Even between acute episodes of bipolar disorder, sleep problems were documented in 70% of the patients studied by Harvey et al. (p. 50). On most sleep measures, the bipolar patients were more similar to patients with diagnosed insomnia than to volunteers with good sleep patterns. The patients differed from the good sleepers in the proportion of time in bed spent asleep, and the reasons for poor sleep most often identified were "I can’t get in a proper routine," "My mind keeps turning things over," and "I am unable to empty my mind." Sleep loss can herald an acute episode of mania or depression, and worrying about this possibility can add to cognitive arousal. Cognitive behavior therapy for insomnia might help bipolar patients sleep better between episodes and thus help maintain their stability.

The cognitive functioning of people who are at higher than average risk for psychosis has been tested, but whether these abilities are linked to the actual development of psychosis is unclear. Brewer et al. (p. 71) have made such a connection. They assessed the cognitive functioning of people considered to be at ultra-high risk of psychosis because of brief or attenuated psychotic symptoms or because of high genetic risk combined with a recent decrease in mental status or functioning. Of the 98 participants, 34 developed psychosis over a follow-up period of at least 1 year. At baseline they had lower scores than healthy subjects on tests of logical memory and visual reproduction, both of which require rapid processing and efficient organization of information for accurate recall. These deficits suggest that impairment in prefrontal networks occurs before the onset of psychosis.

A lack of facial expression, "flat affect," is a common symptom of schizophrenia but seems also to apply to patients with depression. Trémeau et al. (p. 92) compared various aspects of facial expressiveness in patients with schizophrenia, patients with depression, and healthy subjects. Both groups of patients had lower scores than the healthy subjects on all measures of emotion. The two patient groups did not differ from each other in imitating or producing facial expressions of six emotions or in the number or frequency of smiles. The patients with schizophrenia had better scores for spontaneous expressions of selected emotions, while the depressed patients had better scores for overall facial expressiveness and number of words spoken in a 2-minute period. The deficits of both groups reflect motor, affective, and social impairments. These domains suggest brain connections between a premotor region and cortical and subcortical structures involved in emotional and social behavior.

Complications during birth are suspected of raising the risk of schizophrenia in the offspring. Jablensky et al. (p. 79) wondered whether the reverse is true: Does having schizophrenia increase a woman’s chances of experiencing obstetric complications? Some answers were provided by a population-based study in Western Australia. Women with schizophrenia were more likely to have placental abruption and to have babies with cardiovascular birth defects or low birth weight than were women with bipolar disorder, depressed women, and healthy women. Their odds of neonatal complications rose during the winter, and low birth weights peaked markedly in the spring. Serious mental illness itself may raise the risk of obstetric complications, because the rate increased after the onset of mental illness in both the women with schizophrenia and those with bipolar disorder. Pregnant women with serious mental illness thus may benefit from more illness-related help and more prenatal care.

The features of pathological gambling overlap with those of bipolar illness, the two disorders frequently coexist, and some patients with impulse control disorders are helped by antidepressants or mood stabilizers. Hollander et al. (p. 137) conducted a placebo-controlled trial of lithium for patients with pathological gambling plus bipolar II disorder, bipolar disorder not otherwise specified, or cyclothymia. At the end of the study, the mean score for pathological gambling of the patients receiving lithium was about half that of the patients receiving placebo. Over two-thirds of the lithium patients were judged to be responders, compared to one-third for placebo, and improvement in gambling was correlated with improved mania ratings. Lithium’s success indicates the value of subtyping patients with pathological gambling, so that those with bipolar spectrum disorders can be treated most effectively.F1

Frederic William Henry Myers, 1843–1901 (p. 34)F2