[Validation of the Polish version of the Athens Insomnia Scale]

To validate the Polish version of the Athens Insomnia Scale (AIS), the instrument designed for quantitative measurement of the severity of insomnia based on the ICD-10 criteria.

The AIS was administered to 356 subjects: 160 patients from the sleep clinic presenting with ICD-10 non-organic insomnia (90f; mean age: 44.9 +/- 15.7y) and 196 self-defined good sleepers (106f; mean age: 43.9 +/- 13.4y). To assess psychometric properties of the scale, principal component analysis, Cronbach's alpha, Pearson's correlation (test-retest reliability: two weeks interval; n = 48) and sensitivity-specificity analysis were computed.

Using the factor analysis only one factor was extracted, which accounted for 60.2% of the variance. The internal consistency (Cronbach's alpha = 0.90) and the test-retest reliability (r2 = 0.92) of the AIS were found to be very satisfactory. These values remain practically unchanged when any of the items was removed from the analysis. As expected, subjects with insomnia scored significantly higher on the AIS than good sleepers (14.2 +/- 3.9 pts vs. 4.8 +/- 3 pts; t = 24.9; p < 0.001). Considering the balance between sensitivity and specificity, a score of 8 was indicated as the optimum cut-off, with 89% overall correct case identification--94% patients with insomnia and 84% healthy controls.

Standardisation of the diagnostic process of insomnia is a highly important task in clinical practice, epidemiological considerations and treatment outcomes assessment. The AIS is a brief, self-reported measure of insomnia that may improve these assessments. The good psychometric properties of the Polish version of the AIS were confirmed in the present study.