Which people with insomnia face a higher risk of dying?
Only the ones whose sleep was actually measured as short. In this study, people with insomnia symptoms whose overnight lab test showed under 6 hours of sleep had about 157% higher risk of dying from any cause, or roughly 2.6 times the risk (hazard ratio 2.57).
The comparison group was people with no insomnia symptoms and normal measured sleep. People with insomnia who slept a normal amount had no increased risk. People who slept under 6 hours but had no insomnia symptoms had no increased risk either. The danger showed up only when both things were present at once.
Then comes the finding that changes how this should be read. When researchers used what people said about their own sleep instead of what the machine recorded, the link to death disappeared. Insomnia plus self-reported short sleep, defined as under 7 hours, was not associated with mortality at all.
What the data show
Researchers followed 373 adults with insomnia symptoms and 394 adults without them, for a median of 8 years. Insomnia symptoms meant trouble falling asleep, trouble staying asleep, or waking too early, more than 5 times a month. Objective short sleep meant total sleep time under 6 hours on overnight polysomnography, the full sleep study with sensors that record brain waves and breathing.
The insomnia plus measured short sleep group carried the 2.6-fold higher death risk after the researchers adjusted for other factors that could explain it. The confidence range around that number runs from 1.12 to 5.89. In plain terms, the true increase is very likely somewhere between about 12% higher and nearly six times higher. That is a wide range, so the exact size of the risk is uncertain, but the whole range sits above no difference.
Dr. Kumar’s Take
The gap between measured sleep and remembered sleep is the real story here. People are poor judges of their own sleep. Someone who lies awake feeling wrecked may be logging seven hours, and someone who swears they slept fine may be getting five. This study says the body responds to the actual number, not the impression.
That matters because almost every large sleep study in existence relies on a questionnaire. If self-reported short sleep carries no mortality signal in a group where measured short sleep carries a strong one, a lot of sleep research is working with a blurry instrument.
The practical read is narrower than the headlines will make it. Insomnia by itself, with normal sleep time, did not raise death risk here. Short sleep by itself, without insomnia, did not either. The combination is what marked the higher-risk group.
How strong is the evidence?
This is an observational cohort, not a trial. Nobody was assigned to sleep less. That means the study can show that the combination of insomnia and short measured sleep travels with earlier death, but it cannot prove one causes the other. Something else, an undiagnosed illness that both fragments sleep and shortens life, could sit underneath both.
The sample is also modest for a mortality study. Only 767 people were followed, and deaths in a group that size over 8 years are relatively few, which is why the confidence range is so wide. Sleep was measured on a single night in an unfamiliar lab, which is not always a person’s typical night.
What makes the result worth attention is the internal consistency. Four different groupings were tested, and only one lit up. That pattern is harder to explain away as noise than a single lucky comparison would be.
Practical Takeaways
- If you have trouble falling asleep, staying asleep, or waking too early more than five times a month, that is the pattern this study counted as insomnia symptoms, and it is worth raising with your doctor rather than absorbing as normal.
- Your own sense of how long you sleep did not predict risk in this study, so do not use it either to reassure yourself or to frighten yourself.
- A formal sleep study is the only way to know your actual sleep time, and it also picks up sleep apnea, which is a common and treatable driver of broken sleep.
- Treating insomnia with cognitive behavioral therapy for insomnia, the first-line approach, aims at the sleep itself rather than at how you feel about it.
Related Studies and Research
- Sleep duration and dementia risk: 7 hours protects your brain long-term
- Long daytime naps linked to higher death risk in older adults
- Can your sleep study predict your risk of dying?
- One-third of US adults get less than 7 hours of sleep: geographic patterns revealed
FAQs
Why is measured sleep so different from what people report?
Most people estimate their sleep by how long they were in bed and how rested they feel. Both are unreliable. Time spent awake in the dark blurs into sleep in memory, and brief awakenings that a sleep study records are usually forgotten by morning. People with insomnia in particular tend to underestimate how much they slept, while short sleepers without insomnia often overestimate. This study is a reminder that the two numbers are measuring different things, and only one of them tracked with mortality.
Does this mean insomnia is harmless if I sleep enough hours?
Not harmless, but this study did not find a mortality signal for insomnia with normal measured sleep over 8 years. Insomnia still carries a real cost in daytime function, mood, and quality of life, and those are worth treating on their own terms. What the data suggest is that insomnia is not one condition. The version with short measured sleep looks biologically more severe, and the authors describe objective short sleep duration as a marker of that severity.
Should I ask for a sleep study if I have insomnia?
That is a conversation to have with your doctor, and the answer depends on your other symptoms. Sleep studies are usually ordered when there is reason to suspect sleep apnea, such as snoring, witnessed pauses in breathing, or heavy daytime sleepiness, rather than for insomnia alone. This research does raise the question of whether measured sleep time adds information that a questionnaire cannot. It does not settle it, since a single study of 767 people is not enough to change who gets tested.
Bottom Line
In 767 adults followed for a median of 8 years, insomnia symptoms combined with under 6 hours of lab-measured sleep carried about 2.6 times the risk of death from any cause, while insomnia alone, short sleep alone, and insomnia with self-reported short sleep carried no measurable increase. The distinction between what a sleep study records and what a person reports turned out to be the difference between a signal and nothing at all. It is one observational cohort with a wide confidence range, so the size of the risk is unsettled, but the direction is clear enough to take insomnia with truly short sleep more seriously than insomnia with normal sleep.
