Content uploaded by Hitomi Okubo
Author content
All content in this area was uploaded by Hitomi Okubo on Feb 18, 2014
Content may be subject to copyright.
Dietary glycemic index and load in relation to metabolic risk factors
in Japanese female farmers with traditional dietary habits
1–3
Kentaro Murakami, Satoshi Sasaki, Yoshiko Takahashi, Hitomi Okubo, Yoko Hosoi, Hyogo Horiguchi, Etsuko Oguma,
and Fujio Kayama
ABSTRACT
Background: Little is known about the relation of dietary glycemic
index (GI) and glycemic load (GL) to metabolic risk factors, partic-
ularly in non-Western populations.
Objective: We examined the cross-sectional associations between
dietary GI and GL and several metabolic risk factors in healthy
Japanese women with traditional dietary habits.
Design: The subjects were 1354 Japanese female farmers aged
20 –78 y from 5 regions of Japan. Dietary GI and GL were assessed
with a self-administered diet-history questionnaire. Body mass in-
dex (BMI) was calculated as weight (kg) divided by the square of
height (m). Fasting blood samples were collected for biochemical
measurements.
Results: The mean dietary GI was 67, and the mean dietary GL
(/1000 kcal) was 88 (GI for glucose ҃ 100). White rice (GI ҃ 77) was
the major contributor to dietary GI and GL (58.5%). After adjust-
ment for potential dietary and nondietary confounding factors, di-
etary GI was positively correlated with BMI (n ҃ 1354; P for trend ҃
0.017), fasting triacylglycerol (n ҃ 1349; P for trend ҃ 0.001),
fasting glucose (n ҃ 764; P for trend ҃ 0.022), and glycated hemo-
globin (n ҃ 845; P for trend ҃ 0.038). Dietary GL was independently
negatively correlated with HDL cholesterol (n ҃ 1354; P for trend ҃
0.004) and positively correlated with fasting triacylglycerol (P for
trend ҃ 0.047) and fasting glucose (P for trend ҃ 0.012).
Conclusions: Both dietary GI and GL are independently correlated
with several metabolic risk factors in subjects whose dietary GI and
GL were primarily determined on the basis of the GI of white
rice. Am J Clin Nutr 2006;83:1161–9.
KEY WORDS Glycemic index, glycemic load, white rice,
body mass index, triacylglycerol, glucose, glycated hemoglobin,
HDL cholesterol, Japanese women, epidemiology, Japanese Multi-
centered Environmental Toxicants Study, JMETS
INTRODUCTION
Dietary carbohydrates are typically categorized into simple
sugars and complex carbohydrates on the basis of their degree of
polymerization. Their effects on health, however, may be better
categorized according to their physiologic effects, specifically
their ability to raise blood glucose (1), because the blood glucose
response varies substantially among different carbohydrate-
containing foods and cannot be predicted by their chemical com-
position (2). This varied glycemic response is quantified accord-
ing to the glycemic index (GI), which is a measure of how much
each carbohydrate-containing food raises blood glucose com-
pared with a standard food of either glucose or white bread (per
50 g available carbohydrate) (3). In consideration of the amounts
of carbohydrate-containing foods and total dietary carbohydrate,
the concept of glycemic load (GL: GI ҂ available carbohydrate
content) has also been proposed (4, 5).
Recent results from a limited number of observational studies
have suggested that diets with a low GI, a low GL, or both have
a beneficial effect on several metabolic risk factors for cardio-
vascular disease and type 2 diabetes, such as body mass index
(BMI; in kg/m
2
) (6), HDL cholesterol (7–11), triacylglycerol (9,
10, 12), and glycated hemoglobin (Hb A
1c
) (12, 13). However,
almost all studies of dietary GI or dietary GL and metabolic risk
factors have been conducted in Western countries, whereas, to
our knowledge, only one small study (10) was carried out in
Asian countries, including Japan.
For Japanese people, rice is the food that contributes most to
total carbohydrate and energy intake (43% and 29%, respec-
tively), which is a characteristic seldom observed in Western
people (14). Therefore, a different correlation of dietary GI or
dietary GL and metabolic risk factors may exist between Western
and Japanese populations. Additionally, whereas cardiovascular
disease is the second leading cause of all death in Japan (15), the
number of Japanese people with type 2 diabetes is estimated to be
no fewer than 6.8 million (16); thus, as is the case in Western
people, these are serious health problems in Japan. Conse-
quently, we examined the cross-sectional associations between
dietary GI and GL and several metabolic risk factors for cardio-
vascular disease and type 2 diabetes, including BMI, fasting
1
From the National Institute of Health and Nutrition, Tokyo, Japan (KM,
SS, YT, and YH); the Department of Nutrition Sciences, Kagawa Nutrition
University, Saitama, Japan (HO); the Division of Environmental Medicine,
Center for Community Medicine, Jichi Medical University, Tochigi, Japan
(HH, EO, and FK); and Core Research for Evolutional Science and Tech-
nology, Japan Science and Technology Cooperation, Kawaguchi City, Japan
(EO and FK).
2
Supported mainly by grants from the Japanese Ministry of Health, Labor
and Welfare, the Ministry of Agriculture and Forestry, and Core Research for
Evolutional Science and Technology, Japan Science and Technology Coop-
eration.
3
Address reprint requests to F Kayama, Division of Environmental Med
-
icine, Center for Community Medicine, Jichi Medical University, 3311-1
Yakushiji, Minami-Kawachi, Kawachi-gun, Tochigi 329-0498, Japan. E-
mail: kayamaf@jichi.ac.jp.
Received November 28, 2005.
Accepted for publication January 17, 2006.
1161Am J Clin Nutr 2006;83:1161–9. Printed in USA. © 2006 American Society for Nutrition
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
serum triacylglycerol, fasting plasma glucose, Hb A
1c
, and serum
total, HDL, and LDL cholesterol in a group of apparently healthy
Japanese women.
SUBJECTS AND METHODS
Subjects
The subjects in the present study were participants in the Jap-
anese Multi-centered Environmental Toxicants Study (JMETS),
the main purpose of which was to identify the threshold concen-
tration in the dose-response relation of cadmium renal dysfunc-
tion (17, 18). For this purpose, the JMETS was conducted in
female farmers in 4 moderately cadmium-polluted areas and 1
non-cadmium-polluted area in Japan; however, no difference in
the effects of environmental exposure to cadmium was observed
between the 4 polluted areas and 1 nonpolluted area, at least
regarding renal function and bone density (17, 18). Thus, the
study did not identify evidence that environmental exposure to
cadmium, at the level found in the 4 polluted areas, has an adverse
affect on health. The 5 areas surveyed consist of rural agricultural
communities with inhabitants who remain in the community
even after marriage. Thus, most of the farmers in these areas are
assumed to have maintained traditional Japanese dietary pat-
terns, consuming their own crops, including rice, for decades.
During the winters of 2000 and 2001, female farmers in each area
were recruited through the local Agricultural Cooperative to
participate in a medical examination organized for the JMETS.
One week before the examination, group orientations were held
for the study participants, at which the study purpose and proto-
col were explained and written informed consent was obtained
from each participant. In addition, participants were instructed
on how to complete questionnaires regarding diet and other life-
style factors and were asked to bring them to the examination.
The protocol of the JMETS was approved by the ethical com-
mittee of Jichi Medical University. Additional details about the
JMETS were reported elsewhere (17, 18).
A total of 1407 women aged 20 –78 y completed both a med-
ical examination and the lifestyle-related questionnaires. Sub-
jects excluded from the present study were those with previously
diagnosed diabetes (n ҃ 15) or cardiovascular disease (n ҃ 18),
those with extremely low or high energy intakes (쏝600 or 쏜4000
kcal/d; n ҃ 10), and those with missing covariate information
(n ҃ 4). Furthermore, subjects with missing information regard-
ing dependent variables, were excluded from the analysis of LDL
cholesterol (n ҃ 6), glucose (n ҃ 609), and Hb A
1c
(n ҃ 527), and
subjects who ate breakfast before blood was drawn were ex-
cluded from the analysis of fasting triacylglycerol and glucose
(n ҃ 5). Thus, the final sample was 1354 for BMI and serum total
and HDL cholesterol, 1348 for serum LDL cholesterol, 1349 for
fasting serum triacylglycerol, 764 for fasting plasma glucose,
and 845 for Hb A
1c
; however, some subjects were included in
more than one exclusion category. Further exclusion of subjects
with a diagnosis of hyperglycemia, dyslipidemia, hypercholes-
terolemia, or a combination thereof (n ҃ 24 for BMI, cholesterol,
and triacylglycerol and n ҃ 17 for glucose and Hb A
1c
) did not
alter the findings of the present study; therefore, these subjects
were included in the analyses.
Metabolic risk factors
At the medical examination site, each subject’s weight (mea-
sured while wearing light clothes and no shoes) was measured
with a set of balance scales calibrated to 0.01 kg. Body height was
also measured at the site. The BMI of each subject was calculated
as weight (kg) divided by the square of height (m). Peripheral
blood samples were obtained from subjects after an overnight
fast. Blood was collected in evacuated tubes containing no ad-
ditives, allowed to clot, and centrifuged at 3000 ҂ g for 10 min
at room temperature to separate the serum. Blood samples for
blood sugar measurement were collected in hydrogen fluoride–
containing tubes. All of the following biochemical variables of
the samples were assayed at Mitsubishi Kagaku Bio-Clinical
Laboratories Inc (Itabashi, Tokyo, Japan) within3dofcollection
to avoid significant degradation. Total cholesterol, HDL choles-
terol, and triacylglycerol were measured by enzymatic assay
methods. Serum LDL-cholesterol concentrations were calcu-
lated by using the Friedewald equation (19) for subjects with
fasting serum triacylglycerol concentrations 쏝400 mg/dL. Hb
A
1c
was measured by latex agglutination–turbidimetric immu
-
noassay. In-house quality-control procedures for all of the
abovementioned assays were fulfilled at Mitsubishi Kagaku Bio-
Clinical Laboratories Inc.
Dietary assessment
Dietary habits during the past month were assessed with a
self-administered diet-history questionnaire (DHQ) (20–22),
which was completed by each subject at home and was checked
by 욷2 dietitians during the medical examination. The DHQ is a
16-page structured questionnaire that consists of the following 7
sections: general dietary behaviors, major cooking methods,
consumption frequency and portion size of 6 alcoholic bever-
ages, semiquantitative frequency of intake of 121 selected food
and nonalcoholic beverage items, dietary supplements, con-
sumption frequency and amount of 19 staple foods (rice, bread,
noodles, and other wheat foods) and miso (fermented soybean
paste) soup, and open-ended items for foods consumed regularly
(욷1 time/wk) but not appearing in the DHQ. The food and
beverage items and portion sizes in the DHQ were derived pri-
marily from data in the National Nutrition Survey of Japan and
several recipe books for Japanese dishes (20). Measures of di-
etary intake for 147 food and beverage items, energy, fat, total
carbohydrate, alcohol, and dietary fiber were calculated by using
an ad hoc computer algorithm developed for the DHQ, which was
based on the Standard Tables of Food Composition in Japan (23).
Information on dietary supplements and data from the open-
ended questionnaire items were not used in the calculation of
dietary intake. Detailed descriptions of the methods used for
calculating dietary intake and the validity of the DHQ were
published elsewhere (20 –22). Pearson’s correlation coefficients
between the DHQ and 3-d dietary records were 0.48 for energy,
0.55 for fat, and 0.48 for total carbohydrate in 47 women (20). In
addition, Pearson’s correlation coefficients between the DHQ
and 16-d dietary records were 0.79 for alcohol and 0.69 for
dietary fiber in 92 women (S Sasaki, unpublished observations,
2004).
Calculation of dietary GI and GL
The GI of a food is defined as the 2-h incremental area under
the blood glucose response curve after consumption of a food
portion containing a specific amount (usually 50 g) of available
carbohydrate, divided by the corresponding area after consump-
tion of a portion of a reference food (usually glucose or white
1162 MURAKAMI ET AL
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
bread) containing the same amount of available carbohydrate,
and multiplied by 100 to be expressed as a percentage (24). We
calculated dietary GI by multiplying the percentage contribution
of each individual food to daily available carbohydrate intake by
the food’s GI value and summed these products. Available car-
bohydrate was calculated as total carbohydrate minus dietary
fiber (24). We also calculated dietary GL by multiplying the
dietary GI by the total amount of daily available carbohydrate
intake (divided by 100).
To determine the GI value of each food for these calculations,
each food item on the DHQ was directly matched to foods in the
international table of GI (24), in several publications about the GI
of Japanese foods (25–27), and in a recent article about the GI of
potatoes (28). Glucose was used as the reference (GI for glu-
cose ҃ 100). The white bread– based GI values were transformed
into glucose-based GI values by multiplying the white bread–
based GI by 0.7, as in Western studies (24, 28), or by 0.73
[҃ 100/137 (white bread– based GI value of white bread/white
bread– based GI value of glucose)] as in Japanese studies (27).
The white rice– based GI values were transformed into glucose-
based GI values by multiplying white rice– based GI by 0.82 [҃
100/122 (white rice-based GI of white rice/white rice-based GI of
glucose)] (25, 26). When more than one GI value was available,
the mean GI values was used. Ten foods for which a GI value had
not been determined were assigned a value according to the
nearest comparable food, as follows: Chinese noodles were as-
signed the GI of instant noodles, Japanese-style pancakes were
assigned the GI of pizza, jellies were assigned the GI of pudding,
lotus roots were assigned the GI of carrots, vegetable juice was
assigned the GI of tomato juice, curry and roux in stew were
assigned the GI of white rice with curry, nutritional-supplement
drinks were assigned the GI of sports drinks, nutritional supple-
ment bars were assigned the GI of a sports bar, and ground
fish-meat products and boiled-fish, shellfish, and seaweed in soy
sauce were assigned the GI of fish fingers. Although alcoholic
beverages contain little carbohydrate, large quantities of several
alcoholic beverages, such as beer and sake, may raise glucose
concentrations slightly; however, by definition, the GI is based
on 50 g available carbohydrate. Thus, we ignored alcoholic bev-
erages during the calculation of dietary GI and GL. Furthermore,
foods with a very low available carbohydrate content were ex-
cluded because their GI values cannot be tested. The cutoff for
exclusion of foods was set at 3.5 g available carbohydrate per
serving (6). Of the total 147 food and beverage items included in
the DHQ, 6 (4.1%) are alcoholic beverages, 8 (5.4%) contain no
available carbohydrate, and 63 (42.9%) contain 쏝3.5 g available
carbohydrate per serving. The calculation of dietary GI and GL
was thus based on the remaining 70 items with GI values ranging
from 16 to 91. The GI value of each item is presented in Table 1.
In the present study, the available carbohydrate content of these
70 items contributed to 94.0 앐 2.5% (x 앐 SD) of total available
carbohydrate intake, which is comparable with previous studies
(6, 10).
Other variables
Smoking status, menopausal status, dietary supplement use dur-
ing the previous month, and rate of eating were self-reported in
questionnaires. Body weight at age 20 y was also self-reported,
and BMI at age 20 y was computed by dividing self-reported
weight (kg) at age 20 y by the square of current measured height
(m). In addition, the subjects reported the average times per week
spent on 13 activities such as sleeping, household-related activ-
ities, leisure-time sporting activities, and leisure-time sedentary
activities. The reported number of hours spent on each activity
(per week) was divided by 7 to obtain the mean number of hours
per day. For subjects whose recorded total hours per day were 쏝
or 쏜24 h, the total number of hours spent daily were proportion-
ately increased or decreased to equal 24. Each activity was as-
signed a metabolic equivalent (MET) value from a previously
published table (29, 30). The mean number of hours spent per day
on each activity was multiplied by the MET value of that activity,
and all MET-hour products were summed to give a total MET-
hour score for the day. Total energy expenditure was calculated
by multiplying the total MET-hour score by body weight. Phys-
ical activity level was calculated by dividing total energy expen-
diture by basal metabolic rate, which was estimated as standard
values of basal metabolic rate for Japanese women multiplied by
body weight (31).
Statistical analysis
Dietary GI and GL were examined in relation to the 7 meta-
bolic risk factors: BMI; serum total, HDL, and LDL cholesterol;
fasting serum triacylglycerol; fasting plasma glucose; and Hb
A
1c
. We used crude values for dietary GI and energy-adjusted
values for dietary GL (/1000 kcal) because, by definition, dietary
GI is a measure of carbohydrate quality, not quantity, whereas
dietary GL is a measure of the combination of carbohydrate
quality and quantity. The mean (앐SE) values for these metabolic
factors were calculated according to quintiles of dietary GI and
GL after multivariate adjustment for potential confounding vari-
ables. Confounding variables included residential area (5 cate-
gories), age (울39, 40 – 49, 50 –59, 60 – 69, and 욷 70 y), meno-
pausal status (premenopausal or postmenopausal), current
smoking (no or yes), dietary supplement use (no or yes), rate of
eating (fast, medium, or slow), physical activity level (quintiles),
energy intake (quintiles), percentage of energy as fat (quintiles),
alcohol intake (nondrinkers, 쏜0to쏝1% of energy, or 욷1% of
energy), and energy-adjusted intake (g/1000 kcal) of dietary
fiber (quintiles). In the analyses, except for the analysis of BMI,
current BMI (quintiles) and BMI at age 20 y (quintiles) were also
included as confounding variables. Linear trends with increasing
levels of dietary GI and GL were tested by assigning each par-
ticipant the median value for the category and modeling this
value as a continuous variable. All statistical analyses were car-
ried out by using SAS statistical software (version 8.2; SAS
Institute Inc, Cary, NC). All reported P values are 2-tailed, and a
P value 쏝0.05 was considered statistically significant.
RESULTS
Basic characteristics of the 1354 subjects are shown in Table
2. The mean intakes of protein, fat, and carbohydrate were
14.0%, 25.3%, and 59.0% of energy, respectively. The mean
dietary GI was 66.7 and the mean dietary GL was 88.0 (/1000
kcal; crude mean ҃ 167.7). White rice was the major contributor
to dietary GI and GL (58.5%), followed by confectioneries
(10.6%), fruit (6.7%), sugars (5.5%), bread (4.3%), noodles
(3.4%), other rice (3.2%), and potatoes (2.6%). Potential con-
founding variables of the 1354 subjects are shown in Table 3
according to quintiles of dietary GI and GL. Fewer women in the
higher quintiles of dietary GI used dietary supplements and more
were nondrinkers of alcohol. Women in the higher quintiles of
GLYCEMIC INDEX AND METABOLIC RISK FACTORS 1163
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
dietary GI had lower mean energy, fat, and dietary fiber intakes.
In addition, women in the higher quintiles of dietary GL had
higher mean values for age and physical activity level and lower
mean energy, fat, and dietary fiber intakes. Fewer women in the
higher quintiles of dietary GL were premenopausal, current
smokers, and dietary supplement users and more were nondrink-
ers of alcohol. Similar patterns were observed for potential con-
founding variables according to quintiles of dietary GI and GL
among the subjects included in the analyses of serum LDL cho-
lesterol (n ҃ 1348), fasting serum triacylglycerol (n ҃ 1349),
fasting plasma glucose (n ҃ 764), and Hb A
1c
(n ҃ 845) (data not
shown).
Multivariate-adjusted mean values for metabolic risk factors
across quintile categories of dietary GI and GL are shown in
Table 4. After adjustment for potential confounding variables,
dietary GI was significantly positively correlated with BMI
(mean difference between the lowest and highest quintiles ҃ 0.7;
P for trend ҃ 0.017), fasting serum triacylglycerol (mean differ-
ence ҃ 16.0 mg/dL; P for trend ҃ 0.001), fasting plasma glucose
(mean difference ҃ 6.4 mg/dL; P for trend ҃ 0.022), and Hb A
1c
(mean difference ҃ 0.2%; P for trend ҃ 0.038). No correlation
was observed between dietary GI and serum concentrations of
total, HDL, and LDL cholesterol.
In contrast, after control for potential confounding variables,
dietary GL was significantly negatively correlated with serum
HDL cholesterol (mean difference ҃Ҁ6.4 mg/dL; P for trend ҃
0.004) and positively correlated with fasting serum triacylglyc-
erol (mean difference ҃ 14.4 mg/dL; P for trend ҃ 0.047) and
fasting plasma glucose (mean difference ҃ 12.5 mg/dL; P for
trend ҃ 0.012). Other metabolic risk factors examined, including
TABLE 1
Glycemic index (GI) value of each food and beverage item used in the
present study
1
Food and beverage item GI
White rice 77
White rice with barley 67
White rice with germs 66
50% Polished rice 66
70% Polished rice 70
Brown rice 55
Soba (buckwheat noodles) and udon (Japanese wheat noodles) 47
Instant noodles 47
Chinese noodles 47
Spaghetti 46
White bread 74
Cake bread 62
Butter roll 59
Croissant 67
Pizza 51
Japanese-style pancake 51
Pancake 67
Cornflakes 81
Potato chips 54
French fries 70
Other potatoes 78
Sweet potatoes, yams, and taros 51
Jam and marmalade 51
Sugar for coffee and tea 68
Sugar used during cooking 68
Rice crackers 91
Snacks made from wheat flour 63
Japanese sweets with azuki beans 49
Japanese sweets without azuki beans 68
Cakes 46
Cookies and biscuits 59
Chocolates 43
Candies, caramels, and chewing gum 74
Jellies 44
Doughnuts 76
Boiled beans 16
Raisins 64
Canned fruits 49
Fruit juice (100%) 47
Other fruit juice 47
Tomato juice 38
Oranges 39
Bananas 51
Apples 37
Strawberries 40
Grapes 50
Peaches 42
Pears 38
Persimmons 50
Kiwi fruit 53
Melons 42
Watermelons 58
Pumpkins 75
Lotus roots 47
Vegetable juice 38
Curry and roux in stew 67
Cocoa 51
Lactic acid bacteria beverages 42
Soft drinks 61
Nutritional supplement drinks 66
(Continued)
TABLE 1 (Continued)
Food and beverage item GI
Ground fish meat products 38
Boiled fish, shellfish, and seaweed in soy sauce 38
Full-fat milk 27
Low-fat milk 30
Skim milk 32
Yogurt (sweetened) 24
Yogurt (nonsweetened) 36
Yogurt (moderately sweetened) 30
Ice cream 61
Nutritional supplement bars 48
1
GI of glucose ҃ 100. These 70 food and beverage items from the 147
items in the diet-history questionnaire were used for the calculation of GI.
The remaining 77 items not used consisted of 6 alcoholic beverages (beer,
sake, shochu, schochu highball, whiskey, and wine), 8 items containing no
available carbohydrate (oils used during cooking, table salt, salt used during
cooking, sugarless soft drinks, chicken, Chinese soup, noodle soup, and
water), and 63 items containing 쏝3.5 g available carbohydrate per serving
[peanuts, other nuts, konnyaku, butter, margarine, mayonnaise, salad dress-
ing, tofu, tofu products, natto, miso as seasoning, miso in miso soup, carrots,
tomatoes, green peppers, broccoli, green leafy vegetables, salted pickled
plums (umeboshi), other salted pickles, cabbage, cucumbers, lettuce, Chi-
nese cabbage, bean sprouts, radishes, onions, cauliflower, eggplants, bur-
docks, mushrooms, wakame seaweed, laver, ketchup, nonoil salad dressing,
soy sauce, green tea and oolong tea, tea, coffee, dried fish, small fish with
bones, canned tuna, eel, white meat fish, blue-back fish, red meat fish,
shrimp, squid and octopus, oysters, other shellfish, fish eggs, salted fish
intestines, ground beef and pork, pork, beef, liver, ham and sausages, bacon,
eggs (heno and quailo), cheese, cottage cheese, coffee cream, corn soup, and
artificial sweeteners].
1164 MURAKAMI ET AL
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
BMI, serum concentrations of total and LDL cholesterol, and Hb
A
1c
were not significantly correlated with dietary GL. Adjust
-
ment for the percentage of energy from carbohydrate instead of
the percentage of energy from fat did not change the results
materially, which suggests that the observed correlations be-
tween dietary GI and GL and metabolic risk factors are indepen-
dent of carbohydrate intake (data not shown).
DISCUSSION
Because only limited evidence is available regarding associ-
ations between dietary GI and GL and metabolic risk factors,
particularly in Asian populations, we investigated these associ-
ations in the present cross-sectional study of healthy Japanese
female farmers with traditional dietary habits. We found that
dietary GI was positively associated with BMI, fasting serum
triacylglycerol, fasting plasma glucose, and Hb A
1c
after control
for potentially confounding lifestyle and dietary factors
.
We also
found that dietary GL was independently negatively associated
with serum HDL cholesterol and positively associated with se-
rum triacylglycerol and fasting plasma glucose.
Concerns have been expressed regarding the utility of the GI
for mixed meals (32, 33). However, many researchers have
shown that the GI of a mixed meal can be predicted consistently
as the mean of the GI values of each of the component foods,
weighted according to their relative contribution to carbohydrate
intake (34 –36). In reality, studies using standardized techniques
have observed high correlation coefficients between observed
and calculated GI values, ranging from 0.84 to 0.99 (34 –36).
Dietary GI and GL values in the present study were similar when
compared with those in a previous Japanese study (67 compared
with 64 for GI and 168 compared with 150 for GL) (10). How-
ever, the dietary GI and GL values observed in the present and
previous (10) Japanese studies were considerably higher than the
corresponding values in Western countries (48 – 60 for GI and
84 –120 for GL) (4– 6, 7–9, 37– 40). This may have resulted from
the differences in the major food contributors. Dietary GIs and
GLs in Western populations are determined by a variety of food
items, including potatoes (7– 8%), breakfast cereals (4 –7%),
bread (5%), and rice (5%) (41– 43). However, white rice (GI ҃
77) was the major contributor in the present and previous (10)
Japanese studies, accounting for 59% of dietary GI and GL in the
present study.
All self-reported dietary assessment methods are subject to
measurement error and selective underestimation or overestima-
tion of dietary intake (44). In the present study, however, we used
a previously validated DHQ (20 –22) to minimize data inaccu-
racy. Additionally, dietary GI and GL values calculated in the
present study are believed to be relatively accurate because the
major determinant of dietary GI and GL in the present study, rice
(62%), is more accurately reported than are other foods on the
DHQ because it is consumed regularly in relatively fixed
amounts. Moreover, the same tendency was observed in a re-
peated analysis of subjects with a physiologically plausible en-
ergy intake, ie, subjects with a ratio of energy intake to basal
metabolic rate of 1.2–2.5 (45)—앒78% of the subjects included
in the main analysis (data not shown). Thus, we considered that
the correlations observed in the present study reflect true asso-
ciations, not spurious associations resulting from inaccurate di-
etary data.
In the present study, dietary GI was positively correlated with
BMI. A 5-wk crossover, randomized, controlled trial conducted
in overweight nondiabetic men with ad libitum dietary intakes
also showed a significantly lower fat mass and a tendency for a
higher fat-free mass, but not a lower body weight, after a low-GI
diet than after a high-GI diet (46). In contrast, other ad libitum
trials conducted in subjects with type 2 diabetes showed no
significant differences in body weight change between high-GI
and low-GI diets (47– 49). However, in a 10-wk ad libitum,
randomized, controlled trial conducted in healthy overweight
women, decreases in body weight and fat mass were larger in a
low-GI diet group than in a high-GI diet group, although these
differences were not statistically significant (50). Moreover, as
was shown in this study, a recent observational study also showed
a positive association between dietary GI and BMI and no asso-
ciation between dietary GL and BMI (6).
Dietary GL has consistently been shown to be inversely cor-
related with HDL cholesterol in cross-sectional studies (8 –11).
In contrast, the correlation between dietary GI and HDL choles-
terol is not consistent. An inverse correlation has been reported
TABLE 2
Basic characteristics of the 1354 Japanese women
Value
Age (y) 55.3 앐 10.3
1
울39 y 69 (5.1)
2
40–49 y 319 (23.6)
50–59 y 446 (32.9)
60–69 y 440 (32.5)
욷70 y 80 (5.9)
Body height (cm) 152.9 앐 5.9
Body weight (kg) 56.1 앐 8.3
Current BMI (kg/m
2
)
24.0 앐 3.3
BMI at age 20 y (kg/m
2
)
21.7 앐 2.6
Menopausal status
Premenopausal 427 (31.5)
Postmenopausal 927 (68.5)
Current smoking
No 1309 (96.7)
Yes 45 (3.3)
Dietary supplement use
No 954 (70.5)
Yes 400 (29.5)
Rate of eating
Fast 480 (35.5)
Medium 644 (47.6)
Slow 230 (17.0)
Physical activity level 1.84 앐 0.28
Energy intake (kcal/d) 1944 앐 497
Protein intake (% of energy) 14.0 앐 2.2
Fat intake (% of energy) 25.3 앐 5.8
Carbohydrate intake (% of energy) 59.0 앐 7.1
Alcohol intake (% of energy) 0.8 앐 2.3
Nondrinkers 836 (61.7)
쏜0% to 쏝 1% of energy 271 (20.0)
욷1% of energy 247 (18.2)
Dietary fiber intake (g/1000 kcal) 7.6 앐 2.1
Dietary glycemic index
3
66.7 앐 4.0
Dietary glycemic load (/1000 kcal)
3
88.0 앐 15.1
1
x 앐 SD (all such values).
2
n; percentage in parentheses (all such values).
3
Glycemic index for glucose ҃ 100.
GLYCEMIC INDEX AND METABOLIC RISK FACTORS 1165
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
in 3 (7, 8, 10), but not in another 2 (9, 37), cross-sectional studies.
Furthermore, recent randomized controlled trials have not sup-
ported the beneficial effect of a low-GI diet on HDL cholesterol
in contrast with a high-GI diet (46 –50). In the present study, we
also found an inverse correlation between dietary GL and HDL
cholesterol, but no correlation between dietary GI and HDL
cholesterol.
Both dietary GI and GL were positively correlated with fasting
triacylglycerol in 2 cross-sectional studies (9, 10); however, no
association between dietary GI and fasting triacylglycerol was
observed in a study of elderly men (37). In the present study, both
dietary GI and GL were positively associated with fasting triac-
ylglycerol. Several randomized controlled trials have also shown
the beneficial effect of a low-GI diet on triacylglycerol (51),
although the lack of an effect of GI has been observed in subjects
with low triacylglycerol concentrations (52).
We identified a positive correlation between dietary GI and
GL and fasting glucose, whereas no correlation was observed in
a cross-sectional study of elderly men (37). Several prospective
cohort studies (4, 5, 38), but not others (39, 40, 53), in the United
States have shown a positive association between dietary GI, GL,
or both and the incidence of type 2 diabetes, which is not in
conflict with our finding. Recently, several (48, 49), but not all
(46, 47, 50), randomized controlled trials have also shown lower
fasting glucose concentrations after consumption of a low-GI
diet than after a high-GI diet.
TABLE 3
Selected characteristics of the 1354 Japanese women according to quintiles of dietary glycemic index and load
Quintiles of dietary glycemic index or load
P
1
1
(n ҃ 270)
2
(n ҃ 271)
3
(n ҃ 271)
4
(n ҃ 271)
5
(n ҃ 271)
Dietary glycemic index
2
60.8 앐 2.6
3
64.8 앐 0.7 67.0 앐 0.6 68.9 앐 0.6 71.8 앐 1.4
Age (y) 55.8 앐 11.3 54.7 앐 10.4 55.1 앐 10.4 55.4 앐 9.7 55.7 앐 9.4 0.89
Current BMI (kg/m
2
)
23.9 앐 3.3 23.9 앐 3.1 23.8 앐 3.1 24.2 앐 3.3 24.2 앐 3.5 0.23
BMI at age 20 y (kg/m
2
)
21.7 앐 2.8 21.5 앐 2.2 21.6 앐 2.4 21.8 앐 2.5 21.9 앐 2.8 0.22
Premenopausal women (%) 27 31 35 30 34 0.21
Current smokers (%) 633140.13
Dietary supplement users (%) 37 30 31 26 24 0.0005
Rate of eating (%) 0.45
Fast 37 36 33 39 33
Medium 46 51 47 47 47
Slow 17 13 20 14 20
Physical activity level 1.82 앐 0.29 1.84 앐 0.28 1.83 앐 0.28 1.85 앐 0.28 1.84 앐 0.29 0.45
Energy intake (kcal/d) 2171 앐 559 2067 앐 453 1979 앐 467 1809 앐 404 1695 앐 440 쏝0.0001
Fat intake (% of energy) 27.7 앐 5.5 27.0 앐 5.4 25.6 앐 5.2 24.3 앐 5.4 22.2 앐 5.5 쏝0.0001
Alcohol intake (%) 0.0007
Nondrinkers 57 57 62 67 66
쏜 0% to 쏝 1% of energy 20 21 21 19 20
욷1% of energy 23 22 17 14 15
Dietary fiber intake (g/1000 kcal) 9.1 앐 2.3 7.9 앐 2.0 7.6 앐 1.7 7.2 앐 1.7 6.4 앐 1.6 쏝0.0001
Dietary glycemic load (/1000 kcal)
2
67.6 앐 6.7 79.9 앐 2.4 87.4 앐 2.3 95.4 앐 2.5 109.7 앐 8.4
Age (y) 53.5 앐 12.1 54.4 앐 10.6 54.5 앐 10.1 56.9 앐 8.9 56.9 앐 9 쏝0.0001
Current BMI (kg/m
2
)
24.1 앐 3.3 23.8 앐 3.0 24.0 앐 3.4 24.1 앐 3.2 24.0 앐 3.4 0.70
BMI at age 20 y (kg/m
2
)
21.6 앐 2.6 21.7 앐 2.4 21.6 앐 2.6 21.7 앐 2.5 21.9 앐 2.6 0.21
Premenopausal women (%) 38 34 32 28 27 0.002
Current smokers (%) 633320.024
Dietary supplement users (%) 34 32 32 27 22 0.0007
Rate of eating (%) 0.58
Fast 39 32 36 34 37
Medium 47 48 50 45 47
Slow 14 20 14 21 16
Physical activity level 1.79 앐 0.27 1.84 앐 0.29 1.84 앐 0.27 1.86 앐 0.29 1.85 앐 0.30 0.010
Energy intake (kcal/d) 2285 앐 549 2106 앐 438 1926 앐 404 1808 앐 396 1595 앐 378 쏝0.0001
Fat intake (% of energy) 31.8 앐 4.7 28.4 앐 3.0 25.6 앐 3.1 22.4 앐 2.8 18.5 앐 3.5 쏝0.0001
Alcohol intake (%) 쏝0.0001
Nondrinkers 49 54 63 67 75
쏜0% to 쏝1% of energy 17 23 22 21 18
욷1% of energy 34 23 15 12 7
Dietary fiber intake (g/1000 kcal) 8.1 앐 2.3 7.9 앐 2.0 7.6 앐 1.8 7.7 앐 2.1 6.9 앐 1.9 쏝0.0001
1
For continuous variables, tests for linear trend used the median value in each quintile as a continuous variable in linear regression; a Mantel-Haenszel
chi-square test was used for categorical variables.
2
Glycemic index for glucose ҃ 100.
3
x 앐 SD (all such values).
1166 MURAKAMI ET AL
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
We found a positive correlation between dietary GI and Hb
A
1c
. A positive association was also reported in cross-sectional
studies conducted in patients with type 2 diabetes treated by
dietary restriction alone (12) and in patients with type 1 diabetes
(13). Additionally, a low-GI diet reduced Hb A
1c
more than did
a high-GI diet in several randomized controlled trials (48, 49).
Furthermore, a recent meta-analysis of 14 randomized controlled
trials has shown the amelioration of Hb A
1c
through a low-GI diet
(54).
Both total and LDL cholesterol were not correlated with di-
etary GI or GL in the present study, although randomized con-
trolled trials have generally shown that low-GI diets result in
lower total and LDL cholesterol concentrations (54). However,
similar to our findings, no correlation between dietary GI or GL
and total or LDL cholesterol was observed in several cross-
sectional studies (7, 10, 37).
Our results may not be extrapolated into general Japanese
populations because the subjects in the present study were se-
lected female farmers. Additionally, our DHQ, although similar
to most previous epidemiologic studies, was not designed spe-
cifically to measure dietary GI and GL; however, the satisfactory
validity of this DHQ for total carbohydrate (20) provides some
reassurance. Moreover, although we attempted to adjust for a
wide range of potential confounding variables, we could not rule
out residual confounding because of these or other unknown
variables. Furthermore, because the study population consisted
of generally healthy persons, the clinical relevance of our find-
ings remains to be elucidated. However, our results should pro-
vide valuable insight from a prevention perspective.
In summary, after adjustment for a variety of confounding
factors, we observed positive correlations between dietary GI
and BMI, fasting serum triacylglycerol, fasting plasma glucose,
and Hb A
1c
and between dietary GL and fasting serum triacyl
-
glycerol and fasting plasma glucose and negative correlations
between dietary GL and serum HDL cholesterol in healthy Jap-
anese female farmers whose dietary GI and GL were primarily
determined by white rice. Because the cross-sectional nature of
the present study precludes any causal inferences, more obser-
vational and experimental studies are needed before any firm
conclusions can be drawn with regard to the effect of dietary GI
and GL on metabolic risk factors.
We thank Michiko Sugiyama for technical advice regarding the glycemic
index values for Japanese foods.
KM created a table of glycemic index, conducted the statistical analyses,
and wrote the manuscript. SS was involved in the design of the dietary study
and assisted in the creation of the table and the manuscript. YT assisted in the
creation of the table. HO was involved in the management of the dietary
dataset and data collection during the dietary study. YH was involved in the
data collection for the dietary study. HH and EO were responsible for the
research design, data collection, and data management. FK was responsible
for the research design, data collection, and overall management. All authors
provided suggestions during the preparation of the manuscript and approved
TABLE 4
Metabolic risk factors according to quintiles of dietary glycemic index and load in Japanese women
Total n
Quintiles of dietary glycemic index or load
P for
trend
1
1234 5
Dietary glycemic index
2,3
1354 61 (46.1–63.4) 65 (63.5–65.9) 67 (66.0–67.9) 69 (68.0–70.0) 72 (70.1–76.5)
BMI (kg/m
2
)
4
1354 23.7 앐 0.2 (270)
5
23.9 앐 0.2 (271) 23.8 앐 0.2 (271) 24.2 앐 0.2 (271) 24.4 앐 0.2 (271) 0.017
Serum total cholesterol (mg/dL)
4,6
1354 212.1 앐 2.2 (270) 211.8 앐 2.0 (271) 211.6 앐 2.0 (271) 216.5 앐 2.0 (271) 211.5 앐 2.2 (271) 0.74
Serum HDL cholesterol (mg/dL)
4,6
1354 64.7 앐 0.9 (270) 62.5 앐 0.9 (271) 63.0 앐 0.9 (271) 63.8 앐 0.9 (271) 63.6 앐 1.0 (271) 0.58
Serum LDL cholesterol (mg/dL)
4,6
1348 130.0 앐 2.1 (269) 129.4 앐 1.9 (269) 128.2 앐 1.9 (269) 133.2 앐 1.9 (271) 127.3 앐 2.1 (270) 0.73
Fasting serum triacylglycerol (mg/dL)
4,6
1349 87.1 앐 3.0 (269) 99.1 앐 2.8 (270) 101.7 앐 2.7 (270) 98.0 앐 2.8 (270) 103.1 앐 3.0 (270) 0.001
Fasting plasma glucose (mg/dL)
4,6
764 92.9 앐 2.0 (152) 97.0 앐 1.8 (153) 97.0 앐 1.8 (153) 99.8 앐 1.8 (153) 99.3 앐 1.9 (153) 0.022
Glycated hemoglobin (%)
4,6
845 5.0 앐 0.1 (169) 5.1 앐 0.1 (169) 5.1 앐 0.1 (169) 5.2 앐 0.1 (169) 5.2 앐 0.1 (169) 0.038
Dietary glycemic load (/1000 kcal)
2,7
1354 69 (31.1–75.7) 80 (75.8–83.7) 87 (83.8–91.2) 95 (91.3–100.2) 107 (100.3–148.5)
BMI (kg/m
2
)
4
1354 24.2 앐 0.3 (270) 23.8 앐 0.2 (271) 24.0 앐 0.2 (271) 24.2 앐 0.2 (271) 23.8 앐 0.3 (271) 0.48
Serum total cholesterol (mg/dL)
4,6
1354 212.6 앐 3.0 (270) 215.1 앐 2.4 (271) 212.1 앐 2.1 (271) 212.2 앐 2.4 (271) 211.6 앐 3.2 (271) 0.87
Serum HDL cholesterol (mg/dL)
4,6
1354 67.2 앐 1.3 (270) 65.5 앐 1.0 (271) 62.1 앐 0.9 (271) 61.9 앐 1.0 (271) 60.8 앐 1.4 (271) 0.004
Serum LDL cholesterol (mg/dL)
4,6
1348 127.2 앐 2.8 (267) 130.4 앐 2.3 (271) 130.7 앐 2.0 (270) 130.0 앐 2.2 (270) 129.9 앐 3.0 (270) 0.56
Fasting serum triacylglycerol (mg/dL)
4,6
1349 91.0 앐 4.1 (269) 96.8 앐 3.3 (270) 95.6 앐 2.9 (270) 100.1 앐 3.2 (270) 105.4 앐 4.4 (270) 0.047
Fasting plasma glucose (mg/dL)
4,6
764 90.9 앐 2.7 (152) 97.0 앐 2.1 (153) 97.5 앐 1.9 (153) 97.2 앐 2.1 (153) 103.4 앐 2.9 (153) 0.012
Glycated hemoglobin (%)
4,6
845 5.0 앐 0.1 (169) 5.1 앐 0.1 (169) 5.1 앐 0.1 (169) 5.1 앐 0.1 (169) 5.2 앐 0.1 (169) 0.10
1
Linear trends were tested with increasing dietary glycemic indexes and loads by assigning each participant the median value for the category and modeling
this value as a continuous variable.
2
Glycemic index for glucose ҃ 100. Values are medians; ranges in parentheses.
3
The median values shown are the same for BMI, triacylglycerol, and total, HDL, and LDL cholesterol but are different for glucose and glycated
hemoglobin: 61, 64, 67, 69, and 71, respectively.
4
Adjusted for residential area (5 categories), age (울39, 40 – 49, 50–59, 60 – 69, and 욷70 y), menopausal status (premenopausal or postmenopausal),
current smoking (no or yes), dietary supplement use (no or yes), rate of eating (fast, medium, or slow), physical activity level (quintiles), energy intake (quintiles),
percentage of energy as fat (quintiles), alcohol intake (nondrinker, 쏜0% to 쏝1% of energy or 욷1% of energy), and energy-adjusted dietary fiber intake
(quintiles).
5
x 앐 SE; n in parentheses (all such values).
6
Additionally adjusted for current BMI (quintiles) and BMI at age 20 y (quintiles).
7
The median values shown are the same for BMI, triacylglycerol, glycated hemoglobin, and total, HDL, and LDL cholesterol but are different for glucose:
68, 79, 87, 95, and 107/1000 kcal, respectively.
GLYCEMIC INDEX AND METABOLIC RISK FACTORS 1167
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
the final version submitted for publication. None of the authors had any
conflict of interest to declare.
REFERENCES
1. Augustin LS, Franceschi S, Jenkins DJ, Kendall CW, La Vecchia C.
Glycemic index in chronic disease: a review. Eur J Clin Nutr 2002;56:
1049 –71.
2. Wolever TMS. The glycemic index. World Rev Nutr Diet 1990;62:120 –
85.
3. Jenkins DJA, Wolever TMS, Taylor RH, et al. Glycemic index of foods:
a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981;
34:362– 6.
4. Salmeron J, Manson JE, Stampfer MJ, Colditz GA, Wing AL, Willett
WC. Dietary fiber, glycemic load, and risk of non-insulin-dependent
diabetes mellitus in women. JAMA 1997;277:472–7.
5. Salmeron J, Ascherio A, Rimm EB, et al. Dietary fiber, glycemic load,
and risk of NIDDM in men. Diabetes Care 1997;20:545–50.
6. Ma Y, Olendzki B, Chiriboga D, et al. Association between dietary
carbohydrates and body weight. Am J Epidemiol 2005;161:359 – 67.
7. Frost G, Leeds AA, Dore CJ, Madeiros S, Brading S, Dornhorst A.
Glycaemic index as a determinant of serum HDL-cholesterol concen-
tration. Lancet 1999;353:1045– 8.
8. Ford ES, Liu S. Glycemic index and serum high-density lipoprotein
cholesterol concentration among us adults. Arch Intern Med 2001;161:
572– 6.
9. Liu S, Manson JE, Stampfer MJ, et al. Dietary glycemic load assessed by
food-frequency questionnaire in relation to plasma high-density-
lipoprotein cholesterol and fasting plasma triacylglycerols in postmeno-
pausal women. Am J Clin Nutr 2001;73:560 –6.
10. Amano Y, Kawakubo K, Lee JS, Tang AC, Sugiyama M, Mori K.
Correlation between dietary glycemic index and cardiovascular disease
risk factors among Japanese women. Eur J Clin Nutr 2004;58:1472– 8.
11. Slyper A, Jurva J, Pleuss J, Hoffmann R, Gutterman D. Influence of
glycemic load on HDL cholesterol in youth. Am J Clin Nutr 2005;81:
376 –9.
12. Wolever TMS, Nguyen PM, Chiasson JL, et al. Relationship between
habitual diet and blood glucose and lipids in non-insulin dependent
diabetes (NIDDM). Nutr Res 1995;15:843–57.
13. Buyken AE, Toeller M, Heitkamp G, et al. Glycemic index in the diet of
European outpatients with type 1 diabetes: relations to glycated hemo-
globin and serum lipids. Am J Clin Nutr 2001;73:574 – 81.
14. Sasaki S, Takahashi T, Iitoi Y, et al. Food and nutrient intakes assessed
with dietary records for the validation study of a self-administered food
frequency questionnaire in JPHC Study Cohort I. J Epidemiol 2003;
13(suppl):S23–50.
15. Statistics and Information Department, Minister’s Secretariat, Ministry
of Health, Labour, and Welfare, Japan. Vital Statistics of Japan 2003.
Tokyo, Japan: Ministry of Health, Labour, and Welfare, 2003 (in Japa-
nese).
16. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of
diabetes: estimates for the year 2000 and projections for 2030. Diabetes
Care 2004;27:1047–53.
17. Horiguchi H, Oguma E, Sasaki S, et al. Dietary exposure to cadmium at
close to the current provisional tolerable weekly intake does not affect
renal function among female Japanese farmers. Environ Res 2004;95:
20 –31.
18. Horiguchi H, Oguma E, Sasaki S, et al. Environmental exposure to
cadmium at a level insufficient to induce renal tubular dysfunction does
not affect bone density among female Japanese farmers. Environ Res
2005;97:83–92.
19. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concen-
tration of low-density lipoprotein cholesterol in plasma, without use of
the preparative ultracentrifuge. Clin Chem 1972;18:499 –502.
20. Sasaki S, Yanagibori R, Amano K. Self-administered diet history ques-
tionnaire developed for health education: a relative validation of the
test-version by comparison with 3-day diet record in women. J Epide-
miol 1998;8:203–15.
21. Sasaki S, Yanagibori R, Amano K. Validity of a self-administered diet
history questionnaire for assessment of sodium and potassium. Compar-
ison with single 24-hour urinary excretion. Jpn Circ J 1998;62:431–5.
22. Sasaki S, Ushio F, Amano K, et al. Serum biomarker-based validation of
a self-administered diet history questionnaire for Japanese subjects. J
Nutr Sci Vitaminol (Tokyo) 2000;46:285–96.
23. Science and Technology Agency, Japan. Standard tables of food com-
position in Japan. 5th ed. Tokyo, Japan: Printing Bureau of the Ministry
of Finance, 2000 (in Japanese).
24. Foster-Powell K, Holt SH, Brand-Miller JC. International table of gly-
cemic index and glycemic load values: 2002. Am J Clin Nutr 2002;76:
5–56.
25. Sugiyama M, Tang AC, Wakaki Y, Koyama W. Glycemic index of
single and mixed meal foods among common Japanese foods with white
rice as a reference food. Eur J Clin Nutr 2003;57:743–52.
26. Sugiyama M, Wakaki Y, Nakamoto N, et al. The study of rice and
glycemic index. J Jpn Soc Nutr Care Manage 2003;3:1–15 (in Japanese
with English abstract).
27. Hashizume N, Ihara H, Kakinoki T, Inage H, Kimura S, Kimura M.
Response to blood glucose and insulin by Japanese foods in healthy
subjects. J Jpn Soc Clin Nutr 2004;25:222–5.
28. Fernandes G, Velangi A, Wolever TM. Glycemic index of potatoes
commonly consumed in North America. J Am Diet Assoc 2005;105:
557– 62.
29. Ainsworth BE, Haskell WL, Leon AS, et al. Compendium of physical
activities: classification of energy costs of human physical activities.
Med Sci Sports Exerc 1993;25:71– 80.
30. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical
activities: an update of activity codes and MET intensities. Med Sci
Sports Exerc 2000;32(suppl):S498 –504.
31. Ministry of Health, Labour, and Welfare, Japan. Dietary reference in-
takes for Japanese, 2005. Tokyo, Japan: Daiichi Shuppan Publishing Co,
Ltd, 2005 (in Japanese).
32. Coulston AM, Hollenbeck CB, Swislocki AL, Reaven GM. Effect of
source of dietary carbohydrate on plasma glucose and insulin responses
to mixed meals in subjects with NIDDM. Diabetes Care 1987;10:395–
400.
33. Hollenbeck CB, Coulston AM. The clinical utility of the glycemic index
and its application to mixed meals. Can J Physiol Pharmacol 1991;69:
100 –7.
34. Wolever TM, Jenkins DJ. The use of the glycemic index in predicting the
blood glucose response to mixed meals. Am J Clin Nutr 1986;43:167–
72.
35. Chew I, Brand JC, Thorburn AW, Truswell AS. Application of glycemic
index to mixed meals. Am J Clin Nutr 1988;47:53– 6.
36. Wolever TM, Jenkins DJ, Jenkins AL, Josse RG. The glycemic index:
methodology and clinical implications. Am J Clin Nutr 1991;54:846 –
54.
37. van Dam RM, Visscher AW, Feskens EJ, Verhoef P, Kromhout D.
Dietary glycemic index in relation to metabolic risk factors and inci-
dence of coronary heart disease: the Zutphen Elderly Study. Eur J Clin
Nutr 2000;54:726 –31.
38. Schulze MB, Liu S, Rimm EB, Manson JE, Willett WC, Hu FB. Gly-
cemic index, glycemic load, and dietary fiber intake and incidence of
type 2 diabetes in younger and middle-aged women. Am J Clin Nutr
2004;80:348 –56.
39. Meyer KA, Kushi LH, Jacobs DR Jr, Slavin J, Sellers TA, Folsom AR.
Carbohydrates, dietary fiber, and incident type 2 diabetes in older
women. Am J Clin Nutr 2000;71:921–30.
40. Stevens J, Ahn K, Juhaeri, Houston D, Steffan L, Couper D. Dietary fiber
intake and glycemic index and incidence of diabetes in African-
American and white adults: the ARIC study. Diabetes Care 2002;25:
1715–21.
41. Liu S, Manson JE, Buring JE, Stampfer MJ, Willett WC, Ridker PM.
Relation between a diet with a high glycemic load and plasma concen-
trations of high-sensitivity C-reactive protein in middle-aged women.
Am J Clin Nutr 2002;75:492– 8.
42. Liu S, Willett WC, Stampfer MJ, et al. A prospective study of dietary
glycemic load, carbohydrate intake, and risk of coronary heart disease in
US women. Am J Clin Nutr 2000;71:1455– 61.
43. Jonas CR, McCullough ML, Teras LR, Walker-Thurmond KA, Thun
MJ, Calle EE. Dietary glycemic index, glycemic load, and risk of inci-
dent breast cancer in postmenopausal women. Cancer Epidemiol Bi-
omarkers Prev 2003;12:573–7.
44. Livingstone MBE, Black AE. Markers of the validity of reported energy
intake. J Nutr 2003;133:895S–920S.
45. Black AE, Coward WA, Cole TJ, Prentice AM. Human energy expen-
diture in affluent societies: an analysis of 574 doubly-labelled water
measurements. Eur J Clin Nutr 1996;50:72–92.
46. Bouche C, Rizkalla SW, Luo J, et al. Five-week, low-glycemic index diet
1168 MURAKAMI ET AL
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
decreases total fat mass and improves plasma lipid profile in moderately
overweight nondiabetic men. Diabetes Care 2002;25:822– 8.
47. Heilbronn LK, Noakes M, Clifton PM. The effect of high- and low-
glycemic index energy restricted diets on plasma lipid and glucose pro-
files in type 2 diabetic subjects with varying glycemic control. J Am Coll
Nutr 2002;21:120 –7.
48. Jimenez-Cruz A, Bacardi-Gascon M, Turnbull WH, Rosales-Garay P,
Severino-Lugo I. A flexible, low-glycemic index Mexican-style diet in
overweight and obese subjects with type 2 diabetes improves metabolic
parameters during a 6-week treatment period. Diabetes Care 2003;26:
1967–70.
49. Rizkalla SW, Taghrid L, Laromiguiere M, et al. Improved plasma glu-
cose control, whole-body glucose utilization, and lipid profile on a low-
glycemic index diet in type 2 diabetic men: a randomized controlled trial.
Diabetes Care 2004;27:1866 –72.
50. Sloth B, Krog-Mikkelsen I, Flint A, et al. No difference in body weight
decrease between a low-glycemic-index and a high-glycemic-index diet
but reduced LDL cholesterol after 10-wk ad libitum intake of the low-
glycemic-index diet. Am J Clin Nutr 2004;80:337– 47.
51. Ludwig DS. The glycemic index: physiological mechanisms relating to
obesity, diabetes, and cardiovascular disease. JAMA 2002;287:2414–
23.
52. Jenkins DJ, Wolever TM, Kalmusky J, et al. Low-glycemic index diet in
hyperlipidemia: use of traditional starchy foods. Am J Clin Nutr 1987;
46:66 –71.
53. Hodge AM, English DR, O’Dea K, Giles GG. Glycemic index and
dietary fiber and the risk of type 2 diabetes. Diabetes Care 2004;27:
2701– 6.
54. Brand-Miller J, Hayne S, Petocz P, Colagiuri S. Low-glycemic index
diets in the management of diabetes: a meta-analysis of randomized
controlled trials. Diabetes Care 2003;26:2261–7.
GLYCEMIC INDEX AND METABOLIC RISK FACTORS 1169
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from
