Content uploaded by Arvin Paul P. Tuaño
Author content
All content in this area was uploaded by Arvin Paul P. Tuaño on Sep 23, 2016
Content may be subject to copyright.
Philippine Journal of Crop Science (PJCS) April 2016, 41 (1):52-59
Copyright 2016, Crop Science Society of the Philippines
Short-term Satiety of Brown Rice and Milled Rice
Angelina dR. Felix1*, Arvin Paul P. Tuaño2, Aida C. Mallillin3, Trinidad P. Trinidad3 and
Bienvenido O. Juliano2
1Institute of Human Nutrition and Food, College of Human Ecology, University of the Philippines Los Baños,
4031 College, Laguna, Philippines; 2Philippine Rice Research Institute Los Baños, 4031 College, Laguna,
Philippines; 3Food and Nutrition Research Institute, Department of Science and Technology, Bicutan, Taguig
City, Metro Manila 1631, Philippines. *Corresponding author, adrfelix@uplb.edu.ph/adfelix2@up.edu.ph
The study was conducted to determine the short-term satiety of brown and milled rice forms of
modern Philippine rice varieties of wider range of apparent amylose content (AC) types, and
differing in glycemic index (GI). Brown and milled rices representing waxy, low-AC,
intermediate-AC and high-AC types were cooked to comparable hardness by adjusting
water-rice ratio. The rice samples were subjected to estimation of satiety in 12 normal Filipino
subjects using a randomized, crossover study design. Short-term satiety indexes used were: 1)
satiety quotient (SQ) for four parameters, namely hunger, fullness, desire to eat, and
prospective consumption; 2) overall satiety index (SI); and 3) 2-h post meal common cooked
rice intake. SQ for the four parameters and SI were comparable across AC types. Consistent
with earlier studies, milled rice satiety was independent of AC, as well as of its GI. Meanwhile,
there was a tendency for both SQ and SI to be higher in brown rice than in milled rice, and SI
had strong positive correlation with dietary fiber (DF) content. Contrary however to hypothesis,
the higher SQ and SI for brown rice was not translated into lower subsequent food intake after
2 h. Common cooked rice intake was not affected by DF content. Cooking method, water-rice
ratio, water content, hardness (tenderness) and stickiness of cooked rice may have complicated
the findings. These have to be further elucidated, together with the roles of cognitive cues and
sensory qualities. Follow-up studies of longer duration and to include physiological biomarkers
are likewise suggested.
Key words: brown rice, cooked rice intake, milled rice, satiety index, satiety quotient
INTRODUCTION
Diabetes mellitus is considered a global pandemic
with more than 300M cases expected by year 2030
(Shaw et al. 2010). The problem is attributed to the
shift away from traditional lifestyles and dietary
patterns to an increasingly sedentary behavior
coupled with excess intake of energy-dense foods
and the rising rates of obesity (Priebe et al. 2008).
Epidemiological and clinical studies specifically point
to the adverse consequences of foods and diets that
are rich in readily and easily digested carbohydrates
(Sluijs et al. 2010). Among rice consuming
populations, increased consumption of milled rice
reportedly resulted in an increased risk of type 2
diabetes mellitus (Hu et al. 2012). Replacing milled
rice with unpolished rice or brown rice on the other
hand, has been suggested to mitigate adverse
consequences associated with refined rice
consumption (Sun et al. 2010; Dixit et al. 2011).
Regulation of food intake is an important strategy to
prevent and manage diabetes mellitus and its risk
factors like overweight and obesity. Intake of foods
that have high satiating capacity is expected to
reduce feeling of hunger resulting in low food and
energy intakes (Salmenkallio-Martilla et al. 2009).
Early school of thought proposed satiety as
interrelated to glycemic index (GI) – a physiologically-
based classification of carbohydrate foods based on
their ability to release glucose in rapid or sustained
fashion for a period of time (Brouns et al. 2005).
Low-GI foods are regarded to have higher satiating
capacity (Ludwig 2000). A more recent review
(Niwano et al. 2009), however, proposed that satiety
and food and energy intakes are independent of GI.
Rice GI correlates with starch composition,
particularly apparent amylose content (AC) and
dietary fiber (Trinidad et al. 2013). Meanwhile,
researches on rice satiety are limited. Pioneering
studies involved high- and low-AC puffed rice cakes
fed to nine Australian subjects by Holt and Brand
Miller (1995) and the study on high- and very low-AC
cooked milled parboiled rice fed to 23 Bangladeshis
by Tetens et al. (2003). In both studies, high-AC rice
preparation had higher satiety rating either expressed
through weight (Holt and Brand Miller 1995) or energy
(Tetens et al. 2003) of consumed food, but not both.
Recently, Felix et al. (2013) showed similar cooked
rice intakes in a short-term (120 min) satiety study
using seven milled rice samples differing in AC, GI
and dry matter content.
Campaign to popularize brown rice or unpolished rice
consumption in the Philippines is being intensified.
Full Paper
AdR Felix et al.
Brown rice is considered a potential functional food for
the management of diabetes mellitus through its lower
GI, being of higher fiber content than milled rice
(Trinidad et al. 2014). Reports also claimed brown rice
to be more satiating/stomach-filling than milled rice
(BRADS 2009). However, in a study by Wang et al.
(2013), it was reported that brown rice and milled rice
of a single rice variety are equally satiating. In a pilot
study using the low-AC Sinandomeng and the interme-
diate-AC IR64, we also noted no significant differ-
ences in common cooked rice intake at 120 min post
-meal and SQ for hunger and fullness for milled and
brown rices (PhilRice Los Baños 2012–2013, un-
published).
Since AC and dietary fiber content affects GI, we
hypothesized that the two factors also affect the
satiating potential of rice. This study was thus
conducted to: i) compare the short-term satiety of
brown rice of modern Philippine rice varieties,
representing wider range of AC types, with its
corresponding milled form, and ii ) to determine the
relationship of satiety indexes with AC, GI and dietary
fiber content.
MATERIALS AND METHODS
Rice Samples and Sample Characterization
Four rice varieties were selected to represent four AC
types. These varieties include the waxy or glutinous
rice Improved Malagkit Sungsong 2 (IMS2), the
low-AC National Seed Industry Council (NSIC) Rc160,
the intermediate-AC IR64, and the high-AC Philippine
Seed Board (PSB) Rc10. All were grown at the
University of the Philippines Los Baños Central
Experiment Station during the 2012 dry season,
except for NSIC Rc160 which was sourced from the
Philippine Rice Research Institute (PhilRice) Central
Experiment Station in Muñoz, Nueva Ecija, also grown
in 2012 dry season. The rice samples had been
analyzed in detail at the Rice Chemistry and Quality
Laboratory, PhilRice Los Baños in 2013, to ensure
typical properties were present in the varieties
representing the four AC types (Felix et al. 2013).
Rough rice samples were aged for at least two months
and were dehulled in a Satake THU-35A dehuller
(Satake Engineering Co., Ltd., Tokyo, Japan) to
produce brown rice. A part of the brown rice for each
variety was further polished in a McGill miller no. 3
(Dayton Grain Machinery Mfg. Corp., Nialen, FL, USA)
to obtain milled rice. Portion of the milled rice was
ground in a Udy cyclone mill (Udy Corp., Fort Collins,
CO, USA) with 60-mesh sieve and analyzed for AC in
acetate buffer (Juliano et al. 1981) and in ammonium
buffer (Juliano et al. 2012). AC of IMS2 was also
determined in ammonium buffer using the same
concentration of iodine reagent used for non-waxy
samples but absorbance was read after 2 h standing
time to reduce amylopectin-iodine interference (Tuaño
et al. 2014). Triplicate six whole milled rice grains were
soaked in 10 mL 1.7% potassium hydroxide for 23 h at
room temperature and alkali spreading value was
determined based on Little et al. (1958).
Rice samples were cooked to uniform Instron cooked
rice hardness (1.3 ± 0.1 kg cm-2) based on the
procedure described by Felix et al. (2013). Soaking of
brown rice samples 15 min before cooking was also
conducted. Proximate analysis (AOAC 2005) and total
dietary fiber measurement (AOAC 1995) were
determined on all cooked rice samples to compute for
the amount of test rice/preload that would yield 50 g
available carbohydrates (CHO) for GI and satiety
measurements. GI was run at the Food and Nutrition
Research Institute, Department of Science and
Technology (FNRI, DOST) on 12 normal and
apparently healthy Filipino subjects as described by
Trinidad et al. (2014).
Satiety Measurement
The satiety study was implemented in conjunction with
the GI measurements at FNRI, DOST. The study was
conducted according to the guidelines of the
Declaration of Helsinki as well as procedures
approved by the Institutional Human Ethics Committee
of the FNRI, DOST. All subjects gave written voluntary
informed consent.
The subjects consisted of seven male and five female
non-diabetic, nominally healthy adult volunteers, aged
20-50 yr old, recruited from the pool of volunteers
maintained by the FNRI, DOST. All subjects which
had normal glucose tolerance, were performing light to
moderate physical activities and had normal nutritional
status as indicated by body mass index (BMI) within
the acceptable range of 20–25 kg m-2. Half of the 12
human volunteers were part of the previous satiety
study (Felix et al. 2013), but all subjects had similar
baseline characteristics.
The study employed the randomized crossover study
design. Fixed amount of cooked rice preload that
provided 50 g available CHO were given as breakfast
to the volunteers after an overnight fast (10–12 h). The
subjects were randomly assigned to the sequence of
the test rice preload given. Following the GI
measurement protocol, the subjects had a maximum
of two test rice days in a week, with one day interval.
On a separate day, a 240-mL standard glucose drink
(Medic Orange 50 Glucose Tolerance Test Beverage
Product No. 089, Medic Diagnostic Laboratory, Pasig
City, Philippines) was also given to the volunteers as
reference food. This reference food was well tolerated
by the subjects.
Each of the rice preloads as well as the glucose drink
were tested twice. Subjects had orientation on the
study protocol and were accustomed to the test rices
prior to actual feeding. Food intake and activities of the
subjects outside the feeding days were monitored
through a food and activity questionnaire to ensure
uniformity of subject characteristics.
Subjective satiety and 2-h post meal food intake were
used as indexes of rice satiety. In the current study,
subjective satiety were qualified using satiety quotient
(SQ) and satiety index (SI). SI was included to provide
Short-term satiety of brown and milled rice 53
(Eqn. 1)
a meaningful comparison with GI (Salmenkallio-
Marttila et al. 2009), and to determine its utility as
measure of rice satiety, based on ease of use of its
corresponding rating scale (RS).
Satiety Quotient (SQ). At the same time intervals that
the blood samples for GI measurement were collected
(every 15 min during the first hour and every 30 min
during the second hour), the subjective feeling of
satiety was determined. Four satiety parameters,
namely hunger, fullness, desire to eat and prospective
consumption were employed, each measured by a
100-mm visual analogue scale (VAS), i.e. horizontal
lines with extreme sensations anchored at the ends of
the lines (Green and Blundell 1996; Merrill et al. 2002).
VAS ratings for hunger (0=not hungry at all to
100=hungry as ever felt), fullness (0=not full at all to
100=very full), desire to eat (0=very weak to 100=very
strong) and prospective consumption (0=nothing at all
to 100=a large amount) were recorded.
Satiety quotient (SQ) for each of the four VAS
parameters was computed based on the method of
Drapeau et al. (2005) as employed by Felix et al.
(2013). The formula was based on the difference
between fasting VAS and mean post meal VAS within
120 min and the energy content of the samples. The
higher the SQ, the lesser is the feeling of being hungry
and the desire to eat, given the amount of preload rice
consumed. Perceived amount of food that can be
eaten is also lesser at a higher SQ. Meanwhile, feeling
of fullness is expressed in negative value since the
descriptions were anchored opposite the description of
the other three VAS parameters. The absolute value
(without the negative sign) is however used to interpret
SQ fullness, such that the higher the absolute value,
the higher the feeling of fullness.
Satiety Index (SI). At the same time that the subjects
gave the VAS ratings, overall satiety score was also
provided based on a seven-point rating scale (RS) by
Holt and Brand-Miller (1995). The RS ranged from -3
(extremely hungry) through 0 (no particular feeling) to
+3 (extremely full). Each score was expressed as the
change from the fasting rating and the area under the
curve (AUC) within the cumulative 2-h satiety meas-
urement was calculated using the trapezoidal rule,
ignoring negative AUC values. SI was computed as
follows:
Post Meal Common Cooked Rice Intake. Two hours
after giving the preload, ad libitum cooked rice
consumption was measured and used as another
index of short-term satiety. A common
intermediate-AC commercially available rice named
“Angelica” (21.1% AC) was cooked using a water-rice
ratio of 1.5 and was served with sautéed sergestid or
ginisang alamang as viand. All subjects were used to
eating the accompanying viand. Ten grams viand was
provided per 100 g cooked rice. Test foods were
served in uniform food containers, and were
accompanied by drinking water. The amount of
drinking water varied depending on the weight of test
rice being served, such that the final amount of test
rice and drinking water was 240 g (equivalent to the
amount of the reference food). Feeding was done in a
generally serene area to address the potential
confounding effects of environmental factors on satiety
responses. Subjects ate together in a long table, were
discouraged to talk with each other, and given a video
clip to watch while eating. Subjects were instructed to
uniformly chew the food and were given 10 min to
finish eating.
Data Analysis
Complete data were obtained from the 12 subjects.
Mean SQ for the four VAS parameters, SI and rice
consumption data were subjected to repeated
measures of ANOVA to determine differences across
rice forms and AC types. The relationship of the
satiety indexes with AC, GI, dietary fiber content and
other grain properties was tested using simple linear
correlation. All data were analyzed using the Statistical
Analysis System (SAS) Software (SAS version 9.0,
SAS Institute Inc., Cary, NC, USA). Results were
presented as means ± standard deviation (SD) or
standard error of the mean (SEM), as appropriate.
RESULTS AND DISCUSSION
Characteristics of Rice Samples and Rice Meals
Physicochemical and Cooked Properties of Milled
and Brown Rices. The physicochemical properties
and pasting characteristics of the four rice varieties
were within the expected values, including the protein
content of milled waxy rice IMS2 which was
significantly higher in the earlier study (Felix et al.
2013). AC values were also within range but was
slightly lower for the current batch of samples. Values
obtained with ammonium buffer iodine colorimetric
method (Juliano et al. 2012) were also lower than
those with the original acetate buffer method (Juliano
et al. 1981) (Table 1).
Water-rice ratio to obtain Instron cooked rice hardness
of 1.3 ± 0.1 kg cm-2 increased with AC. Similar
water-rice ratios (1.3–2.1) were obtained by Juliano
and Perez (1983) for milled rices. The obtained Instron
hardness value was lower than the previous value of
1.6 ± 0.1 kg cm-2 (Felix et al. 2013) at the same
water-rice ratios for the four milled rices, probably due
to their lower AC. Water-rice ratios increased with
increasing AC. Higher amount of water was used for
brown rice, to compensate for its slower cooking rate
(greater steam loss) than milled rice. Corresponding
water contents of cooked rice were 57.1-67.8% for
milled rice and 60.9-67.6% for brown rice. Water
content was lowest for waxy rice and highest for high-
AC rice. Energy content followed water content
negatively. Cooked rice stickiness significantly
correlated inversely with AC (r=–0.94, n=8, p=0.01)
and was highest for waxy rice (both brown and milled)
and lowest for PSB Rc10 (Table 1). Bulk density was
also highest for the waxy rice having the lowest water
SI =
satiety AUC of the test food
satiety AUC of the reference food
x 100
54 AdR Felix et al.
content but the differences across rice treatments
were insignificant. Brown rice had higher protein and
fat contents than milled rice (Trinidad et al. 2014).
Total dietary fiber was also higher in brown rice than in
milled rice while amount of available CHO was lower
(Table 1). The calculated energy content tended to
decrease with AC but was comparable between brown
and milled rices .
Glycemic Index across AC Types and Rice Forms.
GI decreased with AC for both milled and brown rices
(Table 1). GI was highest for waxy (0–2% AC) rice and
lowest for high-AC rice for both rice forms, but the
decrease in GI with increasing AC was lower for brown
rice than for milled rice. In the study of Trinidad et al.
(2013), Sinandomeng, with the lowest AC of 12.6%
among the six milled, non-waxy rices tested, had a GI
of 75 classified as high, while PSB Rc10, with the
highest AC of 27% with a GI of 50 was classified as
low. In brown rice form, the same study found the low-
AC Sinandomeng to have a 4% higher GI than the
intermediate-AC IR64 (GI of 55 for Sinandomeng
brown rice vs. 51 for IR64 brown rice).
There was also a decrease in GI due to the switch
from milled rice to brown rice. The decrease was
however greatest and significant for waxy and low-AC
rices with high GI (> 70). Intermediate-AC IR64 milled
rice with medium GI (56–69) showed lesser drop and
was insignificant when consumed as brown rice, but
high-AC PSB Rc10 had no significant decrease in GI
from milled rice to brown rice. This trend was
consistent with the study of Trinidad et al. (2013),
which found a much higher decrease in GI for the
low-AC Sinandomeng (75 in milled form to 55 in brown
rice) than the intermediate-AC IR64 (57 to 51).
Meanwhile, the reported decrease in GI from milled
rice to brown rice was from 75.9 ± 6.6 to 61.5 ± 4.7 for
low-AC Japanese rice (Ito et al. 2005), 63 to 59 in
Chinese rice (Zhang et al. 2011) and 66 ± 8 to 58 ± 8
in high-AC IR42 in healthy Canadian subjects and 61
± 8 to 39 ± 8 in diabetic Filipino subjects (Panlasigui
and Thompson 2006).
Test Rice and Post Meal Cooked Rice
Characteristics. The amount of cooked milled rice
and cooked brown rice that were given as test food/
preload ranged from 141-203 g (Table 2). The amount
decreased with increasing AC and was 8-30% higher
in brown rice than the corresponding milled rice. The
computed energy content of the preload was
comparable between brown and milled rices, and had
a range of 5.44-7.28 kJ g-1.
The intermediate-AC rice variety Angelica was used
as post meal rice. It had comparable water content as
the test rice and the same Instron cooked rice
hardness (1.3 ± 0.0 kg cm-2). Its cooked rice stickiness
(123 ± 20 g•cm) was comparable with IR64 and PSB
Rc10 milled and brown rices while its bulk density
(0.90 g mL-1) was similar to that of PSB Rc10 milled
rice.
Short-term Satiety Indexes of Brown Rice and
Milled Rice of Varying AC
Satiety Quotient for Hunger, Fullness, Desire to
Eat and Prospective Consumption. The SQ was
based on the 100-mm VAS. VAS ratings for hunger,
desire to eat and prospective consumption declined
with intake of the test food and steadily increased
through time, from 15 min up to 120 min post meal.
The opposite was noted for VAS rating for fullness.
VAS for the four parameters were comparable across
AC types for both milled and brown rices in each
observation period, but were significantly different
(p < 0.05) from the reference food (data not shown).
As indicated in Table 2, SQ was lowest for the glucose
beverage, which was twice to four times lesser than
Property Milled Rice Brown Rice
IMS2a NSIC Rc160 IR64 PSB Rc10 IMS2a NSIC Rc160 IR64 PSB Rc10
ACb,c (ammonium buffer, %) 0.4 ± 0.1 12.3 ± 0.3 17.6 ± 0.1 22.1 ± 0.4 0.4 11.8 16.9 21.2
AC (acetate buffer, %) 0.6 14.9 19.3 23.8 0.6 14.3 18.5 22.8
Alkali spreading value 6.0 6.0 4.1 4.7 6.0 6.0 4.1 4.7
Water-rice ratio 1.2 1.4 1.5 2.1 1.7 1.8 2.0 2.5
Mean water content (%) 58.4 61.0 62.4 68.2 53.9 59.9 61.4 66.1
Instron hardnessc,d (kg cm-2) 1.4 ± 0.0a 1.3 ± 0.2a 1.3 ± 0.0a 1.2 ± 0.1a 1.3 ± 0.1a 1.2 ± 0.0a 1.2 ± 0.0a 1.3 ± 0.0a
Instron stickinessc,d (g·cm) 200 ± 15a 161 ± 3c 108 ± 21de 101 ± 9e 190 ± 8ab 169 ± 7bc 131 ± 24d 123 ± 20de
Mean bulk densityd (g mL-1) 0.97a 0.81de 0.84cd 0.88b 0.86cd 0.79e 0.82de 0.79e
Dietary fiber content (%) 0.4 1.2 0.5 0.7 3.3 3.4 2.9 2.9
Available carbohydrates (%) 38.2 33.7 33.0 28.7 29.4 29.0 28.5 25.5
Rice with 50 g avail. CHO (g) 131 148 152 174 170 172 175 196
Energy content (kcal g-1) 1.74 1.55 1.51 1.30 1.64 1.57 1.50 1.31
Energy content (kJ g-1) 7.28 6.49 6.32 5.44 6.87 6.57 6.28 5.48
Glycemic indexd,e,f 94 ± 5e 85 ± 3de 69 ± 4bc 59 ± 3a 77 ± 5cd 69 ± 4bc 61 ± 3ab 57 ± 3a
Table 1. Physicochemical and cooked properties and glycemic index of four milled and brown rices differing in apparent
amylose content (AC) in 12 normal Filipino subjects
aIMS2-Improved Malagkit Sungsong 2; bWaxy 0–2%; low 10–17%; intermediate 17–22%; high > 22%. AC of brown rice = 0.96 x AC of milled rice; cMean ±
standard deviation; dMeans in the same row followed by the same letter are not significantly different at P = 5% by LSD test; eMean ± standard error of the mean
(SEM); fHigh > 70; medium 56–69; low < 55.
Short-term satiety of brown and milled rice 55
the SQ for hunger and fullness of milled and brown
rices. This result suggests that liquid foods elicit
weaker satiety signals than solid foods and agrees
with those noted by Flood-Obbagy and Rolls (2009),
Leidy et al. (2010) and Willis et al. (2011). Among
milled samples, SQ for hunger was similar across rice
varieties, confirming earlier results (Felix et al. 2013).
SQ for fullness, desire to eat and prospective
consumption were also comparable across rice types.
The same trend was noted for brown rice. Between
cooked milled rice and cooked brown rice, SQ for the
four subjective measures tended to be higher for
brown rice, though the differences were not significant
(Table 2). These findings were in support of the study
by Wang et al. (2013) which also found higher
fullness, and lower hunger and prospective food intake
satiety ratings 240 min post meal for milled and brown
rice as compared to glucose beverage, and noted no
differences between milled and brown rice preloads.
The trend between milled and brown rices was also
consistent with the results of our earlier pilot study
(PhilRice Los Baños 2012–2013 unpublished).
Correlation analysis revealed the negative association
between SQ for hunger and SQ for fullness (r = –0.72,
n = 8, p = 0.05), while there was a direct association
between SQ for desire to eat and SQ for prospective
consumption (r = 0.75, n = 8, p = 0.05). These trends
were also in accordance to the physiological
explanation of satiety according to the glucostatic
theory (Drapeau et al. 2005). Based on the duplicate
blood glucose and VAS ratings of the 12 subjects for
seven-time periods for the 8 rice treatments and at 5%
level of significance, blood glucose level was found to
be significantly and negatively correlated with VAS
scores for hunger (r = –0.348), desire to eat
(r = –0.381) and perceived amount of next food intake
(r = –0.327). Blood glucose was, on the other hand,
positively correlated with VAS fullness scores
(r = 0.362).
Satiety Index of Brown and Milled Rices. There was
no significant difference in the SI of the four rice types,
either in milled or brown rice form (Figure 1). On the
average, the SI of the four milled rice varieties and the
brown rice samples was 3.7 and 4.2 times higher,
respectively, than the SI (100%) of the standard
glucose drink. Holt and Brand Miller (1995) reported a
comparable SI of 132% for brown rice and 138% for
„white‟ rice, using 1,000 kJ (240 kcal) portion of the
test foods and having white bread as reference food.
Similar to the findings of Holt and Brand Miller (1995),
the SI of brown rice tended to be higher than the
corresponding milled rice, but the difference was not
significant. Interestingly however, the brown rice of the
waxy variety IMS2 had the highest SI followed by the
intermediate-AC NSIC Rc160 milled rice (Figure 1).
This may suggest the complicating effects of cooked
rice texture (hardness or stickiness or both) in satiety
measurements despite the attempt to yield
comparable Instron cooked rice hardness of the
samples in this study. Follow-up study is needed to
elucidate this insight.
SQ and SI are subjective satiety measures that are
independent of each other. SQ is relative to the energy
content of the test food, while SI is relative to the AUC
Property Milled Rice Brown Rice
IMS2a NSIC Rc160 IR64 PSB Rc10 IMS2a NSIC Rc160 IR64 PSB Rc10
Mean satiety quotients (SQ)b,c
Hunger (mm kJ-1) 2.6 ± 0.4a 2.0 ± 0.5a 2.1 ± 0.3a 1.7 ± 0.5a 2.7 ± 0.5a 3.2 ± 0.5a 2.5 ± 0.5a 2.2 ± 0.0a
Fullness (mm kJ-1) -2.1 ± 0.4a -2.0 ± 0.5a -2.4 ± 0.3a -2.0 ± 0.5a -2.4 ± 0.5a -2.5 ± 0.5a -2.4 ± 0.5a -2.7 ± 0.4a
Desire to eat (mm kJ-1) 2.6 ± 0.4a 2.2 ± 0.5a 2.0 ± 0.3a 1.8 ± 0.5a 2.9 ± 0.5a 2.3 ± 0.5a 1.8 ± 0.4a 2.2 ± 0.3a
Prospective consumption (mm kJ-1) 2.0 ± 0.3a 1.8 ± 0.5a 1.3 ± 0.3a 1.6 ± 0.5a 2.2 ± 0.5a 1.8 ± 0.5a 1.7 ± 0.4a 2.1 ± 0.3a
Mean cooked rice intake
Test meald (g) 141 149 148 143 203 178 175 172
Post mealb,e (g) 252 ± 31a 236 ± 41a 200 ± 28a 208 ± 26a 254 ± 49a 258 ± 38a 233 ± 33a 223 ± 28a
Totalb,e (g) 393 ± 31a 385 ± 41a 348 ± 28a 351 ± 26a 457 ± 49a 436 ± 38a 408 ± 33a 395 ± 28a
Table 2. Short-term satiety quotients of four milled and brown rices differing in apparent amylose content (AC) in 12 normal
Filipino subjects and mean cooked rice intakes as satiety indicators
aIMS2-Improved Malagkit Sungsong 2; bMean ± standard error of the mean (SEM). Means in the same row followed by the same letter are not significantly differ-
ent at P = 5% by LSD test; cSQ of standard glucose drink is 0.7 (hunger), -0.9 (fullness), 1.0 (desire to eat) and 1.1 (prospective consumption). SQ of standard
glucose drink is significantly lower than milled and brown rices at P = 5% by LSD test; dWith 50 g available carbohydrates. Recalculated to 61.3% water content
of Angelica cooked milled rice to be comparable with post meal; ePost meal rice intake for standard glucose drink is 342 ± 47 g which is significantly higher than
the post meal intake for milled and brown rices at P = 5% by LSD test;
Figure 1. Satiety index of four Philippine milled and brown
rices differing in apparent amylose content (AC)
type and glycemic index (GI) and of standard
glucose drink in 12 normal Filipino subjects
56 AdR Felix et al.
of the reference food. However, our subjects noted
ease of using the seven-point RS than the 100-mm
VAS that were used to compute SI and SQ,
respectively. The seven-point RS also had a greater
test-retest reproducibility than VAS (Holt and
Brand-Miller 1995), and RS at different time points
correlated well with blood glucose level (p < 0.05), just
like VAS. This may suggest that SI is a more
appropriate measure of rice satiety among Filipinos.
Yet, this has to be validated especially since the
review by Salmenkallio-Marttila et al. (2009)
emphasized that “the wording printed on the scale to
describe the relevant satiety sensations on either side
of the central point or at the extremes of the scale is
more important than the type of scale”. The
applicability of SQ and SI as a screening tool for
satiety in the Philippine rice breeding program has to
be looked into.
2-h Post Meal Common Cooked Rice Intake. The
consumption of common cooked milled rice (Angelica)
120 min after the intake of preload (i.e. after GI
measurements) was highest (342 ± 47 g) for glucose
drink/reference food, and was significantly different (p
< 0.05) with those of the test rices. As presented in
Table 2, post meal common rice intake was
comparable among the four samples of milled rice,
consistent with earlier findings (Felix et al. 2013). This
also concurs with the study of Aston et al. (2009) and
the review of Niwano et al. (2009) which reported no
correlation between food intake and GI. Intake was
also similar across AC types for brown rice, while
viand intake was uniform among rice treatments.
The 2-h post meal common rice intake between milled
and brown rice was also comparable (Table 2). The
higher SQ and SI for brown rice than milled rice was
not translated into lower common cooked rice intake,
contrary to expectation, but in support of our initial
observation (PhilRice Los Baños 2012–2013,
unpublished). The adjustment in cooked rice hardness
via water-rice ratio and presoaking of brown rice
samples would have affected the results, as the
similarity in cooked rice hardness may result in uni-
form extent of digestion (Han and Lim 2009).
Although the differences in post meal common cooked
rice intake between milled and brown rice were not
significant, intake after 2 h tended to be higher with the
brown rice preload. The relatively shorter interval be-
tween breakfast and post meal may be a factor and a
feeding study that will extend up to lunch time and
even dinner will be undertaken. The use of the same
rice variety for the test meal will also be appropriate in
the follow-up study, since rice energy content was not-
ed to decrease with increasing AC (Table 1).
The trend in post meal cooked rice intake also
suggests that other factors may override physiological
hunger that mainly dictates food intake regulation.
Cooked (boiled) brown rice is a less familiar food than
milled rice for Filipinos, and was rated lower than
milled rice in terms of overall quality (p < 0.01). Since
eating brown rice is a less pleasant experience,
subjects may tend to compensate eating satisfaction
by consuming larger amount of the habitually common
milled rice that was served as post meal. This
observation relates to cognitive cues and sensory
specific-satiety that are equally important as physio-
logical influences in regulating food intake
(Salmenkallio-Marttila et al. 2009).
In another study conducted parallel to this research,
which used the same post meal rice as the preload, an
opposite overall trend was found, i.e. lower rice intake
(wet basis) at lunch time (~240 min post meal) for
brown rice, but the difference remained insignificant
(PhilRice Los Baños 2013–2014 unpublished).
Increased chewing was also shown to reduce meal
size and food intake in normal-weight, overweight and
obese adults (Zhu and Hollis 2013). Since subjects
reported less chewing and more swallowing with
brown rice, reduction in food intake may not be
expected, as noted in the current study. Further
studies are needed to validate such insights. The use
of physiological biomarkers of appetite and satiety in
follow-up study is necessary to substantiate findings.
Rice AC, GI and DF Content in Relation to Rice
Short-term Satiety Indexes
The relationship of AC, GI and DF to short-term satiety
indexes namely: SQ, SI and 2-h post meal common
cooked rice intake, was determined using simple
correlation analysis. As earlier reported by Felix et al.
(2013), SQ for hunger and post meal cooked rice
intake were independent of milled rice AC and GI.
Both AC and GI were also found not to be related with
the other SQ parameters namely: fullness, desire to
eat, and prospective consumption. Meanwhile, DF was
positively correlated with SI (r = 0.92, n = 8, p < 0.01),
but not with SQ for all parameters, nor with 2-h post
meal food intake.
Brown rice has lower GI than milled rice even though
they both contain the same starch content (Atkinson et
al. 2008; Trinidad et al. 2014). Brown rice has higher
DF than milled rice. DF causes the slow release of
glucose after food consumption. This may contribute
to increased satiety and lower food intake that are
important in the proper control and management of
overweight/obesity and type 2 diabetes mellitus
(Trinidad et al. 2013). Contrary however to this
expectation, 2-h post meal common cooked rice intake
were comparable between rice forms (milled rice vs.
brown rice), regardless of AC type. Aside from other
possible explanations provided above, it is also most
possible that the additional 2.2-2.9 g dietary fiber in
brown rice is insufficient to affect subsequent food in-
take. The study of Wang et al. (2013) reported similar
findings while the meta-analysis of 22 published
researches (Howarth 2001) showed that a 10%
reduction in energy intake was possible in normal
individuals only with an average 14 g dietary fiber
intake. Most recent review by Clark and Slavin (2013)
found that fibers in the range 30.0-49.9 g decreased
food intake. Both reviews did not include rice as
source of dietary fiber. In addition, not all fibers influ-
ence satiety equally (Willis et al. 2009) and most fibers
Short-term satiety of brown and milled rice 57
do not affect appetite or energy intake in acute study
designs (Clark and Slavin 2013). Clark and Slavin
(2013) also underscored evidences that β-glucan
(from oats or barley), lupin kernel fiber, rye bran,
whole grain rye, or a mixed high-fiber diet may
decrease appetite more frequently than other fiber
types. Additionally, psyllium, whole grain barley, and
wheat bran had a low-to-negative effect on VAS
ratings, food intake, or both in the case of psyllium.
DF content did not positively affect satiating potential
of rice. Switching from milled rice to brown rice did not
lower food intake after 2 h. This finding forwarded the
need for caution in generalizing or claiming the
beneficial effects of brown rice consumption in relation
to food intake, and prevention and management of
diabetes mellitus and its risk factors like overweight
and obesity. Yet, the perceived better satiating
potential of brown rice than milled rice (i.e. higher SQ
and SI) may provide opportunities to maximize the
promotion of brown rice as a functional food, if taken
as entry point for nutrition education and as part of a
holistic dietary intervention. Follow-up studies of
longer duration are needed to establish effects of
brown rice consumption to satiety, food intake and
weight management in relation to type 2 diabetes
mellitus prevention.
CONCLUSION
The study confirmed earlier findings that the short-
term satiety of milled rice is independent of its AC and
GI. Brown rice tended to have higher perceived satiety
than corresponding milled rice. But contrary to
hypothesis, switching from milled rice to brown rice did
not lower food intake after 2 h. Follow-up studies of
longer duration are needed to establish effects of
brown rice consumption to satiety and food intake.
Further researches that will deal on role of cognitive
cues, and sensory attributes on rice satiety are
suggested, together with the use of physiological
biomarkers.
ACKNOWLEDGMENT
This research was funded in part by PhilRice Central
Experiment Station, Maligaya, Muñoz, Nueva Ecija
3119, Philippines. The assistance of Ruffa R. Solano,
Anna Ruby B. Macahiya, Antonieta D. Peñaloza and
Wilson E. Rivano of PhilRice Los Baños Rice
Chemistry and Quality Laboratory, and Mark Ryan Q.
Ibardaloza of FNRI, DOST are acknowledged.
LITERATURE CITED
[AOAC] Association of Official Analytical Chemists.
1995. Official methods of analysis. 991.43
Supplement. Arlington, VA, USA: Assoc Offi-
cial Anal Chemists. 3 p.
[AOAC] Association of Official Analytical Chemists.
2005. Official methods of analysis. 18th ed.
Arlington, VA, USA: Assoc Official Anal
Chemists.
Aston LM, Stokes CS, Jebb SA. 2008. No effect of a
diet with a reduced glycaemic index on sa-
tiety, energy intake and body weight in over-
weight and obese women. Intern J Obesity
32:160-165.
[BRADS] Brown Rice Advocates. 2009. All about
brown rice. Los Baños, Laguna, Philippines:
Brown Rice Advocates. 12 p.
Brouns F, Bjork I, Frayn KN, Gibss AL, Lang V, Slama
G, Wolever TMS. 2005. Glycemic index
methodology. Nutr Res Rev 18: 145-171.
Clark MJ, Slavin JL. 2013. The Effect of Fiber on Sa-
tiety and Food Intake: A Systematic Review.
J Am Coll Nutr 32:200-211.
Drapeau V, Blundell J, Therrien F, Lawton C, Richard
D, Tremblay A. 2005. Appetite sensations as
a marker of overall intake. Brit J Nutr 93:273-
280.
Dixit AA, Azar KM, Gardner CD, Palaniappan LP.
2011. Incorporation of whole, ancient grains
into a modern Asian Indian diet to reduce the
burden of chronic disease. Nutr Rev 69:479-
488.
Felix AdR, Trinidad TP, Tuaño APP, Juliano BO. 2013.
Short-term satiety of cooked Philippine rices
of varying apparent amylose content and gly-
cemic index. Philipp Agric Scientist 96:46-53.
Flood-Obbagy JE, Rolls BJ. 2009. The effect of fruit in
different forms on energy intake and satiety at
a meal. Appetite 52:416-422.
Green SM, Blundell J. 1996. Subjective and objective
indices of the satiating effects of foods. Europ
J Clin Nutr 50:798-806.
Han JA, Lim ST. 2009. Effect of presoaking on textur-
al, thermal, and digestive properties of
cooked brown rice. Cereal Chem. 86:100-
105.
Holt SHA, Brand Miller J. 1995. Increased insulin re-
sponses to ingested food are associated with
lessened satiety. Appetite 24:43-54.
Howarth NC, Saltzman E, Roberts S. 2001. Dietary
fiber and weight regulation. Nutr Rev 59: 129-
139.
Hu EA, Pan A, Malik V, Sun Q. 2012. White rice con-
sumption and risk of type 2 diabetes: meta-
analysis and systematic review. Brit Med J
344(e1454):1-9.
Ito Y, Mizukuchi A, Kise M, Aoto H, Yamamoto S, Yo-
shihara R, Yokoyama J. 2005. Postprandial
blood glucose and insulin responses to pre-
58 AdR Felix et al.
germinated brown rice in healthy subjects.J
Med Invest 52:159-164.
Juliano BO, Perez CM, Blakeney AB, Castillo DT,
Kongseree N, Laignelet B, Lapis ET, Murty
VVS, Paule CM, Webb BD. 1981. Internation-
al cooperative testing on the amylose content
of milled rice. Starch/Stärke 33:157-162.
Juliano BO, Perez CM. 1983. Major factors affecting
cooked milled rice hardness and cooking
time. J Texture Stud 14:235-243.
Juliano BO, Tuaño APP, Monteroso DN, Aoki N, Mes-
tres C, Duldulao JBA, Bergonio KB. 2012.
Replacement of acetate with ammonium buff-
er to determine apparent amylose content of
milled rice. Cereal Foods World 57:14-19.
Leidy HJ, Apotzan JW, Mattes RD, Campbell WW.
2010. Food form and portion size affect post-
prandial appetite sensations and hormonal
responses in healthy, nonobese, older adults.
Obesity 18:293-299.
Little RR, Hilder GB, Dawson EH. 1958. Differential
effect of dilute alkali on 25 varieties of milled
white rice. Cereal Chem 35:111–126.
Ludwig DS. 2000. Dietary glycemic index and obesity.
J Nutr. 130:280S-283S.
Merrill EP, Kramer FM, Cardello AV, Schutz HG.
2002. A comparison of satiety measures. Ap-
petite 39:181-183.
Niwano Y, Adachi T, Kashimura J, Sakata T, Sasaki
H, Sekine K, Yamamoto S, Yonekubo A, Ki-
mura S. 2009. Is glycemic index of food a
feasible predictor of appetite, hunger and sa-
tiety? J Nutr Sci Vitaminol 55: 201-207.
Panlasigui LN, Thompson LU. 2006. Blood glucose
lowering effects of brown rice in normal and
diabetic subjects. Int J Food Sci Nutr 57:151-
158.
Priebe M, Van Binsbergen J, De Vos R, Vonk RJ.
2009. Whole grain foods for the prevention of
type 2 diabetes mellitus. Cochrane database
for systematic reviews 2008. Issue 1. Art. No.
CD006061.doi:10.1002/14651858.CD006061
.pub.2.
Salmenkallio-Martilla M, Due A, Gunnarsdottir I,
Karhunen L, Saarela M, Lyly M. 2009.Satiety,
Weight Management and Foods: Literature
Review. Oslo, Norway: Nordic Innovation
Centre. 164 p.
Shaw JE, Sicree RA, Zimmet PZ. 2010. Global esti-
mates of the prevalence of diabetes for 2010
and 2030. Diabetes Research and Clinical
Practice 87:4-14.
Sluijs I, Beulens JWJ, Van Der Daphne LA, Spijker-
man AMW, Grobbee DE, Van Der Schouw
YT. 2010. Dietary intake of total, animal, and
vegetable protein and risk of type 2 diabetes
in the European Prospective Investigation
into Cancer and Nutrition (EPIC)-NL Study.
Diabetes Care 33:43-48.
Sun Q, Spiegelman D, Van Dam RM, Holmes MD,
Malik VS, WIllett WC, Hu FB. 2010. White
rice, brown rice and risk of type 2 diabetes in
US men and women. Arch Intern Med
170:961-969.
Tetens I, Kabir KA, Parvin S, Thilsted SH. 2003. Differ-
ential rates of energy release from rice and
effects on satiety. In: TM Mew, DS Brar, D
Dave, B Hardy, editors. Rice Science Innova-
tions and Impact for Livelihood. Proc. Interna-
tional Rice Research Conference 2002, Bei-
jing, China. Pp 421-430.
Trinidad TP, Mallillin AC, Sagum RS, Felix AdR, Juli-
ano BO. 2013. The effect of apparent amyl-
ose content and dietary fibre on the glycemic
response of different varieties of cooked
milled and brown rice. Intern J Food Sci Nutr
64:89-93.
Trinidad TP, Mallillin AC, Sagum RS, Felix AdR, Tua-
ño APP, Juliano BO. 2014. Relative effect of
apparent amylose content on the glycemic
index of milled and brown rice. Philipp Agric
Scientist 97:405-408.
Tuaño APP, Aoki N, Fujita N, Oitome NF, Merca FE,
Juliano BO. 2014. Grain and starch proper-
ties of waxy and low-apparent amylose Philip-
pine rices and of NSIC Rc222. Philipp Agric
Scientist 97:329-339.
Wang XS, Neill MO, Thomas W, Slavin JL. 2013.
White and brown rice are equally satiating
and more satiating than glucose beverage. J
Obes Weight Loss Ther 3: 201.
Willis HJ, Eldridge AL, Beiseigel J, Thomas W, Slavin
JL. 2009. Greater satiety response with re-
sistant starch and corn bran inhuman sub-
jects. Nutr Res 29:100-105.
Willis HJ, Thomas W, Willis DJ, Slavin JL. 2011. Fea-
sibility of measuring gastric emptying time
with a wireless motility device, after subjects
consume fiber matched liquid and solid
breakfast. Appetite. 57:38-44.
Zhu Y, Hollis JH. 2014. Increasing the Number of
Chews before Swallowing Reduces Meal
Size in Normal-Weight, Overweight, and
Obese Adults J Acad Nutr Diet 114:926-931.
Short-term satiety of brown and milled rice 59
