SEARCH

SEARCH BY CITATION

This month's Journal contains, as it usually does, a mixture of experimental studies and non-experimental or observational studies, together with some reviews and meta-analyses. Since the paper most likely to attract public attention this month is an observational study, let us first consider the papers using this approach.

Observational studies can have many strengths, especially if they follow a large, representative sample over time. They can have higher ecological validity than experimental designs, and can take into account a range of potential confounders of the cause–effect relationship that we all seek in our research. But they can also contain serious threats to validity.

Psychiatric consequences of abortion

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

Fergusson and colleagues’ new analysis of the Christchurch, New Zealand, longitudinal community survey looks at the psychiatric sequelae of abortion. The authors report that young women who had an abortion by age 25 had elevated rates of psychiatric disorders, compared with women who became pregnant but did not have an abortion and women who had never become pregnant by that age. This is an important topic anywhere in the world, and an incendiary one in the United States.

The paper demonstrates the value of collecting information repeatedly over years or decades on the same individuals, in a large and representative sample, so that readers can have confidence that the findings are likely to generalize to other similar young people. It is also an example of the skill and care with which this group of researchers analyze their complex longitudinal data and present their results.

The paper, however, also demonstrates a fundamental problem in observational (non-experimental) research: the difficulty of determining whether an association is a causal one. Was it the abortion that caused the observed excess of depression, anxiety, and suicidal ideation? Or was there something in these young women's lives that led to both the abortion and the mental health problems independently? Fergusson and colleagues tackle this problem in two ways. First, they include a number of covariates (socio-demographic factors, measures of family functioning, child and adolescent adjustment and attainment, etc.) that might predict risk for both abortion and later mental health problems. After including their set of covariates, the risk of each disorder is still higher in the women who had abortions than in the other two groups. Second, they compare women's mental health measured at ages 21 and 25 as a function of their pregnancy and abortion history up to age 21. The number of later mental health problems is significantly higher in the group that had an abortion by age 21 than in the other two groups, whose mean number of mental health problems does not differ.

Fergusson and colleagues are thus able to use the longitudinal, multivariable nature of their data to narrow down the number of possible explanations of their findings. However, as they are careful to point out in their discussion, there are still several alternative explanations that need further research. First, the range of covariates is limited and may not include key factors that might independently predict both abortion and mental illness. Familial transmission of attitudes to sexuality is one that comes immediately to mind.

Second, the study contains no information at this point about how many of the pregnancies that were not aborted were planned or wanted, and how many were unwanted. We might expect women who wanted their babies to have better mental health in later years than women who found themselves with an unplanned pregnancy, and this could raise the overall scores for the no-abortion group. The key comparison is between women who carried unwanted pregnancies to term and those who had an abortion. Fortunately, Fergusson and colleagues will be able to find out the answer to this question when next they interview their study participants.

Infant development and adult cognitive functioning

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

In another study of a birth cohort now grown up, Murray and colleagues use the Northern Finland 1966 Birth Cohort to add to the growing body of work, under the general title ‘life course epidemiology’, showing that very early development can have considerable consequences for health and functioning decades later. In this instance, the authors make the case that early motor development, indexed by age in months at learning to walk, is a marker of executive function development. They contrast categorization tasks thought to use executive functioning capacities with tasks of verbal learning and memory. There was a significant linear relationship between age at standing unsupported and the former tasks, but not the latter. The authors argue that motor control, like categorization, is sensitive to both frontal and basal ganglia function, and that variations in the functioning of fronto-striatal circuitry could provide a parsimonious explanation of this relationship. Here too the researchers made use of information on confounders to rule out alternative explanations.

Cortisol and psychopathy

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

In another study linking biological functioning and behavior Loney and colleagues replicated the finding of low mean resting cortisol levels in a group of adolescent boys scoring high on a scale of callous/unemotional (CU) symptoms, compared with conduct disordered but not CU boys. No such effect was found in girls. The study advances this area of research in that it uses a general population sample of both sexes, and a measure of testosterone levels as well. This showed no effect of CU score, as the authors predicted.

It is important for researchers to think carefully about the role of ‘biological markers’ in future clinical or preventive work. The authors end their paper by suggesting the ‘low resting cortisol may be a biological marker for male CU traits’. However, what they showed was a low mean score in their CU group, which is different from showing that individuals with low resting cortisol have high CU scores. The positive predictive power of both psychological and biological indicators of risk has been notoriously poor in psychiatry. The authors are right to stress that low resting cortisol should be explored as ‘one of several indicators of risk’.

The symptoms of autism

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

Van Lang et al. use a very special non-experimental sample – a large group of individuals with mild to severe autistic symptoms – to explore the best way to classify the symptoms of autism. As they describe in the introduction to their paper, there have been many attempts over the years to identify the key clusters of symptoms in autism. Their attempt to solve this enigma is backed by a large sample, all assessed using a respected diagnostic interview, and a method of analysis that directly pits the DSM-IV model against another model based on the authors’ earlier work. Clearly, many factors such as the age and intellectual range of the sample contribute to the results of analyses like these.

Preschool bullying

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

Another non-experimental study by Perren and Alsaker examines the relationship between bully/victim characteristics and behavioral and friendship patterns of children in Swiss kindergarten schools. Given photos of their class-mates, they were easily able to put them into friendship groups, and to identify their own best friend. Social isolates could be identified negatively: they were the children who were rarely selected as best friends or members of social clusters.

There are many interesting things in this paper. First, the study shows that it is possible to collect useful information from 4–7-year-olds if the researchers go about it the right way. Among the findings most relevant to treatment and intervention, the authors make the excellent point that for problems like bullying it is not enough to devise treatment programs for individuals. Bullies are socially quite popular, as this study shows, and can create a social culture in which their behavior becomes tolerated or even supported by other children. It is essential to change the culture of the whole class, or school, or community, if bullying is to be eradicated. The authors have some good ideas about how to go about this.

Studies using service settings

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

Two studies this month use health care delivery settings as their focus. Schiffman and Daleiden use 18 months of data from the Hawaii Department of Health's Child and Adolescent Mental Health Division (CAMHD) to show that although children with schizophrenia spectrum disorders are rare, they are expensive. Compared with the 14,000 other children registered for services, the 71 with schizophrenia-spectrum disorders (defined as most psychotic disorders except for affective psychoses) cost almost twice as much on average, largely because they were three times more likely to be placed out of home or to use flex or respite services, and were also more likely to use intensive home and community services. Even so, there was a subgroup (9 out of the 71) for whom no service use was recorded. It would be useful to know more about how children are referred to CAMHD, whether it serves all the referred children in Hawaii or whether some are treated privately despite the islands’ universal health care.

St James-Roberts and Alston use primary health care settings to address the question of screens or tests as indicators of need for targeted interventions. In this case the screen is a brief questionnaire about the pre-language behavior of 10-month-old children. ‘Failure’ on this test is being used in several Health Trusts (primary care practices outside the UK) as grounds for treatment (not described in this paper). This sounds like a costly and labor-intensive undertaking for a Trust, so the authors set out to validate the screen by testing whether infants who failed the screen were also more likely to fail a test of sustained attention in the face of a distraction. They found some evidence to support their hypothesis, but also a poor predictive value positive, with up to a third of the ‘failures’ not meeting that criterion later when tested using other methods. Since the intervention, whatever it is, is likely to consume the resources of both families and Trusts, accuracy at the level of individual children, not just significant group differences, will be an important goal of the screening program.

Experimental studies

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

Three of this month's papers use experimental designs. Two explore how maternal depression transmits risk to children. Silk et al. designed a task highly likely to elicit unhappiness: children of depressed and never-depressed mothers had to wait several minutes while their mother completed a boring task before they could have a cookie or toy that was clearly visible beside the mother. Video tapes of the experience were coded for four different strategies used by children to deal with their frustration. The authors confirmed their hypothesis for one of the four strategies: children of depressed mothers were more likely to focus on the object that they wanted but could not have. However, two important findings emerged from the post-hoc, exploratory analyses. First, daughters of depressed women were less likely to use active distraction to regulate their feelings, and more likely to engage in passive waiting, than were sons. Second, the more episodes of depression the mother reported, the greater was girls’ use of passive waiting and the less their use of active distraction to deal with their feelings. We know that children, and especially daughters, of depressed parents are at heightened risk. This study provides ideas about the nature of the risk and suggests strategies for prevention.

In another study from the same group, and using largely the same subjects, Forbes et al. look at frontal electroencephalogram (EEG) asymmetry as a marker reflecting the balance of activity in the left and right frontal areas of the brain. Left frontal asymmetry indicates greater activation in the left than the right hemisphere frontal areas, and vice versa. Most of the work so far has been with infants and very young children, and suggests an association between right frontal asymmetry and withdrawal, social wariness, and behavioral inhibition. Because infants of depressed mothers have been shown to have right frontal asymmetry, this has been seen as indicating risk for depression. The present study uses an older sample of prepubertal children, and finds no simple relationship between maternal history of depression and right or left frontal asymmetry. Further analysis shows several intriguing interactions, but these are with left rather than right frontal asymmetry. Children of depressed mothers with left frontal asymmetry had high levels of anxious/depressed symptoms according to the parent-completed Child Behavior Checklist. Children with poor affect regulation (a risk for daughters of depressed mothers, as shown in the previous paper) had high anxious/depressed symptoms only if they showed left frontal asymmetry. This paper takes advantage of an unusually large and well-characterized sample, but even so it lacks the power needed to test the three-way and four-way interactions that this work points to if we are to understand the nature of the risk conferred by a maternal history of depression.

In another paper exploring the mechanisms of disorder Bebko and colleagues ask whether a basic deficit in autism has to do with the development of skill, or interest, in paying attention to stimuli in which sound and image are matched. They found no deficit when the stimuli were not linguistic ones, but a marked deficit when the stimuli were verbal. As the authors point out, ‘Discrimination of temporal synchrony may be the first step in developing a capacity to discriminate more complex and specific forms of language or other information that is invariant across sensory modalities.’ The nature of the task, which simply records eye movements while the child sits in front of the televised images, does not distinguish between whether autistic children are less capable than other children at an equivalent level of functioning of matching multimodal inputs, or simply less interested in the results of matching.

Reviews and meta-analyses

  1. Top of page
  2. Psychiatric consequences of abortion
  3. Infant development and adult cognitive functioning
  4. Cortisol and psychopathy
  5. The symptoms of autism
  6. Preschool bullying
  7. Studies using service settings
  8. Experimental studies
  9. Reviews and meta-analyses

The other two papers this month pull together past work of many researchers to clarify what we know about two very different topics. Gathercole and Alloway provide a valuable introduction to problems with short-term and working memory that have implications for many areas of children's lives, including school attainment. Children with short-term memory deficits may give the impression of being inattentive or disobedient when they simply have limited ability to hold a sentence in the ‘phonological loop’ long enough to obey an instruction or complete a task. Children with poor working memory, in contrast, will have problems with tasks that involve both storing and processing information. They may also appear to be inattentive or failing to try, when they have simply failed to hold a complex task in memory long enough to complete it. The authors describe typical memory deficit profiles seen in children with Down syndrome (verbal short-term memory), Williams syndrome (visuo-spatial short-term memory but not verbal short-term memory), and specific learning impairment (verbal short-term and working memory), and provide an interesting discussion of whether the inattentive type of ADHD is associated with a primary deficit in working memory, or ‘intermittent failures to attend to working memory tasks’. The paper provides suggestions for remedial help for both short-term and working memory deficits.

Reyno and McGrath provide a most valuable meta-analytic paper on the questions: What predicts the effectiveness of parent training for children's externalizing problems, and what predicts dropout from these parent training programs? Clearly, these are topics of importance to clinicians as well as researchers. The paper ends with useful suggestions about how to make it easier for parents of behaviorally disordered children – parents who tend to have all sorts of difficulties in everyday survival at the best of times – to get involved in treatment programs and to stick with them.