Visual-spatial attention systems can reportedly detect threat-related stimuli rapidly. The propensity to quickly detect the presence of threatening stimuli, such as snakes and angry faces, may be an important survival and adaptive mechanism. Threat-related stimuli (e.g., threat words or angry faces) may also cause a delay in disengagement [1
], a tendency possibly increased by an individual’s elevated level of state anxiety. Further, using fearful facial expressions as stimuli, Georgiou et al. [2
] showed that high trait-anxious people exhibited extended dwell time to threat-related stimuli. The inability to rapidly disengage from threat-related stimuli may keep cognitive resources focused on the stimuli and result in increased anxiety [1
]. This trend might influence subsequent cognitive and emotional processing, which is likely to play an important role in shaping children’s cognitive representations of themselves, others, and the situation, from their earliest years [3
Even in infancy, humans have been found to orient more quickly to threatening than to nonthreatening stimuli [4
]. Recent studies have demonstrated that 7-month-olds disengaged their fixation significantly less frequently from fearful faces than from happy faces and control stimuli [5
]. Moreover, Peltola et al. [6
] found that the delayed withdrawal of attention reflected not a simple response to fearful wide-open eyes but rather an enhanced sensitivity to facial signals of threat. Fearful expressions also caused greater heart rate deceleration responses in 7-month-old infants during the first 1000
ms of face viewing [7
]. Leppänen et al. [7
] concluded that emotion–attention interactions such as those displayed by adults can also be observed early in life.
Growing evidence suggests that attentional bias to threat plays a causal role in individual differences in emotional vulnerability [8
]. Lonigan et al. [10
] considered temperament’s contributions to childhood disorders. They extracted factors similar to those drawn previously from childhood self-reported items that assess emotionality and attention and described automatic attention allocation mechanisms linked to negative affectivity, which may have an effect on both the daily experiences of children and their proneness to future negative experiences. Thus, the association between negative affectivity and anxiety pathology could be mediated by the attentional bias to threat.
Going beyond the individual mechanism of anxiety-related information processing, Fox et al. [11
] proposed a model of plasticity for affective neurocircuitry, describing how genetic disposition and environmental circumstances may interact. Thus, a child’s fearful temperament elicits and is elicited by the caregiver’s insensitivity and intrusiveness to shape the attentional bias to threat and the neural systems involved in this bias (i.e., ventral prefrontal cortex-amygdala circuitry). Fox et al. [11
] further suggested that exaggerated attentional bias to threat may cause the emergence and maintenance of anxious behaviors.
On the basis of a meta-analysis of 172 studies, Bar-Haim et al. [12
] pointed out that although attentional bias to threat may largely contribute to the development and maintenance of anxiety over time, the possibility of a causal link between the two has been insufficiently investigated. Recently, longitudinal studies of very young children have examined the relation between attentional bias to threat and later socioemotional outcome or risk of psychopathology [13
]. Results indicate the moderating roles that attention played in anxiety development. However, the youngest participants in these studies were 24
months old [13
], and the dot-probe task used for very young children did not necessarily measure the ability to disengage [14
]. Therefore, further research is needed to fill these gaps in the previous studies in order to understand the initial structure and function of anxiety-related information processing.
For the reasons stated above, we conducted a longitudinal study of infants approximately 12–36
months old. Fear develops by the end of the first year of life, and fearful infants show inhibition of motor approach [15
]; hence, we studied infants from the end of their first year. The purpose of this study was, first, to confirm the infants’ greater difficulty in disengaging attention from fearful faces than from happy or neutral faces [5
]. In the overlap task that we used following Peltola et al. [5
], infants were required to disengage their fixation from a centrally presented facial expression and shift attention to a peripheral target.
Second, we examined the relationship between individual differences in fear or negative affectivity and the attentional bias to threat in early life. For this purpose, we examined individual differences in fear or negative affectivity through a revised Infant Behavior Questionnaire (IBQ-R Japanese version [16
]) administered at 12
months and the Early Childhood Behavior Questionnaire (ECBQ Japanese version [17
]) administered at ages 18–36
months. Since an increased number of fixation responses (i.e., no movement) with a centrally presented fearful face has been given as evidence of the effect of fearful faces on attentional disengagement [5
], we computed an index of the attentional bias to threat-related stimuli on the basis of these fixation responses. Negative affectivity (i.e., the reactive component of temperament) is considered relatively easy to change early in life. Effortful control reflects a voluntary component of attention and undergoes significant development in the second
year of life and later [18
]. In the first year of life, the association between a high level of negative affectivity and an attentional bias to threat would involve a mainly reactive temperamental component and would be easier to notice. In the second and third years, however, this connection might be modified by control systems such as effortful control. Children and adolescents, who are considered high in negative affectivity and low in effortful components of temperament, are reported to demonstrate a significant attentional bias in favor of threat stimuli [19
Third, we conducted a moderation analysis to examine if an effortful control moderates the link between attentional bias at 12
months and temperament at 36
months. In infancy, a brain network involved in orienting to sensory events may provide the chief means of self-regulation. This orienting network involves areas of the inferior and superior parietal lobes and the frontal eye fields. Later in childhood, however, the executive attention system, including the anterior cingulate, insula, and areas of the basal ganglia, becomes dominant as a mechanism of self-regulation [20
]. Effortful control as a temperamental construct is considered to reflect the functioning of a neutrally based executive attention. The IBQ-R orienting score measures an orienting attentional network, and the ECBQ effortful score assesses an executive attention network. Therefore, we included both the orienting score of the IBQ-R (at 12
months) and the effortful control score of ECBQ (at 24
months). The interactive (moderating) effect was tested by using centered variables in hierarchical regression.
We hypothesized that effortful control buffers the link between attentional bias and the variables of fear or negative affect. Specifically, we predicted that the correlation between the 12-month attentional bias score and 36-month temperament score (fear or negative affectivity) would be stronger among toddlers low in effortful control. On the other hand, toddlers with high orienting scores might interact in the opposite manner. Because high scores on the latter are partially due to a tendency to focus for long periods, toddlers who attend to threat might be particularly likely to continue focusing on threatening stimuli, thus possibly increasing the likelihood that they will be fearful at 36