Cortisol Levels: Do They Identify Neglected or Abused Children?

Behavior observations tell us a lot about children’s development, abilities, and moods, but to be accurate they require a lot of work from highly trained people. Understanding a child’s past experiences would also help us assess that child’s present needs and capacities, but many-- even most—experiences in the family are private, and we can identify them only if they have consequences that can be observed later on. It’s partly because of these difficulties that there have been a number of suggestions about using hormonal or neurotransmitter testing to try to assess children’s backgrounds and the help they need.

Sometimes these suggestions appear to ignore the fact that a biological measure is only part of the complex picture that represents the child in the family. For example, one recent case study was used to show that termination of parental rights could be decided on the basis of a child’s neurotransmitter levels following a visit with the non-custodial mother (Purvis, K.B., McKenzie, L.B., Kellermann, G., & Cross, D.R. [2010]. An attachment based approach to child custody evaluation: A case study. Journal of Child Custody, 7, 45-60). That study was very much lacking in the kind of evidence that would be needed to conclude that such measurement could be used appropriately.

Of the various biological measures that have received attention as possible indicators of a child’s condition and needs, cortisol, the “stress hormone” that is produced in response to a wide range of experiences, including neglect and abuse. For example, an adoptive mother commenting on http://phtherapies.wordpress.com/2010/09/23/monica-pignotti-responds-to-adoptive-parents-comments explains her child’s distress and difficult behavior by referring to high levels of cortisol.

Increases in cortisol levels following brief, intense stress help the individual to withstand stress effects. Some cortisol production is valuable-- it would be a problem to be without cortisol to help withstand the occasional shocks of normal life. But long periods of stress, with chronic cortisol activation, have a number of negative effects, including cognitive and emotional problems. And some people who have been stressed over long periods begin to show lower levels of cortisol production as a consequence. Clearly, there is no simple one-to-one connection between current cortisol level and history of stress, or between cortisol level and present ability to handle stress.

A recent article in Child Development summarizes some of the complexities of the connections between cortisol levels and experiences of abuse or neglect (Cicchetti, D., Rogosch, F.A., Howe, M.L., & Toth, S.L. [2010]. The effects of maltreatment and neuroendocrine regulation on memory performance. Child Development, 81, 1504-1519). These authors point out that there seem to be no overall differences in cortisol levels between all types of maltreated children and non-maltreated children. However, like so many biological characteristics, cortisols vary with respect to time of day, and the pattern of that variation may be different for children who have been extensively maltreated (both sexual and physical abuse, as well as neglect and emotional abuse) and those who have been physically abused only. Some maltreated children produce high levels of cortisol, some low levels, and some show a normal pattern in spite of their experiences with abuse or neglect. As Cicchetti and his colleagues conclude, “… hypercortisolism [high levels of cortisol] is more likely to occur as a result of acute stress and a shorter time since the onset of stress or maltreatment, whereas hypocortisolism [low levels of cortisol] is likely the result of a long-term process of chronic stress and a later time since the onset of stress or maltreatment” (p. 1506).] Once again, the cortisol measure is not related in a simple way to whether a child has or has not been maltrerated.

Why can’t we just use children’s cortisol levels to decide what kinds of stresses they may have experienced, and at what points in their development? Should we use hypercortisolism to convict a child’s present caregiver of abusive treatment, or hypocortisolism to convict a former caregiver? Or should we tell the adoptive mother cited above that if her child has high cortisol levels (which of course we don’t know), this must be due to the stress of his recent experiences? No, of course not. For one thing, as Cicchetti and his co-authors point out, there is a good deal of variation in children’s cortisol patterns, and no overall difference between maltreated and non-maltreated children. In addition, we need to recall that these are correlational studies, in which there is an association between two measures, but no certain way to decide which caused the other. It is conceivable-- though perhaps remotely—that children with certain daily patterns of cortisol production are more likely to be extensively abused than others, and even that biologically-related caregivers who share certain patterns are more likely to be abusive. Information about stress hormones and neurotransmitters allows us to explore connections between biological factors, abusive treatment, and child outcomes, but at this point such information is only a compass, not a detailed map.