Self-harm in young people

Several papers in the BJPsych this month focus on self-harm, including suicide attempts, in young people, in terms of both predictors of self-harm and subsequent outcomes. Rahman et al (pp. 215–221) linked general practitioner, hospital and education records in Wales and identified an association between declining academic attainment in primary and secondary school and subsequent onset of depression in adolescence, while onset of self-harm was predicted by a decline in attainment during secondary school only. The authors comment on the utility of considering academic decline as a risk factor for future depression and self-harm in young people, along with a range of other established risk factors. In another study investigating risk factors for self-harm, Huang et al (pp. 234–238) found evidence that a diagnosis of attention-deficit hyperactivity disorder in a cohort of adolescents and young adults was associated with suicide attempts and repeated suicide attempts, including in age- and gender-defined subgroups. Treatment with pharmacological agents did not increase risk, while long-term treatment was actually associated with a decreased risk in males.

Examining outcomes following self-harm in young people, Wilkinson et al (pp. 222–226) found that recurrent non-suicidal self-injury (NSSI) occurring before the age of 14 years was associated with depression and eating disorders 3 years later, while sporadic NSSI was associated with later development of anxiety disorders. The associations found were not explained solely by confounding by common environmental risk factors or by underlying latent psychological traits. The authors recommend the development of prevention strategies targeting young people presenting with NSSI.

In an editorial in the BJPsych this month, Pitman (pp. 197–198) highlights research supporting an association between suicide of a relative or partner and subsequent risk of mental illness and suicide in those bereaved. Policy responses to knowledge of this association have been limited to date, and trials of interventions for those bereaved by suicide have been stalled. Pitman calls for an urgent focus on filling gaps in the evidence needed to inform the development of services for the suicide-bereaved.

Cannabis and psychosis

Whereas the association between cannabis and psychosis has been widely studied, establishing the temporal nature of any association remains an important research goal. Mustonen et al (pp. 227–233) used data from the Northern Finland Birth Cohort of 1986 and found that adolescents who had tried cannabis at least five times were more likely to develop later psychosis, even after adjustment for prodromal symptoms, other substance use and parental psychosis. The authors also describe a dose–response relationship between adolescent cannabis use and later psychosis.

In an editorial, Colizzi and Murray (pp. 195–196) argue that psychiatrists need to be more alert to cannabis use and its potential consequences. They describe the variations in legal approaches to cannabis use in North America and the UK, and suggest that, in future, the effects of individual approaches need to be examined and compared if we are to determine the best evidence-based policies.

Influence of terrorism and depression in diabetes

Six months after terrorist attacks occurred in Paris in 2015, Vandentorren et al (pp. 207–214) found evidence that almost one-fifth of civilians reported post-traumatic stress disorder (PTSD) symptoms, almost one-third had anxiety disorders and 10% had depression. Comparative rates for rescue workers were lower (3% PTSD symptoms and 14% anxiety disorders). The authors comment on the significant effects of the terrorist attacks even on civilians not directly threatened and on the range of potential effects, including occupational outcomes for both civilians and rescue workers.

Individuals with diabetes have increased rates of depression, and a range of interventions, including web-based initiatives, have been found to be effective in reducing depressive symptoms. Nobis et al (pp. 199–206) conducted a health economic evaluation alongside a randomised controlled trial of such an intervention and found that, compared with an active control, the web-based intervention was highly likely to be cost-effective from a societal perspective, considering the willingness-to-pay ceiling determined for a treatment response and that determined for a quality-adjusted life year.