Please do not post this version on the web. The official published version of this article can be found at

Transcription

1 Please do not post this version on the web. The official published version of this article can be found at Psychological Reports, 2011, 108, 3, Psychological Reports 2011 Electroencephalographic responses to photographs: A case study of three women with distinct Adult Attachment Interview classifications 1 Kazuko Y. Behrens Texas Tech University Naomi I. Gribneau Bahm University of California, Berkeley Yingli Li Suzhou University Michael W. O Boyle Texas Tech University Summary. This preliminary study explored whether neurophysiological responses to visual stimuli, including attachment-related pictures, differed based on attachment status. Along with self-reported valence ratings and reaction times, recorded electroencephalographic (EEG) responses to a total of 100 images, 25 each of Positive, Negative, Neutral, and Personal (each participant s parents and child), were analyzed within and among three mothers with three attachment statuses (Dismissing, Preoccupied, and Secure), as judged by the Adult Attachment Interview (AAI). All three mothers gave their highest pleasantness ratings for Personal photographs. However, differences emerged when cross-region Alpha2 activation patterns in response to each picture type were compared amongst attachment categories. Alpha2 activation recorded during viewing of the participants children s photographs was similar to viewing Negative pictures for mothers with insecure (Dismissing and Preoccupied) status; whereas the Alpha2 activation of the mother with Secure status towards photographs of her child was similar to Positive pictures. Different patterns of hemispheric asymmetry in Beta1 frequency when processing different picture types were also found. The mother with Dismissing status showed significantly stronger left-hemisphere Beta1 activation across all image types. The Preoccupied mother showed significantly stronger right-hemisphere Beta1 activation for all but the Neutral images, during which activation did not differ between the two hemispheres. The mother with Secure status showed significantly stronger Beta1 activation in the left hemisphere for all but parental Personal photos, during which activation did not differ between the two hemispheres. Implications from the current findings and future research possibilities are discussed. Bowlby s attachment theory (1969/1982) is rooted in evolutionary theory, positing that a child s attachment behaviors represent activation of a universal species-wide behavioral system that enhances the likelihood of survival. Attachment behaviors are then expectably observed during times of distress (e.g., fear, injury, etc.), and deactivation of the system occurs when the child no longer senses threat. Typically, the distress is alleviated when the child seeks protection or comfort from an attachment figure, usually the biological mother. Bowlby further claimed that in hu- 1 Address correspondence to Kazuko Y. Behrens, Department of Human Development and Family Studies, Texas Tech University, Mail Stop 1230, Lubbock, TX or (kazuko.behrens@ttu.edu). DOI / PR ISSN

2 994 K. Y. Behrens, et al. mans, through repeated interactive experiences with the caregiver, the child develops an internal working model which influences the child s perception of self as worthy (or unworthy) of social acceptance (Bretherton & Munholland, 2008). Individual Differences in Research on Attachment Differences in internal working models, or infants different strategies of coping with distress, are believed to be observable in the seminal laboratory measure known as the Strange Situation Procedure (Ainsworth, Blehar, Waters, & Wall, 1978). These different strategies are identified as attachment patterns or categories (i.e., secure, avoidant, resistant) and have been related to the quality of the parent-child relationship (e.g., a secure child is confident in the caregiver s consistent availability in times of distress and can competently explore). Identifying attachment patterns in infancy is important because attachment contributes to and predicts developmental outcomes in varied domains. For example, early attachment security provides children the foundation for competent peer relationships, which then better prepare them for intimate relationships, and later parenting (e.g., Sroufe, Egeland, Carlson, & Collins, 2005). Equally important is to understand how different attachment patterns are formed. Analysis of responses in the Adult Attachment Interview (protocol, George, Kaplan, & Main, 1996; classification system, Main, Goldwyn, & Hesse, 2002) 2 has demonstrated powerful predictability of offspring s attachment quality (van IJzendoorn, 1995). That is, parents current states of mind regarding their own attachment experiences transmit to the next generation in a predictable manner, possibly through unconscious processes manifested during parent-child interactions. The Adult Attachment Interview identifies three major organized states of mind based largely on the coherence of interviewees discourses regarding childhood attachment experiences: Secure, Dismissing, and Preoccupied. Insecure states of mind are further characterized by distinct markers such as high idealization for Dismissing speakers and high involving anger for Preoccupied speakers (Main, et al., 2002). Unresolved or disorganized responses to loss or trauma (indicated by lapses in the monitoring of reasoning or discourse during discussions of the loss/trauma) are analyzed separately from organized states of mind, so for example, a tran- 2 This widely used attachment measure requires an extensive 2-week training institute which is offered by certified trainers several times a year at several locations around the world. After completing the institute, trainees may opt to proceed through a reliability testing process which takes a minimum of 18 months. Thorough training to this extent is necessary in order to consistently and objectively assess interviewees states of mind with respect to attachment to use the measure in scientific research. For this reason, developers of the AAI decided that both the protocol and classification manual should remain unpublished to avoid potential misuse of this system without proper training.

3 EEG Responses, Attachment Status 995 script which exhibits high unresolved scores but is otherwise relatively coherent can be judged Unresolved alternate Secure, referred to as U/Secure. Attachment and Neurophysiology With technological advancements, it is now possible to explore individual differences by correlating psychology with underlying physiology (Fox & Hane, 2008). For example, the association of left hemisphere activation with positive emotional (or approach) responses and right hemisphere activation with negative emotional (or withdrawal) responses has been well-documented by many studies across differing age groups (e.g., Davidson & Fox, 1982; Fox & Davidson, 1984; Henriques & Davidson, 1990; Schmidt, Shahinfar, & Fox, 1996). The electroencephalogram (EEG) is a commonly used technique for measuring brain activation during performance of various cognitive tasks. Compared to other neurophysiological techniques, EEG has high temporal resolution and is suitable for measuring brain activity during long and complex tasks (Springer & Deutsch, 1998). Some evidence linking EEG asymmetry with temperamental dispositions in infants has also been reported. For example, inhibited infants exhibit right frontal EEG asymmetry in the awake, quiet, nondistressed state (e.g., Fox, Calkins, & Bell, 1994). Some temperamental differences have been proposed to explain differences in attachment behaviors during the Strange Situation Procedure (e.g., Calkins & Fox, 1992), although at least one study of behavioral genetics did not support such claims, showing that temperamental differences or similarities were not necessarily correlated with twins attachment concordance (Bokhorst, Bakermans-Kranenburg, Fearon, van IJzendoorn, Fonagy, & Schuengel, 2003). It appears then that attachment and temperament continue to represent separate aspects of development, with neither fully explaining individual differences in the other (Vaughn, Bost, & van IJzendoorn, 2008). Thus, exploring contextual parameters that tune attachment and temperament systems (Vaughn, et al., 2008, p. 210) and the further understanding of how temperament and caregiving behaviors interact, remain important questions to pursue (Fox & Hane, 2008). To date, no direct association between EEG activity or activational asymmetry with specific attachment categories has been reported. According to Coan (2008), much has been learned from neurobiological studies of attachment, but primarily from nonhuman animal work, and studies of neuro-anatomical correlates of individual differences in attachment are rare. Nevertheless, some EEG studies of attachment have emerged but are largely confined to investigations of romantic attachment relationships in humans and often rely on self-report questionnaires (Cohen & Shaver, 2004; Simpson, Collins, Tran, & Haydon, 2007; Rognoni, Galati, Costa, & Crini, 2008). Although informative for understanding

4 996 K. Y. Behrens, et al. how adults reporting differing attachment styles experience close emotional relationships differently, correlated with differing EEG asymmetry patterns, self-report attachment styles often bear little relation to attachment states of mind which predict parent-child attachment quality (Roisman, et al., 2007). That is, those who self-report as having secure attachment styles in close adult relationships via questionnaire may not be judged on the Adult Attachment Interview as having secure states of mind with respect to attachment. The Adult Attachment Interview is the most reliable tool for examining the link between physiological correlates of attachment and the manifestation of attachment in parenting practices. Siegel (1999) discussed possible associations between neurobiology and attachment with a particular focus on how neurology can affect the states of mind with respect to attachment, as elucidated by the Adult Attachment Interview narratives. To date, the work by Gribneau (2006) appears to be the only study exploring a direct empirical link between the Adult Attachment Interview and EEG activation. Specifically, Gribneau explored associations between women s Unresolved status for loss on the Adult Attachment Interview and their event-related potential (ERP) responses to cemetery images. In this study, Gribneau randomly presented four categories of images consisting of social positive, nature positive, blatant death, and cemetery scenes. In the analysis, certain components of Unresolved women s ERPs, but not of non-unresolved women s, reacted to cemetery scenes similarly to the way all participants reacted to blatant death, suggesting that even subtle reminders of death may trigger Unresolved individuals to enter into states of mind that may lead to problematic parenting behaviors when interacting with their own children. Such findings supplement understanding of why Unresolved caregivers may engage in certain detrimental, anomalous caregiving behaviors. Yet, it is still not known if major Adult Attachment Interview statuses have different physiological correlates. Specifically, are the brain activation patterns exhibited by individuals with distinct Adult Attachment Interview categories different while responding to attachment related pictures, particularly photographs of their parents taken during the participant s childhood and current photographs of their own children? Current Study In this preliminary case study, the physiological correlates of attachment for three mothers with differing Adult Attachment Interview classifications were explored. Specifically, the electroencephalogram (EEG) was used to monitor the brain activity of participant mothers while they were rating visual stimuli of differing emotional content. Among the various electrophysiological techniques, EEG is widely used given its high tempo-

5 EEG Responses, Attachment Status 997 ral resolution when measuring brain activation during various cognitive tasks (Springer & Deutsch, 1998). The study design partly follows Gribneau s work (2006) in that participants completed both the Adult Attachment Interview and EEG sessions and their subsequent EEG was monitored in response to a series of visual images. Note, however, that in the present study, the categorization of picture types differed somewhat from the Gribneau experiment. The picture stimuli used here consisted of four types: Positive (e.g., smiling babies, puppies), Negative (e.g., malnourished children, corpses), Neutral (e.g., cups, leaves), and Personal photographs, the latter being further divided into participants parents and children. EEG activation patterns in response to Personal photographs and their relation to pictures of the other types (e.g., Positive, Negative) were of special interest. The following preliminary hypotheses were tested: (H 1 ) all mothers will rate their Personal photographs highest (i.e., most pleasant); (H 2 ) response time to different picture types will differ among mothers with different Adult Attachment Interview statuses; (H 3 ) activation differences in EEG patterns across different hemispheric locations will be observable amongst Adult Attachment Interview statuses when responding to Personal photographs; and (H 4 ) hemispheric differences (left vs right hemispheric dominance) will be observable among Adult Attachment Interview statuses for pictures of differing valence. Method Participants Three mothers (ages 34, 38, and 40 years) participated as part of a larger longitudinal social-emotional development project based on a community sample. All had previously undergone an Adult Attachment Interview, but separately consented to the EEG study and received $40 compensation for their participation. Photographs of their own parents during each mother s childhood, and photographs of their own child were collected in advance and used as visual stimuli. Measures Adult Attachment Interview. The interview is a semi-structured, 1-hr. interview asking questions regarding participants childhood experiences with their parents. The Adult Attachment Interview was administered following the standard procedure (George, et al., 1996), and verbatim transcripts were then coded according to the Adult Attachment Interview classification system (Main, et al., 2002). The Adult Attachment Interview is a well-validated measure with strong predictability for intergenerational transmission of attachment security (van IJzendoorn, 1995) and has also been reported to be unrelated to memory, intelligence, social desirability, or discourse style (Bakermans-

6 998 K. Y. Behrens, et al. Kranenburg & van IJzendoorn, 1993; Sagi, van IJzendoorn, Scharf, Koren-Karie, Joels, & Mayeless, 1994; Crowell, Waters, Treboux, O Connor, Colon-Downs, Feider, et al., 1996). The recent report that 10,000 Adult Attachment Interviews have been administered to date worldwide based on published studies, combining both nonclinical and clinical samples across cultures, attests to the extent to which the Adult Attachment Interview has been successfully used in research (Bakermans-Kranenburg & van IJzendoorn, 2009). Interviewees who generally discuss their childhood experiences with their parents coherently, openly, and truthfully, while demonstrating that they clearly value attachment relationships, are judged to be Secure. In this study, the mother with Secure status was fluent and descriptive about her childhood from the beginning, providing a number of episodic examples supporting her descriptors of her relationships with her parents. She was cooperative with the interviewer throughout the interview and very thoughtful in providing answers. This mother was often animated, openly showing free-flowing speech and clearly valued her relationships with her parents and her child. Interviewees who claim that they do not remember much from childhood or attempt to portray childhood relationships positively without sufficient supporting evidence are judged Dismissing. The Adult Attachment Interview for the mother with Dismissing status was characterized by frequent statements of I don t remember and was very general and superficial throughout the interview. She was unable to provide any specific incidents or memories. Her positive descriptors were often emphasized, such as very, very close to mother, always being loved, and all the time for being hugged or kissed, leading to high idealization scores. Interviewees who angrily discuss their childhood experiences or current relationships with one or both parents are judged Preoccupied. In this study, the mother with Preoccupied status described a relatively recent event in the relationship with her mother with vivid anger. This speaker ended her angry passage by addressing the parent as though she were present, showing a violation of manner with high anger scores. Picture stimuli. A total of 100 picture stimuli was presented for participants to rate in terms of pleasantness (e.g., see Schupp, Cuthbert, Bradley, Cacioppo, Ito, & Lang, 2000, for a similar method). Seventy-five picture stimuli (25 each of Positive, Negative, and Neutral) were selected from the International Affective Picture System (Lang, Bradley, & Cuthbert, 1999). An additional 25 photographs (Personal) consisting of each participant s parents (13) and child (12) were collected prior to the EEG session. All of these picture stimuli were programmed into a random presentation sequence via E-Prime 1.0 software. The primary interest was in learning how similar or dissimilar participants responses to their Person-

7 EEG Responses, Attachment Status 999 al photographs might be relative to their responses to pictures of a specified valence. Therefore, it was important to first confirm an agreed-upon valence for each picture that would be readily recognizable to all participants. To that end, 10 graduate students participated in finalizing the selection of all 75 pictures (in all three categories) with 100% agreement in their classification of each picture into the appropriate category. Each picture was presented on a computer monitor for 3 sec., followed by a 3-sec. blank-screen interval. The total running time for the presentation sequence was 10 min. Separately, a practice run of 10 pictures (Neutral pictures such as a screwdriver and a toothbrush, which were different from those used in the experimental task) was used to familiarize the participant with the rating task. Procedure After signing a consent form, participants were seated in a soundattenuated chamber and faced a computer monitor approximately 60 cm away on which the pictures were visually presented. A 32-channel EEG electrode cap was positioned on the participants scalps, and participants were subsequently briefed about the procedure and instructed to rate the pictures using a scale with anchors 1: Extremely unpleasant and 6: Extremely pleasant. Participants indicated their ratings by pressing a keypad placed in front of them. Two practice runs of the aforementioned extra picture stimuli were conducted to familiarize them with the task, followed by the random sequence of 100 pictures of differing valence. Participants were instructed to view the picture stimuli, decide on the rating value, and press the corresponding number key on the response pad all within the 3-sec. time limit. EEG Acquisition In the present study, EEG data were collected via a 32-channel EEG system (electrode assignments positioned in accordance with the International system; Binnie, Rowan, & Gutter, 1982). Sampling rate was 500 Hz and the impedance of each electrode was kept beneath 5 KΩ. EEG signals were filtered through a low pass (for waves with small variation) filter of 50 Hz and a high pass filter (for waves with large variation) of 0.15 Hz. Participants EEG recording began at the point when the first picture appeared on the screen. Artifacts (segments with EEG power exceeding ± 100 microvolts, such as those contaminated with excessive head movement, eye blinks, muscle tension, etc.) were examined and removed from the data via the standard off-line analysis procedure using Scan 4.4 software. Artifact removal comprised less than 3% of the data. EEG power in six different bandwidths [Delta (0.5 4 Hz), Theta (4 8 Hz), Alpha1 (8 11 Hz), Alpha2 (11 14 Hz), Beta1 (14 25 Hz) and Beta2 (25 35 Hz)] was extracted from the data through the use of a standard Fast Fourier Transform

8 1000 K. Y. Behrens, et al. (FFT) algorithm (Brigham, 2002) and averaged for each of five picture categories (i.e., Positive, Negative, Neutral, and Personal: parents, child). Results Behavioral Data Analyses Pleasant/unpleasant ratings for different picture categories. To test H 1 within each participant, the average response ratings were compared across the four picture categories, using the SAS Proc Mixed procedure, taking into account the fact that all the ratings were obtained from the same participant. The results were statistically significant (F 3,94 = , p <.0001, η 2 = 0.82 for the mother with Dismissing status; F 3,95 = 87.19, p <.0001, η 2 = 0.73 for the mother with Preoccupied status; and F 3,96 = 43.08, p <.0001, η 2 = 0.57 for the mother with Secure status). Descriptive indices of Table 1 Rating of Image Type by Adult Attachment Interview Status Status Positive Negative Neutral Personal M SD M SD M SD M SD Dismissing Preoccupied Secure Note. Image ratings ranged from 1: Very unpleasant to 6: Very pleasant. the three mothers ratings for each of the four categories of pictures are shown in Table 1 (also Fig. 1). All three mothers gave their highest ratings to Personal photographs regardless of their Adult Attachment Inter Ratings Positive Negative Neutral Personal Fig. 1. Rating of image type by Adult Attachment Interview status (1: Very unpleasant, 6: Very pleasant). Dismissing ( ); Preoccupied ( ); Secure ( ).

9 EEG Responses, Attachment Status 1001 Table 2 Reaction Time (msec.) For Rating Image Type by Adult Attachment Interview Status Status Positive Negative Neutral Personal M SD M SD M SD M SD Dismissing 1, , , , Preoccupied 1, , , , Secure 1, , , , view status. Insecure mothers also gave higher ratings for Personal photographs than did the mother with Secure status. Reaction times for different picture categories. To test H 2, the average reaction times were compared across the four picture categories within each subject, using the SAS Proc Mixed procedure, once again taking into account the fact that all the reaction times were from the same participant. The results were statistically significant (F 3,95 = 8.78, p <.0001, η 2 = 0.22 for the mother with Dismissing status; F 3,95 = 14.65, p <.0001, η 2 = 0.32 for the mother with Preoccupied status; and F3,96 = 15.52, p <.0001, η 2 = 0.33 for the mother with Secure status). For the mother with Dismissing status, the reaction time for the Negative pictures was the shortest while the reaction time for the Positive was the longest. For the mother with Preoccupied status, the reaction time for the Negative pictures was the longest while the reaction for the Personal was the shortest. For the mother with Secure status, the reaction time for the Neutral pictures was the shortest while the reaction time for the Personal was the longest. The descriptive analyses of 2,000 1,800 1,600 Reaction Time (msec.) 1,400 1,200 1, Positive Negative Neutral Personal Fig. 2. Reaction time to picture type by Adult Attachment Interview status. Dismissing ( ); Preoccupied ( ); Secure ( ).

10 1002 K. Y. Behrens, et al. reaction times for each picture type for each of the three participants are shown in Table 2 (also Fig. 2). EEG Data Analyses Correlation analysis across six hemispheric locations. To test H 3, correlation analyses of the aforementioned six EEG power bands [i.e., Delta (0.5 4 Hz), Theta (4 8 Hz), Alpha1 (8 11 Hz), Alpha2 (11 14 Hz), Beta1 (14 25 Hz), and Beta2 (25 35 Hz)] across the two hemispheres were conducted. Because this is an exploratory study, no specific prediction was made as to which EEG frequency would be most sensitive to the attachment-related emotional arousal. Therefore, all possible bands from 0.5 to 35 Hz were analyzed. The rationale for conducting this type of correlation analysis was to investigate whether: (1) an individual s EEG power levels in different bands across different hemispheres would show similar or different patterns for the four categories of pictures, and (2) if an Insecure mother, for example, would view her child with an EEG signature similar to that in response to Negative or Positive picture sets. For the latter correlation analysis, EEG power in each of the six bandwidths was averaged for the following six hemispheric locations: left frontal (F3, F7, FC3), right frontal (F4, F8, FC4), left parietal (P3, P7), right parietal (P4, P8), left temporal (T7), and right temporal (T8). Within each EEG power band, correlations among the five picture categories (with Personal including subcategories of parents and child) across each of the six hemispheric locations (left frontal, right frontal, left parietal, right parietal, left temporal, and right temporal) were conducted for each participant. Meaningful results were found in the Alpha2 band (11 14 Hz; see Table 3). The mother with Secure status showed a significant positive correlation between Alpha2 activation for pictures of her child and for Positive pictures (r =.90, p <.05, 95%CI =.33,.99), whereas the mothers with Dismissing and Preoccupied status showed significant positive correlations between their Alpha2 activation for pictures of their children and for Negative pictures (r =.83, p <.05, 95%CI =.06,.98; r =.82, p <.05, 95%CI =.03,.98, respectively). Analysis for left versus right hemispheric activation. Following previous literature, and as stated above, showing left hemispheric dominance for processing positive stimuli and right hemispheric activation for processing negative stimuli (Davidson & Fox, 1982; Fox & Davidson, 1984; Henriques & Davidson, 1990; Schmidt, et al., 1996), the present study explored whether the three mothers showed differences in hemispheric activity when processing the five categories of pictures (the personal pictures being further segmented into parental pictures and child pictures). To test H 4, EEG power in each of the six EEG bandwidths was averaged for the left hemisphere (O1, P7, P3, CP3, TP7, C3, T7, FT7, FC3, F3, F7) and the right hemisphere (O2, P8, P4, CP4, TP8, C4, T8, FT8, FC4, F4, F8). Sub-

11 EEG Responses, Attachment Status 1003 Table 3 Correlations of Alpha2 Cross-region Activation Pattern For Five Picture Types by Adult Attachment Interview Status Status & Picture Type Negative Neutral Parent Child Dismissing Positive Negative * Neutral Parent.28 Child Preoccupied Positive Negative.87*.02.82* Neutral.01.89* Parent.44 Child Secure Positive * Negative Neutral Parent.18 Child *p <.05. sequently, the 5 (picture type) 2 (hemisphere) Mixed Procedure analysis within the SAS System was used. The model consisted of: beta value = intercept + β 10 * image type + β 02 * left/right + β 12 * image type * left/right + ε ij, and was based on data from each participant, with their averaged power from the first second EEG segment (post picture presentation) 3 as the dependent variable. The rationale for analyzing just the first second of the EEG segment rather than the entire 3-sec. acquisition epoch was to eliminate possible contamination from motor responses associated with pressing the ratings key or other later movements not associated with the rating process (e.g., blinking). Note that all participants completed their pleasant/unpleasant ratings within 2 sec. of picture presentation (see Table 2). The SAS Proc Mixed Procedure yielded statistically significant results in participants Beta1 power band (14 25 Hz). Fig. 3 shows three graphs, depicting activation patterns in response to each picture category for each mother. The mother with Dismissing status had significantly greater Beta1 activation in her left hemisphere (F 4,5000 = , p <.0001, η 2 = 0.62). The mother with Preoccupied status showed overall stronger activation in her right hemisphere (F 4,5000 = 18.32, p <.0001, η 2 = 0.41) for all but the Neutral pictures. The mother with Secure status showed overall stronger activa- 3 Because the sampling rate was set at 500, 501 numbers for power variation were collected at each second.

12 1004 K. Y. Behrens, et al. Table 4 Beta 1 Activation (µv) Comparisons Between the Right and the Left Hemisphere For Five Picture Types by Adult Attachment Interview Status Type Hemisphere Dismissing Preoccupied Secure M SD M SD M SD Positive Left Right Negative Left Right Neutral Left Right Parents Left Right Child Left Right tion in the left hemisphere (F 4,5000 = , p <.0001, η 2 = 0.69) for all pictures except for her parents photographs. Beta1 activation comparisons for each Adult Attachment Interview status and each picture category are shown in Table Dismissing Preoccupied Secure Dismissing Preoccupied Microvolts (μv) Secure Dismissing Preoccupied Secure Dismissing Preoccupied Secure Dismissing Preoccupied Secure Positive Negative Neutral Parents Child Fig. 3. Beta 1 activation for 25 different pictures in each type by hemisphere and by Adult Attachment Interview status (25 pictures for Positive, Negative, Neutral, 13 pictures for Parents, 12 pictures for Child). Left ( ); Right ( ). Discussion In this preliminary study, distinct Adult Attachment Interview states of mind were examined to see if they correlated with different EEG activation patterns while viewing a series of pictures, including photographs of

13 EEG Responses, Attachment Status 1005 participants parents and children. The idea that a family photograph may evoke emotional responses with meaningful differences, corresponding to the differences in attachment status, was first suggested by Main and colleagues (Main, Kaplan, & Cassidy, 1985) when they showed 6-year-olds their family Polaroids, following a series of attachment tasks. The present study, however, is the first study to include family photographs as a part of a picture stimulus set and was specifically designed to explore the proposed association. Note that general claims regarding the definitive nature of electrophysiological correlates of attachment would be premature based on the current findings alone. The present study was an attempt to examine activation patterns of the brain in individuals with different Adult Attachment Interview statuses as prompted by pictures of their attachment figures or of their own children. Possible Interpretations Regardless of Adult Attachment Interview status, all three mothers rated personal photographs higher than any other category of pictures, supporting H 1. This indicates that behaviorally, all mothers regard their parents and their children as positive entities and rate them most pleasant, possibly to meet social desirability expectations. Reaction times differed among the individuals with three Adult Attachment Interview statuses, however, supporting H 2. Whereas the mother with Dismissing status took the longest to respond to Positive images, the mother with Preoccupied status took the longest to respond to Negative images. In contrast, the mother with Secure status took the longest to respond to Personal photographs. These differences may be related to the underlying states of mind which led to their differing Adult Attachment Interview classifications. First, since individuals with Dismissing status tend to idealize, creating generalized positive accounts of even negative experiences, it is possible that mothers with this status may be accustomed to distorting their perceptions. Thus, when there is nothing to distort, as in the case of Positive images, the mother may re-adjust her image processing, by taking slightly longer to respond to Positive pictures. Secondly, individuals with Preoccupied status are known to display currently involving anger during the Adult Attachment Interview, sometimes losing themselves in angry speech about negative experiences with attachment figures that occurred long ago. It is possible that the Negative pictures in the sequence, even when these pictures are not attachment-related, may have triggered the mother with Preoccupied status to re-experience the negative emotions surrounding her own attachment experiences, thereby delaying her reaction times as she dwelled on these Negative pictures. The trigger that may have interfered with this cognitive process has also been suggested by Blount-Matthews (2004) who found that when the word mother (but

14 1006 K. Y. Behrens, et al. not neutral words such as basket or table ) was used as a subliminal prime, individuals with Preoccupied status took significantly longer to complete a lexical decision task than did those with Dismissing or Secure status. Remember that individuals with Secure status are generally characterized by coherence of speech and thoughtfulness, often demonstrated by high metacognitive function (Main, et al., 2002). As such, it is possible that the mother with Secure status took longer to view each image of her parents or her child as they were associated with distinctly (reflectively and attentively) processed events. The EEG data revealed some intriguing contrasts among the three mothers in Alpha2 and Beta1 bandwidths. First, in the correlation analyses, Alpha2 activation patterns for both mothers with Dismissing and Preoccupied status for photographs of their children were similar to those for the Negative pictures, supporting the notion that some parents with Insecure status may rebuff or view their child as aversive (Cassidy & Berlin, 1994). In contrast, the Alpha2 activation pattern of the mother with Secure status for photographs of her child was more similar to that for positive pictures. Thus, H 3 was partially supported. For the hemispheric activation analyses, the mother with Dismissing status had significantly stronger brain activation in her left hemisphere regardless of picture type, consistent with the idealization that characterizes a Dismissing status (Main, et al., 2002), and indicative of this mother s disconnect in perceiving even Negative pictures in a positive manner. The mother with Preoccupied status showed stronger activation in her right hemisphere for all but the Neutral pictures, consistent with the involving anger often present in the Preoccupied status (Main, et al., 2002). That is, except for Neutral images which consisted of nonsocial or nonvalence-provoking images, the mother with Preoccupied status may be hyper-reactive in perceiving and processing social images, even positive ones which might remotely remind her of her attachment relationships in a negative way. The mother with Secure status showed stronger activation of her left hemisphere except for Personal parental photographs during which activations did not differ between the two hemispheres. This was partly unexpected, because the Secure mother was anticipated to show stronger activation in the left hemisphere for Positive pictures and stronger activation in the right for Negative pictures (cf. Davidson, 1984), reflecting the positive or negative valence of the respective photographs. Nevertheless, the mother with Secure status did not show an asymmetry for personal photographs, reflecting the more balanced view commonly found for Secure status, wherein individuals are prone to perceiving their attachment figures realistically with neither idealization nor anger (Main, et al., 2002). Thus, H 4 was supported.

15 EEG Responses, Attachment Status 1007 Notably, in previous EEG studies both Alpha2 and Beta1 EEG bands have been shown to be sensitive to emotional valence (e.g., Sachs, Anderer, Dantendorfer, & Saletu, 2004; Wyczesany, Kaiser, & Coenen, 2008, 2009). In the present study, the findings that Alpha2 and Beta1 activations correlated with different attachment statuses when processing Personal photographs (expected to contain emotional valence) further supports that these two EEG bands are adequate for analyzing the human brain processing of emotional stimuli. Although this study was exploratory in nature, the results provide a novel glimpse into the underlying brain states exhibited by three mothers of three differing Adult Attachment Interview statuses. The Adult Attachment Interview is a tool known to potentially surprise the unconscious (Hesse, 2008, p. 555). The gap between what the Insecure mothers tried to consciously present behaviorally by selecting a highly pleasant rating for their child and what was shown by their EEG patterns that specifically resembled their brain responses evoked by Negative pictures, helps better explain the often unintended insensitive behavior toward their children. Certainly, one participant per Adult Attachment Interview classification is not sufficient to make any definitive claims, and a full study with a much larger sample is needed to validate the current findings. Another limitation of the current design includes the choice of picture categories. It is possible that Positive and Negative photographs were not balanced for arousal and relative effectiveness at invoking a particular emotion. Nevertheless, the present data may be regarded as a first step toward providing a framework for integrating the Adult Attachment Interview and electrophysiological brain states, assisting in a better understanding of how individuals with differing Adult Attachment Interview statuses may perceive their attachment figures and their own children, differences that were reflected in both electrophysiological measures and behavioral measures. References Ainsworth, M. D. S., Blehar, M., Waters, E., & Wall, S. (1978) Patterns of attachment: a psychological study of the strange situation. Hillsdale, NJ: Erlbaum. Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (1993) A psychometric study of the Adult Attachment Interview: reliability and discriminant validity. Developmental Psychology, 29, Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2009) The first 10,000 Adult Attachment Interviews: distributions of adult attachment representations in clinical and non-clinical groups. Attachment & Human Development, 11, doi: / Binnie, C. D., Rowan, A. J., & Gutter, T. (1982) A manual of electroencephalographic technology. Cambridge, UK: Cambridge Univer. Press.

16 1008 K. Y. Behrens, et al. Blount-Matthews, K. (2004) Attachment and forgiveness in human development: a multi-method approach. Unpublished doctoral dissertation, Univer. of California, Berkeley. Bokhorst, C. L., Bakermans-Kranenburg, M. J., Fearon, R. M. P., van IJzendoorn, M. H., Fonagy, P., & Schuengel, C. (2003) The importance of shared environment in mother-infant attachment security: a behavioral genetic study. Child Development, 74, Bowlby, J. (1969/1982) Attachment and loss: Vol. I. Attachment. New York: Basic Books. Bretherton, I., & Munholland, K. A. (2008) Internal working models in attachment relationships: elaborating a central construct in attachment theory. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: theory, research, and clinical applications. (2nd ed.) New York: Guilford. Pp Brigham, E. O. (2002) The fast fourier transform and its applications. New York: Prentice Hall. Calkins, S. D., & Fox, N. A. (1992) The relations among infant temperament, security of attachment, and behavioral inhibition at twenty-four months. Child Development, 63, Cassidy, J., & Berlin, L. (1994) The insecure/ambivalent pattern of attachment: theory and research. Child Development, 65, Coan, J. A. (2008) Toward a neuroscience of attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: theory, research, and clinical applications. (2nd ed.) New York: Guilford. Pp Cohen, M. X., & Shaver, P. R. (2004) Avoidant attachment and hemispheric lateralization of the processing of attachment- and emotion-related words. Cognition & Emotion, 18, Crowell, J. A., Waters, E., Treboux, D., O Connor, E., Colon-Downs, C., Feider, O., Golby, B., & Posada, G. (1996) Discriminant validity of the Adult Attachment Interview. Child Development, 67, Davidson, R. J. (1984) Affect, cognition, and hemispheric specialization. In C. E. Izard, J. Kagan, & R. Zajonc (Eds.), Emotion, cognition, and behavior. New York: Cambridge Univer. Press. Pp Davidson, R. J., & Fox, N. A. (1982) Asymmetrical brain activity discriminates between positive and negative affective stimuli in human infants. Science, 218, Fox, N. A., Calkins, S. D., & Bell, M. A. (1994) Neural plasticity and development in the first two years of life: evidence from cognitive and socioemotional domains of research. Development and Psychopathology, 6, Fox, N. A., & Davidson, R. J. (1984) Hemispheric substrates of affect: a development model. In N. A. Fox & R. J. Davidson (Eds.), The psychobiology of affective development. Hillsdale, NJ: Erlbaum. Pp Fox, N., & Hane, A. A. (2008) Studying the biology of human attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: theory, research, and clinical applications. (2nd ed.) New York: Guilford. Pp George, C., Kaplan, N., & Main, M. (1996) Adult Attachment Interview. Unpublished protocols, Univer. of California, Berkeley. Gribneau, N. (2006) Event-related potentials to cemetery images distinguish electroencephalogram recordings for women Unresolved for loss on the Adult Attachment Interview. Unpublished doctoral dissertation, Department of Integrative Biology, Univer. of California at Berkeley.

17 EEG Responses, Attachment Status 1009 Henriques, J. B., & Davidson, R. J. (1990) Regional brain electrical asymmetries discriminate between previously depressed and healthy control subjects. Journal of Abnormal Psychology, 99, Hesse, E. (2008) The Adult Attachment Interview: protocol, method of analysis, and empirical studies. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (2nd ed.) New York: Guilford. Pp Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999) International Affective Picture System (IAPS): instruction manual and affective ratings. (Technical Report No. A-4) Gainesville, Florida: Center for Research in Psychophysiology, Univer. of Florida. Main, M., Goldwyn, R., & Hesse, E. (2002) Adult attachment scoring and classification system. Unpublished manuscript, Univer. of California, Berkeley. Main, M., Kaplan, N., & Cassidy, J. (1985) Security in infancy, childhood, and adulthood: a move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50(1-2, Serial No. 209), Rognoni, E., Galati, D., Costa, T., & Crini, M. (2008) Relationship between adult attachment patterns, emotional experiences and EEG frontal asymmetry. Personality and Individual Differences, 44, Roisman, G. I., Holland, A., Fortuna, K., Fraley, R. C., Clausell, E., & Clarke, A. (2007) The Adult Attachment Interview and self-reports of attachment style: an empirical rapprochement. Journal of Personality and Social Psychology, 92, Sachs, G., Anderer, P., Dantendorfer, K., & Saletu, B. (2004) EEG mapping in patients with social phobia. Psychiatry Research: Neuroimaging, 131, Sagi, A., van IJzendoorn, M. H., Scharf, M. H., Koren-Karie, N., Joels, T., & Mayeless, O. (1994) Stability and discriminant validity of the Adult Attachment Interview: a psychometric study in young Israeli adults. Developmental Psychology, 30, Schmidt, L. A., Shahinfar, A., & Fox, N. A. (1996) Individual differences in temperament. In V. S. Ramachandran (Ed.), Encyclopedia of human behavior. Vol. 2. San Diego, CA: Academic Press. Pp Schupp, H. T., Cuthbert, B. N., Bradley, M. M., Cacioppo, J. T., Ito, T., & Lang, P. J. (2000) Affective picture processing: the late positive potential is modulated by motivational relevance. Psychophysiology, 37, Siegel, D. J. (1999) The developing mind: toward a neurobiology of interpersonal experience. New York: Guilford. Simpson, J. A., Collins, W. A., Tran, S., & Haydon, K. C. (2007) Attachment and the experience and expression of emotions in romantic relationships: a developmental perspective. Journal of Personality and Social Psychology, 92, Springer, S. P., & Deutsch, G. (1998) Left brain, right brain: perspectives from cognitive neuroscience. (5th ed.) New York: W. H. Freeman. Sroufe, L. A., Egeland, B., Carlson, E., & Collins, A. (2005) Placing early attachment experiences in developmental context: the Minnesota longitudinal study. In K. Grossmann, K. Grossmann, & E. Watters (Eds.), Attachment from infancy to adulthood: the major longitudinal studies. New York: Guilford. Pp van IJzendoorn, M. H. (1995) Adult attachment representations, parental responsiveness, and infant attachment: a meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117,

18 1010 K. Y. Behrens, et al. Vaughn, B. E., Bost, K. K., & van IJzendoorn, M. H. (2008) Attachment and temperament: additive and interactive influences on behavior, affect, and cognition during infancy and childhood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: theory, research, and clinical applications. (2nd ed.) New York: Guilford. Pp Wyczesany, M., Kaiser, J., & Coenen, A. M. (2008) Subjective mood estimation covaries with spectral power EEG characteristics. Acta Neurobiologiae Experimentalis, 68, Wyczesany, M., Kaiser, J., & Coenen, A. M. (2009) Cortical lateralization patterns related to self-estimation of emotional state. Acta Neurobiologiae Experimentalis, 69, Accepted June 8, 2011.