REVIEW Effect of sensory perception of foods on appetite and food intake

15 Pages·2003·0.16 MB·English

InternationalJournalofObesity(2003)27,1152–1166
&2003NaturePublishingGroup Allrightsreserved 0307-0565/03 $25.00
www.nature.com/ijo
REVIEW
Effect of sensory perception of foods on appetite and
food intake: a review of studies on humans
LB S(cid:1)rensen1*, P M(cid:1)ller2, A Flint1, M Martens2 and A Raben1
1Department of Human Nutrition, Centre for Advanced Food Studies, The Royal Veterinary and Agricultural University,
Frederiksberg C, Denmark; and 2Department of Dairy and Food Science, The Royal Veterinary and Agricultural University,
Frederiksberg C, Denmark
Objective:Howmuchdothesensorypropertiesoffoodinfluencethewaypeopleselecttheirfoodandhowmuchtheyeat?The
objectiveofthispaperistoreviewresultsfromstudiesinvestigatingthelinkbetweenthesensoryperceptionoffoodandhuman
appetite regulation.
Content of the review: The influence of palatability on appetite and food intake in humans has been investigated in several
studies. All reviewed studies have shown increased intake as palatability increased, whereas assessments of the effect of
palatabilityusingmeasuresofsubjectiveappetitesensationshaveshowndivergingresults,forexample,subjectseitherfeelmore
hungryandlessfullafterapalatablemealcomparedtoalesspalatablemeal,ortheyfeeltheopposite,orthereisnodifference.
Whether palatability hasaneffecton appetite intheperiod following consumption of atest meal isunclear.
Severalstudieshaveinvestigatedwhichsensorypropertiesoffoodareinvolvedinsensory-specificsatiety.Taste,smell,texture
andappearance-specificsatietieshavebeenidentified,whereasstudiesontheroleofmacronutrientsandtheenergycontentof
thefoodinsensory-specificsatietyhavegivenequivocalresults.Differentstudieshaveshownthatmacronutrientsandenergy
content play a role in sensory-specific satiety or that macronutrients and energy content are not a factor in sensory-specific
satiety. Sensory-specific satiety may have an important influence on the amount of food eaten. Studies have shown that
increasingthefoodvarietycanincreasefoodandenergyintakeandintheshorttomediumtermalterenergybalance.Further
knowledgeabouttheimportanceofflavourinappetiteregulationisneeded,forexample,whichflavourcombinationsimprove
satietymost,thepossibleconnectionbetweenflavourintensityandsatiety,theeffectofpersistenceofchemestheticsensation
onpalatabilityandsatiety,andtowhatextentgeneticvariationintastesensitivityandperceptioninfluencesdietaryhabitsand
weightcontrol.
International Journal ofObesity (2003) 27, 1152–1166. doi:10.1038/sj.ijo.0802391
Keywords: appetite;hunger; food intake; palatability; sensory-specific satiety; intensity
Introduction impression people get from the sensory properties of foods
How much do the sensory properties of food influence playsaveryimportantroleinthewaytheyselecttheirfood
eating behaviour? Traditionally, sensory science and nutri- andhowmuchtheyeat.Seeninthelightofthecontinuing
tion,includingtheuncoveringofthemechanismsofhuman increaseintheprevalenceofoverweightandobesityinlarge
appetite regulation, have been two separate fields of parts of the world,1 it isimportant to identify all aspects of
research. However, within the last decades, there has been humanappetiteregulationtounderstandthistrend,andto
increasing interaction between the two fields, which has preventitfromprogressing.Thus,theobjectiveofthispaper
resultedinawiderangeofstudiesthathavecoveredaspects is to review results from studies investigating the link
frombothfields.Resultsfromthesestudiessuggestthatthe between the sensory perception of food and human eating
behaviour. The review begins with a presentation of how
appetiteandpalatabilityaremeasured.
*Correspondence:LBSørensen,DepartmentofHumanNutrition,Centre
for Advanced Food Studies, The Royal Veterinary and Agricultural
University,Rolighedsvej30,DK-1958FrederiksbergC,Denmark.
E-mail:[email protected] Methods
ThestudywassupportedbyOrklaFoodsAS,theDanishDairyResearch
Intheselectionofstudies,wehaveconcentratedonhuman
Foundation(DanishDairyBoard)andtheDanishGovernment.
studies with participation of healthy subjects with a body
Received1July2002;revised11April2003;
accepted7May2003 massindexof18.5–25.0kg/m2.Normal-weightsubjectswere
Sensoryscienceandappetite
LBSørensenetal
1153
chosen because it is important to identify how appetite is manipulations.Thisismadepossiblebyservinganadlibitum
regulatedinnormal-weightpeopleinordertoprovideabasis meal and asking the subjects to eat until they are full and
for understanding whether there is a difference in normal- then recording the amount of food eaten in grams or
weight and overweight people’s eating behaviour. The units.11,12 Stubbs et al13 have reviewed the use of VAS as a
subjects were adults (18–50y) and the women were pre- tool for measuring subjective appetite sensations and they
menopausal. Only studies conducted in Western countries concluded that VAS show some ability to predict aspects of
published in 1980 and onwards were included. Selected feedingbehaviourandactasausefuladjuncttomeasuresof
studies had a link between the sensory perception of food food, energy and nutrient intake; VAS show sensitivity to
and appetite sensations and/or energy intake. Therefore, experimentalmanipulations andshowgoodreproducibility
studies dealing only with macronutrients and appetite provided that they are used in within-subject designs. Flint
sensations and/or energy intake were not included. In the etal14havestudiedthereproducibility,powerandvalidityof
search for relevant studies, we used the databases Medline VASandtheyfoundthatusingapaireddesign,whenlooking
andFoodScienceandTechnologyAbstractsand,inaddition, for an effect size of 10%, 18 subjects are needed to achieve
referencelistswerescannedthoroughly. sufficientstudypowerandtherebyminimisingtheriskofa
type IIerror (‘false-negative’ result). Iftheeffect parameters
ofinterestarelimitedtomeanpostprandialappetiteratings,
Measuringappetiteinhumansubjects eightsubjectsareenough.
To prevent confusion about the terminology used when When investigating human appetite regulation using
describing aspects of appetite regulation, the following will female subjects, it is very important that the women are in
be applied in the review: ‘appetite’ is often referred to as a the same phase in their menstrual cycle on all test days
sensationrelatedtomaintenanceofeating,oftenadesirefor becauseoftheuncertaintyofhowthedifferentphasesofthe
somethingspecific,2butinthisreviewitisusedasageneral menstrual cycle influence appetite regulation. Dye and
term of overall sensations related to food intake. The Blundell15 have reviewed 30 studies from the period 1981
definitionof‘satiety’canbeseparatedintotwofunctionally to 1995 all investigating the connection between the
different terms: ‘satiation’ or ‘within-meal’ satiety, which menstrual cycle and energy intake. With the reservation
referstotheprocessthatleadstotheterminationofameal, fordifferentstudydesignskeptinmind,thegeneralpicture
and‘satiety’or‘between-mealsatiety’,whichisconsideredas shows that women have asignificantly higher spontaneous
the state of inhibition of eating, and therefore related to energy intake in the premenstrual phase (luteal phase)
the time interval until the next eating episode.3–5 ‘Hunger’ comparedtothepostmenstrualphase(follicularphase).
canbedescribedasanagging,irritatingfeelingthatsignifies
food deprivation to a degree that the next eating episode
should take place.2,3,6 The term ‘fullness’ is defined as a Measuringpalatabilityinhumansubjects
sensation of the degree of stomach filling, and ‘prospective There have been several suggestions to define palatability,
foodconsumption’asanindicatorofthesupposedamount but in his review on palatability and appetite Yeomans16
offorthcomingfoodintake. suggests that palatability should be defined as the hedonic
Within appetite research, different scales have been used evaluation of oro-sensory food cues under standardised
for the assessment of subjective appetite sensations. Fixed- conditions. The sensory properties of food determining
point scales (FPS) and visual analogue scales (VAS) are the palatability are taste, smell, texture, temperature, visual
most commonly used rating methods. FPS usually uses appearance, sound and trigeminal input (irritative sensa-
numerical scales, which vary greatly in complexity and tions).17
structure.Thus,theintervalsorpointsmayhavenumerical Palatability has been measured using either VAS or FPS
indications or different wording.7 Most often used in where subjects are asked questions like ‘how pleasant have
appetiteresearchareVAS,intheformofhorizontallinesof you found the food?’, ‘how satisfying have you found the
varyinglength,withwordsanchoredateachend,describing food?’, ‘how palatable is this food?’, ‘how attractive do you
theextremesofaunipolarquestion(forexampleforhunger: findthefood?’or‘howmuchdidyoulikethefood?’.7,8,18,19
‘Ihaveneverbeenmorehungry’/’Iamnothungryatall’).8–11 Palatability is classified according to how pleasant or well
Subjects are asked to make a vertical mark across the line likedthefoodis,themorepleasantorliked,thehigherthe
corresponding to their feeling at the present time. Quanti- palatability. Many investigators have used the terms palat-
fication of the measurement is made by measuring the abilityandpleasantnesssynonymously,butthereisadanger
distance from the left end of the line to the mark. The in this because of different interpretations of the term
subjects are asked about their appetite for a meal or their palatability.11
feelings of hunger, desire to eat, satiety, fullness and pro- The palatability of food is often assessed by taste tests
spective food consumption, with text expressing the most pleasantness ratings, where a small sample of food is tasted
positiveandmostnegativeratinganchoredateachend. and swallowed,20–23 or by using taste-and-spit pleasantness
Besidesusingappetiteratingstoassessappetitesensations, tests,wherethefoodsamplesaretastedandthenspitout.24
food intake has been used to measure the effect of dietary Investigationsofwhetherthesetestscanpredictconsumption
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1154
have shown that taste test pleasantness ratings are poor there were no differences in intake in the semi-liquid and
predictors of the actual amount a person will consume in sandwich meals (see Table 1). In this study, appetite was
comparisontootherpeople,buttheycanbereliableguides measuredusingFPS,andtheresultsshowednodifferencein
toconsumptiondifferenceswithinaperson,25andthattaste- effectofpalatabilityineitherofthemealtypes.Inastudy,by
and-spit pleasantness ratings cannot be used to predict Hill et al,8 where the effect of palatability on appetite was
consumption by individual subjects, but perhaps can be measuredsolelyusingVAS,adifferenceinratingsofdesireto
usedtopredicttheaverageconsumptionofaproduct,26and eat was observed as soon as the subjects saw the food, the
theresultsshouldbetakencautiouslywhentryingtopredict difference mostly being due to an increase in ratings in
ingestion.27 response to the highly preferred meal. The difference
persisted throughout the meal declining only after the
mealhadended8(seeTable1).Inaddition,theinvestigators
Studies investigating the link between the sensory foundthat2hafterthetestmeal,ratingsofdesiretoeatand
properties of food and appetite hunger were significantly higher after consumption of the
Influenceofpalatabilityonappetiteandfoodintake preferredfood.8
Theusualapproachinstudyingtheinfluenceofpalatability Sweetness is a majorcontributor to palatability, although
on appetite has been to offer subjects test foods with thereareindividualdifferencesinoptimallevelofsweetness.
differentlevelsofpalatabilityandcomparetheeffectofthe In two studies, the palatability of the test food was
testfoodsonhunger,satietyand/orsubsequentfoodintake. manipulatedbyusingyoghurtwithdifferentconcentrations
One way is to make use of subjects’ food preferences by of the intense sweetener, aspartame, and sucrose, respec-
making a subjective index of the palatability of test foods tively, hereby making use of the subjects’ preferences for
obtained from each subject using VAS and then comparing sweettaste.31,32Inbothstudies,yoghurtintakewasgreaterat
this index with the amount eaten by the subjects. Another the most preferred concentration than at all the other
way to make use of the subjects’ preferences for foods is by concentrations, and in the study using aspartame, sponta-
comparing the effect of a preferred test meal and a neous 24-h food intake was significantly greater following
nonpreferred test meal on appetite, the test meals being consumption of the preferred yoghurt, however, only in
equicaloricalthoughnotidenticalinmacronutrientcontent. men.31
Studiesusingtheseapproacheshaveshownincreasedintake Speculations as to whether factors other than palatability
aspalatabilityincreased12,28–30(seeTable1).Guy-Grandetal7 could explain differences in appetite can be avoided by
compared the effect of two levels of palatability in three manipulating the palatability of the test foods by factors
different types of test meals (conventional, semiliquid with little effect on the nutritional quality. To manipulate
and sandwich), and they observed that high palatability flavourintestmealsaddingmustard,spicessuchasoregano
increased intake only in the conventional meal, whereas orcumin,citricacidorsalt,orvaryingthequalityofcheese
Table1 Effectofpalatabilityusingpreferredfoods
Study Subjects Testmeal Hungerratings Foodintake
BellisleandLeMagnen28,29 3males 5differentflavouredsandwiches NR Subjectivepalatabilityscores
(1980,1981) 7females plusamixedmealinwhichall correlatedsignificantlywith
Withinsubjects 5flavourswereoffered. numberofunitseatenin5
Mealswereequalinenergy subjectsoutof10
Bellisleetal12(1984) 5males Leastormostacceptablesandwich NR Mostacceptablesandwich
5females type(units)identifiedforeachsubject typeincreasedintake
Withinsubjects
Hilletal8(1984) 12females Ahighlypreferredandaless Increasedafterthesightof NR
Withinsubjects preferredmealmadeforeach preferredfood.Increases
individual.Mealswere 2haftereatingthepreferredmeal
equicaloric
Spiegeletal30(1989) 9females High-andlow-palatablefillingin NR Highpalatabilityincreased
Withinsubjects spiralsofbread intake
Guy-Grandetal7(1994) 7femalesa Conventional,semiliquidand Nodifferences Conventional:highpalatability
Withinsubjects sandwichmealsat2levelsof increasedintake.Semiliquid:
palatability.Mealswere nodifference.Sandwich:
equalinenergy nodifference
NR:notreported.
aThewomenweretestedbetweenthe10thandthe25thdayoftheirmenstrualcycle.
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1155
Table2 Theeffectofpalatability
Study Subjects Preload Testmeal Hungerratings Foodintake
Bobroffand 8males None Banana-coladafrozen Morehungryandless Highpalatabilityincreased
Kissileff25(1986) 8females yoghurt:plain,6gcumin/ satiatedaftercuminmeals intake
Withinsubjects 1500gdrinkor4.5g
cumin/1700gdrink
Bellisleetal35 36malesand None Spinachmousseorbeef NR Highpalatabilityincreased
(1989) females andvegetablejelly:0% intake
Betweensubjects monosodiumglutamate
(MSG),0.6%MSG,1.2%
MSG
Rogersand Inallstudies: Inallstudies: Study1:notestmeal Study1:nodifferences Study1:NR
Blundell36(1990) 13males Nopreloadorbeef Study2:lunch,30min Study2:NR Study2:noeffect
23females consomme´with0%MSG, delay Study3:nodifferences Study3:noeffect
4subjects 10%MSG(study2only), Study3:lunch,2mindelay afterpreload.No
excluded 20%MSG differencesimmediately
Withinsubjects Fixedintake aftertestmeal.Duringnext
3hmorerapidrecoveryin
hungeranddesiretoeat
after20%MSGpreload
Warwicketal10 7males Tasty-CHO, Standardlunch,5hdelay Lesshungryandmorefull Noeffect
(1993) 3femalesa bland-CHO,tasty-fat, aftertastybreakfast
Withinsubjects bland-fat.Fixedintake
Yeomans18(1996) 27males None Pastaintomatosauce:no Nodifferenceexceptfor Highpalatabilityincreased
27females oregano:bland;0.27% fullnessratingsreflecting intake
Betweensubjects oregano:palatable;0.54% theamounteaten
oregano:strong
Yeomansetal33 16males None Pastaintomatosauce: Hungerwaslowerinthe Highpalatabilityincreased
(1997) Withinsubjects bland,nooregano; continuous/bland.Fullness intake
palatable,0.27%oregano. wasunaffectedbytypeof
2eatingstyles: foodeaten
continuous/interrupted
DeGraafetal19 9males Tomatosoup;nocitricacid: Avarietyoffoodproducts, Lowerhungerratingsafter Highpalatabilityincreased
(1999) 29females palatable,0.75gcitric 15or90mindelay thehigherpalatableadlib intakeofadlibpreload
Withinsubjects acid/100g:med. preloads. Noeffectontestmeal
palatable;1.5gcitricacid/ Withfixedpreloadintake intakeineitherthe15or
100g:unpalatable therewasnoeffectof 90mincondition
Adliborfixedintake palatabilityeitherinthe
15or90mincondition
Yeomansand 50males None Pastaincheesesauce:bland Nodifferences Highpalatabilityincreased
Symes11(1999) Withinsubjects withmildtastingcheese, intake
palatablewithamature
cheeseversionand
mustard
Zandstraetal34 11males None Sandwicheswithspread Fortheleastpleasant Palatabilityeffectonthe
(2000) 25females madeofbread:0%NaCl: bread,desiretoeatand firstday.Nodifferenceon
Withinsubjects leastpleasant;4%NaCl: fullnessincreasedover thefifthday
med.pleasant,2%NaCl: days,nodifferencefor
mostpleasant.For3 med.andmostpleasant
weeksF5dayspersalt bread
level
NR:notreported.
aThewomenweretestedduringthefollicularstageoftheirmenstrualcycle.
have all been used.10,11,18,19,25,33,34 In addition, monoso- but one,35 of the studies also investigated the effect of
dium glutamate has been used to enhance palatability.35,36 palatabilityonappetiteusingVASandawiderangeofresults
All, but two,10,36 of the above-mentioned studies show emerged when reviewing the studies: in Bobroff and
increasing intake as palatability increased (see Table 2). All, Kissileff’s25study,subjectsfeltlesshungryandmoresatiated
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1156
after the palatable meal reflecting the larger amount Summary. The results from studies investigating the influ-
eaten. Warwick et al10 found that after eating a palatable enceofpalatabilityonappetitesensationsinhumansgivean
fixed preload, subjects felt less hungry and more full unclearpicture,asincreasingpalatabilityinonestudyresults
compared to eating a less palatable preload. In Yeomans’18 in lower hunger and higher satiety ratings after consump-
study from 1996, subjects felt no difference except for tion,butinotherstudiesresultsinhigherhungerandlower
fullness,butthisreflectedtheamounteatenbythesubjects. fullness ratings or in higher hunger ratings and unaffected
De Graaf et al19 observed lower hunger ratings after the fullnessratings,whereasothersshownodifferenceineffect
higherpalatablepreload,reflectingthelargeramounteaten; of palatability. Studies also measuring the effect of palat-
after eating the fixed preloads, there was no difference in ability on energy intake (the ‘within-meal’ effect) show
effect of palatability. Yeomans and Symes11 observed no increasedintakeaspalatabilityincreases,nomatterwhatthe
differences in ratings in spite of the higher intake of the effectofpalatabilityonthesubjectiveappetitesensationsis.
palatable meal. Zandstra et al34 manipulated pleasantness The effect of palatability on postprandial appetite sensa-
by varying the salt level and studied appetite over five tions after a test meal (the ‘after-meal’ effect) has been
successive days, and at day 1 they observed higher ratings examinedinthreestudies,twoshowingastimulatingeffect
for ‘faint with hunger’, ‘appetite for a snack’ and ‘appetite of palatability on subsequent appetite sensations and one
for a meal’ for low-palatable bread meals than for medium study showing no difference. Two studies investigating the
and high-palatable bread meals, and ratings for fullness effect of palatability on subsequent energy intake showed a
were lower for the low-palatable bread meals, all ratings stimulating effect of palatability on energy intake and no
reflecting the lower intake of low-palatable bread meal. differenceineffectofpalatability,respectively.
During the 5-day period, this was changed and the
investigators concluded that with repeated exposure, the Sensory-specificsatiety
desiretoeatandintakeofalesspreferredfoodcouldincrease When a food is eaten, it drops in ratings of pleasantness
overtime.34 relative to foods that have not been eaten and this
InagreementwiththefindingsofHilletal8andMonneuse phenomenon is referred to as sensory-specific satiety.24
et al31, Rogers and Blundell36 observed that while there Althoughthetermsatietyisbeingused,appetitesensations
were no differences in test meal intake and feelings of are not necessarily assessed. The standard procedure to test
hunger and desire to eat after consumption of preloads sensory-specific satiety is to have moderately hungry
of different palatability, during the subsequent 3h there subjects taste and rate the pleasantness of the sensory
wasamorerapidrecoveryoffeelingsofhungeranddesireto propertiesofsmallportionsoffoods.Followingtheseinitial
eat after the palatable preload compared both with the less ratings,oneofthefoodsisofferedasamealandsubjectseat
palatablepreloadandnopreload.IncontrastDeGraafetal19 as much as they want. At various times following this test
observed that neither hunger ratings nor food intake was meal, they taste and rate the foods tasted initially. Studies
affected by the pleasantness of the test foods 2h after have shown that changes in the pleasantness of the foods
consumption, and Warwick et al.10 observed that although occur immediately after consuming the test food and that
there was a greater decrease in hunger ratings following the magnitude of these changes does not increase over
thepalatablepreloads,thesensorypropertiesofthepreloads time.21,39,40Thisconfirmsthatthedevelopmentofsensory-
did not influence the total caloric intake at lunch served specificsatietyisrelatedprimarilytothesensorystimulation
5hlater. accompanying ingestion as opposed to the postabsorptive
All the above-mentioned studies examining the effect of effects of consuming these foods. The duration of sensory-
palatability on appetite have been carried out in laboratory specific satiety is unknown; several studies show thatdrops
environments, but we do not really know how much in liking persist more than 1h after consumption.39–41
understanding these studies provide about the influence of Sensory-specific satiety may also have an effect on the
palatability on the normal food intake of humans in their decisiontoterminateameal.42
natural everyday environment. To investigate palatability Whatsensorypropertiesoffoodsareinvolvedinsensory-
influences on eating behaviour, De Castro et al37 made 54 specific satiety? The best case is made for the gustatory
French subjects maintain food intake diaries for four 7-day sensationsofsaltinessandsweetness.Eatingasavouryfood
periods. The subjects recorded their intake along with tosatietyinvariablyleadstoadecreaseinthepleasantnessof
palatabilityratingsofeachindividualitemeatenandalsoa that food and other savoury foods.20,23,43–45 The same
global rating of palatability, which was an overall rating of phenomenon occurs when eating a sweet food to sati-
the entire meal. The results showed that higher levels of ety.23,43–45 Other factors that could influence sensory-
palatabilitywererelatedtolargermealsizes.Toextendthese specific satiety have been examined. To study whether
findings, De Castro et al38 conducted a similar study with flavour intensity is involved in sensory-specific satiety, a
564 North Americans and this time the subjects recorded weak tea and a strong tea were compared and the test
palatability only by global rating of the palatability of the showedanequivalentchangeinlikingforbothteas.46
entiremeal.Mealsthatwereratedhigherinpalatabilitywere Thepleasantnessofthesightofdifferentfoodsconsumed
44%largerthanmealslowinpalatability. tosatietydecreasedmorethanthepleasantnessofthesight
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1157
offoodsthathadnotbeeneaten.20Aftereatingsweetsofone approximatelyaslongasitwouldbeinthemouthduringa
colour, the pleasantness of the taste of sweets eaten of that meal.48
particular colour declined more than that of the noneaten The role of different macronutrients in sensory-specific
sweetsofothercolours,althoughthesweetsdifferedonlyin satiety has been examined and different results have been
appearance.47Changesintheshapeofpasta(whichaffected reached. In several studies, no evidence of macronutrient
boththeappearanceandthemouthfeelofthepasta)ledtoa specificitywasobtained,21,40,49–51butJohnsonandVickers52
specificdecreaseinthepleasantnessoftheshapeeaten.47In observedatrendforhigh-proteinfoodstodecreasemorein
another study, texture-specific satiety was observed but not likingthanlow-proteinfoodsandVandewaterandVickers22
to the same extent as taste-specific satiety.23 Taste and foundthathigher-proteinfoods,wheneateninequalenergy
olfactory sensory-specific satiety can be produced by chew- amounts, produced greater sensory-specific satiety than
ing samples of a food or by smelling the food for lower-proteinfoods(seeTable3).Astudyoftheroleoffibre
Table3 Sensory-specificsatietyFeffectofmacronutrients
Study Subjects Preload Liking/pleasantness Hungerratings Typeoftestmeal Foodintake
ratings
Rollsetal21 10female Chicken,pasta, SSSafterallpreloads HungrierafterTurkish Self-selectionlunch Intakewaslessafter
(1988) Withinsubjects creamcheeseon exceptchocolate.No Delight,creamcheese withfoodshighin chickenandpasta
celery,TurkishDelight, differencesin andchocolatethan certainmacronutrients,
chocolate pleasantnessofeaten afterpastaandchicken 2hdelay
anduneatenfoodsof
similarmacronutrient
content
Rollsetal49 16males Chocolatepudding NodifferencesinSSS Nodifferences Self-selectionmeal, Nodifferences
(1989) 16females andstrawberry– 2hdelay
}In2groups bananagelatine
(info,noinfoa) desserts,highandlow
calorie
Johnsonand 13males None Thehigherprotein Theturkeymeal Cornwithbutter,cola NR
Vickers52 25females mealsdroppedmorein suppressedappetite withliquidglucose
(1992) Between likingthanthelower morethantheother polymer,cheese,
subjects proteinmeals.Buttered meals butteredroll,turkey
rollandcoladropped breast,potatochips,
theleast sweets,canned
peachesornomeal
Johnsonand 6males 150 kcalor300kcal: Theeatenfood Pastaandchickenwere Self-selectionlunch, Lowerintakefollowing
Vickers40 8females pastawithtomato appearedtoinfluence moresatiatingthanthe 90mindelay thechickenthanthe
(1993) Withinsubjects sauce,blueberry- likingofuneatenfoods whippedcream whippedcream
flavouredwhipped withsimilarflavours
cream,chicken butnotsimilar
macronutrients
Vandewater 13/11males None SSSafterallpreloads. Thehigh-proteinmeals 1980kJ,sandwich, NR
andVickers22 27/29females Thehigh-protein suppressedappetite highandlow-protein
(1996) Withinsubjects yoghurtwasliked morethanthelow- 3240kJ,yoghurt,high
betterthanthe proteinversions andlow-protein
low-protein,yoghurt.
HigherSSSwith
high-proteinmeals
comparedto
low-proteinversions
Milleretal50 Noinfob: Adlibregularpotato Noeffectofpotato NR Turkey,strawberry Nodifferencesin
(2000) 24males chipsorpotatochips chiptypeonSSS yoghurt,cookieand intake(weight)of
20females withOlestraina carrot potatochipsonSSS
Infob: 10-daysessioneach testdays,orbetween
27males thetwoinformation
24females conditions
Withinsubjects
SSS:sensory-specificsatiety;NR:notreported.
aTheinformedgroupwasinformedofthenumberofcaloriesperservingandalsowhetherthefoodwashighorlowcalorie.
bNoinfo/info ¼ theparticipantswerenotinformed/wereinformedaboutthefatandenergycontentofthechips.
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1158
in sensory-specific satiety showed no effect of a high-fibre diet could develop a long-term form of sensory-specific
treatment vs a low-fibre treatment when muffins or satiety.54
soup were matched closely on sensory characteristics.53 In a study conducted by Kamphuis et al,51 fat-specific
In two studies, Rolls et al41,49 observed that very low- satiety during dinner with oil high in linoleic acid was
caloriefoodsproducedsensory-specificsatiety.Milleretal50 shown after a 2-week treatment, while there was no
found that eating potato chips with high or low energy reductionintotalenergyintake.
contentdidnotinfluencesensory-specificsatietydifferently.
Summary. So far, flavour, texture and appearance-specific
Whiletheseresultsindicatethatsensory-specificsatietydoes
satietieshavebeenidentified,whereasstudiesontheroleof
not depend on the energy content of the food, De Graaf
et al45 and Johnson and Vickers40 observed that the extent macronutrients and the energy content of the food in
sensory-specific satiety have given equivocal results. Differ-
of sensory-specific satiety for an eaten food increased
ent studies have shown that macronutrients and energy
with increases in the amount of energy and/or food eaten
content play a role in sensory-specific satiety or that
(seeTable4).
macronutrients and energy content are not a factor in
In contrast to the studies described above where sensory-
sensory-specificsatiety.
specificsatietyhasbeeninvestigatedwithinasinglemeal,a
long-term form of sensory-specific satiety has also been
studied. Measures of the palatability of foods that the Varietyandfoodintake
subjects had been eaten for many months was compared The sensory-specific satiety effect may have an important
with measures of the palatability of other foods that the influenceontheamountoffoodeaten.Ifsatietyisspecificto
subjects had not been eating; the subjects were refugees individualsensorypropertiesoffoods,thenintakeshouldbe
from Ethiopia. The results indicated that monotony in the greater when a variety of foods are consumed together or
Table4 Sensory-specificsatietyFeffectofenergy
Study Subjects Preload Liking/pleasantness Hungerratings Typeoftestmeal Foodintake
ratings
Rollsetal41 24females TomatosouporJelly, NodifferencesinSSS Nodifferences Cheeseoncrackers, Nodifferencesinthe
(1988) Withinsubjects highandlowcalorie 1hdelay weightsoffoodeaten.
Highertotalenergy
intakeinthehigh-
calorie,soupcondition
Rollsetal49 16males Chocolatepudding NodifferencesinSSS Nodifferences Self-selectionmeal, Nodifference
(1989) 16females andstrawberry– 2hdelay
In2groups bananagelatine
(info,noinfoa) desserts,highandlow
calorie
DeGraaf Experiment1: Sweetandnonsweet NM Exp.1and2:sweet Exp.2:threebuffet Exp.1:NR
etal45 28females foods250,350,450, preloadssuppressed meals:lunch,snack Exp.2:sweetpreloads
(1993) Withinsubjects 550and650kcal appetiteforsomething anddinnerconsisting suppressedtheenergy
Experiment2: Sweetandnonsweet sweetinadose– ofavarietyofsavoury intakefromsweetfood
29females foods250,500and responsemanner. andsweetfoods inadose–response
Withinsubjects 750kcal Thesamecasefor manner,thesamecase
nonsweetpreload fornonsweetpreload
Johnson 6males 150or300kcal:pasta Highercaloriepreloads High-calorieloadswere Self-selectionlunch, Lowerintakefollowing
and 8females withtomatosauce, droppedmoreinliking moresatiatingthan 90mindelay high-calorieloadsthan
Vickers40 Withinsubjects blueberry-flavoured thanlower lowcalorie lowcalorie
(1993) whippedcream,
chicken
Miller Noinfob: Adlibregularpotato Noeffectofpotato NR Turkey,strawberry Nodifferencesin
etal50 24males chipsorpotatochips chiptypeonSSS yoghurt,cookieand intake(weight)of
(2000) 20females withOlestraina carrot potatochipsonSSS
Withinsubjects 10-daysessioneach testdays,orbetween
Infob:27males thetwoinformation
24females conditions
Withinsubjects
SSS:sensory-specificsatiety.
aTheinformedgroupwasinformedofthenumberofcaloriesperservingandalsowhetherthefoodwashighorlowcalorie.
bNoinfo/info¼theparticipantswerenotinformed/wereinformedaboutthefatandenergycontentofthechips.
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1159
consecutively as opposed to when a single food is served. intakeeitheroverthecompletemealorinanyofthecourses
Several studies have been carried out to investigate the afterthevarietywasintroduced.43Inastudyexaminingthe
impact of meals varying in sensory properties such as impact of varying the colour of food, the presentation of a
flavour,textureandappearance(seeTable5).Varyingflavour variety of colours of sweets did not affect food intake
has shown diverse results. Changes in only the flavour of compared with consumption of the subjects’ favourite
sandwich filling led to a 15% enhancement of food intake colour, although sensory-specific satiety was observed.47 In
when all three flavours were presented successively com- the same study, changes in the shape of pasta led to a 14%
pared with intake of the favourite flavour.47 On the other enhancementoffoodintakewhenthreeshapeswereoffered
hand, offering subjects a variety of flavours of yoghurts of compared with intake of the subjects’ favourite shape.47 In
the same colour and texture in succession did not enhance another study, subjects received hors d’oeuvres of one kind
Table5 Effectsofavarietymealcomparedwithano-varietymeal
Study Subjects Testmeal Hungerratings Foodintake
Plineretal55(1980) 72malesF24normal Novariety(onetypehors Nodifferences Varietyincreasedintake
weightnondieters d’oeuvres)orvariety(3kinds)
Bellisleand 3males 5differentflavouredsandwiches NR Varietyincreasedintake
LeMagnen28,29 7females (novariety),amixedmealinwhich
(1980,1981) Withinsubjects all5flavourswereoffered(variety).
Mealswereequicaloric
Rollsetal43(1981) Experiment1: Experiment1:sandwicheswiththe NR Experiments1and2:
36females samefilling(plain)or4different Varietyincreasedintake
Withinsubjects fillings(variety)
Experiment2: Experiment2:yoghurtwith1
12females flavour(plain)or3different
12males flavours(variety)
Withinsubjects
Rollsetal47(1982) Inallexperiments: Exp.1:threeconditionswith NR Experiment1:
12males sweetsFmostpreferredcolour nodifference
12females (plain),adifferentcolouronein Experiments2and3:
Withinsubjects eachcourseandequalmixofall varietyincreasedintake
fourcolours
Exp.2:tomatosaucewithpastain
differentshapesalsointhree
conditionsasexp.1
Exp.3:creamcheeseflavoured
withsalt,lemonessenceand
saccharin,orcurryspreadon
bread.Uniformappearanceof
spreadusingyellowcolour.Three
conditionsasabove
Rollsetal44(1984) 24males Twoconditions:mealwitha NR Varietyincreasedintake
24females differentfoodineachoffour
Withinsubjects courses(varied)orthesamefood
ineachcourse(plain)
Berryetal56(1985) 65males 1flavouricecream,group;1 NR Varietyincreasedintake,
61females flavour,alone;3flavours,group;3 exceptformeneatingina
Betweensubjects flavours,alone group
Wisniewskietal57 40males 4conditions:pizza,cheeseburger, Nodifferences Varietyincreasedintake
(1992) Betweensubjects pizza–pizza/cheeseburger–
cheeseburger,pizza–
cheeseburger/cheeseburger–pizza
Porrinietal58(1997) Exp.1:13males Exp.1:uniformmealFomelette Exp.1:subjectsstoppedeating NR
whensatiatedbutnotcompletely
full
Exp.2:10males Exp.2:variedmealFomelette Exp.2:subjectsstoppedeating
Withinsubjects andbuffet-styleself-selectionmeal whensatiatedandcompletelyfull
NR:notreported.
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1160
in a single condition and three kinds of hors d’oeuvres Differentstimuli
distinctively different in appearance, flavour and texture in Toexaminewhetherthesensorypropertiesoffatalone,that
the variety condition. The subjects consumed greater is ‘the fatty taste’, play a part in appetite regulation, the
amounts in the variety condition than in the no-variety nonabsorbable fat substitute olestra was used to uncouple
condition, yet they did not show greater decreases in the sensory properties of fat from fat absorption and
reported hunger.55 In two studies, serving sandwiches with metabolism.Creamofbroccolisoupswasdevelopedinthree
differentfillingsinducedanincreaseinmealsizecompared versions:fat-free,fat-free+olestraandhighfat.Subsequent
toservingsandwicheswithjustonekindoffilling.28,29,43In energy intake was not affected by the sensory properties or
anotherstudy,yoghurtsdifferinginflavour,appearanceand cognitivecuesassociatedwiththesoups.Thus,noevidence
texturewereservedinthevarietyconditionandtherewasa indicated that fatty taste per se affects the regulation of
12.6% elevation of intake when the amount eaten in the energyintake.60Incontrast,Ceciletal61observedthatwhen
variety condition was compared with the amount eaten of a high-fat and a high-carbohydrate soup were administered
themostpreferredyoghurt.43Inastudy,wheresubjectswere intragastrically there was no difference in ratings of
given a four-course lunch as either a varied meal with hungerandfullnessandfoodintakebetweenthetwosoups
different food in each course, or as a plain meal with the whereaswhenthetwosoupswereingested,thehigh-fatsoup
samefoodineachcourse,energyintakewaselevatedby60% suppressed hunger, induced fullness and tended to reduce
in the varied meal primarily because of increased food energyintakeinasubsequenttestmealmorethanthehigh-
consumption in the third and fourth courses.44 Another carbohydrate soup. The two soups were equienergetic and
studyshowedthatsubjectsatemoreicecreamwhenoffered therewasnodifferenceinpleasantnessofthetwosoups.
threekindsoficecreamdifferinginflavourandcolourthan Cornell et al62 demonstrated that individual’s food con-
when offered only one kind of their own choosing.56 In a sumption could be affected by a period of pre-exposure to
study,theinfluenceofintroducinganewfoodafterrepeated sensoryaspectsoffoodpriortoeating.Thecombinationofa
presentationsofonefoodonfoodconsumptionwasstudied. brieftasteandsightofeitherpizzaoricecreampresenteda
Pizzaorcheeseburgerwasserveddesignedtovaryinsensory fewminutespriortoeatingstimulatedsubjectstoeatbeyond
properties such as flavour, texture and appearance. The the point of satiety. They also demonstrated that priming
presentationofthenewfoodresultedinsignificantlygreater with a highly preferred food enhanced subsequent con-
caloric consumption than another serving of the same sumption of that food relative to another equally preferred
food (540 vs 180kJ) with no influence on hunger or but not primed food, even in satiated subjects. In a study
fullness.57 In yet another study, the investigators observed investigating factors that promote the initiation of eating,
that when a uniform meal was consumed, subjects stopped Marcelino et al63 observed that appetite for pizza increased
eating when they felt satiated but not completely full, with hunger and after being visually exposed to the pizza,
whereaswhenservedavariedmealtheydidnotstopeating both factors acting independently. Desire to eat pizza
untiltheyfeltsatiatedandcompletelyfull.Theinvestigators depended on the visual quality of the pizza and also on
concluded that in this condition, sensory characteristics of hunger.Incontrast,Fedoroffetal64foundthatpre-exposure
foods played an important role in controlling food intake tothesmellofbakingpizzadidnotinfluencehungerratings
and made the uniform meal more satiating thanthe varied andfoodconsumptioncomparedtonoexposureinsubjects
one.58 thathadrefrainedfromeatingfor2hpriortotheexperiment.
Examiningtheeffectofincreasingthevarietyofsensorily Inastudytestingtheeffectoftemperature,vegetablejuice
distinct but nutritionally identical foods on appetite, food was used as a preload. Male subjects experienced a greater
intakeandbodyweight,over7days,foodandenergyintake reduction in their desire to eat following the cold preloads
increasedasthevarietyoffoodsincreased.Intakeamounted thanaftereitherthehotpreloadsornopreloadcondition.65
to1.57,1.76and1.97(cid:2) restingmetabolicrateonthelow-, Themenalsohadasignificantlylowerfoodintakefollowing
medium-andhigh-varietytreatments,respectively,although the cold preload compared to the no-preload condition.
therewasnodieteffectonsubjectivelyratedhunger.59 Temperatureofthepreloadsdidnotaffecteitherdesiretoeat
orfoodintakeinfemalesubjects.
Summary. Increasing the variety of sensorily distinct
foods that are almost identical in composition can increase Sweetness and appetite. Besides the contribution of sweet-
food and energy intake and in the short to medium term nesstopalatability,therehasbeenaninterestintheeffectof
alter energy balance. In the majority of the reviewed sweettasteperseonappetite;however,theroleofsweettaste
studies, subjective appetite sensations were either not in appetite regulation is controversial. One way to investi-
reported or not monitored. Three studies have assessed gate the influence of sweet taste has been to uncouple
subjectiveappetitesensationsandinallthreestudies,variety sweetnessandenergysoastoevaluatetherelativecontribu-
increased intake, while there was no diet effect on hunger tionofeachfactortothecontrolofappetite.Bysubstitution
ratings.55,57,58 ofanartificialsweetener forsugarinabeverageorfood,its
Long-termeffectsofalteringthevarietyoffoodsonenergy energyvalueisreducedwhileitssweetnessismaintainedand
balanceareyettobedetermined. thentheeffectofthesweetenedbeverageorfoodonappetite
InternationalJournalofObesity
Sensoryscienceandappetite
LBSørensenetal
1161
canbecomparedwiththeeffectofunsweetenedbeverageor that after administration of a benzodiazepine-receptor
food(seeTable6).BlundellandHill66reportedanincreasein agonist, the number of eating occasions increased without
motivationtoeatanddecreasedratingsoffullnessfollowing alteringmealsize.Palatability,however,wasnotassessed.
the ingestion of an aspartame solution compared to inges- Studies investigating the hypothesis that palatability and
tion of water, which is consistent with the hypothesis that sweet taste have an effect on cephalic phase insulin release
sweet taste can stimulate hunger. Support for this view has anddiet-inducedthermogenesishaveshowndifferentresults.
come from other short-term studies showing a stimulating In some studies, the cephalic phase plasma insulin levels
effect of sweet taste on hunger ratings when the sweetener measured were lower when subjects were consuming an
wasprovidedintheformofaqueoussolutions,chewinggum unpalatablemealthanwhenconsumingpalatablefood,92–96
orbeverages9,67–69andbothhungerratingsandfoodintake whileinastudyusingshamfeeding,nosignificantdifference
when using fruit yoghurt as avehicle.70In contrast, several inthemagnitudeofcephalicphaseinsulinreleasewasfound
studies showed no effect of sweet taste on either hunger between palatable and unpalatable food.97 Studies on the
ratings or food intake when the sweetener was provided in effect of sweet taste per se on cephalic phase insulin release
theformofgelatine,cornflakesorfromageblanc71–74oras failedtoshowinsulinsecretionafteroralstimulationofsweet
aspartame-orsaccharin-sweeteneddrinks.75–79 taste.98–100Studiesonthe effect of diet-induced thermogen-
Other studies have contributed to elucidate the role of esisofpalatableandunpalatablefoodshaveshowndifferent
sweettasteonappetite.BralaandHagan80showedthatwhen results. In some studies, diet-induced thermogenesis was
subjects’ ability to taste sweetness was attenuated by significantlyelevatedafterconsumptionofapalatablemeal
gymnemic acid, they ate less food after a sweet-tasting in comparison with an unpalatable meal,92,99,101,102 while
preloadthandidsubjectswhosetasteperceptionwasintact. othersshowednodifferenceinthermogenesis.96,103Investi-
Thesefindingssuggestthatasweetoralstimulationinitiates gation of the effect of sweet taste on diet-induced thermo-
reflexesthatincreaseappetite.Rollsetal81havestudiedthe genesisshowedthatsweettastehasnoeffect.104
effect of eating a sweet food either as a first course or as a It has been speculated that sensory-specific satiety is
thirdcourseonfoodintakeduringathree-courselunch.The caused by sensory adaptation, that is, the intensity of a
secondcourseremainedthesameinthetwoconditionsand food’s flavour diminishes when there is a sensory-specific
thetwocoursesthatweretobeinterchangedwereofsimilar decreaseinthepleasantnessofthefood.Thisdoesnotseem
energydensitiesandratedpalatability,onecourseconsisted to be the case, as studies have shown that changes in the
of a chocolate bar and the other was cheese on crackers. pleasantnessofafooddonotappeartoreflectchangesinthe
There was no indication that eating candy as a first course perceivedintensityofthefoods.20,48Insteadofbeingrelated
affectedappetiteandfoodintakedifferentlyfromeatingthe to changes in the sensory processing of responses to foods,
same amount of a cheese on crackers. In two experiments, sensory-specific satiety is more likely related to brain areas
De Graaf et al45 studied the hypothesis that if sweet taste controllingmotivationandtherewardvalueoffoods.105–108
sensation increases appetite then sweet carbohydrates must
belesssatiatingthannonsweetcarbohydrates.Theresultsof
thefirstexperimentwereconsistentwiththehypothesis,but Suggestions for future work
thesecondexperimentdidnotshowanydifferencebetween Results reviewed in this paper suggest that increased
sweetandnonsweetcarbohydrates. palatability increases food intake in the short term. The
contradictory results found when assessing subjective appe-
titesensationsmaybearesultofdifferentstudydesigns,but
Summary. The role of sweet taste per se in appetite
itisworthstudyingwhetherthesediverseresultsinfactare
regulationisnotfullyclarified.Althoughsomestudiesshow
due to different methodologies or whether there is a
a stimulating effect, most studies show no effect of sweet
discrepancy between subjective appetite sensations and
tasteonappetiteandfoodintake.
actualfoodintakewheneatingapalatablemeal.
Sensory-specific satiety may influence the amount of food
Mechanisms eaten,asincreasingthevarietyoffoodsincreasesfoodintake.
What physiological mechanisms are behind the possible Thus,itmaybethatvarietyenhancesthepalatabilityofameal
effect of palatability on appetite and food intake? Study and people therefore eat more. Here, as in the palatability
results imply that endogenous opioid peptides are involved studies, some interesting results appeared when subjective
inprocessesunderlyingthedeterminationofpalatability.In appetite sensations were assessed: in three studies, variety
several studies using opioid antagonists, intake is reduced, increasedfoodintakewithoutinfluencingsubjectivehunger
butmainlyforthefoodsthatareratedashighlypalatable.82–86 ratings.Thissupportstheideathattheamountofpalatable
Hetherington et al,87 however, failed to find an effect of food eaten may not be recognised and therefore subjective
opioidantagonismonpleasantnessratingsandintake.Accord- appetiteratingsdonotreflecttheamountoffoodeaten.
ing to several animal studies, benzodiazepine may enhance Long-termstudiesoftheconsequencesoffoodpalatability
food palatability and thereby increase meal size.88–90 The andvarietyonfoodintakeandenergybalancecarriedoutin
onlystudyonhumans,performedbyHaneyetal,91showed everydaynaturalenvironmentsareneeded.
InternationalJournalofObesity