Monthly Archives: December 2012

How to Talk to Your Child About Complex Subjects


At a gathering of family and close friends recently, a young couple asked me a question about their almost 3-year old son. I had observed the child and had found him to be intelligent, charming, and warmly connected to his parents. He also seemed sensitive, perhaps one of Kagan’s “inhibited” children (Kagan & Moss, 1983). The child, whom I will call “David” (not his real name), had been anxiously asking his parents about whether he might be put in jail, about whether he was a “bad guy”. He seemed to associate these fears to bible stories he was hearing in Sunday school, particularly the stories of Daniel and the Lion’s Den, and the story of Jesus being arrested. This was confusing to his parents, because in their understanding of these narratives, it was the “good guys” who were arrested. His parents told me that they constantly reassure him that he is a good boy, that they love him, and that he is safe. David’s parents further explained that he seems to ask these questions about being thrown in jail whenever a stranger is in the house, asking whether that person is going to throw him in jail. His parents decided to avoid stories about jail and to limit the bible stories since many of the concepts seemed too complex for him to process at this age. They asked me what I thought.

This question is fascinating from several points of view. First of all, it highlights the difference between the meanings an adult makes of certain narratives and those made by a preschool child. Second, it underscores the often discrepant levels of maturation of different developmental capacities in the same child. Third, it reveals aspects of a crucial cultural context that forms the way narratives transmit important beliefs and values in a society. Consideration of these factors may help parents in their efforts to talk to their children about complex subjects. 

What about what Tronick calls the “age possible” meanings that two people of different developmental ages make of the same story (Tronick, 2007)? An example is offered by the 4-year old whose mother was trying to explain to her the generational relationship of people at a family reunion. The mother explained, “Your nana is your daddy’s mommy.” The little girl thought for a moment and then asked in wonder, “But how did she get him into the car seat?” I am also reminded of a little patient, a 3-year old who witnessed people jumping from the World Trade Towers on television and in a play session suggested to me that children could “jump big” in a playground because it is “softer” (Harrison & Tronick, 2007). I took her to be referring to a “soft landing”, though I knew that no number of soft mattresses at the foot of the WTT could have cushioned the fall of the jumpers. In David’s case, the meaning he derived from “being put in jail” was that you were a bad guy, period. He was not able to consider a nuanced meaning in which a good guy was unjustly jailed. It is clear that good guys and bad guys are on his mind these days. Remember the “terrible two’s”?. His age-typical anxiety about the result of noncompliance to parental demands (whether real acts of noncompliance or imagined ones) led him to fear that his “bad guy” feelings and thoughts would brand him as a bad guy and cause someone to throw him in jail. A “stranger” is a preferable enforcer of that terrible punishment, because a stranger can be seen as “all bad”.  If it were one of his beloved parents who threatened him with jail, how could he manage the stress of fearing one that he also loved and depended on? 

The second issue is that of discrepant developmental capacities. Human development is not a smooth, linear process. It occurs in a messy process of hits and misses, halts and bumps forward, and reiterative efforts to master. In many children, this messy process occurs at very different rates and in different ways in different domains of competency. For example, some children have precocious motor coordination but are slow to speak. Others speak sophisticated sentences early but struggle to do one rung of the monkey bars or are insecure about climbing and jumping. If you have significant discrepancies in your developmental capacities, you are left with a subjective sense of inner imbalance, sometimes even of incipient chaos, in the background. It does not always bother you, but when you experience a threat, it can emerge. This could be called “anxiety”, but that is a rather crude description of a complicated subjective experience. I do not know David well enough to guess about whether he has a discrepant developmental profile. The inheritance of “inhibited” genes is another possibility. However, I do know many children who fit this picture of uneven development.

Finally, there is the interesting factor of culture. In an earlier posting, I wrote about how another 3-year old sat through a 6-hour wedding dinner with a minimum of fuss. I described what I saw his French parents do in order to teach him to sit at the table for long stretches. In any culture many core beliefs are transmitted by narratives. Children hear these narratives repeated over and over from early on and learn the culturally shared meanings that their parents convey to them. However, the meaning is not transmitted by language alone. Peter Fonagy talks about this process. He says, “Human communication is specifically adapted to allow the transmission of cognitively opaque cultural knowledge, kind-generalizable generic knowledge, and shared cultural knowledge” (Fonagy, lecture IPMH, May, 2012). This knowledge is transmitted by what he calls “ostensive communication cues” such as eye contact, turn taking with contingent reactivity, and special vocal tones. In a study Fonagy cited, infants of 18 months old were asked by the researcher to pass an object, a doll. In the control group, the researcher gave no cues directed to the infant, but in the study group, the researcher first smiled and said hello to the infant. Then in both groups, the researcher smiled at one doll and made a disgust face at the other. At that point, another researcher came into the room and the baby was asked to give the second person a doll. Only in the group in which the experimenter had smiled and said hello, did the babies give the second person the doll the first experimenter had smiled at, the doll designated as desirable . In other words, the researcher had initiated a relationship with the baby and in that context, the baby attended to the “ostensive cues” (smiling or disgust face) she then gave him. The infant trusted the researcher who smiled and said hello and then judged the information she gave him to be reliable.  

So, in response to my young friends’ question, I would say that I support their decision to protect David from anxiety provoking bible stories for the present. In avoiding certain bible stories they are acknowledging a dysynchrony between the dominant contemporary middle class U.S. culture and a culture in which bible stories are a primary means of transmitting beliefs. In the latter culture, bible stories would not just be read but from early on would be told as stories, with accompanying “ostensive cues”. In that culture, the parent would communicate – with eye contact, turn taking rhythms, and tone of voice – who the bad guys and who the good guys are in every story, over and over.  In that way, David would learn the salient meanings – with associated values – of his culture. Of course, he might still have fears of being a bad guy, because of his age-typical struggles with his aggression and negativity. It is less likely, though, that his fears would focus on bible stories.

Harrison, Alexandra & Tronick, Ed (2007). Now we have a playground: Emerging new ideas of therapeutic action”, J Amer Psychoanal Assoc., 55/3: 853-874.

 Kagan, Jerome & Moss, Howard A (1983) From Birth to Maturity: A Study in Psychological Development, Yale University Press. 

Tronick, Ed (2007). The Neurobehavioral and Social-Emotional of Infants and Children, WW Norton.


photograph by Joshua Sparrow



Read this blog in Spanish.









Murray-Cooper IPMH weekend, Part IV: Anxiety Disorders


Anxiety – Maternal anxiety and its effect on mother-infant interactions

There tends to be a specificity of mode of transmission of various forms of maternal anxiety from mother to child. Murray and Cooper are designing a prospective longitudinal study of two types of anxious mothers – General Anxiety Disorder (GAD) and Social Phobia (SP). First sample, 10-14 months.

They observed mother-infant and mother-infant-stranger interactions. When the mothers with SP are alone with baby, they are not less sensitive, but they show signs of anxiety. They are a bit less engaged with their babies. The subject babies, when they interact with the mother, show no difference from the controls. This is different from the pervasive effects of maternal depression. But when the stranger comes into the room, the SP mother shows more anxiety and does not offer her baby the clear displays of encouragement that the control mother did (for example, making eye contact with the baby, head nod with eye movement oriented to stranger, smile). The socially anxious mothers do not greet the stranger, look away, do not communicate to the baby that this could be a positive experience that the baby might find pleasurable. What they didn’t see (in contrast with other studies in the literature) was an increase in controlling behavior. 

Then they looked at how the infants were doing. The babies whose mothers have SP keep looking at their mothers when the stranger picked them up and showed less social engagement with another person (at 10 weeks). What was it in the mother’s behavior or in the baby’s behavior that can explain this finding? If you take into account the infant’s (irritability) you can see it feeds into the mother’s anxious lack of support for the engagement. They predicted that the mother’s behavior would predict future inhibition on the part of the baby. If you look at the change over time between 10 and 14 months, you do see a difference between the babies of SP mothers and control, but it is in the babies who are also assessed as inhibited. The control group mothers augment their encouraging behaviors, whereas the SP mothers do not do that; they are over protective and almost back off more – “don’t worry; you don’t have to do this”. It is the lack of encouragement on the part of the mother that predicts the increase of avoidant behavior of the babies.

They see these babies as having acquired SP behaviors . Ed wondered if this were lack of encouragement, or anxiety contagion. Yet, the situation is not static; in the future, the mothers have to send the babies to school at 5-yo, so what will happen then? You would like to have a narrative to provide: coherence, a temporal and causal structure, with links between internal states and behaviors, and highlighted salient states. They got the mothers and children to come in a month before going to school for the first time, and removed the text of a “Lucy Goes to School” book so that there were only the pictures. They asked the mother to use the book to talk to their children about what was happening to the children in the book.

The mothers exhibited different behaviors in commenting on the preparation-for-school-book to their children. 

Positive – “You are really going to like school!”

Negative – “There are lots of strange children in the classroom.”Attribution of vulnerability to the child – “You will worry about that.”Emphasis on the child’s dependency on mother – “you will need to hold my hand.” Promotion of avoidance etc.

Maternal anxious cognitions were pretty stable over time. Few SP mothers seek Rx because they have constructed their lives to protect themselves for example by leading relatively isolated lives. The children whose mothers had SP were more likely to give negative reports in doll play about the school experience. The mother’s level of encouragement was important; it was the children of mothers who showed low levels of encouragement that showed the negative reports. For children who were identified at 14 mos as inhibited and also having mothers with SP, the children were rated as depressed by their teachers. Only this combination produced this result. The way the children were thinking about school did seem to be driving their future adjustment at school. 

The way the mothers are talking to their children about this challenge of going to school is related to their own anxiety disorder and also to the way the children are thinking about the challenge. This is also marked by the child’s already having been seen as inhibited as an infant. Finally, there is also the potential moderating role of other family members.

By contrast, mothers with GAD do not show disturbances in their interaction with the stranger or with the infant in the presence of the stranger. 

Anxiety disorders in childhood are common and serious, affecting 5-10% of children. They have a significant adverse impact on the emotional, social, and academic development of the child, and they are often stable and increase the risk of the development of other problems such as substance abuse. Most treatments for Childhood Anxiety Disorder currently are CBT (approx. 55% free of primary Dx following Rx , eg Cartwright-Hatton et al, 2004, James et al, 2007). This is a little deceptive, since it doesn’t say that children are free of anxiety. These treatments tend to be individual 12-16 sessions of CBT for the children. What are the predictors of treatment response – severity, parental emotional distress, in particular parental anxiety. But these treatments are not widely available. In the UK, only ¼ children with mental health problems have seen a mental health professional in the last year. And only ¼ mental health teams had protocols for Rx of child disorders. Half of kids with AD have an anxious parent and half do not. Parental anxiety is a predictor of poor response.

Cresswell and Willets:

1.Introduction to anxiety – possible causes, maintenance cycles, implication for treatment. 2.Cognitive restructuring. 3.Graded exposure with positive reinforcement. 4.Parental responses to anxiety/brave behaviors. 5. Problem solving.

How you get to the end point doesn’t seem to matter; if you get rid of the anxiety, it is good. You can also use these treatments with ASD children. 

In their studies, Murray and Cooper took “primary mental health workers, who have basic training, and trained them to deliver the manual (Murray and Cooper, 2006). The children were aged 5-12 years old referred for anxiety formally assessed. When children met criteria for primary anxiety disorder, they were invited to participate. The PMHW delivered GSH under supervision by clinical psychologist. Then there was a post-Rx assessment of child anxiety, parent and therapist satisfaction. In the study, 56 kids met criteria and were given consent.

They did surprisingly well. 70% were “much” or “very much improved”. 53% were diagnosis free, and 36.7% were completely diagnosis free Parent feedback was remarkably positive. You have given the parents the tools to deal with their child’s anxiety. What the mental health workers have done in this intervention study is to empower the mothers to continue to help their children. This is changing practice in the treatment of anxiety disorders in the UK. It is all being done by the parents. You lose about 50% of the therapeutic benefit by just giving parents the book, because at that point they are skeptical and “want you to fix” the child. You need to have a bit of time to build a relationship and help the parent become open to change. 

Ed wondered if making the parent the therapist is a technique that could be used generally. When we talk about manualized treatments, we are always doing more than we talk about. The CBT person has a relationship with the patient. You may be doing a therapeutic intervention that changes the relationship between parent and child? Couldn’t you conceptualize this as relational therapy? Peter says of course it is. 

In their intervention model, Murray and Cooper challenge the mothers’ low expectations of their children’s behavior using video feedback – encouraging the child to take chances and rewarding him when he does. Treating the mothers for their AD with CBT is successful. However, at 16 weeks the controls – mothers who do not get treatment – get better. Why is that? By this time the children have gotten their own treatment. Maybe the mothers are responding to the children becoming less anxious, and maybe the mothers are paying attention to what the children are doing for their CBT. 

What about the changes in maternal behavior – overprotection and intrusiveness? They are all getting better, including the controls. There is no specific treatment effect. You don’t get a decrease in intrusiveness with CBT. You get most decrease in expressed anxiety with the video Rx (MCI). How do the children do when the mothers alone are treated? There is a benefit, but it is not significant. At the end of Rx, there is no benefit from treating maternal anxiety. There is a benefit of MCI in diminished diagnosis, but still not significant, though almost. The results do tend to improve over time given the new tools mothers and children have gained from the CBT. Yet this is not as compelling as they had thought. Maybe this is because if you effect significant change in child anxiety, this can drive a change in parenting itself. For example, if mothers are very overprotective, it is tempting to say maternal overprotection causes AD, but it may also be true that children’s anxiety elicits overprotection. It seems that there are many ways to get the same output. 

There was a discussion about maternal behavior and maternal depression and anxiety, and child anxiety. Ed asked about working within the behavioral categories – if you effect change in intrusiveness, what change will you achieve in the other categories?   

Cooper responded by comparing Italian vs English mothers in the studies – intrusiveness and controlling behaviors are way up in Italian mothers by comparison with English ones, but this intrusiveness is mediated by warmth in the Italian mothers.  Jeff Cohn and Campbell published on how when (Cohn & Campbell, 1990) depressed mothers had problematic relationships with their babies and then went to work, their relationships improved. Tiffany Field’s work – M-I interaction – child gets adjusted to that style, and when a stranger interacts with them the baby interacts to them in the problematic style he shares with the mother. Then the new person starts to slip into the problematic pattern (Transference). Alice Carter says that depending on your discipline, you will see a child as either being anxious or having a sensory processing disorder. Gergely states that if the adult looks at the baby and raises his eyebrows and then looks at an object, the baby will look at the object more readily than if the adult just looks at the object without engaging the baby. 

Conclusions: Since there is nothing we know that can really prevent PND and there is evidence of the destructive effect of antenatal depression (AND), it might be possible to early on intervene in troubled mother-infant relationships with great benefit. 

Ed talks about how the intractability of mothers’ depression results in the infants being exposed to more sad affect in the parent. He thinks of how much he would like physiological data in Peter and Lynne’s studies. In the bio-psychological meaning in states of consciousness, all states have stability and are not all concordant with one another. So we may be sleepy after eating even though we want to pay attention. They all have local purposes, but those local purposes fit into a higher order. There is lots of room for conflict, messiness, and complexity. In the behavior of depressed mothers of 6 month old infants, you see anger, poking, disengagement, and less positive play. You don’t see disengagement on the part of the nonclinical parents, nor do you see poking behaviors.  The babies show protest, less play, less attention to objects, and more looking away. They are disengaging socially, so they are not getting that, but they are focusing on objects, so they may preserve cognitive development. A lot of this may depend on what the parent may be able to do. You begin to learn and develop patterns with your caregiver. There is an element of chronicity. Small events occur over time, so that the child has experience with some things and not with others. 

This is where culture occurs. You can think of the game of peek a boo as a metaphor for development. The baby at first relies on the adult to initiate the game, then the child begins to initiate the game, and the 14 month old will begin to play both roles. Each child learns peek a boo in a different way because each child learns with a different parent who plays it in a different way. If you think about anxiety as similar to the way you learn peek a boo, what is your parent doing? Is your parent tightly squeezing your hand or calm and casual? This happens again and again. For example, how about building a routine of playing itsy bitsy spider every time after you change your child’s diaper, instead of getting on the floor and playing with your child 20 minutes a day. Let the parent know that whatever you help them do is all going to fit in their daily routine. Ed thinks there is an overemphasis on reflective awareness. For example, parenting from the inside out, Dan Seigel. He is convinced that there are lots of parents who have little reflective functioning but are excellent parenting. Often some athletes are terrific without reflecting on their playing but are fully attuned with their opponent.

Cohn, J & Campbell, S (1990), Face-to-face interactions of postpartum depressed and nondepressed mothers in mother-infant pairs at 2 months, Developmental Psychology, Vol 26 (1), 15-23. 


Creswell, C & Willetts, L (2007)– Overcoming Your Child’s Fears and Worries, Guided CBT Self-Help. 8 sessions with parents, London: Constable and Robinson.

 Cooper, P et al. (2006). Current anxiety disorder among mothers of anxious and non-anxious children. J Affect Disorders.

Read this blog in Spanish.











Murray-Cooper IPMH weekend, Part III: Interventions for PND


The next day, Peter Cooper, focused on interventions. He referred to studies that trained non-professional women in either “non directive counseling” or “CBT” and sent them to the homes of recruited depressed women in the third trimester of pregnancy (Rahmin et al, 2008, Lancet). The results demonstrated a higher rate of recovery in both groups, the CBT group higher than the group with non-directive counselors. Cooper and Murray wanted to know if it were some simple principle of intervention that breaks the isolation, someone to do nonjudgmental listening; they wondered if that were just as good as some specific technique. They had done a study in the early 90’s in which they compared groups of about 40 – controls with routine care from health visitors (the typical system in the U.K., in which minimally trained women are sent out to the home of each mother who has just given birth to act in a supportive and educative role.) – and three kinds of treatment – counseling, CBT, and psychodynamic therapies. The routine care group got no additional treatment; the non-directive counseling group focused on the mother’s mood; the CBT focused on the mother-infant relationship with behavior modification and cognitive interventions; and the psychodynamic focused on the mother-infant relationship with a focus on attachment, through the mother ‘s own personal history (This was inspired by B Cramer, 1993.). 

Here are some of the differences among the interventions that were studied: In the counseling group, the counselor followed a Rogerian (Rogers, 1951) technique in which the counselor provided an ear and encouraged elaboration of how the mother felt. In the CBT group, the counselor took a more concrete problem solving approach, for example, “It sounds as if you need some help in the afternoon so that you can have a bath. What are the possibilities? Do you have friends who can watch the baby? In the psychodynamic approach, following Cramer, the counselor explored the mother’s history with her own mother, starting with a modified AAI (Adult Attachment Inventory, cite). Then whatever issues the mother was finding difficult with her own baby would be considered from the point of view of the AAI material (aspects of her mother’s behavior that this mother was unconsciously repeating with her own child). Finally, there was an IPT group. IPT was developed by Myrna Weissman and Gary Klerman and conceives of depression as a relationship problem that occurs when one is not getting satisfaction from one’s relationships (cite); so the therapy is about examining and renegotiating relationships. The results of the IPT group were as good as the CBT group.  It turns out that both therapies were rather equally effective.

Yet, surprisingly, by 9 months the findings showed that there was no difference between those who did not get treatment and those who did. In other words, despite the fact that the three treatment groups had different foci of attention in their interventions, they did not have a treatment-specific effect. Not only that, but there was no enduring benefit of the treatment on the mother’s depression. The study showed that treatment is highly beneficial in the short run, and all treatments are moderately effective. However, about a third are not helped; this significant minority does not respond and goes on to be depressed and also to have disturbed relationships with their children.

What about medication for PND? There is only one good trail (Appleby et al 1997) in which a group of 87 were randomized to fluoxetine and either had CBT or not, and were then compared with placebo and CBT or not. All groups improved. Fluoxetine is better than placebo, and six sessions of CBT is better than one, but it is not additive. There did not seem to be any advantage to receiving both CBT and fluoxetine!

In 2004, Dennis and Creedy reviewed the studies and concluded that there was not benefit to prevention (Dennis and Creedy, 2004). However, Lynne and Peter felt that if you focused more on the quality of the early relationship and improving it, than you did on treating the mother’s depression, you could have greatest benefit. By creating a “predictive index” that included general adversity factors (poor housing, et.) obstetrical factors, and psychiatric factors, they were able to predict depression but just not very well. Why was that? One possibility is that the neonatal factors that you can’t pick up ante-natally played a important role. High irritability in the baby increases the risk of PND three fold, more than past history of maternal depression. Poor motor control in the infant is also very important as a risk factor. Maternal reports of difficult infant behavior is important too, though less, and maternal “blues” is also significant, though least powerful. Peter says that there is cumulative risk, so for example if you have high risk from previous depression and a highly irritable baby, you are at most risk for PND). Lynne and Peter found that when the mother had good support, the irritability of the baby did not add risk, but irritability plus no support added 4-fold risk. 

Questions and Comments from the Group:

In the US, there has been a focus on the trauma of childbirth (typical childbirth), but also a romanticization of pregnancy and childbirth. There was a discussion about the “grandiosity” of neonatal experience of mothers. There was another question about the effect of physiological factors, such as sleep deprivation. This has not been studied well. In the 30’s to 50’s they studied sleep a lot because they did a lot of EEG studies. Studies showed that during pregnancy the fetus developed two rhythms related to the mother’s rhythm and sleep cycle, and at birth the baby lost one and kept one. In DST terms, one rhythm was well organized and one was not. One might have been a maternal rhythm and one was related to fetal organismicity. If the mother in the pregnancy had a dysregulated sleep cycle, maybe the infant would be born with a really (doubly) dysregulated sleep. 

Reading Trial of Preventive supportive interaction treatment

This was a study of the preventive effect of health visitors to pregnant mothers. The mothers felt supported and understood, and the treatment was well appreciated. Yet, they found that at 8 weeks there was no impact on PND development. In other words, there is no evidential base to support the introduction of preventive therapeutic programs for PND. The best predictors from this questionnaire were marital conflict and previous depression. No studies have found to show that method of delivery apart from past psychiatric history had an effect. 

However, these studies typically eliminate the outliers. It may be that we need alternative models to analyze these data, such as dynamic systems models. For example, birth trauma may be a factor for a particular person. A nonlinear model that takes seriously these outlier situations may be more useful for clinical work. Instead of talking of “cumulative risk”, it may be necessary to dimensionalize the risk factors into factors such as financial factors, loss, etc., and in that way gain significance. Cindy Liu found that that significance varied in terms of ethnicity. For example, the occurrence of loss was much greater in the African American group and so were financial factors (Liu, ). The detailed contextualization is crucial. 

Summary and conclusions – 

PND is generally effectively treated with psychotherapy of a variety of forms; severe depression is less likely to improve and may require antidepressant medication and or longer term psychotherapy, but mothers are likely to experience further episodes. Preventive treatments do not seem to be effective. 

The manner of treatment does not seem to bear upon the outcome, though the issue of whether or not you have treatment at all is important. At 18 months the mothers report better behavior with treatment, and the teacher reports better behavior at 5 years. When mothers were under social stress, treatment improved the quality of mother-infant engagement.

The findings of the Reading study included the facts that the preventive intervention had no impact on the occurrence of depression, the quality of the M-I relationship, the infant security of attachment, or the infant cognitive and emotional outcome

Treatments directed at maternal depression even when successful have a minimal impact on the quality of the -I relationship and child developmental outcome. Treatments directed at the M-I relationship even when successful have no impact on maternal mood. It is therefore necessary that specific treatments be provided for both the maternal mood disorder and the associated disturbed m-child relationship. Depression does seem to be manifested in disturbances in the relationship in various domains. Mothers become depressed, the relationship becomes disturbed in certain ways, and patterns of relationship become established and gain a life of their own. Then the depression goes up and down. In other words, it is the depression that is causal, but once the problematic patterns are established, they just trundle along.

Cramer, Bertrand (1993) Are postpartum depressions a mother-infant relationship disorder, Infant Mental Health Journal, Vo. 14, Issue 4, pp. 283-297.

Dennis, CL & Creedy, D (2004). Psychosocial and psychological interventions for preventing postpartum depression, Cochrane Library, 2007, Issue 4, Wiley.

Rahmin (could not find reference).

Rogers, Carl (1951). Client-Centered Therapy, Cambridge, MA, The Riverside Press


Read this blog in Spanish.


Murray-Cooper IPMH weekend, Part II



Post Natal Depression: Part II

How do mothers affect the development of emotional regulation in infants? Murray showed a film of a study with mothers and 14-month infants in which the mother is first asked not to interact with the child when the researcher puts a toy frustratingly out of reach. The child naturally becomes distressed and remains so. When the mother is allowed to support her child, she responds quickly with an empathic facial expression. Then she says, “Oooh, how do you feel? Perhaps we can do something!” while giving a hopeful and suggestive facial expression. The mother begins to occupy the child by playing a little game with her fingers so that it distracts the child from her frustration at not being able to get the toy. It is clear how this kind of interaction repeated over and over can support the child’s growing tolerance for frustrating experiences.

In another wonderful film, a (non-depressed) mother is giving a bath to a 15-month girl. The mother is washing her face, and the child is tolerating this relatively well, but when she tries to brush her teeth, the child objects forcefully. This child turns out to have a firey temperament with a low threshold for frustration, and the mother must take account of the temperamental characteristics of her baby. The mother pauses a bit. She has a little toy that squirts water and uses it to squirt water on the toy the baby is playing with, making a connection with her. The baby responds happily, and then the mother builds on this connection and focuses on what the baby is doing. Mother pretends to brush her own teeth. The baby smiles and lets the mother brush her teeth, then takes the tooth brush and tries to do it herself. Of course, depressed mothers have little tolerance for this step-by-step reciprocal play at bath time. They are more likely to either forcefully take the toothbrush and get the job done over the objections of the baby, or withdraw. 

What are the mechanisms by which the early mother-child relationship affects the development of depression in the child? First there is emotional contagion (Field, 1984, 1988). There are also effects on the HPA axis (Murray et al, 2010 a). Finally, there is the link between insecure attachment and low resilience (Sroufe et al, 2005). One of my favorite studies, given my interest in vocal tones and non-symbolic communication, is one in which mothers were coded according to their speech tones (Murray, 2010). Falling intonation, non-rhythmicity, and monotony were significantly more prominent in PND mothers. In the non-depressed mothers, there was a rhythmic quality, with variability in pitch. What is the human equivalent of Michael Meaney’s “low licking mothers” whose babies had elevated cortisol levels? It is the low engagement of depressed human mothers. Halligan et al,, in 2004 and 2007) found that the 13-yo children of depressed mothers had elevated morning cortisol levels, which was predictive of depressed state. In one film of a 7-year old child of a PND mother playing a game with a peer, the child was initially triumphant when she was winning, but when she lost the advantage, she also lost her composure. Even when you control for marital conflict, maternal depression at the beginning of the infant’s life proves to be a predictor of depression at 16 years (Murray et al, 2011).

In conclusion, PND is associated with a range of disturbances in the mother-infant relationship. Babies’ outcome is affected in diverse domains- cognition, behavior, and affect, to name several important ones. Each outcome might have a specific developmental trajectory, such that for example, a depressed child may have no cognitive impairment. One of the most important take away points from this presentation is that obstetricians and pediatricians must get better and better at identifying and treating mothers with PND. Another point, though with less immediate importance, is for mental health clinicians learn to elicit the history of a PND in their evaluations of children with psychiatric problems. Finally, it is crucial to remember that nothing is written in stone. These problematic patterns may be interrupted at any point in the process – whether by the early intervention I have just been advocating, psychotherapy in the older child or adolescent, or by positive life events – a new supportive partner for the mother, the healing of an important relationship from the past, a better work situation, a new group of supportive friends for the child, the discovery of a talent in the child that brings positive self esteem to the whole family – that change the family trajectory in a better direction. Pointing out risk factors is never helpful without an equal emphasis on the resiliency factors that create a natural correction on a family life turning off course. 

Murray L, Marwick H, & Arteche A, (2010). Sadness in mothers’ ‘baby-talk’ predicts affective disorder in adolescent offspring, Infant Behavior and Development, 33:361-364.

Murray L, Halligan S, Goodyer I, & Herbert J, (2010 a). Disturbances in early parenting of depressed mothers and cortisol secretion in offspring: A preliminary study, J Affective Disorders, 122:218-223.  

Murray L, Arteche A, Fearon P, Halligan S, Goodyer, I, & Cooper P (2011). Maternal postnatal depression and the development of depression in offspring up to 16 years of age, J Amer Acad Child Adol Psychiat, 50(3):460-470. 

 Photo – Joshua Sparrow, M.D., Nov. 2012 


Read this blog in Spanish.