This article was downloaded by: [University of Sussex Library] On: 02 February 2015, At: 06:48 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Neurocase: The Neural Basis of Cognition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/nncs20
Perception of odor-induced tastes following insular cortex lesion a
bc
Richard J. Stevenson , Laurie A. Miller
& Ky McGrillen
d
a
Department of Psychology, Macquarie University, NSW 2109 Sydney, Australia
b
Neuropsychology Unit, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia
c
Department of Medicine, University of Sydney, NSW 2006 Sydney, Australia
d
Radiology Department, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia Published online: 05 Dec 2013.
Click for updates To cite this article: Richard J. Stevenson, Laurie A. Miller & Ky McGrillen (2015) Perception of odor-induced tastes following insular cortex lesion, Neurocase: The Neural Basis of Cognition, 21:1, 33-43, DOI: 10.1080/13554794.2013.860175 To link to this article: http://dx.doi.org/10.1080/13554794.2013.860175
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Neurocase, 2015 Vol. 21, No. 1, 33–43, http://dx.doi.org/10.1080/13554794.2013.860175
Perception of odor-induced tastes following insular cortex lesion Richard J. Stevensona*, Laurie A. Millerb,c and Ky McGrillend a
Department of Psychology, Macquarie University, NSW 2109 Sydney, Australia; bNeuropsychology Unit, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia; cDepartment of Medicine, University of Sydney, NSW 2006 Sydney, Australia; dRadiology Department, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
(Received 30 October 2012; accepted 11 September 2013) Lesions of the insula can affect olfaction and gustation. Here, we examined the effect of insula lesions on taste and taste-like experiences generated via smelling (i.e., odor-induced tastes) in patients with focal insula lesions and intact olfaction. From a set of 16 patients with lesions to the insula, we found 7 (6 with right-sided lesions) who performed normally on various olfactory measures. These were compared to 42 normal control subjects on tests of gustatory and odor-induced taste perception as well as control measures. The patients were impaired relative to controls on most gustatory measures. They were also impaired on tests of odor-induced taste perception, primarily for stimuli presented on the left side. Examining cases individually revealed evidence of a dissociation: two patients exhibited no impairment in odor-induced taste perception in spite of gustatory deficits. Together, these findings suggest that the insula mediates taste recognition, hedonics, and intensity judgments as well as odor-induced taste perception. However, the areas responsible for aspects of taste perception and those responsible for odor-induced taste do not fully overlap each other and they are also independent of olfactory areas. Keywords: insula lesion; gustation; synesthesia; odor-induced taste; chemical senses
Four types of findings suggest that the insula functions as primary taste cortex. First, lesions to the insula are associated with impairments in gustation (Cereda, Ghika, Maeder, & Bogousslavsky, 2002; Mak, Simmons, Gitelman, & Small, 2005; Pritchard, Macaluso, & Eslinger, 1999). Second, neuroimaging reveals activation of the insula during taste perception in healthy adults (e.g., Cerf-Ducastel, Van de Moortele, MacLeod, Le Bihan, & Faurion, 2001). Third, taste sensitive neurons have been identified in the insula of rats and primates (e.g., Yaxley, Rolls, & Sienkiewicz, 1990). Fourth, neuroanatomical studies have revealed that the insula is the first cortical recipient of afferent taste information (e.g., Pritchard, Hamilton, Morse, & Norgren, 1986). While gustatory perception is typically understood to occur when tastants are applied to the surface of the tongue, taste-like sensations have also been reported when participants smell certain odors (e.g., Harper, Land, Griffiths, & Bate-Smith, 1968). Documented examples cover almost the entire range of primary taste sensations, including sweet (e.g., Stevenson, Prescott, & Boakes, 1995), sour (e.g., Stevenson, Boakes, & Prescott, 1998), bitter (e.g., Yeomans & Mobini, 2006), and salty (e.g., Nasri, Beno, Septier, Salles, & Thomas-Danguin, 2011). These odor-induced taste experiences seem to be perceptually similar to their gustatory counterpart. For example, when a sweet odor is added to a sweet taste the resultant flavor is judged to taste sweeter than the sweet taste alone *Corresponding author. Email: [email protected] © 2013 Taylor & Francis
(e.g., Frank & Byram, 1988). Moreover, the judged sweetness of a sniffed odor predicts the degree to which a sweet taste will be enhanced if that odorant is added to it (e.g., Valentin, Chrea, & Nguyen, 2006). More objective approaches also suggest perceptual concordance. Sweet tastes, but not non-sweet tastes, facilitate the detection of sweet smells (e.g., Dalton, Doolittle, Nagata, & Breslin, 2000) and taste-congruent smells facilitate the identification of their congruent taste and retard the identification of their incongruent taste (White & Prescott, 2007). Furthermore, Prescott and Wilkie (2007) reported that sweet smelling odors increased pain tolerance, just as sweet tastes have been found to do. Comparable findings have also started to emerge from the animal literature (e.g., Gautam & Verhagen, 2010). The neural basis for this commonality between tastantdriven gustation and odorant-induced gustation has not been well explored. Recent neuroimaging work (Veldhuizen, Nachtigal, & Small, 2009) suggests that activation in the insular cortex corresponds with the degree of sweetness reported for food odors. In another study involving six patients with insula lesions, Stevenson, Miller and Thayer (2008) found impairment on tests of gustation and olfaction as well as on three specific tests of odor-induced taste perception. On the first of these tests participants were asked to place four sweet and four non-sweet (food) odors on to a grid, so that each odor’s position reflected its similarity to every other odor. Controls
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
34
R.J. Stevenson et al.
clustered the four sweet-smelling odors more tightly than the four non-sweet smelling odors, relative to patients. We hypothesized that impairments in the perception of odorinduced sweetness were causing a looser clustering of the four sweet odors in patients because they did not smell alike (i.e., the odors did not smell sweet). A second test of odor-induced taste in that study involved discriminating four sweet smells. Here, we expected that if the capacity to smell sweetness was diminished, then the four sweet smells should be more discriminable. Again, this prediction arose because their common quality (sweetness) should no longer be apparent. A trend for this effect was observed. The third test relied upon participants judging the taste-like qualities (e.g., how sweet they smelled) of a range of sweet and non-sweet odors. Again, as predicted, patients were generally less able to perceive these tastelike qualities than controls. While these findings support the idea that the insula is involved in mediating odor-induced taste, it is possible that the observed impairment could result from a more general deficit in olfactory perception. Indeed, there are several reasons for treating this concern seriously. First, the insula appears active in neuroimaging studies that involve olfaction (e.g., Plailly, Radnovich, Sabri, Royet, & Kareken, 2007), suggesting that it may be implicated in multiple odor-related functions. Second, Mak et al. (2005) reported the case of a female patient with a lesion to the left posterior insula who was impaired in her capacity to perceive odor intensity. Third, Metin, Bozluolcay, and Ince, 2007, described a further patient with a flavor-related parosmia following an insula lesion. Fourth, naturally occurring lesions are rarely restricted to insular cortex and often impinge on adjacent areas in the frontal and temporal lobes (e.g., Pritchard et al., 1999), areas that have been extensively associated with olfactory impairment (e.g., Jones-Gotman & Zatorre, 1988; Savage, Combs, Pinkston, Advokat, & Gouvier, 2002). In the present study, we sought to rule out a contribution of olfactory deficits to odor-induced taste performance by selecting insula lesion patients with an intact sense of smell. We then determined whether this subset (as a group and as individuals) evidenced impairments in gustation (intensity, recognition, quality, hedonics, discrimination) and/or odor-induced taste perception on the three tests described above (similarity, discrimination, and quality). Matched visual control measures were used to rule out more general deficits in these cognitive domains. In addition, we also included (1) a neuropsychological test sensitive to language-related impairments that are known to impact upon the ability to complete olfactory identification tasks (see Westervelt, Ruffolo, & Tremont, 2005); (2) a neuropsychological test used to estimate premorbid intelligence; and (3) a screen of current overall cognitive functioning. We were interested in whether we could find individuals who showed dissociations in these three
domains (olfaction, gustation, and odor-induced taste) to suggest functional specialization within the insula. Furthermore, we carefully examined the extent and location of the lesions in each patient to see if extra-insula regions made important contributions to these processes. In sum, we hypothesized (1) that patients with an insula lesion and a demonstrably intact sense of smell, should still show impairments in their ability to experience odorinduced tastes and (2) that there is evidence for functional specialization in the insula for taste, smell, and odorinduced taste.
Method Participants Sixteen patients with a lesion to their insular cortex were recruited over a 6-year period from the neurology department at the Royal Prince Alfred Hospital (RPAH), Sydney, Australia. Data from six patients who had completed our original test battery a few years earlier (MP, SE, RD, RS, FC, SK; see Stevenson et al., 2008) were included here and three of those subjects (SE, FC, SK) agreed to return to perform additional smell identification testing and a control test of visual hedonics. Control subjects were sought to match the patient group in terms of age range, gender distribution, and level of education. Control participants were recruited in two phases: an initial sample of 19 reported in the earlier study (Stevenson et al., 2008) and a further group of 23, who completed the same test battery as well as the additional testing, yielding a total of 42 normal control participants. All subjects spoke English, had no history of psychiatric disorder or traumatic brain injury, and reported no upper respiratory tract infection at the time of testing. The control subjects had no history of neurological disorders. Lesion analysis, based on clinical MRI and CT images, was conducted by a neuroradiologist (KM), who was blind to the test results (see Figure 1 for sample images). KM calculated lesion location, extent, and volume for each patient (see Naidich et al., 2004, for anatomical definitions). Informed consent was obtained from each participant and the protocol was approved by the Macquarie University Human Research Ethics Committee (HREC) and the Sydney South West Area Health Authority HREC.
Materials and measures Olfactory tests Odor name recognition. The Smell Identification Test (formerly known as the University of Pennsylvania Smell Identification Test, UPSIT; Doty, 1995) was employed to measure identification ability for a standard set of odors. Half of this test’s 40 scratch-and-sniff stimuli were presented to the left nostril (always booklets 1 and 4) and the
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
Neurocase
35
Figure 1. Axial MRI scans of patients in the intact olfaction group: From top left to bottom right, the images represent cases RS, DW, FC, NB, MW, MP, and TD. Note that all scan images are left–right reversed.
remaining half to the right (always booklets 2 and 3). The opposite nostril was occluded using a finger pressed to the base of the nostril. Participants selected a name for each odor from the four provided for each smell, providing a score from 0 to 20 for each side tested. Odorant naming and intensity judgment. Four odorants (chocolate sauce [12.5 g; Cadbury], plum essence [0.05 g; Quest], vegemite [12.0 g; Kraft], and oregano oil [0.08 gl Dragoco]) were presented in opaque, visually identical, plastic squeeze bottles. Odors were delivered by giving three puffs with the bottle placed 7 cm below the open nostril while participants sniffed. Using a random presentation order, with alternation between the left and right sides, each odorant was sampled once by each nostril (with the opposing nostril occluded). Participants were asked to complete a number of scales to describe the quality of the odorant (which are described later below). Two measures were used here to assess general olfactory ability. First, mean ratings of odor intensity (how strong? on a scale of Not at all [0] to Very [7]) for each side tested were determined. Second, total number of correct name choices (maximum = 4) from the four available for each trial (choices were chocolate, plum, vegemite, oregano) by side tested was found.
Gustatory testing Tastant evaluations. Four taste solutions were used: sucrose (292 mM), saline (200 mM), citric acid (9 mM; Sigma, Sydney) and quinine (0.2 mM; Sigma, Sydney). Each tastant was applied by cotton swab (Q tip) to each side of the tongue in a fixed random order, alternating between the left and right anterior tip, so that each tastant
was sampled once by each tongue side. Following application of a tastant, the participant completed six ratings concerning the taste; these were (1) the four taste qualities sweet, sour [acidic], salty, bitter (using an eight-point scale from Not at all [0] to Very [7]); (2) taste intensity (strong; same scale); and (3) how much they liked the tastant (hedonics; using an eight-point scale dislike extremely [0] to like extremely [7]. Finally, participants were also asked to select the name of the tastant from four available choices (sweet, sour [acidic], salty, bitter). Between each tastant there was a water rinse. Four scores were then derived from these data for the left and right sides, respectively: (1) A taste name recognition score, this being the sum of correct selections (maximum = 4); (2) A taste intensity score, being the mean of the four taste intensity ratings (maximum = 7); (3) A taste quality score, this being the average of four scores derived for each tastant, namely the target rating (e.g., sweetness for sucrose) minus the mean of the remaining three nontarget ratings (maximum = 7); and (4) A taste hedonic score, this being the hedonic rating for sucrose minus the mean rating for the three remaining tastes (these likely to be liked less; maximum = 7). Taste discrimination test. The four tastants described in the previous test were used here. There were 12 trials, presented in a randomized order, consisting of six presentations of one taste then the same taste again (sucrose vs. sucrose, citric acid vs. citric acid, salt vs. salt, for each side, respectively), and six involving one taste followed by a different one (quinine vs. sucrose, salt vs. sucrose, citric acid vs. sucrose, for each side, respectively). Tastants were applied by stroking a tastant saturated cotton swab (Q tip) across the anterior side (left or right) of the tongue. After
36
R.J. Stevenson et al.
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
each trial participants were asked to respond to “same” or “different,” which was then followed by a water rinse. Side testing alternated between left and right side of the tongue on a trial-by-trial basis. A d’ score was then calculated for each side. Odor-induced taste measures Odor similarity test. Eight different odorants were used, divided into two sets, one composed of four sweet smelling odors and the other of four non-sweet smelling odors. The sweet set was; chocolate sauce (12.5 g; Cadbury), vanilla essence (0.18 g; Dragoco), banana essence (0.15 g; Quest), and plum essence (0.05 g; Quest). The nonsweet sets were vegemite (12.0 g; Kraft), rogan josh curry paste (8.0 g; Patak), garlic paste (2.5 g; Homebrand), and oregano oil (0.08 g Dragoco). Although the taste-like properties of odors are probably acquired, selection of sweet or non-sweet smelling stimuli is not particularly difficult as most participants will have encountered these odors as flavors (i.e., with the respective taste) in commonly eaten foods. This test was conducted for the left nostril first (a fixed order of testing was used so that we could compare any group or person). Participants were presented with a large piece of paper with 81 squares printed on it arranged in a 9 by 9 grid. At the center of this grid, the experimenter placed garlic odorant (this being fixed) and the participant were then presented with the remaining seven odorants in a fixed random order. On each presentation, and starting with garlic, participants sniffed the odor (all odor presentation was done by the experimenter), and then placed that particular odorant (with the exception of garlic which acted as the anchor) on the grid in such a manner that its distance from the other odors represented its similarity to them. Participants were allowed to re-smell the odorants and to move them (with the exception of garlic), until they felt that the resulting layout represented the similarity of each odorant to every other odorant. Within set similarity was then calculated using a Euclidean metric. As the anchored (non-sweet) set had a smaller possible score, relative to the unanchored set (sweet), scores were corrected by dividing them by the maximum possible score obtainable for that set. These scores were then linearly transformed so that they fell between 0 and 1, with a higher score indicating greater dispersal (i.e., less withinset similarity). Subsequently, the same procedure was used for odorants presented to the right nostril. Impairments in odor-induced taste perception were expected to result in lower clustering scores for the sweet odorants relative to the non-sweet ones. Odorant evaluations. Ratings of qualities of the four odorants described above (chocolate, plum, vegemite, and oregano) were obtained here, again separately for the left and
right sides (so that each odorant was sampled once by each nostril), with trials presented in random order with the constraint that nostril tested always alternated between the right and the left on a trial-by-trial basis. Participants rated each odorant on the same set of scales as used for the taste evaluations (i.e., four taste qualities, intensity, hedonics, and naming). An overall odorant quality score was calculated by taking the average of four amalgamated scores derived for each of the odorants (the amalgamated quality score being the rating on the expected (corresponding) taste scale minus the average of the taste qualities that were not expected). Hedonic data were not used as they were highly variable, and intensity and naming data preparation were described above. Sweet odor discrimination test. The four sweet smelling odorants described above (i.e., chocolate sauce, vanilla essence, banana essence, and plum essence) were used here. This test was composed of 24 trials, split over two parts. For each of the parts, six trials involved presentation of the same odor twice (same trials; three to the left nostril and three to the right), and in another six, one odorant was followed by a different odorant (different trials; three to the left nostril and three to the right), with these trials occurring in randomized order, with the constraint that the nostril tested always alternated between the right and the left on a trial-by-trial basis. Over both parts of testing, each nostril received an identical set of same and different trials. Participants were asked to respond “same” or “different” on each trial and a d’ score was calculated to reflect discriminative performance for each side.
Visual control tests The visual similarity test used two sets of four amoeboid shapes and was designed to parallel the olfactory similarity test (see Stevenson et al., 2008 for details). Visual similarity data were scored in the same manner (within set similarity) as the odor similarity data. The visual discrimination test, paralleled the taste and odor discrimination tests, with a d’ score calculated for each participant (see Stevenson et al., 2008 for details). The visual hedonic test used images from the International Affective Picture Series (IAPS; Lang, Bradley & Cuthbert, 2001) and consisted four negative (as judged by the IAPS reference sample) images (2120, angry face; 9290, garbage; 9830, cigarette butts; 1300 pit bull dog), four neutral (8311, golfer; 7180 neon lit building; 7050, hairdresser; 1670, cow), and four positive images (2050, baby; 8190, skiers; 1463, kittens; 5982, sky), with pictures presented in random order. After viewing each image participants used the same hedonic scale described for the odor and taste evaluation tests above. As the neutral images did not
Neurocase significantly differ between patients and controls, just the contrast between the mean hedonic ratings for the pleasant images minus the mean hedonic rating for the unpleasant images is reported.
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
Neuropsychological tests The Mini Mental State Exam (MMSE; Folstein, Folstein, & McHugh, 1975) and the National Adult Reading Test (NART-2; Nelson & Willison, 1991) were completed to provide a basic snapshot of current cognitive functioning and to estimate premorbid IQ, respectively. The Boston Naming Test (BNT; Kaplan, Goodglas, & Weintraub, 1983) is a confrontation naming task in which participants are required to name a series of line drawings of everyday items within a set period of time, after which prompts are provided. The BNT was completed by all participants, as it requires the same general cognitive and language skills as the odor identification task (Westervelt et al., 2005). Procedure Once consent was obtained, subjects were interviewed to gather relevant demographic and medical information. The test battery was then presented in a fixed order, with most of the olfactory and gustatory tasks being divided into different components and interspersed, so as to minimize olfactory and gustatory adaptation. Testing took Table 1.
37
approximately 2.5 hours, with a 30 min break. Additional breaks were permitted if needed.
Analysis and selection of patients We started by identifying patients who had an olfactory impairment. Each patient’s results on the general olfactory performance tests were examined relative to the performance of the control group, which was used to establish the 95% cut-offs (note that four controls fall at or below this cut-off), as normative data were not available for the split UPSIT test as used here. As can be seen in Table 1, 7 of the 16 patients had no apparent olfactory impairment, scoring above the cut-offs on all of the general olfactory performance tests they completed. Poorer odor performance in the non-selected patients was not associated with more general cognitive impairments, as they did not differ from the selected group in terms of the Boston Naming Test, estimated IQ, or MMSE (all ts < 1). The intact olfaction patients had predominantly right-sided insula lesions (6/7), in contrast to the excluded impaired olfaction patients, where only 2/9 had unilateral rightsided lesions. The number of right vs. left-sided (and bilateral) lesions differed significantly between the two groups, Fisher exact test, p < .05. The taste data for the intact olfaction patients were then compared to the normal control participants’ data using a series of mixed design ANOVAs, so as to indicate
Performance on olfactory tasks used to group patients into those with and without general olfactory impairments.
Group/Patient (lesion side)
UPSIT† L
Normal control performance: Mean (SD) 17.3 (2.0) 95% cut-off
Neurocase: The Neural Basis of Cognition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/nncs20
Perception of odor-induced tastes following insular cortex lesion a
bc
Richard J. Stevenson , Laurie A. Miller
& Ky McGrillen
d
a
Department of Psychology, Macquarie University, NSW 2109 Sydney, Australia
b
Neuropsychology Unit, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia
c
Department of Medicine, University of Sydney, NSW 2006 Sydney, Australia
d
Radiology Department, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia Published online: 05 Dec 2013.
Click for updates To cite this article: Richard J. Stevenson, Laurie A. Miller & Ky McGrillen (2015) Perception of odor-induced tastes following insular cortex lesion, Neurocase: The Neural Basis of Cognition, 21:1, 33-43, DOI: 10.1080/13554794.2013.860175 To link to this article: http://dx.doi.org/10.1080/13554794.2013.860175
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Neurocase, 2015 Vol. 21, No. 1, 33–43, http://dx.doi.org/10.1080/13554794.2013.860175
Perception of odor-induced tastes following insular cortex lesion Richard J. Stevensona*, Laurie A. Millerb,c and Ky McGrillend a
Department of Psychology, Macquarie University, NSW 2109 Sydney, Australia; bNeuropsychology Unit, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia; cDepartment of Medicine, University of Sydney, NSW 2006 Sydney, Australia; dRadiology Department, Royal Prince Alfred Hospital, NSW 2050 Sydney, Australia
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
(Received 30 October 2012; accepted 11 September 2013) Lesions of the insula can affect olfaction and gustation. Here, we examined the effect of insula lesions on taste and taste-like experiences generated via smelling (i.e., odor-induced tastes) in patients with focal insula lesions and intact olfaction. From a set of 16 patients with lesions to the insula, we found 7 (6 with right-sided lesions) who performed normally on various olfactory measures. These were compared to 42 normal control subjects on tests of gustatory and odor-induced taste perception as well as control measures. The patients were impaired relative to controls on most gustatory measures. They were also impaired on tests of odor-induced taste perception, primarily for stimuli presented on the left side. Examining cases individually revealed evidence of a dissociation: two patients exhibited no impairment in odor-induced taste perception in spite of gustatory deficits. Together, these findings suggest that the insula mediates taste recognition, hedonics, and intensity judgments as well as odor-induced taste perception. However, the areas responsible for aspects of taste perception and those responsible for odor-induced taste do not fully overlap each other and they are also independent of olfactory areas. Keywords: insula lesion; gustation; synesthesia; odor-induced taste; chemical senses
Four types of findings suggest that the insula functions as primary taste cortex. First, lesions to the insula are associated with impairments in gustation (Cereda, Ghika, Maeder, & Bogousslavsky, 2002; Mak, Simmons, Gitelman, & Small, 2005; Pritchard, Macaluso, & Eslinger, 1999). Second, neuroimaging reveals activation of the insula during taste perception in healthy adults (e.g., Cerf-Ducastel, Van de Moortele, MacLeod, Le Bihan, & Faurion, 2001). Third, taste sensitive neurons have been identified in the insula of rats and primates (e.g., Yaxley, Rolls, & Sienkiewicz, 1990). Fourth, neuroanatomical studies have revealed that the insula is the first cortical recipient of afferent taste information (e.g., Pritchard, Hamilton, Morse, & Norgren, 1986). While gustatory perception is typically understood to occur when tastants are applied to the surface of the tongue, taste-like sensations have also been reported when participants smell certain odors (e.g., Harper, Land, Griffiths, & Bate-Smith, 1968). Documented examples cover almost the entire range of primary taste sensations, including sweet (e.g., Stevenson, Prescott, & Boakes, 1995), sour (e.g., Stevenson, Boakes, & Prescott, 1998), bitter (e.g., Yeomans & Mobini, 2006), and salty (e.g., Nasri, Beno, Septier, Salles, & Thomas-Danguin, 2011). These odor-induced taste experiences seem to be perceptually similar to their gustatory counterpart. For example, when a sweet odor is added to a sweet taste the resultant flavor is judged to taste sweeter than the sweet taste alone *Corresponding author. Email: [email protected] © 2013 Taylor & Francis
(e.g., Frank & Byram, 1988). Moreover, the judged sweetness of a sniffed odor predicts the degree to which a sweet taste will be enhanced if that odorant is added to it (e.g., Valentin, Chrea, & Nguyen, 2006). More objective approaches also suggest perceptual concordance. Sweet tastes, but not non-sweet tastes, facilitate the detection of sweet smells (e.g., Dalton, Doolittle, Nagata, & Breslin, 2000) and taste-congruent smells facilitate the identification of their congruent taste and retard the identification of their incongruent taste (White & Prescott, 2007). Furthermore, Prescott and Wilkie (2007) reported that sweet smelling odors increased pain tolerance, just as sweet tastes have been found to do. Comparable findings have also started to emerge from the animal literature (e.g., Gautam & Verhagen, 2010). The neural basis for this commonality between tastantdriven gustation and odorant-induced gustation has not been well explored. Recent neuroimaging work (Veldhuizen, Nachtigal, & Small, 2009) suggests that activation in the insular cortex corresponds with the degree of sweetness reported for food odors. In another study involving six patients with insula lesions, Stevenson, Miller and Thayer (2008) found impairment on tests of gustation and olfaction as well as on three specific tests of odor-induced taste perception. On the first of these tests participants were asked to place four sweet and four non-sweet (food) odors on to a grid, so that each odor’s position reflected its similarity to every other odor. Controls
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
34
R.J. Stevenson et al.
clustered the four sweet-smelling odors more tightly than the four non-sweet smelling odors, relative to patients. We hypothesized that impairments in the perception of odorinduced sweetness were causing a looser clustering of the four sweet odors in patients because they did not smell alike (i.e., the odors did not smell sweet). A second test of odor-induced taste in that study involved discriminating four sweet smells. Here, we expected that if the capacity to smell sweetness was diminished, then the four sweet smells should be more discriminable. Again, this prediction arose because their common quality (sweetness) should no longer be apparent. A trend for this effect was observed. The third test relied upon participants judging the taste-like qualities (e.g., how sweet they smelled) of a range of sweet and non-sweet odors. Again, as predicted, patients were generally less able to perceive these tastelike qualities than controls. While these findings support the idea that the insula is involved in mediating odor-induced taste, it is possible that the observed impairment could result from a more general deficit in olfactory perception. Indeed, there are several reasons for treating this concern seriously. First, the insula appears active in neuroimaging studies that involve olfaction (e.g., Plailly, Radnovich, Sabri, Royet, & Kareken, 2007), suggesting that it may be implicated in multiple odor-related functions. Second, Mak et al. (2005) reported the case of a female patient with a lesion to the left posterior insula who was impaired in her capacity to perceive odor intensity. Third, Metin, Bozluolcay, and Ince, 2007, described a further patient with a flavor-related parosmia following an insula lesion. Fourth, naturally occurring lesions are rarely restricted to insular cortex and often impinge on adjacent areas in the frontal and temporal lobes (e.g., Pritchard et al., 1999), areas that have been extensively associated with olfactory impairment (e.g., Jones-Gotman & Zatorre, 1988; Savage, Combs, Pinkston, Advokat, & Gouvier, 2002). In the present study, we sought to rule out a contribution of olfactory deficits to odor-induced taste performance by selecting insula lesion patients with an intact sense of smell. We then determined whether this subset (as a group and as individuals) evidenced impairments in gustation (intensity, recognition, quality, hedonics, discrimination) and/or odor-induced taste perception on the three tests described above (similarity, discrimination, and quality). Matched visual control measures were used to rule out more general deficits in these cognitive domains. In addition, we also included (1) a neuropsychological test sensitive to language-related impairments that are known to impact upon the ability to complete olfactory identification tasks (see Westervelt, Ruffolo, & Tremont, 2005); (2) a neuropsychological test used to estimate premorbid intelligence; and (3) a screen of current overall cognitive functioning. We were interested in whether we could find individuals who showed dissociations in these three
domains (olfaction, gustation, and odor-induced taste) to suggest functional specialization within the insula. Furthermore, we carefully examined the extent and location of the lesions in each patient to see if extra-insula regions made important contributions to these processes. In sum, we hypothesized (1) that patients with an insula lesion and a demonstrably intact sense of smell, should still show impairments in their ability to experience odorinduced tastes and (2) that there is evidence for functional specialization in the insula for taste, smell, and odorinduced taste.
Method Participants Sixteen patients with a lesion to their insular cortex were recruited over a 6-year period from the neurology department at the Royal Prince Alfred Hospital (RPAH), Sydney, Australia. Data from six patients who had completed our original test battery a few years earlier (MP, SE, RD, RS, FC, SK; see Stevenson et al., 2008) were included here and three of those subjects (SE, FC, SK) agreed to return to perform additional smell identification testing and a control test of visual hedonics. Control subjects were sought to match the patient group in terms of age range, gender distribution, and level of education. Control participants were recruited in two phases: an initial sample of 19 reported in the earlier study (Stevenson et al., 2008) and a further group of 23, who completed the same test battery as well as the additional testing, yielding a total of 42 normal control participants. All subjects spoke English, had no history of psychiatric disorder or traumatic brain injury, and reported no upper respiratory tract infection at the time of testing. The control subjects had no history of neurological disorders. Lesion analysis, based on clinical MRI and CT images, was conducted by a neuroradiologist (KM), who was blind to the test results (see Figure 1 for sample images). KM calculated lesion location, extent, and volume for each patient (see Naidich et al., 2004, for anatomical definitions). Informed consent was obtained from each participant and the protocol was approved by the Macquarie University Human Research Ethics Committee (HREC) and the Sydney South West Area Health Authority HREC.
Materials and measures Olfactory tests Odor name recognition. The Smell Identification Test (formerly known as the University of Pennsylvania Smell Identification Test, UPSIT; Doty, 1995) was employed to measure identification ability for a standard set of odors. Half of this test’s 40 scratch-and-sniff stimuli were presented to the left nostril (always booklets 1 and 4) and the
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
Neurocase
35
Figure 1. Axial MRI scans of patients in the intact olfaction group: From top left to bottom right, the images represent cases RS, DW, FC, NB, MW, MP, and TD. Note that all scan images are left–right reversed.
remaining half to the right (always booklets 2 and 3). The opposite nostril was occluded using a finger pressed to the base of the nostril. Participants selected a name for each odor from the four provided for each smell, providing a score from 0 to 20 for each side tested. Odorant naming and intensity judgment. Four odorants (chocolate sauce [12.5 g; Cadbury], plum essence [0.05 g; Quest], vegemite [12.0 g; Kraft], and oregano oil [0.08 gl Dragoco]) were presented in opaque, visually identical, plastic squeeze bottles. Odors were delivered by giving three puffs with the bottle placed 7 cm below the open nostril while participants sniffed. Using a random presentation order, with alternation between the left and right sides, each odorant was sampled once by each nostril (with the opposing nostril occluded). Participants were asked to complete a number of scales to describe the quality of the odorant (which are described later below). Two measures were used here to assess general olfactory ability. First, mean ratings of odor intensity (how strong? on a scale of Not at all [0] to Very [7]) for each side tested were determined. Second, total number of correct name choices (maximum = 4) from the four available for each trial (choices were chocolate, plum, vegemite, oregano) by side tested was found.
Gustatory testing Tastant evaluations. Four taste solutions were used: sucrose (292 mM), saline (200 mM), citric acid (9 mM; Sigma, Sydney) and quinine (0.2 mM; Sigma, Sydney). Each tastant was applied by cotton swab (Q tip) to each side of the tongue in a fixed random order, alternating between the left and right anterior tip, so that each tastant
was sampled once by each tongue side. Following application of a tastant, the participant completed six ratings concerning the taste; these were (1) the four taste qualities sweet, sour [acidic], salty, bitter (using an eight-point scale from Not at all [0] to Very [7]); (2) taste intensity (strong; same scale); and (3) how much they liked the tastant (hedonics; using an eight-point scale dislike extremely [0] to like extremely [7]. Finally, participants were also asked to select the name of the tastant from four available choices (sweet, sour [acidic], salty, bitter). Between each tastant there was a water rinse. Four scores were then derived from these data for the left and right sides, respectively: (1) A taste name recognition score, this being the sum of correct selections (maximum = 4); (2) A taste intensity score, being the mean of the four taste intensity ratings (maximum = 7); (3) A taste quality score, this being the average of four scores derived for each tastant, namely the target rating (e.g., sweetness for sucrose) minus the mean of the remaining three nontarget ratings (maximum = 7); and (4) A taste hedonic score, this being the hedonic rating for sucrose minus the mean rating for the three remaining tastes (these likely to be liked less; maximum = 7). Taste discrimination test. The four tastants described in the previous test were used here. There were 12 trials, presented in a randomized order, consisting of six presentations of one taste then the same taste again (sucrose vs. sucrose, citric acid vs. citric acid, salt vs. salt, for each side, respectively), and six involving one taste followed by a different one (quinine vs. sucrose, salt vs. sucrose, citric acid vs. sucrose, for each side, respectively). Tastants were applied by stroking a tastant saturated cotton swab (Q tip) across the anterior side (left or right) of the tongue. After
36
R.J. Stevenson et al.
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
each trial participants were asked to respond to “same” or “different,” which was then followed by a water rinse. Side testing alternated between left and right side of the tongue on a trial-by-trial basis. A d’ score was then calculated for each side. Odor-induced taste measures Odor similarity test. Eight different odorants were used, divided into two sets, one composed of four sweet smelling odors and the other of four non-sweet smelling odors. The sweet set was; chocolate sauce (12.5 g; Cadbury), vanilla essence (0.18 g; Dragoco), banana essence (0.15 g; Quest), and plum essence (0.05 g; Quest). The nonsweet sets were vegemite (12.0 g; Kraft), rogan josh curry paste (8.0 g; Patak), garlic paste (2.5 g; Homebrand), and oregano oil (0.08 g Dragoco). Although the taste-like properties of odors are probably acquired, selection of sweet or non-sweet smelling stimuli is not particularly difficult as most participants will have encountered these odors as flavors (i.e., with the respective taste) in commonly eaten foods. This test was conducted for the left nostril first (a fixed order of testing was used so that we could compare any group or person). Participants were presented with a large piece of paper with 81 squares printed on it arranged in a 9 by 9 grid. At the center of this grid, the experimenter placed garlic odorant (this being fixed) and the participant were then presented with the remaining seven odorants in a fixed random order. On each presentation, and starting with garlic, participants sniffed the odor (all odor presentation was done by the experimenter), and then placed that particular odorant (with the exception of garlic which acted as the anchor) on the grid in such a manner that its distance from the other odors represented its similarity to them. Participants were allowed to re-smell the odorants and to move them (with the exception of garlic), until they felt that the resulting layout represented the similarity of each odorant to every other odorant. Within set similarity was then calculated using a Euclidean metric. As the anchored (non-sweet) set had a smaller possible score, relative to the unanchored set (sweet), scores were corrected by dividing them by the maximum possible score obtainable for that set. These scores were then linearly transformed so that they fell between 0 and 1, with a higher score indicating greater dispersal (i.e., less withinset similarity). Subsequently, the same procedure was used for odorants presented to the right nostril. Impairments in odor-induced taste perception were expected to result in lower clustering scores for the sweet odorants relative to the non-sweet ones. Odorant evaluations. Ratings of qualities of the four odorants described above (chocolate, plum, vegemite, and oregano) were obtained here, again separately for the left and
right sides (so that each odorant was sampled once by each nostril), with trials presented in random order with the constraint that nostril tested always alternated between the right and the left on a trial-by-trial basis. Participants rated each odorant on the same set of scales as used for the taste evaluations (i.e., four taste qualities, intensity, hedonics, and naming). An overall odorant quality score was calculated by taking the average of four amalgamated scores derived for each of the odorants (the amalgamated quality score being the rating on the expected (corresponding) taste scale minus the average of the taste qualities that were not expected). Hedonic data were not used as they were highly variable, and intensity and naming data preparation were described above. Sweet odor discrimination test. The four sweet smelling odorants described above (i.e., chocolate sauce, vanilla essence, banana essence, and plum essence) were used here. This test was composed of 24 trials, split over two parts. For each of the parts, six trials involved presentation of the same odor twice (same trials; three to the left nostril and three to the right), and in another six, one odorant was followed by a different odorant (different trials; three to the left nostril and three to the right), with these trials occurring in randomized order, with the constraint that the nostril tested always alternated between the right and the left on a trial-by-trial basis. Over both parts of testing, each nostril received an identical set of same and different trials. Participants were asked to respond “same” or “different” on each trial and a d’ score was calculated to reflect discriminative performance for each side.
Visual control tests The visual similarity test used two sets of four amoeboid shapes and was designed to parallel the olfactory similarity test (see Stevenson et al., 2008 for details). Visual similarity data were scored in the same manner (within set similarity) as the odor similarity data. The visual discrimination test, paralleled the taste and odor discrimination tests, with a d’ score calculated for each participant (see Stevenson et al., 2008 for details). The visual hedonic test used images from the International Affective Picture Series (IAPS; Lang, Bradley & Cuthbert, 2001) and consisted four negative (as judged by the IAPS reference sample) images (2120, angry face; 9290, garbage; 9830, cigarette butts; 1300 pit bull dog), four neutral (8311, golfer; 7180 neon lit building; 7050, hairdresser; 1670, cow), and four positive images (2050, baby; 8190, skiers; 1463, kittens; 5982, sky), with pictures presented in random order. After viewing each image participants used the same hedonic scale described for the odor and taste evaluation tests above. As the neutral images did not
Neurocase significantly differ between patients and controls, just the contrast between the mean hedonic ratings for the pleasant images minus the mean hedonic rating for the unpleasant images is reported.
Downloaded by [University of Sussex Library] at 06:48 02 February 2015
Neuropsychological tests The Mini Mental State Exam (MMSE; Folstein, Folstein, & McHugh, 1975) and the National Adult Reading Test (NART-2; Nelson & Willison, 1991) were completed to provide a basic snapshot of current cognitive functioning and to estimate premorbid IQ, respectively. The Boston Naming Test (BNT; Kaplan, Goodglas, & Weintraub, 1983) is a confrontation naming task in which participants are required to name a series of line drawings of everyday items within a set period of time, after which prompts are provided. The BNT was completed by all participants, as it requires the same general cognitive and language skills as the odor identification task (Westervelt et al., 2005). Procedure Once consent was obtained, subjects were interviewed to gather relevant demographic and medical information. The test battery was then presented in a fixed order, with most of the olfactory and gustatory tasks being divided into different components and interspersed, so as to minimize olfactory and gustatory adaptation. Testing took Table 1.
37
approximately 2.5 hours, with a 30 min break. Additional breaks were permitted if needed.
Analysis and selection of patients We started by identifying patients who had an olfactory impairment. Each patient’s results on the general olfactory performance tests were examined relative to the performance of the control group, which was used to establish the 95% cut-offs (note that four controls fall at or below this cut-off), as normative data were not available for the split UPSIT test as used here. As can be seen in Table 1, 7 of the 16 patients had no apparent olfactory impairment, scoring above the cut-offs on all of the general olfactory performance tests they completed. Poorer odor performance in the non-selected patients was not associated with more general cognitive impairments, as they did not differ from the selected group in terms of the Boston Naming Test, estimated IQ, or MMSE (all ts < 1). The intact olfaction patients had predominantly right-sided insula lesions (6/7), in contrast to the excluded impaired olfaction patients, where only 2/9 had unilateral rightsided lesions. The number of right vs. left-sided (and bilateral) lesions differed significantly between the two groups, Fisher exact test, p < .05. The taste data for the intact olfaction patients were then compared to the normal control participants’ data using a series of mixed design ANOVAs, so as to indicate
Performance on olfactory tasks used to group patients into those with and without general olfactory impairments.
Group/Patient (lesion side)
UPSIT† L
Normal control performance: Mean (SD) 17.3 (2.0) 95% cut-off
