Neuroscience 295 (2015) 72–79
OLFACTORY SPEED – TEMPORAL ODOR PROCESSING OF PAIRED STIMULI V. A. SCHRIEVER, a,b* C. FRENZEL, a S. WERNECKE, a I. CROY, a,c C. VALDER d AND T. HUMMEL a
INTRODUCTION Our sense of smell is often unappreciated and underestimated. Its true value therefore is commonly not noticed unless a loss or reduction of olfactory function is present (Croy et al., 2014b). This has many reasons. Our attention is rarely focused on olfactory sensation and even when asked to, it is sometimes difficult to direct the attention toward our sense of smell (Sela and Sobel, 2010). Nevertheless, odors have an influence on our life. It has been shown that even not consciously perceived odors influence our behavior (Gelstein et al., 2011). But when it comes to actively including sensory perception in decision making we rely mostly on our other sensory modalities like audition and especially vision (Morrot et al., 2001). This is strengthened by the fact, that when being presented with visual or olfactory stimuli only 3.5% of visual but about 10% of olfactory stimuli are missed (Spence et al., 2001). Not only the accuracy of detection but also the speed of stimulus recognition and perception differs between the sensory modalities. Compared to other sensory modalities, reaction times toward olfactory stimuli are rather long, ranging around 600–1200 ms (Cain, 1976). Several studies have been conducted to study the temporal perception of odor components within an odor mixture (Jinks et al., 1998; Jinks and Laing, 1999). Depending on the component property and concentration, it is therefore necessary to present odors with an interval of more than 600 ms in order to be perceived separately (Jinks and Laing, 1999). Decreasing interstimulus interval (ISI) leads to odors being perceived as mixtures (Jinks and Laing, 1999). Due to peripheral adaptation as well as central habituation processes, the time of regeneration after odorous stimulation is relatively long (Zufall and Leinders-Zufall, 2000). Therefore studies targeting olfaction frequently use long ISI of 20 s or even more to minimize the influence of adaptation and habituation. The effect of short and very short ISIs on the perception and detection of olfactory stimuli of the same odor indicated that an ISI of approximately 2.5 s in a stimulus pair of identical olfactory stimuli is necessary to perceive them as separate stimuli (Wang et al., 2002; Croy et al., 2014a). Nevertheless not much is known about the factors, which influence the time necessary for two odors being perceived separately. Aim of the current study therefore was to study the ISI length between two repetitive stimuli on the decision whether there were two stimuli or one. The following study included three sub-studies, which were designed to examine the influence of the ISI,
a
Smell & Taste Clinic, Department of Otorhinolaryngology, University of Dresden Medical School, Dresden, Germany b Department of Neuropediatrics, University of Dresden Medical School, Dresden, Germany c Department of Clinical Neurophysiology, Sahlgrenska University Hospital, University of Gothenburg, Sweden d
Frey and Lau, Henstedt-Ulzburg, Germany
Abstract—Objectives: Compared to other senses, temporal perception of odors seems fairly slow. In addition it has been shown in previous studies that even not consciously perceived odors could influence our behavior. Aim of the current study therefore was to study the interstimulus interval (ISI) length, which is necessary between two repetitive stimuli to be able to perceive them separately. The additional aim focused on observing central odor processing of not perceived odorous stimuli. Materials and methods: The study was divided into three parts. In each part healthy, normosmic volunteers were included. In part I and II stimulus pairs (CO2, H2S, orange and phenyl ethyl alcohol (PEA)) were presented to the subjects via a computer-controlled olfactometer with short ISI of 0.6–9 s. The decision whether one or two stimuli were perceived was recorded. In addition the influence of odor valence, trigeminallity and concentrations was observed. In part III olfactory event-related potentials (OERPs) to perceived and not-perceived odors were recorded. Results: The two stimuli of a stimulus pair were perceived separately more often with increasing ISI length. This increase was significant until an ISI between the stimuli of 4 s. Odor intensity, pleasantness, trigeminallity and sex had no major influence on this. In addition we were able to observe that OERPs are less often detected in response to not perceived olfactory stimuli. However, the presence of OERP in response to not perceived stimuli in more than half of the cases indicated that even not perceived stimuli are centrally processed. Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved.
Key words: temporal, odor perception, event-related potentials. *Correspondence to: V. A. Schriever, Smell & Taste Clinic, Department of Otorhinolaryngology, University of Dresden Medical School, Fetscherstrasse 74, 01307 Dresden, Germany. Tel: +49-351-458-4189. E-mail addresses: [email protected] (V. A. Schriever), [email protected] (C. Frenzel), [email protected] (S. Wernecke), [email protected] (I. Croy), [email protected] (C. Valder), [email protected] (T. Hummel). Both authors contributed equally to this paper. Abbreviations: ISI, interstimulus interval; LP, late positivity; OERPs, olfactory event-related potentials; PEA, phenylethylalcohol. http://dx.doi.org/10.1016/j.neuroscience.2015.03.029 0306-4522/Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved. 72
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
pleasantness, trigeminallity and odor concentration on the perception of the stimuli. The measurements obtained from these studies will be subjective and rely on the compliance of the participants. To obtain additional information about odor processing with very short ISI between stimulus presentations, a fourth sub-study was designed. Olfactory event-related potentials (OERPs) are commonly used to observe central odor processing. Therefore an electrophysiological approach was taken to observe if and to what extent the second stimulus in a stimulus pair will be centrally processed when being not perceived by the participant.
EXPERIMENTAL PROCEDURES All aspects of the study were performed in accordance to the Declaration of Helsinki. The local Ethics Committee of the Medical Faculty of the Technical University of Dresden approved the study protocol (EK332092011 and EK434112013). The test procedure was explained to the participants. Written informed consent was obtained from all participants. The present investigation consisted of three parts. General setup and odor presentation are described for all parts together being followed by specific details about each part separately. Setup To test the time interval needed between two olfactory or trigeminal stimuli of the same kind to being perceived separately, the following setup was applied. The stimuli were presented in pairs with different ISI. Participants were notified with a written statement on a computer screen to pay attention shortly before (between 2 and 4 s) the stimulus presentation. After each presentation of a stimulus participants were asked whether they perceived one or two stimuli. All responses were recorded using a self-written computer program running on Adobe Flash Player (Adobe System, San Jose, CA, USA). The answer was considered as a correct response, when the two stimuli in a stimulus pair were perceived separately and therefore the participants answered that they perceived two stimuli.
73
by headphones with white noise of approximately 60 dB to prevent responses to the clicking of the valves associated with stimulus presentation. The stimuli were presented in pairs with different ISI as described above. The inter-trial interval in between stimulus pairs was randomized between 28 and 32 s. Part Ia: Effect of ISI, odor and trigeminallity on temporal perception
Participants. Forty participants (20 women, 20 men) were included in this part. The mean age of the participants was 24.0 ± 2.5 years (range 20–29 years). There was no significant difference between the age of women and men (t = 0.82, p = 0.42). All participants were normosmic as indicated by screening with the ‘‘Sniffin’ Sticks’’ 16-item odor identification test (Hummel et al., 1997; Kobal et al., 2000). The mean odor identification score was 14.4 ± 0.96 points (range 12–16 points). Test procedure. The following stimuli were used in this part: PEA, H2S and CO2. The ISI ranged between 2000 and 9000 ms in steps of 1000 ms. Each condition was presented five times in a randomized order. Therefore 40 stimulus pairs were presented for PEA, H2S and CO2, resulting in 120 stimulus pairs in total. For reasons of practicability they were divided into six blocks, each consisting of 20 stimulus pairs. In addition to the question presented after each stimulus pair, participants were instructed to press the left button of the computer mouse after every stimulus as soon as they perceived a stimulus. The testing was divided into two sessions on separate days due to the length of the study. After ascertaining the participants’ normosmia, intensity and pleasantness ratings for the used stimuli PEA, CO2 and H2S were obtained on a visual analog scale (0–100 and 50 to 50 respectively). In addition two of the six blocks for the main experiment were conducted in this session. The remaining four blocks were performed in session two. The duration of each session was approximately 90 min. Part Ib: Effect of very short ISI on temporal perception
Stimulus presentation As olfactory stimuli phenyl ethyl alcohol (PEA; from Sigma, Deisenhofen, Germany) (50%v/v), H2S (from air liquide, Kornwestheim, Germany) (5 ppm), Orange (from Frey & Lau, Henstedt-Ulzburg, Germany) (5 and 50%v/v) and as a trigeminal stimulus CO2 (from air liquide, Kornwestheim, Germany) (35%v/v) were used. All stimuli were presented using the olfactometer OM6b (Burghart, Wedel, Germany). The olfactometer provided a continuous airflow of 7 l/min with heated and humidified air (36.5 °C, 80% relative humidity) and the stimulus duration for all stimuli was set to 400 ms. Stimuli were presented monorhinally and the nasal side of presentation was changed between blocks of stimulation (see below). The participants were acoustically shielded
Participants. Fifteen participants (8w/7m), mean age 25.5 ± 3.6 years (range 21–35 years) were included. All participants were normosmic as indicated by the ‘‘Sniffin’ Sticks’’ 16-item odor identification test (Hummel et al., 1997; Kobal et al., 2000). Participants scored on average 14.07 ± 1.15 points (range 12–16 points). Test procedure. The same stimuli as in part Ia, PEA, H2S and CO2, were used. The following ISI were used in this part: 600, 800, 1200, 1600 and 2000 ms. Again each condition was presented 5 times resulting in a total of 75 stimulus pairs, which were presented in three blocks (15 stimulus pairs each) in one session.
74
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
Part II: Effect of odor concentration on temporal odor perception Participants. Forty participants (22w/18m), mean age 24.1 ± 3.3 years (range 19–35 years) were included. Normosmia was ascertained by the ‘‘Sniffin’ Sticks’’ 16item odor identification test (Hummel et al., 1997; Kobal et al., 2000). On average participants scored 14.18 ± 0.93 points (range 12–16 points). Test procedure. Two concentrations of orange, 5 and 50%v/v, were used as an olfactory stimulus in this part. The ISI were set to: 800, 1200, 1600, 2000, 3000, 4000, 5000 and 6000 ms. Every condition was presented 5 times leading to 80 stimulus pairs in total. The testing took place in two sessions on separate days. In line with part Ia, the stimuli were divided into six blocks, with 20 stimulus pairs in each block. Lateralization task. To be able to compare the results of this part with the results obtained with the olfactory stimuli PEA and H2S, it was necessary to ensure that the orange stimuli produced little or no activation of the trigeminal system. A lateralization task was performed which is based in the assumption that only trigeminally active stimulants can be localized (e.g. (Kobal et al., 1989; Frasnelli et al., 2011)). The participants were presented with two stimuli in squeeze bottles, one for each nostril – one bottle containing the odor (50% orange), the other containing the solvent. Participants were asked to localize the side of odor presentation. A total of 20 stimuli were presented randomized to either the left or the right nostril. Part III: Central processing of perceived and notperceived odorous stimuli Participants. Twenty participants (11w/9m) mean age 22.8 ± 2.4 years (range 19–29 years) from part II were included in this part. Test procedure. ERPs were measured using orange (50%v/v) as an olfactory stimulus. The ISI was chosen according to results from part II at which participants reached approximately 50% of correct answers. Therefore either 2 or 3 s was used as an ISI. A total of 80 stimulus pairs were presented divided into two blocks, with a duration of approximately 22 min each. OERPs were recorded from five electrodes (Fz, Cz, Pz, C3 and C4) according to the international 10–20 system (Jasper, 1958) with linked earlobes (A1–A2) as reference. The data were acquired using a 16-channel amplifier (SIR: Ro¨ttenbach, Germany). EEG-segments of 2048 ms, starting 500 ms before stimulus onset, were recorded at a frequency of 250 Hz using a band-pass filter of 0.2–30 Hz. EEG-data analysis. All EEG analyzing steps were processed using the Matlab (MathWorks, Natick, MA, USA) toolbox Letswave 5 (Mouraux, Brussels, Belgium, http://nocions.webnode.com/letswave). In addition to the
online filter an offline band-pass filter (FFT) of 0.3– 15 Hz was applied. The 500 ms pre-stimulus interval was used for applying a baseline correction. EEG recordings, which contained artifacts, were manually removed, e.g. eye blinks exceeding 50/50 lV at electrode Fp2. Artifact free recordings were averaged for each condition. On the level of individual participants latencies and peaks for N1 and the late positivity (LP) were measured. The largest negative peak between 200 and 700 ms was considered as N1 and the LP peak was measured between 300 and 800 ms (Hummel et al., 2000). The OERPs of the second stimulus were divided into two categories ‘‘perceived’’ and ‘‘not perceived’’ according to the participants’ answer to the question following each odor presentation. Statistical analyses Data were analyzed by means of SPSS 22.0 (SPSS Inc., Chicago, IL, USA). An Analysis of Variance was performed using generalized linear mixed models for part I, II and III. ISI length, odor valence, trigeminallity, stimulus intensity and sex were used in part I and II to observe the decision about the perception of paired stimuli. In part III the factors first stimulus, second stimulus and perceived, not perceived stimuli were used to analyze the effect on the OERP components. T-tests were used for post hoc comparison. Multiple pairwise comparisons were corrected after Bonferroni. The level of significance was set at 0.05.
RESULTS Part Ia and b: Effect of ISI, odor and trigeminallity on temporal perception Intensity ratings for all three stimulants did not differ significantly (F = 0.88, p = 0.36). On the visual analog scale from 0 to 100 average ratings for PEA were 52.8 ± 18.9, for H2S 55.0 ± 19.2 and for CO2 50.9 ± 17.6. In contrast to that, pleasantness ratings on a visual analog scale from 50 to 50 for the stimuli were significantly different (F = 68.6, p < 0.001). PEA was rated as slightly pleasant (3.5 ± 17.1), whereas CO2 ( 15.2 ± 12.8) and H2S ( 26.0 ± 16.7) were rated as unpleasant – values above 0 indicated pleasantness, values below 0 indicated unpleasantness. When the responses to all stimuli were analyzed together a steady increase of correct answers was observed with increasing lengths of the ISI. Meaning that the two stimuli of the stimulus pair were perceived more of separately with increasing lengths of the ISI. At an ISI of 600 ms only 14.7 ± 22.7% of the stimulus pairs were perceived as two separate stimuli. This rate increased to 76.5 ± 26.7% at an ISI of 9000 ms. An analysis revealed a main effect of ISI length on the percentage of correct answers (F = 66.3, p < 0.001) (Fig. 1a). Pairwise Bonferroni-corrected comparisons showed significant increases from intervals between 600 and 4000 ms (3.24 < t < 19.58, 0.029 < p < 0.001). After 4000 ms, the percentage of correct answers still
75
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
Fig. 1. Paired stimulus perception with increasing interstimulus interval. The figure displays the correct answers after pairwise stimulus presentation in relation to the length of the ISI, meaning that the two stimuli were perceived separately. (a) Stimuli are perceived more often as separate at longer ISI. This is uniform for PEA (red dotted line), H2S (orange dashed line) and CO2 (blue continuous line). (b) The increase of perceiving the two stimuli of a stimulus pair separately is displayed. The increase is significant until an ISI of 4000 ms, no significant increase was observed with longer ISI – reaching a plateau. (c) The influence of odor concentration on paired odor stimulus perception is shown. With the higher orange odor concentration (dashed line) stimuli of the pair were perceived more often separately compared to the lower orange odor concentration (continuous line). This was true for overall perception. No difference in correct answers was found at each ISI. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
increased slightly but not significantly reaching a plateau (Fig. 1b). In addition to this, a main effect of stimulus type (PEA, CO2, H2S) was observed (F = 4.78, p = 0.009) with H2S stimulus pairs being perceived more often as separate compared to the other two stimulus types (overall average answer: H2S: 71.2%, PEA = 65.4%, CO2 = 65.4%). Pairwise comparisons (Bonferroni corrected) showed: H2S vs. PEA (t = 2.67, p = 0.023) and H2S vs. CO2 (t = 2.66, p = 0.023). This was only true for the average when all ISI were included. An analysis of each ISI only showed a significance for the very short ISI of 600 ms between the three stimulus types (F = 5.93, p = 0.007) with PEA pairs being perceived as separate stimuli more often than CO2 (t = 2.82, p = 0.014) and H2S (t = 2.48, p = 0.027) respectively. No main effect or interaction of sex between stimulus type and ISI could be observed. In line with the results presented above there was a main effect of ISI when analysis was conducted on a single stimulus type: PEA (F = 9.71, p < 0.001), CO2 (F = 24.8, p < 0.001) and H2S (F = 31.9, p < 0.001). For the two olfactory stimuli, PEA and H2S an increase of correctly perceiving a pair of stimuli as separate stimuli increased up to 3000 ms when compared to an ISI of 9000 ms (PEA all p < 0.001, H2S all p between
OLFACTORY SPEED – TEMPORAL ODOR PROCESSING OF PAIRED STIMULI V. A. SCHRIEVER, a,b* C. FRENZEL, a S. WERNECKE, a I. CROY, a,c C. VALDER d AND T. HUMMEL a
INTRODUCTION Our sense of smell is often unappreciated and underestimated. Its true value therefore is commonly not noticed unless a loss or reduction of olfactory function is present (Croy et al., 2014b). This has many reasons. Our attention is rarely focused on olfactory sensation and even when asked to, it is sometimes difficult to direct the attention toward our sense of smell (Sela and Sobel, 2010). Nevertheless, odors have an influence on our life. It has been shown that even not consciously perceived odors influence our behavior (Gelstein et al., 2011). But when it comes to actively including sensory perception in decision making we rely mostly on our other sensory modalities like audition and especially vision (Morrot et al., 2001). This is strengthened by the fact, that when being presented with visual or olfactory stimuli only 3.5% of visual but about 10% of olfactory stimuli are missed (Spence et al., 2001). Not only the accuracy of detection but also the speed of stimulus recognition and perception differs between the sensory modalities. Compared to other sensory modalities, reaction times toward olfactory stimuli are rather long, ranging around 600–1200 ms (Cain, 1976). Several studies have been conducted to study the temporal perception of odor components within an odor mixture (Jinks et al., 1998; Jinks and Laing, 1999). Depending on the component property and concentration, it is therefore necessary to present odors with an interval of more than 600 ms in order to be perceived separately (Jinks and Laing, 1999). Decreasing interstimulus interval (ISI) leads to odors being perceived as mixtures (Jinks and Laing, 1999). Due to peripheral adaptation as well as central habituation processes, the time of regeneration after odorous stimulation is relatively long (Zufall and Leinders-Zufall, 2000). Therefore studies targeting olfaction frequently use long ISI of 20 s or even more to minimize the influence of adaptation and habituation. The effect of short and very short ISIs on the perception and detection of olfactory stimuli of the same odor indicated that an ISI of approximately 2.5 s in a stimulus pair of identical olfactory stimuli is necessary to perceive them as separate stimuli (Wang et al., 2002; Croy et al., 2014a). Nevertheless not much is known about the factors, which influence the time necessary for two odors being perceived separately. Aim of the current study therefore was to study the ISI length between two repetitive stimuli on the decision whether there were two stimuli or one. The following study included three sub-studies, which were designed to examine the influence of the ISI,
a
Smell & Taste Clinic, Department of Otorhinolaryngology, University of Dresden Medical School, Dresden, Germany b Department of Neuropediatrics, University of Dresden Medical School, Dresden, Germany c Department of Clinical Neurophysiology, Sahlgrenska University Hospital, University of Gothenburg, Sweden d
Frey and Lau, Henstedt-Ulzburg, Germany
Abstract—Objectives: Compared to other senses, temporal perception of odors seems fairly slow. In addition it has been shown in previous studies that even not consciously perceived odors could influence our behavior. Aim of the current study therefore was to study the interstimulus interval (ISI) length, which is necessary between two repetitive stimuli to be able to perceive them separately. The additional aim focused on observing central odor processing of not perceived odorous stimuli. Materials and methods: The study was divided into three parts. In each part healthy, normosmic volunteers were included. In part I and II stimulus pairs (CO2, H2S, orange and phenyl ethyl alcohol (PEA)) were presented to the subjects via a computer-controlled olfactometer with short ISI of 0.6–9 s. The decision whether one or two stimuli were perceived was recorded. In addition the influence of odor valence, trigeminallity and concentrations was observed. In part III olfactory event-related potentials (OERPs) to perceived and not-perceived odors were recorded. Results: The two stimuli of a stimulus pair were perceived separately more often with increasing ISI length. This increase was significant until an ISI between the stimuli of 4 s. Odor intensity, pleasantness, trigeminallity and sex had no major influence on this. In addition we were able to observe that OERPs are less often detected in response to not perceived olfactory stimuli. However, the presence of OERP in response to not perceived stimuli in more than half of the cases indicated that even not perceived stimuli are centrally processed. Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved.
Key words: temporal, odor perception, event-related potentials. *Correspondence to: V. A. Schriever, Smell & Taste Clinic, Department of Otorhinolaryngology, University of Dresden Medical School, Fetscherstrasse 74, 01307 Dresden, Germany. Tel: +49-351-458-4189. E-mail addresses: [email protected] (V. A. Schriever), [email protected] (C. Frenzel), [email protected] (S. Wernecke), [email protected] (I. Croy), [email protected] (C. Valder), [email protected] (T. Hummel). Both authors contributed equally to this paper. Abbreviations: ISI, interstimulus interval; LP, late positivity; OERPs, olfactory event-related potentials; PEA, phenylethylalcohol. http://dx.doi.org/10.1016/j.neuroscience.2015.03.029 0306-4522/Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved. 72
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
pleasantness, trigeminallity and odor concentration on the perception of the stimuli. The measurements obtained from these studies will be subjective and rely on the compliance of the participants. To obtain additional information about odor processing with very short ISI between stimulus presentations, a fourth sub-study was designed. Olfactory event-related potentials (OERPs) are commonly used to observe central odor processing. Therefore an electrophysiological approach was taken to observe if and to what extent the second stimulus in a stimulus pair will be centrally processed when being not perceived by the participant.
EXPERIMENTAL PROCEDURES All aspects of the study were performed in accordance to the Declaration of Helsinki. The local Ethics Committee of the Medical Faculty of the Technical University of Dresden approved the study protocol (EK332092011 and EK434112013). The test procedure was explained to the participants. Written informed consent was obtained from all participants. The present investigation consisted of three parts. General setup and odor presentation are described for all parts together being followed by specific details about each part separately. Setup To test the time interval needed between two olfactory or trigeminal stimuli of the same kind to being perceived separately, the following setup was applied. The stimuli were presented in pairs with different ISI. Participants were notified with a written statement on a computer screen to pay attention shortly before (between 2 and 4 s) the stimulus presentation. After each presentation of a stimulus participants were asked whether they perceived one or two stimuli. All responses were recorded using a self-written computer program running on Adobe Flash Player (Adobe System, San Jose, CA, USA). The answer was considered as a correct response, when the two stimuli in a stimulus pair were perceived separately and therefore the participants answered that they perceived two stimuli.
73
by headphones with white noise of approximately 60 dB to prevent responses to the clicking of the valves associated with stimulus presentation. The stimuli were presented in pairs with different ISI as described above. The inter-trial interval in between stimulus pairs was randomized between 28 and 32 s. Part Ia: Effect of ISI, odor and trigeminallity on temporal perception
Participants. Forty participants (20 women, 20 men) were included in this part. The mean age of the participants was 24.0 ± 2.5 years (range 20–29 years). There was no significant difference between the age of women and men (t = 0.82, p = 0.42). All participants were normosmic as indicated by screening with the ‘‘Sniffin’ Sticks’’ 16-item odor identification test (Hummel et al., 1997; Kobal et al., 2000). The mean odor identification score was 14.4 ± 0.96 points (range 12–16 points). Test procedure. The following stimuli were used in this part: PEA, H2S and CO2. The ISI ranged between 2000 and 9000 ms in steps of 1000 ms. Each condition was presented five times in a randomized order. Therefore 40 stimulus pairs were presented for PEA, H2S and CO2, resulting in 120 stimulus pairs in total. For reasons of practicability they were divided into six blocks, each consisting of 20 stimulus pairs. In addition to the question presented after each stimulus pair, participants were instructed to press the left button of the computer mouse after every stimulus as soon as they perceived a stimulus. The testing was divided into two sessions on separate days due to the length of the study. After ascertaining the participants’ normosmia, intensity and pleasantness ratings for the used stimuli PEA, CO2 and H2S were obtained on a visual analog scale (0–100 and 50 to 50 respectively). In addition two of the six blocks for the main experiment were conducted in this session. The remaining four blocks were performed in session two. The duration of each session was approximately 90 min. Part Ib: Effect of very short ISI on temporal perception
Stimulus presentation As olfactory stimuli phenyl ethyl alcohol (PEA; from Sigma, Deisenhofen, Germany) (50%v/v), H2S (from air liquide, Kornwestheim, Germany) (5 ppm), Orange (from Frey & Lau, Henstedt-Ulzburg, Germany) (5 and 50%v/v) and as a trigeminal stimulus CO2 (from air liquide, Kornwestheim, Germany) (35%v/v) were used. All stimuli were presented using the olfactometer OM6b (Burghart, Wedel, Germany). The olfactometer provided a continuous airflow of 7 l/min with heated and humidified air (36.5 °C, 80% relative humidity) and the stimulus duration for all stimuli was set to 400 ms. Stimuli were presented monorhinally and the nasal side of presentation was changed between blocks of stimulation (see below). The participants were acoustically shielded
Participants. Fifteen participants (8w/7m), mean age 25.5 ± 3.6 years (range 21–35 years) were included. All participants were normosmic as indicated by the ‘‘Sniffin’ Sticks’’ 16-item odor identification test (Hummel et al., 1997; Kobal et al., 2000). Participants scored on average 14.07 ± 1.15 points (range 12–16 points). Test procedure. The same stimuli as in part Ia, PEA, H2S and CO2, were used. The following ISI were used in this part: 600, 800, 1200, 1600 and 2000 ms. Again each condition was presented 5 times resulting in a total of 75 stimulus pairs, which were presented in three blocks (15 stimulus pairs each) in one session.
74
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
Part II: Effect of odor concentration on temporal odor perception Participants. Forty participants (22w/18m), mean age 24.1 ± 3.3 years (range 19–35 years) were included. Normosmia was ascertained by the ‘‘Sniffin’ Sticks’’ 16item odor identification test (Hummel et al., 1997; Kobal et al., 2000). On average participants scored 14.18 ± 0.93 points (range 12–16 points). Test procedure. Two concentrations of orange, 5 and 50%v/v, were used as an olfactory stimulus in this part. The ISI were set to: 800, 1200, 1600, 2000, 3000, 4000, 5000 and 6000 ms. Every condition was presented 5 times leading to 80 stimulus pairs in total. The testing took place in two sessions on separate days. In line with part Ia, the stimuli were divided into six blocks, with 20 stimulus pairs in each block. Lateralization task. To be able to compare the results of this part with the results obtained with the olfactory stimuli PEA and H2S, it was necessary to ensure that the orange stimuli produced little or no activation of the trigeminal system. A lateralization task was performed which is based in the assumption that only trigeminally active stimulants can be localized (e.g. (Kobal et al., 1989; Frasnelli et al., 2011)). The participants were presented with two stimuli in squeeze bottles, one for each nostril – one bottle containing the odor (50% orange), the other containing the solvent. Participants were asked to localize the side of odor presentation. A total of 20 stimuli were presented randomized to either the left or the right nostril. Part III: Central processing of perceived and notperceived odorous stimuli Participants. Twenty participants (11w/9m) mean age 22.8 ± 2.4 years (range 19–29 years) from part II were included in this part. Test procedure. ERPs were measured using orange (50%v/v) as an olfactory stimulus. The ISI was chosen according to results from part II at which participants reached approximately 50% of correct answers. Therefore either 2 or 3 s was used as an ISI. A total of 80 stimulus pairs were presented divided into two blocks, with a duration of approximately 22 min each. OERPs were recorded from five electrodes (Fz, Cz, Pz, C3 and C4) according to the international 10–20 system (Jasper, 1958) with linked earlobes (A1–A2) as reference. The data were acquired using a 16-channel amplifier (SIR: Ro¨ttenbach, Germany). EEG-segments of 2048 ms, starting 500 ms before stimulus onset, were recorded at a frequency of 250 Hz using a band-pass filter of 0.2–30 Hz. EEG-data analysis. All EEG analyzing steps were processed using the Matlab (MathWorks, Natick, MA, USA) toolbox Letswave 5 (Mouraux, Brussels, Belgium, http://nocions.webnode.com/letswave). In addition to the
online filter an offline band-pass filter (FFT) of 0.3– 15 Hz was applied. The 500 ms pre-stimulus interval was used for applying a baseline correction. EEG recordings, which contained artifacts, were manually removed, e.g. eye blinks exceeding 50/50 lV at electrode Fp2. Artifact free recordings were averaged for each condition. On the level of individual participants latencies and peaks for N1 and the late positivity (LP) were measured. The largest negative peak between 200 and 700 ms was considered as N1 and the LP peak was measured between 300 and 800 ms (Hummel et al., 2000). The OERPs of the second stimulus were divided into two categories ‘‘perceived’’ and ‘‘not perceived’’ according to the participants’ answer to the question following each odor presentation. Statistical analyses Data were analyzed by means of SPSS 22.0 (SPSS Inc., Chicago, IL, USA). An Analysis of Variance was performed using generalized linear mixed models for part I, II and III. ISI length, odor valence, trigeminallity, stimulus intensity and sex were used in part I and II to observe the decision about the perception of paired stimuli. In part III the factors first stimulus, second stimulus and perceived, not perceived stimuli were used to analyze the effect on the OERP components. T-tests were used for post hoc comparison. Multiple pairwise comparisons were corrected after Bonferroni. The level of significance was set at 0.05.
RESULTS Part Ia and b: Effect of ISI, odor and trigeminallity on temporal perception Intensity ratings for all three stimulants did not differ significantly (F = 0.88, p = 0.36). On the visual analog scale from 0 to 100 average ratings for PEA were 52.8 ± 18.9, for H2S 55.0 ± 19.2 and for CO2 50.9 ± 17.6. In contrast to that, pleasantness ratings on a visual analog scale from 50 to 50 for the stimuli were significantly different (F = 68.6, p < 0.001). PEA was rated as slightly pleasant (3.5 ± 17.1), whereas CO2 ( 15.2 ± 12.8) and H2S ( 26.0 ± 16.7) were rated as unpleasant – values above 0 indicated pleasantness, values below 0 indicated unpleasantness. When the responses to all stimuli were analyzed together a steady increase of correct answers was observed with increasing lengths of the ISI. Meaning that the two stimuli of the stimulus pair were perceived more of separately with increasing lengths of the ISI. At an ISI of 600 ms only 14.7 ± 22.7% of the stimulus pairs were perceived as two separate stimuli. This rate increased to 76.5 ± 26.7% at an ISI of 9000 ms. An analysis revealed a main effect of ISI length on the percentage of correct answers (F = 66.3, p < 0.001) (Fig. 1a). Pairwise Bonferroni-corrected comparisons showed significant increases from intervals between 600 and 4000 ms (3.24 < t < 19.58, 0.029 < p < 0.001). After 4000 ms, the percentage of correct answers still
75
V. A. Schriever et al. / Neuroscience 295 (2015) 72–79
Fig. 1. Paired stimulus perception with increasing interstimulus interval. The figure displays the correct answers after pairwise stimulus presentation in relation to the length of the ISI, meaning that the two stimuli were perceived separately. (a) Stimuli are perceived more often as separate at longer ISI. This is uniform for PEA (red dotted line), H2S (orange dashed line) and CO2 (blue continuous line). (b) The increase of perceiving the two stimuli of a stimulus pair separately is displayed. The increase is significant until an ISI of 4000 ms, no significant increase was observed with longer ISI – reaching a plateau. (c) The influence of odor concentration on paired odor stimulus perception is shown. With the higher orange odor concentration (dashed line) stimuli of the pair were perceived more often separately compared to the lower orange odor concentration (continuous line). This was true for overall perception. No difference in correct answers was found at each ISI. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
increased slightly but not significantly reaching a plateau (Fig. 1b). In addition to this, a main effect of stimulus type (PEA, CO2, H2S) was observed (F = 4.78, p = 0.009) with H2S stimulus pairs being perceived more often as separate compared to the other two stimulus types (overall average answer: H2S: 71.2%, PEA = 65.4%, CO2 = 65.4%). Pairwise comparisons (Bonferroni corrected) showed: H2S vs. PEA (t = 2.67, p = 0.023) and H2S vs. CO2 (t = 2.66, p = 0.023). This was only true for the average when all ISI were included. An analysis of each ISI only showed a significance for the very short ISI of 600 ms between the three stimulus types (F = 5.93, p = 0.007) with PEA pairs being perceived as separate stimuli more often than CO2 (t = 2.82, p = 0.014) and H2S (t = 2.48, p = 0.027) respectively. No main effect or interaction of sex between stimulus type and ISI could be observed. In line with the results presented above there was a main effect of ISI when analysis was conducted on a single stimulus type: PEA (F = 9.71, p < 0.001), CO2 (F = 24.8, p < 0.001) and H2S (F = 31.9, p < 0.001). For the two olfactory stimuli, PEA and H2S an increase of correctly perceiving a pair of stimuli as separate stimuli increased up to 3000 ms when compared to an ISI of 9000 ms (PEA all p < 0.001, H2S all p between
