'I Barely Slept, but My App Says 7 Hours' — Insomnia and Device-Recorded Sleep Time

Originally written in Korean and translated for international readers.

"I spent most of last night awake, but my app says I slept for 7 hours." Both device readings and the sleep duration perceived by people with insomnia may differ from a clinical sleep study. Insomnia is diagnosed by considering difficulty sleeping together with disruption to daytime life, rather than an app reading.

Sleep symptoms and daytime functioning are central to diagnosing insomnia

The clinical history assessed by a clinician and symptoms reported by the patient form the basis of the major international diagnostic criteria (ICSD-3-TR, DSM-5-TR). Clinicians assess whether difficulty falling asleep or frequent awakenings are accompanied by impaired daytime functioning. Neither the time it takes to fall asleep nor the number of nighttime awakenings must reach a particular numerical threshold, and a clinical sleep study is not always necessary for a routine insomnia diagnosis. When evaluating chronic insomnia, clinicians check whether these problems have occurred at least 3 times a week for 3 months and use a sleep diary kept for 7–14 days as supporting information^[1]. A consumer sleep device display alone cannot determine whether insomnia is present or absent^[2].

Why can wrist-worn recordings differ from actual sleep?

Wrist-worn devices estimate sleep using signals such as movement and heart rate. Their measurements of sleep duration, sleep efficiency (the proportion of time in bed actually spent asleep), and time awake after falling asleep may differ from those of a clinical sleep study. In an analysis combining several wrist-worn devices, the average difference from clinical sleep studies was approximately 17 minutes for sleep duration and approximately 13 minutes for time awake after falling asleep. In that analysis the devices recorded sleep duration as shorter, not longer, than the clinical measurement^[3].

In one study of a Fitbit device, recorded sleep duration was approximately 7 minutes longer in good sleepers and approximately 33 minutes longer in people with insomnia. Sleep efficiency was also overestimated by approximately 8 percentage points in people with insomnia^[4]. These figures come from the specific device examined in that study. They cannot be applied equally to every device or person.

Perceived sleep can also differ from actual sleep

Device errors do not mean that a person's own perception is always accurate. People with insomnia tend to feel they slept less than they actually did. Reporting a shorter sleep duration than measured in a clinical sleep study is common, but it does not occur in everyone^[5].

불면에서 자주 보이는 차이의 방향 — 사람마다 다르다

본인 보고: 실제 잔 시간보다 짧다고 느낄 수 있음
병원검사: 비교의 기준
손목 기기: 실제 잔 시간보다 길게 나타낼 수 있음

Both self-reports and devices have limitations. For this reason, assessment considers sleep symptoms, daytime functioning, and how long the symptoms have persisted, rather than relying on a single night's sleep-duration reading^[1].

What should guide the assessment?

• A normal reading alone does not rule out insomnia. Consumer sleep devices are not diagnostic tools^[2].
• An abnormal device reading alone should not determine treatment. Device data can support a consultation when a clinician interprets them alongside symptoms^[2].
• Keep the clinical history, sleep symptoms, and daytime functioning at the center of the assessment. A sleep diary kept for 7–14 days can also help inform the assessment^[1].

Can feedback on app data help improve insomnia?

The NITE randomized trial evaluated a program that provided sensor data, guidance on interpreting them, and ongoing follow-up. Insomnia severity decreased more in the program group than in the comparison group receiving sleep education and sleep hygiene advice, but the effect was modest^[6].

• Mean reduction in insomnia severity score within the program group: 6.11 points
• Difference after the program, with a lower score than the comparison group (adjusted for baseline scores): 2.03 points

The reduction within the program group was 6.11 points, whereas the difference between the two groups after the program, adjusted for baseline scores, was 2.03 points. However, the 2.03-point difference cannot be interpreted as an effect attributable solely to app data.

The program aimed to help participants interpret the discrepancy between app readings and their own perception of sleep. However, the reduction in that discrepancy was not clearly greater than in the comparison group (between-group P=.23–.93, effect sizes 0.02–0.24)^[6]. The researchers considered this approach alone unlikely to provide sufficient treatment for insomnia. They also interpreted it as likely to be less effective than cognitive behavioral therapy for insomnia (CBT-I).

An insomnia assessment also considers sleep apnea

COMISA refers to coexisting insomnia and sleep apnea. In a meta-analysis, approximately 29–35% of people with insomnia also had sleep apnea, depending on the diagnostic criteria used^[7]. Earlier studies have reported that people with both conditions have greater impairment in daytime functioning and a lower quality of life than those with either condition alone.

The European guideline recommends including an assessment for suspected sleep apnea when evaluating insomnia. If you snore or someone observes pauses in your breathing during sleep, you should check whether testing for sleep apnea is needed. If both are confirmed, which condition to treat first and how to combine treatments is decided together with your clinician^[1].

Read the medical info for this product 스마트워치 수면무호흡·산소포화도 알림, 검사받아야 하나

This article provides general information, not individualized medical advice. If symptoms occur at least 3 times a week for 3 months and interfere with daytime life, they may meet the minimum criteria warranting assessment for chronic insomnia disorder^[1]. In this situation, consult a healthcare professional.