Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/springer-journals/mothers-pupillary-responses-to-infant-facial-expressions-GX4JNZbT2N
- Publisher
- Springer Journals
- Copyright
- Copyright © 2017 by The Author(s)
- Subject
- Biomedicine; Neurosciences; Neurology; Behavioral Therapy; Psychiatry
- eISSN
- 1744-9081
- DOI
- 10.1186/s12993-017-0120-9
- pmid
- 28166792
- Publisher site
- See Article on Publisher Site
Abstract
Background: Human parental care relies heavily on the ability to monitor and respond to a child’s affective states. The current study examined pupil diameter as a potential physiological index of mothers’ affective response to infant facial expressions. Methods: Pupillary time-series were measured from 86 mothers of young infants in response to an array of photo- graphic infant faces falling into four emotive categories based on valence (positive vs. negative) and arousal (mild vs. strong). Results: Pupil dilation was highly sensitive to the valence of facial expressions, being larger for negative vs. posi- tive facial expressions. A separate control experiment with luminance-matched non-face stimuli indicated that the valence effect was specific to facial expressions and cannot be explained by luminance confounds. Pupil response was not sensitive to the arousal level of facial expressions. Conclusions: The results show the feasibility of using pupil diameter as a marker of mothers’ affective responses to ecologically valid infant stimuli and point to a particularly prompt maternal response to infant distress cues. Keywords: Pupil, Emotion, Facial expressions, Attention, Mothers, Infant faces Background To begin examining whether pupil diameter is a sen- Parental care and parent-infant interaction relies heavily sitive index of parents’ physiological responses to chil- on the ability to receive and express nonverbal emotional dren’s affective cues, the present study focused on signals through facial expressions [1]. There is increasing mothers’ responses to infant facial expressions. Despite interest in the neurocognitive bases of these capacities in increasing involvement of fathers in childcare in many parents [2–4] and infants [5], and in the possibility that societies and the need for research on biological bases of subtle variations in emotional signaling may have impor- paternal childcare, there are known intersex differences tant influences on the quality of parent–child attach - in the neural and hormonal bases for caregiving behav- ment [6]. In the current study, we extend these studies iors [3]. For this reason, we limited our current investi- by examining whether mothers’ pupil dilation is sensitive gation on parental responsiveness to infant emotion to to children’s affective cues and could, in future, serve as mothers rather than sampling parents of both sexes. The an accessible marker of interindividual variation in these neural and physiological basis of mothering constitutes a responses. If the pupil is sensitive to mothers’ affective distinct domain of research [7], with potentially impor- responses, such as heightened vigilance towards infant tant implications for maternal and infant mental health. signals of discomfort and distress [1], this index may Previous studies show differential patterns of brain and prove useful in studying and understanding the mecha- behavioral responses elicited by infant as opposed to nisms underlying maternal sensitivity or neglect. adult faces [8–11], especially in women [11]. Further, mothers, compared to nulliparous women, show more marked early frontal (∼100 ms) event-related potentials *Correspondence: santeri.yrttiaho@uta.fi 1 to infant facial expressions, as well as more pronounced Tampere Center for Child Health Research, School of Medicine, University of Tampere, Lääkärinkatu 1, 33520 Tampere, Finland modulation of posterior visual responses to infant faces Full list of author information is available at the end of the article © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 2 of 12 displaying negative emotions [2, 12]. The current study in variable background illumination, and faces of people extends the research on neurophysiological processes with variable ethnicity or hair color. For this reason, an of mothering by examining the pupillary correlates of important challenge for pupillometric studies is to exam- mothers’ responses to infants’ positive and negative facial ine ways to disentangle unavoidable pupillary responses expressions [12, 13]. to luminance changes from those involved in affective Pupil size is largely determined by reflexive control processing. over light entering the eye [14]. However, pupil diam- The current study consisted of two experiments where eter is also influenced by the activity of the sympathetic pupil response was measured in response to infant facial autonomic nervous system (ANS) during emotional stimuli. In Experiment 1, a large sample of mothers of arousal [15, 16]. The pupillary response to a visual stim - young infants (N = 86) was recruited in order to exam- ulus can typically be characterized by two consecutive ine whether pupil constriction and dilation are sensitive phases. First, in response to increased brightness, there physiological measures of mothers’ responses to mild and is a constriction in pupil size around 600–1600 ms after strong instances of infant positive and negative affect. stimulus onset [15]. Following the constriction, the pupil Based on prior studies [15–20], we predicted relatively starts to dilate back to a baseline level over the course larger pupil diameter in response to high-arousal facial of several seconds. For example, an initial constriction expressions (i.e., high intensity positive and negative of the pupil in response to visual stimuli is followed by expressions) in both the constriction and dilation phases. a slow dilation that is augmented for emotionally posi- Our secondary aim was to disentangle affective pupil - tive and negative scenes [15, 17]. Pupil constriction and lary responses to facial expressions from the response dilation per se are brought about by distinct branches of to unavoidable variations in stimulus luminance. To this the nervous system, parasympathetic and sympathetic, end, we carried out Experiment 2, a control experiment, innervating the constrictor and dilator muscles, respec- where participants were presented both with faces (face tively [16]. Pupil size at any time reflects the tone of both condition) and luminance-matched non-face stimuli of these muscles. Therefore, both the constriction and (control condition). We hypothesized that an emotive the dilation phase are susceptible to emotional effects pupil response, a greater pupil diameter during pupil elicited by emotionally arousing scenes and facial expres- dilation and constriction phases, would be manifested in sions [15, 17–19]. Pupil response to emotional factors is the face condition but not in the control condition. thought to be mediated by the modulatory effects of the brain’s noradrenergic system on neural circuitry control- Experiment 1 ling the muscles of the iris [20, 21]. Importantly, larger Methods pupil dilation to emotional stimuli cannot be suppressed Participants voluntarily [22], making it an accessible marker for stud- The participants were mothers of young infants par - ies examining human affective responses in a variety of ticipating in an ongoing longitudinal study examin- contexts. ing the mental health of mother-infant dyads in the Because the dominant source of variability in pupil size Cape Town metropolitan area, South Africa. The par - comes simply from changes in stimulus luminance [20], ticipants had no mental health disorder as assessed pupillometry studies have traditionally required stringent by a psychiatrist through clinical assessment and the control of luminance levels across experimental stimuli Mini International Neuropsychiatric Interview [23]. or, at minimum, equalization of mean luminance across The eye-tracking assessment in mothers was conducted different stimulus categories. While the luminance of as a part of a scheduled immunization visit to a private many types of visual stimuli can, in principle, be easily well-baby clinic at the infant age of 6 weeks. All moth- adjusted to equal mean level, such equalization is not via- ers recruited to this study and tested by the eye-tracking ble for all studies and small deviations from the mean lev- procedure by September 30th, 2015, were included in els will be unavoidable. This poses particular challenges the current analyses. The final sample consisted of 86 for studies examining pupillary responses to ecologi- mothers (Age: M = 32.4 years, SD = 5.4 years) of Cau- cally valid stimuli such as infant facial expressions. The casian (N = 53) and Black (N = 33) ethnicity. The Cau - only available method to capture preverbal infants’ facial casian participants had higher socioeconomic status expressions is to photograph them as they occur sponta- (SES) than the Black participants as indexed by monthly 2 2 neously in variable environments where strict control of income [Χ (3) = 57.91, p < .001], level of education [Χ luminance is difficult or may result in unnatural quality (2) = 66.32, p < .001] and employment [Χ (3) = 66.89, of the stimuli. Corresponding problems exists in other p < .001]. The number of pregnancies of the participants contexts such as studies using facial stimuli presented ranged from 1 to 5 with 68.7% of participants having in live face-to-face or over-the-internet conditions, faces more than one pregnancy. Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 3 of 12 Stimuli (happy), 2 = neutral, and 3 = negative (sad) emotion. The infant face stimuli were obtained from an existing The rating scores varied according to the intended emo - stimulus set [13]. These 36 grayscale photographs were tional category (i.e., SP, MP, MN, SN) of the stimuli [F(1, close-ups of infants with black uniform backgrounds, 24) = 2405.73, p < .001, partial η = .99] in a subset of with infants producing four types of facial expressions randomly selected 25 participants. The rating scores (Fig. 1): mild positive (MP), strong positive (SP), mild increased monotonically between SP, MP, and MN negative (MN), and strong negative (SN). It is noteworthy (ps < .001) and reached a plateau at MN vs. SN (p = .37). that this classification is based on previous work show - In effect, SP was rated as positive (M = 1.0, SD = .04), ing that infants’ emotional expressions during the first MP close to neutral (M = 1.8, SD = .1), and both MN year of life are not readily categorized into discrete emo- and SN as negative (Ms > 2.9, SDs < .2). tion categories, but instead, to primary dimensions of In order to reduce the pupillary light reflex [15] elic - hedonic pleasure and arousal [13]. Nine instances of each ited by the face stimuli, a visual non-face pattern was category were used in the experiment. The selection was generated to be shown during the “pre-stimulus” interval made to obtain sets of images that were matched for face- before each face stimulus. This visual stimulus was pro - background ratio, models’ head orientation, and models’ duced by randomly permuting and then averaging the age as closely as possible. As the original pool of images pixels derived from the entire set of face stimuli in order was taken from separate sources, some variations in to match its grayscale intensity to that of the faces. above characteristics remained in each stimulus category. The infant’s eyes were open in all positively-valenced (MP Procedure and SP) photographs but were closed in 67% of images The participants sat in a dimly lit room with no external with strong (SN) and in 22% of images with mild negative light other than that laminating from the screen. Before (MN) facial expressions. The infants had open eyes with data acquisition, eye-tracking calibration was performed direct gaze in 33, 56, 11, and in 33% of pictures classified for each participant by requiring the participant to fixate as SP, MP, SN, and MN, respectively. All infants depicted on five targets. In instances where the gaze was not found in the stimuli were Caucasian. at these targets, the calibration-procedure was repeated. The current infant stimuli were selected from an The calibration results from a subgroup of participants original pool of 208 images consistently rated (>75%) as are shown in Additional file 1: Figure S2. reflecting one of the four emotions (i.e., MP, SP, MN, or After calibration, participants were presented with a SN) by human judges [12, 13]. However, the emotional series of trials each consisting of (1) a short 1000-ms fore- content of the current face stimuli has been previously period with a black screen, (2) a 2000-ms pre-stimulus validated [12, 13] in a population which is geographically interval with a random visual pattern, (3) a 5000-ms face (Italy vs. South Africa), ethnically, and socioeconomi- stimulus, and (4) a white rectangular border surrounding cally different from the current participants. There - the face for 1000 ms to signal the end of the trial. A short fore, behavioral rating scores were obtained to ensure sound signal, a “notification beep”, was presented 2500 ms that the participants agreed with the emotional valence before each face stimulus. Illustration of the stimulus and arousal previously attributed to the stimuli. Partici- sequence is shown in Fig. 2. The face stimulus for a given pants were asked to judge the emotional valence of the trial was selected randomly, without replacement, from face stimuli on a scale from 1 to 3, where 1 = positive the set of 36 faces (9/category as explained above). The Fig. 1 Examples of infant face stimuli from each stimulus category defined by the intensity and valence of the facial expression of emotion. Rand- omized pixels (on the right), derived from all face stimuli were used as a pre-stimulus display. In the control experiment (Experiment 2) randomized face pixels, derived from each face stimulus individually, were used in place of the face stimuli in the “non-face” condition Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 4 of 12 Fig. 2 Trial structure in the experimental paradigm used to examine pupillary responses to infant facial expressions. Trial events con- sisted of (1) a black screen (duration = 1000 ms, a sound alarm was presented 500 ms prior to the next visual stimulus), (2) pre-stimulus display consisting of randomized pixels from the face images with stimulus-matched luminance (duration = 2000 ms), (3) face stimulus (duration = 5000 ms), and (4) a white border added to the face stimu- lus to signal the end of the trial. Pupil data were analyzed from the period starting from the pre-stimulus interval (displaying randomized pixels) and ending 5 s after the onset of the face stimulus Fig. 3 Single-trial pupil response. Pupil diameter as a function of time is shown from both eyes separately (blue = left, green = right) and from data combined across eyes (red). Valid data frames are indicated by cyan points. Pupil size during invalid data frames were stimuli were presented on a black background. The partic - replaced by interpolation in the pupil size combined across eyes (red). ipants were asked to simply view each stimulus presented The first 2 s of each trial displayed a random visual pattern followed by a face stimulus (2–7 s). Pupil dilation was extracted as the mean without a specific requirement for a response. To collect pupil size during the time-window ranging from 4000 to 5000 ms subjective ratings, the experimenter presented the face after face onset (6000–7000 ms after trial onset, yellow background). pictures to participants as paper print-outs and collected Pupil constriction was extracted as the minimum pupil size from time verbal responses from participants onto separate sheets window spanning 300–1200 ms after face onset (blue background). after the entire sequence of pupil data acquisition. A baseline for pupil size was calculated from −300 to 0 ms relative to face onset (gray background) Acquisition and analysis of pupil diameter data Pupil size was measured with a Tobii X-60 or X2-60 were reached at 300–1200 and 4000–5000 ms after face eye-tracker camera which measures corneal reflection onset, respectively (Fig. 4). These time intervals were of infrared light relative to the image of the pupil. The thus selected for the extraction of pupil constriction acquisition was controlled by custom-written MATLAB scripts, Psychtoolbox, and the Talk2Tobii toolbox, inter- facing with a Tobii (Danderdyn, Sweden) eye-tracker. Pupil size was measured from both eyes during the presentation of the face stimuli as well as during the pre-stimulus interval (with random visual pattern). The pupil data was acquired in conjunction with synchronous point-of-gaze (POG), eye-tracking validity (i.e., “valid” or “invalid”), and stimulus timing data at a sampling rate of 60 Hz. The data on pupil size was preprocessed using gazeAnalysisLib [24] to complete the following steps: averaging the diameter across the left and the right eye, replacing “invalid” frames of pupil size by the means of linear interpolation, median filtering, and baseline cor - rection. The POG data was also combined across eyes and median filtered. The window size of the median filter was 7 frames (ca. 120 ms) for both POG and pupil data. Fig. 4 Grand average pupil response. Pupil time-series were base- line-corrected to the 300-ms interval preceding the face stimulus. The A typical pupil response from a single trial is shown in pupil response is shown for mild negative (MN) and positive (MP) as Fig. 3. well as for strong negative (SN) and positive (SP) stimulus conditions. Pupil response was baseline-corrected by subtract- Face onset is indicated by the vertical line and the time-windows ing the pupil size from a 300-ms pre-face time window. for pupil constriction and dilation by blue and yellow background, Based on visual inspection of the grand average pupil respectively. The shaded are around curves indicate standard error of the mean response, the minimum and maximum pupil diameters Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 5 of 12 and dilation, respectively. While pupil constriction Table 1 Eye-tracking quality was determined as the minimum pupil size during the Constriction Dilation 300–1200 ms time window, pupil dilation was extracted Mean SD Mean SD as the mean pupil size during the latter 4000–5000 ms time-window. Valid eye-tracking (%) 90.5 17.6 87.0 19.8 Quality control (QC) of trial-by-trial pupil data was Longest non-valid streak based on the following factors: (1) participant maintain- Baseline (ms) 28.5 61.0 27.9 60.5 ing gaze within the face stimulus, (2) error-free pupil/ Response (ms) 75.1 141.7 120.3 176.9 eye-tracking, and (3) absence of outlier values. Quantita- Inside AOI, valid frames (%) 96.1 21.6 95.4 23.5 tive indices of the participant gaze coordinates and error- free pupil tracking were acquired on a frame-by-frame basis together with pupil diameter. Using these metrics, t-tests. The emotive pupil response, in turn, was defined trials where participant’s gaze was directed at the loca- as a change in pupil size between conditions differing tion of the face stimulus less than 10% of time, either in in emotive valence or arousal of the stimuli. Emotion- the baseline or in the response interval, were rejected related differences in baseline-corrected pupil diameter from analysis. Similarly, trials with excess of eye/pupil- across stimulus conditions and response time-windows tracking frames labeled as “invalid” by the acquisition were investigated with a Time-window (2) × Arousal software were discarded from the analyses. These trials (2) × Valence (2) repeated-measures Analysis of Vari- were defined as those where the longest streaks of con - ance (ANOVA). The factors comprised constriction vs. secutive non-valid frames exceeded 250 ms within the dilation phase (Time-window), strong vs. mild emotion baseline period, or 900 ms within the time-window used (Arousal), and negative vs. positive emotion (Valence), for extracting the pupil constriction or dilation response. respectively. Both main an interaction effects were ana - In addition to these basic QC measures, we identified lyzed. In case interaction effects were found in the initial the first trial within each measurement as a systematic ANOVA, further comparisons within pupil constric- source of outlier data, and rejected these trials from tion and dilation data, separately, were conducted with further analyses as well (see Additional file 1 for further Arousal × Valence ANOVAs. Any subsequent pairwise details about the rejection criteria used in the study). On tests were conducted with pair-wise t-tests. Effect sizes the average 6.2 (SD = 6.1) trials were rejected (out of 35 are reported using Cohen’s d for t-tests and partial η for available) from each participant (averaged across con- ANOVA throughout the results. striction and dilation). Two participants had less than 3 averaged trials available for the analysis of the effects of Results Arousal and Valence on pupil response, and were thus Following previous research [15–20], the presentation rejected from final statistical analyses. of faces was expected to elicit a typical visually induced The quality of the trials accepted for further analyses pupil response consisting of an initial pupil constriction was then inspected by analyzing the average values of (decrease in pupil diameter Ø) and a subsequent dilation frame-by-frame pupil/eye-tracking data within these tri- (increase in pupil diameter Ø). Further, as an increase in als. The targeted metrics included the average percentage pupil size has been documented for pictures (not always of valid frames for both eyes (during the response time- including faces) with strong vs. mild emotive arousal [15, window), the duration of non-valid streaks during the 17, 18], the pupil size was hypothesized to be larger both baseline and response time-windows, and the percent- during the constriction and the dilation phase for strong age of gaze directed at the location of the face (minimum vs. mild expressions. The corresponding effects were percentage across baseline and response time-windows). investigated also in response to negative vs. positive emo- These statistics are shown in Table 1. In summary, mean tional valence of the facial expressions. valid eye-tracking reached 87–91%, and on the aver- An ample pupil constriction, that is, a decrease in age, longest-non valid data streaks were shorter than pupil diameter during 300–1200 ms after stimulus onset 76–121 ms. (Fig. 4) against baseline level, was found in response to all face stimuli [|ΔØ| > .11 mm, |ts| > 13.18, ps < .001, Statistical analyses |ds| > 1.43]. Pupil constriction was followed by a subse- In order to determine whether significant pupil con - quent pupil dilation (Fig. 4) and the pupil size increased striction and dilation were elicited by the current stim- significantly above the pre-stimulus baseline during the uli, baseline-corrected pupil diameter was contrasted latter 4000–5000 ms time window [.05 < ΔØ < .19 mm, to the zero-level (i.e., no response) with one sample ts > 3.96, ps < .001, .43 < |ds| < 1.44]. Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 6 of 12 Main effects of Time-window [constriction vs . dilation; and written informed consent was obtained from all F(1, 83) = 532.40, p < .001, partial η = .87], and Valence participants. [F(1, 83) = 157.03, p < .001, partial η = .65] were found on baseline-corrected pupil size. These effects were due Stimuli to greater pupil size during dilation (ΔØ = .12 mm) vs. The stimuli in the original pool of infant facial expres - constriction (ΔØ = −.16 mm) and greater pupil size in sions were matched for luminance [12], but as pupil response to negatively-valenced vs. positively-valenced diameter is highly sensitive to stimulus luminance, a fur- stimuli. Interaction effects on pupil size were found ther analysis of brightness was conducted for the pur- between Arousal and Valence [F(1, 83) = 7.76, p < .01], poses of the present study. Optical luminance data were Time-window and Valence [F(1, 83) = 15.03, p < .001], unavailable (requires careful photometric measurements as well as Time-window and Arousal [F(1, 83) = 13.33, with appropriate equipment), but the possible differ - p < .001]. Therefore, the effects of Arousal and Valence ences in luminance were inspected from mean grayscale on pupil size were inspected separately for pupil constric- intensity (0–255) values of the bitmap files. A one-way tion and dilation. ANOVA showed that, overall, there were no statistically Main effects of Valence [F(1, 84) = 83.60, p < .001, par- significant differences between stimulus categories in tial η = .50] and Arousal [F(1, 84) = 14.22, p < .001, par- grayscale intensity values, [F(4,32) = 1.92, p = .13, partial 2 2 tial η = .15] were found on pupil constriction. Contrary η = .19], but inspection of the bitmap intensity for indi- to the hypothesis of increased pupil size in response to vidual images and direct pairwise comparisons showed stimulus arousal, strong stimulus arousal was related to noticeable variation within and across stimulus catego- a decreased pupil size (i.e., increased constriction) within ries in intensity (Fig. 5). Furthermore, while the random this early time window [strong < mild; ΔØ = −.03 mm]. visual pattern presented before each face stimulus had Moreover, there was an interaction between Valence and grayscale intensity equal to the mean intensity across Arousal on pupil constriction [F(1, 84) = 11.64, p < .001], all faces (horizontal line in Fig. 2), it differed somewhat reflecting a significant effect of arousal for the positively- in intensity from each individual face stimulus. Because valenced stimuli [ΔØ = −.05 mm, t(84) = −4.77, p < .001, such differences in stimulus luminance might bias the d = −.52] but not for the negatively-valenced stimuli emotional effects on pupil response, we generated an [ΔØ = −.01 mm, t(84) = −.99, p = .33, d = −.11]. While array of control stimuli by randomly scrambling the the hypothesized emotion-related increase in pupil bitmap matrices of the face images. The resultant con - diameter during constriction was not found for stimu- trol stimuli, thus, had exactly the same mean grayscale lus arousal, decreased pupil constriction (i.e., greater intensity as the face stimuli. These non-face stimuli were pupil diameter), was found for faces with negative vs. presented to a group of new participants (N = 15) who positive emotional valence across both levels of arousal [ΔØs > .05 mm, ts(84) > 6.10, ps < .001, |d|s > .66]. A main effect of Valence [F(1, 83) = 129.31, p < .001, par- tial η = .61] was found on pupil dilation due to a .12-mm increase in pupil size to negatively-valenced (ΔØ = .17) vs. positively-valenced infant facial expressions (ΔØ = .06). Contrary to the hypothesized arousal-related pupil dilation, no effect of arousal on pupil dilation was found [F(1, 83) = 1.41, p = .24]. No interaction between Arousal and Valence [F(1, 83) = 2.54, p = .12] was found on pupil dilation either. Experiment 2 Methods Participants The participants in Experiment 2 consisted of 15 volunteers (9 female, Age: M = 28.9 years, Fig. 5 Grayscale (bitmap) intensity values of the stimuli. Pair-wise range = 24–50 years). None of the participants in Experi- comparisons between means of different stimulus categories indicate ment 2 participated in Experiment 1 and parenthood was higher intensity (and, hence, luminance) for face stimuli with “strong not required for the inclusion of participants into Experi- positive” vs. “negative” emotional expressions. SN strong negative, ment 2. The study was ethically approved by the insti - MN mild negative, SP strong positive, MP mild positive. Horizontal line indicates the mean grayscale intensity across all stimuli tutional review board of the University of Stellenbosch Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 7 of 12 also viewed the original face pictures used in Experi- to qualify the source of the interaction effect. Effect sizes ment 1. The same visual pattern, consisting of randomly are reported as partial η . permuted pixels, was presented during the pre-stimulus interval as in Experiment 1. Results Pupil constriction was elicited across all conditions Procedure (|ΔØ| > .08 mm, |ts| > 4.19, ps < .001, |ds| > 1.08) The stimulation paradigm was identical to that in Experi - including both those with face stimuli and those with ment 1 (Fig. 2). In particular, the sequence and timing of the non-face stimuli (randomly permuted pixels). events within experimental trials was similar across the However, significant pupil dilation above pre-stimu - experiments. However, an additional experimental condi- lus baseline was observed only in the face condition tion was included where the face stimuli were replaced by (ΔØ > .10 mm, 2.24 < ts < 5.16, ps < .05, .58 < ds < 1.33). their pixel-scrambled counterparts. That is, in addition Such pupil dilation was invariantly absent in the non- to face stimuli, random non-face patterns were presented face condition (ΔØ ≤ .06 mm, −1.24 < ts < 2.01, ps > .06, in a separate experimental block preceding or follow- −.32 < ds < .52). ing the face sequence in a counterbalanced order across Main effects of Time-window [F(1, 14) = 80.57, participants. p < .001, partial η = .85], Pixel randomization [F(1, 14) = 13.05, p < .01, partial η = .48], and Valence [F(1, Acquisition and analysis of pupil diameter data 14) = 51.22, p < .001, partial η = .79] were found on The same equipment, software, and parameters were pupil size across all conditions within Experiment 2. used in the acquisition of the pupil data as in Experi- However, interaction effects on pupil size were found ment 1. The pupil constriction and dilation were likewise between Time-window and Pixel randomization [F(1, extracted from the same time-windows spanning 300– 14) = 10.29, p < .01] and between Valence, Time-window, 1200 and 4000–5000 ms after face onset. An average of and Pixel randomization [F(1, 14) = 7.26, p < .05]. There - 6.0 (SD = 6.5) and 5.5 (SD = 7.6) trials were rejected per fore, the effects of Valence on pupil size were inspected participant in the face and the control (pixel-scrambled separately from the two different time-windows (con - face) condition, respectively. However, all participants striction and dilation), and further, separately within the had a sufficient number of averaged trials (≥3) for the face and the non-face condition. The purpose of these statistical analyses of the effects of Arousal and Valence. further analyses was to determine whether the effect of u Th s data from all participants were included in the final Valence on pupil size differed between the face condition, statistical analyses from Experiment 2. with genuine emotional signals, and the non-face condi- tion, without emotional content. No main or interaction Statistical analyses effects of Emotional arousal on pupil size were found. Pupil change from baseline (zero-level) in all stimulus A main effect of emotional Valence was found on pupil conditions separately was first analyzed with one-sample constriction [F(1, 14) = 41.93, p < .001, partial η = .75]. t-tests. Then, emotion-related differences in baseline- A trend for interaction between Valence and Pixel rand- corrected pupil diameter across stimulus conditions omization was further found on pupil constriction [F(1, and response time-windows were compared with a 14) = 3.07, p < .11]. However, the effect of Valence was Pixel randomization (2) × Time-window (2) × Arousal found both within the face condition [F(1, 14) = 12.77, (2) × Valence (2) repeated-measures ANOVA. The p < .001, partial η = .48] and within the non-face condi- repeated-measures factors comprised intact faces vs. tion [F(1, 14) = 35.10, p < .001, partial η = .72] in subse- scrambled non-faces (Pixel randomization), constric- quent tests. As the effect of Valence on pupil constriction tion vs. dilation phase (Time-window), strong vs. mild was found both in the face and in the non-face control emotion (Arousal), and negative vs. positive emotion condition, it cannot be considered to indicate an emotive (Valence), respectively. Both main an interaction effects pupil response (the non-face stimuli consisted of mean- were analyzed. In case interaction effects were found in ingless random patterns). the initial ANOVA, further comparisons within pupil A main effect of emotional Valence was also found constriction and dilation data, separately, were con- on pupil dilation [F(1, 14) = 17.85, p < .001, partial ducted with Pixel randomization × Arousal × Valence η = .56]. Furthermore, a trend level interaction [F(1, ANOVAs. Finally, if an interaction was found between 14) = 3.80, p < .08] was found between Valence and Pixel emotional pupil response (i.e., Arousal or Valence) and randomization on pupil dilation (Fig. 6). This interaction Pixel randomization, Arousal × Valence ANOVAs were was qualified by a significant effect of Valence on pupil conducted separately for the face and the-noise condition dilation in the face condition [F(1, 14) = 23.30, p < .001, Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 8 of 12 further replicated in comparison against a non-face con- trol condition in Experiment 2, a separate control experi- ment. The control experiment further showed that the pupil dilation response triggered by infant distress was dissociable from a response to the brightness of stimuli in this category. In child and adolescent participants, pupil dilation has been previously reported in response to face stimuli with direct as opposed to averted gaze, especially when depicting happy facial expressions [18]. In contrast, greater pupil dilation in response to angry vs. happy or fearful facial expressions has been reported in adult par- ticipants [31, 32]. Furthermore, the sensitivity of the pupil response to direct vs. averted gaze found in typically developing participants was absent in children diagnosed Fig. 6 Pupil diameter during constriction (lower half) and dilation with autism spectrum disorders [18]. Therefore, the sen - phase (upper half) in response to negative and positive facial expres- sitivity of the pupil response to facial expressions of emo- sion of emotion and to luminance-matched non-face stimuli (scram- tion seems to depend on participant population, stimulus bled pixels). Greater pupil dilation was elicited by negatively-valenced than by positively-valenced faces. No such difference in dilation was material, or on their combination [18, 31, 32]. Conse- found across non-face stimuli. While the effect of valence on pupil quently, previous literature on emotive pupil response is size was found also during the constriction phase, this effect was not insufficient to describe the perception and physiological unique to face stimuli. Data from the control experiment (Experiment responsiveness to infant facial expressions in mothers or 2, N = 15). Error bars indicate standard errors of the mean. ***p < .001, to describe how such processes are reflected in the pupil n.s. not significant response. The current study (Experiment 1) is the first pupillometric study using specifically infant facial expres - sions of emotion as stimuli and mothers of young infants partial η = .63] but not in the non-face condition [F(1, as participants. The current results characterize the pupil 14) = 2.78, p < .12, partial η = .17]. In the face condi- response in this particular context and indicate increased tion, greater pupil dilation was found for negatively- autonomic responsiveness in response to infant signals of valenced (ΔØ = .23 mm) than for positively-valenced discomfort and distress. (ΔØ = .11 mm) infant face stimuli (difference = .12 mm). Adaptive infant-caregiver interaction rests on the abil- As the non-face stimuli, consisting of random pixels, ity of the interactants to receive and express emotional may be less motivating for participants to attend to than signals through facial expressions [1]. Mutually positive, faces, it was necessary to ensure that the participants fix - optimally arousing social interaction involves the regula- ated equally on both stimulus types. Equal participant tion of the activity of the autonomic nervous system as attentiveness and data quality in face and non-face con- a component of emotion regulation [25] and correlates dition was indicted by invariable number of acceptable of maternal sensitivity have been found in both sympa- trials across conditions [F(1, 14) = .14, p = .72, partial thetic and parasympathetic activity. Activation of the η = .01]. Therefore, the difference between face and sympathetic nervous system is associated with emo- non-face condition reflect true effects of face and emo - tional arousal and has been previously indicated in moth- tion processing as verified against conceivable data qual - ers’ response to infant cry as indexed by electrodermal ity issues. measures [26, 27]. Despite wide psychophysiological application [20, 28–30], relatively few studies have used Discussion pupillometry to investigate emotional processes evoked Pupillary response, in particular pupil dilation, has been by the perception of facial expressions [18, 31, 32]. Such proposed as an indicator of variable psychophysiological recent studies, using adult faces as stimuli, have indicated states [14, 15, 20]. In the current study, we investigated greater pupil dilation in response to angry vs. happy or whether the emotive pupil response could be used to fearful facial expressions in adult participants [31, 32]. index mothers’ responsiveness to infant non-verbal com- In the current study, we investigated specifically whether munication. To this end, in Experiment 1, we measured the emotional responsiveness in mothers to infant facial pupillary responses in mothers while they viewed infant expressions might be indexed with pupillometry. Our facial expressions. Larger pupil dilation was evoked by results were consistent with the previous studies on infant signals of distress or discomfort than by positively- pupillary responses to facial expressions of emotion by valenced facial expressions. Emotive pupil dilation was Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 9 of 12 indicating an increase in baseline-corrected pupil size specific subcategory of social cognition related to face- in response to emotional face stimuli. Importantly, this to-face interaction and caregiving behaviors. effect was now replicated in the special case of infant Given inter-individual variability in the accuracy to stimuli viewed by mothers. interpret infant facial expressions [13], we used a behav- While autonomic responsiveness may be a prerequisite ioral rating task to assess recognition of infant emotional for adequate mother-infant interaction, overactive sym- signals in the current participants. The results from the pathetic arousal to infant or child cues has been linked rating task indicated high accuracy in the recognition with harsh parenting [26], lower maternal sensitivity of infant facial expressions. In our pupillary analyses, a [27], negative appraisal of children [33], and child abuse distinction between stimuli rated as indicating negative [34]. Also parasympathetic activity reflected in the res - emotion produced a pupil dilation which was larger than piratory sinus arrhythmia, in both baseline level [35] and that elicited by stimuli rated as positive or neutral. Thus, in the regulation of the vagal tone [36], has been linked the pupil response was associated with the subjective to maternal sensitivity. Moreover, the sympathetic and identification of negative vs. positive affect in the infant parasympathetic systems may act in concert in determin- pictures. In future studies, a comparable approach com- ing emotional response in mothers to infant crying and bining pupillometry and behavioral performance could distress [27]. In the current study, emotive response to be used in studies involving specific participant groups infant faces was established in pupil dilation which has with variable social cognitive abilities especially related been associated with sympathetic activity [16]. In this to infant signals of emotion (e.g., from families at-risk for light, the current emotive pupil response is analogous maladaptive infant-caregiver interaction). to the increased skin conductance in mothers elicited by Previous research has demonstrated an effect of emo - sounds of infant cry, which is also attributed to sympa- tional arousal rather than that of valence on pupil dila- thetic arousal [26, 27]. This interpretation is further sup - tion [15]. In contrast, in the current study we found an ported by covariance between emotive pupil dilation and effect of emotional valence on pupil dilation but no effect skin conductance in response to affective pictures [15]. of stimulus arousal. The difference between the findings However, the pupil size during dilation may reflect the may be related to the type of stimuli (face vs. IAPS, not level of parasympathetic activity as well [16]. Based on limited to faces), the type of people depicted (infants vs. findings from autonomic responses to infant cry [26, 27], IAPS, not limited to infants), and the scale used in stimu- the pupil response might index either sufficient or exces - lus classification. Perceptual scaling of any stimulus is sive autonomic arousal to infant negative affect for the inherently arbitrary and heavily influenced by the refer - maintenance of adaptive maternal sensitivity. In future ence stimulus or stimuli [40, 41]. In scaling emotional studies, mapping the maternal pupil response to favora- valence and arousal different sets of stimuli, and hence ble level of autonomic responsiveness to infant cues different reference(s), may have been used in the current might be achieved by relating the response to indices of stimulus set in comparison to IAPS pictures [42]. Thus, caregiver behaviors and maternal sensitivity. the arousal and valence categories used here may be dif- The norepinephrine attentional system of the brain ferent from those used in the IAPS. It further seems pos- originating in the locus coeruleus (i.e., the LC-NE sys- sible that the negatively-valenced stimuli in the current tem) has been suggested to underlie emotional pupil study depicting infant distress or discomfort may signal dilation [20, 37, 38]. Therefore, maternal pupil dilation in (and elicit) stronger emotional arousal than the posi- response to infant negative facial expressions is likely to tively-valenced faces used in the current study. That is, share some common mechanisms with emotional pupil the dimensions of arousal and valence are not orthogonal dilation in general which is elicited by a wide range of as difference in valence between stimuli requires a suffi - stimuli and conditions [20, 28–30]. Yet, there is evidence cient level of arousal to emerge [42]. that effects of social signals of emotion on pupil size In principle, the onset and the time course of pupil dila- may reflect distinct social-cognitive processes. Firstly, tion following an emotional stimulus could be estimated interpersonal mimicry of gestures including mimicry of from the latencies of the LC-NE subsystems and their the pupil size [39] may specifically modulate the pupil influence on pupil size [20]. To our knowledge, such esti - response to faces. Secondly, previous studies suggest mates of the time course of the pupil response have not that there may be a dissociable neurocognitive system been established. In practice, the emotive pupil response involved in monitoring infants’ emotional cues which is has been investigated from different time-windows important for supporting parental caregiving [2, 12, 13]. spanning 500–1300 ms [17], 600–1600 ms [18], 1000– u Th s, while probably mediated by the attentional LC-NE 1300 ms [43], 2–4 s [31, 32], or 2–6 s [15, 17] after stimu- system, the current results may be viewed as indexing a lus onset. The early and late time-windows have been Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 10 of 12 typically selected to cover the constriction and the dila- confounds in emotion research would be to use stimuli tion phase, respectively. In the current study, a relatively with invariable luminance levels as well as matched con- late time-window spanning 4–5 s after stimulus onset trast and spatial frequency profile [46]. However, in many was chosen for the extraction of pupil dilation in order to cases perfect matching between stimulus luminance minimize the contribution of the pupillary light response and other low-level features may be difficult or result in on the estimate. The current results indicate that the unnatural stimulus qualities. In the current study, con- pupil dilation within this time-window was both sensitive founding effects of stimulus luminance (i.e., variable face to the emotive content of the stimuli and independent luminance and the luminance variability between the of stimulus brightness. In future studies using constant stimulus and the pre-stimulus intervals) were controlled light conditions, a more detailed, frame-by-frame, analy- by contrasting pupil responses elicited by face stimuli to sis of the pupil response together with known latencies those elicited by pixel-scrambled version of the same face of the LC-NE system [20] might provide insight into the stimuli. In this control experiment (Experiment 2), the time course of LC activation in the context of emotional stimulus light intensity was exactly matched to the face perception. condition while all facial and emotive cues were removed Pupillary response elicited by emotive face stimuli from the stimuli. If the difference in pupil size across already around 600–1600 ms after stimulus onset in stimulus conditions were to persist in the control con- child participants [18], have been observed in previous dition, such effects could simply be attributed to differ - studies. This latency overlaps with the pupil constriction ences in stimulus luminance. Conversely, if the effects are reflex extracted in the current study. Furthermore, while unique to the face condition, they most likely stem from a study [15] using pictures from the International Affec - genuine emotive processes related to face perception. tive Picture System (IAPS), found no evidence for emo- While pupil constriction was affected by stimulus cat - tional effects in pupil constriction, a more recent study egory in both face and control condition, pupil dilation from the same authors indicated emotional suppression and it’s modulation by emotive stimulus category were of this initial light reflex [17]. In the current study, modu - confined to the face condition only. Thus, we may con - lations of pupil constriction in response to stimulus cat- fidently interpret the current results as indicating emo - egory were found across both experiments. However, tive pupil dilation elicited by infant faces which is further the control experiment (Experiment 2) indicated that intensified by negative emotional expressions. these effects were not specific to emotional content or to faces as they were found for the random non-face stimuli Limitations of the study as well. The difference between the current results from (1) The participants were not perfectly matched across those obtained with IAPS [17] may be related to the type the main and the control experiment. For example, unlike of emotionally salient stimuli used to evoke the auto- in the main experiment, the participants in the control nomic pupil response: in the IAPS study, the largest emo- experiment were both male and female, and parenthood tional suppression of pupil constriction was found for was not required as an inclusion criterion. However, par- erotic and violent scenes which were not used in the cur- ticipants in both experiments were healthy adults and rent study but may elicit CNS [44] and ANS [45] activity manifested very similar pupil dilation in response to distinct from other emotionally equally arousing stimuli. stimuli depicting infant distress or discomfort. (2) The Further, the pupil constriction in the current study may infants depicted in face stimuli were Caucasian while have been partially suppressed by the presentation of the the participants viewing the stimuli where both Black pre-stimulus visual pattern, which might also have sup- and Caucasian, with low and high SES, respectively. pressed emotional effects on this initial light reflex. Thus, Therefore, own-race biases in face processing [47] and further studies may be needed to clarify the modulations SES [48, 49] might have affected the emotive responses of pupil constriction by affective face processing, espe - elicited by the stimuli. (3) The current study focused in cially in the context of maternal responses to infant emo- testing intra-individual variation in pupil response across tive cues. variable infant facial expressions. Therefore, measures of Pupillary responses to emotional cues are relatively inter-individual variations in potentially related variables small and intermixed with the larger effects of stimulus such as maternal sensitivity to infant cues were not pre- or ambient luminance. In the current study, stimuli in sented. As such, positive effect of increased pupil size in different emotion categories were not significantly differ - response to pictures of infant negative affect was found ent with respect to their mean (bitmap) intensity values, within the current sample consisting of healthy mothers but there was a clear trend for both within- and between- (main experiment) and adult controls (control experi- category variability. An ideal solution for avoiding these ment). Future studies are needed to indicate whether the Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 11 of 12 Competing interests pupil response to infant faces is sensitive to inter-individ- The authors declare that the research was conducted in the absence of any ual variations in general and in relation to motherhood in commercial, financial, or other relationships that could be construed as poten- particular. tial competing interests. Availability of data and materials Conclusions The pupil/gaze data presented in the current study have been anonymized Our current results indicate that pupil diameter is a sen- and made publicly available at Zenodo (http://dx.doi.org/10.5281/ zenodo.45989). The analysis toolbox, gazeAnalysisLib [24] is open source and sitive marker of emotional processes elicited by infant custom scripts for analyses together with the currently utilized version of facial expressions in the targeted participant group of gazeAnalysisLib are available at GitHub (https://github.com/infant-cognition- mothers of infant children. While the perception of infant tampere/sa-pupil-analysis). Syntax (IBM SPSS Statistics) for statistical analyses performed for the extracted pupil constriction and dilation are further signals of distress may constitute a specific case of face included within the same repository. processing [2, 12, 13], the current approach may be appli- cable to other domains of social perception as well due Consent for publication Written informed consent for publication of individual person’s data was to a common psychophysiological pathway, the LC-NE obtained from one of the authors. This data was used in visualizing a repre- system. Consequently, it remains possible that compara- sentative single-trial pupil response (Fig. 4). ble pupil response may be elicited by non-infant stimuli Ethics approval and consent to participate as well or in non-mother viewers exposed to affective The study was ethically approved by the institutional review board of the facial stimuli. In order to address the specificity of the University of Stellenbosch and written informed consent was obtained from current emotive pupil response to infant cues, further all participants. studies with both adult and infant face stimuli as well as Funding participants with sufficient inter-individual variability in This research was supported by a joint project grant from the Academy of responsiveness to facial expressions of emotion in both Finland and National Research Foundation, South Africa (# 2501271617). The funders had no role in study design, data collection and analysis, decision to categories are needed. Nevertheless, the principle contri- publish, or preparation of the manuscript. bution of the current study is in indicating the feasibility of pupil diameter as an index of mothers’ perception and Received: 23 June 2016 Accepted: 24 January 2017 responsiveness to infant non-verbal communication. As such, pupil diameter may provide a useful and accessible measure for studies of individual variations in mother- infant interaction [1]. References 1. Strathearn L. Maternal neglect: oxytocin, dopamine and the neurobiol- Additional file ogy of attachment. J Neuroendocrinol. 2011;23(11):1054–65. 2. Peltola MJ, Yrttiaho S, Puura K, Proverbio AM, Mononen N, Lehtimäki T, et al. Motherhood and oxytocin receptor genetic variation are associated Additional file 1. Additional figures. with selective changes in electrocortical responses to infant facial expres- sions. Emotion. 2014;14(3):469–77. 3. Rilling JK. The neural and hormonal bases of human parental care. Neu- Abbreviations ropsychologia. 2013;51(4):731–47. ANOVA: analysis of variance; ANS: autonomic nervous system; AOI: area 4. Swain JE, Lorberbaum JP, Kose S, Strathearn L. Brain basis of early of interest; CNS: central nervous system; GLM: general linear model; IAPS: parent-infant interactions: psychology, physiology, and in vivo functional International Aec ff tive Picture System; LC-NE: locus coeruleus-norepinephrine; neuroimaging studies. J Child Psychol Psychiatry. 2007;48(3–4):262–87. MN: mild negative; MP: mild positive; POG: point of gaze; SES: socioeconomic 5. Yrttiaho S, Forssman L, Kaatiala J, Leppänen JM. Developmental precur- status; SN: strong negative; SP: strong positive. sors of social brain networks: the emergence of attentional and cortical sensitivity to facial expressions in 5 to 7 months old infants. PLoS ONE. Authors’ contributions 2014;9(6):e100811. DN, ET, and JML conceived, designed, and coordinated the study. SY per- 6. Peltola MJ, Forssman L, Puura K, van Ijzendoorn MH, Leppänen JM. formed pupil and eye-tracking analyses, statistical analyses, and drafted the Attention to faces expressing negative emotion at 7 months predicts manuscript. JML and DN helped to draft the manuscript. All authors critically attachment security at 14 months. Child Dev. 2015;86(5):1321–32. reviewed the manuscript and gave final approval for the version to be pub - 7. Barrett J, Fleming AS. Annual research review: all mothers are not cre- lished. All authors read and approved the final manuscript. ated equal: neural and psychobiological perspectives on mothering and the importance of individual differences. J Child Psychol Psychiatry. Author details 2011;52(4):368–97. Tampere Center for Child Health Research, School of Medicine, University 8. Glocker ML, Langleben DD, Ruparel K, Loughead JW, Gur RC, Sachser N. of Tampere, Lääkärinkatu 1, 33520 Tampere, Finland. Department of Psychia- Baby schema in infant faces induces cuteness perception and motivation try, Faculty of Health Sciences, Stellenbosch University, Stellenbosch, South for caretaking in adults. Ethology. 2009;115(3):257–63. Africa. 9. Glocker ML, Langleben DD, Ruparel K, Loughead JW, Valdez JN, Griffin MD, et al. Baby schema modulates the brain reward system in nulliparous Acknowledgements women. Proc Natl Acad Sci USA. 2009;106(22):9115–9. The authors are grateful to Alice Mado Proverbio for the stimulus material, 10. Kringelbach ML, Lehtonen A, Squire S, Harvey AG, Craske MG, Holliday IE, to Gerdia Harvey for assisting in data collection and to Jussi Kaatiala for the et al. A specific and rapid neural signature for parental instinct. PLoS ONE. programming efforts (gazeAnalysisLib). 2008;3(2):e1664. Yrttiaho et al. Behav Brain Funct (2017) 13:2 Page 12 of 12 11. Proverbio AM, Riva F, Zani A, Martin E. Is it a baby? Perceived age affects 32. Kret ME, Stekelenburg JJ, Roelofs K, deGelder B. Perception of face and brain processing of faces differently in women and men. J Cogn Neurosci. body expressions using electromyography, pupillometry and gaze meas- 2011;23(11):3197–208. ures. Front Psychol. 2013;4:28. 12. Proverbio AM, Brignone V, Matarazzo S, Del Zotto M, Zani A. Gender and 33. Lorber MF, O’Leary SG. Mediated paths to over-reactive discipline: moth- parental status affect the visual cortical response to infant facial expres- ers’ experienced emotion, appraisals, and physiological responses. J sion. Neuropsychologia. 2006;44(14):2987–99. Consult Clin Psychol. 2005;73(5):972–81. 13. Proverbio AM, Matarazzo S, Brignone V, Del Zotto M, Zani A. Processing 34. Frodi AM, Lamb ME. Child abusers’ responses to infant smiles and cries. valence and intensity of infant expressions: the roles of expertise and Child Dev. 1980;51(1):238–41. gender. Scand J Psychol. 2007;48:477–85. 35. Musser ED, Ablow JC, Measelle JR. Predicting maternal sensitivity: the 14. Beatty J, Lucero-Wagoner B. The Pupillary System. In: Cacioppo J, Tassi- roles of postnatal depressive symptoms and parasympathetic dysregula- nary LG, Berntson GG, editors. Handbook of psychophysiology. 3rd ed. tion. Infant Ment Health J. 2012;33(4):350–9. New York: Cambridge University Press; 2007. p. 142–62. 36. Moore GA, Hill-Soderlund AL, Propper CB, Calkins SD, Mills-Koonce WR, 15. Bradley MM, Miccoli L, Escrig MA, Lang PJ. The pupil as a measure Cox MJ. Mother-infant vagal regulation in the face-to-face still-face para- of emotional arousal and autonomic activation. Psychophysiology. digm is moderated by maternal sensitivity. Child Dev. 2009;80(1):209–23. 2008;45(4):602–7. 37. Aston-Jones G, Cohen JD. An integrative theory of locus coeruleus-nor- 16. Steinhauer SR, Siegle GJ, Condray R, Pless M. Sympathetic and parasym- epinephrine function: adaptive gain and optimal performance. Annu Rev pathetic innervation of pupillary dilation during sustained processing. Int Neurosci. 2005;28:403–50. J Psychophysiol. 2004;52(1):77–86. 38. Murphy PR, Robertson IH, Balsters JH, O’Connell RG. Pupillometry and 17. Henderson RR, Bradley MM, Lang PJ. Modulation of the initial light reflex P3 index the locus coeruleus-noradrenergic arousal function in humans. during affective picture viewing. Psychophysiology. 2014;51:815–8. Psychophysiology. 2011;48:1532–43. 18. Sepeta L, Tsuchiya N, Davies MS, Sigman M, Bookheimer SY, Dapretto M. 39. Kret ME, Tomonaga M, Matsuzawa T. Chimpanzees and humans mimic Abnormal social reward processing in autism as indexed by pupillary pupil-size of conspecifics. PLoS ONE. 2014;9(8):e104886. responses to happy faces. J Neurodev Disord. 2012;4(1):17. 40. Kreiman J, Gerratt BR, Ito M. When and why listeners disagree in voice 19. Prehn K, Kazzer P, Lischke A, Heinrichs M, Herpertz SC, Domes G. Eec ff ts quality assessment tasks. J Acoust Soc Am. 2007;122:2354–64. of intranasal oxytocin on pupil dilation indicate increased salience of 41. Yrttiaho S, Alku P, May PJ, Tiitinen H. Representation of the vocal rough- socioaffective stimuli. Psychophysiology. 2013;50:528–37. ness of aperiodic speech sounds in the auditory cortex. J Acoust Soc Am. 20. Laeng B, Sirois S, Gredebäck G. Pupillometry: a window to the precon- 2009;125(5):3177–85. scious? Perspect Psychol Sci. 2012;7(1):18–27. 42. Bradley MM, Lang PJ. The international affective picture system (IAPS) 21. Joshi S, Li Y, Kalwani RM, Gold JI. Relationships between pupil diameter in the study of emotion and attention. In: Coan JA, Allen JB, editors. The and neuronal activity in the locus coeruleus, colliculi, and cingulate handbook of emotion elicitation and assessment. New York: Oxford cortex. Neuron. 2016;89(1):221–34. University Press; 2007. p. 33. 22. Loewenfeld I. The pupil: anatomy, physiology, and clinical applications. 43. Bradley MM, Lang PJ. Memory, emotion, and pupil diameter: repetition of Detroit: Wayne State University Press; 1993. natural scenes. Psychophysiology. 2015;52(9):1186–93. 23. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. 44. Schupp H, Cuthbert B, Bradley M, Hillman C, Hamm A, Lang P. Brain The mini-international neuropsychiatric interview (M.I.N.I.): the develop- processes in emotional perception: motivated attention. Cogn Emot. ment and validation of a structured diagnostic psychiatric interview for 2004;18(5):593–611. DSM-IV and ICD-10. J Clin Psychiatry. 1998;59:22–33. 45. Lang PJ, Bradley MM, Cuthbert BN. Motivated attention: Aec ff t, activation, 24. Leppänen JM, Forssman L, Kaatiala J, Yrttiaho S, Wass S. Widely applicable and action. In: Lang PJ, Simons RF, Balaban M, editors. Attention and MATLAB routines for automated analysis of saccadic reaction times. orienting. Mahwah: Erlbaum; 1997. p. 97–135. Behav Res Methods. 2015;47(2):538–48. 46. Link B, Junemann A, Rix R, Sembritzki O, Brenning A, Korth M, et al. Pupil- 25. Koole SL. The psychology of emotion regulation: an integrative review. lographic measurements with pattern stimulation: the pupil’s response Cogn Emotion. 2009;23(1):4–41. in normal subjects and first measurements in glaucoma patients. Invest 26. Joosen KJ, Mesman J, Bakermans-Kranenburg MJ, van Ijzendoorn Ophthalmol Vis Sci. 2006;47(11):4947–55. MH. Maternal overreactive sympathetic nervous system responses to 47. Ekman P, Friesen WV, O’Sullivan M, Chan A, Diacoyanni-Tarlatzis I, Heider repeated infant crying predicts risk for impulsive harsh discipline of K, et al. Universals and cultural differences in the judgments of facial infants. Child Maltreat. 2013;18(4):252–63. expression of emotion. J Pers Soc Psychol. 1987;53:712–7. 27. Leerkes EM, Supple AJ, O’Brien M, Calkins SD, Haltigan JD, Wong MS, et al. 48. Hackman DA, Gallop R, Evans GW, Farah MJ. Socioeconomic status and Antecedents of maternal sensitivity during distressing tasks: integrat- executive function: developmental trajectories and mediation. Dev Sci. ing attachment, social information processing, and psychobiological 2015;18(5):686–702. perspectives. Child Dev. 2015;86(1):94–111. 49. Kraus MW, Cote S, Keltner D. Social class, contextualism, and empathic 28. Laeng B, Orbo M, Holmlund T, Miozzo M. Pupillary Stroop effects. Cogn accuracy. Psychol Sci. 2010;21(11):1716–23. Process. 2011;12(1):13–21. 29. Dabbs JM Jr. Testosterone and pupillary response to auditory sexual stimuli. Physiol Behav. 1997;62(4):909–12. 30. Cassady JM, Farley GR, Weinberger NM, Kitzes LM. Pupillary activity meas- ured by reflected infra-red light. Physiol Behav. 1982;28:851–4. Submit your next manuscript to BioMed Central 31. Kret ME, Roelofs K, Stekelenburg JJ, de Gelder B. Emotional signals from and we will help you at every step: faces, bodies and scenes influence observers’ face expressions, fixations and pupil-size. Front Hum Neurosci. 2013;7:810. • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit
Journal
Behavioral and Brain Functions – Springer Journals
Published: Feb 6, 2017