A Survey of Commercially Available

Isomaltooligosaccharide-Based Food Ingredients

Lee R. Madsen II, Sarah Stanley, Peter Swann, and Jack Oswald

Abstract: Isomaltooligosaccharides (IMOs) are included in many commercially available food products including pro-
tein/fiber bars, shakes, and other dietary supplements. Marketed as “high fiber,” “prebiotic soluble fiber,” and/or as a
“low-calorie, low glycemic sweetener,” IMO may be present in significant amounts, for example, more than 15 g/item or
serving. Herein, high-pressure anion exchange chromatography with pulsed amperometric detection and high-pressure
liquid chromatography with differential refractive index detection are used to compare 7 commercially available IMO-
containing bulk food ingredients. The ingredients are typical of those produced either (a) via bacterial fermentation
(“fermented” IMO or MIMO) of sucrose in the presence of a maltose acceptor mediated by a glucosyltransferase enzyme
(dextransucrase), or (b) via transglycosylation of hydrolyzed starch with α-glucosidase (“industrial” IMO). Analysis of the
results with respect to digestibility suggests that the potential glycemic impact of the ingredients and products containing

Food Chemistry
“industrial” IMO may be inconsistent with the product labeling and/or certificates of analysis with respect to overall
fiber content, prebiotic fiber content, and glycemic response and are thus inappropriate for diabetic patients and those on
low-carbohydrate (for example, ketogenic) diets.

Keywords: carbohydrates diabetes, dietary fiber, oligosaccharides, prebiotics

Practical Application: This analysis comparing 7 commercially available isomaltooligosaccharide-based food ingredients
demonstrates that most of these products are, by way of definition, and particularly with respect to content of “oligosac-
charides” and “dietary fiber,” mislabeled. This is significant because claims, such as “low glycemic,” “zero calorie,” and
the like, are certainly false, and may pose a health hazard to certain populations (diabetic patients and epileptic patients
on ketogenic diets, in particular) while misleading others (those on low carbohydrate diets). We conclude that labeling
requirements should be reconsidered for products of this type.

Introduction with an α-glucosidase enzyme, typically from Aspergillus niger

Oligosaccharides are small polysaccharides ranging in degree spp (Kwon and others 2011). This method yields a mixture of
of polymerization (DP) from 3 to 9 or “oligosaccharides DP > type 1 and type 2 IMOs. This is the method by which “indus-
2” (Jones 2014), or as “carbohydrates consisting of between 2 trial” IMOs (6 exemplars) are produced. Resistant maltodextrin
disaccharides and 10 monosaccharides connected by either an (RMDx) should also be evident as a side product resulting from the
alpha- or beta-glycosidic link” (National Library of Medicine hydrolysis of starch via α-amylase, for example, limit dextrins. Be-
2011). Isomaltooligosaccharides (IMOs) are glucooligosaccharide cause this approach can yield a significant amount of glucose, this
(GlcOS) homooligomers, and can be subdivided into 3 types. Type is typically removed via secondary fermentation with yeast (Zhong
1, known as “6-O-α-isomaltooligosyl-D-maltose,” “IMOMs,” and others 2005). Method B involves the fermentation of sucrose
“maltosylisomaltooligosaccharides,“ or “MIMOs,” are linear α- with a maltose acceptor in the presence of a bacterial organism
1,6 oligoglucan chains terminated to glucose via an α-1,4 glyco- capable of expressing the dextransucrase (1,6-α-D-glucan 6-α-D-
sidic linkage. Panose is the representative trisaccharide, and ho- glucosyltransferase [GT]) enzyme (Day and Chung 2004). This
mologs are referred to as PAN-type oligosaccharides, here. Type method yields IMOs of type 1 (very small amounts), but types 2
2, also known as oligodextrans, are linear α-1,6 oligoglucans that and/or 3 predominate, and broths may contain, depending on the
include isomaltotriose (IMT) as the representative trisaccharide, organism employed, a significant quantity of fructose and/or man-
and are, henceforward, referred to as IMT-type oligosaccharides. nitol (depends on the metabolic profile). The branching pattern,
Type 3 are linear chains of either the type 1 or type 2, but are also or lack thereof, of the oligomeric product is likewise dependent
branched to glucose via α-1,2, α-1,3, or α-1,4 linkages. These on the organism employed, and can be sufficiently specific as to
side chains may extend beyond a single residue, usually via α-1,6 facilitate speciation. Commercial products manufactured this way
linkages, or the novel centose-type (Shi and others 2016), origi- are usually purified to remove color bodies and acidic (lactic, acetic
nally proposed by Goffin and others (2010). acids, and so on) metabolites that can contribute to off-flavors.
The production of IMOs studied herein is accomplished via IMOs produced using the methods A and B (linear) are currently
1 of 2 methods. Method A involves the transglycosylation (TG) available on the commercial market in the United States.
of saccharified vegetable starch (typically corn or cassava/tapioca)
Materials and Methods
JFDS-2016-1643 Submitted 10/5/2016, Accepted 12/19/2016. Authors are Several bulk IMO food ingredients were acquired and used
with ISOThrive, LLC, 9385 Discovery Blvd. Suite 133, Manassas, VA 20109, as-is. The materials included 1 “fermented” IMO and 6 “indus-
U.S.A. Direct inquiries to author Madsen II (E-mail: lmadsen@ISOThrive.com). trial” IMOs. ISOThriveTM Prebiotic Nectar (ISOThrive, LLC.,
Healdsburg, Calif. U.S.A. and Manassas, Va., U.S.A.), the only

C 2017 Institute of Food Technologists

 R

doi: 10.1111/1750-3841.13623 Vol. 82, Nr. 2, 2017 r Journal of Food Science 401
Further reproduction without permission is prohibited
 A survey of Imo-based food ingredients . . .

“fermented” IMO, is made up of a mixture of mainly linear onex ICS-5000+, details below). First, the amount of the primary
type 2/PAN-type oligosaccharides (DP 3 to 8, 85% PAN-type diluted material (typically 15% to 25% RDS) needed was calcu-
IMO) and a very small fraction of type 1/IMT-type (DP 3 to lated as 500/brix. The calculated amount was diluted by mass to
4, <1% IMT-type IMO) prepared via fermentation of a sucrose 1 g in a 1.5 mL Agilent-type autosampler vial (target 0.5% w/w
donor in the presence of a maltose acceptor via dextransucrase RDS, HPLC-RID). Second, 25 μL of the 0.5% solution was
(Leuconostoc citreum NRRL B-742) as described by Madsen and transferred by pipette to a tared 1.5 mL Agilent-type autosam-
others (2015). The 6 “industrial” IMOs manufactured via TG of pler vial. The mass was recorded. To this was added 25 μL of
starch hydrolysate and typically followed by a yeast fermentation internal standard (IS, L-arabinose, Sigma, 99+%, 500 μg/g), and
to remove residual glucose include (product name and description the total mass was recorded. To the latter was added DI water
provided by manufacturer): to a total of 1 g (target 15 to 25 μg/g per analyte, HPAEC-
PAD).
r IMO-900 powder (Baolingbao Biotechnology, Co. LTD., Instrumental analysis: Bulk DP 1 to 3, mannitol, glycerol,
Yucheng Shandong, China, >90% IMO). and organic acids were quantified via HPLC-RID (Agilent 1100,
r AdvantaFiber 90 powder, non-GMO soluble fiber sweetener 20 μL on-column, BioRad Aminex HPX-87H 7 × 300 mm @
(Top Health Ingredients, Inc. (Edmonton Canada) 2014; coun- 65 °C, isocratic, 0.008N H2 SO4 , 0.6 mL/min, runtime 25 min,
try of manufacture: China, 90% IMO). RID @ 45 °C) using external standards (0.2% w/w per analyte,
r
Food Chemistry

Wako IMOs (Wako Pure Chemical Industries, LTD., Osaka, see Figure 3). The sugar alcohols, mono-, di-, and oligosaccha-
Japan). rides, were quantified/confirmed via HPAEC-PAD (ThermoDi-
r FiberYumTM Prebiotic sweetener, sugar free (Raw Indulgence onex ICS-5000+, all PEEK, 10 μL on column, Carbopac PA-
LTD., Hawthorne, N.Y., U.S.A.; packed in the United States, 100 (4 × 250 mm) and guard @ 35 °C, NaOH > pH 12.50,
100% fiber/dry solids based on 5 g total carbohydrate = 5 g 100 mMol, 5 min isocratic then acetate gradient to 250 mM, run-
fiber, from tapioca starch). time 25 min, including 2 min equilibration prior to injection, PAD
r VitaFiberTM prebiotic fiber sweetener, powder from corn at 25 °C with standard gold electrode compared with Ag/AgCl
(China), and syrup from tapioca (Indonesia, Bioneutra Global running the standard Au quad waveform). All compounds listed,
Corp., Edmonton, 96% carbohydrate, 91% fiber). except for PAN-type IMOs ࣙ DP 3 were quantified using L-
arabinose as an IS relative to bona-fide compounds (15 ppmw
Sample preparation: Each IMO exemplar was diluted by a each). It is important to first analyze any new matrix without IS
factor of 4 with deionized water (DI water, 18 M, house system, to make sure that the sample does not contain L-arabinose. PAN-
Hydro Service & Supplies, Inc., Gaithersburg, MD). The refractive type IMOs were quantified by approximation using the relative
dry solids (RDSs) were measured using an Atago PAL1 critical- response factors (RRFs) for the corresponding DP in a known
angle refractometer, as g RDS/100 g material (% w/w) relative to maltodextrin ladder (DP 3 to 7: Supelco Oligosaccharides kit, DP
sucrose (Sigma S7903). Serial dilutions were prepared therefrom, 8 to 10: Elicityl Oligotech).
in order, for analysis via high-pressure liquid chromatography with Electrode surface effects dictate that RRF compared with ex-
differential refractive index detection (HPLC-RID; Agilent 1100, ternal standards can drift over time, so an IS was used. However,
details below) and high-pressure anion exchange chromatography because RRFs are not necessarily the same relative to L-arabinose
with pulsed amperometric detection (HPAEC-PAD; ThermoDi- for a homologous series of oligosaccharides with different linkage

Figure 1–Overlaid (offset by 60 nC)

HPAEC-PAD 
R
chromatograms of 1.
ISOThrive , 2. IMO-900, 3.
AdvantaFiber 90P, 4. Wako IMO, 5.
FiberYum, 6. VitaFiber syrup, and 7,
VitaFiber powder. A: D-mannitol, B:
L-arabinose (IS), C: unknown, possibly
D-mannose, D: D-glucose, E:
D-leucrose, F: isomaltose, G: sucrose,
H: isomaltotriose, I: unknown
disaccharides, J: isomaltotetraose, K:
maltose, L: D-panose, M-Q: PAN-type
IMO DP 4 to 8, and R: resistant
maltodextrins in clusters of increasing
molecular weight.

402 Journal of Food Science r Vol. 82, Nr. 2, 2017

A survey of Imo-based food ingredients . . .

Table 1–HPAEC-PAD results for 7 isomaltooligosaccharide bulk food ingredients. Note that the sucrose observed in all but
ISOThrive appears to be a coeluting unknown. MWD is Mw or mass-average molecular weight of DP 2 to 10 (to avoid skew) and
IMO content the sum of components from DP 3 to 10.

ISOThrive IMO-900 AdvantaFiber Wako FiberYum VitaFiber VitaFiber

Compound, %/brix nectar powder 90P IMO prebiotic IMO syrup IMO powder
Exemplar # 1 2 3 4 5 6 7
brix: 63.06 95.18 96.24 96.98 75.58 75.93 97.49
Erythritol 0.15 0.00 0.00 0.00 0.00 0.00 0.00
Mannitol 6.80 0.00 0.00 0.00 0.05 0.06 0.09
Glycerol 0.40 0.00 0.32 0.00 1.79 1.71 1.86
Glucose 0.90 1.58 1.81 5.32 1.16 0.77 1.36
Fructose 0.12 0.00 0.00 0.09 0.00 0.00 0.01
Leucrose 2.07 0.00 0.00 0.00 0.00 0.00 0.00
Isomaltose 0.13 32.14 29.83 33.62 24.13 23.85 21.84
Sucrose 1.15 10.06 9.73 12.11 6.71 8.20 7.02
Maltose 3.80 3.36 3.27 6.08 5.27 5.15 5.28
Isomaltotriose 0.40 13.98 12.12 14.69 10.41 10.08 8.32
Isomaltotetraose 0.51 9.47 8.53 7.91 5.26 5.21 4.13

Food Chemistry
Maltotriose 3.17 1.18 2.07 1.79 1.41 2.92 0.00
Maltotetraose 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Panose (MIMO-DP3) 11.18 3.76 3.41 7.30 9.49 10.12 11.75
MIMO-DP4 22.48 6.55 5.95 8.24 7.01 7.24 6.76
MIMO-DP5 23.77 2.99 2.65 2.76 2.19 2.93 1.79
MIMO-DP6 14.48 0.77 0.71 0.40 0.38 0.54 0.00
MIMO-DP7 4.58 0.00 0.00 0.00 0.00 0.00 0.00
MIMO-DP8 1.73 0.00 0.00 0.00 0.00 0.00 0.00
MIMO-DP9 0.61 0.00 0.00 0.00 0.00 0.00 0.00
IMT IMO 0.90 23.45 20.66 22.60 15.66 15.29 12.46
PAN IMO 78.83 13.30 12.01 18.30 18.70 20.29 20.30
Total IMO 79.74 36.75 32.67 40.91 34.36 35.58 32.76
Total 98.42 85.07 79.70 99.91 75.27 78.23 70.23
Balance: 1.58 14.93 20.30 0.09 24.73 21.77 29.77
MWD, Da: 779.06 630.18 606.27 562.07 572.46 580.81 536.04

types (despite identical molecular weight and number of reducing 0.71 to 0.75 for the PAN/IMT series DP 3 to 10. We found that
end groups (Goffin and others 2009), an RRF correction factor standards DP 8 to 10 were inadequately fractionated and did not
was based on the peak areas normalized over mass between the reflect the indicated purities (likely done via area % rather than
corresponding DP 3 types (panose in this case, and the calibrating mass), so care must be taken to qualify the individual standards
compound, maltotriose). Typically, this factor is in the range of prior to use.

Figure 2–Overlaid HPAEC-PAD

chromatograms of 1. ISOThrive, 2.
IMO-900, 3. AdvantaFiber 90, 4. Wako
IMO, 5. FiberYum, 6. VitaFiber Syrup, and
7. VitaFiber powder; A-G indicate
resistant maltodextrins in clusters of
increasing molecular weight.

Vol. 82, Nr. 2, 2017 r Journal of Food Science 403

 A survey of Imo-based food ingredients . . .

Refractive components, given in g refractive material/100 g did not contain a significant amount of either isomaltose or IMT-
sample (brix), were measured using an Atago PAL1 refractometer. type oligosaccharides, 0.13%/solids and 0.90%/solids, respectively.
When compared with true dry solids (oven), calculation over brix The IMOs were almost exclusively PAN-type comprising 79% of
(for homologues equivalent to the calibrated standards DP 3 to the solids (80% total IMO). Exemplar 1 was further differenti-
10 only) yields excellent closure on balance of mass, for example, ated by the presence of mannitol (6.8%/solids), a byproduct of the
97% to 103%, and total % w/w is usually within ± 2.5%. fermentation process, and leucrose (5-O-α-D-glucopyranosyl-D-
Upon examination of the results, the authors decided to per- fructose), a byproduct specific to IMOs made via donor–acceptor
form a brief (N = 2) set of experiments comparing the effects chemistry mediated by dextransucrase(s).
of glucose and a representative “industrial IMO” (FiberYum) on In Figure 2, exemplars 2 to 3 (IMO-900, AdvantaFiber) and 5
their blood sugar. Blood sugar was measured using a contour next to 7 (FiberYum, VitaFiber syrup and powder) contain a significant
EZ (Bayer, Ascensia Diabetes Care, Parsippany, NJ) blood glucose quantity of unidentified material evident as clusters of increasing
meter and contour next test strips (#7312). After a 12 h fast, base- molecular weight. This is RMDx. Determined as the difference of
line blood glucose was measured for 1 h in 15 min interval. At the the whole from 100%, exemplars 2 and 3 and 5 to 7 contain 22.3 ±
1 h mark, a 70 g aqueous solution containing either 20.00 ± 0.001 5.5% leading to mass closure of 77.7 ± 5.5%. Interestingly, exem-
g glucose (Fisher D14) or 26.50 ± 0.05 g FiberYum (Raw Indul- plar 4 contained virtually no RMDx and the mass balance closed
gence LTD.) IMO syrup (equivalent to 20.02 ± 0.05 g glucan) to 99.9%, indicating that the missing mass for the other exemplars
Food Chemistry

was rapidly ingested and chased with 50 mL tap water. Blood sugar (2 and 3, 5 to 7) was likely the result of inadequate quantitation
was measured in triplicate every 15 min. Testing was concluded of the resistant components. We suppose that the saccharifica-
once blood sugar either returned to baseline or it became evident tion was more complete, perhaps via use of isoamylase/pullalanase
that it would not. The results (N = 2) were averaged, integrated, debranching enzymes prior to treatment with amylase.
and expressed as glycemic index (GI) where glucose = 100. Within the scope of this work (types 2a and 2b are made from
corn and type 1a uses high-purity maltose), TG type 2c can be fur-
Results and Discussion ther differentiated into products made from corn/maize (TG2cm)
It becomes clear (Figure 1, Table 1) that commercially available and those made from tapioca (TG2ct). The exemplars made from
IMO can be sorted into 1 of 4 discrete types. TG type 2a consists tapioca typically demonstrated a higher molecular weight (from
predominantly of DP 4 and includes exemplars 2 (IMO 900) and 3 DP 6), relative to those made from corn, of about 10 Da. While
(AdvantaFiberTM ). TG type 2b features an even bias on PAN-type the difference is too small, in terms of mass-average molecular
DPs 3 and 4, and includes exemplar 4 (Wako). TG type 2c fea- weight, to decisively differentiate them, the RMDx profiles are
tures a heavy bias on DP 3 (panose), and includes exemplars 5 to 7 both discrete and reproducible (Figure 2).
(FiberYum, VitaFiber syrup and powder). All TG types contained It can be seen from the fingerprint of the RMDx fractions that
a significant amount of isomaltose, for example, 28 ± 5%/solids exemplar 1 (ISOThrive) is clearly discrete (it has no RMDx be-
and a relatively even distribution of IMT- and PAN-type oligosac- cause is not made from starch), exemplars 2 and 3 (IMO-900,
charides, 18.4 ± 4.5%/solids and 17.2 ± 3.6%/solids, respectively AdvantaFiber) are similar (TG2a), exemplar 4 (Wako) is unique
(35.5 ± 3.1% total IMO). GT type 1a, or exemplar 1, appears (TG2b) by virtue of containing little, if any RMDx while belong-
distinct from exemplars 2 to 7 (which are quite similar upon in- ing to the “industrial IMO” group, exemplars 5 and 6 (FiberYum,
spection of gross composition), featuring a predominance of DP VitaFiber syrup) are essentially identical (TG2ct) and discrete from
5, and the presence of oligosaccharides up to DP 9. Exemplar 1 7 (VitaFiber powder, TG2cm). In Figure 2, groups B, D, and E are

Figure 3–HPLC-RID chromatograms of 1.

Standard mixture 0.2% w/w per analyte,
and 0.5% w/w each of 2. VitaFiber
syrup, and 3. FiberYum syrup; where A:
maltotriose, B: maltose, C: leucrose, D:
maltitol, E: glucose, F: fructose, G:
mannitol, H: sorbitol, I: xylitol, J:
erythritol, K: lactic acid, L: glycerol, M:
formic acid, N: acetic acid, O: propionic
acid, P: iso-butyric acid, Q: n-butyric acid,
R: ethanol, S: DP >3, T: isomaltotriose,
and U: isomaltose.

404 Journal of Food Science r Vol. 82, Nr. 2, 2017

A survey of Imo-based food ingredients . . .

Table 2–Comparison of available analytical data with this work.

ISOThrive IMO-900 AdvantaFiber Wako FiberYum VitaFiber VitaFIber

Constituent group nectar powder 90P IMO Prebiotic IMO powder IMO syrup
Solids, % 63-67i , 63.1 >95b , 95.2 95.3h , 96.2 97.0 75.6 96f , 97.5 75g , 75.9
Glucose, % <1.5i , 0.9 <5a ,4b ; 3.8c , 1.6 1.6h , 1.8 5.3 1.2 <5f , 1.4 <5g , 0.8
Fructose, % <0.2i , 0.1 n/a 0.0 0.1 0.0 0.0 0.0
Isomaltose %: 0.1 22.8c , 32.1 29.8 33.6 24.1 21.8 10-15a , 24.0
Isomaltose + panose + isomaltose 11.7 >45b,d , 49.9 45.0 55.6 44.0 42.0 44
DP 3 IMT, % 0.4 14.0 12.1 14.7 10.4 8.3 10.1
DP 3 PAN, % 11.2 3.8 3.4 7.3 9.5 11.8 10.1
DP 3 total, % 11.6 16.7c , 17.8 15.5 22.0 19.9 20.1 20-25a , 20.2
DP 4 IMT, % 0.5 9.5 8.5 7.9 5.3 4.1 5.2
DP 4 PAN, % 22.5 6.6 6.0 8.2 9.5 6.8 7.2
DP 4 total, % 23.0 17.7c , 16.1 14.5 16.2 14.8 10.9 15a , 12.4
DP 5, % 23.8 7.2c , 3.0 2.6 2.8 2.2 1.8 7.9a , 2.9
DP 6, % 14.5 1.7c , 0.8 0.71 0.4 0.4 0.0 4-5a , 0.5
DP 7, % 4.6 0.0 0.0 0.0 0.0 0.0 0.0
DP 8, % 1.7 0.0 0.0 0.0 0.0 0.0 0.0

Food Chemistry
DP 9, % 0.6 0.0 0.0 0.0 0.0 0.0 0.0
RMDx† 1.6 14.9 20.3 0.1 24.7 29.8 21.8
Total IMO DP 3 to 9 >75i , 79.7 >90a,b , 37.5 90.8h , 33.4 n/a∗ ,41.3 100e‡ , 34.7 91f , 32.8 93.3g , 36.1
Total IMO DP 2 to 9 79.8 69.6 63.2 74.9 58.8 54.6 60.1
Total IMO DP 2 to 9 + RMDx 81.4 84.5 83.5 75.0 83.5 84.4 81.9
TOTAL, all compounds 98.4 85.1 79.7 99.9 75.3 70.2 78.2
Mw, Da 779 630 606 562 572 536 581
DP 4.72 3.80 3.65 3.37 3.44 3.21 3.49
∗
no available data, † RMDx approximated by difference based on total of all compounds by mass, ‡ derived from given information, for example, carbohydrate 5g and fiber 5g from
product label.
a
Bioneutra Table 3 GRAS document (Zhu, March 12, 2005)22 ; b Baolingbao Biology Co., LTD23 ; c Kaneko and others18 (1992); d Dancheng Caixin Sugar Industry Co. LTD,
specification of isomaltooligosaccharide (IMO)24 ; e Packed in the United States for Raw Indulgence LTD. FiberYum, Nutritional Facts from product container; f Bioneutra
Nutritional Data (May 22, 2013) VitaFiber-IMO powder, declared as “VitaSugar Isomalto-oligosaccharide food ingredient,” on bill of lading (non-GMO corn25 , China); g syrup
(tapioca26 , Indonesia); j Bioneutra Table 3 GRAS document (Zhu, March 12, 2005)22 ; h Top Health Ingredients, Inc. AdvantaFiber 90P, Certificate of Analysis #1090-14071631,
July 17, 2014, non-GMO corn27 , manufactured by Baolingbao Biology Co. Ltd, China; i ISOThrive, LLC, ISOThrive Prebiotic Nectar, GRAS internal document available upon
request with NDA.

Table 3–Comparison of oligosaccharide descriptions given in the the consistency in glycerol quantity suggests that these products
product profiles with this work. were manufactured using a similar enzyme preparation.
IMO powder IMO syrup Syrup solids
From product profiles
Discussion of claims and nutritional information
Nondigestible oligos∗ 55 42 55 Because MIMO (as in exemplar 1, ISOThrive) only became
Digestion resistant oligos∗∗ 16 13 17 available for purchase in 2016, most of the prior literature regard-
Digestible oligos† 20 15 20 ing digestibility/fermentability of IMO dealt with products similar
Dietary fiber (TDF) 91 70 92 to exemplars 2 to 7. Of these, IMO-900 (2) and Wako IMO (4)
This work
Nondigestible oligos∗ 12.7 12.1 15.9 claim the greatest coverage. With respect to popularity and online
Nondigestible + RMDx 42.5 28.6 37.7 presence, however, VitaFiber is perhaps the best known IMO-
Digestion resistant oligos∗∗ 20.1 15.3 20.2 based food ingredient and FiberYum is gaining popularity because
Digestible oligos† 21.8 18.1 23.8 it is less expensive in bulk. Bioneutra claims that “oligosaccha-
Dietary fiber (TDF) 84.4 62.1 81.7
rides are short-chained carbohydrates composed of DP 2 to 7,
or more, glucosyl units” (BioNeutra). Furthermore, they claim
that said oligosaccharides “constitute nondigestible soluble dietary
diagnostic for differentiating TG2ct from TG2cm, and may be suf- fiber” by citing AACC (March 2001) that, interestingly, defines
ficiently discrete as to be fingerprints of the specific process/place oligosaccharides as “chains with a DP between 3 and 10,” (AACC
of manufacture, and/or of the lot. That is, the signature may vary 2001). Thus, by way of proper definition, exemplars 2 to 7 do not
due to slight inconsistencies in the saccharification process. It ap- contain >90% IMO. If one takes the definition of oligosaccharide
pears that exemplars 5 and 6 may have been manufactured in the to include only DP > 2, then these products contained approxi-
same Indonesian factory (6 is known to be of Indonesian origin). mately 36% (of both IMT and PAN types via HPAEC-PAD). It is
Accordingly, the fingerprints suggest that exemplars 2 and 3 were further claimed that the composition is prebiotic (Bioneutra citing
likely manufactured either in the same plant, or using the same Rycroft and others 2001) by virtue of demonstrated bifidogenesis
IP (Baolingbao Biology Co. Ltd 2015). Additionally, in Figure 3, in-vitro (Bioneutra citing Kohmoto and others 1988) and in-vivo
it was noted that samples 5 to 7 also contained similar amounts using 20 g/d doses (containing 13.5 g IMO at purity described in
of glycerol, 1.79%/solids, 1.86%/solids, and 1.71%/solids, respec- the paper, but only 7.35 g where 4.69 g is IMT type, 2.66 g is
tively. Unlike sample 1, where some glycerol remains as a metabolic PAN type, and the MW = 619.54 Da or DP = 3.73, according to
byproduct, its presence in samples 5 to 7 likely originated with the this work) of Isomalto-900(R) composition.
enzyme cocktails (glycerol, sorbitol, and so on) (Elliot and McKay Interestingly, it was later determined by Oku and Nakamara
2002) used to perform the TG. Because most enzyme cocktails (2003) via testing of breath hydrogen and methane that a 20 g bo-
are proprietary formulations (and can vary widely in composition), lus dose of an IMO composition (Isomalto-900(R), Showa Sangyo

Vol. 82, Nr. 2, 2017 r Journal of Food Science 405

 A survey of Imo-based food ingredients . . .

Co., Ltd, Tokyo, Japan, composition verbatim from Kaneko and VitaFiber: “VitaFiberTM is a sweet natural fiber providing
others (1995) and similar to IMO type 2a; IMO-900) “hardly low calorie and soluble prebiotic fiber for human digestive
reaches the large intestine and does not produce gas.” This corrob- health . . . VitaFiber is greater than 90% soluble fiber . . . a prebi-
orated the work of Kohmoto and others (1988) who estimated, via otic . . . Maintain healthy blood sugar levels . . . low GI” (Bioneu-
studying digestibility of 13 C-labeled IMOs, that the tested mate- tra.ca/products 08-09-2016).
rial (prepared in-house and congruent with Isomalto-900 [Showa AdvantaFiber: “Adds prebiotic, soluble fiber . . . 90% IMO (Iso-
Sangyo Co., Ltd]) was largely digestible and delivered 70% to 80% malto oligosaccharide) Fiber” (http://www.tophealthingredients.
of the calorific value of an equivalent dose of maltose. Kaneko and com/products/advantafiber/ 08-09-2016).
others (1995) discovered via “Rat Jejunum Loop Method,” that FiberYum: “super low-glycemic alternative sweetener with
IMOs with increasing molecular weight became increasingly in- awesome amounts of soluble prebiotic fiber . . . Make special foods
digestible, and that “the digestibility of disaccharides in IMO (for for glucose intolerance . . . low glycemic . . . sugar free . . . 5 g sol-
example, isomaltose) is similar to those of sucrose or maltotriose.” uble fiber per serving” (a serving is 5 g; http://shop.rawrev.com/
It can thus be implied that IMOs with higher molecular weight product-p/clearancefiberyum2.5a.htm 08-09-2016).
would be less glycemic. ISOThrive: “Naturally Fermented Prebiotic Soluble Fiber . . . a
These authors also noted that in large doses, the brush-border type of complex carbohydrate you cannot digest . . . has no calories
enzymes (sucrase-isomaltase) would saturate with the preferred and does not cause blood sugar spikes” (https://www.isothrive.
Food Chemistry

substrate (isomaltose) allowing some IMO to pass. So, consid- com 08-09-2016).
ering digestibility, the 20 g dose administered by Kohmoto and It appears that all of the commercially available products are
others (1988) would have delivered, at best, a potential prebiotic making similar claims that are in line with the beneficial effects
dose of, perhaps, 1.5 to 2.2 g. Inasmuch as prebiotic specificity is of prebiotics on human health (indeed, the first use of the term
known, the dose administered would have been of relatively low “prebiotic”) described by Gibson and Roberfroid (1995).
molecular weight, according to Hu and others (2013), who tested Therefore, when comparing the analytical results from this work
VitaSugar (Bioneutra Inc., Edmonton) as a sole carbon source with the declared fiber content within the context of the given
for Lactobacillus reuteri and Bifodobacteria spp., and found that DP claims, the data suggest that not all ingredients labeled as “IMOs”
ࣘ 4 would favor Lactobacillus spp. only, and thus it is less likely are quantitatively equal. It appears that most manufacturers have
representative of a bifidogenic prebiotic. adopted a loose definition that is neither based on an accurate
It is also claimed, based on a publication in the Chinese Journal quantitative analysis nor consequence of metabolic impact.
of Clinical Nutrition (Bioneutra citing Sheng and others 2006), that In Table 2 below, values available from product certificates of
the composition exhibits low GI and is thus suitable for use by analysis (COA)/specifications or abstracted from Table 1 are com-
those on restrictive diets (diabetics (Evert et al. 2014), low-carb pared with this work, given in bold numbers.
ketogenic (Lee and Kossof 2011), and so on). However, as noted In general, for all components up to DP 4, the analytical values
above, these compositions, particularly in the large doses needed to from this work were similar to the values reported from the various
reach the colon (effectively saturating the brush border enzymes), cited sources. For DP 5 and higher, our results were always signif-
can be expected to have a GI of at least 70% of an equivalent icantly lower (factor of 2.43 ± 0.31) than indicated. Ketabi and
dose of maltose (which is completely digestible), for example, 14 others (2011) noted similar findings where the starting material
g maltose/20 g dose IMO or about 56 kcal/serving. “IMO obtained from BioNeutra Inc.” indicated that a composi-
These are claims typical of those purveyed by IMO manufac- tion of where DPs 2 to 8 were present in (DP2) 18% to 25%, (DP3)
turers/importers including: 15% to 23%, (DP4) 14% to 22%, (DP5) 8% to 10%, (DP6) 6% to

Figure 4–Effect of glucose and IMO on blood

sugar over time in 2 human subjects.
Triplicate measurements (each subject and
time) demonstrated that the test meter/strip
returned standard deviations ranging from ±
0.71 to 6.24 g/dL (average ± 3.05).

406 Journal of Food Science r Vol. 82, Nr. 2, 2017

A survey of Imo-based food ingredients . . .

Table 4–Average (N = 2) results and integrated values normalized to glucose (GI = 100).

GLUCOSE GLC eq., g: 20.00 FiberYum GLC eq., g: 20.02

averages glc g/dL stdev g/dL stdev averages glc g/dL stdev  g/dL stdev
0 89.18 1.32 0.0 0.0 0 89.84 0.58 0.0 0.0
15 108.83 1.18 19.6 2.5 15 117.00 7.07 27.2 6.5
30 132.00 2.83 42.8 1.5 30 128.67 7.07 38.8 7.7
45 134.00 6.60 44.8 5.3 46 122.17 0.24 32.3 0.8
60 112.83 10.61 23.6 9.3 74 94.00 3.77 4.2 4.4
78 92.17 4.48 3.0 5.8 89 89.17 5.42 −0.7 4.8
95 93.08 3.65 3.9 5.0 98 82.67 0.47 −7.2 1.1
integral 2470 744 integral 2074 361
GI, % glc 100 GI, % glc 83.97

8%, (DP7) 2% to 4%, and (DP8) 2% to 3%, respectively, was used glucan made from starch via heat, acid, and/or α and β-amylases,
to formulate rat-chow and found only DPs 3 to 5 and a trace of (2) other enzymes that transglycosylate starch to yield a variety
6, as indicated in this work. Thus, we conclude that the analytical of linkage types (Lee et al. 2013), or (3) known as “limit dex-

Food Chemistry
method that is being use to qualify the IMO product (and likely trins,” are relatively small, glycogen-like α-(1,4) glucans that are
the others, given similarity in the COAs) does not fractionate the α-(1,6) branched so as to resist the further action of α-amylase
individual IMO types by DP, and includes the RMDx in with the (α-limit dextrin can also be made). α-D-1,6-glucan-6-hydrolases
IMT and PAN-type oligosaccharides of similar molecular weight. (dextranase), pullulan α-1,6-glucanhydrolases (pullulanase), and so
Further, this becomes evident when we add everything up, for on, should be able to quickly reduce it into a digestible form, for
example, total IMO DP 3 to 9 = 32.9%. Adding the isomaltose example, oligo amylose (maltodextrin). Given the distribution of
gives us 60.1%. Adding RMDx (determined by difference from products that results, it appears that the RMDx found in these
100 of the sum of all accountable components given in Table 1) products is of the latter type, for example, limit dextrin. Thus,
brings the balance to reasonable closure at 84.4%. Clearly, this the RMDx described here is digestible, and is not likely to be a
is an approximation, as direct quantitation of a similar product constituent of the fiber content, soluble, or not.
(exemplar 4) containing little if any RMDx closes to 99.9%. This Therefore, we conclude that the definition of “IMOs” must not
helps us to better understand the information given in BioNeutra’s include digestible constituents including disaccharides (isomaltose)
product profiles, for example, Table 3. and α-limit dextrins. That is, the definition of “oligosaccharide”
Here, it appears evidently that the convention seems to follow must be made to include DP ࣙ3, only. This is the difference
Ketabi and others (2011) (citing Kohmoto and others 1992 and between a COA that indicates >90% soluble fiber for a product
Kaneko and others 1995) assertion that “Isomaltose is hydrolyzed that actually contains approximately 36% DP 3–10 (actually DP
by the brush border enzymes in the intestinal epithelium, the di- 3 to 5 with an average of 3.8). Inclusion of DP 3 constituents into
gestibility of IMT and panose is unclear, and longer chain oligosac- this definition requires further study, but the literature suggests that
charides are considered nondigestible.” When tested, “digestible these are digestible as well. In order to test this, the authors studied
oligos” are similar to the given values when our results for isoma- the effect, relative to equivalent glucose, of low molecular weight
ltose are used. It is important to differentiate disaccharides such as (labeled as 5 g dietary fiber per 5.5 g syrup dose, FiberYum) IMO
isomaltose from the oligosaccharide class. Biologically, there are on fasting blood glucose levels.
discrete enzymes for dealing with disaccharides (disaccharidases
such as sucrase-isomaltase, glucoamylase, cellobiase, and so on) Postprandial blood glucose levels
and for dealing with oligosaccharides, for example, oligo glucan- The authors’ results were averaged (N = 2) and the increase in
α-1,6-glucosidases, and that the former is found in abundance in measured blood glucose over time is shown in Figure 4 and the
the small intestine. “Digestion-resistant oligos” are likewise sim- tabulated results, including integrals relative to glucose (GI = 100)
ilar because sucrase-isomaltase is somewhat promiscuous in that are shown in Table 4.
it can hydrolyze a glucose unit from either IMT or panose to Although an N = 2 experiment is not statistically significant,
yield isomaltose or maltose, respectively, and that both products the standard deviations (between subjects) were relatively small and
are completely digestible, and therefore glycemic. This explains permitted some comparison. Our observations suggest that the GI
why digestibility decreases with increasing molecular weight (that of the “industrial” IMO exemplar (5), which is typical of the com-
is, kinetically limited). positions commonly encountered on the commercial market, is at
Interestingly, though, adding up all DPs ࣙ 4 gives a low result least 80% (84% observed here, see Table 4) as digestible, and hence,
compared to the given data for all exemplars 2 to 7. Adding the glycemic, as anhydrous glucose on an equivalent weight/weight
approximated value for RMDx to this amount gives us a far more basis. Based on transit time, absorption likely began in the stomach
reasonable result with respect to the given data. When summed, (15 min postprandial) and the observed amount was unlikely to
we get reasonable closures of mass, for example, 84.4%/solids and have reached the colon (absorption complete within 90 min). An-
81.7%/solids. other interesting feature that was observed in both subjects with
A review of the literature differentiated the digestibility of IMO is that the “sugar crash” was significantly larger and more
RMDx from resistant starch, a high molecular weight polyglucan prolonged than that elicited via equivalent dosage of glucose.
that is either retrograded, granular (and hence resistant to amy- Using scientifically sound definitions facilitates fair comparison
lolytic activity), or chemically modified so as to be indigestible between products allowing downstream manufacturers to make
(Sajilata and others 2006; Fuentes-Zaragoza and others 2010). improved formulations while providing a more accurate account-
RMDx can be subdivided as (1) an indigestible, highly branched ing of ingredients and nutritional information. The latter is critical

Vol. 82, Nr. 2, 2017 r Journal of Food Science 407

 A survey of Imo-based food ingredients . . .

for consumers who require a low-glycemic product. If the nutri- References

tional information given for bulk ingredient is incorrect, it can Baolingbao Biology Co., LTD. 2015. Certificate of Analysis. IMO-900 powder, lot # 15070331.
Arrived with product.
lead to downstream products that are mislabeled. The modern BioNeutra North America Inc. (Edmonton, Canada). 2013. VitaFiberTM IMO Powder, product
consumer gauges what they can eat (this is especially true of those profile issued 2013 May 22. Arrived with product.
BioNeutra North America Inc. (Edmonton, Canada). 2016. Selected Scientific Literature. Avail-
on restrictive diets) by reading the ingredients and nutritional in- able from: http://www.bioneutra.ca/_pdf/Structure-FunctionClaims.pdf. Accessed August
formation given on product labels. It is therefore imperative that 08.
they are correct. Dancheng Caixin Sugar Industry Co. LTD. 2013. Specification of isomaltooligosaccharide
(IMO), IMO900. Available from: http://www.caixin-sugar.com/product/114.html. Accessed
2016 March 24.
Conclusion Day DF, Chung C-H, inventors. 2004. Isomaltooligosaccharides from Leuconostoc as neutraceu-
ticals. US Patent 7291607 B2.
This analysis comparing 7 commercially available IMO-based Elliot R, McKay B, inventors. 2002. Compositions containing enzymes stabilized with certain
food ingredients demonstrates that most of these products are, osmo-protectants and methods for using such compositions in personal care. US Patent
US20030175232 A1.
by way of definition, and particularly with respect to content of Evert AB, Boucher JL, Cypress M, Dunbar SA, Franz MJ, Mayer-Davis EJ, Neumiller JJ,
“oligosaccharides” and “dietary fiber,” mislabeled. When com- Nwankwo R, Verdi CL, Urbanski P, Yancy WS. 2014. Nutrition therapy recommendations
for the management of adults with diabetes. Diab Care 37(Suppl. 1):S120–43. Available from:
paring the analytical results from this work with the declared https://doi.org/10.2337/dc14-S120.
fiber content within the context of the given claims, the data Fuentes-Zaragoza E, Riquelme-Navarrete MJ, Sanchez-Zapata E, Perez-Alvarez JA. 2010.
Resistant starch as a functional ingredient: a review. J Food Res Intl 43(4):931–42. DOI:
suggest that not all ingredients labeled as “IMOs” are quantita-
Food Chemistry

10.1016/j.foodres.2010.02.004.
tively equal. It appears that most manufacturers have adopted a Gibson GR, Roberfroid MB. 1995. Dietary modulation of the human colonic microbiota:
introducing the concept of prebiotics. J Nutr 125(6):1401–12.
loose definition that is neither based on an accurate quantitative Goffin D, Robert C, Wathelet B, Blecker C, Malmendier Y, Paquat M. 2009. A step-forward
analysis nor consequence of metabolic impact. This is significant method of quantitative analysis of enzymatically produced isomaltooligosaccharide prepara-
tions by AEC-PAD. Chromatographia 69(3/4):287–93. DOI: 10.1365/s10337-008-0875-0.
because claims, such as “low glycemic,” “zero calorie,” and the Goffin D, Wathelet B, Blecker C, Deroanne C, Malmendier Y, Paquot M. 2010. Comparison
like, are certainly false, and may pose a health hazard to certain of the glucooligosaccharide profiles produced from maltose by two different transglycosidases
from Aspergillus niger. Biotechnol Agron Soc Environ 14(4):607–16.
populations (diabetic patients and epileptic patients on ketogenic Hu Y, Ketabi A, Buchko A, Ganzle MG. 2013. Metabolism of isomalto-oligosaccharides by Lac-
diets, in particular) while misleading others (those on low car- tobacillus reuteri and bifidobacteria. Lett Appl Microbiol 57:108–14. DOI: 10.1111/lam.12076.
Jones JM. 2014. CODEX-aligned dietary fiber definitions help to bridge ‘fiber gap’. Nutr J
bohydrate diets). Therefore, we conclude that the definition of 13:34. DOI: 10.1186/1475-2891-13-34.
“IMOs,” and certainly, “fiber,” must not include digestible con- Kaneko T, Yokoyama A, Suzuki M. 1995. Digestibility characteristics of isomaltooligosaccharides
stituents such as disaccharides (isomaltose) and α-limit dextrins. in comparison with several saccharides using the rat jejunum loop method. Biosci Biotechnol
Biochem 59(7):1190–4. DOI: 10.1271/bbb.59.1190.
That is, the definition of “oligosaccharide” must be made to in- Ketabi A, Dieleman LA, Ganzle MG. 2011. Influence of isomalto-oligosaccharides on intestinal
clude DP ࣙ 3, only. This is the difference between a COA that microbiota in rats. J Appl Microbiol 110:1297–306. DOI: 10.1111/j.1365-2672.2011.04984.x
Kohmoto T, Fukui F, Takaku H, Machida Y, Arai M, Mitsuoka T. 1988. Effect of
indicates >90% soluble fiber for a product that actually contains isomalto-oligosaccharides on human fecal flora. Bifidobacteria Microflora 7(2):61–9. DOI:
approximately 36% DP 3 to 10 (actually DP 3 to 5 with an av- 10.12938/bifidus1982.7.2_61.
Kohmoto T, Keisuke T, Toshiyuki K, Shota M, Fukui F, Takaku H, Nakagawa Y, Ichikawa
erage of 3.8). Inclusion of DP 3 constituents into this definition T, Kobayashi S. 1992. Metabolism of 13C-isomaltooligosaccharides in healthy men. Biosci
requires further study, but the literature suggests that these are di- Biotechnol Biochem 56(6):937–40. DOI: 10.1271/bbb.56.937.
Kwon H-K, Jeong H-S, Lee J-H, inventors. 2011. Production of isomaltooligosaccharides and
gestible, as well. We conclude that labeling requirements should uses therefore. US Patent US8637103 B2.
be reconsidered for products of this type. If the nutritional in- Lee B-H, Yan L, Phillips RJ, Reuhs BL, Jones K, Rose DR, Nichols BL, Quezada-Calvillo R,
Yoo S-H, Hamaker BR. 2013. Enzyme-synthesized highly branched maltodextrins have slow
formation given for bulk ingredient is incorrect, it can lead to glucose generation at the mucosal α-glucosidase level and are slowly digestible in vivo. PloS
downstream products that are mislabeled. The modern consumer One 8 (4):e59745.
Lee PR, Kossoff EH. 2011. Dietary treatments for epilepsy: management guidelines for the
gauges what they can eat (this is especially true of those on restric- general practitioner. Epilepsy Behav 21(2):115–21. doi:10.1016/j.yebeh.2011.03.008.
tive diets) by reading the ingredients and nutritional information Madsen LR, Oswald J, Day DF, Moon Y-H, inventors. 2015. Process for the production of
isomaltooligosaccharides. US Patent Applied for PCT/US2015/046441.
given on product labels. It is therefore imperative that they are National Library of Medicine - Medical Subject Headings. 2011. Oligosaccharides. Defini-
correct. tion. Tree number D09 698:629. Available from: https://www.nlm.nih.gov/cgi/mesh/2011/
MB_cgi?mode=&term=Oligosaccharides. Accessed 2016 August 9.
Oku T, Nakamara S. 2003. Comparison of digestibility and breath hydrogen gas excretion
Acknowledgments of fructo-oligosaccharide, galactosyl-sucrose, and isomalto-oligosaccharide in healthy human
subjects. Eur J Clin Nutr 57:1150–6. DOI: 10.1038/sj.ejcn.1601666.
The authors report a business relationship as partners in Report of the Dietary Fiber Definition Committee to the Board of Directors of the Ameri-
ISOThrive LLC, a manufacturer of one of the ingredients an- can Association of Cereal Chemists. 2001. The Definition of Dietary Fiber. Am Soc Ce-
alyzed in the manuscript. The report is intended to provide an real Chem 46(3):112–126. Available from: http://www.aaccnet.org/initiatives/definitions/
Documents/DietaryFiber/DFDef.pdf. Accessed 2016 August 9.
accurate accounting of the composition of each product, while Rycroft CE, Jones MR, Gibson GR, Rastall RA. 2001. A comparative in vitro evaluation of the
bringing to light and averting potential health hazards to certain fermentation properties of prebiotic oligosaccharides. J Appl Microbiol 91(5):878–87. DOI:
10.1046/j.1365-2672.2001.01446.x.
patient populations due to the discovered mislabeling of ingredi- Sajilata MG, Singhal RS, Kulkarni PR. 2006. Resistant starch—a review. Compr Rev Food Sci
ents. Every effort has been made to maintain unbiased reporting Food Safety 5:1–17. DOI: 10.1111/j.1541-4337.2006.tb00076.x.
Sheng GE, Dong-lian CAI, Wan L-L. 2006. Determination of glycemic index of xylitol and
of facts with respect to all ingredients analyzed. The authors have isomaltooligosaccharide. Chin J Clin Nutr 14(4):235–7.
listed methods so that results can be replicated by any interested Shi Q, Hou Y, Juvonen M, Tuomainen P, Kajala L, Shukla S, Goyal A, Maaheimo H, Katina
K, Tenkanen M. 2016. Optimization of isomaltooligosaccharide size distribution by acceptor
party. We would like to acknowledge Savana Gilman for adminis- reaction of Weissella confusa dextransucrase and characterization of novel α-(1,2)-branched iso-
trative assistance in the preparation of this manuscript. maltooligosaccharides. J Agric Food Chem 64(16):3276–86. DOI: 10.1021/acs.jafc.6b01356.
Top Health Ingredients Inc. (Edmonton, Canada). 2014. Certificate of Analysis, AdvantaFib-
erTM 90P, COA#1090-14071631, Lot#14071631, issued July 17, 2014. Arrived with
Author Contributions product.
Zhong Z, Zhu J, Li X, Xu X-Y, Mu X, inventors. 2005. Method for the removal of monosac-
L. Madsen designed the study, collected test data, interpreted charide in oligosaccharides production. US Patent 7,906,314.
the results, and drafted the manuscript; S. Stanley collected test Zhu J. 2005. GRAS exemption claim. Bioneutra Inc. Available from: http://www.fda.gov/
downloads/Food/IngredientsPackagingLabeling/GRAS/NoticeInventory/ucm268863.pdf.
data; P. Swann interpreted the results and edited the manuscript; Accessed 2016 August 08.
and J. Oswald assisted in the design of the study, interpreted the
results, and edited the manuscript.

408 Journal of Food Science r Vol. 82, Nr. 2, 2017