Transgenerational effects of maternal depression on affect recognition in children

Highlights

• In remission, mothers with depression show a negative processing bias.

• Their children show parallel alterations in their affect recognition.

• Mothers with remitted depression with or without history of abuse do not differ.

Abstract

Background

The association between maternal depression and adverse emotional and behavioral outcomes in children is well established. One associated factor might be altered affect recognition which may be transmitted transgenerationally. Individuals with history of depression show biased recognition of sadness. Our aim was to investigate parallels in maternal and children's affect recognition with remitted depressed mothers.

Methods

60 Mother–child dyads completed an affect recognition morphing task. We examined two groups of remitted depressed mothers, with and without history of physical or sexual abuse, and a group of healthy mothers without history of physical or sexual abuse. Children were between 5 and 12 years old.

Results

Across groups, mothers identified happy faces fastest. Mothers with remitted depression showed a higher accuracy and response bias for sadness. We found corresponding results in their children. Maternal and children's bias and accuracy for sadness were positively correlated. Effects of remitted depression were found independent of maternal history of physical or sexual abuse.

Limitations

Our sample size was relatively small and further longitudinal research is needed to investigate how maternal and children's affect recognition are associated with behavioral and emotional outcomes in the long term.

Conclusions

Our data suggest a negative processing bias in mothers with remitted depression which might represent both the perpetuation of and vulnerability to depression. Children of remitted depressed mothers appear to be exposed to this processing bias outside acute depressive episodes. This may promote the development of a corresponding processing bias in the children and could make children of depressed mothers more vulnerable to depressive disorders themselves.

Introduction

There is ample empirical evidence for an association between maternal depression and a range of adverse child behavioral (Lee and Gotlib, 1991) and emotional outcomes (for review, see: Beardslee et al., 2011, Goodman et al., 2011). Children of depressed parents often show diminished affect regulation and cognitive and language functioning when their mothers are depressed (e.g., Carter et al., 2001; Cicchetti et al., 1998; Rohde et al., 2005). They are at risk for developing depressive disorders themselves (Sim and England, 2009) and other internalizing and externalizing disorders (Lyons-Ruth et al., 1997). However, the factors contributing to the transgenerational transmission of depression have not been fully elucidated. One psychological factor contributing to the risk of transmission might be altered affect recognition in mothers, such as difficulties in correctly interpreting and responding to facial expressions.

In adults, a number of studies have investigated alterations in facial affect recognition in depression. These studies consistently showed an increased accuracy of, and increased response bias for sad faces (Arteche et al., 2011, Gollan et al., 2010, Gollan et al., 2008, Gur et al., 1992, Hale, 1998, Lee et al., 2008, Mandal and Bhattacharya, 1985, Sterzer et al., 2011, Surguladze et al., 2004). These emotional-cognitive deficits seem to persist during remission (Bhagwagar and Cowen, 2008, Bouhuys et al., 1999, Fuhr et al., 2014, Joormann and Gotlib, 2007, Lange et al., 2012). The question arises whether biased affect recognition might be transmitted from affected mothers to their children. Investigations addressing this question have only recently begun and suggest that alterations in affect recognition may also exist in children of depressed mothers. The alterations seem mainly to concern sad affects (Joormann et al., 2010; Lopez‐Duran et al., 2013). To the best of our knowledge, to date no studies exist that have directly investigated parallels in affect recognition in mother–child dyads.

We here studied affect recognition in both mothers with remitted depression and their children to address the question of whether biased affect recognition in remitted depressed mothers corresponds with respective alterations in their children.

Patients with depression often report a history of traumatization, such as physical or sexual abuse (Heim et al., 2008, Lizardi et al., 1995, McCabe et al., 2009). Based on neurobiological (Heim et al., 2008, Vythilingam et al., 2002) and treatment studies (Nemeroff et al., 2003), it has been suggested that depressed patients with and without trauma represent two distinct subgroups of depression. This distinction may be critical for the investigation of affect recognition, as history of trauma has – independently of depression - been associated with altered affect recognition. However, these alterations seem to concern biased anger and fear processing mainly (Dannlowski et al., 2012, Poljac et al., 2011, Rauch et al., 2000, Shin et al., 2005, van Harmelen et al., 2010). To account for effects of trauma, we here distinguished between remitted depressed mothers with and without history of trauma (i.e., physical or sexual abuse).

We tested two hypotheses in the present study. First, we predicted that mothers with remitted depression would exhibit higher accuracy and a response bias for sad faces. Second, we hypothesized that children of remitted depressed mothers would show corresponding alterations.

To study these hypotheses we performed an affect recognition task in mothers with remitted depression and their children. We tested for parallels between mothers and children both at group and at dyadic level. At group level, we investigated whether group differences observed between depression and control mothers were analogously found in children. At dyadic level, we examined whether maternal and children's abilities in affect recognition for sad facial expressions were correlated.