Applying individual participant data meta-analysis to 3 pivotal randomized clinical trials in acute schizophrenia (n = 611) and 3 randomized clinical trials in predominantly negative schizophrenia (n = 475), Furukawa and colleagues found that for both types of schizophrenia, the more severe the baseline symptoms, the greater the differences between the active treatment and placebo. Towards the mildest end of the spectrum, patients benefit less in terms of symptom improvement but may experience full adverse effects of antipsychotics.
Kiosses and colleagues found that in 74 older adults with major depression and cognitive impairment to the level of moderate dementia, problem-adaptation therapy, a home-delivered psychosocial intervention, increased emotion regulation and improved depression and disability more than supportive therapy. Thirty-eight percent of participants in problem-adaptation therapy achieved full remission of depression and an additional 25% achieved partial remission. In addition, problem-adaptation therapy improved depression in patients who had failed 1 adequate trial of an antidepressant (n = 31).
Using data from a 10-year longitudinal study of preschool-onset major depressive disorder, Belden and colleagues examined anterior insula volume as a biomarker for recurrent depression. Schoolchildren with a history of preschool-onset major depressive disorder had significantly reduced insula volume compared with healthy peers. The association between preschool-onset major depressive disorder and reduced insula volume was driven by the symptom of early pathological guilt. Smaller anterior insula volume predicted an increased likelihood of major depressive disorder recurrence postscan.
Hahn and colleagues applied machine learning to functional magnetic resonance imaging data to predict treatment response to cognitive behavioral therapy in panic disorder with agoraphobia. Prognostic markers were derived from regional and whole-brain gaussian process classifiers using a nested leave-one-out cross-validation. Integrating regional classifiers yielded good predictive performance (accuracy, 82 %; sensitivity, 92 %; and specificity, 72 %), showing that response prediction based on functional magnetic resonance imaging data is possible with high accuracy on a single-subject level.
Safer and colleagues analyzed national data of physician office–based visits for youth, finding that DSM-IV subthreshold not otherwise specified (NOS) diagnoses rose significantly between 1999 and 2010 for composite mood disorder NOS, anxiety disorder NOS, and bipolar disorder NOS compared with those psychiatric diagnoses achieving full criteria. When attention-deficit/hyperactivity disorder diagnoses were excluded, NOS diagnoses from 2007 to 2010 accounted for 56% of all psychiatric diagnoses in youth.