Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/springer-journals/difficulty-in-disengaging-from-threat-and-temperamental-negative-Z8J0wDfFuO
- Publisher
- Springer Journals
- Copyright
- Copyright © 2012 by Nakagawa and Sukigara; licensee BioMed Central Ltd.
- Subject
- Biomedicine; Neurosciences; Neurology; Behavioral Therapy; Psychiatry
- eISSN
- 1744-9081
- DOI
- 10.1186/1744-9081-8-40
- pmid
- 22897933
- Publisher site
- See Article on Publisher Site
Abstract
Background: Attention disengagement is reportedly influenced by perceiving a fearful facial expression even in the first year of life. In the present study, we examined whether individual differences in disengaging from fearful expressions predict temperamental negative affectivity. Method: Twenty-six infants were studied longitudinally at 12, 18, 24, and 36 months, using an overlap paradigm and two temperament questionnaires: the Japanese versions of the revised Infant Behavior Questionnaire and Early Childhood Behavior Questionnaire. Results: The infants fixated significantly more frequently to fearful than to happy or neutral faces. The attentional bias to threat (i.e., the number of fixed responses on fearful faces divided by the total number of fixed responses on faces) at 12 months was significantly positively correlated with negative affect at 12 months, and its relations with negative affect measured later in development was in the expected positive direction at each age. In addition, a moderation analysis indicates that the orienting network and not the executive network marginally moderated the relation between early attentional bias and later fear. Conclusions: The results suggest that at 12 months, infants with more negative affectivity exhibit greater difficulty in disengaging their attention from fearful faces. We also found evidence that the association between parent- reported fear and disengagement might be modulated in the second year, perhaps because of the differences in temperamental control networks. Keywords: Attention, Infant, Negative affect, Longitudinal study, Temperament Background keep cognitive resources focused on the stimuli and re- Visual-spatial attention systems can reportedly detect sult in increased anxiety [1]. This trend might influence threat-related stimuli rapidly. The propensity to quickly subsequent cognitive and emotional processing, which is detect the presence of threatening stimuli, such as likely to play an important role in shaping children’s snakes and angry faces, may be an important survival cognitive representations of themselves, others, and the and adaptive mechanism. Threat-related stimuli (e.g., situation, from their earliest years [3]. threat words or angry faces) may also cause a delay in Even in infancy, humans have been found to orient disengagement [1], a tendency possibly increased by an more quickly to threatening than to nonthreatening individual’s elevated level of state anxiety. Further, using stimuli [4]. Recent studies have demonstrated that 7- fearful facial expressions as stimuli, Georgiou et al. [2] month-olds disengaged their fixation significantly less showed that high trait-anxious people exhibited frequently from fearful faces than from happy faces and extended dwell time to threat-related stimuli. The inabil- control stimuli [5]. Moreover, Peltola et al. [6] found that ity to rapidly disengage from threat-related stimuli may the delayed withdrawal of attention reflected not a sim- ple response to fearful wide-open eyes but rather an enhanced sensitivity to facial signals of threat. Fearful * Correspondence: nakagawa@hum.nagoya-cu.ac.jp expressions also caused greater heart rate deceleration Graduate School of Humanities and Social Sciences, Nagoya City University, 1, Yamanohata, Mizuho-cho, Mizuho-ku, Nagoya 467-8501, Japan © 2012 Nakagawa and Sukigara; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 2 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 responses in 7-month-old infants during the first overlap task that we used following Peltola et al. [5], 1000 ms of face viewing [7]. Leppänen et al. [7] con- infants were required to disengage their fixation from a cluded that emotion–attention interactions such as those centrally presented facial expression and shift attention displayed by adults can also be observed early in life. to a peripheral target. Growing evidence suggests that attentional bias to Second, we examined the relationship between indi- threat plays a causal role in individual differences in vidual differences in fear or negative affectivity and the emotional vulnerability [8,9]. Lonigan et al. [10] consid- attentional bias to threat in early life. For this purpose, ered temperament’s contributions to childhood disor- we examined individual differences in fear or negative ders. They extracted factors similar to those drawn affectivity through a revised Infant Behavior Question- previously from childhood self-reported items that assess naire (IBQ-R Japanese version [16]) administered at emotionality and attention and described automatic at- 12 months and the Early Childhood Behavior Ques- tention allocation mechanisms linked to negative tionnaire (ECBQ Japanese version [17]) administered at affectivity, which may have an effect on both the daily ages 18–36 months. Since an increased number of fix- experiences of children and their proneness to future ation responses (i.e., no movement) with a centrally negative experiences. Thus, the association between presented fearful face has been given as evidence of negative affectivity and anxiety pathology could be the effect of fearful faces on attentional disengagement mediated by the attentional bias to threat. [5], we computed an index of the attentional bias to Going beyond the individual mechanism of anxiety- threat-related stimuli on the basis of these fixation related information processing, Fox et al. [11] pro- responses. Negative affectivity (i.e., the reactive compo- posed a model of plasticity for affective neurocircuitry, nent of temperament) is considered relatively easy to describing how genetic disposition and environmental change early in life. Effortful control reflects a volun- circumstances may interact. Thus, a child’s fearful tem- tary component of attention and undergoes significant perament elicits and is elicited by the caregiver’s in- development in the second year of life and later [18]. sensitivity and intrusiveness to shape the attentional In the first year of life, the association between a high bias to threat and the neural systems involved in this level of negative affectivity and an attentional bias to bias (i.e., ventral prefrontal cortex-amygdala circuitry). threat would involve a mainly reactive temperamental Fox et al. [11] further suggested that exaggerated at- component and would be easier to notice. In the sec- tentional bias to threat may cause the emergence and ond and third years, however, this connection might maintenance of anxious behaviors. be modified by control systems such as effortful con- On the basis of a meta-analysis of 172 studies, trol. Children and adolescents, who are considered Bar-Haim et al. [12] pointed out that although atten- high in negative affectivity and low in effortful compo- tional bias to threat may largely contribute to the devel- nents of temperament, are reported to demonstrate a opment and maintenance of anxiety over time, the significant attentional bias in favor of threat stimuli possibility of a causal link between the two has been in- [19]. sufficiently investigated. Recently, longitudinal studies of Third, we conducted a moderation analysis to examine very young children have examined the relation between if an effortful control moderates the link between atten- attentional bias to threat and later socioemotional out- tional bias at 12 months and temperament at 36 months. come or risk of psychopathology [13,14]. Results indicate In infancy, a brain network involved in orienting to sen- the moderating roles that attention played in anxiety de- sory events may provide the chief means of self-regula- velopment. However, the youngest participants in these tion. This orienting network involves areas of the studies were 24 months old [13], and the dot-probe task inferior and superior parietal lobes and the frontal eye used for very young children did not necessarily measure fields. Later in childhood, however, the executive atten- tion system, including the anterior cingulate, insula, and the ability to disengage [14]. Therefore, further research is needed to fill these gaps in the previous studies in areas of the basal ganglia, becomes dominant as a mech- order to understand the initial structure and function of anism of self-regulation [20]. Effortful control as a tem- peramental construct is considered to reflect the anxiety-related information processing. For the reasons stated above, we conducted a longitu- functioning of a neutrally based executive attention. The dinal study of infants approximately 12–36 months old. IBQ-R orienting score measures an orienting attentional network, and the ECBQ effortful score assesses an ex- Fear develops by the end of the first year of life, and fearful infants show inhibition of motor approach [15]; ecutive attention network. Therefore, we included both hence, we studied infants from the end of their first year. the orienting score of the IBQ-R (at 12 months) and the The purpose of this study was, first, to confirm the effortful control score of ECBQ (at 24 months). The infants’ greater difficulty in disengaging attention from interactive (moderating) effect was tested by using cen- fearful faces than from happy or neutral faces [5]. In the tered variables in hierarchical regression. Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 3 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 We hypothesized that effortful control buffers the link Facial expression overlap task between attentional bias and the variables of fear At the beginning of each trial, the experimenter dis- or negative affect. Specifically, we predicted that the cor- played a central fixation attractor on the monitor in relation between the 12-month attentional bias score front of each infant or toddler by pressing a key. While and 36-month temperament score (fear or negative the infant looked at the attractor, the experimenter affectivity) would be stronger among toddlers low in presses another key, causing the facial expression stimu- effortful control. On the other hand, toddlers with high lus (i.e., a happy, fearful, or neutral face) to replace the orienting scores might interact in the opposite manner. fixation attractor in the center. The central stimulus Because high scores on the latter are partially due to a remained visible throughout the trial. Subsequently, tendency to focus for long periods, toddlers who attend (200 ms afterward), a peripheral stimulus was presented to threat might be particularly likely to continue focus- for 2600 ms at approximately 30 degrees to either the ing on threatening stimuli, thus possibly increasing the left or right of the central fixation point. The experiment likelihood that they will be fearful at 36 months. comprised a total of 24 trials, with 12 left and 12 right targets presented in a pseudorandom order. Method Stimuli Participants The central fixation attractors and peripheral targets Twenty-six infants (15 boys, 11 girls) with no history of comprised brightly colored abstract figures, which were perinatal or postnatal difficulties were recruited through animated and subtended at a visual angle of 5 degrees. local maternity groups in Nagoya, Japan’s third largest Each central fixation attractor was accompanied by a metropolitan area, located near the center of the coun- sound. The central facial stimuli used were from Japa- try. Criteria for participation in the study were no nese and Caucasian Facial Expressions of Emotion [21]. known complications of birth or other causes, having For each expression (i.e., happy, fearful, or neutral), color been carried to full term (more than 37 weeks gestation), images of two male and two female models were dis- and normal birth weight (2500 g–4000 g). The infants played on the monitor, and the peripheral stimuli were were longitudinally assessed four times (at 12, 18, 24, reflected in a first-surface mirror on the left or right side and 36 months of age) through eye movement record- at approximately 30 degrees from the central fixation ings and behavior questionnaires filled out by their care- point. The facial stimuli appeared at a vertical and hori- givers. All caregivers were Japanese mothers who gave zontal visual angle of 9 and 6 degrees, respectively. informed consent on behalf of their infants before the experiments. The study was approved by the Ethics IBQ-R and ECBQ Committee of Nagoya City University (No. 07007) and The following instruments were used to assess the fre- accorded with the ethical standards specified in the 1964 quency of occurrence, over the previous one or two Declaration of Helsinki. One participant at 18 months, weeks, of temperament-related behaviors on a 7-point four at 24 months, and one at 36 months of age could Likert scale ranging from never to always. The IBQ-R not participate in the experiments for personal reasons was administered at 12 months, and the ECBQ at 18, 24, (e.g., birth of a sibling or father’s work reassignment). and 36 months. The IBQ-R included 191 items yielding 14 scales, while the ECBQ consisted of 201 items con- Procedure taining a total of 18 scales. Factor analyses of the IBQ-R Eye movement recording and ECBQ scale scores identified a three-factor solution: Each participant sat in a baby chair in a semidark area, Positive Emotionality/Surgency, Negative Affectivity, and 65 cm from the color monitor of an AV tachistoscope Regulatory Capacity/Orienting [22]. The following αs are (IS-702). During the eye movement recording, the parti- based on Nakagawa and Sukigara [16] and Sukigara cipant’s mother was nearby but stayed out of sight. An et al. [17]. In the IBQ-R, scales of Distress to Limitations experimenter outside the semidark area monitored the (α = .82), Sadness (α = .83), Falling Reactivity (negatively; participant’s eye movements through a low-angle CCD α = .80), and Fear (α = .92) primarily loaded on Negative near-infrared video camera (ELMO CN43H) positioned Affect. The Orienting factor of IBQ-R was primarily in front of the participant and controlled the stimulus contributed by Attention/Duration of Orienting (α = .74), presentation by means of a microcomputer (FMV-S167). Low Intensity Pleasure (α = .78), Soothability (α = .61), The stimuli presented were superimposed synchronously and Cuddliness (α = .77). In the ECBQ, scales of Frustra- on video images of the infant’s eye movements by a tion (α = .80), Sadness (α = .78), Discomfort (α = .65), digital image processor (FOR-A, MF-310) and were then Motor Activation (α = .69), Fear (α = .72), Soothability recorded on videotape (SONY DSR-11), which was sub- (negatively; α = .87), Shyness (α = .82), and Perceptual sequently used for off-line video coding. Sensitivity (α = .81) primarily loaded on Negative Affect. Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 4 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 Moreover, the effortful control factor of ECBQ was pri- An examination of the inter-rater reliability between marily defined by Attention Focusing (α = .89), Inhibitory the two coders showed .96 agreement regarding their Control (α = .85), Attention Shifting (α = .66), Low Inten- classification of responses (i.e., disengagement responses sity Pleasure (α = .71), and Cuddliness (α = .74). to targets, responses to stay fixated on the face, and fail- At the end of each experimental session (at 12, 18, 24, ures to respond to targets) and .94 correlation between and 36 months of age), the mothers were given a tem- the RTs of the disengagement responses. Because partici- perament questionnaire (IBQ-R or ECBQ) to complete pants with less than two scorable responses (responses at home and mail back. to targets or fixation responses) under any of the experi- mental conditions were excluded from the analysis, three infants at 12 months, one at 18 months, three at Data analysis for video coding 24 months, and one at 36 months were deemed ineli- Videotapes of the eye movements were coded off-line by gible as they showed excessive fussiness or crying. two independent coders not directly involved in the ex- periment. The number of scorable trials was calculated Results for each participant of each age. On average, the follow- To treat the uneven distribution of the dependent vari- ing numbers of trials were excluded from the 24 trials ables, we applied a log linear transformation to latencies owing to the infant’s failure to fixate on the central facial and an arcsine transformation to response probabilities. stimuli before the target presentation: 2.95 ± 2.68 for In addition, a number of correlations were determined in 12-month-olds, 2.20 ± 2.75 for 18-month-olds, 2.26 ± 2.32 this study; hence, to control for Type 1 error, a p-value of for 24-month-olds, and 1.43 ± 1.75 for 36-month-olds. less than .01 was considered to indicate a statistically sig- Most of these cases involved failure to fixate on the first nificant correlation. central attractor presented for 7 sec. Table 1 presents the means and standard deviations of Of the scorable trials, those in which the infant gaze both the probabilities of responses to targets and those followed the direct path from the central face to the per- of fixation responses on the face in each experimental ipheral target were coded as responses toward targets. If condition across age. Since the probabilities of responses an infant did not move his/her eyes from the central face to targets (disengagement) were relatively low and they during the trial, the response was coded as a fixation re- varied across conditions, the latency data, which are sponse on the face. There were also failures to respond based on a low number of trials per cell and not com- to targets (e.g., infants looked to the opposite side of the parable across conditions, are of limited use for a longi- target or made eye movements with latencies < 200 ms tudinal analysis. Because the probability of fixation after the target onset). Probabilities of both the responses responses on the face (no movement) and that of toward the targets and fixation responses were calculated responses to the target are inherently inversely related, by dividing the number of times the response occurred we used fixations as a metric of infants’“failure to disen- by the number of scorable trials. gage.” This also excludes failures to respond to targets The reaction time (RT: in video frames, 33 ms per mentioned above. frame) was recorded by carefully choosing the first frame in which an eye movement to a target was detected. We Probabilities of fixation responses on the face defined response latency as the elapsed time between To confirm that infants have greater difficulty in disen- the presentation of the peripheral target and the begin- gaging attention from fearful faces than from happy or ning of the movement of the infant’s gaze toward the neutral faces, we conducted a 3 (facial expression: fear- peripheral target. Head movements were not coded in ful, happy, neutral) × 4 (age: 12, 18, 24, and 36 months) this study. within-subject analysis of variance (ANOVA; Table 1) on Table 1 Mean probabilities of responses to targets and fixation responses (Following Arcsine Transformation) Age Expression 12 months 18 months 24 months 36 months Probability of responses to peripheral target Fearful .410 (.392) .431 (.286) .513 (.373) .683 (.408) Happy .513 (.372) .507 (.241) .469 (.245) .733 (.450) Neutral .621 (.390) .600 (.258) .705 (.382) .745 (.368) Probability of remaining fixed Responses Fearful .522 (.376) .514 (.261) .546 (.316) .295 (.313) Happy .435 (.282) .395 (.156) .455 (.285) .234 (.181) Neutral .352 (.285) .328 (.148) .334 (.311) .260 (.224) Note. Standard deviations are given in parentheses. Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 5 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 Table 3 Intercorrelations between temperamental scores the probability of fixation responses on the face. The across time main effects of facial expression and age were signifi- 18 months 24 months 36 months cant (F(2, 30) = 12.91, p < .001; F(3, 45) = 2.98, p < .05, respectively). Infants fixated with greater probability Fear scale 12 months .477 .020 .359 on fearful (M = .472) than on neutral (M = .319; 18 months - .292 .338 p < .001) or happy faces (M = .382; p<.05).Further,no 24 months - - .218 significant difference was found between neutral and ** Average score of 12 months .584 .370 .344 happy conditions. Regarding the main effect of age, a scales loading on ** ** 18 months - .670 .578 significant difference existed only between 24 months Negative Affect ** 24 months - - .735 (M = .447) and 36 months (M = .263; p < .05). ** p < .01. Intercorrelations across ages .434 (SD = .254) for 18 months, .469 (SD = .312) for Table 2 presents the stability of individual variations in 24 months, and .238 (SD = .199) for 36 months. Table 4 attention disengagement (the probabilities of fixation presents the correlation coefficients between the index responses) from 12 to 36 months. We found significant of attentional bias to threat and the temperamental fear correlations between 18 and 24 months and between or negative affectivity scores on the IBQ-R or ECBQ 24 months and 36 months in the fearful condition. [22]. As a result, the relationship between the present at- Table 3 shows the correlations between temperamental tentional bias index toward threat and the degree of scores (Fear, Negative Affectivity) from 12 to 36 months. negative emotions was significant at 12 months (.527). Regarding Negative Affectivity, all correlations between The 95% confidence interval for this is .144 to .770. each age and the next proved significant and there was Results indicate that 12-month-old infants with more also a correlation between 18 and 36 months. Our negative affectivity show more difficulty in disengaging results demonstrated some stability in the temperamen- attention from fearful faces. Attentional bias to threat at tal scores over these ages. 12 months was related not to parent-reported fear but to a broad factor of negative affectivity. Relation between attentional bias and temperament As the response probability was relatively high at To examine the relation between individual differences 36 months (Table 1), we applied latency in the fearful in the attentional bias to fearful faces and fearfulness or expression condition as an attentional bias index at negative affectivity, correlation coefficients were calcu- 36 months (M = 1150.85 ms, SD = 47.04). While no rela- lated. The relevant index for attentional bias to threat tionship was observed with negative affect, we found could be response variables to fearful expression, con- that latency in the fearful expression condition was posi- trolling for the overall level of the infants’ failure to tively correlated with the scores on the Fear scale disengage from a face regardless of its expression. (r = .516, p = .012). Therefore, we calculate the following proportion score as an index of attentional bias to threat: the number of remaining fixed responses on fearful faces divided by the Moderating role of temperamental control total number of remaining fixed responses on faces. We examined whether temperamental control moder- Average indices were .363 (SD = .166) for 12 months, ates the relation between fearful attentional bias at 12 months and temperament at 3 years. In hierarchical multiple regression analyses, attentional bias to threat Table 2 Intercorrelations between probabilities of fixed (the number of fixed responses on fearful faces divided responses across time by the total number of fixed responses on faces) at 18 months 24 months 36 months 12 months, orienting score at 12 months, and effortful Fearful 12 months −.180 −.009 −.112 control score at 24 months were investigated as predic- ** 18 months - .587 .324 tors of the temperament at 36 months. Independent ** 24 months - - .577 variables were centered at their means prior to the ana- Happy 12 months .137 .112 .106 lysis. Table 5 summarizes these results. As can be seen in Table 5, in a main effects model of 18 months - .344 −.007 fear, effortful control only showed a trend toward relat- 24 months - - .113 ing to fearful temperament (b = −.77, t = −2.03, p = .059). Neutral 12 months −.132 −.036 −.254 Although the addition of interaction terms resulted in 18 months - .057 −.051 nonsignificant change in the model (ΔR = .140, p = .20), 24 months - - .492 only the attentional bias to threat × orienting interaction ** p < .01. term was marginal (b = 5.45, t = 1.80, p = .093). Thus, Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 6 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 Table 4 Correlation coefficients between temperamental scores and attentional bias index to fearful expression Attentional bias index to fearful expression 12 months 18 months 24 months 36 months Fear scale (IBQ-R or ECBQ) 12 months −.020 - - - 18 months .240 −.231 - - 24 months −.100 −.335 −.262 - 36 months .174 −.463 .105 .017 Average score of scales loading on Negative Affect 12 months .527** - - - 18 months .403 −.083 - - 24 months .222 −.286 −.023 - 36 months .309 −.276 −.125 .146 ** p < .01. toddlers who showed higher effortful control were less to fear at 36 months. Yet, as depicted in Figure 1, the fearful, but effortful control at 24 months did not mod- trend of the interaction effect indicates that the atten- erate how attentional bias to threat at 12 months related tional bias to threat at 12 months predicted fear at Table 5 Moderation Analysis: Temperamental control variables as moderators in the relationship between attentional bias to threat at 12 months and temperamental score of fear and negative affects at 36 months Predictor Fear at 36 months b(SE) β t 2 2 Step 1 (ΔR = .309, R = .309, F(3,19) = 2.38) Attentional Bias to threat at 12 M 1.05 (1.03) .21 1.01 Orienting at 12 M −.26 (.39) −.15 −.67 Effortful control at 24 M −.77 (.37) −.45 −2.03 2 2 Step 2 (ΔR = .140 R = .449, F(5,19) = 2.28) Attentional Bias to threat at 12 M 2.07 (1.35) .41 1.52 Orienting at 12 M −.15 (.38) −.08 −.39 Effortful control at 24 M −.95 (.40) −.56 −2.37 Attentional Bias to threat at 12 M × Orienting/Regulation at 12 M 5.45 (3.02) .44 1.80 Attentional Bias to threat at 12 M × Effortful control at 24 M 2.18 (5.19) −.11 .42 Predictor Negative affects at 36 months b(SE) β t 2 2 Step 1 (ΔR = .213, R = .213, F(3,19) = 1.44) Attentional Bias to threat at 12 M 1.03 (.69) .33 1.48 Orienting at 12 M −.12 (.26) −.11 −.47 Effortful control at 24 M −.30 (.25) −.28 −1.19 2 2 Step 2 (ΔR .0, R = .213, F(5,19) = .76) Attentional Bias to threat at 12 M 1.04 (1.01) .33 1.02 Orienting at 12 M −.12 (.28) −.11 −.43 Effortful control at 24 M −.30 (.30) −.28 −1.01 Attentional Bias to threat at 12 M × Orienting/Regulation at 12 M .06 (2.26) .00 .02 Attentional Bias to threat at 12 M × Effortful control at 24 M .03 (3.89) .00 .00 * † p < .05; p < .10. Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 7 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 the fixated responses to threat (Table 4). These results might be due to the development of attentional control over a relatively reactive component of temperament. 3.5 The subjects’ probability of responding to peripheral stimuli was higher at 36 months than earlier (Table 1), which is consistent with the results of ANOVA for prob- abilities of fixation responses on the face. Therefore, we 2.5 applied RT as an index and found a positive correlation between the Fear scale and RT in the fearful condition. Rothbart et al. [23] mentioned that younger infants present relatively undifferentiated distress, but later it is 1.5 possible to differentiate anger/frustration from fear. Thus, it is understandable that at an early stage, atten- tional bias is related not to fear but to overall negative Low Orienting Score 0.5 affectivity. These data also suggest that if we apply ad- High Orienting Score equate individual indices for each age, the relation be- 0 tween temperament and attentional bias to threat is Low High found across the age range of 12–36 months. In any Attentional Bias Attentional Bias case, just as Kiel and Buss [13] found attention toward Figure 1 Joint effect of attentional bias to threat and threat in the toddler years to be a predictor of social in- temperamental orienting score on 36-month fear. hibition in kindergarten, the present study revealed a predictive role of attentional bias to threat at 12 months. 36 months with a higher level of orienting score. With a Temperamental control systems may play an import- lower level of orienting score, the attentional bias to ant role in moderating the link between attentional bias threat at 12 months did not predict fear at 36 months. to threat and later fear or negative affectivity. Posner This trend is consistent with our expectation. Models et al. [20] argued that during infancy, control is princi- for negative affectivity are nonsignificant, and neither pally carried out by the orienting brain network, whereas the main effect nor the interaction term is significant. by three to four years of age this control shifts primarily to the executive network. With regard to effortful con- Discussion trol, we only found a main effect relating to fear at In this study, we investigated relations among disengage- 36 months. A negative relation between effortful control ment from threatening stimuli, negative affectivity, and and negative affect has been reported consistently [23]. effortful control in infants across the second and third The orienting attention network marginally moderated years of life. The infants demonstrated greater difficulty the association between early attentional bias to threat in disengaging their attention from fearful faces than and later fearful temperament (Figure 1). As expected, from happy or neutral faces. The ability to disengage attentional bias at 12 months seems to predict fear at from fearful faces at 12 months was significantly related 36 months for children who had high orienting scores to negative affect only at 12 months, although it was on the IBQ-R but not for those with low scores. Since positively related with negative affect across the age high orienting scores are partially attributed to a ten- range of 12–36 months (Table 4). In addition, our results dency to focus for a long period (measured by the sub- marginally indicate that it was not the executive network scale of the duration of orienting), toddlers who attend but rather the orienting network that moderated the as- to threat might be particularly likely to continue inter- acting with threatening stimuli, thus possibly increasing sociation between early attentional bias and later fear. These results indicate that infants with higher negative the likelihood that they will be fearful at 36 months. emotionality exhibit more extended dwell time for fear- One possible reason that we found no interaction of ef- fortful control is that at 36 months, both orienting and ful facial expressions at 12 months. Individuals with strong attentional bias to threat might experience more executive networks perform regulatory functions, while stimuli as aversive, or their negative reactivity might be the executive network was subdominant earlier in life. The current results were consistent with those of Pel- triggered by more stimuli [10]. In that case, earlier inter- vention—namely, the caregiver’s efforts in soothing at an tola et al. [5,6] and Leppänen et al. [7], indicating that, earlier stage—might be necessary for infants high in as in adults, attention disengagement was influenced by fearful faces in 7-month-old infants (e.g., average disen- negative affectivity. On the other hand, at 18, 24, and 36 months, no significant positive correlation was gagement latencies for fearful face, M = 674 ms; happy observed between temperament and the probability of face, M = 554 ms; neutral face, M = 540 ms; [6]). These Fear Nakagawa and Sukigara Behavioral and Brain Functions 2012, 8:40 Page 8 of 8 http://www.behavioralandbrainfunctions.com/content/8/1/40 studies showed the face alone for 1000 ms at the center References 1. Fox E, Russo R, Bowles RJ, Dutton K: Do threatening stimuli draw or hold before presenting the peripheral stimulus, which was ap- visual attention in sub-clinical anxiety? JExp PsycholGen 2001, 130:681–700. proximately 14 degrees to either the left or right of the 2. Georgiou GA, Bleakley C, Hayward J, Russo R, Dutton K, Eltiti S, Fox E: center. On the other hand, the current study presented Focusing on fear: attentional disengagement from emotional faces. Vis Cogn 2005, 12:145–158. the face alone for 200 ms in the center and the periph- 3. Derryberry D, Reed MA: Regulatory processes and the development of eral stimuli at approximately 30 degrees from the center. cognitive representations. Dev Psychopathol 1996, 8:215–234. Two hundred milliseconds is a psychophysical threshold 4. LoBue V, Deloache JS: Superior detection of threat-relevant stimuli in infancy. Deve Sci 2010, 13:221–228. of face visibility during infancy [24]. These differences in 5. Peltola MJ, Leppänen JM, Palokangas T, Hietanen JK: Fearful faces experimental conditions may have influenced RTs in the modulate looking duration and attention disengagement in experiment as it took longer for the participants to dis- 7-month-old infants. Dev Sci 2008, 11:60–68. 6. Peltola MJ, Leppänen JM, Vogel-Farley VK, Hietanen JK, Nelson CA: Fearful engage attention from the central faces (at 36 months, faces but not fearful eyes alone delay attention disengagement in fearful face, M = 1150 ms; happy face, M = 1023 ms; neu- 7-month-old infants. Emotion 2009, 9:560–565. tral face, M = 930 ms). 7. Leppänen JM, Peltola MJ, Mäntymaa M, Koivuluoma M, Salminen A, Puura K: Cardiac and behavioral evidence for emotional influences on attention The present results should be considered in light of sev- in 7-month-old infants. Int J Behav Dev 2010, 34:547–553. eral limitations. First, because many of the infants partici- 8. Hazen RA, Vasey MW, Schmidt NB: Attentional retraining: a randomized pated in very few trials, our data may not adequately clinical trial for pathological worry. J Psychiat Res 2009, 43:627–633. 9. MacLeod C, Rutherford E, Campbell L, Ebsworthy G, Holker L: Selective represent a child’s attentional bias or difficulty to disen- attention and emotional vulnerability: Assessing the causal basis of their gage from a threat. Second, to control for Type I error association through the experimental manipulation of attentional bias. regarding correlation coefficients, we decide to treat a J Abnorm Psychol 2002, 111:107–123. 10. Lonigan CJ, Vasey MW, Phillips BM, Hazen RA: Temperament, anxiety, and p-value less than .01 as statistically significant. However, the processing of threat-relevant stimuli. J Clin Child Adolescent Psychol this is less conservative than applying the Bonferroni cor- 2004, 33:8–20. rection. Further, our sample size was rather small. 11. Fox NA, Hane AA, Pine DS: Plasticity for affective neurocircuitry. Curr Dir Psychol Sci 2007, 16:1–5. 12. Bar-Haim Y, Lamy D, Pergamin L, Bakermans-Kranenburg MJ, van IJzendoorn MH: Threat-related attentional bias in anxious and nonanxious Conclusion individuals: a meta-analytic study. Psychol Bull 2007, 133:1–24. 13. Kiel EJ, Buss KA: Toddlers’ duration of attention toward putative threat. In summary, the longitudinal study of infants in the age Infancy 2011, 16:198–210. range of 12–36 months revealed that the attentional bias 14. Pérez-Edgar K, Reeb-Sutherland BC, McDermott JM, White LK, Henderson to threat is preserved over time in infancy and childhood. HA, Degnan KA, Hane AA, Pine DS, Fox NA: Attention biases to threat link behavioral inhibition to social withdrawal over time in very young The results also indicated that individual differences in the children. J Abnorm Child Psychol 2011, 39:885–895. strength of that bias depend upon parent-reported tem- 15. Rothbart MK: Temperament and the development of inhibited approach. peramental negative affectivity or fearfulness in early life. Child Dev 1988, 59:1241–1250. 16. Nakagawa A, Sukigara M: How are cultural differences in the Moreover, the attentional bias to threat may possibly interpretation of infant behavior reflected in the Japanese revised Infant interact with the temperamental orienting network. Fu- Behavior Questionnaire? Jpn J Educ Psychol 2005, 53:491–503. ture research should examine whether individual differ- 17. Sukigara M, Nakagawa A, Mizuno R: Studying toddler temperament via Japanese Early Childhood Behavior Questionnaire [abstract]. Jnp Psychol ences in attentional bias to threat emerging early in Ass 2007, 71:1095. infancy predict later emotional traits. Investigation of the 18. Posner MI, Rothbart MK: Educating the human brain. Washington, DC: period before the effortful control becomes dominant American Psychological Association; 2007. 19. Lonigan CJ, Vasey MW: Negative affectivity, effortful control, and might provide valuable information for parenting, early attention to threat-relevant stimuli. J Abnorm Child Psychol 2009, education, and child psychiatry. 37:387–399. 20. Posner MI, Rothbart MK, Sheese BE, Voelker P: Control networks and Competing interests neuromodulators of early development. Dev Psychol 2012, 48:827–835. The authors declare that they have no competing interests. 21. Matsumoto D, Ekman P: Japanese and Caucasian Facial Expressions of Emotion (JACFEE) and Neutral Faces (JACNeuF). San Francisco, CA: University Authors’ contributions of California, Human Interaction Laboratory; 1988. Both authors conceived of the study and conducted the experiments. MS 22. Putnam SP, Ellis LK, Rothbart MK: The structure of temperament from administered the instruments and analyzed data. AN mainly collected data infancy through adolescence.In Advances in research on temperament. and prepared the initial draft of the manuscript. Both authors approved the Edited by Eliasz A, Angleitner A. Germany: Pabst Scientist Publisher; final manuscript. 2001:165–182. 23. Rothbart MK, Sheese BE, Rueda MR, Posner MI: Developing mechanisms of Acknowledgements self-regulation in early life. Emot Rev 2011, 3:1–7. The present study was supported by a Grant-in-Aid (No. 17330143) for 24. Gelskov SV, Kouider S: Psychophysical thresholds of face visibility during Scientific Research from the Ministry of Education, Culture, Sports, Science infancy. Cognition 2010, 114:285–292. and Technology of Japan. The authors thank all the infants and families who generously took part in this project. Oshio Shinji is gratefully acknowledged doi:10.1186/1744-9081-8-40 for providing statistical support. We are also indebted to Ms. Kazuyo Cite this article as: Nakagawa and Sukigara: Difficulty in disengaging Kanayama for her assistance in conducting the longitudinal research. from threat and temperamental negative affectivity in early life: A longitudinal study of infants aged 12–36 months. Behavioral and Brain Received: 15 February 2012 Accepted: 9 August 2012 Functions 2012 8:40. Published: 14 August 2012
Journal
Behavioral and Brain Functions – Springer Journals
Published: Aug 14, 2012