Major depressive disorder is preventable in 25% of cases with use of cognitive behaviour therapy (CBT).1 However, prevention of depression is hugely challenging. Most depression cases have an onset in young people unaware that they are at risk;2 furthermore, the causal pathways are poorly understood and are not associated with any specific biomedical or psychosocial marker, doctors do not routinely detect depression risk, help seeking is low,3 and CBT can be difficult to access.
One promising strategy to prevent depression in people at risk might be to target insomnia, in view of the high co-occurrence of the two disorders. Evidence suggests that insomnia is associated with, but distinct from, depression, and is also linked to other psychiatric and behavioural disorders, such as schizophrenia, anxiety, and suicide.4 Nevertheless, depression is the highest comorbid disorder:5 35–47% of people with insomnia have clinically significant depression and 60–84% of people with major depressive disorder have significant insomnia symptoms.6, 7 Insomnia is also a strong risk factor for the development of depression, commonly preceding its development.8 Insomnia complicates depression treatment,9 does not necessarily remit when depression improves,8 and increases risk of depression relapse,10 suggesting that it is an independent problem. Insomnia treatments for people with insomnia can be effective in reducing depressive symptoms independently of their direct effects on sleep symptoms, although the evidence is not conclusive.
To date, there have been four trials11, 12, 13, 14 of CBT for insomnia (CBT-I) interventions targeting insomnia in patients with residual or concurrent depression. The main aim of these trials was to establish whether residual insomnia symptoms could be relieved in individuals with diagnosed depression. Two trials11, 12 reported statistically significant reductions in both insomnia and depression. The other two trials13, 14 had no significant findings, but showed effects in the predicted direction. In a trial15 of 43 adults that examined the effectiveness of CBT-I versus depression CBT, both delivered via the internet, in patients with a dual diagnosis of major depressive disorder and insomnia, CBT-I was as effective as depression CBT in reducing depression symptoms. Despite these findings, the mechanisms by which insomnia treatment might affect depression are unclear. The treatment might act artifactually by simply reducing insomnia symptoms that contribute to measured depression scores, or it might reverse prolonged disruption of 24 h sleep–wake, activity, and neurohormonal and metabolic cycles.
Research in context
Evidence before this study
We searched PubMed between Jan 1, 2012, and Aug 12, 2015, for new randomised controlled trials of depression prevention in adults published after Van Zoonen and colleagues' systematic review in 2012. Our search terms were (((((((((((((((depressive disorder[Title]) OR depressive[Title]) AND disorder[Title]) OR depression[Title]) OR depressive)[Title]) AND (prevention[Title]) AND control)[Title]) OR preventive[Title]) OR prevention[Title]) AND randomized controlled trial[Publication Type]) AND clinical trial[Publication Type]) AND (“2012/03/01”[Date - Publication] : “2015/08/12”[Date - Publication])) AND English[Language])) AND depress*[Title]. We also searched for randomised controlled trials in which an insomnia treatment was given to individuals diagnosed with depression from 2012 onwards, to ascertain whether any trials were missed in our original search. Search terms for this additional search were ((((((((insomnia[Title]) OR sleep*[Title]) AND depress*[Title]) AND randomized controlled trial[Publication Type]) OR clinical trial[Publication Type]) AND English[Language]) AND depress*[Title]) AND sleep*[Title]) AND (“2012/01/01”[Date - Publication] : “2015/08/12”[Date - Publication]). We identified no trials that focused on an insomnia intervention to prevent depression.
Added value of this study
We have shown for the first time that depression, suicidal ideation, and anxiety symptoms can be reduced in individuals at risk for depression by use of a behavioural online sleep program.
Implications of all the available evidence
An insomnia intervention could be the preferred method to prevent depression in individuals with insomnia because it is effective, has low stigma, and can be disseminated through the internet to the appropriate population. Long-term follow-up results will provide more evidence for any continuation of the preventive effect. Larger trials are needed to establish effects on the incidence of major depressive disorder.
No trial has yet assessed whether depression can be prevented in people with both subclinical depression symptoms and insomnia. Thus, we did the GoodNight Study to establish whether insomnia could be a primary preventive target for individuals with high depression symptomatology, but not meeting criteria for a diagnosis of major depressive disorder. To maximise access and the potential for dissemination, we used an internet-based automated intervention, SHUTi, which focuses on sleep restriction and consolidation, rather than the direct modification of depression, anxiety, or rumination.16 We speculated that an insomnia intervention would be more acceptable to the public, on the basis of evidence that people frequently seek help from family doctors for insomnia.17 Although several reasons exist as to why people with insomnia might not seek help, stigma is reported to be the reason in only 17% of cases.18 By contrast, stigma is the most frequently cited reason for not seeking help for depression.3
We recruited adults with depression symptom levels that were elevated but subclinical (indicated sample), who concurrently had insomnia (selective sample). We predicted that people given the SHUTi intervention would have fewer depression symptoms at the end than at the beginning of the intervention, and at 6 weeks and 6 months, and that the intervention would reduce major depressive disorder, suicide risk, and anxiety, and improve disability at these timepoints.