 33 .46 4 .8
Protein and carb meal + fructose ‡ 23 .65 5 .0
‘Protein and carb meal’ predicted ‡ 18 .44 9 .8
Sorbitol 35 11 3 .93 .9
Predicted sum for mixture 22.25 3 .6
Observed for mixture 10.05 7 .5
Observed/predicted value 0.45 1 .07
M, male; F, female; DM, diabetes mellitus; carb, carbohydrate.
* Insulinaemic index values are from Table 5 or calculated from information in the references cited in Table 6. Cases 1–7 corre spond to the glycaemic data in
Table 6.
† Insulin load = intake /H11003index/100.
‡ Insulinaemic loads calculated assuming the difference in glycaemic response between the sucrose and fructose meal was equal t o the difference in insulinaemic
loads from sucrose (intake of 35g /H11003insulinaemic index of 43 divided by 100) and fructose (intake of 35g /H11003insulinaemic index of 15 divided by 100).

glycaemic carbohydrate in the context of a high-protein
mixed meal, reducing plasma glucose and insulin require-ments by 50 % compared with sucrose.
Glycaemic control in groups of normal, type 1 and type 2
diabetes mellitus subjects
Markers of glycaemic control include fasting plasma glu-
cose (FPG), glucose tolerance (GT) or 2 h post GT during a75 g oral GT test, HbA
1c(glycosylated or glycated) concen-
trations and appearance of urinary glucose, all of which fallwith improvement in glycaemic control (Alberti & Zimmet,1998; Bastyr et al . 2000; Wang et al . 2002). The glucose
response after a mixed meal (or meal GT) also provides ananalogous measure to GT or 2 h post GT during a 75 g oralGT test. It is a relevant measure in longitudinal nutritionalstudies, though HbA
1c(and fructosamine as another marker
of protein glycation) is probably the most relevant overallmarker of glycaemic control and is now commonly used forthis purpose. It is well established that both HbA
1cand
fructosamine concentrations are reduced in diabetics by theconsumption of low-glycaemic-carbohydrate diets (Jenkins
et al. 2002), possibly more so when taken at each meal of
the day.
When taken orally with meals at a readily tolerated dose,
polyols may help to improve long-term glycaemic controlin type 2 DM patients, as expected for low-glycaemic car-bohydrates. Thus polyols have provided an example of howa low-glycaemic carbohydrate can benefit type 2 DMpatients. A 12-week randomised controlled study of theimpact of 6 g isomalt per meal (24 g daily) was undertakenon twenty-four subjects (twelve control and twelve parallelreceiving isomalt). Measurements were made (Pometta et
al. 1985) of HbA
1c(glycosylated) and FPG. In addition, the
change in mealtime glycaemia was calculated by takingpre-treatment FPG as the baseline (change in this resultthen reflects the overall improvement due to the sum ofchronic changes in FPG, meal GT and GI due to carbohy-drate replacement).
The following data were subsequently ascertained by the
present author’s analysis. For the control group (no drugs,regular diet treatment alone) the underlying trend was for176 G. Livesey
Table 8. Snack meals, foods, sugars and polyols ranked by insulinaemic index (II)
Meal item Carbohydrate intake (g) II GI II–GI Reference
1 Cereal and milk 25 127 26 101 *
2 Chocolate confection 31 102 58 44 †3 Glucose – 100 100 04 White bread – 92 74 18 ‡5 Cheese, bread and milk 25 89 8 81 *6 Peanut butter, bread and milk 25 88 14 73 *7 Chocolate milk (drink) 25 81 24 57 *8 Ice cream 26 79 52 27 †9 Milk chocolate (bar) 25 79 25 54 *
10 Milk chocolate (bar) 26 78 23 56 *11 Maltitol chocolate 25 82 30 52 §12 Sucrose-based chocolate 25 76 30 46 §13 Banana 32 68 58 10 †14 Grapes 15 68 54 14 †15 Yoghurt 25 63 35 28 *16 Fried chipped potato 36 .55 1 3 8 1 3 *
17 Peanut butter cup 25 51 10 41 *18 Oranges 50 .65 0 3 0 2 0 †
19 Potato chips (crisps) 25 49 23 26 *20 Apples 18 49 38 11 †21 Popcorn 27 .44 5 4 5 0 †
22 Maltitol syrup (regular) – 44 52 –8 ||23 Sucrose – 43 65 –22 ||24 Maltitol syrup (high-polymer) – 31 36 –5 ||25 Maltitol – 27 35 –8 ||26 Polyglycitol – 23 39 –16 ||27 Peanuts 5 .41 7 9 8 †
28 Isomalt chocolate 31 16 13 3 §29 Fructose – 15 23 –8 ||30 Sorbitol – 11 9 2 ||31 Xylitol – 11 13 –2 ||32 Isomalt – 6 9 –3 ||33 Lactitol – 4 6 –2 ||34 Erythritol – 2 0 2 ||35 Erythritol chocolate 37 2 – – §
36 Mannitol – 0 0 0 ||
GI, glycaemic index; II, insulinaemic index.
* Computed from Shively et al. (1986).
† Computed from Holt et al. (1997).
‡ Computed from Jenkins et al. (1981), Wolever & Bolognesi (1996 a), and Lee & Wolever (1998).
§ Computed from Pelletier et al. (1994), Gee et al. (1991), and Bornet et al. (1996 a).
|| See Table 5.

glycaemic control to become progressively worse, though
only slightly at an average rate of rise of HbA1cof 0·022
(SE0·006) % of the basal value per week ( P= 0·035). This
compares favourably with a worsening of twice this rate atapproximately 0·05 % of the basal value per week calcu-lated for conventionally controlled type 2 DM patients inother studies (UK Prospective Diabetes Study Group, 1998;Wallace & Matthews, 2000). The isomalt treatment groupby contrast maintained or improved HbA
1cconcentrations.
The differences in the mean of treatment outcomes in thepresent study were expressed as a percentage of the averageof means (Fig. 6), an appropriate statistical method forresults comparisons (Altman, 1991). Mealtime glycaemiawas immediately lower due to treatment with isomalt, by12·5 (
SE2·7) %, a difference that tended to widen with time
to 20 % lower after 3 months (Fig. 6) due to the combinedimprovement in FPG and meal GT. Relative to the control,the FPG fell at a significant rate of 0·5 (
SE0·1) % per week,
while the corresponding fall for HbA1cwas at a significant
rate of 0·4 ( SE0·02) % per week. The relative falls in FPG
and HbA1cwere progressive with time and appeared not to
have reached completion. These data contribute to theweight of data (Jenkins et al. 2002) showing that low-gly-
caemic carbohydrate ingestion by type 2 DM patients canimprove blood glucose control.
Other long-term studies on the effects of polyols in nor-
mal individuals and diabetic patients have been undertaken.Many predate current concepts in glycaemic control and sorequire fresh interpretation. The objective of the studies ofearly design was usually to establish whether or not thepolyols had adverse influences on metabolism, such ascausing FPG, cholesterolaemia or triacylglycerolaemia toincrease. For example: no such adverse effects were foundin healthy individuals, a mixed group of mainly olderschoolchildren (aged > 13 years) with some adults, when
exchanging 50 g xylitol for sucrose for 2 years (Huttunen et
al. 1975). The reduced GL due to this exchange is esti-
mated to be 30 g daily, which is substantial. Reduced FPGwas not observed, which suggests that the difference in GLis not of great importance in children or possibly youngadults with a healthy metabolism. In another study,Abraham et al . (1981) investigated the exchange of 26 g
sucrose for 30 g maltitol syrup for 4 weeks in type 2 DMpatients. No adverse effects were observed and there wasalso no improvement in glycaemic control as indicated byHealth potential of polyols 177
010203040
–60 0 60 120 180 240 300 360
Time (min)Plasma insulin (mU/l)
Breakfast Snack Lunch
Fig. 5. Insulin demand between meals is reduced using a polyol-
based snack food (Bornet et al. 1996b). ( /H17033), Sucrose-based
chocolate snack between meals; ( /H17034), erythritol-based chocolate
snack between meals. (–), Erythritol treatment group afteradjustment upwards to account for differences in treatment-groupmean responses to the breakfast (1·48 ×area above the basal
insulin concentration after breakfast).–5–4–3–2–10048 1 2Time (weeks )Glycosylated haemoglobin
(% change)
–8–6–4–20048 1 2Fasting glucose
(% change)
–25–20–15–10–50048 1 2Postprandial glucose
(% change)
Fig. 6. Improvement in glycaemic control with 6 g isomalt per meal
in type 2 diabetes mellitus (DM) patients. Isomalt was fed to twenty-four subjects, twelve controls and twelve type 2 DM patients, at arate of 6 g per meal (24 g daily). Percentage change is 100 ×
treatment means difference/treatment means average. Meansdifferences discounted the minor difference between treatmentgroups immediately after randomisation. The regression lines were:% Glycosylated haemoglobin = –0·4 (
SE0·02) % per week, P=
0·0004;Fasting glucose = –0·5 (
SE0·1) % per week, P= 0·04;
Postprandial glucose = (–12·5 ( SE2·7) %)* + (–0·5 ( SE0·4) % per
week)†.* P= 0·04, † P= 0·3. Data for these calculations were from the
study of Pometta et al. (1985), which reported the data as figures;
tabulated means data were kindly supplied by Palatinit GmbH(Mannheim, Germany).

either FPG or fasting insulin, which can reflect the degree
of insulin sensitivity and/or ß-cell function (Matthews et al.
1985). This is not surprising given the difference in GLbetween the treatments, estimated at present at just 2 gdaily (due to sucrose, GI = 65 and 26 g intake daily v.
maltitol syrup, GI = 50 at 30 g intake daily). Other studieshave examined mainly type 1 DM children. The treatmentswere usually polyol v.‘no polyol’ and the outcomes were
usually no adverse effects on FPG and urinary glucose, forexample with sorbitol (Steinke et al. 1961). Assuming the
study followed the controlled plan, the supplementary sor-bitol treatment group would have had an extra GL of 5 gdaily; thus the study provided no information about therelationship of GI or GL to the degree of control of carbo-hydrate metabolism. Thannhauser & Meyer (1929) andMehnert et al . (1960) undertook similar studies (of early
design) in type 2 DM patients. Again no adverse effects ofsorbitol (40 g) were observed, but again the experimentaldesigns did not allow an assessment of the relationshipbetween GL and the control of glucose metabolism.Another study was undertaken with parenteral xylitol (30 gdaily for 1 week) because of expectations of reducedrequirements for insulin secretion. Such treatment with xyl-itol lowered the FPG in some individuals of a mixed popu-lation of type 1 and 2 DM patients (Yamagata et al. 1965,
1969); amongst the type 2 DM patients the present authornotes the xylitol to have consistently reduced urinary excre-tion of glucose, and this almost quantitatively in accor-dance with the degree of glucosuria observed beforetreatment.
Scope for replacement of sugars, maltodextrins and
glucose syrups
Even quite small differences in GL due to carbohydrate
exchange appear to be important. Thus in well-controlledtype 2 DM patients a residual deterioration in plasmaHbA
1coccurs at an average rate of 0·2 ( SEM 0·04) units
HbA1c % per year (0·1, 0·2 and 0·3 % per year in Pomettaet al. 1985; Orchard et al. 1990; Wallace & Matthews, 2000
respectively). Intervention studies with low-GI diets showthe reversal of deterioration during the period of study.Assuming linear responses, the minimum change in GLthrough change in carbohydrate quality needed to reversethe average deterioration is just 12 (
SEM 2) g/d. Estimates
for individual studies are 11, 8, 13, 19, 9, 14 and 12 g/d (forJenkins et al . 1988; Brand et al . 1991; Wolever et al .
1992 a,b(two treatments); Frost et al . 1994; Järvi et al .
1999; Giacco et al . 2000 respectively). Such reversal is
seen with the polyol isomalt consumed at 24 g daily (Fig.6). A similar conclusion arises from the examination of theupper quintiles of GI and advent of DM in men (Salmerónet al. 1997 a) and in women (Salmerón et al. 1997 b; Meyer
et al. 2000), and CHD in women (Liu et al. 2000 b). Thus a
change in GL due to carbohydrate quality (not quantity) of10 g glucose/d corresponds to a change in disease advent of6, 27, 10 and 33 % respectively, with a mean of 19 (
SEM7)
% (P< 0·05). Such a change in GL by exchanging carbohy-
drates could readily be achieved by replacing some sugars,maltodextrins and glucose syrups with tolerable amounts ofpolyols.The consumption of sucrose in one population of US
women ranged from the lowest quintile median of 26 g/d tothe highest of 57 g/d, with a similar range for glucose andfructose combined (Meyer et al . 2000). This corroborates
similar findings from elsewhere with men consuming moreby weight than women in accordance with higher energyintakes (Glinsmann et al. 1986; Henderson et al. 2003). In
terms of macronutrient exchange or replacement, it is morerelevant to consider intakes per meal (rather than per d)because it is the meal that initiates an impulse to whichmetabolism responds (Livesey, 2000 b). For an average
three meals per d these sucrose consumption data corre-spond to an average meal sucrose intake of 8 to 19 g/meal;comparative values for glucose are from 4 to 10 g/meal(Meyer et al. 2000). Such quantities as polyol are tolerable
and many individuals can tolerate more (Livesey, 2001;Marteau & Flourié, 2001). There is, therefore, a realisticpotential for sugar replacers to exchange with sugars mak-ing a useful contribution towards a smaller glycaemicresponse to diet as a whole among those who would choosethis approach.
The total replacement of dietary sugars nevertheless
would be neither realistic nor expected, and in practice thepotential benefit would probably be limited to reducing theupper range of sugar intakes. The range between the lowerand upper quintiles in the study of Meyer et al. (2000) was
just 10 g/meal for sucrose and 7 g/meal for glucose, a largepart of which could potentially be replaced by polyolswhenever desirable.
Food manufacturers will, however, consider foods not
meals as products of their manufacture; likewise consumersbuy food items, for which there is scope for sugar replace-ment to achieve reduced glycaemia. About 25 g per servingin foods is practical; however, usage of lower amountsacross a broader range of food products may be more satis-factory. Unfortunately, this possibility is hampered at pre-sent by history; regulatory provisions in Europe currentlylimit the scope of use of sugar replacers (categorised assweeteners and additives) but not other carbohydrates,which are considered as ingredients (Barlow, 2001;Howlett, 2001). This situation tends to limit the use ofsugar replacers to confections and baked goods, and to ele-vate their content in such foods. A regulation permitting thebroader use of polyols, as for low-digestible sugars of simi-lar tolerance, would deserve consideration.
Assignment of polyols and foods to glycaemic index bands
Foods have GI values that span a continuous broad normal
distribution, which can be divided into narrower bands (forexample, very low, low, intermediate, high; Table 9).Banding can make it easier in practice for users to selectappropriate diets, as noted by Black & Rayner, for theCoronary Prevention Group (1992), or appropriately low-glycaemic diets for diabetes control (Brand et al . 1991).
Brand-Miller et al. (1999) suggest that GI > 70 would indi-
cate a high-GI food while GI < 55 indicates a low-GI food,with intermediate GI being 55 to 70. These bands havebeen demonstrated in practice to be helpful in the selectionof a low-GI diet (Brand et al. 1991), which without setting
a precise value would just fall into the low-GI band. To178 G. Livesey

make a more stringent target for formulating low-GI foods,
Bär (2000) recently suggested GI < 40, which here is called‘very-low GI’ to avoid confusion with the low-GI band ofBrand et al. (1991), Brand-Miller et al. (1999) and which
coincidently occurs approximately at the mean less onestandard deviation for the normal distribution of food GIvalues (G. Livesey, unpublished results). Some foods havesuch low GI values that the carbohydrate assessedinevitably includes unavailable or so-called ‘non-gly-caemic’ carbohydrate (Jenkins et al . 1987; Food and
Agriculture Organization, 1998), including resistant starch(Björck et al. 2000) and some polyols. Certain polyols (sor-
bitol, xylitol) additionally cause low increments in plasmaglucose due to slow absorption and metabolism in the liverand, although glucogenic, they give only low glycaemicresponses.
The use of nutrient banding to communicate nutritional
value is still in its infancy (Black & Rayner, for theCoronary Prevention Group, 1992). Table 9 simply mapsthe polyols, fruits, sugars, and candies and snacks to thepresently used bands and suggests an additional very-lowband for GI based on currently available information.
Regular, intermediate- and higher-maltitol syrups fall
into the low-GI band while other polyols fall into the very-low-GI band (erythritol, xylitol, sorbitol, mannitol, isomalt,lactitol, maltitol, high-polymer maltitol syrup, polyglyci-tol). There is an absence from Table 9 of information onpolyols in goods other than confections, such as bakedgoods or jams. Reduced glycaemia and insulinaemia hasbeen demonstrated in such products (Bakr, 1997) but thereis inadequate information across the time course for the cal-culation of GI and II. It is possible to replace sucrose (andsome maltodextrins and glucose syrups) with polyols in
baked goods, preserves and candies, but it is not possible todo this with intense sweeteners which lack both volume orbulk mass.
Mixtures of polyols with sugars, fats, starch-based foods
and protein-based foods were shown in the present reviewto yield lower GI than predicted for the component GI val-ues. Until such time as a method is established to predictsuch lower GI values for the mixture, it is suggested thatfood products might, when polyol based, have GI valuesthat are estimated from the GI values of the ingredients;this in the same way as GI values of meals are calculatedfrom the GI values of the component foods.
Food energy values of polyols
Various articles concerned with blood glucose control and
dental health report energy values for polyols incorrectly as17 kJ (4 kcal)/g. This value was a supposition based on theapproximate heats of combustion of polyols and anassumption that each polyol was fully absorbed and used inmetabolism. Numerous investigations have now beenundertaken and the polyols have been found to have differ-ent values lower than their heats of combustion (Table 10).The basis of derivation of polyol food energy is that carbo-hydrate absorbed via the small intestine and not excreted inthe urine is fully available as energy, while carbohydrateentering the colon and fully fermented is only 50 % avail-able as energy. This basis has widespread support, and sovarious reviewing bodies have derived similar (though notidentical) energy values to those shown in Table 10 (seeLivesey et al. 2000). Values obtained by indirect calorimetry
Health potential of polyols 179
Table 9. Glycaemic index (GI) bands and assignment of polyols, fruits, sugars, and candies and snacks by GI shown*
Band Polyols GI Fruits GI Sugars GI Candies and snacks GI
High GI (GI >70–140) Dates (dried) 103 Maltose 105 Jelly beans 87
Watermelon 72 Glucose 100 Pretzels 83
Corn chips 72
Intermediate GI (GI >55–70) Pineapple 66 Sucrose 65 Regular candy 70
Banana 55 Honey 58 Fruit chews 70
Almond bar 68Power chocolate bar 58Chocolate confection 58
Low GI (GI >40–55) Maltitol syrups Lactose 46 Ice-cream 52
Intermediate 53 Grapes 54 Chocolate 49Regular 52 Oranges 50 Yoghurt 46High 48 Popcorn 45
Chocolate coated toffee
and cookie bar 44
Chocolate peanut
confection 41
Very low GI (GI 0–40) Polyglycitol 39 Plum 39 Fructose 23 Fried chipped potato 38
Maltitol syrup Apple 38 Maltitol chocolate 30
(high-polymer) 36 Cherries 22 Potato chips (crisps) 23
Maltitol 35 Peanuts 14Xylitol 13 Isomalt chocolate 14Isomalt 9 Erythritol chocolate 2Sorbitol 9Lactitol 6Erythritol 0
Mannitol 0
* For references, see Table 8 footnotes and Foster-Powell et al. (2002).

corroborate the formula approach (see also Livesey,
2002 b). Values accepted in the USA under the process of
‘self determination’ are in reasonable agreement, whileEuropean regulations (European Communities, 1990) pre-scribe a single value for all permitted polyols.
The USA, Canada and Australia considered whether a
single value for all polyols might be misleading to the pub-lic and allocate separate values to each polyol. In the con-text of the labelling of individual foods, and in the contextof individual food products meeting energy-reduced claimsin respect of low-energy food regulations (for example,Codex Alimentarius Commission, 1991), a single energyvalue is not easily sustainable.
For low-glycaemic foods or dental-remineralising can-
dies and chewing gums made with polyols, it follows thatsuch foods and dentifrices would also be lower in energythan the corresponding product made with sugars, mal-todextrins or starches. Thus a candy of 25 g portion sizeand made with a polyol of 8 kJ/g, which may be consumedbecause it is tooth friendly or low glycaemic or both, wouldhave a food energy content of 200 kJ compared with 425 kJfor similar candies based on sugars (> 50 % energy reduc-tion). For a snack food of 1000 kJ with the same 25 g ofcarbohydrate this ingredient exchange would be a 20 %reduction in energy.
Dental aspects of polyols
The role of polyols in reducing dental caries may be
regarded as a benefit to part of the digestive system and soan aspect of digestive health. Other such aspects are consid-ered further later (p. 182).
Polyols are a poor source of energy for micro-organisms
of the oral cavity. Sucrose, other sugars and high-GIstarches, by contrast, are readily fermented by oral micro-organisms. Such carbohydrates are acidogenic and causetooth decay (dental caries), whereas polyols effectively donot. For this reason polyols have been described as ‘tooth
friendly’ and are permitted ingredients in sugar-free prod-ucts (European Communities, 1994).
Five key factors are involved in dental caries: teeth, bac-
teria, sugar or starch, time and saliva. Bacteria in the mouthreside mainly in dental plaque. Many species reside therebut few continue to ferment once a critical low pH of 5·7 isreached. In the main, mutans streptococci ( Streptococcus
mutans and S. sobrinus ) and lactobacilli are involved in aci-
dogenesis (British Nutrition Foundation, 2000). Salivadelivers amylase that may facilitate acidogenesis fromstarch, but also provides buffer capacity to wash away solu-ble carbohydrate, acids and immunoglobulins that aggre-gate bacteria. Other agents in saliva are effective inprotecting the body from harmful pathogens: lysozymedigests certain bacteria, lactoferrin binds and deprives bac-teria of Fe, sialoperoxidase reacts with H
2O2and salivary
thiocyanate to form a potent antibacterial agent, andhypothiocyanite (British Nutrition Foundation, 2000).Saliva also provides Ca, which supports remineralisation ofdemineralised teeth. Increased salivary buffer capacity onmastication might contribute to reduced caries incidenceand the sweetness of polyols and sugar-free chewing fre-quency have each been implicated in salivation rate (Rugg-Gunn, 1989; Birkhed & Bär, 1991; Dodds et al . 1991;
Mäkinen et al. 1995, 1996) though direct evidence for this
is scant.
Although dental caries has a multifactorial aetiology
(Burt & Ismail, 1986) and has decreased in prevalence fromvalues 40 years ago (König, 1990), it is still a highly preva-lent disease. Current evidence indicates that it does notdevelop without either sugars and starches or bacteria in themouth (National Research Council, 1989; British NutritionFoundation, 2000), and cannot occur without an increase inacid production (Bibby, 1975; Burt & Ismail, 1986).Acidogenesis in human volunteers is measurable routinelyby interdental-plaque-pH telemetry (Mühlemann, 1971).180 G. Livesey
Table 10. Food energy values of polyols (reference: sucrose, maltodextrins, starch at 17 kJ (4 kcal)/g)
Formula based on
Potential energy current availability Indirect US ‘self determined’ European
(heat of combustion)* data† calorimetry‡ and LSRO§ regulations||
kJ/g kcal/g kJ/g kcal/g kJ/g kcal/g kJ/g kcal/g kJ/g kcal/g
Erythritol 17 .24 .11 0 .2n a n a 1 0 .2
Isomalt 17 4 .19 2 .18 2 8 2
Lactitol 17 4 .18 2 8 1 .98 1 .9
Maltitol 17 4 .11 1 2 .71 1 ¶ 2 .6¶ 9 2 .1
Maltitol syrups
Regular, intermediate, high 17 .14 .1 12 3 na na 10 2 .4
High-polymer 17 .14 .11 2 2 .81 1 ¶ 2 .6¶ 13 3
Polyglycitol 17 .14 .11 2 2 .8n a n a
Mannitol 16 .74 .06 1 .5 na na 7 1.6
Sorbitol 16 .74 .01 0 2 .5n a n a 1 1 2 .7
Xylitol 17 4 .1 12 3 na na 10 2 .4
LSRO, Life Sciences Research Office; na, information not available.
* Potentially available had the polyol been fully available. Heats of combustion are calculated (Livesey, 1992).† Formula value = heat of combustion /H11003(available carbohydrate + 0.5 x fermentable carbohydrate) using data from Table 3.
‡ For studies on indirect calorimetry, see van Es et al. (1986), Sinaud et al. (2002) and Livesey (2002 a).
§ US
‘self determined’ labelling values are given with support from LSRO (Life Sciences Research Office, 1994, 1999).
|| European Communities (1990).¶ Deduced from a high-polymer syrup and its polymer fraction based on Sinaud et al. (2002).




The technique is used for investigation of the compliance
of tooth-friendly products with the requirements of theauthority in Switzerland, where sugar-free products are themajor form of confectionery (Imfeld, 1983, 1993).
Lack of acidogenic potential in polyols is the major
mechanism minimising caries development in polyol-basedcandies and sweet goods (Table 11). It appears there are noreal concerns about adaptation, that is, selection of polyol-fermenting acidogenic organisms (Table 11). Adaptation isnot completely absent, but does not occur to any extent thatwould risk caries formation from acidogenesis (Toors,1992). Acid production in plaque after sugar ingestion fol-lows a characteristic curve, a rapid fall in pH followed by aslow rise, called a Stephan curve. A fall below the criticalpH of 5·7 puts teeth under carious attack. According to thisapproach, Imfeld (1993) in his review was able to classifythe polyols as either having ‘no cariogenic potential’ orhaving ‘virtually no cariogenic potential’ (see Table 11).
Caries prevention using polyols has been described as a
‘passive process’ as it is the absence of acidogenic sub-stance rather than the presence of an active or bacteriostaticsubstance that is important (Imfeld, 1993). However, xyli-tol may also be bacteriostatic on one and possibly morestrains of S. mutans (Waaler et al. 1992). The mechanism
proposed was the reversible inhibition of essential meta-bolic pathways including the accumulation of xylitol-5-phosphate, an inhibitor of phosphoenolypyruvateproduction. The clinical significance has been reported as areduction in virulence of S. mutans and modification of the
plaque ecosystem including reductions in plaque quantityand adhesivity (reduced ability to adhere to the hard tis-sues). The quantitative contribution this makes to cariesreduction is reported as unclear by some authors (Isokangaset al . 1991; Scheie et al . 1998; Alanen, 2001).
Nevertheless, xylitol is commonly associated with reducednumbers of S. mutans (Hayes, 2001; Mäkinen et al. 2001),
appears more effective than erythritol in reducing the massof plaque in human subjects (Mäkinen et al. 2001), and is
more effective than sorbitol in caries prophylaxis (Mäkinenet al . 1996). A difficulty with the interpretation of these
comparisons is a lack of quantitative information on theseparate roles of saliva stimulation and microbiological fac-tors (Alanen, 2001). Interestingly, the reduced transmissionof S. mutans from mother to offspring may explain a lower
caries incidence in 2- to 5-year-old children after maternalxylitol consumption when the children were aged 3–24months (Isokangas et al. 2000).
Less well known than the virtually non-acidogenic
potential of polyols as sugar replacers is their limitation ofplaque formation. Plaque is a conglomerate of bacteria andpolysaccharides where acidogenesis takes place. The poly-saccharides synthesised by oral bacteria bulk out theplaque, which in turn harbours these organisms and retainsfermentation products, so depressing the pH further andreducing the ability of saliva to wash the organisms andacid away (Newbrun, 1982; Rolla et al. 1985). By contrast
polyols are not substrates for polysaccharide and plaquesynthesis. Isomalt, while not supplying substrate for poly-saccharide synthesis (Bramstedt et al . 1976; Ciardi et al .
1983), might also inhibit this process from sucrose, as evi-dent for some of the longer-chain hydrogenated isomalto-oligosaccharides (Tsunehiro et al . 1997). Polysaccharide
synthesis is also lower with xylitol, lactitol, mannitol andsorbitol than with sucrose (Grenby et al. 1989).
Polyols also reverse the initial stages of dental caries by
promoting remineralisation. This is preferable to toothrestoration except on advanced lesions (Featherstone,2000). Stimulation of salivary flow facilitates remineralisa-tion when induced between meals by confections contain-ing a polyol; this is evident because the repair of earlylesions is greater when such products are ingested thanwhen no food is consumed (Leach, 1987). A recent andimportant observation is that polyols both slow deminerali-sation of tooth enamel and accelerate remineralisation of
Health potential of polyols 181
Table 11. Cariogenic potential, bacteriostasis, inhibition of polysaccharides synthesis, remineralisation and adaptation
Minor (active) mechanisms
Major (passive) mechanism: Inhibition of
cariogenic potential* polysaccharide Promotion Concerns: significant
(based on acidogenesis) Bacteriostasis synthesis of remineralisation adaptation
Erythritol None to virtually none* – – – –
Xylitol None Yes† – Yes§ None||Sorbitol Virtually none – – – None¶Mannitol Virtually none – – – –Maltitol Virtually none – – – –Isomalt None – Suggested‡ Yes§ None**Lactitol None – – – None††
Regular maltitol syrup Virtually none – – – None‡‡
* After Imfeld (1993), except eythritol for which a preliminary classification is given here. This reflects the practical inabili ty of oral bacteria to use these carbo-
hydrates for acid production (or for plaque polysaccharide synthesis).
† Waaler et al. (1992). The quantitative contribution of this bacteriostatic mechanism to clinical outcome is unknown, though may explain adv antages of xylitol over
sorbitol and erythritol (see p. 181).
‡ Ciardi et al. (1983), Bramstedt et al. (1976). Quantitative contribution to clinical outcome is unknown.
§ Takatsuka (2000), Mäkinen et al. (1995).
|| Toors (1992), Gehring et al. (1975).
¶ Toors (1992), Cornick & Bowen (1972).** Van der Hoeven (1979, 1980).†† Havenaar et al. (1978).
‡‡ Rugg-Gunn (1989).

demineralised lesions. Xylitol and particularly isomalt may
be effective in this regard (Takatsuka, 2000).
On the basis of substantial studies in human subjects
regarding caries the prophylactic properties of lactitol, iso-malt and xylitol have been recommended (Imfeld, 1993;Featherstone, 1995; Mäkinen et al. 1996). Clinical trials on
sorbitol (Birkhed & Bär, 1991), maltitol syrup (Rugg-Gunn, 1989) and xylitol (Mäkinen et al. 1996) indicate that
they are non-cariogenic. Erythritol has been advanced as apotential new caries preventative (Kawanabe et al . 1992;
Mäkinen et al. 2001).
Colonic health aspects
The colonic environment
Due to their ease of fermentation by gut flora, low-
digestible carbohydrates are very important in humanhealth. Such carbohydrates contribute fundamentally to theestablishment of an anaerobic and acidic environment inthe colon. Their fermentation enables saccharolytic anaer-obes and aciduric organisms to grow in preference overputrefying, endotoxic, pathogenic, and procarcinogen-acti-vating aerobic organisms (Hawksworth et al. 1971; Brown
et al . 1974; Gracey, 1982; Hill, 1985; Hill et al . 1987;
Rowland, 1991; Mitsouka, 1992; Screvola et al . 1993 a,b;
Mital & Garg, 1995).
Low-molecular-weight carbohydrate (lactulose) and
polyol (lactitol) have long been acknowledged for theirability to reduce circulating levels of NH
3and toxic micro-
bial substances, the clinical utility of which is the treatmentof hepatic encephalopathy (Blanc et al. 1992).
The acidic conditions associate with or normalise epithe-
lial functions resulting in fewer pathologies and their mark-ers, such as aberrant crypts (Samelson et al . 1985), large
adenomas (Roncucci et al . 1993; Ponz de Leon &
Roncucci, 1997; Biasco & Paganelli, 1999) and possiblytumours (Thornton, 1981). Lactic acid is of particular note;it is generated from all fermentable carbohydrates but espe-cially those that readily undergo microbial glycolysisincluding polyols. A slow removal of lactic acid from thecolon would help to maintain acidity and the growth ofaciduric organisms such as the lactic acid bacteria, whichare now widely promoted as probiotics. Butyric acid, whichcan be generated from polyols, sometimes in large amounts(Mortensen et al. 1988; Clausen et al. 1998), and possibly
due to secondary fermentation of lactic acid, is widelyrecognised for its probable role in maintaining a healthycolonic epithelium. It is also recognised for its improve-ment of inflammatory conditions of the colonic mucosa(Roediger, 1990; Scheppach et al. 1995) and anti-neoplasic
activity (Velazquez et al. 1996; Scheppach et al. 2001; for a
review, see Brouns et al. 2002). Although faecal butyrate is
not especially prominent amongst black South Africanswho are renowned for their healthy colons, raised concen-trations of short-chain organic acids (Segal et al. 1995) and
acidity (Levy et al . 1994) are found in these individuals.
This has been attributed to increased fermentation and ahigher than usual entry into the colon (than in Westerners)of osmotic carbohydrate (Veitch et al. 1998; Segal, 2002).
These responses can generally be attributed to saccha-rolytic fermentation, ease of fermentation, and water entry
into the colon with osmotic carbohydrates; thus responseshave been reported for a wide range of low-digestible andfermentable carbohydrates including polyols in human sub-jects. For example, responses have been reported for lactu-lose (MacGillivary et al. 1959), lactitol (Felix et al. 1990;
Screvola et al. 1993 b; Ravelli et al. 1995; Tarao et al. 1995;
Ballongue et al. 1997), isomalto-oligosaccharides (Kaneko
et al . 1994), lactosucrose (Teramoto et al . 1996), and
fructo-oligosaccharides (Gibson & Roberfroid, 1995;Tuohy et al . 2001). Gibson & Roberfroid (1995) have
reported responses for inulin, Zhong et al. (2000) for poly-
dextrose and Bird et al. (2000) for some resistant starches.
For individuals with disaccharidase deficiencies, similarreports appear for lactose (Segal, 1998, 2002) and sucrose(Veitch et al . 1998; Segal, 2002), and incomplete absorp-
tion of fructose (Segal, 1998). Short-chain organic acidsmay also modify gastrointestinal motility and so could havea role in maintaining a regular bowel habit (Cherbut et al.
1998; Piche et al. 2000).
Constipation and laxation
Constipation may be defined most simply as ‘less than three
bowel movements per week’ and is the most common gas-trointestinal complaint in Western cultures, triggering con-siderable use of over-the-counter laxatives and consultationswith medical practitioners (Royal College of GeneralPractitioners, 1986; Sandler et al. 1990; Sweeney, 1997). It
is particularly common in the elderly (Koch & Hudson,2000), diabetics (Haines, 1995), children (Guimaraes et al.
2001) particularly those with developmental and neurologi-cal disability (Staiano et al . 2000; Tse et al . 2000), preg-
nancy (Signorelli et al . 1996), and in those with reduced
food intake (anorexia, weight reduction, hospitalisation). Itis also common in numerous other less prevalent circum-stances (Baker et al. 1999; Nurko et al. 2001). Some drugs
are causative, including the commonly used Al antacids anddietary Fe supplements (Baker et al. 1999).
Laxation is the ‘gentle stimulation of the bowel to render
the motion slightly soft without causing any gripes’(Macpherson, 1990). Laxative action has been establishedfor acceptable intakes of xylitol, sorbitol, mannitol, isomalt,lactitol, maltitol and erythritol (Brin & Miller, 1974;Sheinin et al . 1974; Ornskov et al . 1988; Livesey, 2001;
Marteau & Flourié, 2001). All act to promote hydration ofthe colonic contents. Usefully, polyols are obtainable by thepublic in tasty food items such as sugar-free, reduced-energy candies and other products. Studies have demon-strated the efficacy of polyols (crystalline or syrupformulations) in the elderly (Lederle et al. 1990) and in a
multicentred study of the elderly both hospitalised and out-patients (Delas et al . 1991; Sacchetta et al . 2000), and in
children (Ornskov et al . 1988; Pitzalis et al . 1996). Data
from Spengler et al . (1987) indicate approximately 30 %
less constipation even amongst young adults with ‘healthycolonic function’ consuming up to 48 g isomalt daily (thirtysubjects, 84 d each), and without excess laxation. Also, lac-titol and lactulose each show a reduced likelihood of slowtransit occurring in physically inactive hospitalised individ-uals with healthy gastrointestinal tracts (Pontes et al. 1995).182 G. Livesey

Adequate drinking water has been recommended along
with dietary fibre to enhance laxation (Gray, 1995; Anti et
al. 1998). However, even this dual action may be inade-
quate (Benton et al. 1997). Rural South African and Asian
diets are thought ideal for optimal stool formation.However, osmotic carbohydrate in these diets may be justas important as dietary fibre due to sucrase and lactase‘deficiency’ in these populations (Veitch et al. 1998; Segal,
2002) promoting an adequate hydration of colonic contents.Westerners without lactase and sucrase deficiency could,logically, achieve the same goal with appropriate intakes ofpolyols.
A consensus of food and nutritional scientists and physi-
cians has been established for polyol consumption: ‘Eachindividual may experiment with intake amounts and makeadjustments based on their own experience – as they maydo routinely with everyday foods having the same effectswhen eaten to excess’ (Salford Symposium Consensus,2001). This was recommended because individuals vary inthe magnitude of their response to polyol ingestion, asindeed they do in the degree to which constipation is expe-rienced. Physicians have also recommended that individu-als ‘adjust the dose of polyol to a daily bowel movementfor 1 to 2 months’ (Baker et al. 1999; Nurko et al. 2001).
Tolerance
Low- and very-low glycaemic-carbohydrate foods can be a
cause of unwanted gastrointestinal responses in sensitiveindividuals or when ingested to excess due to their reach-ing the colon. Increased gastrointestinal awareness is com-monly experienced with high-fibre foods, some of whichare low-glycaemic-carbohydrate foods such as beans,lentils and legumes. Other foods include cabbage, Brusselssprouts, brown bread, oatmeal porridge, rough-seededfruits, honey, tamarinds, figs, prunes, raspberries, straw-berries, stewed apples, aloes, rhubarb, cascara and senna(Macpherson, 1990; Friedman, 1991) and modest levels offibre supplements (Stevens et al. 1987). Similar responses
can occur without a change in food source by lowering ofthe GI using pharmacological means; the sucrase inhibitoracarbose results in elevated flatulence in up to 43 % ofconsumers and osmotic diarrhoea or laxation in up to 27 %(Sels et al. 1998). All such foods and carbohydrates can be
a cause of increased colonic fermentation, flatulence,bloating and cramp. Feelings of bloating (as opposed tomeasurements of abdominal distension) are probably morecommon after overingestion of food in general, which isall too common. Furthermore, cramp appears to be sec-ondary to faecal impaction in those with a poor bowelhabit (McRorie et al. 2000), or in individuals with irritable
bowel syndrome (Briet et al . 1995). In contrast to infec-
tious diarrhoea, watery stools due to colonic fermentationof low-digestible carbohydrates are not a medical issue,and intakes of polyols comparable or greater than normalfor dietary fibre are possible (Steinke et al. 1961; Sheinin
et al . 1974; Spengler et al . 1987; Sinaud et al . 2002; A
Lee, DN Storey, F Bornet and F Brouns, unpublishedresults).
Rapid transition from a diet that encourages constipation
(diets low in polyols, dietary fibre and some slimmingdiets) to ones that promote laxation (high polyol, dietary
fibre and high food intakes) may be a transient cause of dis-comfort (see McRorie et al. 2000). This may be avoided by
varying the daily intake of polyol-based foods graduallyover a period of 1 to 4 weeks (see Steinke et al . 1961;
Baker et al . 1999; Salford Symposium Consensus, 2001;
Nurko et al. 2001). Adaptation to polyols usually improves
gastrointestinal tolerance (Tucker et al. 1981; Pometta et al.
1985; Briet et al. 1997) and may in part be psychological
(Tucker et al . 1981) and occur with an increasing experi-
ence of fermentable carbohydrate consumption (Briet et al.
1997). Tolerance and intakes are greatest when polyols areconsumed at regular intervals throughout the day (Livesey,2001; Sinaud et al. 2002) as may be desirable in some dia-
betics (Warshaw & Powers, 1999). Consuming polyols inor with other foods will also improve tolerance by delayingstomach emptying (Livesey, 1990 a, 2001; Marteau &
Flourié, 2001). In this respect the co-ingestion of a high-cereal-fibre diet may be useful as it provides a matrix withwhich water combines to be retained in the large bowel.Some individuals are sensitive to polyols and should reduceor even avoid such foods altogether (Salford SymposiumConsensus, 2001). Children more than younger or olderadults are likely to consume larger amounts of freely avail-able polyols; there is, however, no evidence that childrenare less able to tolerate polyols than are adults in terms ofthe weight of polyol per meal or d (Spengler et al . 1987;
Paige et al. 1992; A Lee, Salford University, personal com-
munication).
The scientific interpretation of consumer responses to
polyols is difficult. Consumers generally indicate that theyhave diarrhoea whenever they notice a softening of theirstool independently of whether it is inconveniencing andsome 98 % of such occurrences do not meet commonlyaccepted criteria for clinical diarrhoea (McRorie et al .
2000). In agreement, a market survey of 1000 consumers ofsugar-free products (polyols) has indicated that as little as0·5 % of individuals make unprompted claims to the expe-rience of adverse gastrointestinal responses (Stewart,2001). This coincides with the rate observed in the absenceof polyol consumption (Steinke et al. 1961; Spengler et al.
1987). Reported responses to polyols are often based onquestionnaires that prompt volunteers to notice symptomsof intolerance, and so may be biased; thus when promptedsuch claims may increase five-fold (Stewart, 2001). Alsothe interpretation of scientific studies in a laboratory settingcan be difficult due to substantial inter-individual variationin gastrointestinal responses to polyols (Livesey, 2001),adaptation (Marteau & Flourié, 2001) and other reasons(Barlow, 2001).
There are probably more non-diabetics who consume
polyols than diabetics, though the latter have often been thesubject of study. The American Diabetes Association(2001) has suggested, bearing in mind the varied responsesamong individuals, that the choice to consume particulartypes of carbohydrate including polyols must be an individ-ual one, taking account of global dietary guidance and indi-vidual metabolic needs. Constipation can be common indiabetics and older individuals (Wegener et al . 1990) and
older diabetics may indicate that polyol consumptionimproves bowel habit (Pometta et al . 1985). Idiopathic
Health potential of polyols 183

diarrhoea also occurs in diabetics, but is not due to the
increased use of polyols, and polyols are not contraindi-cated in diabetics when consumed in moderate amounts(Verina et al . 1995). Individuals with type 2 DM tolerate
polyols equally as well as normal individuals (Zumbé &Brinkworth, 1992; Verina et al. 1995).
Conclusion
Polyols are found to provide acknowledged examples of clini-
cal benefits in the treatment and regulation of bowel habit,and in the conditioning of the colonic environment.Intriguingly, appropriate consumption of low-digestibleosmotic carbohydrates may be critically important inWesterners to achieve stool consistencies comparable withthose of rural South Africans. These benefits add to theacknowledged properties of polyols as reduced-energy carbo-hydrates and to the benefits of tooth friendliness, where poly-ols may have a role in the repair as well as the prevention ofcaries. The low- to very-low-glycaemic and insulinaemicproperties of polyols offer further potential health benefits onreplacement of bulk in sugars, syrups and maltodextrins infoods for individuals with both normal and abnormal carbo-hydrate metabolism. Scope exists for such benefit within gas-trointestinal tolerances, which can be improved by attentionto the dose, timing, and diet during polyol consumption.
Information on the glycaemic and insulinaemic
responses to polyol-based foods is scarce compared withinformation on polyols alone. Nevertheless, it is evidentthat interactions between polyols and macronutrients tendto reduce postprandial glycaemia, and interactions betweensugars and fats that elevate postprandial insulinaemia canbe attenuated or almost abolished using polyols. There is noreason to suppose that long-term use of polyols elevatesprotein glycation, a marker of glycaemic control, as dohigh-glycaemic carbohydrates, and there is evidence thatthe consumption of a polyol might reduce protein glyca-tion, adding to similar observations for other low-gly-caemic-carbohydrate diets.
On a technical note, as with carbohydrate foods tabulated
in the international tables of GI (Foster-Powell et al. 2002),
where data are available on polyols it is found acceptable topool information on GI values from normal, type 1 and2 DM patients to obtain a single value for each polyolapplicable in all these conditions. Similarly, there is nomore dose-dependency of GI values for polyols than forother carbohydrates.
Acknowledgements
This analysis and review was commissioned by the European
Polyol Association, to whom the author is grateful for sup-port. Thanks are due to Keir J. Livesey, Independent NutritionLogic, for support and discussion of statistical issues.
References
Abraham RR, Davis M, Yudkin J & Williams R (1981) Controlled
clinical trial of a new non-calorigenic sweetening agent.Journal of Human Nutrition 35, 165–172.Adcock LH & Grey CH (1957) The metabolism of sorbitol in the
human subject. Biochemical Journal 65, 554–560.
Akgün S & Ertel NH (1980) A comparison of carbohydrate
metabolism after sucrose, sorbitol and fructose meals in normaland diabetic subjects. Diabetes Care 3, 582–585.
Alanen P (2001) Does chewing explain the caries-preventative
results with xylitol. Journal of Dental Research 80, 1600–1601.
Alberti KG & Zimmet PZ (1998) Definition, diagnosis and classi-
fication of diabetes mellitus and its complications. Part 1:Diagnosis and classification of diabetes mellitus provisionalreport of a WHO consultation. Diabetes Medicine 15, 539–553.
Altman DG (1991) Practical Statistics for Medical Research .
London: Chapman and Hall.
American Association of Cereal Chemists (2001) The Definition
of Dietary Fiber . Report of the Dietary Fiber Definition
Committee. St Paul, MN: American Association of CerealChemists.
American Diabetes Association (2001) Postprandial blood glucose
(a consensus statement). Diabetes Care 24, 775–778.
American Diabetes Association (2002) Evidence-based nutrition
principles and recommendations for the treatment and preven-tion of diabetes and related complications. Diabetes Care 25,
S50–S60.
Anti M, Pignataro G, Armuzzi A, Valenti A, Iascone E, Marmo R,
Lamazza A, Pretaroli AR, Pace V , Leo P, Castelli A &Gasbarrini G (1998) Water supplementation enhances the effectof high-fibre diet on stool frequency and laxative consumptionin adult patients with functional constipation.Hepatogastroenterology 45, 727–732.
Australia New Zealand Food Authority (2001) Inquiry Report:
Derivation of Energy Factors . Canberra, Australia: ANZFA.
Bachmann W, Haslbeck M, Spengler M, Schmitz H & Mehnert H
(1984) Untersuchungen zur Stoffwechselbeeinflussung durchakute Palatinitgaben. Vergleich zu Fructose und Saccharose beiTyp-II-Diabetes (Investigations of the metabolic effects of acutedoses of Palatinit. Comparison with fructose and sucrose inType II diabetes). Aktulle Ernährungsmedizin 9, 65–70.
Baker SS, Liptak GS, Colletti RB, Croffie JM, Di Lorenzo C,
Ector W & Nurko S (1999) Constipation in infants and chil-dren: evaluation and treatment. A medical position statement ofthe North American Society for Pediatric Gastroenterology andNutrition. Journal of Pediatric Gastroenterology and Nutrition
29, 612–626.
Bakr AA (1997) Application potential for some sugar substitutes
in some low energy and diabetic foods. Nahrung-Food 41,
s170–s175.
Ballongue J, Schumann C & Quignon P (1997) Effect of lactulose
and lactitol on colonic microflora and enzymatic activity.Scandinavian Journal of Gastroenterology 32, Suppl. 222,
41–44.
Bär A (1990) Factorial calculation model for the estimation of the
physiological caloric value of polyols. In Caloric Evaluation of
Carbohydrates , pp. 209–257 [N Hosoya, editor]. Tokyo:
Research Foundation for Sugar Metabolism.
Bär A (2000) Foods Intended for Use in a Carbohydrate
Controlled Diet . Position paper for German Diätvertband.
Basel, Switzerland; Bioresco.
Barlow S (2001) Workshop: regulatory affairs. British Journal of
Nutrition 85, Suppl. 1, S63–S64.
Bastyr EJ, Stuart CA, Brodows RG, Schwartz S, Graf CJ, Zagar A
& Robertson KE (2000) Therapy focussed on lowering post-prandial glucose, not fasting glucose, may be superior for low-ering HBA
1c. Diabetes Care 23, 1236–1241.
Beaugerie L, Fourié B, Marteau P, Pellier P, Franchisseur C &
Rambaud J-C (1990) Digestion and absorption in the humanintestine of three sugar alcohols. Gastroenterology 99,
717–723.184 G. Livesey

Bellisle F (2001) Glycaemic Index and Health: the Quality of the
Evidence . Montrouge, France: John Libbey, Eurotext.
Benton JM, O’Hara PA, Chen H, Harper DW & Johnston SF
(1997) Changing bowel hygiene practice successfully: a pro-gram to reduce laxative use in a chronic care hospital. Geriatric
Nurse 18, 12–17.
Bernier JJ & Pascal G (1990) The energy value of polyols (sugar
alcohols). Medicine et Nutrition 26, 221–238.
Bernt WO, Borzelleca JF, Lamm G & Munro IC (1996)
Erythritol: a review of biological and toxicological studies.Regulatory Toxicology and Pharmacology 24, s191–s197.
Biasco G & Paganelli GM (1999) European trials on dietary sup-
plementation for cancer prevention. Annals of the New York
Academy of Sciences 889, 152–156.
Bibby B (1975) The cariogenicity of snack foods and confections.
Journal of the American Dental Association 90, 121–132.
Bird AR, Brown IL & Topping DL (2000) Starches, resistant
starches, the gut microflora and human health. Current Issues in
Intestinal Microbiology 1, 25–37.
Birkhed D & Bär A (1991) Sorbitol and dental caries. World
Review of Nutrition and Dietetics 65, 1–37.
Björck I, Liljeberg H & Östman E (2000) Low-glycaemic index
foods. British Journal of Nutrition 83, Suppl. 1, S149–S155.
Black A & Rayner M, for the Coronary Prevention Group (1992)
Just read the label: understanding nutrition information innumeric, verbal and graphic format. London: H.M. StationeryOffice.
Blanc P, Daures JP, Rouillon JM, Peray, P, Pierrugues R, Larrey
D, Gremy F & Michel H (1992) Lactitol or lactulose in thetreatment of chronic hepatic encephalopathy: results of a meta-analysis. Hepatology (Baltimore) 15, 222–228.
Bornet FRJ, Blayo A, Dauchy F & Slama G (1996 a) Plasma and
urine kinetics of erythritol after oral ingestion by healthy humans.Regulatory Toxicology and Pharmacology 24, s220–s285.
Bornet FRJ, Blayo A, Dauchy F & Slama G (1996 b)
Gastrointestinal response and plasma and urine determinationsin human subjects given erythritol. Regulatory Toxicology and
Pharmacology 24, s296–s302.
Bornet FRJ, Costagliola D, Rizkalla SW, Blayo A, Fontvieille
AM, Haardt MJ, Letanoux M, Tchobroutsky G & Slama G(1987) Insulinemic and glycemic indices of six starch-richfoods taken alone and in a mixed meal by type 2 diabetics.American Journal of Clinical Nutrition 45, 588–595.
Bramstedt F, Gehring F & Karle EJ (1976) Comparative Study of
the Cariogenic Effects of Palatinit, Xylitol and Saccharose inAnimals . Würzburg, Germany: University of Würzburg.
Brand J, Colagiuri S, Crossman S, Allen A, Roberts D & Truswell
S (1991) Low-glycemic index foods improve long-term gly-caemic control in NIDDM. Diabetes Care 14, 95–101.
Brand-Miller J, Wolever TMS, Colagiuri S & Foster-Powell K
(1999) The Glucose Revolution . New York: Marlow &
Company.
Briet F, Achour L, Fourié B, Beaugerie L, Pellier P, Franchisseur
C, Bornet F & Rambaud JC (1995) Symptomatic response tovarying levels of fructooligosaccharides consumed occasionallyor regularly. European Journal of Clinical Nutrition 49,
501–507.
Briet F, Pochart P , Marteau P, Flourie B, Arrigoni E & Rambaud
JC(1997) Improved clinical tolerance to chronic lactose inges-
tion in subjects with lactose intolerance: a placebo effect? Gut
41, 632–635.
Brin M & Miller OM (1974) The safety of oral xylitol. In Sugars
in Nutrition , pp. 591–606 [HL Sipple and KW McNutt, edi-
tors]. New York: Academic Press.
British Nutrition Foundation (2000) Oral Health, Diet and Other
Factors . The Report of the British Nutrition Foundation’s Task
Force. Amsterdam: Elsevier.Brooks SPJ (1995) Report on the Energy Value of Sugar Alcohols .
Ottawa, Canada: Ministry of Health.
Brouns F, Kettlitz B & Arrigoni E (2002) Resistant starch and the
‘butyrate revolution’. Trends in Food Science and Technology
13, 251–261.
Brown R, Gibson JA, Sladen GE, Hicks B & Dawson AM (1974)
Effects of lactulose and other laxatives on ileal and colonic pHas measured by radiotelemetry device. Gut15, 999–1004.
Burt A & Ismail AI (1986) Diet, nutrition, and food cariogenicity.
Journal of Dental Research 65, 1475–1484.
Buyken AE, Toeller M, Heitkamp G, Irsiglert C, Hollert F,
Santeusanio F, Stehle P, Fuller JH and the EURODIAB IDDMComplication Study Group (2000) Carbohydrate sources andglycaemic control in type 1 diabetes mellitus. Diabetic
Medicine 17, 351–359.
Buyken AE, Toeller M, Heitkamp G, Karamanos B, Rottiers R,
Muggeo M, Fuller JH and the EURODIAB IDDMComplications Study Group (2001) Glycaemic index of the dietof European outpatients with type-1 diabetes: relations to gly-cated hemoglobin and serum lipids. American Journal of
Clinical Nutrition 73, 574–581.
Canadian Diabetes Association (2000) Guidelines for the manage-
ment of diabetes mellitus in the new millennium. Canadian
Journal of Diabetes Care 23, 56–69.
Cherbut C, Ferrier L, Roze C, Anini Y , Blottiere H, Lecannu G &
Galmiche JP (1998) Short-chain fatty acids modify colonicmotility through nerves and polypeptide YY release in the rat.American Journal of Physiology 275, G1415–G1422.
Ciardi J, Bowen WH, Rolla G & Nagorski K (1983) Effects of
sugar substitutes on bacterial growth, acid production and glu-can synthesis. Journal of Dental Research 62, 182.
Clausen MR, Jørgensen J & Mortensen PB (1998) Comparison of
diarrhea induced by ingestion of the fructo-oligosaccharideidolax and the disaccharide lactulose. Digestive Diseases and
Sciences 43, 2696–2707.
Codex Alimentarius Commission (1991) Codex Standard for
Formula Foods for Use in Weight Control Diets. CodexStandard 181. Rome: Food and Agricultural Organization.
Collier GR, Greenberg GR, Wolever TMS & Jenkins DJA (1988)
The acute effect of fat on insulin secretion. Journal of Clinical
Endocrinology and Metabolism 66, 323–326.
Collier GR, Wolever TMS, Wong GS & Josse RG (1986)
Prediction of glycaemic responses to mixed meals in non-insulin dependent diabetic subjects. American Journal of
Clinical Nutrition 44, 349–352.
Cornick DER & Bowen WH (1972) The effect of sorbitol on the
dental plaque in monkeys (Macacac Irus). Archives of Oral
Biology 17, 1637–1648.
Delas N, Gislon J, Glikmanas M, Henri-Biabaud E, Lemerez M,
Licht H, Slama JL & Gillaume PN (1991) Lactitol in the treat-ment of constipation in the adult. Open, non-comparative studyof its efficacy and its clinical and biological tolerance. Annals
of Gastroenterology and Hepatology (Paris) 27, 231–233.
Diabetes UK (2000) Diet in diabetes care.
www.diabetes.org.uk/summer 00/diet.htm
Diabetes UK (2002) Nutritional guidelines in diabetes care.
www.diabetes.org.uk/infocentre/carerec/nutrition.htm
Dills WL (1989) Sugar alcohols as bulk sweeteners. Annual
Review of Nutrition 9, 161–186.
Dodds MW, Hsieh SC & Johnson DA (1991) The effect of
increased mastication by daily gum-chewing on salivary glandoutput and dental plaque acidogenicity. Journal of Dental
Research 70, 1474–1478.
Doorenbos H (1977) Metabolism of Lactitol . Arkelsedijk, The
Netherlands: PURAC Biochem BV , Gorinchem.
Drost H, Gierlich P, Spengler M & Jahnke K (1980) Blutglucose
and Seruminsulin nach oraler Applikation von Palatinit imHealth potential of polyols 185

Vergleich zu Glucose bei Diabetikern vom Erwachsenentyp
(Blood glucose and serum insulin after oral administration ofPalatinit in comparison with glucose in diabetics of the late-onset type). Verhandlungen der Deutschen Gesellschaft für
innere Medizin 86, 978–981.
Drost H, Spengler M, Kleophas W, Schmitz H & Jahnke K (1985)
Comparative study of the effect of a standardised breakfast con-taining sorbitol, fructose or sucrose in type-II diabetes mellitus.Aktulle Ernährungsmedizin 10, 195–198.
Dutch Nutrition Council (1987) The Energy Values of Polyols .
Recommendations of the Committee on Polyols. The Hague:Nutrition Council.
Ellis FW & Krantz JC (1941) Sugar alcohols XXII. Metabolism
and toxicity studies with mannitol and sorbitol in man and ani-mals. Journal of Biological Chemistry 141, 147–151.
Ellis FW & Krantz JC (1943) Sugar alcohols XXIV . The metabo-
lism of sorbitol in diabetes. Annals of Internal Medicine 18,
792–796.
European Association for the Study of Diabetes (1995)
Recommendations for healthcare professionals in the nutritionalmanagement of patients with diabetes. Diabetes, Nutrition and
Metabolism 8, 186–189.
European Association for the Study of Diabetes (2000)
Recommendations for the nutritional management of patientswith diabetes mellitus. European Journal of Clinical Nutrition
54, 353–355.
European Communities (1990) Directive 90/496/EC: nutrition
labelling for foodstuffs. Official Journal of the European
Communities L276 , 40–44.
European Communities (1994) Directive 94/35/EC: sweeteners
for use in foodstuffs. Official Journal of the European
Communities L237 , 2–12.
European Diabetes Policy Group (1999 a) A desktop guide to type-
1 (insulin dependent) diabetes mellitus. Diabetes Medicine 16,
253–266.
European Diabetes Policy Group (1999 b) A desktop guide to type-
2 diabetes mellitus. Diabetes Medicine 16, 716–730.
Featherstone JDB (1995) Effect of isomalt sweetener on the caries
process: a review. Journal of Clinical Dentistry 5, 82–85.
Featherstone JDB (2000) The science and practice of caries preven-
tion. Journal of the American Dental Association 131, 887–899.
Felber JP, Tappy L, V ouillamoz D, Radin JP & Jéquier E (1987)
Comparative study of maltitol and sucrose by means of contin-uous indirect calorimetry. Journal of Enteral and Parenteral
Nutrition 11, 250–254.
Felix YF, Hudson MJ, Owen RW, Ratcliffe B, van Es AJH, nan
Velthuijsen JA & Hill MJ (1990) Effect of dietary lactitol on thecomposition and metabolic activity of the intestinal microflorain the pig and in humans. Microbial Ecology Health and
Disease 3, 259–267.
Food and Agriculture Organization (1996–1999) Food and
Nutrition Paper no. 52, addenda 4–7. Rome: Food and
Agriculture Organization.
Food and Agriculture Organization (1998) Carbohydrates in
Human Nutrition. Food and Nutrition Paper no. 66. Rome:
Food and Agriculture Organization.
Ford ES & Liu S (2001) Glycemic index and serum high-density
lipoprotein cholesterol concentration among US adults.Archives of Internal Medicine 161, 572–576.
Foster-Powell K, Holt SH & Brand-Miller JC (2002) International
table of glycaemic index and glycaemic load: 2002. American
Journal of Clinical Nutrition 76, 5–56.
Friedman G (1991) Diet and the irritable bowel syndrome.
Gastroenterology Clinics of North America 20, 313–324.
Frost G, Leeds A, Trew G, Margara R & Dormhorst A (1998)
Insulin sensitivity in women at risk of coronary heart disease andthe effect of a low-glycaemic diet. Metabolism 47, 1245–1251.Frost G, Leeds AA, Doré CJ, Maderios S, Brading S & Dornhorst
A (1999) Glycaemic index as a determinant of serum HDL cho-lesterol concentration. Lancet 353, 1045–1048.
Frost G, Wilding J & Beecham J (1994) Dietary advice based on
the glycaemic index improves dietary profile and metabolic con-trol in type 2 diabetic patients. Diabetes Medicine 11, 397–401.
Gee JM, Cooke D, Gorick S, Wortley GM, Greenwood RH,
Zumbé A & Johnson IT (1991) Effects of conventional sucrose-based, fructose-based and isomalt-based chocolates on post-prandial metabolism in non-insulin-dependant diabetics.European Journal of Clinical Nutrition 45, 561–566.
Gehring F, Mäkinen KK, Larmas M & Scheinin A (1975) Turku
sugar studies. X. Occurrence of polysaccharide forming strepto-cocci and ability of mixed plaque microbiota to ferment variouscarbohydrates. Acta Odontologica Scandinavica 70, Suppl.,
223–237.
Giacco R, Parillo M, Rivellse AA, Lasorella G, Giacco A,
D’Episcopo L & Richardi G (2000) Long-term dietary treat-ment with increased amounts of fibre rich low glycaemic nat-ural foods improves blood glucose control and reduces thenumber of hypoglycaemic events in type 1 diabetic patients.Diabetes Care 23, 1461–1466.
Gibson GR & Roberfroid MB (1995) Dietary modulation of the
human colonic microbiota. Introducing the concept of prebiotic.Journal of Nutrition 125, 1401–1412.
Gilbertson HR, Brand-Miller JC, Thorburn AW, Evans S,
Chondros P & Werther GA (2001) The effect of flexible lowglycemic index dietary advice versus measured carbohydrateexchange diets on glycemic control in children with type 1 dia-betes. Diabetes Care 24, 1137–1143.
Glinsmann WH, Irausquin H & Park YK (1986) Evaluation of
health aspects of sugars contained in carbohydrate sweeteners.Report of Sugars Task Force, 1986. Journal of Nutrition 116,
Suppl. 11, s1–s216.
Gracey M (1982) Intestinal microflora and bacterial growth in
early life. Journal of Pediatric Gastroenterology and Nutrition
1, 13–22.
Gray DS (1995) The clinical uses of dietary fiber. American
Family Physician 51, 419–426.
Grenby TH, Phillips A & Mistry M (1989) Studies of the dental
properties of lactitol compared with five other bulk sweetenersin vitro. Caries Research 23, 315–319.
Grimble GK, Patil DH & Silk DBA (1988) Assimilation of lacti-
tol, an unabsorbed disaccharide, in the normal human colon.Gut29, 1666–1671.
Guimaraes EV , Goulart EM & Penna FJ (2001) Dietary fiber
intake, stool frequency and colonic transit time in chronic func-tional constipation in children. Brazilian Journal of Medical
Biology and Research 34, 1147–1153.
Haines ST (1995) Treating constipation in the patient with dia-
betes. Diabetes Education 21, 223–232.
Hassinger W, Sauer G, Cordes U, Krause U, Beyer J & Baessler
KH (1981) The effect of equicaloric amounts of xylitol, sucroseand starch on insulin requirements and blood glucose levels ininsulin-dependent diabetes. Diabetologia 21, 37–40.
Havenaar R, Huis in’t Veld JHJ, Baker-Dirks O & Stoppelaar JD
(1978) Health and sugar substitutes. In Proceedings of the
ERGOB Conference on Sugar Substitutes, Geneva , pp. 192–218
[B Guggenheim, editor]. Basel, Switzerland: Karger.
Hawksworth G, Drasar BS & Hill MJ (1971) Intestinal bacteria
and the hydrolysis of glycosidic bonds. Journal of Medical
Microbiology 41, 451–459.
Hayes C (2001) The effect of non-cariogenic sweeteners on the
prevention of dental caries: a review of the evidence. Journal of
Dental Education 65, 1106–1109.
Henderson L, Gregory J, Irving K & Swan G (2003) The National
Diet and Nutrition Survey: Adults Aged 19–64 Years . London:
TSO.186 G. Livesey

Herman RH (1974) Hydrolysis and absorption of carbohydrates,
and adaptive responses of the jejunum. In Sugars in Nutrition ,
pp. 145–172 [HL Sipple and KW McNutt, editors]. New York:Academic Press Inc.
Hill MJ (1985) Bacteria and colorectal adenomas. Topics in
Gastroenterolology 13, 237–252.
Hill MJ, Melville D, Lennard-Jones J, Neale K & Richie JK
(1987) Faecal bile acids, dysplasia and carcinoma in ulcerativecolitis. Lancet ii, 185–186.
Holt SHA, Miller JCB & Petocz P (1997) An insulin index of
foods: the insulin demand generated by 1000-kJ portions ofcommon foods. American Journal of Clinical Nutrition 66,
1264–1276.
Howlett J (2001) Low-digestible carbohydrates – the regulatory
framework. British Journal of Nutrition 85, Suppl. 1, S55–S58.
Hyams JS (1983) Sorbitol intolerance: an unappreciated cause of
functional gastrointestinal complaints. Gastroenterology 84,
30–33.
Huttunen KJ, Mäkinen KK & Scheinin A (1975) Effects of
sucrose, fructose and xylitol diets on glucose, lipid and uratemetabolism. Acta Odontologica Scandinavica 70, 239–245.
Imfeld T (1983) Identification of Low Caries Risk Dietary
Components . Basel, Switzerland: Karger Verlag.
Imfeld T (1993) Efficacy of sweeteners and sugar substitutes in
caries prevention. Caries Research 27, Suppl. 1, 50–55.
International Diabetes Institute Australia (2002) Diabetes preven-
tion programs: eat well live well. www.diabetes.com.au/liv-ing_with/healthpromotion.htm
Ishikawa M, Miyashita M, Kawashima Y , Nakamura T, Saitou N
& Modderman J (1996) Effects of oral administration of ery-thritol on patients with diabetes. Regulatory Toxicology and
Pharmacology 24, s303–s308.
Isokangas P, Söderling E, Pienihäkkinen K & Alanen P (2000)
Occurrence of dental decay in children after maternal consump-tion of xylitol chewing gum, a follow-up from 0 to 5 years ofage. Journal of Dental Research 79, 1885–1889.
Isokangas P, Tenovuo J, Söderling E, Männistö H & Mäkinen KK
(1991) Dental caries and mutans streptococci in the proximalarea of molars affected by the habitual use of xylitol chewinggum. Caries Research 25, 444–448.
Järvi AE, Karlström BE, Granfeldt YE, Björck IE, Asp N-G &
Vessby BOH (1999) Improved glycemic control and lipid pro-file and normalized fibrinolytic activity on a low-glycemicindex diet in type 2 diabetic patients. Diabetes Care 22, 10–18.
Jenkins DJ, Jenkins AL, Wolever TMS, Collier GR, Rao A V &
Thompson LU (1987) Starchy foods and fiber: reduced rate ofdigestion and improved carbohydrate metabolism.Scandinavian Journal of Gastroenterology 22, 132–141.
Jenkins DJ, Wolever TM, Buckley G, Lam KY , Giudici S,
Kalmusky J, Jenkins AL, Patten RL, Bird J, Wong GS & JosseRG (1988) Low-glycemic-index starchy foods in the diabeticdiet. American Journal Clinical Nutrition 48, 248–254.
Jenkins DJA, Kendall CWC, Augustin LSA, Franceschi S,
Marchie A, Jenkins AL & Axelsen M (2002) Glycemic index:overview of implications in health and disease. American
Journal of Clinical Nutrition 76, 266S–273S.
Jenkins DJA, Wolever TM, Taylor RH, Barker H, Fielden H,
Baldwin JM, Bowling AC, Newman HC, Jenkins AL & GoffDV (1981) Glycemic index of foods: a physiological basis forcarbohydrate exchange. American Journal of Clinical Nutrition
34, 362–366.
Kamoi M (1974) Study on metabolism of maltitol. Part 2. Clinical
experiments. Journal of the Japanese Diabetes Society 18,
451–460.
Kandelman D (1997) Sugar, alternative sweeteners and meal fre-
quency in relation to caries prevention: new perspectives.British Journal of Nutrition 77, Suppl. 1, S121–S128.Kaneko T, Kohmoto T, Kikuchi H, Shiota M, Iino H & Mitsukoka
T (1994) Effects of isomalto-oligosaccharides with differentdegrees of polymerisation on human fecal bifidobacteria.Bioscience Biotechnology and Biochemistry 58, 2288–2290.
Kaspar L & Spengler M (1984) Wirkung oraler Gaben von
Palatinit auf den Insulinverbrauch bei Typ-I-Diabetikern (Effectof oral doses of Palatinit on insulin requirements in type I dia-betics). Aktulle Ernährungsmedizin 9, 60–64.
Kapur A & Kapur K (2001) Relevance of glycemic index in the
management of post-prandial glycaemia. Journal of the
Association of the Physicians of India 49, 42–45.
Kawanabe J, Hirasawa M, Takeuchi T, Oda T & Ikeda T (1992)
Noncariogenicity of erythritol as a substrate. Caries Research
26, 358–362.
Kearsley MW, Birch GG & Lian-Loh RHP (1982) The metabolic
fate of hydrogenated glucose syrups. Starch 8, 279–283.
Keller U & Froesch ER (1972) Vergleichende Untersuchungen
über den Stoffwechsel von Xylit, Sorbit und Fruktose beimMenschen (Comparative investigations on the metabolism ofxylitol, sorbitol and fructose in humans). Schweizerische
Medizinische Wochenschrift 102, 1017–1022.
Keup U & Püttner J (1974) Serumglucose- und -insulinverlauf bei
gesunden Probanden nach einmaliger oraler Palatinit- bzw.Saccharosebelastung (Determination of blood sugar and plasmainsulin in healthy patients having orally absorbed an oral doseof palatinit or saccharose). Bayer AG, Pharma-Bericht no. 4781vom 01·07.
Khaw KT, Wareham N, Luben R, Bingham S, Oakes S, Welch A
& Day N (2001) Glycated haemoglobin, diabetes, and mortalityin men in Norfolk cohort of European Prospective Investigationof Cancer and Nutrition (EPIC-Norfolk). British Medical
Journal 322, 1–6.
Koch T & Hudson S (2000) Older people and laxative use: litera-
ture review and pilot study report. Journal of Clinical Nursing
9, 516–525.
König KG (1990) Changes in the prevalence of dental caries: how
much can be attributed to dietary change. Diet, Nutrition and
Dental Caries 24, Suppl. 1, 16–18.
Leach SA (1987) Sugar substitutes and remineralisation. Deutsche
Zahnarztliche Zeitschrift 42, S135–S138.
Lederle FA, Busch DL, Mattox KM, West MJ & Aske DM (1990)
Cost-effective treatment of constipation in the elderly: a ran-domised double-blind comparison of sorbitol and lactulose.American Journal of Medicine 89, 597–601.
Lee BM & Wolever TMS (1998) Effect of glucose, sucrose and
fructose on plasma glucose and insulin responses in normalhumans: comparison with white bread. European Journal of
Clinical Chemistry 52, 924–928.
Levy RD, Segal I, Hassan H & Saadia R (1994) Stool weight and
faecal pH in two South African populations with a dissimilarcolon cancer risk. South African Journal of Surgery 32,
127–128.
Life Sciences Research Office (1994) The Evaluation of the
Energy of Certain Sugar Alcohols Used as Food Ingredients .
Bethesda, MD: Life Sciences Research Office, Federation ofAmerican Societies for Experimental Biology.
Life Sciences Research Office (1999) Evaluation of the Net
Energy Value of Maltitol . Bethesda, MD: Life Sciences
Research Office, Federation of American Societies forExperimental Biology.
Liu S, Manson JE, Stampfer MJ, Rexrode KM, Hu FB, Rimm EB
& Willett WC (2000 a) Whole grain consumption and risk of
ischemic stroke in women: a prospective study. Journal of the
American Medical Association 284, 1534–1540.
Liu S, Willett WC, Stumper MY , Hun FIB, Franz M, Sampson
L, Heinekens CH & Manson JED (2000 b) A prospective study
of dietary glycaemic load, carbohydrate intake, and risk ofHealth potential of polyols 187

coronary heart disease in US women. American Journal of
Clinical Nutrition 71, 1455–1461.
Livesey G (1990 a) The impact of the concentration and dose of
PalatinitRin foods and diets on energy value. Food Sciences
and Nutrition 42F, 223–243.
Livesey G (1990 b) On the energy value of sugar alcohols with the
example of isomalt. In International Symposium on Caloric
Evaluation of Carbohydrates , pp. 141–164. Kyoto, Japan: The
Japan Association of Dietetic and Enriched Foods.
Livesey G (1992) Energy values of dietary fibre and sugar alco-
hols for man. Nutrition Research Reviews 5, 61–84.
Livesey G (1993) Comments on the methods used to determine the
energy values of carbohydrates: dietary fibre, sugar alcohols andother bulking agents . International Journal of Food Sciences
and Nutrition 44, 221–241.
Livesey G (2000 a) Studies on Isomalt – Published and
Unpublished . Wymondham, UK: Independent Nutrition Logic.
Livesey G (2000 b) The absorption of stearic acid from triacyl-
glycerols: an inquiry and analysis. Nutrition Research Reviews
13, 185–214.
Livesey G (2001) Tolerance of low-digestible carbohydrates – a
general view. British Journal of Nutrition 85, Suppl. 1, S7–S16.
Livesey G (2002 a) Thermogenesis associated with fermentable
carbohydrate in humans, validity of indirect calorimetry, andimplications of dietary thermogenesis for energy requirements,food energy and body weight. International Journal of Obesity
26, 1553–1569.
Livesey G (2002 b) Approaches to health via lowering postpran-
dial glycaemia. British Journal of Nutrition 88, 741–744.
Livesey G, Buss D, Coussement P, Edwards DG, Howlett J, Jones
DA, Kleiner JE, Müller D & Sentko A (2000) Suitability of tra-ditional energy values for novel foods and food ingredients.Food Control 11, 250–289.
Livesey G, Johnson IT, Gee JM, Smith T, Lee WA, Hillan KA,
Meyer J & Turner SC (1993) ‘Determination’ of sugar alcoholand Polydextrose
Rabsorption in humans by the breath hydro-
gen (H2O) technique: the stoichiometry of hydrogen produc-
tion and the interaction between carbohydrates assessed in vivo
and in vitro . European Journal of Clinical Nutrition 47,
419–430.
Livesey G, Wilson PDG, Roe MA, Faulks RM, Oram LM, Brown
JC, Eagles J, Greenwood RH & Kennedy H (1998) Splanchnicretention of intraduodenal and intrajejunal glucose in healthyadults. American Journal of Physiology 38, E709–E716.
MacDonald I, Keyser A & Pacy D (1978) Some effects, in man, of
varying the load of glucose, sucrose, fructose or sorbitol on var-ious metabolites in blood. American Journal of Clinical
Nutrition 31, 1305–1311.
MacGillivary PC, Finley HVL & Binns TB (1959) Use of lactu-
lose to create a preponderance of lactobacilli in the intestine ofbottle fed infants. Scottish Medical Journal 4, 182–189.
McNaught AD (1996) Nomenclature of carbohydrates (JCBN).
Pure and Applied Chemistry 68, 1919–2008. www.chem.qmul.
ac.uk/iupac/2carb/
Macpherson G (1990) Black’ s Medical Dictionary . London: Black
A & C.
McRorie J, Zorich N, Riccardi K, Filloon T, Wason S & Giannalla
R (2000) Effect of olestra and sorbitol consumption on objec-tive measures of diarrhea: impact of stool viscosity on commongastrointestinal symptoms. Regulatory Toxicology and
Pharmacology 31, 59–67.
Mäkinen KK, Isotupa KP, Kivilompolo T, Mäkinen PL, Toivanen
J & Soderling E (2001) Comparison of erythritol and xylitolsaliva stimulants in the control of dental plaque and mutansstreptococci. Caries Research 35, 129–135.
Mäkinen KK, Mäkinen PL, Pape HR, Allen P, Bennett CA,
Isokangas PJ & Isotupa KP (1995) Stabilisation of rampantcaries: polyol gum and arrest of dentine caries in two long-term
cohort studies in young subjects. International Dental Journal
45, 93–107.
Mäkinen KK, Makinën PL, Pape HR Jr, Peldyak J, Hujoel P,
Isotupa KP, Soderling E, Isokangas PJ, Allen P & Bennett C(1996) Conclusion and review of the Michigan XylitolProgramme (1986–1995) for the prevention of dental caries.International Dental Journal 46, 22–34.
Marteau P & Flourié B (2001) Tolerance to low-digestible carbo-
hydrates: symptomatology and methods. British Journal of
Nutrition 84, Suppl. 1, S17–S21.
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF
& Turner RC (1985) Homeostasis model assessment: insulinresistance and beta-cell function from fasting plasma glucoseand insulin concentrations in man. Diabetologia 28, 412–419.
Mehnert H (1971) The relative value of sugar substitutes and arti-
ficial sweeteners in the diet of diabetics. Deutsche Gesselschaft
für Ernärung 20. Darmstadt: Steinkopff.
Mehnert H, Struhlfauth K, Mehnert B, Weiner L & Hoelflmayr X
(1960) Uber die Moglichkeiten der Verabreichung hoher per-oraler Gaben von Fructose, Sorbit oder Fructose/Sorbit-gesmishan Diabetiker (On the possibility of the administration bymouth of fructose, sorbitol or fructose/sorbitol gesmish in dia-betics). Müchener Medizin Wochenschrift 102, 1–11.l.
Meyer KA, Kushi LH, Jacobs DR Jr, Slavin J, Sellers TA &
Folsom AR (2000) Carbohydrates, dietary fiber, and incidenttype 2 diabetes in older women. American Journal of Clinical
Nutrition 71, 921–930.
Mimura G, Koga T, Oshikawa K, Kido S, Sadanaga T, Jinnouchi
T, Kawagchi K & Mori N (1972) Maltitol tests with diabetics.Journal of Japanese Nutrition 30, 145–152.
Mital BK & Garg SK (1995) Anticarcinogenic, hypocholestero-
laemic, and antagonistic activities of Lactobacillus acidophilus.Crotatian Review of Microbiology 21, 175–214.
Mitsouka T (1992) Intestinal flora and aging. Nutrition Reviews
50, 438–446.
Morgan LM, Tredger JA, Hampton SM, French AP, Peake JCF &
Marks V (1988) The effect of dietary modification and gly-caemia on gastric emptying and gastric inhibitory polypeptide(GIP) secretion. British Journal of Nutrition 60, 29–37.
Mortensen PB, Holtug K & Rasmmusen HS (1988) Short-chain
fatty acid production from mono- and disaccharides in a fecalincubation system: implications for colonic fermentation ofdietary fiber in humans. Journal of Nutrition 118, 321–325.
Mühlemann HR (1971) Intra-oral radio telemetry. International
Dental Journal 21, 456–465.
Müller-Hess R, Geser CA, Bonjour J-P, Jequier E & Felber J-P
(1975) Effects of oral xylitol administration on carbohydrateand lipid metabolism in normal subjects. Infusiontherapie 2,
247–252.
Nasrallah SM & Iber FL (1969) Mannitol absorption and metabo-
lism in man. American Journal of Medical Sciences 258,
80–88.
Natah SS, Hussein KR, Touominen JA & Koivisto V A (1997)
Metabolic response to lactitol and xylitol in healthy men.American Journal of Clinical Nutrition 65, 947–950.
National Research Council (1989) Diet and Health: Implications
for Reducing Chronic Disease Risk . Washington, DC:
Committee on Diet and Health, Food and Nutrition Board.Commission on Life Sciences. National Research Council.National Academy Press.
Newbrun E (1982) Sucrose in the dynamics of the carious
process. International Dental Journal 32, 13–23.
Nguyen NU, Dumoulin G, Henriet M-T, Berthelay S & Regnard J
(1993) Carbohydrate metabolism and urinary excretion of cal-cium and oxalate after ingestion of polyol sweeteners. Journal
of Clinical Endocrinology and Metabolism 77, 388–392.188 G. Livesey

Nilsson U & Jägerstad M (1987) Hydrolysis of lactitol, maltitol
and Palatinint by human intestinal biopsies. British Journal of
Nutrition 58, 199–206.
Noda K, Nakayama K & Oku T (1994) Serum and insulin levels
and erythritol balance after oral administration of erythritol inhealthy subjects. European Journal of Clinical Nutrition 48,
286–292.
Nurko S, Baker SS, Colletti RB, Lorenzo CD, Ector W & Liptak
GS (2001) Contemporary pediatrics® archive. www.contpeds.com/past issues/Dec 2001/CME
Oku T & Noda K (1990) Erythritol balance study and estimation
of metabolisable energy of erythritol. In Caloric Evaluation of
Carbohydrates , pp. 65–75 [N Hosoya, editor]. Tokyo: Research
Foundation for Sugar Metabolism.
Orchard TJ, Dorman JS, Maser RE, Becker DJ, Ellis D, LaPorte
RE, Kuller LH, Wolson SK & Drash AL (1990) Factors associ-ated with avoidance of severe complications after 25 yrs ofIDDM: Pittsburgh Epidemiology of Diabetes ComplicationsStudy 1. Diabetes Care 13, 741–747.
Ornskov F, Nielsen CB, Nielsen ML & Christophersen SJ (1988)
Peroral mannitol in whole-gut irrigation for chronic constipa-tion in children. Ugeskr Laeger 150, 847–849.
Paige DM, Bayless TM & Davies LR (1992) Palatinit® (isomalt)
digestibility in children. Nutrition Research 12, 27–37.
Pelletier X, Hanesse B, Bornet F & Derby G (1994) Glycaemic
and insulinaemic responses in healthy volunteers upon inges-tion of maltitol and hydrogenated glucose syrups. Diabetes and
Metabolism 20, 291–296.
Petzoldt R, Lauer P, Spengler M & Schöffling K (1982 a)
Palatinit bei typ-II Diabetiken: Wirkung auf blutglucose, seur-minsulin, C-peptide und freie Fettsauren im verlagleich mitglucose (Palatinit® in type II diabetics: effect on bloodglucose, serum insulin, C peptide and free fatty acids incomparison with glucose). Deutsche Medizinesche
Wochenschrift 107, 1910–1913.
Petzoldt R, Müller-Siebert A, Schöffling K & Spengler M (1982 b)
Zur Wirkung von Saccharose, Fruktose und Sorbit auf denKohlenhydrate-, fett- und Purinstoffwechsel bei Typ II-dia-betikern (On the effect of saccharose, fructose and sorbitol oncarbohydrate-, fat-, and purine metabolism in type II diabetics).Aktulle Ernährungsmedizin 7, 151–156.
Piche T, Zerbib F, Varannes SB, Cherbut C, Anini Y , Roze C, le
Quellec A & Galmiche JP (2000) Modulation by colonic fer-mentation of LES function in humans. American Journal of
Physiology 278, G578–G584.
Pitzalis G, Deganello F, Mariani P, Chiarini-Testa MB, Virgilii F,
Gasparri R, Calvani L & Bonamico M (1996) Lactitol inchronic idiopathic constipation in children. La Pediatria
Medica e Chirurgica 17, 223–226.
Pometta D, Trabichet C & Spengler M (1985) Effects of a 12-
week administration of isomalt on metabolic control in type-II-diabetics. Aktulle Ernährungsmedizin 10, 174–177.
Pontes FA, Silva AT & Cruz AC (1995) Colonic transit times and
the effect of lactulose or lactitol in hospitalized patients.European Journal of Gastroenterology and Hepatology 7,
441–446.
Ponz de Leon M & Roncucci L (1997) Chemoprevention of colo-
rectal tumors: role of lactulose and of other agents.Scandinavian Journal of Gastroenterology 222, Suppl., 72–75.
Ravelli GP, White A, Spencer R, Hotton P, Harbron C & Keen R
(1995) The effect of lactitol intake upon stool parameters andthe faecal bacterial flora in chronically constipated women.Acta Therapeutica 21, 243–255.
Rizkalla SW, Luo J, Wils D, Bruzzo F & Slama G (2002)
Glycaemic and insulinaemic responses to a new hydrogenatedstarch hydrolysate in healthy and type 2 diabetic subjects.Diabetes and Metabolism 28, 385–390.Roediger WEW (1990) The starved colon – diminished mucosal
nutrition, diminished absorption, and colitis. Diseases of the
Colon and Rectum 33, 858–862.
Rolla G, Scheie AA & Ciardi JE (1985) Role of sucrose in plaque
formation. Scandinavian Journal of Dental Research 93,
105–111.
Roncucci L, Di Donato P, Carati L, Carati L, Ferrari A, Perini M,
Bertoni G, Bedogni G, Paris B, Svanoni F, Girola M & Ponz deLeon M (1993) Antioxidant vitamins or lactulose for the preven-tion of the recurrence of colorectal adenomas. Colorectal CancerStudy Group of the University of Modena and the Health CareDistrict 16. Diseases of the Colon and Rectum 6, 227–234.
Rowland IR (1991) Nutrition and gut microflora metabolism. In
Nutrition, Toxicity and Cancer , pp. 113–136 [IR Rowland, edi-
tor]. Boston, MA: CRC Press.
Royal College of General Practitioners (1986) Morbidity Statistics
from General Practice – Third National Study, 1981–1982 .
Series MB5 (1). London: H.M. Stationery Office.
Rugg-Gunn AJ (1989) Lycasin and the prevention of dental caries.
In Progress in Sweeteners , pp. 311–329 [T Grenby, editor].
Amsterdam: Elsevier.
Sacchetta A, Bottini C, Guarisco R, Candiani C & Brambilla M
(2000) Acceptability, efficacy and tolerability of lactitol syrupin chronic or hospitalisation-related constipation. European
Bulletin of Drug Research 8, 1–6.
Salford Symposium Consensus (2001) Consensus statements from
participants of the International Symposium on Low DigestibleCarbohydrates. British Journal of Nutrition 85, Suppl. 1, S5.
Salmerón J, Ascherio A, Rimm EB, Colditz GA, Spiegelman D,
Jenkins DJ, Stampfer MJ, Wing AL & Willet WC (1997 a)
Dietary fibre, glycaemic load, and risk of NIDDM in men.Diabetes Care 20, 545–550.
Salmerón J, Manson JE, Stampfer MJ, Colditz GA, Wing AL &
Willet WC (1997 b) Dietary fibre, glycaemic load, and risk of
non-insulin-dependent diabetes in women. Journal of the
American Medical Association 277, 472–477.
Salminen S, Salminen E & Marks V (1982) The effects of xylitol
on the secretion of insulin and gastric inhibitory polypeptide inman and rats. Diabetologia 22, 480–482.
Samata A, Burden AC & Jones GR (1985) Plasma glucose
responses to glucose, sucrose, and honey in patients with dia-betes mellitus: an analysis of glycaemic and peak incrementalindices. Diabetic Medicine 2, 371–373.
Samelson SL, Nelson RL & Nyhus LM (1985) Protective role of
faecal pH in experimental colon carcinogenesis. Journal of the
Royal Society of Medicine 78, 230–233.
Sandler RS, Jordan MC & Shelton BJ (1990) Demographic and
dietary determinants of constipation. American Journal of
Public Health 80, 185–189.
Scheie AA, Fejerskov O & Danielsen B (1998) The effect of xyli-
tol-containing chewing gums on dental plaque and acidogenicpotential. Journal of Dental Research 77, 1547–1522.
Scheppach W, Bartram P & Richer F (1995) Management of
diversion colitis, pouchitis and distal ulcerative colitis. InPhysiological and Clinical Aspects of Short-chain Fatty Acids ,
pp. 353–360 [JH Cummings, JL Rombeau and T Sakata, edi-tors]. Cambridge, UK: University Press.
Scheppach W, Luehrs H & Menzel T (2001) Beneficial health
effects of low digestible carbohydrate consumption. British
Journal of Nutrition 85, Suppl. 1, S23–S30.
Screvola D, Bottari G, Oberto L, Monzillo V , Perversi L &
Marone P (1993 a) Intestinal bacterial toxins and alcohol liver
damage: effect of lactitol, a synthetic disaccharide. La Clinica
Dietologica 20, 297–314.
Screvola D et al (1993 b) The role of lactitol in the regulation of
intestinal microflora in liver disease. Giornale di Malattie
Infettive e Parassitarie 45, 906–918.Health potential of polyols 189

Secchi A, Pontiroli AE, Cammille L, Bizzi A, Cini M & Pozza G
(1986) Effects of oral administration of maltitol on plasma glu-cose, plasma sorbitol, and serum insulin levels in man.Klinische Wochenschrift 64, 265–269.
Segal I (1998) Rarity of colorectal adenomas in the African black
population. European Journal of Cancer Prevention 7, 387–391.
Segal I (2002) Physiological small bowel malabsorption of carbo-
hydrates protects against bowel diseases in Africans. Journal of
Gastroenterology and Hepatology 17, 249–252.
Segal I, Hassan H, Walker AR, Becker P & Braganza J (1995)
Fecal short chain fatty acids in South African urban Africansand whites. Diseases of the Colon and Rectum 38, 732–734.
Sels JP, Verdonk HE & Wolffenbuttel BH (1998) Effects of acar-
bose (Glucobay) in persons with type 1 diabetes: a multicentrestudy. Diabetes Research and Clinical Practice 41, 139–145.
Sheinin A, Makinën KK & Ylitako K (1974) An intermediate
report on the effects of sucrose, fructose and xylitol diets on thecaries incidence in man. Acta Odontologica Scandinavia 32,
383–412.
Shively CA, Apgar JL & Tarka SM (1986) Postprandial glucose
and insulin responses to various snacks of equivalent carbohy-drate content in normal subjects. American Journal of Clinical
Nutrition 43, 335–342.
Signorelli P, Croce P & Dede A (1996) A clinical study of the use
of a combination of glucomannan with lactulose in the consti-pation of pregnancy. Minerva Ginecologica 48, 577–582.
Sinaud S, Montaurier C, Wils D, Vernet J, Brandolini M,
Boutloup-Demange C & Vermorel M (2002) Net energy valueof two low digestible carbohydrates, Lycasin HBC and thehydrogenated polysaccharide constituent of Lycasin HBC inhealthy human subjects and their impact on nutrient digestiveutilisation. British Journal of Nutrition 87, 131–139.
Slama G (1989) Study of the effects on glycaemia and insuli-
naemia in normal subject and non-insulin dependent diabeticsof three hydrogenated derivatives: Palatinit®, Maltisorb® andLycasin®. Hotel-Dieu, Paris: Laboratory Services forDiabetology.
Spengler M, Somogyi JC, Pletcher E & Boehme K (1987)
Tolerability, acceptance and energetic conversion of isomalt(Palatinit) in comparison with sucrose. Aktulle
Ernährungsmedizin 12, 210–214.
Staiano A, Simeone D, Del Giudice E, Miele E, Tozzi A &
Toraldo C (2000) Effect of the dietary fiber glucomannan onchronic constipation in neurologically impaired children.Journal of Pediatrics 136, 41–45.
Steinke J, Wood FC, Domage L, Marble A & Renold AE (1961)
Evaluation of sorbitol in the diet camp of diabetic children atcamp. Diabetes 10, 218–227.
Stevens J, Levitsky DA, VanSoest PJ, Robertson JB, Kalkwarf HJ
& Roe DA (1987) Effects of psyllium and wheatbran on sponta-neous energy intake. American Journal of Clinical Nutrition 46,
812–817.
Stewart D (2001) Consumption and consumer perceptions: report
of a workshop. British Journal of Nutrition 85, Suppl. 1,
S61–S62.
Stratton IM, Adler AI, Neil AW, Matthews DR, Manley SE, Cull
CA, Hadden D, Turner RC & Holman RR on behalf of the UKProspective Diabetes Study Group (2000) Association of gly-caemia with macrovascular and microvascular complications oftype 2 diabetes (UKPDS 35): prospective observational study.British Medical Journal 321, 405–412.
Sweeney M (1997) Constipation. Diagnosis and treatment. Home
Care Provider 2, 250–255.
Sydney University’s Glycaemic Index Research Service (2002)
Glycaemic Index Report – Isomalt . Sydney, Australia: Sydney
University’s Glycaemic Index Research Service (SUGiRS),University of Sydney.Takatsuka T (2000) Influence of Palatinit® and xylitol on dem-
ineralisation/remineralisation on bovine enamel. Cariology
Today 1, 37–40.
Tarao K, Tamai S, Ito Y , Okawa S & Hayashi M (1995) On
changes in faecal bacteria flora by administration of lactitol inliver cirrhosis patients with hepatic encephalophathy. Journal of
the Japanese Society of Gastroenterology 92, 1037–1050.
Teramoto F, Rokutan K, Kawakami Y , Fujimura Y , Uchida J, Oku
K, Oku M & Yoneyama M (1996) Effect of 4
G-ß-D-galactosyl
sucrose (lactosucrose) on fecal microflora in patients withchronic inflammatory bowel disease. Journal of
Gastroenterology 31, 33–39.
Thannhauser SJ & Meyer KH (1929) Sorbit (Sionin) als
Kohlehydraterstz für den Diabeteskranken (Sorbitol (Sionin) ascarbohydrate for diabetics). Müchener Medizinische
Wochenschrift 76, 356–360.
Thiébaud D, Jacot E, Schmitz H, Spengler M & Felber JP (1984)
Comparative study of isomalt and sucrose by means of continu-ous indirect calorimetry. Metabolism 33, 808–813.
Thornton JR (1981) High colonic pH promotes colorectal cancer.
Lancet i, 1081–1082.
Tong Z-H, Gu W-Z & Gen Z (1987) Effect on plasma glucose and
insulin after xylitol loading in 30 normal adults. Zhonghau
Neike Zazhi 26, 420–422.
Toors FA (1992) Chewing gum and dental health – literature
review. Revue Belge de Medecine Dentaire 47, 67–92.
Tse PW, Leung SS, Chan T, Sien A & Chan AK (2000) Dietary fibre
intake and constipation in children with severe developmentaldisabilities. Journal of Paediatric and Child Health 36, 236–239.
Tsuji K, Osada Y , Shimada N, Nishimura R, Kobayashi S, Tomio
Ichikawa T & Hosoya N (1990) Energy value of sorbitol andmaltitol in healthy men and rats. In Caloric Evaluation of
Carbohydrates , pp. 77–90 [N Hosoya, editor]. Tokyo: Research
Foundation for Sugar Metabolism.
Tsunehiro J, Matsukubo T, Shiota M & Takaesu Y (1997) Effects
of a hydrogenated isomaltooligosaccharide mixture on glucansynthesis and on caries development in rats. Bioscience
Biochemistry and Biotechnology 61, 2015–2018.
Tucker DM, Sandstead HH, Logan GM, Klevay LM, Mahalko J,
Johnson LK, Inman L & Inglett GE (1981) Dietary fibre andpersonality factors as determinants of stool output.Gastroenterology 81, 879–883.
TuohyKM, Kolida S, Lustenberger AM & Gibson GR (2001) The
probiotic effects of biscuits containing partially hydrolysedguar gum and fructo-oligosaccharides – a human volunteerstudy. British Journal of Nutrition 86, 341–348.
UK Prospective Diabetes Study Group (1998) Intensive blood-
glucose control with sulphonylureas or insulin compared withconventional treatment and risk of complications in patientswith type 2 diabetes (UKPDS 33). Lancet 352, 837–854.
Unwin N, Shaw J, Zimmet P & Alberti KGMM (2002) Impaired
glucose tolerance and impaired fasting glycaemia: the currentstatus on definition and intervention. Diabetes Medicine 19,
708–723.
Vaaler S, Bjørneklett A, Jelling I, Skrede G, Hannsen K, Fausa O &
Aggenæs Ø (1987) Sorbitol as a sweetener in the diet of insulin-dependent diabetes. Acta Medica Scandinavica 221, 165–170.
Van der Hoeven JS (1979) Influence of disaccharide alcohols on
oral microflora. Caries Research 13, 301–306.
Van der Hoeven JS (1980) Cariogenicity of disaccharide alcohol
in rats. Caries Research 14, 61–66.
van Es AJH, De Groot L & V ogt JE (1986) Energy balance of
eight volunteers fed on diet supplemented with either lactitol orsaccharose. British Journal of Nutrition 56, 545–554.
van Velthuijsen JA (1990) Physiology and metabolic energy of
lactitol. In Caloric Evaluation of Carbohydrates , pp. 124–139
[N Hosoya, editor]. Tokyo: Research Foundation for SugarMetabolism.190 G. Livesey

Veitch AM, Kelly P, Segal I, Spies SK & Farthing MJ (1998)
Does sucrase deficiency in black South Africans protect againstcolonic disease. Lancet 351, 183.
Velazquez OC, Lederer HM & Rombeau JL (1996) Butyrate and
the colonocyte. Implications for neoplasia. Digestive Diseases
Science 41, 727–739.
Verina P, Frandina C, Bilotta T, Ricciardi MR, Villotti G &
Fallucca F (1995) Sorbitol malabsorption and non-specificabdominal symptoms in type II diabetics. Metabolism 44,
796–799.
Waaler SM, Assev S & Rolla G (1992) Xylitol-5-P formation by
dental plaque after 12 weeks’ exposure to xylitol/sorbitol con-taining chewing gum. Scandinavian Journal of Dental
Research 100, 319–321.
Wallace TM & Matthews DR (2000) Poor glycaemic control in
type 2 diabetes: a conspiracy of disease, suboptimal therapy andattitude. Quarterly Journal of Medicine 93, 369–374.
Wang W, Lee ET, Fabsitz R, Welty TK & Howard BV (2002)
Using HbA(1c) to improve efficacy of the American DiabetesAssociation fasting plasma glucose criterion in screening fornew type 2 diabetes in American Indians: the strong heart study.Diabetes Care 25, 1365–1370.
Wang Y-M & van Eys J (1981) Nutritional significance of fructose
and sugar alcohols. Annual Review of Nutrition 1, 437–475.
Warshaw HS & Powers MA (1999) A search for answers about
foods with polyols (sugar alcohols). The Diabetes Educator 25,
307–310, 315, 321.
Wegener M, Borsch G, Schffstein J, Luerweg C & Leverkus F
(1990) Gastrointestinal disorders in patients with insulin-treateddiabetes mellitus. Digestive Diseases 8, 23–26.
Wheeler ML, Finberg SE, Gibson R & Fineberg N (1990) Metabolic
response to oral challenge of hydrogenated starch hydrolysate ver-sus glucose in diabetes. Diabetes Care 13, 733–740.
Willibald-Ettle I & Schiweck H (1996) Properties and applica-
tions of isomalt and other bulk sweeteners. In Advances in
Sweeteners , pp. 134–149 [TH Grenby, editor]. London: Blackie
Academic & Professional.
Wolever TM (2000) Dietary carbohydrates and insulin action in
humans. British Journal of Nutrition 83, Suppl. 1, S97–S102.
Wolever TM & Bolognesi C (1996) Source and amount of carbo-
hydrate affect postprandial glucose and insulin in normal sub-jects. Journal of Nutrition 126, 2798–2806.
Wolever TM & Jenkins DJ (1986) The use of glycaemic index in
predicting the blood glucose response to mixed meals.American Journal of Clinical Nutrition 43, 167–172.
Wolever TM, Jenkins DJ, Jenkins AL & Josse RG (1991) The gly-
caemic index: methodology and clinical implications. American
Journal of Clinical Nutrition 54, 846–854.
Wolver TM, Jenkins DJ, Josse RG, Wong GS & Lee R (1987) The
glycaemic index: similarity of values derived in insulin-depen-dent and non-insulin-dependent diabetic patients. Journal of the
American College of Nutrition 6,295–302.
Wolever TM, Jenkins DJ, Vuksan V , Jenkins AL, Buckley GC,
Wong GS & Josse RG (1992 a) Beneficial effects of a low gly-
caemic index diet in type 2 diabetes. Diabetes Medicine 9,
451–458.
Wolever TM, Jenkins DJ, Vuksan V , Jenkins AL, Wong GS &
Josse RG (1992 b) Beneficial effect of low-glycemic index diet
in overweight NIDDM subjects. Diabetes Care 15, 562–564.
Yamagata S, Goto Y , Ohneda A, Anzai M, Kawashima S, Chiba
M, Maruhama Y & Yamauchi Y (1965) Clinical effects of xyli-tol on carbohydrate and lipid metabolism in diabetes. Lancet ii,
918–921.
Yamagata S, Goto Y , Ohneda A, Anzai M, Kawashima S, Kikuch
J, Chiba M, Marumama Y , Yamauchi Y & Toyota T (1969)Clinical applications of xylitol in diabetics. In Pentoses and
Pentitols , pp. 316–325 [BL Horecjker, K Lang and Y Takagi,
editors]. Berlin: Springer-Verlag.
Zaal J & Ottenhof A (1977) Influence of Lactitol on Blood Sugar
Levels after Sucrose Intake , TNO report R5443. Zeist, The
Netherlands: TNO, Centraal Instituut voor V oedingsonderzoek.
Zhong J, Luo BY , Xiang MJ, Liu HW, Zhai ZK, Wang TS & Craig
SAS (2000) Studies on the effects of polydextrose intake onphysiologic functions in Chinese people. American Journal of
Clinical Nutrition 72, 1503–1509.
Ziesenitz SC & Siebert G (1987) The metabolism and utilization
of polyols and other bulk sweeteners compared with sugars. InDevelopments in Sweeteners , vol. 3, pp. 109–149 [TH Grenby,
editor]. London and Amsterdam: Elsevier Applied SciencePublishing.
Zumbé A & Brinkworth RA (1992) Comparative studies of gas-
trointestinal tolerance and acceptability of milk chocolate con-taining either sucrose, isomalt or sorbitol in healthy consumersand type II diabetics. Zeitschrift Ernährungswissschaft 31,
40–48.
Zumbé A, Lee A & Storey D (2001) Technological properties of
low digestible carbohydrates. British Journal of Nutrition 85,
Suppl. 1, S31–S45.Health potential of polyols 191