Reviews of Note

In an occasional column, NEJM Journal Watch Psychiatry editors comment briefly on review articles.

Internet treatments for eating disorders. In this review of eight randomized, controlled trials involving 609 patients,1 eating-disorder symptoms and, often, quality of life and symptoms of depression and anxiety improved in six studies of Internet-based interventions using principles of cognitive-behavioral therapy and involving guided self-help programs or self-help books backed by supportive e-mails. However, in two randomized, controlled trials, no significant treatment effects were seen with an intervention using writing tasks or with one that was based on e-mail and lacked a structured program.

This is your brain on stress. In this comprehensive review of mostly animal research, the authors examine how the prefrontal cortex — both the dorsolateral part dedicated to cognition and planning, and the ventromedial part, dedicated to emotion regulation — responds to stress and the moderating effects of age and sex.2 Stress can shrink dendrites, and these effects can be reversed with recovery from stress, but this resilience lessens with age. Estrogen has a potent effect in ameliorating stress-induced changes. Corresponding molecular, neurochemical, and neuroendocrine changes accompany these structural dendritic changes. In humans, exercise and mindfulness have been shown to affect dendritic growth, which suggests that changes in behavior may moderate and counteract stress effects on the brain.

Implications of attachment style for outcomes in psychosis. Based on a review of 22 publications describing 20 heterogeneous studies involving 1453 patients, of whom 58% had schizophrenia, these authors found small-to-moderate associations between insecure attachment (marked by anxious or avoidant coping styles) and positive and negative symptoms, mood problems, interpersonal problems, indications of trauma, and greater difficulties in engaging with services.3 Considering the implications of attachment style for outcomes in psychosis makes sense, because of parallel associations between insecure attachment and worse outcomes in other medical and psychiatric conditions. But before fully accepting the formulation that attachment-related difficulties lead to worse outcomes, we also need to understand how psychopathologic features of psychosis might themselves distort the responses to the attachment measures used in these studies. To what extent does being psychotic lead to the manifestation of insecure attachment in cross-sectional assessments? Bring on the prospective designs.

The DSM-5's new approach to diagnosing substance use disorders. The DSM-5 has opted to combine alcohol or substance abuse and dependence disorders into an overarching category of “use disorders.”4 These authors detail the rationales and evidence supporting these changes, which were proposed by the DSM-5 Substance-Related Disorders Work Group. Several other adopted changes include the additions of cannabis withdrawal and caffeine withdrawal syndromes and the decision to move gambling disorder to the section on substance use disorders.

Neurotransmitter switches in light and dark. These authors5 place into context recent research into the ability of neurons to switch the type and amount of neurotransmitters released in response to environmental conditions.6 Rats, which are naturally nocturnal, were exposed to a week of daily cycles having either 19 hours of dark and 5 hours of light or the reverse. During the long-light days, the rats produced more stomatostatin than dopamine, which led to increased production of corticotropin-releasing factor and corticosteroids. With this pattern of neurotransmission, the rats displayed anxious and depressive behaviors. Exposed to short-light days, these same animals had higher hypothalamic production of dopamine than stomatostatin, with subsequent decreased corticosteroids and corticotropin-releasing factor and lack of depressed/anxious behaviors. These paradigms may have parallels in human seasonal and shift-work disorders.

Prolactin levels with aripiprazole treatment. These authors discuss data on prolactin levels in diverse clinical populations receiving aripiprazole and compared these levels to norms in healthy populations.7 An age differential exists in aripiprazole-associated decreases in prolactin. Low prolactin levels occurred in 60% of children younger than 13 years on aripiprazole and 30% of adolescents on aripiprazole, compared with 8% of untreated children. Data on adults suggested lower, but not subnormal, prolactin levels due to aripiprazole. Low prolactin in women is associated with an inability to lactate, whereas in men it is associated with dysfunctional sperm, erectile difficulties, and infertility. Longitudinal effects of low prolactin levels in early childhood are not known. Physicians need to inform both adult patients and parents of pediatric patients about potential problems with aripiprazole-associated low prolactin.