Relaxation training made people worse, and a circadian app cut mood episode recurrence threefold

The edition in brief

A Bayesian individual participant data component network meta-analysis pooled 34 randomised trials of task-shared psychosocial interventions for common mental disorders, with individual data from 30 trials and 10,612 participants. Strengthening social support was the most beneficial component (incremental mean difference -9.48, 95% credible interval -13.29 to -6.60), followed by behavioural activation (-4.15, -7.48 to -0.05) and problem management (-4.08, -5.37 to -2.83). Relaxation was detrimental (+7.97, 3.35 to 12.11), as was cognitive reframing less clearly (+5.17, 0.78 to 9.15). Component effects varied by baseline severity. In 676 adolescents with subthreshold depressive and anxiety symptoms, the Unified Protocol for Adolescents was compared with protocolised progressive relaxation as an active control. Both improved symptoms with no significant difference on the primary outcomes; the Unified Protocol was better on hyperactivity, conduct problems and total difficulties, and marginally on depressive symptoms at 12 months. A mega-analysis of individual data from 1,189 participants across 18 magnetic resonance spectroscopy studies found antipsychotic non-responders had raised medial frontal glutamate, glutamate plus glutamine, choline and myo-inositol compared with responders, with small effect sizes. Raised glutamate plus glutamine was seen prospectively in first-episode psychosis. A randomised study of 1,012 US teenagers and young adults found health warnings on a mock social media page were rated more discouraging than a neutral message and more effective than the screen-time break warnings platforms currently use voluntarily. A double-blind sham-controlled trial of a circadian rhythm feedback app in 93 adults with major depressive or bipolar disorder found the sham group had 3.39 times the recurrence rate over 12 months.