Neuroscience of human social interactions and adult attachment style

Abstract

Since its first description four decades ago, attachment theory (AT) has become one of the principal developmental psychological frameworks for describing the role of individual differences in the establishment and maintenance of social bonds between people. Yet, still little is known about the neurobiological underpinnings of attachment orientations and their well-established impact on a range of social and affective behaviors. In the present review, we summarize data from recent studies using cognitive and imaging approaches to characterize attachment styles and their effect on emotion and social cognition. We propose a functional neuroanatomical framework to integrate the key brain mechanisms involved in the perception and regulation of social emotional information, and their modulation by individual differences in terms of secure versus insecure (more specifically avoidant, anxious, or resolved versus unresolved) attachment traits. This framework describes how each individual's attachment style (built through interactions between personal relationship history and predispositions) may influence the encoding of approach versus aversion tendencies (safety versus threat) in social encounters, implicating the activation of a network of subcortical (amygdala, hippocampus, striatum) and cortical (insula, cingulate) limbic areas. These basic and automatic affective evaluation mechanisms are in turn modulated by more elaborate and voluntary cognitive control processes, subserving mental state attribution and emotion regulation capacities, implicating a distinct network in medial prefrontal cortex (mPFC), superior temporal sulcus (STS), and temporo-parietal junction (TPJ), among others. Recent neuroimaging data suggest that affective evaluation is decreased in avoidantly but increased in anxiously attached individuals. In turn, although data on cognitive control is still scarce, it points toward a possible enhancement of mental state representations associated with attachment insecurity and particularly anxiety. Emotion regulation strategies such as reappraisal or suppression of social emotions are also differentially modulated by attachment style. This research does not only help better understand the neural underpinnings of human social behavior, but also provides important insights on psychopathological conditions where attachment dysregulation is likely to play an important (causal) role.

Keywords: adult attachment style; cognitive affective neuroscience; emotional versus cognitive mentalization; functional neuroanatomical framework; human social interactions.

Figure 1

Functional neuroanatomical model of the influence of adult attachment style on social processing. Two core component networks mediate relatively automatic affective evaluations versus more controlled cognitive processes, broadly corresponding to emotional versus cognitive mentalization mechanisms proposed in other models (Fonagy and Luyten, 2009). The affective evaluation component further comprises social approach (purple) versus aversion (blue) systems, whereas the cognitive control component comprises distinct systems implicated in emotion regulation (orange) and mental state representation (red). We assume “push-pull” effects between approach versus aversion modules (green arrow), which might be jointly influenced by learning as well as genetic factors (e.g., neuromodulator systems listed in the gray box). In addition, more complex reciprocal influence may exist between the affective evaluation and cognitive control components (turquoise arrows). The possible influence of attachment avoidance (AV) or anxiety (AX) on activity of each of these networks is depicted by (downward or upward) arrows (red boxes) representing relative hypo- or hyper-activation, respectively. For details, please refer to text. (DL)PFC = (dorsolateral) prefrontal cortex; OFC = orbitofrontal cortex; (p)STS = (posterior) superior temporal sulcus; TPJ = temporo-parietal junction; aSTG = anterior superior temporal gyrus.

Figure 2

Modulation of social aversion and approach activations by adult attachment style. Adapted from Vrtička et al. (2008). (A) Participants performed a visual task, while receiving feedback from virtual partners about their performance. Feedback was composed of words reflecting outcome (“Won” if correct and “Lost” if incorrect response was given), associated with either smiling or angry faces, inducing the perception of supportive “friends” (congruent word-face combinations) or hostile “opponents” (incongruent combinations). (B) Top: reward-related areas (left: ventral striatum; right: ventral tegmental area) were activated during the perception of positive social feedback (“Won” paired with a smiling face; SF-W), but this effect was modulated by avoidant attachment style. Bottom: negative relation between avoidant attachment style (AVS) scores and the ventral striatum response. (C) Top: the central amygdala was activated by the perception of social punishment (“Lost” paired with an angry face; AF-L), and this effect was modulated by anxious attachment style. Bottom: positive relation between anxious attachment style (AXS) scores and amygdala response. BOLD signal is depicted in arbitrary units.

Figure 3

Modulation of social emotion perception and regulation by adult attachment style. Adapted from Vrticka et al. (2012a). (A) Bilateral amygdala activation to social (versus nonsocial) emotional scenes perception, regardless of valence (positive or negative) and task. (B) Positive correlation between avoidant attachment (AV) scores and activity in left amygdala for social negative images during reappraisal (averaged across voxels). (C) Median split illustrating data for high (red; N = 5) versus low (blue; N = 8) avoidantly attached participants in left amygdala, showing a decrease in activity to social negative images during reappraisal for the low but not high avoidant group (^**indicate the differential response accounting for significant effects in the correlation analysis). (D) Right amygdala activation to negative scenes showing a significant modulation by anxious attachment (AX) scores during natural viewing conditions. (E) Negative correlation between anxious attachment scores and response to nonsocial negative scenes in the right amygdala. (F) Median split illustrating data for high (red; N = 8) versus low (blue; N = 9) anxiously attached participants in right amygdala, showing that activation to negative scenes was greater for nonsocial content in those with lower AX scores, but greater for social content in those with higher AX scores. (^**indicate the differential response accounting for significant effects in the whole-brain correlation analysis). NAT = natural viewing, REAP = reappraisal, ESUP = expressive suppression. BOLD signal is depicted in arbitrary units, and error bars represent +/− 1 standard error from mean.