The study participants lived in three cities in the State of São Paulo and presented a similar profile to that of other national studies evaluating pregnant and puerperal women (Beltrami, Moraes, & Souza, 2013; Morais, Lucci, & Otta, 2013).
Mothers with depressive symptoms presented lower levels of education and fewer of them lived with a partner compared with anxious mothers and the control group. The association between depressive symptoms, both prenatal and postnatal, and lower levels of education and socioeconomic status is recurrent in the literature, as are conflictive marital relationships and the lack of a partner (Gelaye et al., 2016; Morais et al., 2013).
In the interactive episode, reciprocal relationships were observed between the children and their mothers, confirming the mutuality and bidirectional character of the interaction during the first year of life (Cassiano & Linhares, 2015; Feldman, 2015). More sensitive, stimulating, and positively affective mothers had children that were more involved and integrated in the interaction. They displayed more positive affect with smiles, enthusiasm, and physical displays of affection and few episodes of negative affect, such as crying, negative vocalizations, anger, and hostility. Conversely, the children of more disengaged and intrusive mothers interacted less and were less involved. According to Piccinini et al. (2014), the intrusive mother does not respect the child’s autonomy, attempts to control their behavior with unnecessary interventions, and ends up impairing the experience of new acquisitions.
Corroborating other studies (Cassiano & Linhares, 2015; Ribeiro et al., 2014), behaviors of positive affect predominated in the recorded episodes, with few episodes of negative affect, on the part of both the mother and the child. It seems probable that minimal hostility and demonstrations of positive affect are characteristic of mother-child interactions in the age range studied (Cassiano & Linhares, 2015); however, since the observations occurred in the presence of a stranger, in particularly small rooms, this could have inhibited the emission of socially censored negative affect behaviors.
Concerning the relationship between mental health and the interaction, the initial hypothesis was confirmed. There were significant differences in the behaviors emitted during the episode of interaction by mothers with chronic depressive symptoms, those with chronic anxiety symptoms, and control group mothers. Mothers of the group with chronic depressive symptoms were significantly less sensitive, cognitively stimulated their children less, had more restrained demonstrations of positive affect, and presented more negative affect than those in the anxiety and control groups. When asked to play with their children as they would normally, they often stated that they did not play with them. This profile corresponds to the description of mothers with chronic depressive symptoms reported in the literature: mothers unresponsive or not attentive to the needs of the child and rarely ever involved in close exchanges (Herba, Glover, Ramchandani, & Rondon, 2016; Murray et al., 2010). However, the data also showed that mothers with chronic depressive symptoms were more intrusive than those in the other two groups. Servilha and Bussab (2015) observed that mothers with postpartum depression presented more intrusive behaviors and were less responsive, which made the contacts of these dyads asynchronous regarding affection and attention. The literature reports that mothers with symptoms of depression can present different styles of interaction, with some being disengaged, uninvolved in interactions, less communicative, and showing less synchronous interaction, although they are more empathic with the child (Cornish et al., 2005; Feldman et al., 2009). Others are intrusive, are less affective when dealing with their children, and are more hostile, impatient, and coercive (Murray et al., 2010; Piccinini et al., 2014). There is a third group of mothers who, despite the depression, still find pleasure in their relationship with the child, with a moderate amount of smiles, touches, and responsive behavior, thus impairing the interaction less (Fonseca, Silva, & Otta, 2010).
In contrast, mothers with chronic anxiety symptoms presented interactive behaviors similar to those of the control group. Studies related to the interactive characteristics of anxious mothers are still scarce and the results are inconsistent. Some authors reported that anxious mothers were less sensitive than those in the control group and more intrusive than depressed mothers (Feldman et al., 2009; Nardi, Rodrigues, Melchiori, Salgado, & Tavano, 2015). Kaitz et al. (2010) did not observe sensitivity deficits and intrusiveness in anxious mothers, only that they acted in a more exaggerated way with their 6-month-old children in the free play situation. Murray, Cooper, Creswell, Schofield, and Sack (2007) also observed small differences in some interactive situations of mothers with social phobia and their 10-week-old children, but not in others. In this study, although mothers with chronic anxiety symptoms scored higher in the intrusiveness category, the scores did not differ significantly from those of mothers without a mental disorder. However, the anxious mothers provided their children with more cognitive stimulation than the depressed mothers, with possible developmental gains, as previously observed by de Fraga, Linhares, Carvalho, and Martinez (2008).
The children of mothers with chronic depression symptoms interacted significantly less than those of the other two groups, in agreement with other studies (Feldman et al., 2009; Murray et al., 2010). Research on how children of depressed mothers behave during the interaction presents contradictory results. In her review article, Field (2010) concluded that disturbances in the child’s interaction with mothers who have symptoms of depression are a universal phenomenon. Other studies, however, have shown that children of depressed mothers can maintain the same quality of interaction as those of mothers without such symptoms (Fonseca et al., 2010; Ribeiro et al., 2014). In this study, the chronicity of the condition may have acted as an aggravating factor for the lower percentage of interactive behaviors of the children of depressed mothers (Stein et al., 2014). Similarly, Cornish et al. (2005) observed that mothers with chronic depression had a negative perception of their children, complained about having to take care of them, and that their hostile behaviors inhibited the child’s participation.
Although the children of mothers with symptoms of depression interacted less, there were no significant differences among the three groups regarding the other interactive characteristics of the children. Regardless of maternal mental health, the children of the three groups presented demonstrations of positive affect (positive vocalizations, smiles, laughter, hugs, and kisses), good involvement, and minimal negative affect, with rare demonstrations of crying, agitation, discontent, or hostility. The children’s involvement in the interaction, regardless of the mother’s mental health, with numerous demonstrations of positive affect, seems to indicate that the relationship between the mother with depression and the child has no linear causality. It is a complex process and the final outcome involves other variables, including the individual characteristics of the mother and of the child. It is also influenced by the surrounding environment, for example, the presence of the father, which was the case for the majority of the children in this study (Goodman et al., 2011; Mendonça, Bussab, Lucci, & Kärtner, 2015; Morais et al., 2013).
If evidence in the literature indicates that the lack of a partner aggravates the effects of the maternal depressive condition (Lovejoy, Graczyk, O’Hare, & Neuman, 2000; Murray et al., 2010; Stein et al., 2014), Goodman et al. (2011) showed that the presence of the father alleviated the overload related to child care of the depressed mother and could represent a potentially healthy alternative form of care for the child. Similarly, Parfitt et al. (2014) and Mendonça et al. (2015) observed that the presence of a father who was involved with the child could compensate for and reduce the impact of maternal depression on the child.
Another variable that may have favored the involvement of the children of the three groups in the interaction was their age at the time of the evaluation. Lovejoy et al. (2000) observed that although younger children were more sensitive to the effects of maternal disengagement, older children, around one and a half years of age, were less dependent on maternal initiatives and were already able to establish some social reciprocity. In pleasurable situations, as in free play, they initiated the interaction, compensating for the lack of maternal involvement and creating conditions for the mother to participate in the activities. In this study, faced with the disengagement of the mother, the children often used gestures and vocalizations to attract her attention. According to Feldman (2015), these behaviors can elicit more parental investment and better interaction and development conditions.
According to Morais et al. (2013) and Goodman et al. (2011), full comprehension of the involvement of children of mothers with depression in the interaction requires that we consider two other variables: resilience and temperament. The resilience of the child has shown compensatory effects in adverse environments when the child was cared for by mothers with chronic depression (Frizzo & Piccinini, 2007). Regarding temperament, there is strong evidence that parental care does not produce the same effects in all children; some are more susceptible to adverse effects and inadequate care from the parents, as they benefit more from protective environments (Parfitt et al., 2014). It has also been identified that children with easy temperaments are less vulnerable to inadequate parental care provided by mothers with depression (Goodman et al., 2011). Studies on the differential susceptibility of children to adverse factors and their role in the interaction, especially in relation to maternal depression, are relatively recent, and new research should assist in identifying children who are more vulnerable so that interventions can be more accurately directed.
Some limitations need to be taken into consideration in the analysis of the results. Regarding the size of the sample, despite various efforts, there were several losses in the selection of the participants and few mothers presented the stipulated criterion of chronicity. The small number of participants limited the possibility of some statistical analyses, such as structural equation modeling, which could provide a better understanding of the complementary effects between the variables, from a bidirectional perspective.
Regarding the interactive episode, the analyses were limited to a single episode for each dyad, which does not allow the behaviors observed to be classified as stable relationship styles. Some mothers with chronic depressive symptoms, on other occasions, could possibly show greater sensitivity toward the child, avoiding impairments in the interaction due to the depression. However, the instrument used to analyze the interaction showed a high level of concordance between observers and it was considered adequate for observing dyads with children of this age group, allowing the bidirectional relationship between the partners of the dyad to be captured.
Regarding maternal mental health, mothers with chronic depressive symptoms also scored high for anxiety. The overlap between symptoms of anxiety and depression is recurrent in the literature, which possibly reflects the psychometric limitations of the scales used for the evaluation, especially the STAI and BDI, which present high correlation (Andrade & Gorenstein, 2008). It can also be assumed that anxiety and depression are components of the same process of general psychological stress, since there is data showing that 30% to 58% of patients with symptoms of depression present comorbidity with symptoms of anxiety (Field, 2010). Since comorbidity in the group of mothers with depressive symptoms may have affected the comparison between the groups, a group of mothers with chronic anxiety symptoms without comorbidity was included. The results of this group reinforce the data that describe anxious mothers as more intrusive, though with a profile very close to that of mothers without mental health problems, which causes less impairment in the development of their children.