Maternal depression is associated with difficulties in understanding and adequately responding to children’s emotional signals. Consequently, the interaction between mother and child is often disturbed. However, little is known about the neural correlates of these parenting difficulties. Motivated by increasing evidence of the amygdala’s important role in mediating maternal behavior, we investigated amygdala responses to child sad and happy faces in mothers with remitted major depression disorder (rMDD) relative to healthy controls.

We used the sensitivity subscale of the emotional availability scales and functional magnetic resonance imaging in 61 rMDD and 27 healthy mothers to examine the effect of maternal sensitivity on mothers’ amygdala responses to their children’s affective facial expressions.

For mothers with rMDD relative to controls, we observed decreased maternal sensitivity when interacting with their child. They also showed reduced amygdala responses to child affective faces that were associated with lower maternal sensitivity. Connectivity analysis revealed that this blunted amygdala response in rMDD mothers was functionally correlated with reduced activation in higher-order medial prefrontal areas.

Our results contribute toward a better understanding of the detrimental effects of lifetime depression on maternal sensitivity and associated brain responses. By targeting region-specific neural activation patterns, these results are a first step toward improving the prediction, prevention, and treatment of depression-related negative effects on mother–child interaction.

Fear responses are particularly intense and persistent in post-traumatic stress disorder (PTSD), and can be evoked by unspecific cues that resemble the original traumatic event. Overgeneralisation of fear might be one of the underlying mechanisms. We investigated the generalisation and discrimination of fear in individuals with and without PTSD related to prolonged childhood maltreatment.

Sixty trauma-exposed women with (N = 30) and without (N = 30) PTSD and 30 healthy control participants (HC) underwent a fear conditioning and generalisation paradigm. In a contingency learning procedure, one of two circles of different sizes was associated with an electrical shock (danger cue), while the other circle represented a safety cue. During generalisation testing, online risk ratings, reaction times and fear-potentiated startle were measured in response to safety and danger cues as well as to eight generalisation stimuli, i.e. circles of parametrically varying size creating a continuum of similarity between the danger and safety cue.

The increase in reaction times from the safety cue across the different generalisation classes to the danger cue was less pronounced in PTSD compared with HC. Moreover, PTSD participants expected higher risk of an aversive event independent of stimulus types and task.

Alterations in generalisation constitute one part of fear memory alterations in PTSD. Neither the accuracy of a risk judgement nor the strength of the induced fear was affected. Instead, processing times as an index of uncertainty during risk judgements suggested a reduced differentiation between safety and threat in PTSD.