Degree of Starch Gelatinization, Digestion Rate of Starch in Vitro, and Metabolic Response in Rats

Abstract

Glycemic response after ingestion of starchy foods varies. Starch in many common ready-to-eat foods is only partly gelatinized. In view of this, the relationships among degree of starch gelatinization, in vitro digestion rate, and in vivo metabolic response in rats were studied. Wheat starch with different degrees of gelatinization was used in the experiments. Plasma glucose and insulin responses as well as the rate of in vitro hydrolysis with alpha-amylase were strongly correlated to the degree of starch gelatinization (r = 0.88, r = 0.90, and r = 0.96, respectively). Plasma glucose and insulin responses were also positively correlated to the rate of hydrolysis with alpha-amylase in vitro (r = 0.98 and r = 0.76, respectively). These results suggest that the degree of starch gelatinization is an important determinant both for the rate of starch hydrolysis in vitro and for the metabolic response in vivo.

Full text

1010 Am J C/in Nuir l988;47: 1010-6. Printed in USA. © 1988 American Society for Clinical Nutrition
Degree of starch gelatinization, digestion rate of starch
in vitro, and metabolic response in rats1’2
J#{246}rgenHo/rn, MSc; Jngmar Lundquist, MD, PhD; Inger Bj#{246}rck,PhD;
Ann-Charlotte Eliasson, PhD; and Nils-Georg Asp, MD, PhD
ABSTRACI’ Glycemic response after ingestion of starchy foods varies. Starch in many
common ready-to-eat foods is only partly gelatinized. In view ofthis, the relationships among
degree of starch gelatinization, in vitro digestion rate, and in vivo metabolic response in rats
were studied. Wheat starch with different degrees ofgelatinization was used in the experiments.
Plasma glucose and insulin responses as well as the rate ofin vitro hydrolysis with a-amylase
were strongly correlated to the degree ofstarch gelatinization (r =0.88, r=0.90, and r=0.96,
respectively). Plasma glucose and insulin responses were also positively correlated to the rate
ofhydrolysis with a-amylase in vitro (r =0.98 and r=0.76, respectively). These results suggest
that the degree of starch gelatinization is an important determinant both for the rate of starch
hydrolysis in vitro and for the metabolic response in vivo. Am J C/in Nuir 1988;47:
1010-6.
KEY WORDS Starch gelatinization, starch digestion rate, a-amybolysis, plasma glucose,
plasma insulin
Introduction
The variable glycemic response after ingestion of
starchy foods has been the subject of much interest re-
cently, especially in relation to diabetes. So far, however,
not much attention has been paid to the importance of
the degree ofstarch gelatinization (DG).
In plant cells starch is present as granules. Starch poby-
mers (amybose and amybopectin) are tightly packed in
granules with a high degree of molecular order and are
associated by hydrogen bonding. Raw granules contain
highly crystalline regions and are birefringent in polar-
ized bight. The granules are insoluble in cold water.
When exposed to heat in the presence ofwater, the starch
granules undergo an irreversible swelling and destruction
of the internal crystalline structure and birefringence is
lost. This transformation is termed gelatinization. With
excessive treatment the granules may even rupture and
disintegrate and a fraction of the starch is then sobubi-
lized.
Raw starch is only slowly digested by enzymes in vitro
whereas cooking increases the susceptibility considera-
bly because ofthe rupture and disintegration ofthe com-
pact crystalline granular structure (1-4). Furthermore,
the glucose and insulin responses in vivo are significantly
greater after ingestion of cooked compared with raw
starches(3-7). Consequently, DG is an extremely impor-
tant factor in the rate ofstarch hydrolysis and metabolic
response.
However, in many common plant foods the starch is
only partly gelatinized, because ofthe limited water con-
tent during processing. The starch granules are only
slightly swollen and the internal structure is partly intact.
Examples of such foods are breakfast flakes of cereal
grains (3) and several baked products (8- 1 1). The starch
granules in legumes may swell incompletely during pro-
cessing (12, 13). In view of the increased attention to
starchy foods and the nutritional advantages of carbohy-
drates that are slowly digested and absorbed (14-19), the
nutritional properties of incompletely gelatinized starch
are of interest.
This paper describes the close relationship between the
DG of pure wheat starch and the enzymatic susceptibil-
ity in vitro as well as the glucose and insulin responses
to starch in rats. The results are discussed in terms of a
IFrom the departments ofFood Chemistry and Food Technology,
University of Lund, Lund, and the Department ofCell Biology, Urn-
versity ofLink#{246}ping, Link#{246}ping,Sweden.
2Address reprint requests to J#{246}rgenHolm, Department of Food
Chemistry, Chemical Center, University ofLund, P 0 Box 124, S-22l
00 Lund, Sweden.
ReceivedFebruary2, 1987.
Accepted for publication June 16, 1987.
by guest on July 11, 2011www.ajcn.orgDownloaded from

STARCH GELATINIZATION AND DIGESTION RATE bOll
TABLE 1
Physical and morphological characteristics ofwheat-starch samples used in the experiments
Treatment
temperaturet Heat ofgelatinization En)
Degree of gelatinization
Swelling
power [nJ Solubility [n] Microscopical
appearance
Calorimetric
method
(DSC)[nJ Enzymatic
method [n]
C J/g(cal/g) % % %
-
47
50
53
56
59
65
100
10.52 ± 0.29 (2.5 1 ± 0.07) [3]
9.02 ± 0. 17 (2. 15 ± 0.04)[4)
6.64 ±0.25 (1 .59 ±0.06) [5]
3.04 ± 0.03 (0.73 ±0.01 ) [3]
1.59 ±0.02 (0.38 ±0.0 1) [21
0.42 (0. 10) [2]
0 [2]
0 [2]
Ot [3]
l4.2 [4]
36.9t [5]
7 1.1f [3]
85.0 [21
96.Of [2]
l00 [2]
l00 [21
0. 1 ± 0.4t [4]
14. 1 ±1.3f [4]
47.3 ±1.9f (6]
91 .9 ±0.4f [4J
97.2 ±0.7f [2]
97.8f [2]
-
-
2.0 [4]
2.4 [4]
2.8 [3)
3.9 [31
5.4 ±0. 1 [3]
6. 1 ± 0.2 [3]
6.9 ±0.2 [3]
-
0.2 [4J
0.4 ±0. 1 [4J
0.6 ±0. 1 [3]
0.8 ±0. 1 [3]
1.4 ±0. 1 [3]
1.6 [3J
2.8 [3)
35.0 ±4.0(3J
Intact birefringent
granules
Increased swelling and
decreased number of
granules showing
birefringence
Extensively swollen
nonbirefringent
granules
Granules disintegrated
Mean±SEM.
tRaw.
tPartly gelatinized.
§Completely gelatinized.
physical and morphological characterization of the
starch.
Materials and methods
Preparation ofstarch samples
Raw commercial wheat starch was obtained from KEBO
Lab AB (Stockholm, Sweden). The starch content was 98.2 g/
100 g (polymer weight, dry basis), determined as glucose after
incubating the sample with a thermostable a-amylase at 95 #{176}C
and amyloglucosidase at 60 #{176}C(20). To obtain starch with
different DOs, 100 g/L suspensions ofwheat starch in distilled
water were heated under gentle agitation at 47, 50, 53, 56, 59,
or 65 C. After 20 mm the samples were cooled to room tern-
perature. A boiled sample was prepared by boiling a 50 g/L
suspension for 20 mm.
Degree ofstarch gelatinization
The DO was measured by two different methods. Duplicate
analyses were performed on each preparation. With the cabori-
metric method (differential scanning calorimetry [DSC]) we
measured the heat ofgelatinization (H) ofthe starch samples,
ie, the energy that has to be supplied to obtain complete starch
gelatinization. Ten to 15 mg of a bOO g/L suspension was
heated from 22 to 80 #{176}Cat a scanning rate of 10 C/min in
coated sample pans. H was calculated from the peak area of
the thermogram. The instrument used was a Perkin-Elmer
DSC-2 (Perkin-Ebmer Corp, Eden Prairie, MN). The DG was
calculated by comparing iH for raw starch with that for heat-
treated starch:
DG (%) =(1 -x100 (1)
The enzymatic method is based on the principle that gelatin-
ized starch is easily digested by glucoamybase to form glucose.
A small amount of the heat-treated starch suspensions corre-
sponding to 20 mg of starch was withdrawn and analyzed ac-
cording to Chiang and Johnsson (21). DO was expressed as the
percentage ofthe total starch content that was immediately sus-
ceptible to glucoarnylase.
Swelling and solubi/itypauerns
The heat-treated suspensions were carefully transferred to
weighed centrifuge bottles and diluted with distilled water to
30 g/L starch concentration. A suspension of raw starch was
kept for 20 mm at room temperature under gentle agitation.
The samples were centrifuged for 15 mm at 700 Xg. The super-
natant was decanted and evaporated to dryness in the oven and
the amount ofsolute was measured. Granule swelling and solu-
bility were calculated according to Leach et al (22):
Swelling power
=weight ofgel/(total starch -weight ofsolutes) (g/g) (2)
Solubility (%)
Microscopy
=100 X weight ofsolutes/total starch (gig) (3)
A few drops ofa 10 g/L suspension were examined by light
microscopy and by polarized light microscopy to further char-
acterize the morphology ofthe granules.
Experiments in vitro
A porcine pancreatic a-amylase preparation (27 mg protein/
mL, 1200 units/mg; Sigma Chemical Co, St Louis, MO) was
diluted 1:20 or 1:400 before incubation. The heat-treated sam-
ples were diluted with distilled water to 50 gIL. Forty-five milbi-
liters of0.05 mob/L Na,K-phosphate buffer(0.025 mol/L each
of KH2PO4 and Na2HPO4) containing 0.4 g/L NaC1 (pH 6.9)
and 1.25 mL ofdiluted a-amylase preparation were added to a
lO-mL subsample corresponding to 500 mg starch. Samples
were taken before and after 5-60 mm incubation (under gentle
by guest on July 11, 2011www.ajcn.orgDownloaded from

‘I)
Cl)
>‘
0
I
80
70
60
50 -
40
30
20
10
0-
05 15 30 60
1012 HOLM ET AL
agitation at 37 #{176}C)and analyzed with dinitrosalicylic acid for
content ofreducing sugar (23). A standard curve was prepared
using maltose. The extent of hydrolysis (the proportion of
starch degraded to maltose, or percent maltose equivalents)
was calculated as 100 times milligrams of maltose equivalents
times 0.95 divided by milligrams ofstarch in sample.
Experiments in vivo
The availability of starch for digestion and absorption in
vivo in male rats (Sprague-Dawley, 120-150 g) was studied by
analyzing concentrations of glucose and insulin in plasma at
various time intervals after gastric intubation. The starch sam-
ples were diluted with a NaCl solution to 40 gIL. The final con-
centration of NaCI was 9 gIL. After being starved for 24 h un-
anesthetized rats were intubated with a starch suspension (2.5
mL) corresponding to 100 mg starch. Nine rats were used for
each sample tested. Blood was obtained from the rats by multi-
pIe serial sampling from the retroorbital venous plexus. After
135 mm the rats were killed by a blow on the head and a piece
of the liver (‘-0.5 g) was taken from the ventral part of the
left lateral lobe for analysis ofglycogen content. Estimation of
plasma glucose concentration was done with a glucose-oxidase
method (24). Plasma insulin levels (IRI) were analyzed with
a radioirnmunoassay (25). Liver glycogen concentration was
determined by the method ofRerup and Lundquist (26).
St atistical evaluation
The results are given as mean ± SEM. The significance of
differences was tested with Student’s ttest.
Results
The physical and morphological characterization of
the starch samples used in the in vitro and in vivo experi-
ments are presented in Table 1. With increasing DG
there was an increased susceptibility to glucoamylase as
well as a gradual boss ofbirefringence, indicating that the
highly ordered structure within the granules was de-
stroyed to different extents. The soluble starch fraction
was only slightly increased for the partly gelatinized sam-
ples and for the completely gelatinized sample heated at
65 #{176}C(just above the gelatinization temperature range).
In contrast, after boiling, a large fraction of the starch
was soluble, indicating disintegration of the granular
structure.
The susceptibility of starch with different DGs to hy-
drobysis by porcine pancreatic a-amylase (added at a
concentration of200 units/g starch where 1 unit liberates
1mg maltose from soluble starch in 3 mm at pH 6.9 and
at 20 #{176}C)is shown in Figure 1. Raw starch was hydro-
lyzed most slowly and the susceptibility to a-amylase in-
creased with DO. The two completely gelatinized sam-
ples(DG =100%)and starch with a DG of96% displayed
the highest availability; no significant difference was oh-
served between these three samples.
When a large excess ofenzyme was used by increasing
the enzyme concentration 20-fold to 4050 units/g starch,
the initial hydrolysis rate increased as did the extent of
hydrolysis (Fig 2). However, the differences in availabil-
ity between the samples, especially in the initial hydroly-
Incubation time (mm)
FIG 1. Hydrolysis ofstarch with different DOs by porcine pancreatic
a-amylase added at a concentration of 200 units/g starch. Percentage
hydrolysis is expressed in terms of maltose equivalents. DG (%): #{149},0
(raw); A, 14.2; #{149},36.9; 0, 71.1; t, 85.0. The bars represent the interval
obtained with 96.0, lOO(heated at 65 ‘C,just above gelatinization tem-
perature range), and 100% (boiled). The SEM did not exceed 3.4 for
any experimental point. n=2-3.
sis rate, were still very pronounced. The boiled sample
was hydrolyzed to 75% (maltose equivalents) within 5
mm and reached a plateau at 80% after 30 mm, whereas
the hydrolysis values at 5 mm for raw starch and starch
with DGs of 14 and 37% were b2, 30, and 48%, respec-
tively.
The course of the plasma glucose and insulin re-
sponses in rats after intubation with starch with different
DGs are shown in Figures 3 and 4 and the area under
the curves are presented in Table 2. The plasma glucose
response after intubation with boiled starch (DG
=100%) was much greater than after raw starch (DG
=0%). The responses to the partly gelatinized products
were intermediate. The response to the partly gelatinized
starch with a DG of 37% was greater than to that with a
DG of 14%. With raw starch the glucose peak was de-
bayed (peak value at 60 mm), whereas the peak values
for the partly and completely gelatinized starch samples
were obtained after only 15 mm. The insulin responses
were closely associated with the glucose responses. The
rise in plasma glucose (0-60 mm) and plasma insulin (0-
30 mm) reflect the early rates of digestion and absorp-
tion, where the largest differences are found. The liver
glycogen contents 1 35 mm after intubation were 4.1
by guest on July 11, 2011www.ajcn.orgDownloaded from

90
‘I)
‘I)
>
0
>‘
I
80
70
60
50
40
30
20
10
0I
___0 5 15 30 60
Incubatmon tmme (1Tn)
smgnifmcant dmfferences
_i aQ (mm)
.-Ap<0.00 1 p <0.001
.-Up<0.00 1 p <0.00 1p<0.001
.- 0 p<0.00 1p<0.00 1po.oo i
A-I p<0.01
A- 0p< 0001 p< 0.001 p< 0.05
1-0 p<o.o5
-I
0
E
E
0)
Cl)
0
C)
a)
E
(0
CO
0
10
9,
8
7
6
5
4
3,
1#{149} I I I I
015 30 60 120
STARCH GELATINIZATION AND DIGESTION RATE 1013
starch was hydrolyzed almost completely within 5 mm,
barge differences in the rate of hydrolysis remained (Fig
2). Starch hydrolysis with a-amylase results in the forma-
tion ofglucose, maltose, maltotriose, and alimit dextrins
(branched obigosaccharides with four glucose monomers
or more). Therefore, 75-80% hydrolysis expressed as
maltose equivalents corresponds to an almost complete
hydrolysis. Once the highly organized structures within
the swollen granules were completely destroyed, as in the
sample treatedjust above the gelatinization temperature
range, the availability ofstarch was not increased further
by swelling, rupture, and disintegration of the granular
structure (boiled sample) (Fig 1). The increased suscepti-
bility to a-amylase was not rebated to increased sobubility
because the soluble starch fraction did not exceed 3% in
any sample, except for boiled starch.
The plasma glucose and insulin concentrations in rats
were raised to different extents depending on the DG. An
already bow DG (14%), which caused only slight swelling
and internal disorganization of the granules (Table 1),
raised the glycemic response far above that ofraw starch.
The decline in plasma glucose concentration was slower
for partly gelatinized starch than for completely gelatin-
FIG 2. Hydrolysis ofstarch with different DGs using a large excess
of porcine pancreatic a-amylase (4050 units/g starch). DG (%): #{149},0
(raw); A, 14.2; ,36.9; 0, 100 (boiled). The SEM did not exceed 3.4
for any experimental point. n=2-3.
± 0.7 (raw), 6.9 ± 0.8 (DG =14%), 6.1 ± 1.1 (DG
=37%), and 6.3 ± 0.5 (boiled) mg/g wet wt ofliver. The
glycogen content was significantly bower with raw starch
than with boiled starch or starch with DG of 14% (p
<0.05). The liver glycogen content in control rats(n =4)
intubated with 0.9% NaCb solution was 0.3 ±0.1.
The plasma glucose and insulin responses were posi-
tiveby correlated with the rate of hydrolysis with a-
amylase in vitro, and both the plasma glucose and insu-
lin responses as well as the rate ofhydrolysis with a-amy-
lase were positively correlated with the DG (Table 3).
Discussion
During gelatinization inter- and intramolecular hy-
drogen bonds are broken. This results in a loosening up ___________________________________
of the compact granular structure and allows different
degrees of swelling and absorption of water, fully hy-
drated starch molecules leach from the granule. Conse- Time after intubation (mm)
quently, the availability of the starch granules to diges- FIG 3. Plasma glucose responses at various intervals in rats given
tive enzymes increases to different levels with increasing h with different DOs (100 mg as a 40 g/L suspension in 9 g/L
DG. Even when we used an enzyme concentration in our NaCI). DG (%): #{149},0 (raw); A, 14.2; #{149},36.9; 0, 100 (boiled). Nine rats
in vitro system that was so high that the boiled wheat were used for each sample tested. (Mean ±SEM.)
by guest on July 11, 2011www.ajcn.orgDownloaded from

0.3
significant differences
i. 122 (mm)
#{149}-A p<0.05
#{149}-o p<0.01
A-D p<0.01
1-0 p<0.05
TABLE 3
Correlation coefficients between DO, digestion rate ofstarch in vitro,
and plasma glucose and plasma insulin responses in rats
0.2
r
DO with digestion rate (n =7)t 0.96
DO with plasma glucose(n =4) 0.88
DO with plasma insulin (n =4) 0.90
Digestion rate with plasma glucose (n =4) 0.98
Digestion rate with plasma insulin (n =4) 0.76
I I I I
120
0
015 30 60
Time after intubation (mm)
FIG 4. Plasma insulin (IRI) responses at various intervals in rats
given starch with different DOs (100 mg as a 40 g/L suspension in 9
g/L NaG). DG (%): #{149},0(raw); A, 14.2; I, 36.9; 0, 100 (boiled). Nine
rats were used for each sample tested. (Mean ±SEM.)
ized starch, which confirms that digestion and absorp-
tion are prolonged when swelling is limited. Raw starch
was absorbed very slowly and elicited a flattened plasma
glucose curve with a delayed and less pronounced peak
and a very slow decline.
The effect ofcompbete gelatinization on metabolic re-
sponse has been studied in humans, both in diabetic and
nondiabetic subjects(5-7)and in animals(3, 4). Potatoes
and corn starch give much lower postprandial glucose
and insulin responses in raw form than after cooking (5-
7). In fact, raw corn starch has been used clinically to
TABLE 2
Area under the curve after gastric intubation ofrats with starch with
different DG5
*DO was calculated from calorimetric measurement ofthe heat of
starch gelatinization. Digestion rate =degree ofhydrolysis after 60 mm
incubation with a-amylase (added at a concentration of 200 units/g
starch). Plasma glucose =area under the curve, 0-60 mm. Plasma insu-
lin =area under the curve, 0-30 mm.
tn=number of starch samples with different DGs on which the
calculations were based.
provide glucose with prolonged absorption in the treat-
ment of type 1 glycogenosis (27). The effect of raw
starchy foods on the glycemic response may be more fa-
vorable than supplementation with gel-forming types of
dietary fiber. The use of raw plant food in the nutrition
ofdiabetic patients was discussed (28).
Despite a low enzymatic availability, raw wheat starch
is almost completely digested and absorbed in the rat
small intestine (29). Raw potato starch, on the other
hand, is poorly digested (30).
In some processed food products the starch is only
partly gelatinized with a slightly swollen granular struc-
ture and the internal granular organization partly intact.
For example, some baked products and breakfast cereals
contain incompletely gelatinized starch granules, which
is attributed to the limited water available (10-60 g/lOO
g) when these products are produced. Cereal flakes are
produced by steaming whole kernels and then flaking
them between rollers. Starch in breakfast cereals or pre-
cooked convenience foods produced under more severe
conditions (elevated temperatures, pressures, and shear
forces, eg, extrusion cooking or popping) is normally
completely gelatinized despite the bow water content. A
recent study (3) found that the plasma glucose and insu-
bin responses in rats were lower after ingestion of wheat
flakes with a DG of45%, as measured by DSC, than after
boiled completely gelatinized whole-grain wheat flour.
The in vivo responses were closely related to the in vitro
digestibility with pepsin and a-amylase.
According to Wootton and Chaudhry (1 1) DGs in
short bread, hard sweet cookies, soda crackers, crisp-
bread, wafer, fruit cake, and bread crumbs were 1,2, 3,
33, 40, 50, and 60%, respectively. The corresponding in
vitro digestibility values with porcine pancreatic a-
amybase were 18, 25, 33, 43, 56, 66, and 69%, respec-
tively whereas pregelatinized wheat starch was digested
to 90%. The variations in gelatinization were explained
in terms ofprebaking water content, baking time at high
moisture bevel favoring high DG, and the presence of
added ingredients that restrict gelatinization. Lineback
Area under the curve
Plasma glucoset Plasma glucose Plasma insulin
DO (0-60 mm) (0-120 mm) (0-30 mm)
%mmol/L Xmm mmol/L Xmm nmo!/L Xmm
0
14
37
100
l26±l3-
r2ll±lOIt
111248±15 It
§1t-293±l1Jt
282±27
1381±19 t
§1465±22 t
t1512±25 t
-0.31±0.66
0.35±0.53
f-0.25±0.55
§1 2.12±0.68 §
IMean ±5EM; n=9. For each bracket, values without footnotes were sig.
nificantly different from values with footnotes.
t <0.001.
tp <0.01.
§p<0.O5.
1014 HOLM ET AL
-a
0
E
C
a:
by guest on July 11, 2011www.ajcn.orgDownloaded from

STARCH GELATINIZATION AND DIGESTION RATE 1015
and Wongsrikasem (10) found DOs between 4 and 97%
in several commercial U.S. baked products. Hagander et
al (3 1) observed that extruded crispbread elicited larger
glucose and insulin responses than a conventionally
baked bread in diabetic patients, possibly because the
conventionally baked bread had a bower DG.
Snow and O’Dea (1) obtained a more rapid hydrolysis
of starch in vitro in commercial breads compared with
homemade breads and suggested that this might have
been due to a greater exposure to heat in the commercial
breadmaking process. L#{252}deret al (32) reported signifi-
cantby lower blood glucose bevels in healthy subjects after
ingestion of 85% rye-15% wheat bread with a baking
time of35 mm compared with bread with a baking time
of 45, 55, or 65 mm. These differences in glycemic re-
sponse were probably caused by differences in DO that
were caused by the variations in baking time.
The extent ofstarch gelatinization also varies within a
given bakery product because ofa moisture gradient (8).
Thus, starch gelatinization in white pan bread ranged
from 33% in the crust to 70% in the center ofthe crumb
(8). The crust contained many unswollen birefnngent
starch granules whereas starch granules in the crumb
center were deformed and only slightly birefringent. The
differences in DG between crumb and crust is the proba-
ble reason for differences in postprandial hyperglycemia
after ingestion ofbread with different crust-to-crumb ra-
tios (33). Bread rolls with high crust-to-crumb ratios
raised the blood glucose level more slowly with a later
peak and a slower decline than loaves with bow crust-to-
crumb ratios.
Dreher et al (34) obtained a higher in vitro digestibility
and more extensive gelatinization of starch in potato
chips compared with starch in baked potato. The in vitro
digestibility of potato chips, baked potato, and raw po-
tato were 66.3, 53.6, and 228%, respectively.
Differences in the extent of gelatinization after pro-
cessing legumes in different ways was suggested as a plau-
sible explanation for differences in in vitro starch digest-
ibility (2). It has been suggested that the low metabolic
response and in vitro starch digestibility of legumes is
caused by the entrapment of starch in the cells (12, 13,
35). The cell walls may limit the hydrolysis rate because
of steric hindrance to enzymatic attack (13). Further-
more, the cell walls may also inhibit starch gelatinization
by physically limiting the degree ofswelling ofthe starch
granules and by limiting the water transport necessary
for complete gelatinization (12, 13). Grinding raw be-
gumes, thus disrupting the cell walls, followed by cook-
ing resulted in an increased swelling ofthe granules, and
it was proposed that this could increase the rate of starch
digestion as well as the glycemic response (12, 13).
DSC is a calorimetric method based on the physical
course of gelatinization but it requires equipment not
usually available in most nutritional laboratories. How-
ever, the values for DU obtained with the simple and
rapid enzymatic method using glucoamylase were
closely correlated (r =0.99) with those obtained with
DSC (Table 1).
The characterization ofthe starch, especially regarding
the degree ofgebatinization, would be desirable in studies
concerned with the rate of digestion and absorption of
starchy foods. Light microscopy in combination with po-
barized light microscopy as webb as enzymatic methods
like that used in this study are rapid methods that are
well suited for this purpose.
In conclusion, the rate ofstarch hydrolysis in vitro and
the glucose and insulin responses in vivo are increased to
different extents, depending on DO. Hence, starchy
foods with low DOs may be beneficial nutritionally be-
cause they favor a reduced rate ofstarch uptake. 13
References
1. Snow P, O’Dea K. Factors affecting the rate ofhydrolysis of starch
in food. Am J Gin Nutr 198l;34:272l-7.
2. El Faki HA, Venkataraman LV, Desikachar HSR. Effect of pro-
cessing on the in vitro digestibility of proteins and carbohydrates
in some Indian legumes. Qual Plant 1984; 34:127-33.
3. Holm J, Bj#{246}rckI, Asp N-G, Sj#{246}bergL-B, Lundquist I. Starch avail-
ability in vitro and in vivo after flaking, steam-cooking and pop-
ping ofwheat. J Cereal 5th 1985;3: 193-206.
4. Lee PC, Brooks SP, Kim 0, Heitlinger LA, Lebenthal E. Digestibil-
ity of native and modified starches: in vitro studies with human
and rabbit pancreatic amylases and in vivo studies in rabbits. J
Nutr 1985; 1 15:93-103.
5. Collings P, Williams C, MacDonald I. Effects ofcooking on serum
glucose and insulin responses to starch. Br Med J198 l;282: 1032.
6. Vaaler 5, Hanssen KF, Aagenaes 0. The effect ofcooking upon the
blood glucose response to ingested cai-rots and potatoes. Diabetes
Care l984;7:221-3.
7. Chen YT, Muenzer J, Sidbury JB. A study ofabsorption of differ-
ent starches with equivalent as cooked food. Gin Res l983;3l:
615A(abstr).
8. Varriano-Marston E, Huang VKEG, Ponte J. Comparison of
methods to determine starch gelatinization in bakery foods. Cereal
Chem l980;57:242-8.
9. Wootton M, Weeden D, Munk N. A rapid method for the estima-
tion ofstarch gelatinization in processed foods. Food Technol Aust
197l;23:6l2-5.
10. Lineback DR. Wongsrikasem E. Gelatinization ofstarch in baked
products. J Food Sci 1980;45:71-4.
1 1. Wootton M, Chaudhry MA. Gelatinization and in vitro digestibil-
ity ofstarch in baked products. J Food 5th l980;45:l783-4.
12. Thorne Mi, Thompson LU, Jenkins DJA. Factors affecting starch
digestibility and the glycemic response with special reference to le-
gumes. Am J Clin Nutr l983;38:48l-8.
13. W#{252}rschP. Del Vedovo S, Koellreutter B. Cell structure and starch
natureas keydeterminantsofthedigestion rate ofstarch in legume.
Am J Gin Nutr 1986;43:25-9.
14. Crapo PA, Insel J, Sperling M, Kolterman 0G. Comparison of se-
rum glucose, insulin, and glucagon responses to different types of
complex carbohydrate in noninsulin-dependent diabetic patients.
AmJClinNutr l98l;34:l84-90.
15. Jenkins DJA, Wolever TMS, Taylor RH, et al. Glycemic index of
foods: a physiological basis for carbohydrate exchange. Am J Clin
Nutr l98l;34:362-6.
16. Jenkins DJA, WoleverTMS, Kalmusky J. Lowglycemic index car-
by guest on July 11, 2011www.ajcn.orgDownloaded from

1016 HOLM ET AL
bohydrate foods in the management ofhyperlipidemia. Am J Can
Nutr l985;42:604-l7.
17. Brudevold F, Goulet D, Tehrani A, Attarzadeh F, van Houte J.
Intraoral demineralization and maltose clearence from wheat
starch.CariesRes 1985; 19:136-44.
18. Pyorala K. Relationship ofglucose tolerance and plasma insulin to
the incidence of coronary heart disease: results from two popula-
tion studies in Finland. Diabetes Care l979;2:l31-4l.
19. Heaton KW. Fibre, satiety and insulin-a new approach to over-
nutrition and obesity. In: Heaton KW, ed. Dietary fibre: current
developments ofimportance to health. London: John Libbey & Co
Ltd. 1978:141-8.
20. Holm J, Bj#{246}rckI, Drews A, Asp N-G. A rapid method for the anal-
ysis ofstarch. Starch/St#{228}rke 1986; 38:224-6.
21. Chiang Y, Johnson JA. Measurement of total and gelatinized
starch by glucoamylase and toluidine reagent. Cereal Chem
l977;54:429-35.
22. Leach HW, McCowen LD, Schoch TJ. Structure ofthe starch gran-
ule. 1 Swelling and solubility patterns of various starches. Cereal
Chem l959;36:534-44.
23. Hostettler F, Borel E, Deuel H. Uber die Reduktion der 3,5-
dinitrosalicylsliure durch zucker. Helv Chim Acts 195 l;34:
2 132-9.
24. Bruss ML, Black AL. Enzymatic microdetermination of glycogen.
Anal Biochem l978;84:309-l2.
25. Heding L. A simplified insulin radioimmunoassay method. In: Do-
nato L, Milhaud 0, Sirchis J, eds. Labelled proteins in tracer stud-
ies. Brussels: Euratom, 1966:345-SO.
26. RerupC, Lundquistl. Non-specific reaction ofcurrentglucose oxi-
dase preparations with glycogen and its application for glycogen
determinations in tissue. Acta Pharmacol Toxicol l967;25:4l-53.
27. Wolfsdorf JI, Laffel LB, Crigler JF Jr. Uncooked cornstarch: re-
quirements for biochemical control in Type I glycogenosis. Gin
Res 1984; 32:764A(abstr).
28. DougiassJM. Raw diet and insulin requirements. Ann Intern Med
1975;82:6 1-2.
29. Bj#{246}rckI, Nyman M, Pedersen B, Siljestrom M, Asp N-G, Eggum
BO. On the digestibility ofstarch in wheat bread-studies in vitro
andinvivo.JCerealSci 1986;4:l-ll.
30. Holm J, Bj#{246}rckI, Ostrowska S, et al. Digestibility of amylose-lipid
complexes in-vitro and in-vivo. Starch/St#{228}rke l983;35:294-7.
31. Hagander B, Bj#{246}rckI, Asp N-G, Ct al. Hormonal and metabolic
responses to breakfast meals in NIDDM: comparison ofwhite and
whole-grain wheat bread and corresponding extruded products.
Human NutrAppl Nutr l985;39A:1 14-23.
32. L#{252}derW, Aust L, Vetter K. Zum Verlaufder alimentAren Hyper-
glyk#{228}mienach Brotverzehr. Die Nabrung 1969; 13:19 1-8.
33. Glatzel H, Rettenmaier 0. Ern#{228}hrungsphysiologische Brotver-
gleiche. 4 Der Verlaufder aliment#{228}ren Hyperglyk#{228}mie. Nutr Dicta
Eur Rev Nutr Diet l962;4:283-96.
34. Dreher ML, Breedon C, Orr PH. Percent starch hydrolysis and di-
etary fibercontent ofchipped and baked potatoes. Nutr Rep Intern
l983;28:687-9 1.
35. Tappy L, Wursch P, Randin JP, Felber JP, J#{233}quierE. Metabolic
effect ofpre-cooked instant preparationsofbean and potato in nor-
mal and in diabetic subjects. Am J Gin Nutr l986;43:30-6.
by guest on July 11, 2011www.ajcn.orgDownloaded from