|Home | About | Journals | Submit | Contact Us | Français|
Classical psychophysical studies have provided significant information on the psychophysical functions for taste stimuli. With the advent of fMRI, studies are being conducted that provide insight into central processing of gustation in humans. However, fMRI experiments impose physical limitations that influence stimulus delivery. The present study compared psychophysical functions relating perceived intensity to concentration derived from previous studies which used the traditional “sip and spit” and dorsal flow delivery methods to functions generated using a technique that simulated a technique used to deliver minute quantities of taste stimulus to the dorsal surface of the tongue in an fMRI scanner, termed the simulated stimulus delivery technique (SSDT). As hypothesized, results indicate that slopes of intensity functions were dependent on the type of stimulus delivery technique. The simulated scanning technique resulted in slopes that were more similar to those generated by dorsal flow than by sip and spit stimulus delivery techniques, suggesting the importance of considering the influence of stimulus delivery on psychophysical response in designing and interpreting experiments.
Psychophysical research on taste has consistently employed two techniques for stimulus delivery, sip and spit and dorsal flow. The classical sip and spit technique involves the participant sampling 5–20 ml of a liquid taste stimulus, swirling the solution around the mouth, and expectorating the contents (Bartoshuk & Cleveland, 1977; Meiselman, 1968, 1971; Moskowitz, 1970; Murphy & Cain, 1980; Murphy, Cain, & Bartoshuk, 1977; Schiffman & Clark, 1980; Stevens, 1969). This technique stimulates the whole mouth, which closely represents the way humans consume food (Meisleman, 1971). Dorsal flow delivers the stimulus over the extended anterior dorsal surface of the tongue while the lips are closed to prevent stimulation of the remainder of the oral cavity (Bartoshuk, 1975; Dubose, Meiselman, Hunt, & Waterman, 1977; McBurney, Kasschau, & Bogart, 1967; Meiselman, 1971; Meiselman, Bose, & Nykvist, 1972; Meiselman & Dubose, 1976; Smith, 1971). This technique prevents whole mouth stimulation and limits saliva from mixing with the tastant, which has been shown to affect receptor binding (Meiselman, 1971).
The slope and intercept of a psychophysical function characterize the relationship between perceived intensity of the stimulus and its concentration. The slope reflects the rate of growth in intensity as a function of concentration. Previous studies suggest that slopes of functions relating perceived taste intensity to concentration are dependent on the technique used to deliver the stimulus and the area of tongue stimulated (Meiselman, 1971; Bartoshuk, 1968; Lawless, 1987; Collings 1974). Specifically, psychophysical functions for intensity of taste stimuli differ between sip and spit and dorsal flow techniques (Meiselman, 1971).
Slopes of functions relating perceived intensity to concentration may differ as a function of concentration (Smith, 1971). Moreover, the slopes of functions relating perceived intensity to concentration may vary as a function of the type of stimulus delivery technique (Meiselman, 1971). Lastly, the relationship between concentration, perceived intensity, and stimulus delivery technique is dependent on the quality of the stimulus. Thus, differences in the psychophysical function of perceived intensity to concentration for taste stimuli between sip and spit and dorsal flow may be a result of the relationship between the stimulus quality, concentration, and stimulus delivery technique.
Functional Magnetic Resonance Imaging (fMRI) holds great promise for elucidating central mechanism of taste. In the context of fMRI experiments, the confines of being in the scanner impose a number of physical limitations that influence the stimulus delivery technique (e.g., swallowing while in the supine position and limiting head movement) resulting in stimulus delivery techniques that are significantly different than those used in behavioral experiments. However, stimulus concentrations used in fMRI experiments are often based on sip and spit and dorsal flow techniques. Expanding the current chemosensory stimulus delivery techniques to new methodologies [e.g., taste stimulation during fMRI] is important for the advancement of taste research. This requires applying valid and reliable psychophysical techniques such as sip and spit and dorsal flow in novel ways that will help meet the demands of new methodologies in taste research. Thus, it is important to understand how the confines of fMRI affect the relationship among stimulus quality, concentration and stimulus delivery technique within the context of fMRI.
The present experiment aimed to examine psychophysical functions using a taste stimulus delivery technique that simulated a delivery procedure that has been utilized within the confines of fMRI experiments (Haase, Cerf-Ducastel, Buracas, & Murphy, 2007; Haase, Cerf-Ducastel, & Murphy, 2009; Stice, Spoor, Bohon, & Small, 2008), termed the simulated stimulus delivery technique (SSDT). During the psychophysical testing in the current experiment, the participant was in an upright position. The SSDT involved the participant placing a plastic syringe on the anterior dorsal surface of the tongue, approximately 0.5cm away from the tip. The stimuli (presented 0.3ml in boluses of 50 μl) were allowed to flow to the lateral edges of the tongue and were swallowed. The aim was to simulate the arrival of the stimulus and the psychophysical experience. Other aspects of stimulus delivery in the scanning experiment (computer control of taste stimuli, liquid delivery inside the magnet, swallowing, head motion) were addressed elsewhere (Haase et al., 2007).
The present study compared psychophysical functions relating perceived intensity to concentration from previous experiments that employed the sip and spit and dorsal flow stimulus delivery techniques to the technique used in the current study, SSDT, to determine if the SSDT produce slopes for psychophysical functions that are similar to those generated in previous psychophysical experiments. We hypothesized that solutions delivered with SSDT would be perceived as less intense than solutions delivered with the sip and spit method, and that the sip and spit method would overestimate the slopes of intensity functions relating perceived intensity to concentration found with SSDT. We further hypothesized that the slopes of intensity functions related perceived intensity to concentration reported in previous dorsal flow literature would more accurately predict the slopes of intensity functions obtained with the SSDT. Another objective was to investigate the response function for various taste stimuli in order to identify specific concentrations for stimuli that would be optimal for future fMRI experiments.
Twenty healthy young adults, 10 females and 10 males, ranging in age from 18 – 29 years of age (M = 22.65, SD = 2.942) participated in the study after giving informed consent. The Institutional Review Board at San Diego State University approved the study for the participation of human subjects. Each subject was screened for ageusia and anosmia with taste threshold and odor threshold tests (Cain, Gent, Catalanotto, & Goodspeed, 1983, modified as in Murphy, Gilmore, Seery, Salmon, & Lasker, 1990) within one week of psychophysical testing. Additional exclusionary criteria consisted of upper respiratory infection or allergies within the prior two weeks.
Participants were presented with 27 aqueous solutions consisting of nine different stimuli each in three different concentrations and were asked to rate the intensity of each stimulus. To rate intensity, the participants used the Labeled Magnitude Scale (LMS, Green, Shaffer, & Gilmore, 1993; Green, Dalton, Cowart, Shaffer, Rankin, & Higgins, 1996, Bartoshuk, Duffy, Green, Hoffman, Ko, Lucchina et al., 2004). Intensity was compared to the “strongest imaginable” taste. The participants were asked to rate the intensity of each stimulus relative to all other tastes that they had experienced.
Stimuli were presented in random order, at room temperature as 0.3 ml, with a 3 cc/ml sterile plastic syringe. The amount of solution given to the subject was equal to the amount delivered in prior fMRI experiments during one stimulation period (Haase, et a., 2007; Cerf-Ducastel & Murphy, 2001). Before the stimulus presentation began, the participant rinsed the mouth thoroughly with distilled water for 10 s and expectorated into a cup. The participant was then instructed to place the plastic syringe, filled with 0.3 ml of stimulus, approximately 0.5 cm away from the tip of the tongue, allowing the stimulus to flow over and around the edges of the tongue and be swallowed. Between stimuli the participant rinsed the mouth with distilled water. Each stimulus was separated by a minimum of 30 s to minimize adaptation (Bartoshuk, McBurney, & Pfaffmann, 1964; McBurney et al., 1967).
The following stimuli were presented in aqueous solutions, sucrose: 0.16M, 0.32M, and 0.64M, saccharin: 0.007M, 0.01M, 0.028M, aspartame: 0.001M, 0.002M, and 0.004M, quinine: 0.001M, 0.0033M, and 0.01M, caffeine: 0.01M, 0.02M, and 0.04M, NaCl: 0.08M, 0.16M, and 0.32M, citric acid: 0.01M, 0.02M, and 0.04M, monosodium glutamate (MSG): 0.001M, 0.005M, and 0.025M, and guanosine 5′-monophosphate (GMP): 0.001M, 0.005M, and 0.025M.
The present experiment represents a 9X3X2 mixed design, with stimuli and concentration as within group variables and gender as a between group variable. A multivariate analysis of variance (MANOVA) was conducted with stimulus, concentration, and gender as the independent variables and perceived intensity scores as the dependent variable. One-sample t-tests were run to compare the slopes of perceived intensity in response to concentration found in the current study to the slopes of perceived intensity from previous literature where dorsal flow and sip and spit methods were employed (Table 1).
The comparisons in the current study were limited to those studies in the literature that 1) reported slopes of perceived intensity to concentration using magnitude estimation, 2) presented stimuli that were dissolved in water (not dissolved in flavored beverages or in food), and 3) presented pure tastants and not mixtures. In order to compare results from the present study with those in the published literature geometric means of the perceived intensity for each gustatory stimulus were graphed on log-log plots. To find the slope of the function relating perceived intensity to concentration, linear regressions were run for each stimulus by taking the logs of perceived intensity and plotting them against the logs of the stimulus concentration. This method was chosen so that all slopes could be compared on the same scale, for comparison to previous literature.
Psychophysical screening demonstrated that the participants had normal functioning for taste (M = .00884, SE = .0042) and olfactory (Left nostril, M = 7.1, SE = .347; Right nostril, M = 6.9, SE = 3.47) thresholds.
As expected, there was a main effect for stimuli, [F (8,144) = 35.42, p < .001]. A Newman-Keuls Multiple Range test demonstrated that MSG, GMP, and caffeine were perceived as less intense (p < .05) than saccharin, NaCl, citric acid, and quinine. Sucrose was perceived as significantly less intense (p < .05) than NaCl, citric acid, and quinine. Saccharin was significantly less intense (p < .05) than citric acid and quinine. Also, NaCl was perceived as being less intense (p < .05) than quinine. There was a main effect of concentration, F (2,36) = 38.85, p < .001. A Newman-Keuls Multiple Range test indicated that the first of three concentrations was perceived as less intense than the second and third (p < .05). Similarly, the second concentration was perceived as less intense than the third (p < .05). There was no main effect of gender, F (1) = .57, p = .46. There were no significant interactions for stimulus by concentration, F (16) = 1.63, p = .06, stimuli by gender, F (8) = 1.20, p = .31, or concentration by gender, F (2) = .64, p = .54.
Overall, comparison of slopes of intensity functions for: sucrose, NaCl, citric acid, saccharin, caffeine, and quinine in the present study indicate that using the SSDT resulted in slopes of intensity functions that were more similar to those found with the dorsal flow technique relative to the sip and spit technique (Table 1).
The slope of the psychophysical function relating perceived intensity to stimulus concentration for sucrose was not significantly different from slopes of perceived intensity in previous studies that employed the dorsal flow technique (Bartoshuk, 1975, 1977; Meiselman et al., 1972) and was not significantly different from one of two sip and spit experiments (Stevens, 1969) (Table 1, Figure 1). However, the slope for sucrose was significantly lower than the slopes for sucrose determined by sip and spit and reported by (Bartoshuk, 1977). The slope of perceived intensity for NaCl was not significantly different from slopes reported in the majority of dorsal flow studies (Smith, 1971; Bartoshuk, 1975). However, the slope of perceived intensity for NaCl found in the present study using SSDT was significantly smaller than the slope for perceived intensity determined by the dorsal flow method, reported by Meiselman (1972), and it was smaller than the slope reported in a previous study using the sip and spit technique (Stevens, 1969; Table 1, Figure 2). The slope for perceived intensity generated using the SSDT technique for citric acid was not significantly different from the slope of the intensity function for citric acid determined by dorsal flow and reported by Smith (1971); however, the slope for citric acid was significantly different from those reported by Bartoshuk (1977, dorsal flow, Table 1, Figure 3). The slope of the function for perceived intensity of saccharin in the present study was not significantly different from the slope reported by Smith (1971, dorsal flow), but was significantly different from the slope reported by Moskowitz (1970, sip and spit; Table 1, Figure 4). The slope of the intensity function for QHCI was significantly different from the slope determined by dorsal flow (Smith, 1971; Bartoshuk, 1975, 1977; Meiselman, 1972) and by sip and spit (Stevens, 1969) (Table 1, Figure 5). Additionally, the slope for caffeine was not significantly different from that reported by Bartoshuk (1977) for dorsal flow; Table 1, Figure 7).
The slope of the intensity function for GMP was not significantly different from the slope determined using the sip and spit method and reported by Rifkin & Bartoshuk (1980, Table 1). The slope for MSG was not significantly different from slopes reported in the literature determined by the sip and spit method (Schiffman & Clark, 1980; Rifkin & Bartoshuk, 1980; Table 1, Figure 7). To our knowledge, no information was available regarding the slope of the perceived intensity functions for MSG and GMP using the dorsal flow procedure.
The objective of the present study was to examine the differences in the slopes of the psychophysical functions relating perceived intensity to concentration for taste stimuli, generated by typical psychophysical stimulus delivery techniques and by a stimulus delivery technique that mimicked critical features of one utilized within the fMRI scanner (SSDT). The results demonstrated that slopes of psychophysical functions generated using the SSDT were more similar to those generated by dorsal flow stimulation than by sip and spit stimulation (Figures 1–7). This finding is not inconsistent with results previously reported by Meiselman (1971) that demonstrated differences in intensity functions generated with sip and spit and dorsal flow techniques.
The present findings are particularly important for the design and implementation of gustatory stimulus delivery techniques in the fMRI environment. As observed in the current study, employing a stimulus delivery technique adapted to the confines of fMRI may produce perceived intensity functions for taste stimuli that differ from perceived intensity functions generated using whole mouth sip and spit stimulation in typical behavioral experiments. The observed differences may be a result of differences in the amount of stimulus delivered in fMRI experiments. Because fMRI image quality can be compromised by significant head movement due to swallowing large quantities of liquid, the amount of stimulus delivered in the fMRI environment may be limited to amounts in the range of 0.3 ml per bolus, while in sip and spit experiments, the amount of liquid in each sample is more typically 5–10 ml. Previous research has demonstrated that the area of the tongue stimulated affects taste perception (Miller, 1986; Zuniga, Davis, Englehardt, Miller, Schiffman, Phillips, 1993). Further, differences in perceived intensity are associated with number of taste buds that are stimulated (Miller & Reedy, 1990; Zungia et al., 1993) and taste bud density varies, increasing as much as five-fold from the anterior region to the midregion of the tongue (Miller, 1986). Collings (1974) reported that the slope of perceived intensity varied with the stimulus and the area of stimulation. For example, slopes for quinine were found to be steeper on the vallate papillae than on the fungiform papillae (Collings, 1974). Thus the area and the region of the tongue and oral cavity stimulated are important factors in psychophysical response (Miller, 1988), and to the extent that restricting the amount of stimulus delivered affects the regions of the oral cavity and tongue that are stimulated, this may be an important consideration in the fMRI environment.
In the current study, saccharin and quinine had high perceived intensities with slopes for intensity functions that were low relative to those previously reported in the literature for sip and spit and dorsal flow methods. The slopes of functions relating perceived intensity to concentration for quinine and saccharin may have been influenced by a number of factors: saturation, the high concentrations used, or limitations in the range of concentrations. Quinine is a bitter stimulus and saccharin is an artificial sweetener; however, at high concentrations saccharin is known to produce a bitter taste (Bartoshuk, 1979; Schiffman, Gatlin, Sattely-Miller, Graham, Heiman, Stagner, & Erickson, 1994; Schiffman, Booth, Losee, Pecore, & Warwick, 1995). Specifically, the bitterness function for saccharin grows at a lower rate than the perceived intensity function for sucrose then decreases and plateaus as concentration increases (Moskowitz 1970). Moskowitz concluded that this results from an increase in bitterness and a decrease in sweetness associated with high concentrations of saccharin. Thus, in the current study, the concentration series used for saccharin may have been high enough to produce substantial bitterness resulting in a shallow slope relating perceived intensity to concentration, similar to quinine.
For the present purposes, we limited comparisons of the current data to studies that used magnitude estimation (Marks, 1974; Stevens, 1969) for rating intensity (Bartoshuk, 1975; Bartoshuk & Cleveland, 1977; Meiselman et al., 1972; Moskowitz, 1970; Rifkin & Bartoshuk, 1980; Smith, 1971; Stevens, 1969), rather than including studies that used other scaling methods, such as category scales. The psychophysical functions produced using the LMS have been shown to be equivalent to magnitude estimation in response to oral sensation (taste, temperature and pain; Green et al., 1993) and gustatory and olfactory stimuli (Green et al., 1996). Future studies might address whether data from other scaling methods, such as category scales, produce similar results.
Finally, fMRI studies of gustatory function require presentation of the stimuli within a scanning environment where other aspects of stimulus delivery and behavioral performance (computer control of taste stimuli in event-related designs, liquid delivery inside the magnet, swallowing while supine, head motion, producing and recording on-line psychophysical assessments during fMRI scans) are important. We, and others, have addressed these issues elsewhere and the interested reader is referred to those studies for further information (Haase et al., 2007; Haase et al., 2009; Stice et al., 2008).
In conclusion, the present study found that the simulated scanning technique produced psychophysical functions with slopes that were generally lower than those reported in the literature from experiments conducted with the sip and spit technique and that were similar to slopes of intensity functions reported in the literature for experiments conducted with the dorsal flow procedure. Choosing stimuli and stimulus concentrations for taste fMRI studies that are based on behavioral data collected with a stimulus amount and stimulation technique that mimics the stimulus delivery used inside the scanner may be particularly useful. The unique requirements of stimulus delivery within the scanning environment require an understanding of the inter-dependent associations among stimulus quality, concentration, stimulus amount, and stimulus delivery technique, to ensure that perceptions of the stimuli can be appropriately interpreted. More broadly, the present study highlights the importance of precise psychophysical experimentation in conjunction with fMRI studies in order to facilitate interpretation of investigation of the associations between brain activation, perception, and cognition.
This study was funded by NIH grant AG04085 to Claire Murphy. We thank Paul Gilbert, Ph.D. for comments on the manuscript, Aaron Jacobson and Craig Mellinger for their assistance in data collection.
Lori Haase, Department of Psychology, San Diego State University, San Diego, CA. University of California San Diego-San Diego State University Joint Doctoral Program in Clinical Psychology, San Diego, CA.
Barbara Cerf-Ducastel, Department of Psychology, San Diego State University, San Diego, CA.
Claire Murphy, Department of Psychology, San Diego State University, San Diego, CA. Department of Head and Neck Surgery, University of California San Diego School of Medicine, San Diego, CA. University of California San Diego-San Diego State University Joint Doctoral Program in Clinical Psychology, San Diego, CA.