in healthy subjects12
Jun Johannes Rumessen, Fructans Susan Bode, Ole Hamberg, and Eivind Gudmand-H#{248}yer
are naturally occurring plant oligo- ten incompletely absorbed (5, 6) and which has been promoted properties. Fructans(FAs) isolated as an alternative dietary sweetener (7, 8). Unabsorbable storage from Jerusalem artichokes (He/ianthus tuberosus) were studied and structural polysaccharides may lower BG and insulin rewith respect to intestinal handling and influence on blood glu-sponses to a test meal (9-12) and dietary fiber may increase malabsorption (I 3). We therefore also studied the effects cose (BG), insulin, and C-peptide responses in eight healthy starch subjects. The responses were compared with those for fructose on absorption and metabolic responses of the addition of FAs ingestion. The effect of FAs added to a wheat-starch meal wasto a starch meal. also studied. Standardized breath-hydrogen excretion mdicated that FAs were completely malabsorbed and, after a 20-g Subjects and methods dose, traces of FA were detected in 24-h urine collections in one subject only. Orocecal transit times were longer for FAsSubjects than for lactulose and fructose. The BG and insulin increments Eight healthy subjects (two female, six male) aged 23-33 y were very low after FA ingestion, lower than after fructose inparticipated. The subjects were of ideal body weight 10% gestion, whereas hydrogen production was much higher. Areas (median and interquartile range, 74 kg and 63-76 kg, respecunder BG curves tended to be smaller when 10 g FA was added tively) and had no history of diabetes or of gastrointestinal or to a 50-g wheat-starch meal, but there was no apparent interferpulmonary disease. They were not under any medication, and ence with starch absorption. Am J Clin Nutr l990;52:675blood screening tests were normal. The study protocol apwas 81. proved by the Copenhagen County Medical Ethics Committee. KEY WORDS Breath tests, dietary carbohydrates, dietary Fructans (fructose o/igosaccharides) fiber, fructan, fructose, hydrogen, inulin, oligosaccharidcs, The FAs used in this study were unbranched chains of fructostarch
saccharides with sweetening furanose units in /3-(2l)-glycosidic binding [(2l)-f3-D-fructo-
ABSTRACt
Introduction
containing a single glucose moiety (2). For the study, FAs were purified from Jerusalem-artichoke
pur-
Fructose polymers containing a single glucose moiety [fruc- tubers in December by means ofa newly developed procedure tans (FAs)] are widespread in various plants, particularly those (Danisco A/S, Danish Sugar Corp, Nakskov, Denmark). of the Compositae and Graminae families Aiium [eg, sp, artiThe cleaned tubers were sliced and extracted in hot water. chokes, asparagus, cereals (wheat and rye), and Dahlia spj (1, The cxtraxt was purified by lime treatment, ion exchange, and 2). Synthetic FAs with sweetening properties and a low degree activated charcoal. The purified juice was concentrated by of polymerization (chain length of three to five sugar units) evaporation, and the final dry, white powder was produced in have been developed through the use ofisolated fructosyltransa vacuum dryer (patent application 1 592/88-K4). ferases (3, 4). We studied naturally occurring fructans of longer The chain lengths of FAs and the sugar residues were deterchain lengths, isolated and purified from the tubers of Jerusa- mined by high-pressure liquid chromatography (HPLC) equip1cm artichokes (He/ianthus tuberosus) by a newly developed ment (Waters Millipore, Inc, Copenhagen) with two columns procedure. Naturally occurring storage polysaccharides from plants, such as FAs, may share properties with resistant starch C From the Laboratory of Gastroenterology and Clinical Nutrition, and easily fermentable dietary fiber. Such polysaccharides Department of Internal Medicine F, and the Department of Clinical could have several potential nutritional advantages as a dietary Chemistry, Gentofte Hospital, University of Copenhagen. supplement and as a sweetening agent for diabetics. 2 Address reprint requests to JJ Rumessen, Department of Medical We studied the absorption, transit, and fermentation of FAs Anatomy C, Panum Institute, University ofCopenhagen, Blegdamsvej as well as their influence on blood glucose (BG), insulin, and3, DK-2200 N, Copenhagen, Denmark. C-peptide responses in healthy subjects. The properties of FAs Received August 4, 1989. were compared with their monomer fructose, which is very of- Accepted for publication October 1 1, 1989.
AmJC/in
Nuir
1990;52:675-81.
Printed
in USA.
1990
American
Society
forClinical
Nutrition
675
676 TABLE
Physical Texture Density 1 properties
RUMESSEN
El
generate
AL
a significant, air were days A/S) in challenges on different (Danisco sustained of 10 ppm given separated 50 in mL 200 tap in 200 mL rise in hydrogen concentration
ofthe
fructans
studied
in basis
end-expiratory
(14).
by 72
Furthermore,
order h, within on
the
fol-
(gJcm2)
Water
solubility
(gIL)
White, amorphous powder 0.75 200-300 at 30 *C; 400 at 50 C 1/3 x sucrose 130-180 0.002-0.004 (at 100-300 g/L and at 20 1 h 1%, 2 h 3%, 3 h 4%, 4 h 6%
lowing 5 g FA
in random
a single-blind
water; 2)
mL C) starch tional
lowing
tap
water; 3) 20 g FA
SAD)
tap
mL
water;
tap
(D-fructopyranose,
water; 5) 50
as bread Institute
carbohydrate
gluten-containing wheat flour (NaScience, Copenhagen) with the fol(% of dry wt): FAs 1 .4%, starch
According
3.9%; C for
process.
up salt,
in the
g wheat through
supplemented
the
in series (Biorad aminex HPX-42C, Bie & Berntsen, Copenhawithin
meal
with
15 mm.
10 g FA in 100 mL
were
eaten
tests were performed before and during all gen). Samples were dissolved in deionized water to 5 g/l00 mL Breath-hydrogen after a 12-h overnight fast and a chlorhexidine mouthand analyzed in 15-zL samples at 85 C and a flow rate of 0.3meals Duplicate samples were taken every 15 mm until orocemL/min. The frequencies of different chain lengths and resi-wash. cal transit times were determined and every 30 mm until 12 h dues in the product were as follows: fructose 3.8%; glucose ingestion of the test meal, as previously described (145.1%; sucrose 14.5%; and three-sugar units 10.6%, four units after concentrations were analyzed on an exhaled10.2%, five units 7.5%, six units 7.0%, seven units 6.1%, eight 16). Hydrogen monitor (Gas Measurements Ltd, Renfrew, Scotunits 5.3%, and more than eight units 29.9%. Enzymatic deter- hydrogen land) ( 14). Standardized symptom scores were completed mination (with kits from Boehringer, Mannheim, FRG)glu- of all tests and comprised the occurrence of flatucose and fructose in the product revealed that glucose was 1.8%throughout distension, borborygmia, abdominal pain, and diarrhea. and fructose, 2.4%. This discrepancy from the HPLC analysis lence, is due to a slight hydrolytic activity of the column toward the All symptoms were rated as none (0), mild (1), moderate (2), terminal glucose-fructose glycosidic bond, and the enzymatic or severe (3) by the subjects at fixed time intervals, and a total score for each ofthe tests was calculated. determinations are considered most accurate. Other physical symptom properties ofthe product arc shown in Table 1. In all studies, except study 1, concentrations of BG, serum
insulin Study design (IRI), and C-peptide (ICP) were determined before
meals from
ing nase
by a glucose
1 10 g galactose
and
60
mg
lactose/L.
All subjects
were
able
to IRI
and
FRG). at -20
for by
ppm
FA
Hours
1
6 and
11
10 12 after 10 g lactu-
FIG 1. Median incremental breath-hydrogen lose (L, 0), 20 g fructose (F, 0), 5 g fructan
INTESTINAL TABLE 2
(OCTTs) fructose, and hydrogen and different production doses offructan
HANDLING
OF
FRUCTANS
<
677
after in
production
Total
AUCt
SPeak
ppm 36 (27-47) 21 (10-61) 29(28-32) 51 (38-64) 65 (36-78)
H2
90 (60- 120) 25(15-225) 270(235-300) 175 (155-270) 145 (85-180) interquartile ranges
significantly slower transit time than 10 g lactuloseP ( Pratts test), and 20 g FA had a significantly faster than 5 g FA (P < 0.0 1, Pratts test) (Table 2). ANOVA ITs of 10 g lactulose, 5 g FA, 10 g FA, and 20 g significant difference (P < 0.01). Multiple-comparison dures showed that there were significant differences g lactulose and 5 g FA < 0.01) and between (P 10 and 1OgFA(P<0.05). Fasting concentrations of BG, IRI, or ICP did nificantly for the different substrates (Table ). The 3
increments in BG were significantly larger for
vary
sig-
maximal
g fructose
Medians;
than for 20 g FA(P < 0.02, Pratts test). Similarly, both net areas and positive incremental areas under the BG-vs-time curves were larger for 20 g fructose than for 20 g(PFA 0.04 = and P = 0.08, respectively)(Table 3). Peak incremental IRI was also significantly larger for 20 g fructose than for 20 g FA (P
<
incremental areas under the hydrogen-concentration-vstime curves from OCTTs. :1: aximal incremental M hydrogen concentration from lowest ous values. From the four subjects with detectable fructose malabsorption.
t Total
0.05,
Pratts
test) and 20
although
AUCs
were
not
different
(Table
previ-
4).
20
Peak
g fructose
incremental
ICP
g FA
and
(Table
AUCs
4). There
were
was
the
no
same
detectable
for
between 10 g and 20 g FA for BG, (Tables 3 and 4). With 10 g lactulose showed that both peak incremental
incremental AUCs
<
radioimmunoassay
(RIA-gnost,
Behringswerke,
Marburg,
BG were significantly of
but not for 20 g FA (P
>
larger for
0.10) FRG). After ingestion of2O g FA (study 3) urine was collected (Table 3). for 24 h and the presence ofFA in urine was analyzed by cnzyMalabsorption after ingestion of 50 g wheat starch was dematic determination offructose after perchloric acid hydrolysis tected in six subjects (calculated malabsorbed amounts 2- 1 1 ofFA (Boehringer Mannheim) (17). g). For these subjects OCITs were 6-8 h (Table 5). After 50 g wheat starch with 10 g FA the total hydrogen production Data analysis (AUCs) was not significantly different from that following the starch and 10 g (P > 0. 10, Pratts FA test); For breath-hydrogen tests orocecal transit times (OCIT), sum of 50 g wheat OCTIs for 50 g wheat starch with 10 g FA were shorter compeak incremental hydrogen concentrations (speak ), H2 and pared with the value for wheat starch alone < 0.05, (P Pratts total incremental areas under the hydrogen-concentration-vstime curves (AUC) were calculated as previously described (15,test). the addition of 10 g FA to the starch meal, a trend to1 6). The malabsorbed amounts of fructose and wheat starch After lower BG values, calculated as the positive or net increwere estimated by means of total AUCs of the 10-g lactulose ward areas under the curves, was apparent (Fig 2, Table 6) standards ( 16). In a few instances hydrogen concentration did mental respectively; Pratts test). There was not reach baseline after a 12-h period and the small additional (P = 0. 13 and P = 0.04, no difference between the magnitude or the timing ofthe peak areas were determined by extrapolation. BG values (Fig 2, Table 6). There was no detectFor BG, IRI, and ICP, fasting concentrations (ic, baseline), incremental able difference between peak incremental values or net and peak incremental values, and areas under the concentrationfructose 0.05) vs-time curves were calculated. Areas were calculated both
(P
positive incremental areas above baseline and net incremental areas (total areas minus baseline times 180 mm). Nonparametric statistics [Pratts test (1 8) and Friedmans two-way analysis ofvariance (ANOVA) with multiple-comparison procedures ( 19)] were used and, consequently, all results are expressed as medians and intcrquartile ranges. values P
<
areas 7).
under
the
IRI
and
ICP
concentra-
TABLE
Glycemic lactulose
3 responses ofdifferent and fructose in eight doses offructan subjects compared with
0.05
were
considered
significant.
PeakBGt
mmo//L
NetAUC
mmo/. L
Positive
. mm
mmo/.
AUC
L . mm
Results
The AUCs increased in proportion to increasing doses (Fig 1, Table 2). The AUCs of 10 g FA were not significantly different from AUCs of 10 g lactulose. Hydrogen production Medians; interquartile ranges in parentheses. after 20 g fructose was much lower than after 20 g FA 0.01, (P < t Maximal incremental rise ofblood glucose from fasting values. Pratts test). Four subjects had detectable malabsorption of the t Total areas under the blood-glucose-vs-time curves minus [180 20-g-fructose dose (calculated amounts 2-5 g). The OCITs de-mm times fasting values]. creased with increasing doses of FA (Table 2); 10 g FA had a Positive incremental areas above fasting values.
-56(-l l2to -36) -45 (-95 to 34) -14(-52to 18) -38 (-69 to -20)
0(Oto 24 (2 to 9(Oto 2 (0 to
1) 47) 18) 6)
678
RUMESSEN
4 (IRI)
El
AL
TABLE Insulin
and
C-peptide
(ICP)
responses
to fructans
compared
with
fructose
in eight
subjects
IRI SPeak
IRIt Net AUC Positive pmol. AUC L . mm Peak pmo//L 167 ( 100-200) 100(33-133) 133 (67-200) ICPt Net pmo/.
ICP
AUC L . mm
Positive pmo/.
AUC L . mm
pmo//L
20 g fructose lOgfructan 20 g fructan
pmol.
L . mm
86 1 (4 16- 10 19)
Medians;
t Maximal
interquartile incremental
Only mild flatulence and borborygmia were reported except diabetes mellitus was observed (22). The interesting dietary and for one subject who complained of moderate flatulence after clinical aspects of FA prompted the production of synthetic 10 g FA (Table 8). The difference between 20 g FA and 20 fructooligosaccharides. g Studies of Neosugar (Meiji Seika Co, fructose or between 50 g wheat starch and 50 g wheat starch Japan), a mixture ofFAs containing two to four fructose moleplus 10 g FA was not significant (P = 0.08 and P = 0. 16, respeccules, suggest that it is not hydrolyzed by digestive enzymes and tively; Pratts test). ANOVA revealed no significant differences increases fecal weight and excretion of volatile fatty acids in rats (4, 23). Studies in humans have suggested that Neosugar is between the symptom scores of all challenges. Diarrhea ab- or dominal pain was not reported. completely malabsorbed (24). It may promote bifidobacteria Urine analyses for FAs in 24-h collections after a 20-g dose in colonic microflora and does ot raise n BG or insulin upon were negative in seven subjects. A concentration ofO.2 corg/L ingestion (25). Yamashita et al (26) found that Neosugar lowresponding to a 24-h excretion of I 32 mg FA (< 1% ofthe load) ered fasting BG concentrations in non-insulin-dependent diawas found in one subject. betics(26). The FAs studied by us occur naturally in Jerusalem artichokes and they have longer chain lengths than Neosugar (50% have more than four fructose units). Discussion The standardized hydrogen production and the urine analyses following a 20-g-FA load, which showed traces ofFA in one FAs arc fructo-oligosaccharides with different degrees of poonly, suggest complete malabsorption of FA. lymerization that occur in varying proportions in different person larger hydrogen response to FA than to fructose plant species and with characteristic seasonal variations (20). The much lower glycemic responses and inThe most well-described FA is inulin, which has -35 fructose and the fact that FA induces sulin peaks than does fructose may indicate thatgastric acid units (2 1). FAs with 50 fructose units have been identified in or small intestinal hydrolysis of FA to fructose monomers is garlic (20). insignificant. This agrees well with in vitro studies of hydrolysis In 1874 inulin was found not to be degraded, hydrolyzed, of cereal FA by human gastric juice (27). The very modest or detected in urine or feces after ingestion in humans (22). A after FA intake seems well acbeneficial effect on urinary glucose excretion in patients with rise in BG seen in our study counted for by the sucrose and monosaccharide contents of the product. FA is transported more slowly in the upper intestine when TABLES compared with fructose or the disaccharide lactulose. upThe Orocecal transit times (OCTTs) and hydrogen responses after 50 g per intestinal transit of FA may be slower than the transit of wheat starch, 10 g fructan, and the combined meal in eight subjects equivalent amounts of nonabsorbable monoor disaccharides
Hydrogen OCTT mm SO g wheat starch l0gfructan 50 g wheat response
TotaIAUC
ppm.min.102
PeakH2 TABLE
ppm Blood
6
glucose (BC) responses after a starch
meal with
and
without
addition
425(375-475)t 175 (155-270) 29(6-60) 109 (77-138) 24(11-37) 51 (38-64)
of 10 g fructan SPeak
AUC
L . mm
starch
+ lOg fructan
mmo//L
SOgwheatstarch 50 g wheat starch + lOgfructan
134(37-191) 90 (32-155)
300(300-325) interquartile
141(70-178)
53(48-60)
Medians;
t From
Medians;
interquartile
INTESTINAL
HANDLING
OF
ABG
mmol/1 has
non-insulin-dependent
1.
effect (7, 8, 22, 29-34). Because of the lack of FA absorption, the possible metabolic disadvantages of fructose ingestion (35) could be avoided. Larger doses of FA are necessary, however, to achieve equal sweetness compared with fructose. Malabsorption of fructose is frequent (6, 36) and it may produce abdominal symptoms in susceptible subjects (37). Chronic intake ofNeosugar may give rise to gas problems (24). Only negligible symptoms were provoked by FA in the present study, but the abdominal symptoms and metabolic consequences of chronic
ingestion of FA with longer chain lengths need further study.
FA may parallel effects of nonabsorbable starch components and certain types ofdietary fiber 13, 38, 39). ( FA does not seem to influence wheat-starch absorption, as we described previously for different types ofdietary fiber (13). Mm. FA may, however, modify the BG responses to ingested starch. BG values seemed to return to baseline earlier with a more attenuated hypoglycemic phase when FA was given. This was not attributable to differences in IRI or ICP responses and was present well before the meal reached the cecum. An effect ofFA on gastric emptying is possible and may also account for some of the shortening ofOCIT. The effect may depend on the dietary -1 formulation, and diabetics may respond differently from nondiabetics. Complete malabsorption of FA reduces the caloric FIG 2. Median incremental blood glucose values(BG) in eight subvalue of the product compared with that of well-absorbed jects after ingestion of 10 g fructan (0), 50 g wheat starch and the (#{149}), combined meal (A) monitored for 180 mm. carbohydrates such as sucrose. This may be relevant for
\N:i
obese that
Metheocarbohy-
because
osmotic transit
ofthe
effect. times
larger
The ofthe
molecular
OCTTs FA studied
size ofFA
of Neosugar by us (24).
and
seem
a less prominent
shorter than
retical
after
complete
The apparent
colonic
bacterial
fermentation
energy of
processes
Neosugar in
metabolizable
to be two-thirds of the combustion energy Malabsorbed FA is vigorously fermented by the colonic rats was estimated or less (44). flora. Compared with equivalent amounts oflactulose, the hybenefits ofFA as a therapy for diabetes (22, 26) drogen production following FA was ofsimilar magnitude even The potential confirmation. Because of its sweetening properties (although the contents ofactual fructose polymers in the product need though only one-third as sweet as sucrose), FA may be used as was not 100%. The more prolonged OCITs of FA compared additive in sweets, soft drinks, and food. For diabetics, with lactulose is in itself unlikely to influence FA-induced hy-a dietary increase palatability and compliance with insignifidrogen production (1 5). A larger colonic bacterial hydrogen FA could production after ingestion of the FA substrate compared with cant glycemic effect and possibly nutritionally advantageous The properties of FA products with even longer chain lactulose therefore cannot be excluded. Complete or nearly effects. extracted from other plant species need to be investicomplete fermentation of cereal FA and inulin was demon- lengths gated. Long-term studies of FA ingestion are needed to further strated in rat hindgut (28).
TABLE
Insulin
7 (IRI)
and
C-peptide
(ICP)
responses
after
SO g wheat
starch
and
a combined
meal
of 10 g fructan
and
50 g wheat
starch
in eight
subjects
IRI SPeak
pmo//L SOgwheat starch 50 g wheat starch + lOgfructan Medians; 193 (90-274) 201(73-263) ranges
ICP Positive
pmol.
IRI
pmo/.
Net AUC
L . min
AUC
L. mipr
SPeak
pmo//L 800(333-1
ICP
pmo/. 133)
Net AUC
L . mm . iO
Positive
pmo/.
AUC
.
L . mm
io
867(400-967)
interquartile
680
TABLE 8 Abdominal symptom scores recorded during
RUMESSEN
El
AL
Symptoms
Substrate
Median 6
2 2.5 7.5 3.5 0 5.5 as the sum
score
Interquartile 1-9
1-11 0-11 4-14 1-9 0-5 2-9
range
None 2
2 4 0 2 5 1
Mild 6
6 3 8 6 3 7 during the
Moderate 0
0 1 0 0 0 0 12-h tests.
Severe 0
0 0 0 0 0 0
10 g lactulose
SOgfructan l0gfructan 2Ogfructan 20 g fructose 50 g wheat starch SOgwheatstarch+
Total
scores
ofall
registered
symptoms
in well-defined
intervals
assess the physiological and nutritional effects jects and in patients with diabetes mellitus.
in healthy
sub-
ofbreath hydrogen (H2) analysis Evaluation ofa H2-monitor and interpretation ofthe H2 breath test. Scand J Clin Invest Lab
aspects l987;47:555-60. E. Influence of oro-
We are grateful to G Bischoff, E Borresen, B Fallesen, 0 Hansen, L Krogh, and I Staack for technical assistance. Peters and JO Rumessen for comments on the manuscript.
LM Hansen, 15. Rumessen ii, Hamberg 0, Gudmand-Heyer We thank S caecal transit time on hydrogen excretion absorption. Gut 1989;3O:8l 1-4.
16.
after carbohydrate
mal-
References
1. Archbold HK. Fructosans in the monocotyledons.
The mechanisms offructosan
A review.
Rumessen JJ, Hamberg 0, Gudmand-Hyer E. Interval sampling ofend-expiratory hydrogen (H,) concentrations to quantitate carbohydrate malabsorption by means of lactulose standards. Gut l990;3l:37-42. New 17. Beutler H-O. Inulin. In: Bergmeyer HU, ed. Methods of enzymatic
analysis. 3rd ed. Basel: VerlagChemie, Weinheim, 1984:41-5. 18. Rahe AJ. Tables ofcritical values for the Pratt matched pair signed in higher plants as exemplified He/ianihus in tuberosus. New Phyrank statistic. J Am StatAssoc l974;69:368-73. tol l968;67:5l7-3l. NJ Jr. Nonparametric statistics for the behav3. Shiomi N, Izawa M. Purification and characterization of sucrose: 19. Siegel S, Castellan ioral sciences. 2nd ed. New York: McGraw-Hill, 1988. sucrose- 1-frutosyltransferase from the roots of asparagus Aspara( 20. Darbyshire B, Henry Ri. Differences in fructan content and syngus officinalis L.). Agric Biol Chem I980;44:603-l4. thesis in some allium species. New Phytol 198 l;87:249-56. 4. Oku T, Tokunaga T, Hosoya N. Nondigestibility of a new sweet21. Bacon JSD, Edelman J. The carbohydrates ofthe Jerusalem artiener, Neosugar, in therat. J Nutr 1984; 1 14:1574-81. choke and other Compositae. Biochem J 1951;48:l 14-26. 5. Ravich Wi, Bayless TM, Thomas M. Fructose: incomplete intesti22. K#{252}lz On the pathology and therapy E. of diabetes mellitus. Marnal absorption in humans. Gastroenterology l983;84:26-9. burg: NG Elwerts Verlag, 1874 (in German). 6. Rumessen ii, Gudmand-H#{248}yer E. Absorption capacity of fructose
metabolism in healthy adults. Comparison with sucrose and its constituent monosaccharides. Gut 1986;27:1161-8. 7. Crapo PA, Koltermann 00, Olefsky JM. Effects oforal fructose normal, diabetic and impaired glucose tolerance subjects. Diabetes Care 1980;3:575-82. 23. in 24. Tokunaga T, Oku T, Hosoya N. Influence ofchronic intake of new
sweetener
fructooligosacchaiide (Neosugar) on growth and gastroinfunction ofthe rat. Nutr Sd Vitaminol J l986;32:l 1 1-2 1.
T, Levitt MD. Gaseous response to ingestion fructo-oligosaccharide sweetener. Am J Clin Nutr of a
8. Crapo PA, Koltermann 00, of two-week fructose feeding l986;9:1 11-9. 9. Jenkins DJA, Leeds AR, Gassull
Decrease in postprandial
l987;46:6
1-5.
of
by26.
Hidaka H, Eida T, Takizawa T, Tokunaga T, Tashiro Y. Effects fructooligosaccharides on intestinal flora and human health. Bifidobact Micro l986;5:37-50. Yamashita K, Kawai K, Itakura M. Effects of fructo-oligosaccha-
guarand
10.
pectin.
Ann
Intern
Med
TMS, tolerance:
fibres,
fibre
Br Med 27.
and serum
lipids
in diabetic
subjects.
fructans:
Nutr
in
ofviscosity.
Nilsson U, Oste R, J#{228}gerstadM, Birkhed D. Cereal vitro and in vivo studies on availability rats and humans. in
J Nutr
inulin in
1 1. Hagander
B, Asp
N-G,
I,
28. 29.
1988; 118:132S-30.
Nilsson U, Bj#{246}rck Availability I. of cereal fructans and
of pan-
tract.
NH.
J Nutr and
1988; 1 18:1482-6.
of carbohydrate metabolism
creatic and gastrointestinal hormones. Diabetes Res 1986; 3:91-6. 12. Hamberg 0, Rumessen ii, Gudmand-H#{248}yer E. Reduced blood glucose response to pea fiber. Comparisons with sugar-beet fiber and wheat bran. Am J Clin Nutr 1989; 50:324-8. 13. Hamberg 0, Rumessen ii, Gudmand-Hyer E. Inhibition starch absorption by dietary fibre. A comparative study of wheat bran, sugar-beet fibre, and pea fibre. Scand J Gastroenterol l989;24: 103-9.
14. Rumessen ii, Kokholm G, Gudmand-H#{248}yer E. Methodological
A comparison
after
sucrose,
sorbitol,
Care
fructose
3:582-5.
meals
in normal
and diabetic
subjects.
Diabetes
1980;
30. Crapo PA, Scarlett JA, Koltermann 00. Comparison ofthe metsbolic responses to fructose and sucrose sweetened foods. Am J Clin of Nutr 1982; 36:256-61. 31. Akg#{252}n Ertel NH. The effects ofsucrose, 5, fructose and high-fructose corn syrup meals on plasma glucose and insulin in noninsulindependent diabetic subjects. Diabetes Care l985;8:279-83.
INTESTINAL 32. Bantle dietary JP, Dawn C, Lame RD. Thomas JW. fructose and sucrose in types I and Metabolic II diabetic Metabolic
OF
FRUCTANS
681
JAMA
33. Osei K,
l986;256:324l-6.
Falko J, Bossetti BM, Holland GC.
Shetty PS, Kurpad AV. Increasing starch intake in the human diet increases fecal bulking. Am J Clin Nutr l986;43:2l0-2. 39. Levitt MD, Hirsch P. Fetzer CA, Sheahan M, Levine AS. H2 excreof complex carbohydrates. Gastroenterology of tion after ingestion
fructose
tory 55. 34.
as a natural sweetener in the physiologic meals of ambulaobese patients with type II diabetes. Am J Med l987;83:249C, Rizkall SW, Halfon P, et al. Lack ofdetectable ofdailyfructose ingestion Care 1988; 1 1:546-50.
Grigoresco
terious
35. 36.
effects
on metabolic
patients.
control
Diabetes
2 mo in NIDDM
Sestoft L. Fructose and the dietary therapy of diabetes Diabetologia 1979; 17:1-3. Rumessen JJ, Gudmand-Hyer E. Malabsorption of fructose-sorbitol mixtures. Interactions causing abdominal distress. Scand Gastroenterol l987;22:43 1-6. Rumessen absorption
1987;92:383-9. Thi#{233}baud Jacot E, Schwitz D, H, Spengler M, Felber JP. Comparative study of isomalt and sucrose by means ofcontinuous indirect calorimetry. Metabolism l984;33:808-l 3. dde41. Van Es AJH, De Groot L, Vogt JE. Energy balances ofeight volunfor teers fed on diets supplemented with either lactitol or saccharose. BrJ Nutr I986;56:545-54. mellitus. 42. Grimble GH, Patil DH, Silk DBA. Assimilation of Lactitol, an 40.
unabsorbed disaccharide in the normal human colon. Gut
37.
bitol
and
l988;29: 1666-7 1. J 43. McBurney MI, Thompson LU, Jenkins DJA. Colonic fermentstion of some breads andits implication for energy availability in JJ, Gudmand-H#{248}yer E. Functional bowel disease: malman. NutrRes l987;7:l229-4l. and abdominal distress after ingestion of fructose, sor- 44. Tokunaga T, Oku T, Hosoya N. Utilization and excretion ofa new fructose-sorbitol mixtures. Gastroenterology I988;95: sweetener, fructooligosaccharide (Neosugar), rats. J Nutr in
1989; 1 19:553-9.
694-700.