In This Issue

Terrorism, war, death, bullying, and having to cope with lifelong illness are painful aspects of human life to which our patients are often exposed. This issue of the Journal brings important new evidence to help clinicians better understand and treat the consequences of these grave problems.

Can exposure to terrorism breed violence? The study by Heinrich and colleagues (p. 619) followed 362 Israeli adolescents living in an area that put them at risk of exposure to rocket attacks from 2008 through 2011. The authors show that exposure to rocket attacks has lasting effects on the mental health of adolescents. The most sobering finding is that these adolescents were 67% more likely to commit acts of violence for each rocket-related exposure they had experienced, even after adjusting for adolescents’ risk of violence before exposure. The findings of this study deserve to be read widely, particularly by those who seek to understand and deal with intergenerational cycles of violence in conflict-stricken areas around the world.

Another, frequently underappreciated, form of aggression is bullying. Lereya and colleagues (p. 608) used a community-based birth cohort to investigate the association between early bullying exposure and later self-harm. They found that having been bullied from 7 to 10 years of age was a direct predictor of self-harm at 16 to 17 years, even after adjusting for other factors, such as maladaptive parenting. In addition, exposure to bullying predicted self-harm indirectly through depression. Also, bullying seemed partly to explain the association between an adverse family environment and self-harm. The authors suggest that tackling bullying at a public health level could have a major impact on the decrease of future self-harm. Clinicians can work with young people, parents, and schools to help prevent bullying or treat its consequences.

Death leads to grief, but up to 10% of children experience such grief longer and more intensely. Children with prolonged grief reactions are at increased risk for depression and impairment; however, identifying those children is not straightforward. Melhem and colleagues (p. 599) compared the utility of the Inventory for Complicated Grief–Revised for Children (ICG-RC) with the proposed DSM-5 criteria for persistent complex bereavement-related disorder for identifying prolonged grief reactions. They followed up a sample of children at 9, 21, and 33 months after they had lost a parent to sudden death. They identified a threshold score at which the ICG-RC had remarkably high sensitivity (0.94) and specificity (0.97) in differentiating cases with prolonged grief reactions from noncases at 9 months. The ICG-RC by far outperformed the utility of the proposed DSM-5 criteria. In addition, the authors propose a 5-question screen that seems promising in identifying bereaved children. It asks about longing and yearning, shock, disbelief, loneliness, and a changed view of the world. These instruments could be of great help to clinicians seeing bereaved children.

We commonly diagnose autism spectrum disorder (ASD) early in our patients’ lives, but we rarely see them in their middle age. The study by Howlin and colleagues (p. 572) reports on the social outcomes of people first diagnosed with ASD as young children (average nonverbal IQ ≥70) and who are now in their mid-40s. The authors found that all patients still met ASD criteria as adults, although the severity of their autism symptoms had declined over time. However, 60% were rated as having “poor” or “very poor” outcomes. More than half the people in the sample had never worked or had been unemployed long term, and more than 30% lived in specialist placements. The Reciprocal Social Interaction domain score of the Autism Diagnostic Interview was the strongest predictor of adult outcome. A key question for clinicians will be whether by detecting children with ASD early and offering specialist care from early on we can help them have a better future social adjustment.

Abstract Thinking

Gene Effects Cross the Boundaries of Psychiatric Disorders

Genes are a major cause of mental illness. This knowledge has influenced how we think as clinicians and the way we interpret our own research findings. However, genetic findings have not yet revolutionized psychiatry in the same way as in other medical disciplines. Think of the ground-breaking development of imatinib to treat chronic myeloid leukemia. This drug was specifically designed to inhibit an aberrant protein produced by a genetic defect. Imatinib effectively cures the disease in most patients. Of course, the disorders we deal with in psychiatry are more complex. Although chronic myeloid leukemia is typically caused by a single chromosomal translocation, common psychiatric disorders (and other medical illnesses) are multigenetic and caused by many genes, each of small effect. Discovering these genes is particularly challenging.

A new study—the largest to date in psychiatric genetics—injects fresh hope to the field by identifying specific areas of the human genome (risk loci) that show strong associations with 5 major psychiatric disorders: autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder, bipolar disorder, major depressive disorder, and schizophrenia.¹ This unique study identified 4 loci that were significantly associated with mental illness, including 2 loci within previously identified gene regions that code for calcium channels (CACNA1C and CACNB2). A key finding was that these genes were associated with more than 1 of the disorders studied. These so-called pleiotropic effects were most pronounced for the calcium channel-related genes, which contributed to bipolar disorder, schizophrenia, and major depressive disorder.

The findings of this study demonstrate convincingly that the genetic architecture of psychiatric illness does not follow the boundaries of current psychiatric nosology. The authors suggest that their findings may allow us to move beyond descriptive syndromes in psychiatry toward a classification based on etiology. The findings certainly point to a future when classification could be based on etiology and pathophysiology. However, there are several challenges ahead. First, the effect of each of these genes explains a relatively small part of the etiology of the diseases studied (see Figure 3 in that article). The second is conceptual: differences in clinical presentation—the fact that we and to an extent our patients distinguish between, say, bipolar disorder and schizophrenia—may be important. Take autoimmune diseases as an analogy, where genes also have pleiotropic effects. Rheumatoid arthritis and celiac disease share genome loci,² but this does not mean that physicians should stop classifying them by their clinical presentation. Having said that, we may want to draw lessons from the genetic findings and be prepared to refine our clinical descriptions and accept that, for some purposes (e.g., drug development), a gene-based classification may prove more yielding. It may seem just a dream right now, but just imagine if our patients could benefit from psychiatry’s own imatinib, say, a class of drugs targeting aberrations in calcium channels.