am j clin nutr 1997 holt 1264 76

7/18/2019 Am J Clin Nutr 1997 Holt 1264 76

http://slidepdf.com/reader/full/am-j-clin-nutr-1997-holt-1264-76 1/13

A BSTRAC T The aim of this study was to system atically

c om p ar e p os tp ra nd ia l in su lin r es po ns es to is oe ne rg etic 1 00 0-U

(240-kcal) portions of several com mon foods. C orrelations w ith

n ut rie nt c on te nt w er e d et er mi ne d. T hir ty -e ig ht f oo ds s ep ar at ed in to

s ix f ood cat egori es ( fr ui t, bake ry produc ts , s nacks , car bohydra te

r ic h f oo ds , p ro te in -ri ch fo od s, a nd b re ak fa st c er ea ls ) w er e fe d t o

g ro up s o f 1 1â €” 13 ea lt hy s ub je ct s. F in ge r-p ri ck b lo od s am pl es w er e

obtained every 15 m m over 120 m m. A n insulin score w as calcu

lated from the area under the insulin response curve for each food

w ith use of w hite bread as the reference food (score = 100 ).

S ig nifica nt d iffe re nce s in in su lin sc ore w ere fou nd b oth w ith in a nd

among the food categories and also among foods containing a

sim ila r am ou nt o f ca rb oh yd rate. O ve rall, g lu cose an d insu lin

scores were highly correlated (r = 0.70, P < 0.001, n = 38).

H o we ve r, p ro te in -r ic h f oo ds a nd b ak er y p ro du cts ( ric h in f at a nd

r ef ined car bohydra te ) e li ci ted i ns ul in r es ponses t ha t w ere d is p ro

portionately higher than their glycemic responses. Total carbohy

drate (r = 0.39, P < 0.05, n = 36) and sugar (r = 0.36, P < 0.05,

n = 36 cont en ts were posi ti ve ly r el at ed t o t he mean insul in s core s

whereas fat (r â€”¿0.27,S, n 36) a nd protein (r â€”¿0.24,S,

n = 3 8 c on te nts w ere n eg ative ly re la ted . C on side ra tio n o f in su lin

scores m ay be relevant to the dietary m anagem ent and pathogen

e si s o f n on -i ns ul in -d ep en de nt d ia be te s m e ll it us a nd h yp er li pi de m ia

an d m ay h elp in crease th e accu racy o f estim atin g prep ran dial

in su lin req uirem en ts. A m J C lin N utr l9 97 ;6 6:l2 64 â€ ”76 .

KEY W ORDS Insulin, glycem ic index, NIDDM , non

in su lin -d ep en de nt d ia be te s m e ff itu s, d ia be tic d ie t, h yp er lip id

e m ia , c arb oh yd ra te , in su lin s co re , g lu co se s co re , a re a u nd er th e

c ur ve , h um an s

INTRODUCTION

The insulinem ic effects of foods m ay be relevant to the

tre atm en t a nd p re ve ntio n o f w eig ht g ain , n on -in su lin -d ep en

dent d iabet es m e ll it us ( N ID D M) , and ass oci at ed com pl icat ions.

R ec en t stu die s h av e s ho wn th at c arb oh yd ra te -ric h d ie ts , w hic h

result in high p ostp rand ia l g luc ose an d insu lin resp on ses, are

a sso cia te d w ith u nd es ira ble lip id p ro file s (1 , 2 ), g re ate r b od y

fa t (3 â€ ”5 ), nd th e d ev elo pm en t o f in su lin re sista nc e in ra ts (6 )

and hum ans (7, 8). Both obesity and N JD DM are associated

w ith v ary in g d eg re es o f in su lin re sista nc e a nd fa stin g h yp erin

s ul in em ia . P ro lo ng ed o r h ig h d eg re es o f p os tp ra nd ia l i ns ulin e

m ia are thought to contribute to the developm ent of insulin

re sis ta nc e an d asso cia te d d ise ase s (9 â€ ”1 7).T here fore , th e clas

sific atio n o f th e re la tiv e in su lin em ic e ffe cts o f d iffe re nt fo od s

is of both theoretical and practical significance.

Postprandial blood glucose responses have been the focus of

much research because of their im portance for glycem ic con

trol in patients with diabetes. It is now well accepted that

d iffe re nt fo od s c on ta in in g e qu al a mo un ts o f c arb oh yd ra te c an

produce a w ide range of blood glucose responses. T he glyce

m ic index (G I) m ethod w as developed to rank foods according

to th e ex te nt to w hich th ey in crease b lo od g lu co se co ncen tra

tions (18). Tables of GI values of com m on carbohydrate

containing foods are a useful guide to help people with diabetes

ch oo se fo ods that p ro du ce sm aller glyce mic resp on ses. H ow

ever, the G I concept does not consider concurrent insulin

responses and few studies have reported GI values and their

accom panyi ng i ns u li n r es ponses.

T he extent to w hich different dietary factors affect post

prandial insulinem ia has not been w ell researched because

insulin secretion is largely assum ed to be proportional to

postprandial glycem ia. Furtherm ore, hyperglycem ia is

th ou gh t to b e m ore re lev ant to th e se con dary co mp lication s

of NIDDM because the abnormal insulin secretion or action

in p eo ple w ith d iab etes is co ntro lled w ith ex og eno us in sulin

or m edications that counteract insulin resistance. H ow ever,

know ledge of factors that influence both postprandial gly

cem ia and insulin secretion in nondiabetic persons is re

quired to devise treatm ent strategies that w ill com pletely

norm alize m eal-related glycem ia (19).

T o e xp lo re th e im po rta nc e o f d ie ta ry h ab its a nd p os tp ra nd ia l

in sulin em ia in th e etio lo gy an d treatm ent o f N ID DM , w e n eed

to be able to systematically rate insulin responses to comm on

foods. If we are to compare insulin responses to foods, what is

the best basis of com parison? Should we com pare insulin

responses to portions of food representing a normal serving

size, portions containing an equal amount of carbohydrate, or

portions containing an equal am ount of energy? 01 tables

represent the glycem ic effects of equal-carbohydrate portions

I From the Hum an Nutrition Unit Department of Biochem istry The

U ni ve rs ity o f S yd ne y; a nd th e S ch oo l o f M a th em ati ca l S ci en ce s, T he

Univers i ty of Technology, Sydney, Austra l ia .

Supported by research grants from The University of Sydney and

Kel lo g g' s Au s tral ia Pty Ltd.

3 Address reprint requests to JC Brand Miller Human Nutrition Unit

D e pa rt m en t o f B i oc he m is tr y 0 08 , T h e U n iv er si ty o f S y dn ey , N S W 2 00 6,

Australia.

R e ce iv ed N o ve m be r 2 1, 1 99 6.

A c ce pt ed f or p ub li ca ti on M a y 2 2, 1 99 7.

1264

A m J C lin N utr 1 99 7;6 6:1 26 4â €” 76 .P rin ted in U SA . Â ©1 99 7 A meric an S ocie ty fo r C lin ica l N utr itio n

An insulinindexof foods: the insulindemand generated by

1 kJ portions of common foods13

Susanne H A H olt Janette C B rand M iller and P eter P etocz



Food Variety, manufacturer, or place of purchase Preparation

IN SU LIN IN DEX OF FOOD S

1265

TABLE

Desc riptionandpreparationof the testfo ods

Fruit

Black grapes

Apples

Oranges

Bananas

Bakery products

Croissants

Chocolate cake with

frosting

Doughnuts with cinnamon

sugar

Chocolate chip cookies

Water crackers

Snack foods and confectionery

Mars Bar

Yogurt

Ice cream

Jellybeans (assorted colors)

Peanuts

Potato chips

Popcorn

Protein-rich foods

Cheese

Eggs

L e n t i l s

Baked beans

Beefsteak

W hi te f is h

Carbohydrate-rich foods

White bread

W ho le -m ea l b re ad

Grain bread

W hi te ri ce

Brown rice

White pasta

Brown pasta

Potatoes

Waltham cross

Reddelicious

Navel

Cavendish

Fresh, stem removed, served whole

Fresh, unpeeled, cut into eight segments

Fresh, peeled, cut into eight segments

Fresh, peeled, cut into quarters

Defros ted, reheated a t 180Â°Cor 6 mm, and served warm

Prepared according to manufacturer's directions, stored at

4 Â°Cup to 2 d before serving at room temperature

Prepared by supermarket from standard recipe, defrosted

overnight, reheated at 180Â°Cor 5 mm, and served

warm

Served crisp at room temperature, stored in airtight

container

Served crisp at room temperature

Cut into four standard pieces and served at room

temperature

Stored at 4 Â°C,erved cold

Stored frozen and served cold

Served at room temperature, stored in airtight container

Served at room temperature, stored in airtight container

Served from freshly opened packet

Prepared according to manufacturer's directions

immediately before serving

A ll s erv ing s cut fro m sam e larg e blo ck, s to red at 4 Â °C,

served cold

Poached the day before serving, stored at 4 Â°Cvernight,

reheatedin microwaveoven for 1.5 mm immediately

before serving

Prepared in bulk according to recipe, stored at 4 Â°Cor up

to 2 d, re he ate d i n a m ic ro wav e o ve n f or 2 m m

immediatelybeforeserving

Heated on s tove for5 mm immediate ly before serv ing

Gril led the day before serving, cut into standard bite-s ized

pieces, and stored at 4 Â°Cvernight; reheated in

microwave oven for 2 mm immediately before serving

Steamed the day before serving, stored at 4 Â°Cvernight,

cut into bite-s ized pieces, and reheated in microwave

oven for 2 miii immediately before serving

Served fresh and plain at room temperature



Bo iled 12 mii i and s tored overnight a t 4 Â°C,eheated in

microwave oven for 1.5 mm immediately before serving

Bo iled 12 mm and stored overnight a t 4 Â°C,eheated in

microwave oven for 1.5 mm immediately before serving

Boiled8 mm and storedovernightat4 Â°C

Re he ate d i n m ic ro wav e o ve n f or 1 .5 m m im me diate ly

before serving

Peeled,boiledfor 20 mm, andstoredat 4 Â°Cvernight;

reheated in a microwave oven for2 mm immediate ly

before serving

P urc has ed i n bul k f ro m s upe rm arke t and s to re d f ro ze n

White Wings Foods, Smithfie ld, Sydney, Australia

Purchased in bulk from supermarket and stored frozen

Anion's Biscuits Ltd. Homebush, Sydney, Australia

Grocery Wholesalers Ltd, Yennora, Australia

Mars Confectionary Australia, Ballarat, Australia

Strawberry fruit yogurt; Australian Co-operative

Foods,' Wetherill Park, Sydney, Australia

Vanil la ice cream; Dairy Bell, Camperdown, Sydney,

Australia

Grocery Wholesalers Ltd

Salted roasted peanuts; Grocery Wholesalers Ltd

Crinkle cut chips; Smith's Snackfood Company,

Chatswood, Sydney, Australia

Microwave cooked popcorn; Uncle Toby's Company

Ltd, Wahgunyah, Australia

M ature c he ddar c he es e; G ro ce ry W ho le sa le rs L td

Poached hens eggs

Served in tomato sauce2

Canned navy beans in tomato sauce; Franklins,

Chullora, Sydney, Australia

Le an to ps ide be ef f il le ts bo ug ht i n bulk f ro m

supermarket, trimmed and stored frozen

Ling f ish i fi le ts bought in bulk from Sydney f ish

markets, trimmed and stored frozen

Fresh sliced wheat-flour bread; Quality Bakers

Australia Ltd. Eastwood, Sydney, Australia

Fresh sliced bread made from whole-meal wheat flour;

Riga Bakeries, Moorebank, Sydney, Australia

Fresh sliced rye bread containing 47% kibbled rye; Tip

TopBakeries,Chatswood,Sydney,Australia

Cairose rice (Sunwhite), Ricegrowers' Co-operative

Ltd. Leeton, Australia

Calrose rice (Sunbrown), Ricegrowers' Co-operative

Ltd

Spirals

W ho le -m eal s pi ral s; S an R em o P as ta C om pany ,

Auburn, Sydney, Australia

Russet potatoes



FoodVariety, manufacturer, or place o furchasePreparationFrench

friesPrefried oven-baked French fries; M cCain's Foods

( A us tra li a) , C as tl e H il l, S y dn ey , A us tra li aS to re d

frozen, cooked in conventional oven for 15 mm

immediately beforeervingBreakfast

cereals3CornflakesKellogg's

Australia Pty Ltd, Pagewood, Sydney,

Australiaâ€”Special

KToasted

f lakes made from wheat and rice f lour, high in

protein; Kellogg's Australia Plytdâ€”HoneysmacksPuffed

whole-wheat grains with a honey-based coating;

Kellogg's Australia Plytdâ€”SustainA

mixture of wheat, corn, and rice flakes; rolled oats;

dried fruit; and flaked almonds; Kellogg's Australia

Ptytdâ€”All-BranA

high-fiber cereal made from wheat bran; Kellogg's

Australia Ptytdâ€”Natural

muesliBased on raw rolled oats, wheat b ran, dried fruit, nuts,

and sunflower seeds; Uncle Toby 's Company Ltd.

Wahgunyah,ustraliaâ€”PorridgeUncle

Toby's Company Ltd. Wahgunyah, AustraliaRaw

ro lled oats cooked in a microwave oven according to

m anuf ac ture r's di re cti ons and s erv ed w itho ut s w ee te ne r

1266

HOLT ET L

T ABL E 1

Continued

1 N ow Dairy Farmer s.

2 Recipe: 15 m L olive oil 350 g dried green lentils 410 g canned tomatoes 120 g onion 1 clove garlic and 1 tsp pepper

3 Al l cereals were served fr esh wi th 125 mL fat-reduced (1.5 fat) mi lk.

of foods However carbohydrates not the only s timulusfor

i ns ul in s ecre ti on . Prote in -ri ch f oods or the addi ti on of prote in

to a carbohydrate-rich meal can stimulate a modest rise in

i ns ul in s ecre ti on w i thout i ncreas ing blood g lucose concentra

tions , particularly in subjects wi th diabetes (20â€”22).Similarly,

adding a large amount of fat to a carbohydrate-rich meal

increases insulin secretio n ev en tho ugh plasma g lucose re

spo ns es are re duc ed (2 3, 2 4).

T hus, pos tprandi al i ns ul in responses are not al w ays propor

tional to blood glucose concentrations or to a meal's total

carbohydrate content. Several insul inotropic factors are known

to po tentiate the stimulato ry effec t o f gluco se and mediate

po stprandial insulin se cre tio n. These inc lude fructo se, c ertain

am ino ac ids and fatty acids, and gastro inte stinal ho rmo nes such

as g as tric inhibito ry pe ptide , g luc ag on, and c ho le cy sto ki in

25 , 2 6 . T hu s, p ro te in - a nd fa t-rich fo od s m ay in du ce s ub sta n

ti al i ns ul in s ecre ti on desp ite produci ng re lati ve ly s mall b lood

g luc ose re spo nse s. W e the re fo re de cide d that c om paring the

insulinemic effects of foods on an isoenergetic basis w as a

logical and practical approach.

The aim o f this study w as to sy stematically c ompare po st

prandia l i ns ul in responses to i soenerge ti c porti ons o f a range of

common foods. A n insulin score (IS) w as calculated for each

f ood on the bas is o f i ts i ns ul inemic e ff ec t re lati ve to a re ference

food. Thirty-eight foods, categorized into six different food

groups, w ere studied to determine w hich foods w ithin the same

food group were most insulinogenic. We hypothesized that

po stprandial ins ulin re spo nse s are no t c lo se ly re late d to the

carbohydrate content or glycemic effects of some foods.

SU BJECTS A ND METHOD S

Test fo ods

Thirty-eight foods were tested and were grouped into six

fo od cate go ries : 1) fruit: grapes, bananas, apples, and o rang es;

2 ba ke ry products: croissants, chocolate cake w ith icing,

doughnuts w ith cinnamon sugar, chocolate chip cookies, and

w ate r c rac ke rs; 3 ) snac k fo ods and c onfe ctio ne ry : M ars B ar

c andy bar ( Mars Co nf ec tio nary A us tral ia, B al larat, A us tral ia) ,

s traw be rry y og urt, v ani ll a i ce c re am , je lly be ans , s al te d ro as te d

pe anuts , plain po tato c hips, and plain po pco rn; 4) pro tein-ric h

f oo ds : c he dd ar c he es e po ac he d e gg s b o ile d le nt ils in a to m at o

sauce, baked beans in a tomato sauce, grilled beef steak, and

s te am ed w hi te f is h; 5 ) c arbo hy drate -ric h f oo ds : w hite bre ad,

w ho le -m eal bre ad, ry e-g rain bre ad, w hite ric e, bro wn ric e,

w hite pasta, bro wn pas ta, bo ile d po tato es , and o ve n-bake d

Fre nch frie s; and 6) breakfast ce reals: Cornflakes (Ke llo gg's

A ustralia Pty Ltd. Pag ew oo d, A ustralia), S pec ial K (Ke llo gg's

Australia Pty Ltd), Honeysmacks (Kellogg's A ustralia Pty

Ltd), Sustain (Kellogg's Australia Pty Ltd), All-Bran

(Kellogg's A ustralia Pty Ltd), natural muesli, and oatmeal

porridge.

Each food was served plain as a 1000-U portion with 220

mL w ater. W hite bread w as used as the reference food for each

fo od g ro up. The fo ods w ere selected to represe nt a rang e o f

natural and pro cesse d fo ods co mm only e aten in industrializ ed

so cie ties . D etails o f the fo ods and the ir preparatio n me tho ds are

listed in Table 1. Foods were bought in bulk to minimize

variations in composition and w ere served in standard-sized

pieces. The nutritional composition ofeach food per 1000 U as

cal cu lated f rom A us tra li an f ood tabl es or manufac turers ' data

is shown in Table 2.

Subjects

S eparate g ro ups of healthy subje cts (n = 1 1â€”13 )w ere re

cruited to test each category of foods. V olunteers w ere ex

cluded i f they were smokers or taking prescription medications ,

had a f am ily his to ry o f di abe te s o r o be si ty , w ere die ting , o r had

i rregul ar eati ng habits . In to tal , 41 s ubjects parti ci pated. O ne

subject consumed all of the test foods and 15 other subjects



ServingCarbohydrateEnergyFood

weight FatProtein

SugarStarch

FiberWater

density

INSULIN INDEX OF FOODS

1267

TABLE 2

Nuthtional composition of the test foods per 1000-U serving as calculated from Australian food tables or manufacturers' data'

g g g g g g

@I/gFruitGrapes3950.43.256.90.03.6317.02.5Bananas2790.34.747.28.46.1210.13.6Apples4350.01.356.52.29.1360.92.3Oranges6250.66.950.60.012.5539.41.6Bake

productsCroissants6114.46.13.118.61.813.516.4Cake26411.94.320.110.50.710.715.6Doughnuts6513.44.38.917.01.416.115.4Cookies25110.92.418.716.21.02.

andonfectioneryMars

Barr549.42.936.71.11.73.518.5Yogurt@2415.311.837.60.00.5187.04.2Ice

cream12013.45.225.80.00.074.28.3Jellybeans880.05.344.611.50.012.211.4Peanuts3820.19.61.73.72.40.626.3Potatochips24416.22.70.222.12.41

foodsCheese5920.015.00.10.00.020.916.9Eggs15917.919.60.50.00.0119.46.3Lentils2534.619.44.224.91

1.4222.03.9Baked

beans3511.716.116.123.216.8267.12.8Beef

steak1587.742.00.00.00 .0104.36.3Fish3331.056.30.00 .00.0250.03.0Carbohydrate-rich

foodsWhite

bread2942.18.51.844.13.336.110.6Whole-meal

bread21012.67.61.743.76.640.39.9Grain

bread21085.49.42.437.66.541.49.3White

rice22030.55.00.156.00.4140.04.9Brown

rice21482.15.20.552.61.493.96.8White

pasta2010.87.82.047.13.5134.85.0Brown

pasta22181.611.30.747.810 .9132.64.6Potatoes3681.010.03.145.99.2290 .82.7French

fries2938.73.91.135.43.533.810.7Breakfast

cerealsCornflakes21702.18.410.236.11.5110.95.9Special

K21722.115.314.027.21.4111.25.8Honeysmacks21722.28.731.117.02.61

15.05.8Sustain21683.19.713.729.13.2119.15.9Muesli21756.110.717.119.86.61

14.15.7Pomdge23836.210.97.529.04.7333.72.6All-Bran21742.911.713.929.414.1111.05.7

I M ar s Bar , M ars Confectionar y Austral ia, Bal larat, Austral ia; Comf iakes, Special K, H oneysmacks, Sustain, and Al l -Bran: Kel logg s Austral ia Pty L td.

Pagewood, Australia.

2 Nutr ient composi tion calculated fr om manufacturer s data.

completed two or more food categories. A ll of the subjects approved by the Medical Ethical Review Committee of the

w ere uni ve rs ity s tude nts ; re le vant c harac te ris tic s o fthe s ubje cts U niv ers ity o f S ydne y.

are listed in Table 3. The mean body mass index (BMI, in

kg/m2) of the 41 subjects w as 22.7 Â ±0.4 (range: 19â€”29).Three P1@ Ot(W Ol

subjects had a B MI > 25 but tw o of these w ere short, stocky Each subject first consumed a 1000-U portion of w hite bread

males w ho had ex cess muscle rather than fat. Female subjects (4 5.9 g carbo hy drate) to co nfirm normal g luco se to lerance.

did not participate during their menstrual period or if they W hite bread w as also used as the reference food (IS = 100%)

experienced adverse premenstrual symptoms. Informed con- against which all other foods were compared, similar to the

sent w as o btained from all o f the subjects and the study w as method used fo r c alculating GI v alues o f fo ods (1 8). The use of



FoodroupAgeBMI2yFruit

(n = 5 F, 6 M)22.9

Â ±3.9 22 .9

Â ±.4Bakery

produc ts (n = 6 F, 6 M )2 2.2

Â±3.7 23 .1 Â ±.7Snacks

and confectionery(n = 5 F, 7 M)21.0

Â±1.2 22. 9 Â ±.5Protein-rich

foods (n 5 F, 6 M)22.4 Â±2.824.3

Â ±.1Carbohydrate-rich

foods (n = 5 F, 8 M)21.0

Â±1.923.0 Â±.9Breakfast

cereals (n = 5 F, 6 M)22.8

Â±3.922.8

Â±1.4

1268

HOLT ET AL

tube radioimmunoassay kit (Coat-A -Count; D iagnostic Prod

ucts Corporation, Los A ngeles). For both plasma glucose and

ins ul in analy si s, all nine plas ma s am ple s f or a partic ular s ub

ject's test w ere analyzed w ithin the same run to reduce any

er ro r in t ro d u ce d b y in t e ra s s a yva r ia t io n . W h e n p o s s i b le , a ll

plasma samples fo r a particular subject w ere analyz ed for

insulin w ithin the same run. For the insulin analysis, the mean

w ithin-assay CV w as 5% and the mean betw een-assay CV w as

7

Stati stical analys is

Cumulativ e chang es in po stprandial plasma g luc ose and

insulin re spo ns es fo r e ac h fo od w ere quantifie d as the inc re

mental area under the 120-mn response curve (A UC), w hich

w as c alc ulate d by us ing the trape zo idal rule w ith fas ti ng c on

centrati ons as the bas el ine and truncated at zero . A ny negati ve

are as te nde d to be s mall and w ere ig no re d. Fo r e ac h s ubje ct, an

IS ( %) w as c alc ulate d f or e ac h te st fo od by div iding the ins uli n

A U C v alue fo r the te st fo od by the ins ulin A U C v alue fo r w hite

bread (the reference food), and expressed as a percentage as

follows:

I S ( )

A re a unde r the 1 20 -m m i ns uli n re spo ns e

c urv e fo r 1 00 0 U te s t fo od

Area underthe 120 -mminsulinresponse curve

f or 1000 Id w hi te bread

TABLE 3

Characteristics of each group of subjects '

â €˜ ¿ Â ± S D

2 I n kg rn2

a re fe re nc e fo od c ontro ls fo r inhe re nt diffe re nc es be tw ee n

indi vi duals that af fe ct ins uli n s ens itiv ity , s uc h as bo dy w eig ht

and ac ti vi ty l eve ls .

Subjects were fed 1000-U portions of the test foods in a

rando m order on separate mo rning s after a 1 0-h o vernig ht

fast. W ithin e ac h fo od g ro up, e ac h subje ct ac te d as his o r he r

o wn contro l, being tested at the same time o f day and under

as similar conditions as possible. S ubjects w ere asked to

refrain fro m unusual ac tiv ity and fo od intake patterns, to

abstain fro m alco ho l and leg umes the day before a te st, and

to eat a similar meal the night before each test. When

subjects arrived at the lab in the mo rning , they co mple ted a

s ho rt que stio nnaire as se ss ing re ce nt f oo d intake and ac tiv ity

patte rns. A fas ting fing er-pric k blo od s am ple w as c olle cte d

and subjects w ere then given a test food and 220 mL water

(0 mm). W hen possible, foo ds w ere prese nted under a larg e

o paque plastic ho od w ith a ho le thro ugh w hich vo lunteers

pulled out pieces of the test food one at a time. This was an

atte mpt to m inim iz e be tw ee n- subje ct v ariatio n in c ephalic

phas e ins ulin s ec re tio n aris ing f ro m the s ens ory s ti mul atio n

associated w ith the anticipation and act of eating (27).

Ho w ev e r, th is w as n o t fe as ib le fo r th e liq u id fo o d s (yo g u rt

and ice cream), foods served in a sauce (baked beans and

lentils), or w ith milk (all of the breakfast cereals), w hich

were presented in standard bowls w ithout the hood.

S ubje cts w ere as ke d to e at and drink at a c om fo rtable rate .

Im me diate ly afte r finishing the test fo od, subjec ts re co rded the

time taken to eat the food and completed a questionnaire

as se ss ing v ari ous appe ti te re spo ns es and the f oo d's palatabil ity .

[These results are reported in a separate paper (28).] S ubjects

re maine d s eate d at table s in a qui et e nv iro nm ent and w ere no t

permitted to eat or drink until the end of the session (120 mm).

Fi ng er- pri ck blo od s am ple s ( 1.5 â€ ”2 .5m L) w ere c olle cte d

from warmed hands immediately before the meal (0 mm) and

15, 30, 45, 60, 75, 90, 105, and 120 mm after the start of the

meal (into plastic tubes that had been kept o n ice ) w ith use o f

an auto matic lanc et de vic e ( Auto cli x; B oe hring er M annhe im

A ustralia, Castle Hill, A ustralia). B lood samples w ere centri

fuged immediately after collection (1 miii at 12 500 X g at

ro om te mpe rature ) and pl as ma w as pipe tte d into c hill ed tube s

and im me diate ly s to re d at â€ ”¿2 0° Cuntil analy ze d (< 1 m o).

P las ma glucose concentrati ons w ere anal yzed i n dupl icate w ith

a C obas F ara auto matic s pe ctro pho to me tric analy ze r ( Ro che

D iagnostica, B asel, S witzerland) and the glucose hexokinase

enzymatic assay. The mean w ithin-assay and betw een-assay

pre cisio ns (CV s) w ere bo th < 6 %. Plasm a insulin c onc entra

ti ons w ere meas ured i n dupli cate by us ing an anti body-coated

X100 (1)

This equation is similar to that developed by W olever and

Jenkins (29) for calculating GI values. A glucose score (GS)

(not the same as a GI score, w hich is based on a 50-g carbo

hydrate portion) for each food w as also calculated by using the

same e quatio n w ith the co rrespo nding plasma g luco se re sults.

A nalysis of variance (A NOV A) and Fisher's probable least

s ignif icant-di fference test for multiple comparisons were used to

d ete rm in e s tati sti cal d if fe re nc es am on g th e f oo ds w i th in e ac h f oo d

group (S TA TV IEW S TU DEN T S OFFW ARE; A bacus Concepts

mc, Berkley, CA). Linear-regressionanalysis was used to test

as so ci ati on s b etw e en g luc os e and i ns ul in re sp ons es and n utri ti on al

indexes (MINITAB DATA ANALYSIS SOFFWARE, version

7.0; M initab Inc, S tate College, PA ). Test foods not containing a

p arti cu lar n utri en t w e re e xc lu de d f ro m th es e an al ys es . T he re fo re ,

s am ple s iz es f or the c orre lati ons be tw ee n i ndi vidual nutrie nts and

the mean G Ss and ISs vari ed f rom 32 to 36 . Mean resu lts f or w hi te

bread for each fo od g ro up w ere inc lude d in so me statistical anal

y se s, so these co rrelations w ere made w ith 4 3 v alues. One subje ct

f ro m the pro te in- ric h f oo d g ro up di d no t c om pl ete the f is h te st and

one s ubjec t f rom the breakfas t cereal group did not compl ete the

S ustain tes t. There fore , in to tal, 5 03 indiV idU al te sts w ere fully

completed.

S tepw ise-m ultiple-reg res sio n analy sis w as used to e xamine

the extent to w hich the different macronutrients and GSs ac

counted for the variability of the ISs (MINITA B DATA

ANALYSIS SOFTWARE). For this analys is , the indiv idual

w hite bread OS and IS results w ere included for the carbohy

drate-rich food group only; therefore, this analysis w as per

f orm ed w ith 4 46 i ndiv idual o bs erv atio ns fo r 3 8 f oo ds . Inc lud

ing the w hite bread results for each food group (n = 503)

sug gests that independent repeat tests w ere do ne fo r w hite



Food Glucose A UCInsulin

AUCInsulin AUC:

glucose AUCInsulin

A U C pe r g

carbohydrateInsulin

A UC per g

serving weightGlucose scoreInsulincoremol-miniLpmolmin/LpmolminL'g'pmolminL'g'%%


1269

TA BLE 4

A reas under the 1 20 -m m plas ma g luco se and insulin res po nse curv es (A UCs), ratio o f ins ulin A UC to g luc ose A UC, the ins ulin A UC per g

carbohydrate and perg serving weight, and mean glucose and insulin scores'

BreakfasterealsWhitebread156Â±2113557Â±1756108Â± 19295Â±38144Â±19100Â±0100Â ±0All-Bran59

Â±94 299 Â±61287 Â±1599 Â ±1425

Â±340 Â ±732 Â±Porridge80

Â±95093 Â±49374 Â±11139

Â±1313

Â±I60 Â±1240 Â±Muesli65Â±126034Â±813118Â±18163Â±2234Â±543Â±746Â±5Special

K1 6 Â±1 4 8 3 8 Â±6 3 5 9 5 Â±1 4 1 9 5 Â± 1 5 4 7 Â±4 7

Â±9 6 6 Â±Ho n e y s ma c k s 9 1Â±1 9 1 2Â±1 5 6 1 8Â±1 2 1 8 9 Â±3 1 5 3Â±9 6 Â±7 6 7Â±6Su s t a i n 9 3Â±8 8 9 3 8Â±7 5 7 1 2Â±9 2 9Â±1 8 5 3Â±4 6 6Â±6 7 1Â±6Co r n f l a k e s 1

1 Â±1 1 8 7 6 8

Â±62388

Â±5 1 8 9 Â± 1 3 5 2 Â±4 7 6

Â±1175

Â±Gr o u p me a n â € ” 7 1 8 3 Â± 3 5 7 9 2 Â± 5 1 6 9 Â± 8 3 9 Â± 2 5 9 Â± 3 5 7 Â± 3 Ca r b o h y d r a t e r i c h

foodsWhitebread120Â±1312882Â±1901112Â±15281Â ±41137Â±20100Â±0100Â± 0Whitepasta50Â±114456Â±453156Â± 4891Â±922Â±246Â± 1040Â±5

p a s t a 7 4 Â±7 4 5 3 5 Â±5 7 4 6 7 Â± 1 9 3 Â± 1 2 2 1 Â±3 6 8 Â± 1 4 Â±Gr a i n b r e a d 6 8Â±9 6 6 5 9 Â±8 3 7 1 6Â±1 2 1 6 6Â±2 1 6 2Â±8 6 Â±1 2 5 6Â±6Br own r i c e 1 1 3Â±1 3 6 2 4 Â±6 1 6 5 8 Â±5 1 1 7Â±1 1 4 2Â±4 1 4Â±1 8 6 2Â±1 1 F r e n c h

f r i e s 7

Â±1 1 7 6 4 3 Â±713146 Â±292 9 Â± 1 9 8 2 Â±8 7 1 Â± 1 6 7 4 Â±2Wh i t e r i c e 1 2 9Â±1 6 8 1 4 3Â±6 8 3 6 9 Â±5 1 4 5Â±1 2 4 Â±3 1 1 Â±1 5 7 9Â±1 2Who l e me a l

b r e a d 1 6 Â± 1 4 1 1 2 3 Â± 1 4 2 1 2 2 Â±2 2 4 7 Â±3 1 1 1 1 Â± 1 4 9 7 Â± 1 7 9 6

Â±2 P o t a t o e s 1 4 8 Â± 2 4 1 3 9 3 Â± 1 4 6 7 1 2 Â± 1 9 2 8 4 Â± 3 3 8 Â± 4 1 4 1 Â± 3 5 1 2 1 Â± 1 1 Gr o u p me a n â € ” 8 4 1 Â± 4 6 1 1 6 Â± 8 1 8 2 Â± 1 6 2 Â± 5 8 8 Â± 6 7 4 Â± 8 P r o t e i n r i c

foodsWhitebread121Â±1917438Â±3154177Â±35387Â ±63185Â±33100Â±0100Â ±0Eggs36

Â± 1 1 4 7 4 4 Â± 1 1 7 1 3 5

Â±92934

Â±1 8 4 5 3

Â±642

Â±1631

Â±C h e e s e 4 2

Â±1 5 9 9 4

Â± 1 5 9 2 6 8 Â± 1 5 3 6 4 2 5 7 Â± 1 5 1 3 1 6

Â±2755

Â± 1 8 4 5 Â±3 Be e f 1 8 Â± 6 7 9 1 Â ±2 1 9 3 1 5 8 3 Â± 9 3 9 â € ” 5 Â± 1 4 2 1 Â± 8 5 1 Â± 1 6 Le n t i l s 6 3

Â±1 7 9 2 6 8

Â±2 1 7 4 3 7 Â± 1 3 3 2 5 Â±6 8 3 7

Â±9 6 2 Â±2 2 5 8

Â±2 Fi s h 2 9

Â± 1 4 9 3 5 Â±2 5 5 7 7 5 Â±5 2 â € ” 2 8 Â±6 2 8 Â± 1 3 5 9

Â±8Ba k e d b e a n s 1 1 Â±1 4 2 1 6 Â±3 7 7 6 1 8 3Â±4 4 5 4Â±8 7 5 7Â±1 1 1 1 4 Â±1 8 1 2 Â±1 9Gr o u p

me a n â € ” 9 9 8 3 Â± 1 3 2 5 8 5 Â±6 1 1 8 6 7 Â±5 4 5 6 5 3

Â±654

Â±7 6 1 Â±F r u i t Wh i t e

b r e a d 1 7 1 Â± 1 9 1 5 5 6 3 Â± 1 6 3 2 1 5 Â± 1 8 3 3 9 Â±3 6 1 6 6 Â±171 Â± 1 Â±Ap p l e s 8 3Â±7 8 9 1 9Â±9 1 1 1 8Â±1 8 1 5 2Â±1 5 2 Â±2 5 Â±6 5 9Â±4Or a n g e s 6 6Â±1 1 9 3 4 5Â±1 7 4 1 6 6Â±2 3 1 8 5 Â±2 1 1 5Â±2 3 9Â±7 6 Â±3Ba n a n a s 1 3 3

Â±1 2 1 2 4 4 5 Â± 1 3 5 3 1 8 Â±2 2 2 2 4 Â±2 4 4 5

Â±5 7 9 Â± 1 8 1 Â±Gr a p e s 1 2 6Â±1 4 1 2 2 9 3Â±1 1 9 1 1 3Â±1 9 2 1 6 Â±2 1 3 1Â±3 7 4Â±9 8 2Â±6Gr o u pme a n â € ” 1 7 5 1Â±6 5 1 2 4Â±1 1 9 4Â±1 1 2 8Â±2 6 1Â±5 7 1Â±3Sn a c k s

andonfectioneryWhitebread159Â±2915592Â±2376104Â±24340Â±52166Â±25100Â±0100Â±0Peanuts20

Â±73 47

Â±8 2 8 2 1 4 Â±8 8 5 6 4 Â± 1 5 3 8 Â±2 2 1 2 Â±4 2 Â±Po p c o r n 7 1Â±1 2 6 5 3 7Â±6 7 9 1 9Â±3 2 2 3 9 Â±2 5 1 3 9Â±1 4 6 2Â±1 6 5 4Â±9Po t a t o c h i p s 7 7Â±1 5 8 1 9 5 Â±1 5 7 7 1 6 9Â±7 8 3 6 7Â±7 1 1 8 6Â±3 6 5 2Â±9 6 1Â±1 4 I c e

c r e a m9 3

Â± 1 7 1 2 3 4 8 Â± 1 8 6 7 1 7 2 Â±3 8 4 7 9 Â±7 2 1 3 Â± 1 6 7 Â± 1 9 8 9 Â±3Yo g u r t 8 8Â±2 3 1 5 6 1 1Â±1 8 8 1 6 7Â±3 3 4 1 5 Â±4 8 6 5Â±7 6 2Â±1 5 1 1 5Â±1 3Ma r s B a r 9 8Â±1 1 6 6 8 2 Â±1 8 9 6 2 1 8Â±6 5 4 4 1 Â±5 3 9Â±3 5 7 9Â±1 3 1 2 2Â±1 5 J e l l y b

productsWhitebread129Â±1517599Â±3058188Â±64383Â ±67187Â±33100Â±0100Â ±0Doughnuts78Â±1412445Â±2402113Â± 21480Â±93191Â±3763Â±

Â±

3 5 Â±3 4 7 2 1 7 8 Â±5 4 4 6 7

Â± I 1 3 2 2 3 Â±5 4 5 6 Â± 1 4 8 2 Â±2 C r a c k e r s 1 3 9

Â±2 6 1 4 6 7 3 Â±2 6 8 6 3 3 1

Â± 1 4 3 5 4 Â±6 5 2 5 3 Â±4 6 1 1 8 Â±2 4 8 7 Â±2Co o k i e s 9 2Â±1 2 1 5 2 2 3Â±3 8 2 2 Â±5 7 4 3 6Â±1 1 2 9 8Â±7 5 7 4Â±1 1 9 2Â±1 5Gr o u pme a n â € ” 1 2 6 8 1Â±1 3 2 5 2 6 1Â±5 6 4 6 8Â±4 7 2 3 6Â±2 4 7 7Â±7 8 3Â±5

I 1 SE M. M ars Bar M ars Confectionary Australia Ballarat Australia; All-Br an Special K H oneysmacks Sustain and Cornf lakes: Kellogg s

Australia Pty Ltd. Pagewood, Australia.



2 7

HO LT ET A L

bread, w hich artificially increases the accuracy of any calcu

lation involving w hite bread.

RESULTS

Fa s tin g g lu co s e a nd in s ulin c on ce n tra tio ns

W ithin each food group, the subjects' average fasting

plasma glucose and insulin concentrations were not signif

icantly different am ong the foods. M ean fasting plasma

glucose concentrations did not vary significantly among the

six food groups, w hereas mean fasting insulin concentra

tions were more variable, ranging from â€œ¿42o 120 pm ol/L.

Fasting insulin concentrations were not m ore variable in

fem ales than in males and there were no significant differ

ences at various stages of the m enstrual cycle. A significant

correlation was found between m ean fasting insulin concen

trations and m ean BM I values for the six groups of subjects

r â €” ¿. 81 , P < 0 . 05 , n 6 ) .

Pos tp r and i al g l u c o seand i n s u li n r es ponses

A s w ith any biological response, there w as betw een-subject

variation in the glucose and insulin responses to the sam e food.

Two-way ANOVA was used to examine the ranking of each

subject's responses to the different test foods within a food

group (ie, interindividual variation). There w ere significant

differences am ong the subjects in the rank order of their glu

cose AUC responses except within the fruit and protein-rich

food groups. T here w ere also significant differences am ong the

subjects' rank order of insulin AUC responses w ithin all food

groups. H ow ever, individual subjects w ithin each food group

consistently produced relatively low , m edium , or high insulin

responses. F urtherm ore, subjects produced their low est insulin

responses for the least insulinogenic foods and their highest

insulin responses for the m ost insulinogenic foods w ithin each

f oo d g ro up .

10 0

There were large differences in mean glycemic and insulin

responses to the foods, both within and between food groups.

M ean glucose and insulin AUC results, mean GSs and ISs, and

the mean ratios of insulin to glucose AUCs (the amount of

insulin secretion in relation to the blood glucose response) are

listed in Table 4. M ean GSs and ISs were calculated for each

fo od g rou p b y av erag in g the sco res fo r all test fo ods w ith in the

food group. On average, the snack food group produced the

highest food group IS (89 ) , followed by bakery products

(8 3 ), carbo hy drate-rich fo od s (74 ), fruit (71 ), p ro tein

rich foods (61 ), and breakfast cereals (57 ). Average GSs

for the food groups did not follow the same rank order (Figure

1 ). T h e c arb o hy dra te -ric h fo od g ro u p p ro d uc e d th e h ig h e s t

average GS (88 ), followed by bakery products (77 ), snack

foods (65 ), fruit (6 1 ), breakfast cereals (59 ), and protein

ric h fo od s (5 4 ). In te re s tin g ly , th e G S ra n k o rd er is n ot p ro

portional to the average total carbohydrate content of each food

group, w hich highlights the influence of other food factors (eg,

fiber and processing) in determ ining the rate of carbohydrate

d ig es tio n a nd a bs or pt io n.

Overall, am ong the 38 test foods, jellybeans produced the

highest m ean IS (160 Â ±16 ), eightfold higher than the low est

IS (for peanuts: 20 Â ±5 ) (F igure 2). W hite bread, the stan

dard food, consistently produced one of the highest glucose and

insulin responses (peak and AUC) and had a higher IS than

m ost of the other foods (84 ). A ll of the breakfast cereals w ere

significantly less insulinogenic than w hite bread (P < 0.001).

All-Bran and porridge both produced a significantly lower IS

than the other cereals (P < 0.001), except muesli. Despite

containing more carbohydrate than porridge and m uesli, A ll

Bran produced the lowest G S. Baked beans, which contain

considerably m ore carbohydrate than the other protein-rich

foods, produced a significantly higher GS and IS (P < 0.001).

O n a ve ra ge , fis h e lic ite d tw ic e a s m uc h in su lin se cre tio n a s d id

the equivalent portion of eggs. W ithin the fruit group, oranges

and apples produced a significantly lower GS and IS than

lucoseco re

Insulincore

8 0

60 J

40

20

0

4)

U

â€˜¿ )

C

4)

4)

E

0.

0

B reak fas t

cerea ls

C arb oh yd ra te - B ak ery

ric h fo o d s p ro d uc ts

Protein

ric h fo o ds

Fru it S n a c ks a n d

con fec t ionery

FIGURE 1. Mean(Â±SEM)glucoseand insulinscoresforeach food group.



White

breac . IPrth ifr@c.

Doughnu t s

Croissants

Ca k e

Crack e r s

Cook ie s

PeanutsI=@lIIIâ€•;â€•t-1@PopcornI=â€•iiPotato

chipsI='=â€•iIcecream lâ€•;lIY ogurtI

lip

IN SUL IN IN DEX OF FOO DS

1271

I@ 1

ft

All-Bran

Porridge

Mues l i

OUSL4I J

Eg @

Cheese

Bee f

Lentils

Fish

e ns

Apples

O r a n g e s

Bananas

Gr a p e s

B r ow n p as ta

W h i te p as ta

Grain bread

r o wn r i c e

Frenchfries

Whit e r ic e

Who l e -mea l b r e ad

-i

.1

1=1

Je l lybeans

1 0 0 2 0 0

In su lin s co re ( )

F IG U R E 2 . M e an ( Â ±S E M ) i ns ul in s co re s f o r 1 00 0- Id p ort io ns o f t he t es t f o od s. W h it e b r ead w a s t he re f ere nc e f o od ( in su li n s co re = 1 00 ) . A l l -B r an

c ere al, S p ec ial K c ere al, H on ey sm ac ks c ere al , S u st ain c ere al , an d C orn flak es , K e ll og g's A u strali a Pty L td . Pag ew o od , A u strali a; M a rs B ar c an dy b ar, M a rs

Confectionary A ustralia, Ballarat, A ustralia.

grapes and bananas (P < 0.05 to P < 0.001), despite contain

ing a sim ilar am ount of carbohy drate.

Potatoes produced signif icantly higher G Ss and IS s than all

of the other carbohy drate-rich foods. W hite bread produced a

higher GS and IS than grain bread (P < 0.05 and P < 0.001

respectiv ely ), but w hole-m eal bread and w hite bread had sim

ilar scores. W hite and brow n rice had sim ilar GS s and IS s, as

did w hite and brow n pasta. A m ong the bak ery products, crack

ers produced a signif icantly higher G S than the other test f oods,

but there w ere no signif icant dif f erences in IS s w ithin this

group (all tended to be high). A m ong the snack foods, jelly

beans produced a signif icantly higher G S and IS than the other

foods in this group. Despite containing sim ilar am ounts of

carbohy drate, jelly beans induced tw ice as m uch insulin secre

tion as any of the four f ruits. T he candy bar and y ogurt, w hich

both contained large am ounts of sugar in com bination w ith

f at or protein, produced relativ ely high IS s. Popcorn and potato

chips elicited tw ice as m uch insulin secretion as peanuts

P < 0.05 and P < 0.01, respectively .

S ignif icant dif f erences w ere found both w ithin and am ong

the food groups w hen the insulin A U C responses w ere

ex am ined as a function of the food's carbohydrate content

(T able 4). On av erage, protein-rich foods produced the

highest insulin secretion per gram of carbohy drate (food

group m ean: 18 607 pm ol . m m . L ' . g@ 1) (because of

their m ostly low carbohy drate contents), follow ed by bak ery

products (468 pm ol . m m . L @ . g 1), snack foods (416

pm ol . m m . L â€˜¿g 1) f ruit (194 pm ol . m m . L â€˜¿g 1),

carbohy drate-rich foods (182 pm ol . m m . L @ . g'), and

break fast cereals ( 169 pm ol . m m . L â€˜¿g â€˜¿ . hen the



1272

HOLT ET A L

S

insulin A U C re spo nse w as e xam ine d in re latio n to the fo od's

serving size (g), the bakery products were the most insuli

nogenic (food group mean: 236 pmol . mm . L â€˜¿g 1), fol

lowed by snack foods (163 pmol . mm . L â€˜¿g 1), carbo

hydrate-rich foods (62 pmol . mm . L â€˜¿g 1 ), pro te in- ri ch

foods (53 pmol . m m . L â€˜¿g 1), breakfast cereals (39

pmol . mm . L â€˜¿g â€˜¿ ,nd fruit (28 pmol . mm . L â€˜¿g 1).

The se re sults re fle ct the insulino ge nic e ffe cts o f pro te in and

fat.

In su lin re s po ns e s in re la tio n to g lu co se re s po ns e s

Overall, mean glucose and insulin AUC values were posi

tively correlated (r = 0.67, P < 0.001, n = 43), as were the

peak glucose and insulin values (r = 0.57, P < 0.001, n = 43).

He nc e, the m ean GS s and IS s w ere hig hly c orre late d (r = 0 .7 0,

P < 0.001, n = 38) F igure 3). The peak glucose concentration

(c ha ng e fro m fa s tin g) c orre la te d p os itiv e ly w ith g lu co s e A UC

values (r = 0.74, P < 0.001, n = 43) and peak insulin

c onc entratio ns w ere pro po rtio nal to the ins ulin A UC v alue s

r 0.95, P < 0.001, n 43 . In addition, the observed GSs

fo r 1 00 0-U po rtio ns o f the fo ods c orre late d w ith pre vio usly

publis he d G I v alue s bas ed o n po rtio ns o f fo ods c ontaining 5 0 g

carbohydrate (r = 0.65, P < 0.001, n = 32). Six test foods

(chocolate chip cookies, eggs, cheese, beef, fish, and Hon

e y s m a c k sc e r e a l w e r e n o t i n c l u d e d in t h is a n a l y s isb e c a u s eG I

v al ue s w ere no t av ail abl e.

Insulin AUC values were divided by glucose AUC values to

determine w hich foo ds w ere marke dly insulino genic relativ e to

th eir g ly c e m ic e ffe ct (T a ble 4 a n d Fig u re 4 ). O n a v e ra g e , th e

p r o t e in r ic h fo o d s s t im u la t e da la r g e a m o u n t o f in s u lin s e c r e

ti on re lati ve to thei r g lycemic respons e, f ol low ed by the bakery

products, snack foods, fruit, carbohydrate-rich foods, and

b re a k f a s t c e r e a ls .

4)

1 @

0

4)

4)

20 0

FIGURE 3. Relationbetweenthe meanglucoseandinsulinscores(r =

0.74, P < 0.001, n = 38).

R e la t io n s b e t w e e n m e t a b o lic r e s p o n s e sa n d n u t rie n t

contents of the foods

C orre lati ons betw een the macronutri ent compos iti ons o f the

te st fo ods and the m ean IS s are s ho wn in Fig ure 5 . The po rtio n

size (energy density: kJ/g), water, and fiber contents of the

foods were not significantly related to the mean ISs. The

re latio n be tw ee n pro te in c onte nts and IS s w as ne gativ e but no t

significant (r â€”¿0.24, = 38). The mean IS s w ere positively

related to total carbohydrate content (r = 0.39, P < 0.05, n =

36) and sugar content (r = 0.36, P < 0.05, n = 36), but were

no t s ig nif ic antl y re late d to s tarc h c onte nt ( r = â€ ”¿0 .0 9, = 3 0).

Fat c onte nt w as ne gativ ely re late d to the m ean IS (r = â€ ”¿0.27,

N S, n = 3 6). W hen e xpressed as a percentag e o f to tal e nerg y,

fat (r = â€”¿0.27,S, n = 36) and protein (r = â€”¿0 .2 4,S , n =

38) w ere negatively associated w ith the mean IS , w hereas total

carbo hydrate w as po sitive ly related (r = 0 .3 7, P < 0.0 5, n

36).

R el ati ons be tw een the G Ss and the nutri ents l argel y f ol low ed

the sam e dire ctio ns as the IS c orre latio ns. M ean GS s w ere no t

sig nific antly related to the fo ods' se rving siz es o r w ater o r fiber

c onte nts . M ean G Ss c orre late d ne gativ ely w ith f at ( r = â€ ”¿0.38,

P < 0.05, n = 36) and protein r = â€” ¿0.38,< 0.05, n = 38)

c onte nts , and po siti ve ly w ith to tal c arbo hy drate c onte nt ( r =

0.32, NS, n = 36). Unlike the ISs, the GSs were significantly

related to starch content (r = 0.43, P < 0.05, n = 30) but not

sug ar c ontent (r = â€”¿0.07,S , n = 36). When expressed as a

percentageof total energy, fat (r = â€”¿0.38, < 0.05, n = 36)

and protein (r = â€”¿0.39, < 0.05, n = 38) w ere negatively

as soc iated w ith mean G Ss , w hereas to tal carbohydrate content

was positively related (r = 0.46, P < 0.01, n = 36).

Stepw is e- multi pl e- regress ion anal ys is o f the 446 i ndiv idual

re sults fo r the 3 8 f oo ds w as pe rf orm ed to de te rm ine the e xte nt

to w hic h the m ac ro nutrie nts and G Ss ac co unte d fo r the v an

abil ity o f the ISs . U nf ortunate ly , i t w as not pos si bl e to generate

a s ing le m ul tiple -re gre ss io n e quati on that inc lude d all o f the

macronutrients because some pairs of nutrients w ere highly

corre lated ( eg , f at and prote in , f iber and w ater, to tal carbohy

drate and s ugar or s tarch, and s ugar and s tarch) . T he regress ion

e quatio n that inc luded all o f the m ac ronutrie nts had unac ce pt

ably high variance inflation factors. Therefore, tw o separate

regression equations w ere generated that w ere limited to the

facto rs that w ere me asured and no t inte rde pendent. Equatio n 2

inc lude s fat but no t pro te in, w he re as e quatio n 3 i nc lude s pro

te in but no t f at:

IS = 72.4 + 0.383 GS â€”¿.88 f at â€”¿.1 03 w ate r

+ 0 .5 09 s u ga r â €”¿.4 2 1 s ta rc h (2 )

fo r w h ic h S D = 3 7 .3 4 ,R 2 = 3 3 .1 ,a n d a d ju s t e dR 2 = 3 2 .4 .

P values significance found in the linear-regression analysis

f or the as soc iati ons be tw een the i nd iv idual nutri ents and the IS)

are as follows: OS and water (P < 0.000), fat (P < 0.001),

sugar (P < 0.005), and starch (P < 0.036).

IS = 23.2 + 0.383 05 + 0.785 protein â€”¿.0 98 w ate r

+ 1.2 9 s ug a r + 0 .3 7 7 s ta rc h (3 )

for w hich SD = 37.42, R2 = 32.8%, and adjusted R2 = 32.1%.

P values are as follows: GS, water, and sugar P < 0.000);

protein (P < 0.003); and starch (P < 0.02).

.

S

â € ¢ ¿ â € ¢

S

0

0

G lu co s e s c ore ( )




1273

Porridge

All-Bran

Comfiakes

S pe c ia l K

S u s t a i n

Honeysmacks

Mues l i

Brownrice

Brownpa s t a

White rice

Gra in b r e ad

Whitebread

Potatoes

Whole-meal bread

Frenchries

White pasta

Bananas

Grapes

Apples

Oranges

Popcorn

Jellybeans

Yogurt

Chips

Icecream

Peanuts

MarsBar

Doughnuts

C a k e

Coo id e s

Crackers

Croissant

Eggs

Beans

Cheese

Lentils

Fi s h

Be@

I. .

I-

I- i

@ 1

-I

-t

500 1000 1500 2000 2500

3000

Insu linAU C/G iuco seUC

FIGURE 4. Ratio of insulin area underthe curve (AUC) to glucose AUC responses.I Â±SEM. All-Brancereal, Special K cereal, Honeysmacks cereal,

Sustaincereal,and Cornflakes,Kello gg'sAustraliaPty Ltd.Pag ew oo d,Aus tralia;MarsBar candybar, MarsConf ectionaryAus tralia,Ballarat,Australia.

Linear-regression analysis of the individual OS and IS re

s ults had an R2 value of 2 3%. There fore , the g ly ce mic respo nse

w as a significant predictor of the insulin response, but it

ac co unte d f or o nly 2 3% o f the v ariability in ins uli ne mi a. The

macronutri ents ( prote in or f at, w ater, s ugar, and s tarch) w ere

also significant predictors but together accounted for only

ano the r 1 0% o f the v ariabil ity o f the i ns ulin re spo ns es . Thus ,

we ca n e xp la in on ly 33 o f th e va ria tio n o f th e in s ulin re

s po nse s to the 3 8 fo ods s tudie d.

DISCUSSION

The re sults o f this s tudy c onf irm and als o c halle ng e s om e o f

o ur bas ic as sum ptio ns abo ut the re lati on be tw ee n f oo d intake

and insuline mia. W ithin each foo d g roup, there w as a w ide

range o f insulin res po nse s, despite similarities in nutrie nt co rn

position. The important Western staples, bread and potato,

w ere among the most i ns ul inogen ic f oods . S imil arl y, the h ighl y

ref ined bakery products and snack foods induced substantial ly

m ore in su lin s e c re tio n p e r kilo jo u le o r p er g ra m o f fo od th a n

did the o the r te st fo ods . In c ontras t, pas ta, o atm eal po rridg e,

and A l l- Bran cereal produced re lati ve ly l ow i ns ul in respons es ,

de spi te the ir hig h c arbo hy drate c onte nts . C arbo hy drate w as

quanti tati ve ly the major macronutri ent f or most f oods . T hus , i t

is no t surpris ing that w e o bs erv ed a stro ng c orre latio n betw ee n

G Ss and IS s (r = 0 .7 0, P < 0 .0 01 ). H ow ev er, s om e pro te in

and fat-ric h fo ods (e gg s, be ef, fis h, le ntils , c he ese , c ake , and

do ug hnuts) induced as much insulin sec retio n as did so me



0

U)

0 20 40 60

P ro t e i n g l s e rv l n g )

10 0

S S

0 20 40

60

1274

H OL T ET A L

20 0

10 0

0

14 0

S

S 5

S

S

S

S

S .

â € ¢ ¿ ,.. S

S Iâ€¢

S

0 10 20

F ib e r 9 / s e r v in g )

20 0

10 0

S

S

S

U)

C

a

S

U)

U)

C

S

@ f S 5 5

I

S t a rc h 9 /s e rv in g )

S S

200W

10 0

0

@ S

S

S 55

55

S S

1@ b5 s

S

F a t 9 /s e rv in g )

S

0 20 40 60

T otal carbo hy drate (9 /se rv in g)

0 i I

0 20 40 60

S u g a r 9 / s e r v i n g )

14 0

S

10 0

S

S

S

S

S

S S

60

7c1

0

0 10 20 30

F IG U R E 5 . R e lat io ns b et w ee n t he n ut ri en t c on te nt s o f t he t es t f o od s an d t he m e an i ns ul in s co re s. F ib er: r = â€ ”¿0.10,S , n = 32; p rotein: r = â€”¿0.24,

N S , n = 3 8; to tal c arb oh yd rate :r = 0 .3 9, P < 0 .0 5, n = 3 6; s ug ar: r = 0 .3 6, P < 0 .0 5, n 3 6; s tarc h:r = â€ ”¿0.09,S , n = 3 0; f at: r = â€”¿0.27,<

0.05, n = 36.

c arb oh yd rate -ric h f oo ds (e g, b ee f w as e qu al to b ro w n ric e an d

f ish w as e qu al to g rain b read ). A s h yp othesiz ed , sev eral f oo ds

w ith sim ilar G S s h ad disp arate IS s (eg , ice cre am an d y og urt,

b ro w n ric e an d b ak ed be ans, cak e an d ap ples, an d do ug hnu ts

an d b ro w n p as ta). O v erall, th e f ib er c on te nt d id n ot p re dic t th e

m ag nitu de o f th e in su lin re sp onse. S im ilar IS s w e re o bserv ed

f or w h ite an d b ro w n p asta, w h ite an d b ro w n rice, an d w h ite and

w hole-m eal bread. A ll of these foods are relatively ref ined

c om p are d w i th th eir t rad itio nal c ou nte rp arts . C o lle ct iv e ly , th e

fm dings im ply that ty pical W estern diets are lik ely to be

sig nif ican tly m o re in sulin og en ic th an m o re trad itio nal d iets

based on less ref in ed f ood s.

I n t hi s s tu dy , we c ho se t o t es t i so en erg et ic p ort io n s o f f o od s

rat he r t han e qu al- carb oh y drate s erv in gs to d ete rm i ne th e in su




1275

lin respo nse to all of the nutrients in the foo ds as normally

consumed. A standard portion size of 1000 kJ was chosen

because this resulted in realistic serving sizes for most of the

foods except apples, oranges, fish, and potatoes. A lthough

some of the protein-rich foods may normally be eaten in

s malle r quantiti es , f is h, be ef , c he es e, and e gg s s till had larg er

ins ul in re spo ns es pe r g ram than did m any o f the f oo ds c ons is t

ing pre do minantly o f c arbo hy drate . A s o bs erv ed in pre vio us

studies, consumption of protein or fat w ith carbohydrate in

creases insulin secretion compared w ith the insulinogenic ef

fect of these nutrients alone (22, 30â€”32). T his may partly

e xplain the m arke dly hig h ins ulin re spo ns e to bake d be ans.

D ri ed hanc ot be ans , w hic h are s oake d and bo ile d, are li ke ly to

hav e a lo w er IS than c om me rc ial bake d be ans , w hic h are m ore

readily digestible.

T he resu lts conf irm that i ncreas ed i ns ul in s ecre ti on does not

ac co unt fo r the lo w g ly ce mic re spo nse s pro duc ed by lo w-G I

fo od s s uch a s p a s ta , p orrid g e , a nd All-Bra n c e re a l 33 ). F u r

thermore, equal-carbohydrate servings of foods do not neces

sarily stimulate insulin sec retio n to the same ex tent. Fo r ex am

ple , i soenerge ti c s erv ings o f pas ta and potatoes both contai ned

=,%50g carbohydrate, yet the IS for potatoes w as three times

g reate r than that fo r pasta. S im ilarly , po rridg e and y og urt, and

w ho le -g rain bre ad and bake d be ans , pro duc ed disparate IS s

despi te their s imilar carbohydrate contents . These f indings , l ike

o thers , chal lenge the s ci enti fi c bas is o f carbohydrate exchange

tabl es , w hi ch as sum e that po rti ons o f di ff ere nt f oo ds c ontaini ng

1 0â€ ”15g carbo hy drate w ill hav e equal phy sio lo gic e ffec ts and

w ill re quire e qual am ounts o f e xo ge no us insulin to be m etab

o liz ed. It is po ssible that preprandial insulin dos es fo r patients

with NID DM could be more scientifically estimated or

m atc he d o n the basis o f a m eal's av erag e ins uline mic e ffe ct in

healthy indiv iduals, rather than o n the basis o f the meal's

carbohydrate content or 01 . Further research i s required to te st

this hy po the sis. The adve nt of intensive insulin the rapy and the

added risk of hypoglycemia increases the urgency of this

re se arc h ( 34 ).

Our study w as undertaken to test the hy po thesis that the

po stprandi al ins ul in re spo ns e w as no t ne ce ss ari ly pro po rti onal

to the blo od g luco se respo nse and that nutrients o ther than

c arbo hydrate influe nc e the o ve rall le ve l o f insulinem ia. M ulti

pIe -re gre ss io n analy si s o f the i ndi vidual re sul ts s ho we d that the

g ly ce mic re spo ns e w as a sig nific ant pre dic to r o f the ins ulin

respo nse, but it acco unted fo r o nly 2 3% o f the variability in

i ns ul inemia. T he macronutri ents ( prote in or f at, w ater, s ugar,

and s tarc h) w e re als o s ig nif ic ant pre dic to rs , but to ge the r ac

counted for only another 10% of the variability of the insulin

re spo nse s. Thus, w e c an e xplain o nly 3 3% o f the v ariatio n o f

the ins ulin re spo nse s to the 3 8 fo ods unde r e xaminatio n. The

lo w R 2 v alue indic ate s that the m ac ro nutrie nt c om po sitio n o f

fo od s h as re la tive ly lim ite d p owe r fo r p re dic tin g th e e xte nt o f

po stprandial ins uline mia. The rate o f starc h dig estio n, the

am ount o f rapidly av ai lable g luc os e and re sis tant s tarc h, the

deg re e ofo sm olality , the v isco sity o fthe g ut's co nte nts , and the

rate of g as tric em pty ing m ust be o ther im po rtant facto rs influ

e nc ing the de gre e o f po stprandial i ns ulin s ec re tio n. F urthe r

research i s required to examine the re lati on betw een pos tpran

dial insulinemia, food form, and various digestive factors for a

m uc h larg er rang e o f fo ods to pro duc e a re gre ss io n e quatio n

w ith g re ate r pre dic tiv e v al ue .

Dietary guidelines for healthy people and persons with

NIDDM have undergone considerable change and will con

tinue to be mo dified as o ur understanding of the relatio ns

be tw ee n di etary patte rns and dis eas e im pro ve s. The re is c on

c ern that hig h- carbo hy drate di ets m ay i nc re as e triac yl gl yc ero l

concentrations and reduce high-density l ipoprote in concentra

tions (35, 36). The use of diets high in monounsaturated fat is

an atte mpt to o ve rc om e the unde sirable e ffe cts o f s om e hig h

carbohydrate d ie ts on plas ma l ip ids ( 37â€”39).H ow ever, d ie ts

high in monounsaturated fat are unlikely to facilitate w eight

lo ss . A lo w -f at die t bas ed o n le ss -re fine d, c arbo hy drate -ric h

foods w ith relatively low IS s may help enhance satiety and aid

w eight loss as w ell as improve blood glucose and lipid control

(4).

The re sults o f this s tudy are pre lim inary but w e ho pe the y

s tim ulate dis cus si on and f urthe r re se arc h. A ddi tio nal s tudi es are

n ee de d to d ete rm in e w he th er th e IS c on ce pt i s u se fu l, re pro duc ib le

around the w orld, predictable in a mix ed-me al c ontex t, and cliii

i cal ly us ef ul i n th e tre atm ent o f d iabe te s m el li tu s, h ype rl ip id em ia,

and overweight. Studies examining the relation between postpran

dial insul inemia and the storage and oxidation of fat, prote in, and

c arb oh yd rate m ay p ro vi de f urth er i ns ig ht i nto th e re lati on b etw e en

fu e l m e t a bo lis m a n d s a t ie ty , a n d e s t a b lis h w h e th e r lo w -in s u lin e

mic diets can facilitate greater body fat loss than isoenergetic

high- insWinemic die ts .

We thank Efi Farmakalidisfor her assistancein the planningof this

study and NatashaPorter forher technical assistance with the experimental

work for the carbohydrate-rich food group.

REFERENCES

1 . Jenkins DJA, Wolever TMS, Kalmusky J, e t a l. Low g lycemic index

carbohydrate foods in the management of hyperlipidemia. Am J Clin

Nu t r 1 9 8 5; 4 2 : 6 4 â €” 17

2 . Je nk in s D JA , W o le ve r T MS , V uk san V . e t al . N i bb li ng v s g org ing :

metabolic advantages of increased meal frequency . N EngI J Med

1989;32l:929â€”34.

3. Bymes SE, Brand Miller JC, Denyer GS. D evelopment of insulin

re si sta nc e i n rats af te r l ow - v s hi gh- am yl os e di ets . B ac he lo r o f S ci enc e

H on ou rs T he si s, D e partm en t o f B io ch em is try , T he U ni ve rs ity o f

Sydney , 1993 .

4 . S labber M , B arnard HC, Kuy l JM , D annhaus er A , S chall R. Effects o f

low-insulin-response,energy-restricted dieton weight loss and plasma

insulin concentrations in hyperinsulinemic obese females . Am J Clin

Nutr l994;60:48â€”53.

5. Lerer MetzgerM Rizkalla SW Luo J et al. Effects of long term

low-g lycaemic index s tarchy food on plasma g lucose and l ipid con

centrations and adipose tissue cellularity in normal and diabetic rats .

Br J Nutr 1996;75:723â€”32.

6. Bymes SE, Brand MillerJC, DenyerGS. Amylopectinstarch pro

motes the development of insulin resistance in rats. J N utr

1995;125:1430â€”7.

7 . S a lmÃ© ro n,As che r ioA, Rimin EB, e t a l .Die ta ry f ibe r , g lycem ic load ,

and risk o f N DDM in men. Diabete s Care l997 ;20 :545-50.

8 . S al mÃ © ro nJ, M ans on JE , M ei r J, e t al . D ie tary f ibe r, g ly ce mi c l oad, a nd

ri sk o f non-insul in-dependent diabe te s me ll itus in women. JAMA

l 9 9 7; l 2 : 4 7 2 â € ”7

9. Modan M, Halkin H, Almog 5, et al. Hyperinsutinemia:a link between

hy pertensio n, o be sity and gluco se into le ranc e. J Clin Inve st

l985;75:809â€”17.

1 0. Z av aro ni I, B o no ra E , P ag li ara M , e t a . R is k f ac to rs f or c oro nary

artely disease in healthy persons w ith hyperinsulinaemia and normal

glucose tolerance. N Engi J Med l989;320:702â€”6.



1276

HOLT ET AL

11 . DeFronzo RA, Ferrannim E. Insulin resi stance . A multi face ted syn

drome responsible for NIDDM, obesity , hypertension, dyslipidemia,

and atherosc lero ti c card iovas cu lar di seas e. D i abe te s C are

199l;14:173â€”94.

1 2. D eFro nzo RA , B onado nna RC, Fe rrannini E. Patho gene sis of

NIDDMâ€”abalanced overview. Diabetes Care 1992;15:3l8â€”68.

13 . Coldi tz GA, Manson JE, StampferMi , RosnerB, Willett WC, Speizer

FE. D iet and risk of clinical diabetes in women. Am J Clin N utr

1992;55:lOl8â€”23.

14. HaffnerS M, Valdez BA, Hazuda HP,et al. Prospective analysis of the

insulin-resistance syndrome (syndrome X). Diabetes l992;41:715-22.

1 5. L il li oja 5 , M o ft D M , S prau l M , C t al . In su li n re si stan ce an d i ns ul in

secretory dysfunction as precursors of non-insulin-dependentdiabetes

me ll itus. N Eng l J Med 1993;329:1988â€”92 .

16. StoutR. Hypennsulinaemiaas a riskfactorforcardiovasculardisease.

hi t D ia be te s F ed B ul l 1 99 4;3 9:l 8â€ ”9 .

17 . Zavaroni I, Bonini L, Fantuzz i M, e t al . Hyperinsul inaemia, obes ity ,

and syndrome X. J Intern Med l994;235:51â€”6.

1 8. W o le ve r T MS , Je nk in s D JA , Je nk in s A L , Jo ss e R G. T he g ly ce mi c

index : me tho do lo gy and c linic al implic atio ns. A m J Clin N utr

l99l;54:846-54.

19. ServiceFJ, Pizza PA, HallLD,Ctal.Prandialrequirementsn insulin

de pe nd en t d iab eti cs : e ff ec ts o f s iz e, ti me o f day , an d s eq ue nc e o f

m eal s. J C li ii E ndo cri no l M etab 1 98 3;5 7:9 31 -. 6.

2 0. S im ps on R W, M cD onald J, W ahl qv is t M L, A lte y L, O utc h K . M a

c ro nutri ents hav e di ff ere nt m etabo li c e ff ec ts i n no ndi abe ti cs and di n

betics. Am J Clii i Nutr 1985;42:449â€”53.

2 1. N uttal l FQ , M oo radi an A D , G anno n M C, e t al . E ff ec t o f pro te in

ingestion on the g lucose and insul in response to a s tandardized oral

glucose load. Diabetes Care l984;7:465â€”70.

22. Krezowski P, Nuttall FQ, Gannon MC, Bartosh NH. The effect of

protein ingestion on the metabolic response to oral glucose in normal

individuals . Am J Clin Nutr l986;44:847â€”56.

2 3. C oll ie r G , M cLe an A , O 'D ea K . E ffe ct o f c o- ing es ti on o f f at o n the

metabolic responses to slowly and rapidly absorbed carbohydrates .

Diabetologia 1984;26:50-.4.

2 4. Ganno n M C, N uttall FQ, W es tphal S A, S eaquist ER. The effect o f fat

and carbohydrate on plasma glucose, insulin, C-peptide, and triglyc

erides in normal male subjec ts . J Am Coll Nutr l993 ;12 :36 -41 .

25. Nuttall FQ, Gannon MC. Plasma glucose and insulin response to

macronutrients in nondiabetic and NIDDM subjects . Diabetes Care

l991;l4:824â€”38.

2 6. M org an L Insulin sec retio n and the entero -insular ax is . In: Rau PR,

e d. N utrie nt re gulatio n o f ins uli n s ec re tio n. Lo ndo n: Po rtl and Pre ss

Ltd. 1992:1â€”22.

2 7. Teff K. Cephalic phase insulin release in humans: mechanism and

function. In: Fernstrom J, Miller G, eds . Appeti te and body we ight

regulation. Sugar, fat and macronutrient substitutes . London: CRC

Press, 1993:37â€”49.

28. Holt SHA, Brand MillerJC, Petocz P. Farmakalidis E. A satiety index

o f common foods. Eur J Clin Nutr 1995 ;49 :675â€”90 .

2 9. W ol ev er T MS , Je nki ns D JA . T he us e o f the g ly ce mi c i nde x i n pre

dicting the blood glucose response to mixed meals. A m J Clin N utr

l986;43:167â€”72.

3 0. Es tri ch D , R av ni k A , S chli ef G , Fukay am a 0 , K ins el l L E ff ec ts o f

c o- ing es ti on o f f at and pro te in upo n c arbo hy drate -i nduc ed hy pe rg ly

cemia. Diabetes 1967;16:232â€”7.

3 1. R as mu ss en 0 , W hi th er E , A rn fre d J, H erm an se n K . C om pari so n o f

blood glucose and insulin responses in non-insulin dependent diabetic

patients . EurJ Clin Nutr l988;42:953â€”61.

32. Peters AL, Davidson MB. Protein and fat effects on glucose responses

and insulin requirements in subjects with insulin-dependent diabetes

me ll itus. Am J Clin Nutr 1993 ;58 :555-60 .

33 . Wolever TMS, Katzman-Rel le L, Jenkins AL, e t al . Glycaemic index

of 102complex carbohydrate foods in patients with diabetes . NutrRes

l994;l4:651â€”69.

3 4. The D iabe tes Co ntro l and Co mplicatio ns Trial Res earch Gro up. The

e ff ec t o f i nte ns iv e tre atm en t o f d iab ete s o n th e d ev el op me nt an d

progression of long-term complications in insulin-dependent diabetes

mellitus . N EngI J Med 1993;329:977â€”86.

35 . Coulston AM, Hollenbeck CB, Swislocki ALM, Reaven GM. Pers is

tence of hypertrig lyceridaemic effect of low-fat, high-carbohydrate

diets in NIDDM patients. Diabetes Care 1989;l2:94â€”lOl.

36 . Rive llese A, Giacco R, Genovese 5 , e t al . Effec ts o f changing amount

of carbohydrate in diet on plasma lipoproteins and apolipoproteins in

ty pe II diabetic patients. D iabe tes Care 1 99 0;1 3:4 46 -8 .

3 7. G arg A , G ru nd y S M , U ng er R H. C om pari so n o f e ff ec ts o f h ig h an d

low carbohydrate diets on plasma lipoproteins and insulin sensitivity in

patients with mild NIDDM. Diabetes l992;4l:1278â€”85.

38 . Garg A, Bantle JP, Henry RR, e t al . Effec ts o f vary ing carbohydrate

content of diet in patients with non-insulin-dependent diabetes. JAMA

1994;271:1421â€”8.

3 9. C am pbe ll L V , M arm ot P E, D y er JA , B o rk man M , S to rl ie n L H. T he

hi gh- mo no uns aturate d f at di et as a pra cti cal al te rnati ve f or N ID D M.

Diabetes Care 1994;17:177â€”82.

am j clin nutr 1997 holt 1264 76

Documents