Are Insomnia and Depression Actually the Same Sleep Disturbance in Different Degrees?

This piece covers a specific meta-analysis comparing polysomnographic sleep-disturbance measures across insomnia, depression, and narcolepsy. It does not cover consumer wearable sleep tracking, which uses different measurement methods than the polysomnography this research is based on.

A meta-analysis of controlled polysomnographic studies tested whether insomnia, major depression, and narcolepsy have genuinely distinct sleep-disturbance profiles, as they're usually clinically classified, or whether they represent the same underlying disturbance at different degrees of severity. Combining five polysomnographic variables into a single sleep disturbance index, the analysis found the three conditions arrayed along a simple continuum of progressively worse sleep disturbance, rather than showing categorically different patterns. The finding suggests polysomnographic sleep architecture may be a single good-to-bad axis rather than a set of condition-specific signatures, at least for these three conditions as measured this way.

The assumption behind treating these as separate conditions

Insomnia, major depression, and narcolepsy are clinically distinguished partly by their different reported polysomnographic profiles, the assumption being that each condition disrupts sleep architecture in its own characteristic way. That assumption underlies a lot of how sleep research and diagnosis approaches these conditions: find the profile, match it to the condition.

This meta-analysis set out to directly test whether the data actually supports that assumption, or whether an alternate, simpler explanation fits better: that these conditions don't have qualitatively different sleep disturbance signatures at all, just progressively more severe versions of the same underlying disturbance.

What 'a single axis of good-to-bad sleep' actually implies

The paper's own conclusion is careful and specific: these findings suggest sleep can be disturbed in only a limited number of ways, and that polysomnographic measures of sleep architecture may not be able to elaborate much beyond one axis running from good to bad sleep. That's a meaningfully different claim than 'insomnia and depression are the same disorder.' It's a claim specifically about what polysomnographic architecture measurements can and can't distinguish, not about the disorders' full clinical picture, which includes far more than sleep architecture alone.

The practical implication is narrower than it might first sound: if someone's polysomnogram shows a moderate level of sleep disturbance, that data point alone may not reliably tell a clinician whether they're looking at insomnia, depression, or somewhere on the path toward narcolepsy, because the underlying architecture measures don't cleanly separate by diagnosis. Other clinical information, not sleep architecture data alone, is doing more of the diagnostic work than the assumption of condition-specific profiles would suggest.

A different question from what wearables actually measure

It's worth being explicit that this research is built entirely on polysomnography, the lab-based, EEG-driven gold standard for sleep staging. That's a fundamentally different measurement approach from what a consumer wearable does, which relies primarily on movement and heart-rate-derived signals rather than direct brain-wave monitoring, a distinction covered in how wearables track sleep stages. Nothing here says anything about whether a wearable's sleep-stage estimate could detect this kind of severity continuum; the underlying measurement methods aren't comparable.

This also sits alongside, but is distinct from, the roster's coverage of whether treating insomnia improves heart health, which is about a downstream health outcome of insomnia treatment. This piece is about the underlying sleep-architecture classification question itself, one step earlier in the chain.

Common questions