Ask most people what insomnia looks like and they will describe someone lying awake for hours. But a new study published in JMIR Formative Research suggests the core problem may not be how little insomnia patients sleep on average — it is how wildly their sleep fluctuates from one night to the next.
Researchers at Washington State University's Sleep and Performance Research Center tracked 112 adults — 83 with diagnosed chronic insomnia and 29 healthy controls — across eight consecutive weeks of sleeping in their own beds. The results redefine what clinicians should be measuring.
Same Total Sleep, Different Experience
The most striking finding: the insomnia group averaged roughly the same total sleep time as healthy controls — about 6.6 hours per night in both groups.
Where the groups diverged was consistency. Participants with chronic insomnia demonstrated significantly higher night-to-night variability in sleep efficiency, sleep onset latency, and intermittent wakefulness compared to controls (all P<.001). One night they might fall asleep in 10 minutes; the next it could take an hour. Sleep efficiency could swing from 90% to below 70% within the same week.
The control group, by contrast, showed tight clustering around their averages. Their sleep was not necessarily longer — it was predictable.
Why This Matters for Diagnosis
The study, led by Devon A. Hansen, used a contactless radiofrequency monitoring device (SleepScore Max) that tracked sleep objectively each night without requiring participants to visit a lab or wear a sensor to bed. This allowed the researchers to capture patterns that a single-night sleep study would miss entirely.
"Objectively and longitudinally capturing this variability in an ecologically valid manner is key to investigating and understanding the disorder," the researchers wrote, arguing that night-to-night variability is a hallmark of the chronic insomnia phenotype.
This has direct implications for how insomnia is evaluated. A patient who sleeps six and a half hours in a sleep lab might appear unremarkable on paper. But if that same patient swings between four and eight hours across a normal week — with corresponding swings in how long it takes to fall asleep and how often they wake up — the data tells a very different story.
The Limits of Single-Night Studies
The finding also raises questions about the standard diagnostic approach. Most clinical sleep evaluations rely on polysomnography conducted over one or two nights in a laboratory setting. While polysomnography remains essential for diagnosing conditions like sleep apnea and periodic limb movement disorder, this study suggests it may be poorly suited for capturing insomnia's defining characteristic.
The eight-week monitoring period revealed that variability patterns stabilized and became statistically detectable after about two weeks of continuous tracking. A single night — or even a few nights — in an unfamiliar laboratory may not capture the signal at all, particularly given the well-documented "first-night effect" in which sleep architecture changes when patients sleep away from home.
A New Target for Treatment
If variability is indeed a core feature of insomnia, it could also serve as a treatment target. Current insomnia therapies, including cognitive behavioral therapy for insomnia (CBT-I) and pharmacological approaches, are typically evaluated by improvements in average sleep metrics — total sleep time, sleep efficiency, or subjective sleep quality.
But a treatment that improves average sleep time without reducing variability might leave patients still experiencing the unpredictability that drives much of insomnia's daytime burden: the anxiety of not knowing whether tonight will be a good night or a bad one.
Conversely, a therapy that stabilizes the pattern — even without dramatically increasing total sleep — might deliver outsized improvements in quality of life. Sleep restriction therapy, a component of CBT-I that deliberately narrows the sleep window to build consistent sleep pressure, may work in part through exactly this mechanism.
What This Means for Patients
If you have been told your sleep study looks "normal" despite feeling exhausted and frustrated, this research may explain the disconnect. Insomnia's signature may not show up in a single snapshot — it lives in the pattern across weeks.
Patients interested in tracking their own variability can use validated consumer sleep monitors over at least two weeks, looking not just at average sleep time but at the consistency of sleep onset, efficiency, and overnight awakenings. Sharing this longitudinal data with a sleep specialist may provide a clearer picture than any single night in a lab.
The full study is available in JMIR Formative Research.
