you can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

8
Appetite (2000) 35, 153–160 doi:10.1006/appe.2000.0348, available online at http://www.idealibrary.com on 1 Original Article You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly M.L. Pelchat Monell Chemical Senses Center, Philadelphia (Received15 August1996, revision 8 June 2000, accepted in revised form12 June 2000, published electronically16 August 2000) It can be difficult to document effects of age-related olfactory loss on eating behavior. However, we have reported that elders, especially those with poor olfaction, were more willing to accept novel foods than were younger adults. It was also found that elderly subjects were more willing to accept foods with unpleasant odors than were young subjects. Because there is often confounding between a food’s odor pleasantness and its familiarity, the purpose of this study was to separate the effects of familiarity and odor pleasantness on food acceptance by the elderly. There was no evidence for effects of age or olfactory sensitivity on food neophobia. However, elderly subjects with poor olfaction showed less reluctance to try the unpleasantly smelling foods than did other subjects. These results suggest that increased willingness to try novel foods among elderly subjects with poor olfaction in earlier work was due to decreased rejection of foods with unpleasant odors and not due to decreased food neophobia per se. # 2000 Academic Press Introduction Food neophobia is a reluctance to try or a dislike for the flavor of novel foods. A large number of studies show that people’s liking or acceptance is greater for familiar than for novel foods (Birch, 1979a,b; Birch & Marlin, 1982; Birch et al., 1987; Meiselman & Waterman, 1978; Pliner, 1982; Torrance, 1958). (However, exces- sive exposure can be associated with decreased liking, Schutz & Pilgrim, 1958; Stang, 1975.) While hesitation to sample new foods is clearly valuable in some settings, such as when unfamiliar mushrooms are encountered in the wild, those settings have little nutritional relevance for most of us. In the nutritionally relevant settings in which we do most of our eating, culture has already defined ‘‘foods’’ as being safe to eat. So, most of the time, food neophobia in humans is unnecessary and can even be maladaptive because it can hamper attempts to bring about diet change and limits dietary variety. Athough there is a general tendency for people to show food neophobia, there are group and individual differences in its intensity (Pliner & Hobden, 1992). One group factor for which there is a fair amount of evidence is age. In contrast to the stereotype that ‘‘you can’t teach an old dog new tricks’’, there have been many reports (most focus on children) of decreasing levels of food neophobia with increasing maturity. The mechanism for the decrease, however, is not known. In a multi-dimen- sional scaling study, it was found that for 3-year-olds, familiarity may be the most important determinant of food preference but that for 4-year-olds, sweetness may surpass familiarity in importance (Birch, 1979b). Birch et al. (1987) reported that 11 out of 16 2-year-olds refused to try a novel food while no five-year-olds refused. Pelchat & Pliner (1995) found that primary school-aged children were more willing to try a novel food than were preschoolers and that high school stu- dents were more willing to try new foods than were younger adolescents. Otis (1984) reported that stated willingness to try unfamiliar foods increased with age in subjects ranging from 17–50 years of age. She did not find a relationship between the subjects’ ages and familiarity with the rated foods which suggests that the age effect was really due to differences in food neopho- bia rather than to differences in experience with the foods. Pliner and Hobden (1992) found decreases with Supported by NIH AG09892. Address correspondence to: Marcia Levin Pelchat, Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA 19104-3308, U.S.A. E-mail: [email protected] 0195–6663/00/050153+08 $35.00/0 # 2000 Academic Press

Upload: ml-pelchat

Post on 16-Oct-2016

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

Appetite (2000) 35, 153±160doi:10.1006/appe.2000.0348, available online at http://www.idealibrary.com on

1

Original Article

You can teach an olddog new tricks: olfactionandresponses to novel foods by the elderlyM.L. PelchatMonell Chemical Senses Center, Philadelphia

(Received15August1996, revision 8 June 2000, acceptedinrevised form12 June 2000, publishedelectronically16 August 2000)

It can be difficult to document effects of age-related olfactory loss on eating behavior. However, we have reportedthat elders, especially those with poor olfaction, were more willing to accept novel foods than were younger adults. Itwas also found that elderly subjects were more willing to accept foods with unpleasant odors than were youngsubjects. Because there is often confounding between a food's odor pleasantness and its familiarity, the purpose ofthis study was to separate the effects of familiarity and odor pleasantness on food acceptance by the elderly. Therewas no evidence for effects of age or olfactory sensitivity on food neophobia. However, elderly subjects with poorolfaction showed less reluctance to try the unpleasantly smelling foods than did other subjects. These results suggestthat increased willingness to try novel foods among elderly subjects with poor olfaction in earlier work was due todecreased rejection of foods with unpleasant odors and not due to decreased food neophobia per se.

# 2000 Academic Press

Introduction

Food neophobia is a reluctance to try or a dislike forthe flavor of novel foods. A large number of studiesshow that people's liking or acceptance is greater forfamiliar than for novel foods (Birch, 1979a,b; Birch &Marlin,1982;Birchetal., 1987;Meiselman&Waterman,1978; Pliner, 1982; Torrance, 1958). (However, exces-sive exposure can be associated with decreased liking,Schutz & Pilgrim, 1958; Stang, 1975.) While hesitationto sample new foods is clearly valuable in some settings,such as when unfamiliar mushrooms are encountered inthe wild, those settings have little nutritional relevancefor most of us. In the nutritionally relevant settings inwhich we do most of our eating, culture has alreadydefined `̀ foods'' as being safe to eat. So, most of thetime, food neophobia in humans is unnecessary and caneven be maladaptive because it can hamper attempts tobring about diet change and limits dietary variety.

Athough there is a general tendency for people toshow food neophobia, there are group and individual

differences in its intensity (Pliner &Hobden, 1992). Onegroup factor for which there is a fair amount of evidenceis age. In contrast to the stereotype that `̀ you can't teachan old dog new tricks'', there have been many reports(most focus on children) of decreasing levels of foodneophobiawith increasingmaturity. Themechanism forthe decrease, however, is not known. In a multi-dimen-sional scaling study, it was found that for 3-year-olds,familiarity may be the most important determinant offood preference but that for 4-year-olds, sweetness maysurpass familiarity in importance (Birch, 1979b). Birchet al. (1987) reported that 11 out of 16 2-year-oldsrefused to try a novel food while no five-year-oldsrefused. Pelchat & Pliner (1995) found that primaryschool-aged children were more willing to try a novelfood than were preschoolers and that high school stu-dents were more willing to try new foods than wereyounger adolescents. Otis (1984) reported that statedwillingness to try unfamiliar foods increased with age insubjects ranging from 17±50 years of age. She did notfind a relationship between the subjects' ages andfamiliarity with the rated foods which suggests that theage effect was really due to differences in food neopho-bia rather than to differences in experience with thefoods. Pliner and Hobden (1992) found decreases with

Supported by NIH AG09892.Address correspondence to: Marcia Levin Pelchat, Monell

Chemical Senses Center, 3500 Market Street, Philadelphia,PA 19104-3308, U.S.A. E-mail: [email protected]

0195±6663/00/050153+08 $35.00/0 # 2000 Academic Press

Page 2: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

age among adults in a trait measure of food neophobia.Although each of these studies focuses on a smallportion of the lifespan, they are consistent in showing arelationship between increased age and decreased rejec-tion of novel foods. It is unfortunate that none of thesestudies includes elderly subjects. It is important to knowabout responses to novel foods in the elderly because theelderly are such a rapidly growing portion of the popu-lation (Havlik, 1992) and are widely believed to be atincreased nutritional risk (Betts, 1988; Kohrs, 1982).Given the literature reviewed above, it seems reason-

able to think that the trend toward decreased foodneophobia would continue into old age. One possiblemechanism is age-related sensory change. Specifically,ortho-nasal olfaction is a distance sense that permitsexploration of novel foods prior to ingestion. Further,olfactory loss is common among the elderly. Comparedto young adults, elderly individuals are less able to detectweak odors (Cain & Stevens, 1989; Chalke &Dewhurst,1957; Cowart, 1989), rate detectable odors as beingless intense (Cain & Stevens, 1989; Cowart, 1989; Enns& Hornung, 1988) and show deficits in identificationof blended foods and food odors (Murphy, 1985;Schiffman, 1977; Schiffman & Pasternak, 1979).

There is some evidence that response to novel foodsis influenced by olfactory sensitivity (Pelchat & Stoess,1991). In a study in which young adult and elderlysubjects smelled and sampled a group of nine foodswhich were chosen to vary in odor pleasantness and infamiliarity, elderly subjects were more likely than youngadult subjects to choose novel foods [rice crackerswrapped in seaweed, a goat's milk cheese coated withash (montrachet), and lychees] to take home as leftovers.Also, elderly subjects with poor olfaction were sig-nificantly more likely to take home novel leftovers thanwere elderly subjects with good olfaction. The latterfinding is consistent with the idea that reduced olfactorysensitivity, not just simple aging, is associated withreduced rejection of novel foods.

However, the neophobia difference may be part of amore general decrease in finickiness that affects bothfamiliar and novel foods. In that same study, ratings ofwillingness to try the three most pleasant smelling andthe three most unpleasant smelling foods were com-pared. For young subjects, willingness to try the moreunpleasant smelling foods was significantly less thanwillingness to try themore pleasant smelling foods,whilefor elderly subjects, odor pleasantness didn't matter. So,the elderly subjects seemed to be less particular than theyoung subjects. One possible mechanism for decreasedfood selectivity by elderly hyposmic subjects is that theycould not detect unpleasant aspects of food odors.Schiffman and Covey (1984) made a similar suggestionbased on subjects with self-reported olfactory loss.

The odors of novel foods also tend to be unpleasant.Thus individuals with poor olfactory sensitivity could bemore willing to try such foods because their unpleasantodors would be less noticeable. One reason that theodors of novel foods tend to be unpleasant is that foodswith odors that are considered to be unpleasant outsideof an eating context (e.g. strong fish or cheese) could beserved less frequently than less offensive foods. In otherwords, bad-smelling foods are served less frequently.Another mechanism for a relationship between noveltyand odor unpleasantness points the causal arrow inthe other direction: all other things being equal, novelstimuli are liked less than are familiar stimuli (e.g.Zajonc, 1968). [Although there is a literature demon-strating a relationship between odor-exposure andhedonic tone, these studies all used relatively familiarodors and are therefore of limited applicability in thiscontext (Beebee-Center, 1932; Cain & Johnson, 1978;Murphy, 1982, 1983)]. Therefore, in the current studies,we attempted to separate the effects of familiarity witha food and of the pleasantness of its odor on willingnessto sample the food by young and elderly adults.

Experiment I

MethodMethod

Subjects

A total of 51 healthy, community-dwelling adultsparticipated in the study. They were divided into threegroups based on age and on olfactory threshold testing(see below): Normosmic young adults (N� 16; 18±25years old), elderly adults with good olfaction (N� 15)and elderly adults with poor olfaction (N� 20). Theelderly subjects ranged from 62±85 years of age. Therewere more females than males in the study (31 female,20 male) however, the proportion of females wasconstant across all three groups. Subjects were paid fortheir participation.

Materials

olfactory threshold stimuli

A pyridine series and a phenylethyl alcohol (PEA)series were prepared for olfactory threshold testing.The pyridine series contained 19 binary dilution stepsstarting from 0�097% pyridine and the PEA seriescontained 19 half-log dilution steps starting from100% PEA. The blank and diluent in the pyridineseries were mineral oil and in the PEA series they wereglycerol. Twenty-millilitre samples were presented tosubjects in polypropylene squeeze bottles.

154 M.L. Pelchat

Page 3: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

food stimuli

Four food stimuli were developed through extensivepre-testing. All four were sauces made by combiningflavorings with an unflavored instant pudding (pro-vided by Kraft-General Foods). The novel food with apleasant odor (NP) was a guanabana-passion fruitsauce. The guanabana is a tropical fruit (Floribbean).The passion fruit flavor was an extract (Bikford). Thenovel food with an unpleasant odor (NU) was awakame-shrimp paste sauce. Wakame (Wel-pak) is atype of seaweed. The familiar food with a pleasantodor (FP) was a strawberry sauce made with straw-berry preserves (Kraft), and the familiar food with anunpleasant odor (FU), was a cheese sauce made withparmesan cheese. A total of 10 young adult pilotsubjects (who did not participate in the main experi-ment) rated odor intensity and odor pleasantness ofthe sauces without being told the names of the flavor-ings. They were then given the name of each sauce andrated their familiarity with the flavoring. All ratingsused 7-point scales. There were no significant differ-ences across foods in odor intensity F(3, 27)� 1�46,NS.The two foods designated as being pleasant were ratedas being significantly more pleasant-smelling (Mean�5�05) than the two foods designated as being unplea-sant (Mean� 3�55; F(1, 9)� 21�32, p� 0�001). It isimportant to note that 5�05 is above neutral on theseven point hedonic scale and that 3�55 is belowneutral. The familiar and unfamiliar foods did not dif-fer in odor pleasantness. The foods designated as famil-iar were rated as significantly more familiar (Mean�6�6) than were the foods designated as unfamiliar(Mean� 1�05; F(1, 9)� 22.63, p< 0�001). There was nosignificant difference in familiarity between foodsdesignated as pleasant-smelling and unpleasant smell-ing. Therefore, the designations NP, NU, FP, FU arebased on the ratings of the 10 young adult subjects.These classifications, along with age/olfactory group,served as the independent variables in the study.

Procedure

Olfactory threshold determinations were made by aforced-choice staircase technique (see Cowart, 1989).Thresholds are given in dilution steps, so a highernumber indicates higher sensitivity. The PEA andpyridine thresholds were significantly correlated(r� 0�49, p< 0�05) and were added together to give anolfactory function score. The elderly subjects weredivided into a good olfaction group (OG) and a poorolfaction group (OP) based on these scores such thatthe subjects in the OG group were equivalent to theyoung, normosmic (Y) subjects in olfactory sensitivity

(i.e. had the same group mean) and the subjects in theOP group were significantly less sensitive (F(2, 48)�26�8, p< 0�001, Tukey tests show that OP is signifi-cantly different from OG and Y; see Table 1).

Following the olfactory threshold determinations,each subject was presented with the four food stimulione at a time and in random order. The subject was toldthe name of each stimulus (i.e. `̀ This is a strawberrysauce'', `̀ This is a parmesan cheese sauce'', `̀ This is aguanabana sauce'', or `̀ This is a wakame sauce''). Foodswere presented to the subjects in uncovered transparentcontainers. Therefore, the foods could be seen andsmelled (subjects were asked to smell them beforegiving ratings). Subjects first rated their familiaritywith each food on a 7-point scale ranging from1 : extremely unfamiliar to 7 : extremely familiar. Theynext rated willingness to taste each food on a 7-pointscale ranging from 1 : extremely unwilling to7 : extremely willing. They were led to believe that, ifwilling, they would have to taste the food at the end ofthe study, but, in fact, the foods were never tasted. Thesubjects also rated the pleasantness of each food's odorat the end of the session.

The main dependent measure was the willingness-to-try rating. The data were analysed in a 2 (pleasant±unpleasant odor) by 2 (familiar or unfamiliar foodname) by 3 (age/olfaction group) analysis of variancewith pleasantness and familiarity as repeated measures.Again, the independent variables, pleasantness andfamiliarity, were based on the prior ratings of the inde-pendent panel of young subjects who did not participateas subjects in this study. All post hoc tests are Tukeytests.

Results and discussionResults and discussion

Overall, there was no difference across the age/olfactory groups in willingness to taste the sauces(see Fig. 1; see also Table 2 for ANOVA results). Therewas a significant effect of food familiarity on will-ingness to taste. Subjects were more willing to try the

Table1. Olfactory function scores, Experiment 1

Old poorolfaction

Old goodolfaction

Young

Olfactory function scoremean�SD

11�3� 0�9 20�2� 1�1 20�1� 1�1

Olfactory function scorerange

1�75±15�5 15�8±29�3 10�0±28�0

The olfactory function score is the sum of the threshold forphenylethyl alcohol and the threshold for pyridine expressed indilution steps. A higher score indicates greater sensitivity.

Olfaction and responses to novel foods by the elderly 155

Page 4: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

familiar foods than the novel foods. In other words,there was a significant neophobia effect. However,there was not a significant interaction between age/olfactory group and familiarity. What the latter meansis that there was no age difference in the response tonovelty per se: the elderly were no more nor lessneophobic than the young adults.

There was also a significant main effect of pleasant-ness: subjects were more willing to try the pleasantsmelling than the unpleasant smelling foods. Qualifyingthis effect was a significant interaction between pleas-antness and age/olfactory group. Tukey tests showedthat ratings of the foods classified as unpleasant were nodifferent for young adults and for elderly adults withgood olfaction. However, members of the OP groupwere significantly more willing to try the unpleasant-smelling foods than were the young adults (see Fig. 1).This is consistent with a general decrease in foodfinickiness in individuals with poor olfaction. However,

we cannot rule out some sort of cohort effect based onthese results because there was not a significant differ-ence between the older subjects with good olfactionand the older subjects with poor olfaction. Therefore,Experiment 2 was proposed to more definitively estab-lish whether olfaction played a role in the effects that weobserved.

As indicated by the group� pleasantness� famil-iarity interaction, the difference across groups wasmost apparent for the NU food. Post hoc testsdemonstrated that whereas the three groups of subjectshad similar willingness to try the FU food, the OP sub-jects were significantly more willing than the youngadults to sample the NU food. It may generally be thecase that differences based onolfactory function are seenmore readilywith novel thanwith familiar foods becausewith familiar foods, people may fill in from memoryformissing sensory information. It also seems that whensubjects are told the identity of the parmesan cheese(FU), they rate its odor pleasantness higher than dosubjects who are not told what they are smelling (The 10subjects in the pilot study who were not given the namesof the foods that they were smelling, rated the parmesancheese odor as below neutral in pleasantness and thesubjects in the study who were told that it was parmesanrated the cheese as above neutral). Therefore, the lack ofdifferences across groups in willingness to try the FUfood is understandable: the elderly subjects with poorolfaction would have given high willingness ratingswhether they knew what they were smelling or not andthe others gave high ratings because they knew that theywere smelling a familiar cheese.

The pleasantness ratings given by the subjectsclosely paralleled their willingness ratings (Fig. 2).

1

7

Age/olfaction group

Mea

n w

illi

ngn

ess

Old/Poor Old/Good

6

5

4

3

2

Young

Figure 1. Average willingness to try the sauces by subjects in Experiment 1 who received the stimuli in uncovered,transparent containers. 1� extremely unwilling, 7� extremely willing. Old/Poor, elderly with poor olfaction; Old/Good,elderly with good olfaction. FP, familiar-pleasant; NP, novel-pleasant; FU, familiar-unpleasant; NU, novel-unpleasant. fp,&;np, ; fu, ; nu, &.

Table 2. ANOVA table for willingness ratings,Experiment 1

F df p

Age/olfactory group 1�1 2, 48 NSPleasantness 36�9 1, 48 < 0�001Familiarity 25�7 1, 48 < 0�001Age/olfactory groupby pleasantness

5�9 2, 48 < 0�01

Age/olfactory groupby familiarity

0�14 2, 48 NS

Pleasantness by familiarity 8�5 1, 48 < 0�01Age/olfactory groupby pleasantnessby familiarity

3�7 2, 48 < 0�02

156 M.L. Pelchat

Page 5: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

Here, pleasantness is a dependent variable based onratings given by the subjects when they sniffed the novelfoods. In fact, the figures showing subjects' ratings ofwillingness to try and showing subjects' ratings ofpleasantness are almost superimposable. The correla-tion between the willingness rating and the pleasant-ness rating was statistically significant at the 0�001level, or better for all four foods (r: strawberry� 0�55;guanabana� 0�44; parmesan� 0�64; wakame� 0�64).The importance of odor pleasantness as a mediatingvariable is supported by the close parallel in this studybetween subjects' willingness and pleasantness ratings.

Experiment 2

This study is identical to Experiment 1 except thathere, the sauces could not be smelled. Therefore, thepurpose of this subtraction study is to further elucidatethe role of olfaction in responses to novel foods. Thehypothesis is that, in the absence of olfactory cues,there will no longer be any differences across thegroups of subjects in willingness to try the novel,unpleasant-smelling food (NU).

MethodMethod

Subjects

A total of 45 healthy, community-dwelling adultsparticipated in the study. None of these subjects hadparticipated in the first experiment. As in Experiment1, they were divided into three groups based on ageand on olfactory threshold testing: Young adults(N� 15; 19±34 years old), elderly adults with good

olfaction (N� 15) and elderly adults with poorolfaction (N� 15). The elderly subjects ranged from64±85 years of age. There were more females thanmales in the study (28 females, 17 males) however, theproportion of females was constant across all threegroups. Subjects were paid for their participation.

Materials

The food stimuli and olfactory testing stimuli wereidentical to those used in Experiment 1.

Procedure

Olfactory testing was conducted as described inExperiment 1. The PEA and pyridine thresholds were,again, significantly correlated (r� 0�47, p< 0�05) andwere added together to give an olfactory functionscore. As before, the olfactory function scores of thesubjects in the OG group were equivalent to the young(Y) subjects (i.e. had the same group mean) and thesubjects in the OP group were significantly lesssensitive (F(2, 42)� 19�1, p< 0�001, Tukey tests showthat OP is significantly different from OG and Y; seeTable 3).

1

7

Age/olfaction group

Mea

n p

leas

antn

ess

Old/Poor Old/Good

6

5

4

3

2

Young

Figure 2. Average pleasantness ratings of odors of sauces by subjects who received the stimuli in uncovered, transparentcontainers. 1� extremely unpleasant, 7� extremely pleasant. Old/Poor, elderly with poor olfaction; Old/Good, elderly withgood olfaction. FP, familiar-pleasant; NP, novel-pleasant; FU, familiar-unpleasant; NU, novel-unpleasant. fp, &; np, ;fu, ; nu, &.

Table 3. Olfactory function scores, Experiment 2

Old poorolfaction

Old goodolfaction

Young

Olfactory function scoremean�SD

13�5� 0�7 18�8� 0�7 20�0� 0�9

Olfactory function scorerange

6�25±15�5 15�75±26�5 14�2±25�5

The olfactory function score is the sum of the threshold forphenylethyl alcohol and the threshold for pyridine expressed indilution steps. A higher score indicates greater sensitivity.

Olfaction and responses to novel foods by the elderly 157

Page 6: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

As in Experiment 1, each subject was presented withthe four food stimuli one at a time, in random order andthe subject was told the name of each stimulus. In con-trast to Experiment 1, foods were presented to the sub-jects in covered transparent containers so that the foodscould be seen but not smelled. Subjects then rated theirfamiliarity with each food and their willingness to tasteeach food. As before, they were led to believe that, ifwilling, they would have to taste the food at the end ofthe study, but, in fact, the foods were never tasted.

The main dependent measure was the willingness-to-try rating. The data were analysed in a two (pleasant±unpleasant odor) by two (familiar or unfamiliar foodname)by three (age/olfactiongroup)analysisofvariancewith pleasantness and familiarity as repeated measures.All post hoc tests areTukey tests. Again, the independentvariables, pleasantness and familiarity, were based onthe prior ratings of the independent panel of youngsubjects who did not participate as subjects in this study.

Results and discussionResults and discussion

As in Experiment 1, there was no difference across theage/olfactory groups in willingness to taste the sauces(see Fig. 3; see also Table 4 for ANOVA results). Therewas, again, a significant main effect of food familiarityon willingness to taste (i.e. a significant neophobiaeffect; see Fig. 3). Again, there was not a significantinteraction between age/olfactory group and familiar-ity. Therefore, it can be concluded that the elderly wereno more nor less neophobic than the young adults.

There was a significant main effect of pleasantness;subjects were more willing to try the pleasant smellingthan the unpleasant smelling foods. This might seem

surprising given that subjects could not smell the foodsin this study. However, qualifying this effect was a sig-nificant interaction between pleasantness and novelty.Tukey tests showed that for the familiar foods, will-ingness was related to odor pleasantness, but that for thenovel foods, odor pleasantness did not matter. So,subjectswere presumably relying onmemory to fill in forthe missing sensory information for the familiar foods.Post hoc tests also indicated that willingness to try theNP foodwas significantly less than willingness to try theFP food, but in Experiment 1, willingness to try the twopleasant-smelling foods was equal (see Figs 1 and 3).These results are to be expected given that subjects in thecovered condition had no information other than thename for the NP food and that in the uncovered con-dition, they could respond to its pleasant odor.

Most critically, the interaction between age/olfactorygroup and odor pleasantness was not significant here, incontrast to Experiment 1. This shows that an odor cue

1

7

Age/olfaction group

Mea

n w

illi

ngn

ess

Old/Poor Old/Good

6

5

4

3

2

Young

Figure 3. Average willingness to try the sauces by subjects in Experiment 2 who received the stimuli in covered, transparentcontainers. 1� extremely unwilling, 7� extremely willing. Old/Poor, elderly with poor olfaction; Old/Good, elderly withgood olfaction. FP, familiar-pleasant; NP, novel-pleasant; FU, familiar-unpleasant; NU, novel-unpleasant. fp, &; np, ;fu, ; nu, &.

Table 4. ANOVA table for willingness ratings,Experiment 2

F df p

Age/olfactory group 1�8 2, 42 NSPleasantness 12�9 1, 42 < 0�001Familiarity 47�9 1, 42 < 0�001Age/olfactory groupby pleasantness

0�6 2, 42 NS

Age/olfactory groupby familiarity

0�1 2, 42 NS

Pleasantness byfamiliarity

5�4 1, 42 < 0�03

Age/olfactory groupby pleasantnessby familiarity

3�0 2, 42 NS

158 M.L. Pelchat

Page 7: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

was necessary for this effect in Experiment 1 and pro-vides further evidence for the hypothesis that the inter-action observed in Experiment 1 was due to differencesin olfactory sensitivity.

General discussion

These studies provide no evidence for age-differencesamong adults in their reactions to the idea of tryingnovel foods. However, they do provide evidence for amechanism by which elderly individuals with poorolfactory sensitivity could be more willing to try novelfoods. It is likely that this mechanism is not specific tounfamiliar foods, but rather is a failure to detectunpleasant aspects of food odors in general. The mainpieces of evidence for a role for olfaction are: (1) inExperiment 1, elderly subjects with poor olfactionwere more willing to try unpleasant-smelling foodsthan were young subjects; (2) there were close parallelsbetween subjects' willingness-to-try ratings and sub-jects' odor-pleasantness ratings in Experiment 1;(3) there were no differences in willingness acrossolfactory sensitivity groups to try foods with unplea-sant odors in Experiment 2 when olfactory cues wereabsent. This shows that olfactory cues were necessaryfor the differences observed in Experiment 1.

It should be noted that there are reports of a linkbetween high olfactory sensitivity and high levels offood acceptance (Frank & van der Klaauw, 1994;Raudenbush et al., 1995). Although this may seem tobe in direct contrast to our hypothesis, we may belooking at two sides of the same coin. Indeed, closeexamination of Frank and van der Klaauw's (1994)data indicate that while subjects who gave low odor-intensity ratings also gave lower pleasantness ratings topleasant-smelling foods, they gave higher pleasantnessratings to foods with unpleasant odors (e.g. Brusselssprouts). This is consistent with Wysocki and Pelchat's(1993) analysis of some of the questions from theNational Geographic Smell survey: ratings of odor-pleasantness and willingness-to-eat for samples thatsmelled like bananaor cloveswent downwith decreasingodor intensity, whereas willingness to eat samples withfood-inappropriate odors went up as odor intensitydecreased. There is a trend in the present study forsubjectswith the highest olfactory sensitivity to show thehighest willingness to try theNP food. Along these lines,Schiffman and Covey (1984) reported that individualswith self-reported poor olfaction liked pleasant smellingfoods less than others. So, it appears that pleasant odorsbecome less pleasant with decreasing intensity whereasunpleasant odors become less unpleasant with decreas-ing intensity. In other words, odors seem to movetoward hedonic neutrality as they become less intense.

Therefore, we conclude that the increased willingnessto try novel foods seen previously in elderly hyposmicsubjects (Pelchat & Stoess, 1991) was due to an inabilityto detect unpleasant aspects of food odors and was notdue to decreased food neophobia per se. Caution mustbe exercised in making predictions of real-world eatingbehavior based on these data. These laboratory-basedresultsapply tosituations inwhich the subject canchooseto accept or reject a series of foods but where factors likeavailability, cost, and health concerns do not come intoplay. So, it is clear that a broader range of issues needs tobe addressed in future studies in order to fully under-stand responses to unfamiliar foods by the elderly.

The mechanism for increased willingness to tryunfamiliar foods described in this paper is unlikely toaccount for the decreases in food neophobia seen duringchildhood and young-adulthood because there do notseem to be significant age-related declines in olfactorysensitivity in these age groups. Therefore, other, moregeneral factors have to be postulated. One promisingexplanation, at least for older children andyoung adults,is that the accumulation of previous positive experienceswith novel foods disconfirms the expectation that novelfoods will be bad-tasting, and leads to increased will-ingness to try still other novel foods (Loewen & Pliner,1999; Pliner et al., 1993).

References

Beebee-Center, J.G. (1932). The psychology of pleasantnessand unpleasantness. New York: Van Nostrand.

Betts, N.M. (1988). Nutrition perspectives on aging. Ameri-can Behavioral Scientist 32, 17±30.

Birch, L.L. (1979a). Preschool children's food preferencesand consumption patterns. Journal of Nutrition Education11, 189±192.

Birch, L.L. (1979b). Dimensions of preschool children's foodpreferences. Journal of Nutrition Education 11, 77±80.

Birch, L.L. & Marlin, D.W. (1982). Effects of exposure ontwo year old children's food preferences. Appetite 3,353±360.

Birch, L.L., McPhee, L., Shoba, B.C., Pirok, E. &Steinberg, L. (1987). What kind of exposure reduceschildren's food neophobia. Appetite 9, 171±178.

Cain, W.S. & Johnson, F., Jr (1978). Lability of odor pleas-antness: influence of mere exposure. Perception 7,459±465.

Cain, W.S. & Stevens, J.C. (1989). Uniformity of olfactoryloss in aging. Annals of the New York Academy ofSciences 561, 29±38.

Chalke, H.D. & Dewhurst, J.R. (1957). Accidental coal-gaspoisoning. British Medical Journal 915±917.

Cowart, B. (1989). Relationships of taste and smell across theadult life span. Annals of the New York Academy ofSciences, 564, 39±55.

Enns, M.P. & Hornung, D.E. (1988). Comparisons of theestimates of smell, taste and overall intensity in youngand elderly people. Chemical Senses 13, 131±139.

Olfaction andresponses to novel foods by the elderly 159

Page 8: You can teach an old dog new tricks: olfaction and responses to novel foods by the elderly

Frank, R.A. & van der Klaauw, N.J. (1994). The contribu-tion of chemosensory factors to individual differences inreported food preferences. Appetite 22, 101±123.

Havlik, R.J. (1992). Health statistics on older persons. InH. Munro & G. Schleirf (Eds), Nutrition of the elderly,Nestle nutrition workshop series Pp. 7±16. New York:Vevy/Raven Press.

Kohrs, M.B. (1982). A rational diet for the elderly. Journal ofClinical Nutrition 36, 796±802.

Loewen, R. & Pliner, P. (1999). Effects of prior exposure topalatable and unpalatable novel foods on children's will-ingness to taste other novel foods. Appetite 32, 351±366.

Meiselman, H.L. & Waterman, D. (1978). Food preferencesof enlisted personnel in the armed forces. Journal of theAmerican Dietetic Association 73, 621±629.

Murphy, C. (1982). Effects of exposure and context on hedon-ics of olfactory-taste mixtures. In J.T. Kuznicki,R.A. Johnson & A.F. Rutkiewic (Eds), Selected sensorymethods: problems and approaches to measuring hedonics.Pp. 60±70. Philadelphia, PA: American Society for Test-ing Materials.

Murphy, C. (1983). Age-related effects on the threshold,psychophysical function, and pleasantness of Menthol.Journal of Gerontology 38, 217±222.

Murphy, C. (1985). Cognitive and chemosensory influenceson age-related changes in the ability to identify blendedfoods. Journal of Gerontology 40, 47±52.

Otis, L.P. (1984). Factors influencing the willingness to tasteunusual foods. Psychological Reports 54, 739±745.

Pelchat, M.L. & Pliner, P. (1995). `̀ Try it. You'll like it.''Effects of information on willingness to try novel foods.Appetite 24, 153±166.

Pelchat, M.L. & Stoess, C. (1991). Aging, olfaction and foodpreferences. Paper presented at the Meeting of the Asso-ciation for Chemoreception Sciences, Sarasota.

Pliner, P. (1982). The effect of mere exposure on liking foredible substances. Appetite 3, 283±290.

Pliner, P. & Hobden, K. (1992). Development of a scale tomeasure the trait of food neophobia in humans. Appetite19, 105±120.

Pliner, P., Pelchat, M.L. & Grabski, M. (1993). Reduction ofneophobia in humans by exposure to novel foods.Appetite 20, 111±123.

Raudenbush, B., van der Klaauw, N.J. & Frank, R.A.(1995). The contribution of psychological and sensoryfactors to food preference patterns as measured by FoodAttitudes Survey (FAS). Appetite 22, 1±15.

Schiffman, S. (1977). Food recognition by the elderly.Journal of Gerontology 32, 586±592.

Schiffman, S. & Covey, E. (1984). Changes in taste and smellwith age: nutritional aspects. In J.M. Ordy, D. Harman &R. Alfin-Slater (Eds). Nutrition in Gerontology. Pp. 43±64. New York: Raven Press.

Schiffman, S. & Pasternak, M. (1979). Decreased discrimina-tion of food odors in the elderly. Journal of Gerontology34, 73±79.

Schutz, H.G. & Pilgrim, F.J. (1958). A field of food monot-ony. Psychological Reports 4, 559±565.

Stang, D.J. (1975). When familiarity breeds contempt,absence makes the heart grow fonder: effects of exposureand delay on taste pleasantness ratings. Bulletin of thePsychonomic Society 6, 273±275.

Torrance, E.P. (1958). Sensitization versus adaptation in pre-paration for emergencies: prior experience with an emer-gency ration and its acceptability in a simulated survivalsituation. Journal of Applied Psychology 42, 63±67.

Wysocki, C.J. & Pelchat, M.L. (1993). The effects of aging onthe human sense of smell and its relationship to foodchoice. Critical Reviews in Food Science and Nutrition33, 63±82.

Zajonc, R.B. (1968). Attitudinal effects of mere exposure.Journal of Personality and Social Psychology MonographSupplement 9, 1±26.

160 M.L. Pelchat