Eyes wide open: sleep disturbance, not melatonin, is the real issue to address for suicide risk

In this issue of the Journal of Clinical Sleep Medicine, Høier et al attempt to address associations between melatonin use and suicidal behavior, based on data from a query of a National Prescription Registry in Denmark.¹ Adjusting for age group and sex, they identified a 4-fold higher rate of suicide and a 5-fold higher risk of first suicide attempts in individuals prescribed melatonin compared with individuals not prescribed melatonin.

Two major strengths of their study are as follows: (1) an expansive dataset from a Denmark national health prescription registry of a very large population sample and (2) the rarely afforded opportunity for systematic evaluation of melatonin use, which is available only by prescription in Denmark. Additionally, the findings of this study corroborate current knowledge in the field about psychiatric disturbances, sleep disturbances, and sleep aids across multiple classes being associated with suicidal behaviors.

However, there are 2 main issues to explore about the identified association between suicide and treatment with melatonin in the Høier et al study. First, relative to the sample size, there were a limited number of individuals treated with melatonin who attempted or died by suicide. Of almost 6 million individuals in the registry, approximately 6,000 individuals died by suicide, 22 of whom were taking melatonin. Over 25,000 individuals were reported to have a first suicide attempt, of whom 134 were taking melatonin. Put another way, 99.5% of individuals with suicidal behaviors were not in treatment with melatonin, while only 0.5% of them were in treatment with melatonin. Second, and most notably, almost all individuals with suicidal behavior in treatment with melatonin had a comorbid mental disorder. Unfortunately, the observational design of the study precluded any opportunity to account for the severity of mental disorder or sleep disturbance symptoms.

The known associations between suicide risk and sleep disturbance (with or without mental/mood disorders) easily obviate an association between suicide risk and melatonin use. Sleep disturbance, short sleep duration, and delayed sleep timing are associated with increased suicidal risk independently of and synergistically with psychiatric disturbances.^2,3 A recent prospective cohort study spanning 20 years identified that insomnia with short sleep time not only increased suicidality directly, but also indirectly, by affecting the risk of onset or recurrence of depression.⁴ Amygdala activation in response to negative emotional stimuli has been shown to be significantly greater with insufficient sleep compared with when well-rested.⁵ Because circadian rhythm modulates positive moods and reward-seeking,⁴ it is not surprising that (1) multiple psychiatric disorders are associated with genetic mutations in circadian clock genes,⁶ (2) evening chronotype and delayed bedtimes are positively correlated with mood disorders,^7,8 and (3) suicide risk is higher at night.^3,9,10

Most evidence supporting associations between sleep disturbance, psychiatric disturbance, and suicidality are observational studies that cannot establish causality. However, recent studies using the analytic method of Mendelian randomization provide the opportunity to explore direct causal effects between sleep disturbance, psychiatric disturbances, and suicidality.^11–13 One such study showed bidirectional causal associations between insomnia and depressive symptoms, potentially through shared wake-promoting mechanisms resulting in hyperarousal.¹⁴ Another recent study found insomnia to be a causal risk factor for suicidal behavior independently of depression.¹¹

Similar to the findings by Høier et al, much of the evidence about increased suicide risk associated with hypnotic medication use has not adequately controlled for psychiatric disorders linked to insomnia. However, recent data from 2 nationally representative samples identified that multiple classes of prescription medications for insomnia were associated with suicidal thinking and behaviors, even when adjusting for measures of mental health.¹⁵

The shared pathways of psychiatric disorders and increased suicide risk being associated with nocturnal wakefulness are thought to be driven by alterations in executive functions, including prefrontal disinhibition, attentional biases, altered reward processing, as well as negative affect.^2,9 Hypnotic medications, particularly benzodiazepine receptor agonists, may exacerbate these effects, as individuals in treatment for sleep disturbance are more likely to have nocturnal wakefulness and psychiatric disturbance.

Given the conundrum that (1) effective pharmacological treatment of insomnia has been shown to have a positive effect on depressed mood,¹⁶ (2) nocturnal wakefulness increases suicidality, but (3) hypnotics across several classes are associated with suicidality, what is the best risk-to-benefit ratio: (1) avoid pharmacological treatment of insomnia, (2) first prescribe melatonin or other nonbenzodiazepine receptor agonists, or (3) prescribe benzodiazepine receptor agonists for particularly worse insomnia based on increased efficacy? While cognitive behavioral therapy for insomnia is first-line treatment for insomnia, it may not be available, appropriate, or timely, particularly in the circumstance of increased suicide risk.

The association of benzodiazepine use with suicide attempts (but not suicidal ideation or planning) corroborates that disinhibition and impulsivity from central nervous system depressant effects may exacerbate nocturnal alterations in executive functions and further contribute to suicidality.¹⁵ Conversely, the association of trazodone and so-called “Z” drugs with suicidal ideation¹⁵ and the findings from Høier et al may indicate that (1) non-benzodiazepine hypnotics and melatonin do not as effectively reduce nocturnal wakefulness and (2) may not reduce sleep disturbance enough to reduce suicidality.¹⁶ It is also possible that ineffective treatment of insomnia contributes to a sense of hopelessness (a known modifiable factor associated with increased suicide risk).¹⁷

This point has been corroborated by several well-designed clinical trials in which all participants were treated for depression and randomized to a Z drug or placebo for insomnia. Insomnia severity correlated with severity of suicidal ideation, even when controlling for depression symptoms.^7,18,19 Although treatment effects for reducing suicidal ideation could not be firmly established, Z drugs were superior to placebo in reducing insomnia symptoms, particularly for participants with severe insomnia at baseline.¹⁷

Although Høier et al conclude that “direct causal association between melatonin and suicide cannot be excluded,” we suggest that a direct causal association should not be assumed or even considered a likely major factor, based on much more robust evidence that sleep disturbances and mental health disorders are (1) independently associated with suicide risk, (2) additive in their associations, and (3) explain the reason for treatment with melatonin in the first place.

DISCLOSURE STATEMENT

The authors report no conflicts of interest.