AIS (Athens Insomnia Scale): Take It, Score It, Understand Your Results
The Athens Insomnia Scale (AIS) is a brief 8-item self-report questionnaire that measures how much your sleep has been disturbed over the past month. Developed by Soldatos and colleagues in Athens and validated against ICD-10 criteria for nonorganic insomnia, the AIS takes about 3 minutes to complete and produces a 0-24 score. A score of 6 or higher is the validated positive screen for insomnia (sensitivity 93%, specificity 85% against an ICD-10 nonorganic insomnia diagnosis). The AIS is a screening tool, not a diagnosis of any specific sleep disorder: results should be reviewed with a healthcare professional.
How the AIS is scored: eight items, 0 to 24
The Athens Insomnia Scale was developed by Constantin Soldatos, Dimitrios Dikeos and Thomas Paparrigopoulos and published in the Journal of Psychosomatic Research in 2000. Its distinguishing feature is its anchor: the items map directly onto the ICD-10 diagnostic criteria for insomnia, rather than onto a general notion of poor sleep. That is what separates it from most other sleep questionnaires.
Each of the eight items is rated 0 to 3, where 0 means no problem and 3 means the most severe form of that problem. The total is the simple sum, giving a range of 0 to 24. The recall window is the past month.
One instruction is easy to miss and changes the result: an item should only be scored above zero if the problem occurred at least three times per week during that month. That frequency threshold is not incidental. It is ICD-10 criterion B, and it is what stops the scale from counting a handful of bad nights as insomnia.
The eight areas
The first five items cover the night itself, and the last three cover what the night cost you the following day. That split is deliberate: insomnia is defined by daytime consequence, not by hours slept.
| # | Item | Domain |
|---|---|---|
| 1 | Sleep induction (time to fall asleep) | Night |
| 2 | Awakenings during the night | Night |
| 3 | Final awakening earlier than desired | Night |
| 4 | Total sleep duration | Night |
| 5 | Overall quality of sleep | Night |
| 6 | Sense of wellbeing during the day | Day |
| 7 | Functioning capacity during the day | Day |
| 8 | Sleepiness during the day | Day |
Score bands
| Score | What it suggests |
|---|---|
| 0–5 | Below the screening threshold |
| 6–9 | Probable insomnia; the validated positive-screen cutoff is 6 |
| 10–15 | A more severe presentation, meeting ICD-10 criteria more strongly |
| 16–24 | Severe |
The cutoff most often cited is 6 or above. Soldatos and colleagues reported in 2003 that at that threshold the AIS achieved 93% sensitivity and 85% specificity against ICD-10 insomnia criteria, with a Cronbach’s alpha of approximately 0.89.
The AIS-5 short form
A five-item version exists, covering only the night-time component. It is used where even eight items is too many, typically for rapid screening in primary care. The eight-item version is the one used here, because the daytime items are the ones that distinguish insomnia from short sleep.
What a positive screen means: a symptom is not a disorder
An AIS score describes subjective severity against ICD-10 criteria. It is not a diagnosis, and the distinction that matters most is between insomnia as a symptom and insomnia as a disorder.
Insomnia as a symptom is ordinary. A stressful stretch at work, a time-zone change, a painful injury, or an infection can all produce broken sleep that resolves when the cause does. Most people experience this repeatedly across a lifetime and need nothing but time.
Insomnia disorder has formal criteria under DSM-5 and ICSD-3, and an elevated AIS does not by itself meet them:
- Difficulty initiating sleep, maintaining it, or waking earlier than intended and being unable to return to sleep.
- Occurring at least three nights per week.
- Present for at least three months.
- Causing clinically significant distress or impairment.
- Occurring despite adequate opportunity to sleep.
- Not better explained by another sleep disorder, a medical condition, a medication, or substance use.
That last condition is the one a questionnaire cannot check, and it is the reason a high score is a starting point rather than a conclusion.
What a clinician does next
A high AIS typically leads to a sleep history covering how long you take to fall asleep, how often and for how long you wake, your usual bed and rise times, and what you do when you cannot sleep. A sleep diary kept for one to two weeks is the practical standard for characterising the pattern, because retrospective recall is unreliable in exactly the direction insomnia distorts it.
Conditions that must be ruled out first
Several sleep disorders produce what looks like insomnia but need entirely different treatment. Missing them is the most consequential error at this stage.
- Obstructive sleep apnea. Breathing pauses fragment the night. The clues are loud snoring, witnessed pauses, excessive daytime sleepiness, and treatment-resistant high blood pressure. It is assessed with a sleep study, in a laboratory or at home.
- Restless legs syndrome. An uncomfortable urge to move the legs, worse at rest and at sleep onset, relieved by movement. It is associated with low iron stores, so ferritin is often checked.
- Circadian rhythm disorders. The problem is not the inability to sleep but the timing of the body clock: delayed phase in adolescents and young adults, advanced phase in older people. Shift work and jet lag belong here, not under insomnia.
- Narcolepsy and the parasomnias (sleepwalking, night terrors, REM sleep behaviour disorder).
Comorbidity worth checking
Roughly half of people with chronic insomnia also have depression, and the relationship runs in both directions. Item 3 of the PHQ-9 asks about sleep, so completing both instruments gives a fuller picture than either alone: the AIS measures insomnia severity, the PHQ-9 measures depression.
Anxiety contributes through rumination and night-time cognitive arousal. Chronic pain fragments sleep mechanically. An overactive thyroid can present as insomnia with palpitations and night sweats, which is why TSH is sometimes checked. Menopause, COPD, reflux and heart failure all have well-described night-time components.
Limitations, and what the AIS cannot see
- It does not measure objective sleep architecture. Sleep stages, efficiency, and true latency require polysomnography or actigraphy.
- It does not diagnose insomnia disorder, and it does not separate primary from secondary insomnia.
- It does not rule out apnea, restless legs, circadian disorders, narcolepsy or parasomnias.
- It does not assess hypersomnia, where excessive daytime sleepiness is the main complaint.
Several things shift the score without shifting the underlying condition. Acute stress and recent life events inflate it temporarily. Stimulants, corticosteroids, bronchodilators, beta blockers and SSRIs can all cause medication-induced insomnia. Caffeine, nicotine and alcohol near bedtime affect sleep quality even when they shorten the time to fall asleep. And self-report bias is well documented and directional: people with insomnia tend to overestimate how long they lay awake and underestimate how long they slept.
When to talk to a clinician
Consider making an appointment if your score is 6 or above, if the difficulty has lasted more than three months, or if there is meaningful daytime cost such as fatigue, poor concentration, irritability or near-misses while driving.
Move sooner if there is loud snoring, witnessed breathing pauses, restless legs at sleep onset, or heavy daytime sleepiness. Those point at a primary sleep disorder and usually warrant a sleep study rather than insomnia treatment.
If sleep difficulty sits alongside persistent low mood or marked anxiety, completing the PHQ-9 as well gives a clinician more to work with.
If you are in crisis
If you are having thoughts of suicide or of harming yourself, do not wait for an appointment. In the US, call or text 988 (Suicide and Crisis Lifeline). In the UK and Ireland, call Samaritans on 116 123. In the EU, the emergency number is 112.
First-line treatment
Both the American Academy of Sleep Medicine and NICE recommend cognitive behavioural therapy for insomnia (CBT-I) as first-line treatment for chronic insomnia, ahead of medication. It is more effective than sleeping tablets over the long term and does not carry their tolerance and dependence risks.
This page does not cover doses for any sleep medication, does not make recommendations about cannabis or cannabidiol, and does not give instructions for stopping a hypnotic on your own, which carries a real risk of rebound insomnia. Those are conversations for the person prescribing.
Frequently asked questions
What is the Athens Insomnia Scale?
It is an eight-item self-report questionnaire measuring insomnia severity, developed by Soldatos, Dikeos and Paparrigopoulos and published in 2000. Its items map onto the ICD-10 diagnostic criteria for insomnia, which is unusual among sleep questionnaires and is the main reason it is used in sleep clinics and research.
How is the AIS scored and what is a normal score?
Each of the eight items scores 0 to 3, giving a total from 0 to 24 over the past month. A score below 6 is below the screening threshold. 6 or above is the validated positive-screen cutoff, and 10 or above indicates a more severe presentation.
Does a high AIS score mean I have insomnia disorder?
No. It means insomnia severity is in a range that warrants a proper assessment. Insomnia disorder requires symptoms at least three nights a week for at least three months, with daytime consequences, despite adequate opportunity to sleep, and with other sleep disorders ruled out. A questionnaire cannot establish any of that.
How does the AIS compare with the ISI and the PSQI?
All three measure sleep, differently. The AIS is anchored to ICD-10 criteria. The Insomnia Severity Index measures perceived severity and is the most common outcome measure in insomnia trials. The Pittsburgh Sleep Quality Index is broader and longer, covering sleep quality overall rather than insomnia specifically. The ISI and the PSQI are both licensed instruments and are not offered as tools here; the AIS is free to use, which is why it is the one you can complete on this page.