March IPMH Meeting: Epigenetics II


(Continuation of Barry Lester’s presentation with discussion)

What are some long-term prenatal stressors that affect placental genes? These risk factors may play a role in future problems but do not predict the future. What they can do is alter the HPA system set points and affect the way the newborn responds to stress. This in turn could influence the baby’s regulation.  One could hypothesize that potentially this could lead to dis-inhibition, psychopathology, cognitive problems, and adolescent substance abuse. Of course, there are many steps along the way and many forks in the road. We are only talking about risk factors in a very complex set of processes. We don’t know what these pathways are that lead to problems down the road. We only see correlations. 

Let’s look at cocaine using mothers. It becomes difficult to isolate a particular risk factor in the lives of these women. But the cocaine exposed babies showed higher reactivity to stress. What are the other stressors in the picture? In addition to cocaine use, there is also a relationship of the babies’ high reactivity to the number of caregiver changes. The important point is that you are looking at how the drug interacts with adversity in the postnatal environment, not just the effect of the drug itself.

Which epigenetic changes will be passed on?  Ed (Tonick) says that nutritional studies show that it takes two more generations after the parent generation to clear out some of the nutritional effects in the generation of the grandparents. Do these pass through the mother’s line only? (There are also those who dispute the intergenerational transmission.) If the changes that are passed on represent a guess about the environment, then that is important. How stable are these epigenetic changes anyway? It is possible that they are short lived but that the environment reinstates them. Suppose you are a 2-yo and you are neglected and so you methylate your GR receptors (stress regulatory system). Then you get neglected again, so you methylate them again. There then may be changes in your hippocampus that in effect become a damaged part of your stress regulatory system. Is it causality or association? It seems unlikely that methylation of one gene is going to cause the changes we are talking about. 

It is essential to keep in mind that the connections we are talking about do not represent the actual events of the real world. Ed says that “simple” and “sovereign” is always wrong. We tend to find a new paradigm, and the situation always becomes more complicated the more you look at it. We always find a new paradigm and it gets more complicated the more you look at it. If you look at the Dutch famine study, think about the mothers who were pregnant during a war in which the whole country was starving. The famine ends, but people have died, fetuses have been aborted, the babies who are born are smaller and more irritable. The women who mother them are traumatized. We talk about the famine but do not seriously take into consideration all the other horrible factors that were involved. These amplifying and reinforcing factors were still going on even after the war.

The magnitude of effects – the effect size of all these factors – is small.  It is interesting that you get this variability in the relationship between methylation and behavior in healthy babies, and it makes you wonder what would happen if you look at “at risk” populations; in that case, do you magnify the effects? Within the normal range you can find the same relationship between birth rate and behavior in a study done by Ed Tronick and Barry Lester. You need to consider the whole range – babies who weigh 8 pounds and those who weigh 6 pounds – will the bigger babies have a little better organized behavior? 

There is a new NICU at Brown with single rooms. The changes taking place include more breast feeding, more kangaroo care, etc. An Italian colleague of Ed’s, Rosario, did a study in which he looked at the quality of care in 24 different NICU’s in Italy. He categorized them and gave them assessments, and the babies in the better NICU’s had better scores than those in the least good ones. You are then discharging a baby who is medically in better shape and also neurobehaviorally in better shape. Also, the higher the level of neurodevelopmental care, the lower the incidence of depression in the mothers. All of these NICU’s ascribe to a particular care policy, but in fact they vary. Some of the things that are done in developmental care are thought of as “neuro-protective”. Ideas of developmental care have shifted. 

Schizophrenia – what is the epigenetic issue? There are people who are studying epigenetic changes related to schizophrenia and autism. There is not a lot published yet, but it seems there are prenatal effects that are related to schizophrenia. We also need to look at the relationship between epigenetic changes and genotype. 

Steve Suomi  has done cross fostering studies of rhesus monkeys. He took inhibited and uninhibited babies and cross fostered them, and found that there was evidence of temperament coming through in addition to the environment. Barry Lester thinks of temperament as a protective factor. Nancy Snidman asked about individual differences in the pups – not all the pups get licked, do they? Everyone agreed that there are probably individual differences among the pups that influence how much they get licked. Ed pointed out that just because these models have to do with stress, it doesn’t mean that stress is all bad. We don’t know what appropriate levels of stress are. What about the stress and temperament interaction – is it possible that stress for a highly reactive kid can lead to a blow out, whereas stress for a low reactor can be facilitating? Yes, but it also depends on what you mean by stress. These kids have different thresholds for reactivity. Some people get a rush that is positive, and others feel a negative reaction immediately.

If you have an acute stressful event and it finishes and is done, and yet you ruminate about it, that self-amplifies the stress.  This is where a psychodynamic factor plays an important role. In the case that the individual makes a harmful, self-critical meaning of the stress, then when stress happens, things deteriorate. Stress reactivity by itself, the cortisol effect, is a nonspecific model that could go many different ways. The Kagan model is more specific – it is reaction to novelty – across the lifespan. Stress reactivity and care is a different model. There are many paths that fussiness in a baby can take. Nancy and Kagan were looking at reaction to novelty, and they wanted to get the system aroused to see the physiological systems respond. There were a lot of things they could not include that have to do with caretaking. The amygdala approach-withdrawal reaction – is specific in its relation to novelty. They started with older children. How could they bring that reaction down to infancy and what was going on in the brain? They do see SES differences. Nancy reminds us that most kids are a mix of the temperamental features they are talking about. They were studying mainly the extremes. 

This kind of research is also constrained by the use of checklists. Remember that it is not only parental translation to what these words like “seldom” and “often” mean; it is also what we mean. In the checklist we have to interpret the answers, and what was “trouble” when the checklist was developed and what it is now sometimes has changed. The reason you use CBCL is that the parent has a thick relationship with the child and you may not see all these things for 15 min in the lab. It is also true that the more extreme anything is, the greater effect you will see from it. So that you will see a greater effect the more abnormal the caretaking is. In the case of the relationship between temperament and training in dogs, it is harder to train certain breeds to be aggressive than others, but you can do it. You could probably override most traits. 

In the discussion, one fellow talked about a home placement program in which a child who has not been able to develop certain skills is put into a supportive foster home. Then, when he is better, he is sent back into the home, and they cannot manage. It is typical for kids to do well in structured environments but do poorly at home, and the school says. There is a tendency to dismiss the fact that the child does better in one context because of the regulatory support context.

In the next blog, I will report on the presentation about Attachment Theory. 

Read this blog in Spanish.


Leave a Reply

Your email address will not be published. Required fields are marked *