Journal of Psychosomatic Research 48 (2000) 555 ± 560
Athens Insomnia Scale: validation of an instrument
based on ICD-10 criteria
Constantin R. Soldatos*, Dimitris G. Dikeos, Thomas J. Paparrigopoulos
Sleep Research Unit, Department of Psychiatry, Athens University Medical School, Athens, Greece
Received 21 June 1999
Abstract
Objectives: To describe and validate the Athens Insomnia
Scale (AIS). Methods: The AIS is a self-assessment psychometric
instrument designed for quantifying sleep difficulty based on the
ICD-10 criteria. It consists of eight items: the first five pertain to
sleep induction, awakenings during the night, final awakening,
total sleep duration, and sleep quality; while the last three refer to
well-being, functioning capacity, and sleepiness during the day.
Either the entire eight-item scale (AIS-8) or the brief five-item
version (AIS-5), which contains only the first five items, can be
utilized. The validation of the AIS was based on its administration
to 299 subjects: 105 primary insomniacs, 144 psychiatric patients
and 50 non-patient controls. Results: Regarding internal con-
Keywords: Scale; Validation; Insomnia; Sleep disorders; ICD-10
Introduction
Problems with poor sleep are quite frequent in the
general population; their prevalence ranges from 10% to
30% [1 ± 5] and it is much higher among patients with
psychiatric disorders [5 ±8]. In most studies assessing sub-
jective sleep difficulties, self-devised questionnaires and
sleep diaries are utilized. Such instruments may be adequate
for the purposes of a given study allowing comparisons
between conditions or groups of individuals, but they are
usually not standardized or uniform Ð making the integra-
tion of results from various studies impossible. Some
standardized instruments, however, have been developed
for the measurement of sleep difficulties in clinical settings:
the Leeds Sleep Evaluation Questionnaire [9,10], the St.
Mary's Hospital Sleep Questionnaire [11,12], the Sleep
* Corresponding author. Eginition Hospital, 72-74 Vas. Sophias Ave.,
11528 Athens, Greece. Tel.: +30-301-7289324; fax: +30-301-7251312.
E-mail address: egslelabath@hol.gr (C.R. Soldatos).
0022-3999/00/$ ± see front matter D 2000 Elsevier Science Inc. All rights reserved.
PII: S 0 0 2 2 - 3 9 9 9 ( 0 0 ) 0 0 0 9 5 - 7
sistency, for both versions of the scale, the Cronbach's a was
around 0.90 and the mean item ± total correlation coefficient was
about 0.70. Moreover, in the factor analysis, the scale emerged as
a sole component. The test ± retest reliability correlation coeffi-
cient was found almost 0.90 at a 1-week interval. As far as
external validity is concerned, the correlations of the AIS-8 and
AIS-5 with the Sleep Problems Scale were 0.90 and 0.85,
respectively. Conclusion: The high measures of consistency,
reliability, and validity of the AIS make it an invaluable tool in
sleep research and clinical practice. D 2000 Elsevier Science Inc.
All rights reserved.
Problems Scale [13], the Pittsburgh Sleep Quality Index
[14], and the Karolinska Sleep Diary [15,16].
The evolution of diagnostic concepts regarding insomnia
is reflected in the various classification systems, which have
been in use over the last two decades [17 ± 21]. The
classification published in 1979 by the Association of Sleep
Disorders Centers required a sleep laboratory based criterion
of reduced sleep quantity for the diagnosis of insomnia [17].
In drafting the DSM-III-R, the clinical approach prevailed,
according to which the diagnosis of insomnia is based on
the patient's subjective perception of unsatisfactory sleep
quantity and/or quality [18]. Ever since, for the diagnosis of
insomnia impaired sleep quality is given equal importance
as that of reduced sleep quantity and the patient's subjective
assessment is not discarded [19 ± 21].
Following the publication of the ICD-10 diagnostic
criteria for insomnia [20], the need arose to develop a scale
as a tool to assist clinicians in assessing the severity of
insomnia based on these globally accepted criteria. The
rationale for the development of such a tool, the Athens
Insomnia Scale (AIS), has been presented previously [22].
556 C.R. Soldatos et al. / Journal of Psychosomatic Research 48 (2000) 555±560

The purpose of the present paper is to provide a detailed its simple and straightforward format, which would be
description of the AIS and documentation of its validation. expected to pose no major problems for the translation.
Then, one of the authors (D.G.D.) translated the scale into
Greek. The Greek translation was examined by the bilingual
Methods group as well as by a three-member unilingual group, who
proposed only a few minor modifications of the Greek
Description of the scale wording. The modified Greek text was back-translated into
English by the third author (T.J.P.). Finally, the back-
The AIS is a self-administered psychometric instrument translation was carefully examined by the bilingual group,
consisting of eight items (Appendix A). The first five items who decided that the Greek translation of the AIS Ð as had
of the AIS (assessing difficulty with sleep induction, awa- been amended by the unilingual group Ðwas equivalent to
kenings during the night, early morning awakening, total the original English text.
sleep time, and overall quality of sleep) correspond to
criterion A for the diagnosis of insomnia according to Sample characteristics
ICD-10, while the requirements of a minimum frequency
(at least three times a week) and duration (1 month) of any Following the approval of the project by the institutional
complaint correspond to criterion B of the ICD-10 [20]. The review committee, the AIS was given to 105 primary health
ICD-10 requirements of marked distress caused by the sleep care patients presenting with a complaint of insomnia not
problem and/or interference with ordinary activities of daily attributed to any obvious underlying cause (primary insom-
living (criterion C) are covered through the strictly sub- nia), 100 psychiatric outpatients, 44 psychiatric inpatients,
jective nature of the response options for every item of the and 50 non-patient controls. Thus, the total sample consisted
scale as well as through the content of the last three items of 299 subjects. As Table 1 shows, 42% of the subjects were
pertaining to the next day consequences of insomnia (pro- males and 58% females with a similar sex ratio across the
blems with sense of well-being, functioning, and sleepiness four subgroups of subjects. The age distribution differed
during the day). somewhat among these subgroups; primary insomniacs were
Each item of the AIS can be rated 0 ± 3, (with 0 the oldest and psychiatric inpatients the youngest (Table 1).
corresponding to ``no problem at all'' and 3 ``very serious All subjects were informed on the research nature of the
problem''). The responders are requested to rate positive if project and all agreed to participate. Primary health care
they had experienced the sleep difficulty described in each patients were recruited through general practitioners and
item at least three times a week during the last month or other non-psychiatric physicians functioning in out-patient
some other period of time, whose length depends on the services of the Hellenic National Health System or in
purpose of a given study. private medical practice settings. Psychiatric patients were
Two versions of the scale can be used. Either the entire those attending the facilities of the Department of Psy-
eight-item scale (AIS-8) Ð with a total score ranging from 0 chiatry of the University of Athens. Finally, non-patient
(denoting absence of any sleep-related problem) to 24 controls were recruited from among the researchers'
(representing the most severe degree of insomnia) Ð or friends and acquaintances.
the brief five-item version (AIS-5), which is limited to the
first five items Ð with a total score ranging from 0 to 15. Assessment procedures
The AIS-8 is intended for use in a fully developed clinical
setting, while the AIS-5 is mainly used when there is a need The subjects were asked to record on the AIS their
to focus just on difficulty with sleep quantity and quality. estimate of any sleep difficulty they might have experienced
The last three items of the AIS-8 refer to daytime symptoms
that often emerge as a consequence of nocturnal sleep
Table 1
disturbance in insomniac patients. However, these symp-
Characteristics of the sample in terms of sex and age
toms may be also caused by sleep disorders other than
Age
insomnia, such as narcolepsy and obstructive sleep apnea.
The AIS was first developed in English by the senior Male Female Mean ‹ S.D. Range
author (C.R.S.), who was the expert responsible for drafting Primary insomniacs 43 62 58.9 ‹ 10.2 21 ± 79
the original diagnostic criteria for the sleep disorders section (n = 105) (41.0%) (59.0%)
Psychiatric outpatients 41 59 43.0 ‹ 13.3 19 ± 78
of ICD-10 [20]. For the purposes of the present study,
(n = 100) (41.0%) (59.0%)
however, the AIS had to be translated in Greek. Thus, in Psychiatric inpatients 20 24 32.3 ‹ 14.2 18 ± 70
order to secure equivalence of the Greek translation to the (n = 44) (45.5%) (54.5%)
original English text the guidelines of the World Health Non-patient controls 23 27 37.2 ‹ 10.5 22 ± 63
Organization were followed [23]. According to these guide- (n = 50) (46.0%) (54.0%)
Total sample (n = 299) 127 172 46.0 ‹ 15.7 18 ± 79
lines, a bilingual group of three experts examined and
(42.4%) (57.6%)
approved the conceptual structure of the English text, noting
 C.R. Soldatos et al. / Journal of Psychosomatic Research 48 (2000) 555±560 557

Table 2 Table 4
Internal consistency of the AIS (Cronbach's a) Test ± retest reliability (Pearson's correlation coefficients)
AIS-8 AIS-5 AIS-8 total score 0.89
AIS-5 total score 0.88
All subjects (n = 299) 0.89 0.87
Primary insomniac patients (n = 105) 0.90 0.85
Single items
Psychiatric outpatients (n = 100) 0.86 0.81
Sleep induction 0.86
Psychiatric inpatients (n = 44) 0.85 0.86
Awakenings during the night 0.80
Controls (n = 50) 0.75 0.75
Final awakening 0.80
Total sleep duration 0.82
Sleep quality 0.70
during the period of the last month. To evaluate test ±retest Well-being during the day 0.85
reliability, all psychiatric patients and the controls (N = 194) Functioning capacity during the day 0.79
were assessed through the AIS at two time points a week Sleepiness during the day 0.77
apart from each other. The controls and psychiatric inpati-
p < 0.001 for each one of the above correlations.
ents (N = 94) were also administered the Sleep Problems
Scale [13] as an external validator.
remained practically unchanged when any one of the items
Data analysis was removed from the calculation (a if item deleted ranging
from 0.83 to 0.89). Also, the mean item ±total correlation
The internal consistency of the scale was measured based coefficients were high (0.67 for AIS-8 and 0.69 for AIS-5),
on Cronbach's a [24] for the whole scale, as well as for all- with individual item values ranging from 0.53 to 0.75 ( p <
but-one items if any one was deleted (``a if item deleted''). 0.001 for each one of the correlations). For both versions of
The correlation of each item score to the total scale score the AIS, the factor analysis demonstrated that the entire
(``item ± total correlation'') was also measured through Pear- scale emerged as a sole component (i.e., only one compo-
son's correlation coefficients. Additionally, the scale was nent had an eigenvalue greater than 1.0) with high percen-
subjected to factor analysis; in the extraction method of the tages of variance explained and all items contributing
principal component analysis, the eigenvalue threshold was practically equally to the component either in the case of
set at 1.0. Pearson's correlation coefficients were computed AIS-8 or in that of the AIS-5 (Table 3).
for the evaluation of test ± retest reliability and for the The correlations of either the AIS-8 or the AIS-5 score
external validation of the AIS against the Sleep Problems with the Sleep Problems Scale score were found to be very
Scale [13]; the respective correlations pertained to total scale strong, with Pearson's correlation coefficients of 0.90 and
scores as well as to single-item ratings. All assessments 0.85, respectively ( p < 0.001 in either case). Finally, the
were performed for both the AIS-8 and the AIS-5 and were test ± retest reliabilities for both the AIS-8 and AIS-5 total
executed through the use of the Statistical Package for scores, as well as for every individual item, were also found
Social Sciences (SPSS-8.0). to be very satisfactory (Table 4).

Results Discussion

The internal consistency measures of AIS-8 and AIS-5, The rationale for the development of the AIS, a scale for
based on Cronbach's a, are quite high for the total group the assessment of insomnia according to ICD-10 principles,
and each subgroup of subjects (Table 2). Cronbach's a is based on the necessity to give appropriate importance to
subjective sleep difficulty and to specify a time-frame as
Table 3 well as a minimum frequency for various sleep related
Factor analysis: eigenvalue, variance explained, and factor loadings problems [22]. In this paper, we present the validation of
AIS-8 AIS-5 the AIS, based on a large sample of subjects including many
users of health services with frequent sleep problems. The
Eigenvalue 4.56 3.29
Percentage of variance explained 56.9% 65.8% administration and assessment of AIS showed that this scale
is very reliable, practical, and easy to use. Its utilization
Factor loadings of the individual scale items provides a reliable measure of the intensity of insomnia.
Sleep induction 0.74 0.78 Moreover, the consistency, reliability, and validity of the
Awakenings during the night 0.71 0.78
scale were found to be very satisfactory.
Final awakening 0.81 0.85
Total sleep duration 0.79 0.85
Sleep quality 0.82 0.80 Consistency of the AIS
Well-being during the day 0.78
Functioning capacity during the day 0.75 The internal consistency of the AIS, either in the AIS-8
Sleepiness during the day 0.62
version or the AIS-5 version, was found to be very
558 C.R. Soldatos et al. / Journal of Psychosomatic Research 48 (2000) 555±560
satisfactory (Cronbach's a: 0.89 and 0.87, respectively).
This demonstrates the very high degree of homogeneity of
the AIS, which is further supported by the finding that a
does not practically change if any one of the items is
deleted and by the high correlation coefficients of the
item ±total correlations. As expected, the lowest internal
consistency measures for both AIS-8 and AIS-5 were
found for the subgroup of healthy controls. Even then,
however, the Cronbach's a values were quite high (0.75
for either the AIS-8 or the AIS-5).
The internal consistency of the AIS is very satisfactory
not only by itself, but also in comparison to other standar-
dized scales for the measurement of sleep difficulty in
clinical settings. Indeed, the Cronbach's a for the AIS is
higher than that for the Pittsburgh Sleep Quality Index
[14,25], the Sleep Problems Scale [13], and the Karolinska
Sleep Diary's sleep quality index component [26]. The
mean item ± total correlation coefficient of the AIS was
found to be 0.67 for the AIS-8, and 0.69 for the AIS-5,
while the mean component ± total correlation coefficient of
the Pittsburgh Sleep Quality Index was 0.58 [14]. The mean
item ±total correlation of the Sleep Problems Scale (0.79)
was somewhat higher than that of the AIS, but the calcula-
tion for the Sleep Problems Scale was based on a homo-
geneous sample of air-traffic controllers to whom a shorter
form of the scale, which included only four items closely
related to sleep difficulty, was administered [13].
The factor analysis provided further evidence for the
homogeneity of the AIS. From this analysis, the whole
scale emerged as only one component; the eigenvalue was
4.6 for AIS-8 with nearly 60% of the variance explained
and all item loadings being equally high. Similarly
satisfactory were the eigenvalue, percentage of variance
explained, and the item loadings for the AIS-5. Three
other sleep scales have been subjected to factor analysis:
the Karolinska Sleep Diary, the St. Mary's Hospital Sleep
Questionnaire, and the Leeds Sleep Evaluation Question-
naire. The Karolinska Sleep Diary was found to consist of
two well-defined components that explained the 68% of
the total variance with eigenvalues 2.9 and 1.8 [26]. The
St. Mary's Hospital Sleep Questionnaire was also found
to consist of two components (eigenvalues 4.6 and 1.4),
which actually could not be clearly defined [27]. The
Leeds Sleep Evaluation Questionnaire was found to be
even less compact, having four separate components that
pertain to sleep latency, quality of sleep, awakening from
sleep, and behavior following wakefulness [28]. Thus, the
AIS emerges as more compact than all previously pub-
lished scales. However, the high degree of homogeneity
of the AIS shown in the present study may be due to the
fact that the patients presented either with primary in-
somnia or with psychogenic sleep difficulties. It is quite
possible that a lower degree of homogeneity of the AIS
could have been found among patients suffering from a
variety of other sleep disorders, such as narcolepsy, sleep
apnea, etc. In this context, it would be useful to test the
performance of the AIS in patient samples different than
those utilized in the present study.
Reliability of the AIS
The test ±retest reliability of the AIS was very satisfac-
tory both for the whole scale in its two versions (0.89 for the
AIS-8 and 0.88 for the AIS-5) and for each individual item
(ranging from 0.70 to 0.86). The test ± retest correlation
coefficient of the total score of the Pittsburgh Sleep Quality
Index (0.85) and the range of single-item coefficients of the
St. Mary's Hospital Sleep Questionnaire (from 0.70 to 0.96)
were similarly high [12,14]. The total score test ± retest
coefficients for the Sleep Problems Scale were much lower
(0.59 at the 3-month retest and 0.42 at the 6-month retest),
but the reexamination was performed at a much later time
compared to the other scales [13].
Validity of the AIS
The external validity of the AIS was measured against
the total score of the Sleep Problems Scale [13]. The
correlation with that scale was very high, with coefficients
being 0.90 and 0.85 for the AIS-8 and AIS-5, respectively.
The Sleep Problems Scale is similar to the AIS in that it is
comparatively simple and contains about the same items,
although for its rating, it is based on frequency of occur-
rence, rather than severity, of the sleep problems [13]. To
our knowledge, the AIS is the only scale that has been
correlated with another sleep scale. The Sleep Problems
Scale was correlated with psychopathology scales not spe-
cifically pertaining to sleep [13]. Polysomnographic mea-
sures have been used for the external validation of the
Karolinska Sleep Diary and the Pittsburgh Sleep Quality
Index; the correlations of the Karolinska Sleep Diary scores
with the polysomnographic measures were satisfactory [26],
while those of the Pittsburgh Sleep Quality Index were not
[14]. Another important measure of external validity of a
given scale is its sensitivity to the change of severity of the
condition it assesses. For example, the effects of a hypnotic
therapy should be detected through a significant change in
the scale score. Such an external validity measure is
obtained for AIS, based on a study in which a hypnotic
drug of known potency is used (publication in preparation).
Applicability of the AIS
The AIS can be utilized in clinical practice to measure
reliably the intensity of perceived sleep-related problems by
a given patient. More specifically, the first five items
provide a measure of difficulty with sleep quantity and
quality, while a measure of the daytime consequences of
insomnia is provided by the last three items. Besides its use
in clinical practice, the AIS can be also utilized as a research
tool. The usual time-frame of assessment according to ICD-
10, is 1 month. However, when the scale is used for specific
 C.R. Soldatos et al. / Journal of Psychosomatic Research 48 (2000) 555±560 559

research purposes, this period can be modified according to
the needs of the specific research project.
The AIS is devised to measure purely subjective esti-
mates of sleep difficulty. Should the need arise for the AIS
to be rated by an interviewer (e.g., when dealing with
illiterate or sight-impaired subjects), the rater must always
bear in mind that only the subjective feelings of the
interviewee are considered for the rating. Thus, the severity
of the interviewee's sleep difficulty as estimated by the
interviewer (based on his/her own evaluation of any addi-
tional information, e.g. number of hours of sleep or approx-
imate duration of sleep onset latency, etc.) should not be
taken into account since it may actually be at variance with
the interviewee's subjective assessment.
The Leeds Sleep Evaluation Questionnaire, the St.
Mary's Hospital Sleep Questionnaire, and the Karolinska
Sleep Diary [9± 12,15,16] are relatively simple in their use,
but they are designed to evaluate mainly the sleep of the
previous night and their purpose is not to assess sleep
difficulties over periods of time comparable to those re-
quired in most diagnostic classification systems in use [18 ±
21]. Similarly to the AIS, the Sleep Problems Scale and the
Pittsburgh Sleep Quality Index have wide applications,
including their use in a variety of clinical and research
settings [13,14]. Compared to the Pittsburgh Sleep Quality
Index, the AIS is much simpler and easier to use. The
Pittsburgh Sleep Quality Index has 18 questions to be
answered by the subject, plus an additional six by a room-
mate [14]. To score the Pittsburgh Sleep Quality Index,
many relatively complex calculations are needed, which
make it rather cumbersome for clinicians and could sub-
stantially increase the time and cost of any investigation in
which it is utilized.
In conclusion, the AIS is a scale for use in a large variety
of clinical and research settings where the quantification of
sleep problems is required. The AIS compares favorably
against other scales regarding all measures of consistency,
reliability, and validity. These properties, of the AIS together
with its simplicity, make it an invaluable psychometric
instrument in sleep research and clinical practice.
Acknowledgments
The authors wish to express their appreciation to the
members of the bilingual group of experts (Drs. V. Mavreas,
G. Papadimitriou and P. Sakkas) and those of the unilingual
group (Drs. D. Pappa, A. Politis and C. Psarros) for their
assistance in the establishment of the equivalence of the
Greek translation of the AIS to its original English text.

Appendix A.
Athens Insomnia Scale
ID:__________ Age:_________ Sex:_________ Date:________

Instructions: This scale is intended to record your own assessment of any sleep difficulty you might have experienced. Please, check (by circling the
appropriate number) the items below to indicate your estimate of any difficulty, provided that it occurred at least three times per week during the
last montha

Sleep induction (time it takes you to fall asleep after turning-off the lights)
0: No problem 1: Slightly delayed 2: Markedly delayed 3: Very delayed or did not sleep at all

Awakenings during the night

0: No problem 1: Minor problem 2: Considerable problem 3: Serious problem or did not sleep at all

Final awakening earlier than desired

0: Not earlier 1: A little earlier 2: Markedly earlier 3: Much earlier or did not sleep at all

Total sleep duration

0: Sufficient 1: Slightly insufficient 2: Markedly insufficient 3: Very insufficient or did not sleep at all

Overall quality of sleep (no matter how long you slept)

0: Satisfactory 1: Slightly unsatisfactory 2: Markedly unsatisfactory 3: Very unsatisfactory or did not sleep at all

Sense of well-being during the day

0: Normal 1: Slightly decreased 2: Markedly decreased 3: Very decreased

Functioning (physical and mental) during the day

0: Normal 1: Slightly decreased 2: Markedly decreased 3: Very decreased

Sleepiness during the day

0: None 1: Mild 2: Considerable 3: Intense
a
The period of the self-assessment may vary, depending on the design of a given study. Whenever the self-assessment pertains to a period other than that of
the last month, the second sentence of the instructions should be rephrased accordingly.
560 C.R. Soldatos et al. / Journal of Psychosomatic Research 48 (2000) 555±560
References
[1] Karacan I, Thornby JI, Anch M, Holzer CE, Warheit GJ, Schwab JJ,
Williams RL. Prevalence of sleep disturbance in a primarily urban
Florida county. Soc Sci Med 1976;10:239 ± 44.
[2] Bixler EO, Kales A, Soldatos CR, Kales JD, Healey S. Prevalence of
sleep disorders in the Los Angeles metropolitan area. Am J Psychiatry
1979;136:1257 ± 62.
[3] Partinen M, Kaprio J, Koskenvuo M, Langinvainio H. Sleeping ha-
bits, sleep quality, and use of sleeping pills: a population study of
31,140 adults in Finland. In: Guilleminault C, Lugaresi E, editors.
Sleep/Wake Disorders: Natural History, Epidemiology, and Long-
Term Evolution. New York: Raven Press, 1983. pp. 29 ± 35.
[4] Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances
and psychiatric disorders: an opportunity for prevention? JAMA,
1989;262:1479 ± 84.
[5] Soldatos CR. Insomnia in relation to depression and anxiety: epide-
miologic considerations. J Psychosom Res S1 1994;38:3 ± 8.
[6] Sweetwood H, Grant I, Kripke DF, Gerst MS, Yager J. Sleep disorder
over time: psychiatric correlates among males. Br J Psychiatry
1980;136:456 ± 62.
[7] Liljenberg B, Almwvist M, Hetta J, Roos B-E, Agren H. Affective
disturbance and insomnia: a population study. Eur J Psychiatry
1989;3:91 ± 8.
[8] Quera-Salva MA, Orluc A, Goldenberg F, Guilleminault C. Insom-
nia and use of hypnotics: study of a French population. Sleep 1991;
14:386 ± 91.
[9] Hindmarch I. A 1.4 benzodiazepine, temazepam (K 3917), its effect
on some psychological parameters of sleep and behaviour. Arzneim-
Forsch 1975;25:1836 ± 39.
[10] Parrot AC, Hindmarch I. The Leeds Sleep Evaluation Questionnaire in
psychopharmacological investigations Ð a review. Psychopharmaco-
logy 1980;71:173 ± 79.
[11] Murphy F, Bentley S, Ellis BW, Dudley HAF. Sleep deprivation in
patients undergoing operation: a factor in the stress of surgery. Br Med
J 1977;2:1521 ± 2.
[12] Ellis BW, Johns MW, Lancaster R, Raptopoulos P, Angelopoulos N,
Priest RG. The St. Mary's Hospital Sleep Questionnaire: a study of
reliability. Sleep 1981;4:93 ± 7.
[13] Jenkins CD, Stanton B-A, Niemcryk SJ, Rose RM. A scale for the
estimation of sleep problems in clinical research. J Clin Epidemiol
1988;41:313 ± 21.
[14] Buysse DJ, Reynolds CF III, Monk TH, Berman SR, Kupfer DJ. The
Pittsburgh Sleep Quality Index: a new instrument for psychiatric prac-
tice and research. Psychiatry Res 1989;28:193 ± 213.
[15] Akerstedt T, Hume K, Minors D, Waterhouse J. The meaning of good
sleep: a longitudinal study of polysomnography and subjective sleep
quality. J Sleep Res 1994;3:152 ± 8.
[16] Akerstedt T, Hume K, Minors D, Waterhouse J. The subjective mean-
ing of good sleep, an intraindividual approach using the Karolinska
Sleep Diary. Percep Mot Skills 1994;79:287 ± 96.
[17] Association of Sleep Disorders Centers. Diagnostic classification of
sleep and arousal disorders. Sleep 1979;2:1 ± 137.
[18] American Psychiatric Association. Diagnostic and Statistical Manual
of Mental Disorders, 3rd edn. Washington, DC: APA Press, 1987
Revised (DSM-III-R).
[19] American Sleep Disorders Association. International Classification of
Sleep Disorders. Rochester, MN: ASDA, 1990.
[20] World Health Organization. The ICD-10 Classification of Mental
and Behavioural Disorders: Clinical Descriptions and Diagnostic
Guidelines. Geneva: WHO, 1992.
[21] American Psychiatric Association. Diagnostic and Statistical Manual
of Mental Disorders, 4th edn. Washington, DC: APA Press, 1994. pp.
317 ± 91 (DSM-IV).
[22] Soldatos CR. The assessment of insomnia: rationale for a new scale
based on ICD-10 principles. In: Szelenberger W, Kukwa A, editors.
Sleep: Physiology and Pathology. Warszawa: Elma Books, 1995.
[23] Sartorius N, Janca A. Psychiatric assessment instruments developed
by the World Health Organization. In: Thornicroft G, Tansella M
editors. Mental Health Outcome Measures. Berlin: Spinger, 1996.
pp. 153 ± 77.
[24] Cronbach LJ. Coefficient alpha and the internal structure of tests.
Psykometrika 1951;16:297 ± 333.
[25] Carpenter JS, Andrykowski MA. Psychometric evaluation of the Pitts-
burgh Sleep Quality Index. J Psychosom Res 1998;45:5 ± 13.
[26] Keklund G, Akerstedt T. Objective components of individual differ-
ences in subjective sleep quality. J Sleep Res 1997;6:217 ± 20.
[27] Leigh TJ, Bird HA, Hindmarch I, Constable PDL, Wright V. Factor
analysis of the St. Mary's Hospital Sleep Questionnaire. Sleep
1988;11:448 ± 53.
[28] Parrott AC, Hindmarch I. Factor analysis of a sleep evaluation ques-
tionnaire. Psychol Med 1978;8:325 ± 29.