- Email: [email protected]
How do individuals sleep around the world? Results from a single-day survey in ten countries
Sleep Medicine 6 (2005) 5–13 www.elsevier.com/locate/sleep
How do individuals sleep around the world? Results from a single-day survey in ten countries Constantin R. Soldatosa,*, Franc¸ois A. Allaertb, Tatsuro Ohtac, Dimitris G. Dikeosa a
Sleep Research Unit, Department of Psychiatry, University of Athens, Vas Sofias 72, Athens 11528, Athens, Greece b Department of Epidemiology and Public Health, McGill University, Montreal, Canada c Graduate School of Medicine, Nagoya University, Nagoya, Japan Received 28 May 2004; received in revised form 11 October 2004; accepted 14 October 2004
Abstract Background and purpose: To describe between-country differences in both the prevalence and type of sleep disorders seen across the globe, and to provide information on how impaired sleep impacts daytime functioning. Patients and methods: The study is a large-scale, global cross-sectional survey conducted on International Sleep Well Day (March 21), 2002. A standardized questionnaire was used in 10 countries under the guidance of local survey managers. In addition, the Athens Insomnia Scale (AIS) and the Epworth Sleepiness Scale (ESS) were completed. Subjects included in the study were adults from 10 countries representing different continents with clear variations in lifestyle. Results: The total number of questionnaires collected was 35,327. Overall, 24% of subjects reported that they did not sleep well. According to self-assessments using the AIS, 31.6% of subjects had ‘insomnia’, while another 17.5% could be considered as having ‘subthreshold insomnia’. According to ESS scores, 11.6% of subjects were found to be ‘very sleepy’ or ‘dangerously sleepy’ during the day. Conclusions: Although there seem to be important global variations in the prevalence of insomnia, its symptoms and their management, about one in four individuals do not think they sleep well. Moreover, self-reported sleep problems could be underestimated in the general population. Overall, there is a need for increased awareness of the importance of disturbed sleep and the improved detection and management of sleep disorders. q 2004 Elsevier B.V. All rights reserved. Keywords: Sleep disorders; Insomnia; Survey; Sleep; Prevalence; Daytime sleepiness
1. Introduction Sleep troubles are a common complaint that can have a huge impact on patients’ quality of life and ability to function. However, a recent review of the literature covering more than 50 studies of insomnia, based on data collected in various representative community-dwelling populations, showed that estimates of prevalence vary widely, from less than 5 to 40% [1]. Prevalence estimates can be affected by a number of factors, such as the characteristics of the population sampled, the definition of insomnia, and regional perceptions and management practices regarding sleep * Corresponding author. Tel.: C30 210 728 93 24; fax: C30 210 725 13 12. E-mail address: [email protected] (C.R. Soldatos). 1389-9457/$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2004.10.006
disorders. Few epidemiological surveys have been conducted that have used a single method to evaluate impaired sleep prevalence simultaneously in different countries. As a result, it is not clear precisely how much the prevalence of insomnia varies between global regions. It is not known whether the variability seen in studies performed to date reflects real regional differences or simply different definitions and/or methodologies. Janson et al. [2] studied the geographic variation in sleep complaints to identify risk factors for sleep disturbances in 2202 subjects from three Northern European countries (Sweden, Iceland and Belgium) using a standardized questionnaire and sleep diaries. However, this study selected only middle-aged adults (aged 20–45 years) and did not refer to a validated insomnia scale, making regional comparisons difficult. Chevalier et al. [3] conducted
6
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
a survey of severe insomnia and its effect on quality of life and healthcare consumption in five Northern European countries using a 4-item questionnaire, reporting prevalence rates of severe insomnia ranging from 4 to 22%. Finally, a telephone survey conducted by Ohayon and Roth [4] in 24,600 general population-based subjects aged R15 years from six European countries indicated a global insomnia prevalence of 16.8%, according to the criteria of the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). However, only overall survey results were presented and no between-country comparisons were available. The objectives of the present investigation were to describe between-country differences in both the prevalence and type of sleep disorders seen across the globe, and to gain information on how impaired sleep generally impacts daytime functioning.
2. Methods
represented Northern Europe, Portugal and Spain represented Southern Europe, and Slovakia represented the Central European population. China and Japan were studied in Asia, Brazil in South America, and South Africa in the African continent. Brazil and South Africa were taken to represent the Southern hemisphere. The total number of questionnaires collected was 35,327, ranging from 202 in South Africa to 10,424 in Japan (Table 1). The global mean age of surveyed individuals was 39.0 years (range 15–99, median 36 years), with 7.9% aged over 65 years. Approximately half of them (49.8%) were men and 64.1% were in the workforce (employed or currently unemployed but looking for work). However, this sociodemographic profile differed between countries. There were more men in the Spanish sample (68.7%) and more women in the Belgian, Slovakian and South African samples (57.8, 57.9 and 58.2%, respectively). Other notable regional differences included a higher mean age in the German sample (49.8 years), a lower than average proportion of individuals in the workforce in Portugal (33.1%), and higher proportions in Japan (74.8%) and South Africa (77.2%).
2.1. Study design and data collection 2.3. Assessments A cross-sectional survey was conducted on International Sleep Well Day (March 21), 2002. A common standardized questionnaire was used in all countries under the guidance of local survey managers, who ensured the feasibility of data collection according to local practices. Methods used to collect data in the different regions included: China, Japan, Slovakia, South Africa—collection from a general population in public booths in various locations (e.g. streets, railway stations, shopping centers); Austria, Brazil, Germany, Portugal, Spain—collection by telephone screening; Japan—by internet poll and at the work place; and Belgium—by internet poll and at the work place, as well as (4.2% of the Belgian questionnaires) at waiting rooms of primary care physicians and/or sleep laboratories. In all settings, the participants gave informed consent after the procedure was fully explained to them. 2.2. Study participants Subjects included in the study were adults from 10 countries representing four different continents with clear variations in lifestyle. Austria, Belgium and Germany
In all 10 countries, the survey was conducted with the same standardized case report form translated into local languages. The form included the following categories: a socio-demographic profile evaluation, a sleep outcomes evaluation that assessed sleep habits, subjective perception of sleep quality under the form of a general question: “Overall, do you consider that you sleep well?”, and, if applicable, a questionnaire-based assessment of sleeppromoting actions taken by subjects. All subjects also completed the Athens Insomnia Scale (AIS) [5] and the Epworth Sleepiness Scale (ESS) [6]. These scales were chosen because they appear to be the simplest scales that could be quickly completed, given the way the data were collected. However, due to time constraints, no cultural validation of the two scales could be implemented in all 10 countries. The AIS is intended to record the individual’s own assessment of any sleep disorder he/she might have experienced, provided that it occurred at least three times per week during the last month. It contains eight questions rated from 0 to 3. If the total score of the AIS is 6 or higher,
Table 1 Demographic data
Number % Male Mean age (GSD) Percentage in the workforce
Total/ average
Austria
Belgium
Brazil
China
Germany
Japan
Portugal
Slovakia
South Africa
Spain
35,327 49.8 39 G15.3 64.1
490 52.0 44.3 G18.8 49.8
6832 42.2 40.0 G16.3 62.6
1999 48.7 37.4 G15.0 61.6
10,079 51.6 37.1 G14.9 64.3
2016 46.5 49.8 G18.2 48.6
10,424 50.9 36.1 G10.7 74.8
784 48.2 44.5 G22.8 33.1
502 42.1 38.4 G16.0 57.6
202 41.8 37.7 G15.6 77.2
1999 68.7 47.7 G18.4 46.8
22.9 30 25.0 60 12.0 60 25.5 45
13.4 60
22.0 30
8h00 22.9G31.5 8h00 34.4G42.1 7h00 22.2G21.4
32.6 60 42.4 40 19.4 30 23.1 60
21.5 60
8h00 15.2G18.4 7h30 24.8G26.5
8h00 26.5G29.4
8h00 26.1G30.0
8h00 28G27.4
7h30 22.1G25.7
8h20 22.7G27.2
8h00 17.9G21.1
478G79 467G84 413G69 480G71 454G83
478G80
464G92
484G79
452G71
504G89
471G82
7:00 AM (33.0%) 6:00 AM (38.5%) 6:00 AM (44.8%) 6:00 AM (40.7%) 6:00 AM (43.4%) 6:00 AM (35.2%)
7:00 AM (35.0%)
6:00 AM (37.8%)
7:00 AM (36.6%)
6:00 AM (42.1%)
7:00 AM (43.6%)
10:00 PM (39.3%)
12:00 PM (40.6%)
South Africa
10:00 PM (32.5%) 11:00 PM (33.3%)
12:00 PM (27.2%)
Japan
10:00 PM (41.5%) 11:00 PM (30.5%) 10:00 PM (29.3%)
Brazil Belgium
11:00 PM (35.8%) 10:00 PM (44.9%)
Austria
11:00 PM (32.0%)
Average Table 2 Sleep habits during the week
7
the patient is diagnosed as insomniac [7]. For the purposes of the present study, we also assessed separately those subjects with an AIS score of 4 or 5, who were considered as presenting with ‘sub-threshold insomnia’. The ESS evaluates the likelihood to doze off or to fall asleep in eight different situations. The answers are rated from 0 to 3 and a total score is calculated. The individual is considered sleepy if the score is greater than 6 and below 10, very sleepy if it is greater than 10 and below 16 and dangerously sleepy if it is 16 or over [6]. 2.4. Data management and statistical analyses Data were collected in each country using the same format and software. All data were sent to a single university statistical center, where they were analyzed according to a pre-defined statistical plan using SAS 8.2 software. Between-country differences were assessed using the c2 test and analysis of variance (ANOVA), followed by the post-hoc Dunnett’s test. A P-value of less than 0.05 was considered to be statistically significant.
3. Results 3.1. Sleep habits
Time of going to bed (mode, and percentage of subjects reporting this value in brackets) Time of waking up (mode, and percentage of subjects reporting this value in brackets) Sleep duration (meanGSD, in min) Sleep duration (median, in h) Time to fall asleep (meanGSD, in min) Regular nap (% of subjects) Nap duration (median, in min)
China
Germany
Portugal
Slovakia
Spain
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
The average (mode) bedtime on weekdays was 11 PM, with 32.0% of individuals reporting that went to bed at this time (Table 2). The Spanish and the Portuguese reported later bedtimes (12 PM). On weekends, the average bedtime was slightly later, at 12 PM. The average (mode) waking time during the week was 6 AM, with 35.2% of participants reporting that they awoke at this time, and 31.6% at 7 AM. On weekends, the average wake-up time was 8 AM. The mean sleep duration during the week was 454 min (median 450 min [7.5 h]). The shortest sleep duration was reported in Japan (413 min) and the longest in Portugal (504 min). An average of 31.4% (range 25.9–36.4%) of responders reported that they slept for 7–8 h per night, and 25.7% (15.6–36.5%) for 6 to 7 h. The global mean sleep latency (the time taken to fall asleep) during the week was 24.8 min (range 15.2– 34.4 min), with a median value of 15 min. Approximately 40% of subjects reported that they needed between 0 and 15 min to fall asleep, and 20% reported a latency period of 15–30 min. At the other end of this spectrum, 39.4% of participants (range 22.1–50.2%) required more than 30 min to fall asleep, and 13.1% needed at least an hour. Approximately one quarter of the overall study population (23.1%) reported that they regularly nap on weekdays (Table 2). The highest napping rate (42.4%) was reported in Brazil and the lowest in Japan (12.0%) and Portugal (13.4%). In those who reported a regular nap, the median duration was 1 h.
8
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
Fig. 1. Percentage of subjects reporting that that ‘they do not sleep well’.
3.2. Sleep disturbances 3.2.1. Global self-assessment To the global question: “Overall, do you consider that you sleep well?” 24% answered ‘No’ (Fig. 1). Percentage of subjects answering ‘No’ to this question ranged from
10.4% in Austria to 32.2% in Belgium; overall the difference between countries was highly statistically significant with a c2 value of 50,264 (9 df) and a P! 0.0001. For subjects who answered ‘No’, the mean duration of sleep disturbances was 5.7 years (median 3 years).
Fig. 2. Percentage of subjects by AIS score category.
O0.0001 O0.0001 1280,62 323,50 4.3 7.3 10.1 20.2 4.2 9.0 2.7 7.9 5.5 12.2 2.5 5.1 25.2 16.3 8.2 9.4 8.0 10.5
12.6 12.9
6.5 8.0
O0.0001 O0.0001 O0.0001 785,40 750,49 1594,46 10.4 7.5 6.9 24.6 19.7 15.2 10.6 6.8 10.0 9.3 5.6 2.9 11.5 10.9 4.9 7.2 4.9 4.4 29.5 26.5 26.7 6.3 5.3 8.0 13.4 12.4 8.1
18.1 16.9 13.9
11.4 11.6 5.0
O0.0001 O0.0001 O0.0001 705,30 1488,29 767,29 12.4 15.4 10.5 25.5 33.2 20.0 11.1 15.3 12.3 9.1 9.7 9.4 8.6 3.8 5.2 8.9 10.9 9.3 11.7 12.8 9.3 27.8 16.8 17.3 7.1 7.6 6.3
Sleep induction (%) Night awakenings (%) Final awakening earlier than desired (%) Total sleep duration (%) Overall quality of sleep (%) Sense of well-being during the day (%) Functioning during the day (%) Sleepiness during the day (%)
12.1 12.2 10.1
14.8 22.2 16.8
China Brazil Belgium Austria Total/ Average
Table 3 Percentage of subjects with AIS symptoms rated 2 or 3 (moderately and severely disturbed)
Germany
Japan
Portugal
Slovakia
South Africa
Spain
Overall c2
P (dfZ9)
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
9
3.2.2. Self-assessment based on the AIS As shown in Fig. 2, 17.5% of the participants exhibited an AIS score of 4 or 5 (‘sub-threshold insomnia’), and 31.6% had a score of 6 or higher (‘insomnia’). Large global variations were observed. The prevalence rates of ‘sub-threshold insomnia’ (i.e. AIS score 4 or 5) ranged from 13.9% (Belgium and Brazil) to 25.7% (Slovakia). Rates of ‘insomnia’ (AIS score 6 or greater) ranged from 17.4% (Germany) to 79.8% (Brazil); the overall c2 of the difference between countries was 2676.10 (9 df, P!0.0001) and of the difference between Brazil and all other countries taken together was 2276.69 (dfZ2, P!0.0001). Table 3 describes percentages of individuals who quoted grade 2 or 3 (‘markedly’ or ‘very affected’) for each of the AIS items. On average, 12.1% (nZ4161; range 7.1–27.8%) of subjects considered their sleep induction to be delayed; 12.2% (3.8–33.2%) reported frequent night awakenings; 10.1% (5.2–20.0%) had final awakening earlier than desired; 13.4% (6.3–29.5%) found their sleep duration was insufficient; and 12.4% (4.9–26.5%) said they were not satisfied with the quality of their sleep. Finally, 8.1% of the participants (4.4–26.7%) reported a decreased sense of wellbeing during the day; 8.0% (2.5–25.2%) said their daytime functioning was decreased; and 10.5% (5.1–20.2%) said they felt sleepy during the day. For all but two of the individual items on the AIS, subjects from Brazil recorded the highest ratings. Problems with sleep latency was the only complaint in 12.1% (nZ1,252) of subjects rating 2 or 3 for any insomnia symptom (nZ10,344). Similarly, as isolated complaints, any problem with sleep maintenance (nightly awakenings and/or early awakening) was observed in 20.8%, insufficient sleep duration in 12.8%, and poor sleep quality in 6.5%. Of the responders who rated even one item on the AIS as 1 or greater, 52.1% had isolated symptoms while 47.9% had more than one symptom. 3.2.3. Global vs. AIS self-assessment When the results of the global question on sleep were compared with the AIS classification of subjects, 91.6% of people who thought they did not sleep well had AIS scores of at least 4, indicating at least ‘sub-threshold insomnia’, and 77.8% had scores of at least 6, indicating a diagnosis of insomnia. Among those who thought they did sleep well, 82.8% had an AIS score of less than 6 and 65.4% a score of less than 4. These results indicate that some insomniacs may underestimate their sleep problems. 3.3. Effect of aging The frequency of a grade 2 or 3 rating on the various AIS items in subjects aged 18–65 years was compared with that of subjects older than 65 years (Table 4). The frequency of four items rated as grade 2 or 3 was significantly higher in elderly than in younger subjects: delayed sleep induction
10
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
Table 4 Percentage of subjects with AIS symptoms quoted 2 or 3 (moderately or severely impaired) as a function of age Age (years)
Sleep induction
Awakenings during the night
Final awakening earlier than desired
Total sleep duration
Overall quality of sleep
Sense of well-being during the day
Functioning during the day
Sleepiness during the day
!65 O65 P-value
11.6 18.3 !0.0001
11.5 21.9 !0.0001
9.7 15.1 !0.0001
13.6 11.6 !0.01
12.6 12.2 n.s.
8.1 9.0 n.s.
7.9 10.0 !0.0001
10.7 8.6 !0.001
n.s., not statistically significant.
(18.3 vs. 11.6%); awakenings during the night (21.9 vs. 11.5%); final awakening earlier than desired (15.1 vs. 9.7%); and functioning during the day (10.0 vs. 7.9%). In contrast, older subjects complained less frequently of insufficient total sleep duration (11.6 vs. 13.6%) and sleepiness during the day (8.6 vs. 10.7%).
the highest rates in Portugal (45.7%) and in South Africa (52.8%). Herbal teas were used to improve sleep by 22.1% of subjects (4.2% in Japan to 48.0% in Slovakia), 19.4% took alcohol (9.8% in Austria to 30.3% in Japan), and 40.0% modified their consumption of tea or coffee (10.0% in Japan to 49.0% in Austria).
3.4. Prevalence based on DSM-IV definition 4. Discussion In an attempt to compare the results with data in the literature that usually refer to the DSM-IV definition of primary insomnia, percentages of subjects presenting with both night-time and daytime symptoms on the AIS are presented in Table 5. On average, 12.1% of subjects met this criterion. The highest percentage was reported in Brazil (31.8%), and the lowest in Germany (5.2%). 3.5. Daytime sleepiness The results of the ESS scores indicated that 11.6% of subjects could be considered very sleepy or dangerously sleepy (ESS score O10). Rates ranged from 6.2% in China to 24.5% in South Africa. The c2 value for the differences between the ten countries for the comparison between subjects who were ‘not sleepy’ and all others was 1138.38, yielding (for 9 df) a P-value of less than 0.0001. Fig. 3 shows percentages of subjects reporting the four categories of sleepiness in the 10 countries. 3.6. Actions taken for the management of sleep problems Among those who thought they did not sleep well, 30.7% (8.0% in Japan to 55.5% in Portugal) reported that they had visited a physician regarding their sleep impairment. In addition, 31.4% of them said they had taken sleep medications. The lowest rates of medication were reported in Austria (9.8%) and Japan (15.3%), and
To our knowledge, this is the first international study attempting to assess sleep behaviors on the same day and with the same standardized questionnaire and scales, and covering 10 countries in four continents. Only three multicountry European studies of the general population have been published [2–4], while the World Health Organization Collaborative Study that involved 14 countries in three continents assessed the burden of sleep problems in primary care patients rather than the general population [8]. Thus, the present survey is probably the largest ever conducted on a general population sample and its results should provide useful insights into world-wide differences in sleep behaviors. There are a number of weaknesses related to the methods of data collection in the different countries that may weaken the extrapolation of findings in the regional samples listed here to their respective national populations, and which limit the relevance of inter-country comparisons. Variability in sample sizes and the ways in which data were collected was considerable across countries. The method which was used for the recruitment of the participants is vulnerable to several biases, such as that individuals with sleep disturbances are more likely to accept to answer the questionnaire, and that people with physical disabilities or limitations are less likely to be included (for example, when the questionnaires were administered in the streets, railway stations or shopping centers). In addition, significant
Table 5 Number and percentage of subjects with at least one night-time symptom (rated 2 or 3) and one daytime symptom (rated 2 or 3) in the AIS Number (%)
Total
Austria
Belgium
Brazil
Yes No
4,087 (12.1) 48 (9.8) 1,023 (16.0) 633 (31.8) 952 (10.0) 102 (5.2) 1,021 (10.3) 49 (6.2) 55 (11.1) 39 (20.5) 165 (8.2) 29,661 (87.9) 442 (90.2) 5,368 (84.0) 1,359 (68.2) 8,592 (90.0) 1,876 (94.8) 8,865 (89.7) 735 (93.8) 439 (88.9) 151 (79.5) 1834 (91.8)
Missing dataZ1579; Overall c2Z1044.08, dfZ9, P!0.0001.
China
Germany
Japan
Portugal
Slovakia
South Africa
Spain
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
11
Fig. 3. Percentage of subjects by ESS score category.
differences between the countries existed with regard to gender ratios, mean ages, and proportions of people in the workforce. It is difficult to distinguish which of these differences were due to local social structures and which were due to dissimilar recruitment conditions. Moreover, the ‘collection day’ was March 21, which is at the end of winter in north hemisphere and at the end of summer in the south hemisphere. This seasonal difference might play a role in the heterogeneity of the results. However, since only about 7% of the total study population lived in the southern hemisphere, it should not have substantially influenced the overall results. Also, gender differences do not seem to have played any major role in the explanation of the variability among countries, since country differences in male:female ratio were not deviating considerably (with the exception of Spain) from the expected 49:51; besides, they were much smaller than differences in the rates of any insomnia-related characteristic among countries. Nonetheless, the main outcome of this study appears to be worthy of interest, if one takes into account that the data constitute an ‘instant picture’ of sleep behaviors around the world. 4.1. Sleep habits Overall, sleep behaviors were similar around the world. Although bedtimes and waking times differed between countries, possibly due to climatic conditions and cultural traditions, total sleep durations were similar and seemed to correspond to the same physiological need across countries.
More surprising was the frequency of napping, which was common in all countries, including the most industrial ones, in contrast to the widespread perception that napping is a ‘warm country’-specific habit. 4.2. Sleep disturbances The global rate of people thinking they do not sleep well is high (about one in four subjects) and appears comparable from one country to another. Globally, this fits into the range of insomnia prevalence rates reported in the literature (range 9–36%), considering either all 10 countries assessed here or only the six European ones [1,4,8,9]. Specific country comparisons with previously reported population surveys conducted in adults of all ages are possible only for seven countries. Rates of insomnia reported in the current study are similar to those reported previously in Portugal [10], Spain [10,11], Brazil [12,13] and Japan [14,15]. The rate reported in Austria (10%) was lower than that reported previously in the literature (25–32%) [16,17]. In contrast, the rate reported here for Germany (16%) is higher than that reported in a recent review (8.5%) [1]. However, this recent review evaluated patients with ‘insomnia symptoms plus daytime consequences’ [1], which probably represents a more severe form of insomnia than the general sleep impairments seen in our survey. Accordingly, Hajak et al. [18] reported a prevalence rate of 4% of ‘severe insomnia’ in the German population. In German primary care facilities, insomnia has been
12
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
reported in 19.3–29% of patients [19–21]. In Belgium, Hajak et al. [18] reported a 9% prevalence of ‘severe insomnia’, well below the rate of people who reported that they ‘do not sleep well’ in the present study (32%). The different rates should be attributed to the difference between the two studies in the level of severity of insomnia required for a subject to be considered a case, since the number of people in our study who were recruited from medical settings and sleep laboratory waiting rooms was 4.2% of the total number of subjects ascertained in Belgium (total nZ6832) and it might not have interfered significantly with the results. No prior data on sleep disturbances in Slovakia, China or South-Africa were available, and the data in the current study pertaining to these countries should be considered original. If one considers the association of at least one marked night-time AIS symptom (moderate or severe) with at least one daytime AIS symptom of similar intensity as a close image of the DSM-IV definition diagnosis of insomnia, then the percentages we obtained in Germany (5.2%), in Portugal (6.2%) and in Spain (8.3%) are similar to those reported in the literature (4–6%, 8.7 and 6.0%, respectively) [1,10,18]. It must be noted, however, that insomnia according to the DSM-IV can be diagnosed even without the presence of a compromised feeling of well-being or functioning and/or somnolence during the day. For the diagnosis to be established, a sleep problem and preoccupation with it is sufficient. In that sense, many more subjects can be considered to be insomniacs than those who report the daytime consequences assessed in the AIS, and those who report insomnia symptoms plus daytime consequences may represent cases suffering from a more severe form of insomnia. Comparisons between the rates of people who thought they did not sleep well and the prevalence of insomnia as diagnosed with the AIS (total score 6 or higher) provided interesting data. In most European countries (5 out of 6), Japan and China, both rates were very similar, indicating agreement between participants’ subjective perceptions of sleep and the semi-subjective assessment provided by the AIS. However, in Brazil, Austria and South Africa, rates of those who claimed they did not sleep well were much lower than prevalences of insomnia indicated by the AIS. The most striking difference was seen in Brazil, where only 19% of subjects reported that they did not sleep well, whereas the prevalence of AIS ‘insomnia’ was close to 79%. Some adaptation of the AIS might be required for certain populations. For example, the cut-off level of the AIS score might need to be higher than 6 for some countries where heat may have a heavy influence on the quality of the sleep. Nonetheless, the 32% global prevalence of AIS-defined insomnia in the current study is consistent with that reported in the literature [1,4,8,9]. Specific country rates were similar to published data for Portugal and Spain [10,11] slightly lower than those published for Austria [16,17], and higher than those published for Brazil and Japan [12–15]. If one
considers together AIS ‘sub-threshold insomnia’ and ‘insomnia’ (scores R4), the 49% rate is higher than all that previously reported. Thus, the present study supports previous suggestions that, at least for the majority of the world population, an AIS score of 6 or higher should be considered as the cut-off defining insomnia [7]. As well as AIS scores, types of sleep disturbances reported differed between countries. Austrian, Portuguese and Chinese subjects reported more ‘delayed sleep induction’, whereas subjects in Belgium, Germany, Slovakia, Spain and South Africa reported more ‘frequent night awakenings’. Brazilian and Japanese subjects most frequently reported ‘insufficient sleep duration’. Finally, our results regarding the influence of aging on sleep problems are consistent with previous findings that certain sleep components are altered in the elderly [22]. The subjects in our sample who were older than 65 years had significantly higher AIS ratings of sleep induction, awakenings during the night, early final awakening, and functioning during the day, compared with younger subjects. Overall quality of sleep and well-being during the day were similar in the two age groups. Elderly subjects complained less (although the statistical significance of this difference was not as robust) about sleep duration and daytime sleepiness, which might be due to the less demanding lifestyles of the elderly. 4.3. Daytime sleepiness One of the most important points addressed in this intercountry comparison is the contrast between the intensity of the symptoms according to the AIS and the daily consequences evaluated by the ESS. For example, Brazilians, who describe the most severe symptoms in the AIS, only rank fifth for being dangerously sleepy during the day, while Belgians, who describe moderate symptoms in the AIS, are the most likely to feel dangerously sleepy during the day. It should be noted, however, that percentages of subjects in all countries who were considered ‘very sleepy’ or ‘dangerously sleepy’ on the ESS were very similar to percentages of subjects with ‘moderate’ or ‘severe’ daytime sleepiness problems on the AIS. We may hypothesize that being sleepy during the day is more acceptable in some countries than in others because of the heat. The results may also indicate that some local adaptations of the ESS might be required concerning the criteria taken into account, some of which specifically pertain to industrial countries (e.g. sleeping in front of the television or in a car). Alternative criteria should be defined to better correspond to the daily lives of subjects in different countries. 4.4. Management of sleep problems The management of sleep disturbances also appeared to vary widely between the countries in this study. Only a third
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
of people who reported that they ‘do not sleep well’ had visited a physician to discuss their sleep problem. This rate appeared to be much lower in Asia than in Europe. Of those subjects who reported sleep disturbances, about a third had taken sleep medications. More detailed information on sleep disorder management will form the subject of a publication separate from this study.
5. Conclusion Although there seem to be important global variations in the prevalence of insomnia, its symptoms and their management, this survey shows that about one in four individuals world-wide do not think they sleep well. Moreover, self-reported sleep problems could be underestimated in the general population. Sleep disorders seem to be considered as normal by many individuals, rather than as a real pathology that may affect their lives and, potentially, jeopardize their health. Overall, there is a need for increased awareness of the importance of disturbed sleep and the improved detection and management of sleep disorders. The under-recognition of sleep problems in primary care could be aided by the routine use of screening tools such as the AIS, which proved to be an effective and convenient scoring instrument in this study.
Acknowledgements The following SLE-EP survey investigators coordinated the data collection in their respective countries: Professor R. Cluydts (Belgium), Professor M. Zhang (China), Professor T. Ohta (Japan), Dr A. Atalaia (Portugal), and Professor E. Estivill (Spain). The Sanofi-Synthelabo Group provided organizational support.
References [1] Ohayon M. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 2002;6:97–111. [2] Janson C, Gislason T, De Backer W, Plaschke P, Bjornsson E, Hetta J, et al. Prevalence of sleep disturbances among young adults in three European countries. Sleep 1995;18:589–97. [3] Chevalier H, Los F, Boichut D, Bianchi M, Nutt DJ, Hajak G, et al. Evaluation of severe insomnia in the general population: results of a European multinational survey. J Psychopharmacol 1999;13(Suppl 1): S21–S4.
13
[4] Ohayon MM, Roth T. What are the contributing factors for insomnia in the general population? J Psychosom Res 2001;51:745–55. [5] Soldatos CR, Dikeos DG, Paparrigopoulos TJ. Athens insomnia scale: validation of an instrument based on ICD-10 criteria. J Psychosom Res 2000;48:555–60. [6] Johns MC. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540–5. [7] Soldatos CR, Dikeos DG, Paparrigopoulos TJ. The diagnostic validity of the Athens Insomnia Scale. J Psychosom Res 2003;55: 263–7. [8] Ustun TB, Privett M, Lecrubier Y. Form, frequency and burden of sleep problems in general health care: a report from the WHO collaborative study on psychological problems in general health care. Eur Psychiatry 1996;11(Suppl 11):5S–110. [9] Hajak G. Insomnia in primary care. Sleep 2000;23(Suppl 3):S54–S63. [10] Ohayon G, Paiva T, Sagales T. Comparison of insomnia prevalence in the general population of Portugal and Spain. Sleep 2001;24(Suppl): A345. [11] Vela-Bueno A, de Iceta M, Fernandez C. Prevalencia de los trastornos del sueno en la ciudad de Madrid. Gac Sanit 1999;13:441–8. [12] Souza CJ, Magna LA, Reimao R. Insomnia and hypnotic use in Campo Grande general population, Brazil. Arq Neuro-Psiquiatr 2002; 60:702–7. [13] Rocha FL, Guerra HL, Lima-Costa MF. Prevalence of insomnia and associated socio-demographic factors in a Brazilian community: The Bambui study. Sleep Med 2002;3:121–6. [14] Doi Y, Minowa M, Okawa M, Uchiyama M. Prevalence of sleep disturbance and hypnotic medication use in relation to sociodemographic factors in the general Japanese population. J Epidemiol 2000;10:79–86. [15] Kim K, Uchiyama M, Okawa M, Liu X, Ogihara R. An epidemiological study of insomnia among the Japanese general population. Sleep 2000;23:41–7. [16] Zeitlhofer J, Schmeiser-Rieder A, Tribl G, Rosenberger A, Bolitschek J, Kapfhammer G, et al. Sleep and quality of life in the Austrian population. Acta Neurol Scand 2000;102:249–57. [17] Tribl GG, Schmeiser-Rieder A, Rosenberger A, Saletu B, Bolitschek J, Kapfhammer G, et al. Sleeping habits in the Austrian population. Sleep Med 2002;3:21–8. [18] SINE Study Group. Hajak G. Epidemiology of severe insomnia and its consequences in Germany. Eur Arch Psychiatry Clin Neurosci 2001; 251:49–56. [19] Backaus J, Junghanns K, Mueller-Popkes K, Broocks A, Riemann D, Hajak G, et al. Short-term training increases diagnostic and treatment rate of insomnia in general practice. Eur Arch Psychiatry Clin Neurosci 2002;252:99–104. [20] Weske G, Voderholzer U, Riemann D, Berger M. Sleep disorders. Differential diagnosis and therapy in general practice. MMW Fortschr Med 2001;143(Suppl 2):5–8. [21] Wittchen HU, Krause P, Hofler M, Pittrow D, Winter S, Spiegel B. NISAS 2000: The Nationwide Insomnia Screening and Awareness Study. Prevalence and interventions in primary care. Fortschr Med Orig 2001;119:9–19. [22] Schubert CR, Cruickshanks KJ, Dalton DS, Klein BE, Klein R, Nondahl DM. Prevalence of sleep problems and quality of life in an older population. Sleep 2002;25:889–93.
How do individuals sleep around the world? Results from a single-day survey in ten countries Constantin R. Soldatosa,*, Franc¸ois A. Allaertb, Tatsuro Ohtac, Dimitris G. Dikeosa a
Sleep Research Unit, Department of Psychiatry, University of Athens, Vas Sofias 72, Athens 11528, Athens, Greece b Department of Epidemiology and Public Health, McGill University, Montreal, Canada c Graduate School of Medicine, Nagoya University, Nagoya, Japan Received 28 May 2004; received in revised form 11 October 2004; accepted 14 October 2004
Abstract Background and purpose: To describe between-country differences in both the prevalence and type of sleep disorders seen across the globe, and to provide information on how impaired sleep impacts daytime functioning. Patients and methods: The study is a large-scale, global cross-sectional survey conducted on International Sleep Well Day (March 21), 2002. A standardized questionnaire was used in 10 countries under the guidance of local survey managers. In addition, the Athens Insomnia Scale (AIS) and the Epworth Sleepiness Scale (ESS) were completed. Subjects included in the study were adults from 10 countries representing different continents with clear variations in lifestyle. Results: The total number of questionnaires collected was 35,327. Overall, 24% of subjects reported that they did not sleep well. According to self-assessments using the AIS, 31.6% of subjects had ‘insomnia’, while another 17.5% could be considered as having ‘subthreshold insomnia’. According to ESS scores, 11.6% of subjects were found to be ‘very sleepy’ or ‘dangerously sleepy’ during the day. Conclusions: Although there seem to be important global variations in the prevalence of insomnia, its symptoms and their management, about one in four individuals do not think they sleep well. Moreover, self-reported sleep problems could be underestimated in the general population. Overall, there is a need for increased awareness of the importance of disturbed sleep and the improved detection and management of sleep disorders. q 2004 Elsevier B.V. All rights reserved. Keywords: Sleep disorders; Insomnia; Survey; Sleep; Prevalence; Daytime sleepiness
1. Introduction Sleep troubles are a common complaint that can have a huge impact on patients’ quality of life and ability to function. However, a recent review of the literature covering more than 50 studies of insomnia, based on data collected in various representative community-dwelling populations, showed that estimates of prevalence vary widely, from less than 5 to 40% [1]. Prevalence estimates can be affected by a number of factors, such as the characteristics of the population sampled, the definition of insomnia, and regional perceptions and management practices regarding sleep * Corresponding author. Tel.: C30 210 728 93 24; fax: C30 210 725 13 12. E-mail address: [email protected] (C.R. Soldatos). 1389-9457/$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2004.10.006
disorders. Few epidemiological surveys have been conducted that have used a single method to evaluate impaired sleep prevalence simultaneously in different countries. As a result, it is not clear precisely how much the prevalence of insomnia varies between global regions. It is not known whether the variability seen in studies performed to date reflects real regional differences or simply different definitions and/or methodologies. Janson et al. [2] studied the geographic variation in sleep complaints to identify risk factors for sleep disturbances in 2202 subjects from three Northern European countries (Sweden, Iceland and Belgium) using a standardized questionnaire and sleep diaries. However, this study selected only middle-aged adults (aged 20–45 years) and did not refer to a validated insomnia scale, making regional comparisons difficult. Chevalier et al. [3] conducted
6
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
a survey of severe insomnia and its effect on quality of life and healthcare consumption in five Northern European countries using a 4-item questionnaire, reporting prevalence rates of severe insomnia ranging from 4 to 22%. Finally, a telephone survey conducted by Ohayon and Roth [4] in 24,600 general population-based subjects aged R15 years from six European countries indicated a global insomnia prevalence of 16.8%, according to the criteria of the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). However, only overall survey results were presented and no between-country comparisons were available. The objectives of the present investigation were to describe between-country differences in both the prevalence and type of sleep disorders seen across the globe, and to gain information on how impaired sleep generally impacts daytime functioning.
2. Methods
represented Northern Europe, Portugal and Spain represented Southern Europe, and Slovakia represented the Central European population. China and Japan were studied in Asia, Brazil in South America, and South Africa in the African continent. Brazil and South Africa were taken to represent the Southern hemisphere. The total number of questionnaires collected was 35,327, ranging from 202 in South Africa to 10,424 in Japan (Table 1). The global mean age of surveyed individuals was 39.0 years (range 15–99, median 36 years), with 7.9% aged over 65 years. Approximately half of them (49.8%) were men and 64.1% were in the workforce (employed or currently unemployed but looking for work). However, this sociodemographic profile differed between countries. There were more men in the Spanish sample (68.7%) and more women in the Belgian, Slovakian and South African samples (57.8, 57.9 and 58.2%, respectively). Other notable regional differences included a higher mean age in the German sample (49.8 years), a lower than average proportion of individuals in the workforce in Portugal (33.1%), and higher proportions in Japan (74.8%) and South Africa (77.2%).
2.1. Study design and data collection 2.3. Assessments A cross-sectional survey was conducted on International Sleep Well Day (March 21), 2002. A common standardized questionnaire was used in all countries under the guidance of local survey managers, who ensured the feasibility of data collection according to local practices. Methods used to collect data in the different regions included: China, Japan, Slovakia, South Africa—collection from a general population in public booths in various locations (e.g. streets, railway stations, shopping centers); Austria, Brazil, Germany, Portugal, Spain—collection by telephone screening; Japan—by internet poll and at the work place; and Belgium—by internet poll and at the work place, as well as (4.2% of the Belgian questionnaires) at waiting rooms of primary care physicians and/or sleep laboratories. In all settings, the participants gave informed consent after the procedure was fully explained to them. 2.2. Study participants Subjects included in the study were adults from 10 countries representing four different continents with clear variations in lifestyle. Austria, Belgium and Germany
In all 10 countries, the survey was conducted with the same standardized case report form translated into local languages. The form included the following categories: a socio-demographic profile evaluation, a sleep outcomes evaluation that assessed sleep habits, subjective perception of sleep quality under the form of a general question: “Overall, do you consider that you sleep well?”, and, if applicable, a questionnaire-based assessment of sleeppromoting actions taken by subjects. All subjects also completed the Athens Insomnia Scale (AIS) [5] and the Epworth Sleepiness Scale (ESS) [6]. These scales were chosen because they appear to be the simplest scales that could be quickly completed, given the way the data were collected. However, due to time constraints, no cultural validation of the two scales could be implemented in all 10 countries. The AIS is intended to record the individual’s own assessment of any sleep disorder he/she might have experienced, provided that it occurred at least three times per week during the last month. It contains eight questions rated from 0 to 3. If the total score of the AIS is 6 or higher,
Table 1 Demographic data
Number % Male Mean age (GSD) Percentage in the workforce
Total/ average
Austria
Belgium
Brazil
China
Germany
Japan
Portugal
Slovakia
South Africa
Spain
35,327 49.8 39 G15.3 64.1
490 52.0 44.3 G18.8 49.8
6832 42.2 40.0 G16.3 62.6
1999 48.7 37.4 G15.0 61.6
10,079 51.6 37.1 G14.9 64.3
2016 46.5 49.8 G18.2 48.6
10,424 50.9 36.1 G10.7 74.8
784 48.2 44.5 G22.8 33.1
502 42.1 38.4 G16.0 57.6
202 41.8 37.7 G15.6 77.2
1999 68.7 47.7 G18.4 46.8
22.9 30 25.0 60 12.0 60 25.5 45
13.4 60
22.0 30
8h00 22.9G31.5 8h00 34.4G42.1 7h00 22.2G21.4
32.6 60 42.4 40 19.4 30 23.1 60
21.5 60
8h00 15.2G18.4 7h30 24.8G26.5
8h00 26.5G29.4
8h00 26.1G30.0
8h00 28G27.4
7h30 22.1G25.7
8h20 22.7G27.2
8h00 17.9G21.1
478G79 467G84 413G69 480G71 454G83
478G80
464G92
484G79
452G71
504G89
471G82
7:00 AM (33.0%) 6:00 AM (38.5%) 6:00 AM (44.8%) 6:00 AM (40.7%) 6:00 AM (43.4%) 6:00 AM (35.2%)
7:00 AM (35.0%)
6:00 AM (37.8%)
7:00 AM (36.6%)
6:00 AM (42.1%)
7:00 AM (43.6%)
10:00 PM (39.3%)
12:00 PM (40.6%)
South Africa
10:00 PM (32.5%) 11:00 PM (33.3%)
12:00 PM (27.2%)
Japan
10:00 PM (41.5%) 11:00 PM (30.5%) 10:00 PM (29.3%)
Brazil Belgium
11:00 PM (35.8%) 10:00 PM (44.9%)
Austria
11:00 PM (32.0%)
Average Table 2 Sleep habits during the week
7
the patient is diagnosed as insomniac [7]. For the purposes of the present study, we also assessed separately those subjects with an AIS score of 4 or 5, who were considered as presenting with ‘sub-threshold insomnia’. The ESS evaluates the likelihood to doze off or to fall asleep in eight different situations. The answers are rated from 0 to 3 and a total score is calculated. The individual is considered sleepy if the score is greater than 6 and below 10, very sleepy if it is greater than 10 and below 16 and dangerously sleepy if it is 16 or over [6]. 2.4. Data management and statistical analyses Data were collected in each country using the same format and software. All data were sent to a single university statistical center, where they were analyzed according to a pre-defined statistical plan using SAS 8.2 software. Between-country differences were assessed using the c2 test and analysis of variance (ANOVA), followed by the post-hoc Dunnett’s test. A P-value of less than 0.05 was considered to be statistically significant.
3. Results 3.1. Sleep habits
Time of going to bed (mode, and percentage of subjects reporting this value in brackets) Time of waking up (mode, and percentage of subjects reporting this value in brackets) Sleep duration (meanGSD, in min) Sleep duration (median, in h) Time to fall asleep (meanGSD, in min) Regular nap (% of subjects) Nap duration (median, in min)
China
Germany
Portugal
Slovakia
Spain
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
The average (mode) bedtime on weekdays was 11 PM, with 32.0% of individuals reporting that went to bed at this time (Table 2). The Spanish and the Portuguese reported later bedtimes (12 PM). On weekends, the average bedtime was slightly later, at 12 PM. The average (mode) waking time during the week was 6 AM, with 35.2% of participants reporting that they awoke at this time, and 31.6% at 7 AM. On weekends, the average wake-up time was 8 AM. The mean sleep duration during the week was 454 min (median 450 min [7.5 h]). The shortest sleep duration was reported in Japan (413 min) and the longest in Portugal (504 min). An average of 31.4% (range 25.9–36.4%) of responders reported that they slept for 7–8 h per night, and 25.7% (15.6–36.5%) for 6 to 7 h. The global mean sleep latency (the time taken to fall asleep) during the week was 24.8 min (range 15.2– 34.4 min), with a median value of 15 min. Approximately 40% of subjects reported that they needed between 0 and 15 min to fall asleep, and 20% reported a latency period of 15–30 min. At the other end of this spectrum, 39.4% of participants (range 22.1–50.2%) required more than 30 min to fall asleep, and 13.1% needed at least an hour. Approximately one quarter of the overall study population (23.1%) reported that they regularly nap on weekdays (Table 2). The highest napping rate (42.4%) was reported in Brazil and the lowest in Japan (12.0%) and Portugal (13.4%). In those who reported a regular nap, the median duration was 1 h.
8
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
Fig. 1. Percentage of subjects reporting that that ‘they do not sleep well’.
3.2. Sleep disturbances 3.2.1. Global self-assessment To the global question: “Overall, do you consider that you sleep well?” 24% answered ‘No’ (Fig. 1). Percentage of subjects answering ‘No’ to this question ranged from
10.4% in Austria to 32.2% in Belgium; overall the difference between countries was highly statistically significant with a c2 value of 50,264 (9 df) and a P! 0.0001. For subjects who answered ‘No’, the mean duration of sleep disturbances was 5.7 years (median 3 years).
Fig. 2. Percentage of subjects by AIS score category.
O0.0001 O0.0001 1280,62 323,50 4.3 7.3 10.1 20.2 4.2 9.0 2.7 7.9 5.5 12.2 2.5 5.1 25.2 16.3 8.2 9.4 8.0 10.5
12.6 12.9
6.5 8.0
O0.0001 O0.0001 O0.0001 785,40 750,49 1594,46 10.4 7.5 6.9 24.6 19.7 15.2 10.6 6.8 10.0 9.3 5.6 2.9 11.5 10.9 4.9 7.2 4.9 4.4 29.5 26.5 26.7 6.3 5.3 8.0 13.4 12.4 8.1
18.1 16.9 13.9
11.4 11.6 5.0
O0.0001 O0.0001 O0.0001 705,30 1488,29 767,29 12.4 15.4 10.5 25.5 33.2 20.0 11.1 15.3 12.3 9.1 9.7 9.4 8.6 3.8 5.2 8.9 10.9 9.3 11.7 12.8 9.3 27.8 16.8 17.3 7.1 7.6 6.3
Sleep induction (%) Night awakenings (%) Final awakening earlier than desired (%) Total sleep duration (%) Overall quality of sleep (%) Sense of well-being during the day (%) Functioning during the day (%) Sleepiness during the day (%)
12.1 12.2 10.1
14.8 22.2 16.8
China Brazil Belgium Austria Total/ Average
Table 3 Percentage of subjects with AIS symptoms rated 2 or 3 (moderately and severely disturbed)
Germany
Japan
Portugal
Slovakia
South Africa
Spain
Overall c2
P (dfZ9)
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
9
3.2.2. Self-assessment based on the AIS As shown in Fig. 2, 17.5% of the participants exhibited an AIS score of 4 or 5 (‘sub-threshold insomnia’), and 31.6% had a score of 6 or higher (‘insomnia’). Large global variations were observed. The prevalence rates of ‘sub-threshold insomnia’ (i.e. AIS score 4 or 5) ranged from 13.9% (Belgium and Brazil) to 25.7% (Slovakia). Rates of ‘insomnia’ (AIS score 6 or greater) ranged from 17.4% (Germany) to 79.8% (Brazil); the overall c2 of the difference between countries was 2676.10 (9 df, P!0.0001) and of the difference between Brazil and all other countries taken together was 2276.69 (dfZ2, P!0.0001). Table 3 describes percentages of individuals who quoted grade 2 or 3 (‘markedly’ or ‘very affected’) for each of the AIS items. On average, 12.1% (nZ4161; range 7.1–27.8%) of subjects considered their sleep induction to be delayed; 12.2% (3.8–33.2%) reported frequent night awakenings; 10.1% (5.2–20.0%) had final awakening earlier than desired; 13.4% (6.3–29.5%) found their sleep duration was insufficient; and 12.4% (4.9–26.5%) said they were not satisfied with the quality of their sleep. Finally, 8.1% of the participants (4.4–26.7%) reported a decreased sense of wellbeing during the day; 8.0% (2.5–25.2%) said their daytime functioning was decreased; and 10.5% (5.1–20.2%) said they felt sleepy during the day. For all but two of the individual items on the AIS, subjects from Brazil recorded the highest ratings. Problems with sleep latency was the only complaint in 12.1% (nZ1,252) of subjects rating 2 or 3 for any insomnia symptom (nZ10,344). Similarly, as isolated complaints, any problem with sleep maintenance (nightly awakenings and/or early awakening) was observed in 20.8%, insufficient sleep duration in 12.8%, and poor sleep quality in 6.5%. Of the responders who rated even one item on the AIS as 1 or greater, 52.1% had isolated symptoms while 47.9% had more than one symptom. 3.2.3. Global vs. AIS self-assessment When the results of the global question on sleep were compared with the AIS classification of subjects, 91.6% of people who thought they did not sleep well had AIS scores of at least 4, indicating at least ‘sub-threshold insomnia’, and 77.8% had scores of at least 6, indicating a diagnosis of insomnia. Among those who thought they did sleep well, 82.8% had an AIS score of less than 6 and 65.4% a score of less than 4. These results indicate that some insomniacs may underestimate their sleep problems. 3.3. Effect of aging The frequency of a grade 2 or 3 rating on the various AIS items in subjects aged 18–65 years was compared with that of subjects older than 65 years (Table 4). The frequency of four items rated as grade 2 or 3 was significantly higher in elderly than in younger subjects: delayed sleep induction
10
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
Table 4 Percentage of subjects with AIS symptoms quoted 2 or 3 (moderately or severely impaired) as a function of age Age (years)
Sleep induction
Awakenings during the night
Final awakening earlier than desired
Total sleep duration
Overall quality of sleep
Sense of well-being during the day
Functioning during the day
Sleepiness during the day
!65 O65 P-value
11.6 18.3 !0.0001
11.5 21.9 !0.0001
9.7 15.1 !0.0001
13.6 11.6 !0.01
12.6 12.2 n.s.
8.1 9.0 n.s.
7.9 10.0 !0.0001
10.7 8.6 !0.001
n.s., not statistically significant.
(18.3 vs. 11.6%); awakenings during the night (21.9 vs. 11.5%); final awakening earlier than desired (15.1 vs. 9.7%); and functioning during the day (10.0 vs. 7.9%). In contrast, older subjects complained less frequently of insufficient total sleep duration (11.6 vs. 13.6%) and sleepiness during the day (8.6 vs. 10.7%).
the highest rates in Portugal (45.7%) and in South Africa (52.8%). Herbal teas were used to improve sleep by 22.1% of subjects (4.2% in Japan to 48.0% in Slovakia), 19.4% took alcohol (9.8% in Austria to 30.3% in Japan), and 40.0% modified their consumption of tea or coffee (10.0% in Japan to 49.0% in Austria).
3.4. Prevalence based on DSM-IV definition 4. Discussion In an attempt to compare the results with data in the literature that usually refer to the DSM-IV definition of primary insomnia, percentages of subjects presenting with both night-time and daytime symptoms on the AIS are presented in Table 5. On average, 12.1% of subjects met this criterion. The highest percentage was reported in Brazil (31.8%), and the lowest in Germany (5.2%). 3.5. Daytime sleepiness The results of the ESS scores indicated that 11.6% of subjects could be considered very sleepy or dangerously sleepy (ESS score O10). Rates ranged from 6.2% in China to 24.5% in South Africa. The c2 value for the differences between the ten countries for the comparison between subjects who were ‘not sleepy’ and all others was 1138.38, yielding (for 9 df) a P-value of less than 0.0001. Fig. 3 shows percentages of subjects reporting the four categories of sleepiness in the 10 countries. 3.6. Actions taken for the management of sleep problems Among those who thought they did not sleep well, 30.7% (8.0% in Japan to 55.5% in Portugal) reported that they had visited a physician regarding their sleep impairment. In addition, 31.4% of them said they had taken sleep medications. The lowest rates of medication were reported in Austria (9.8%) and Japan (15.3%), and
To our knowledge, this is the first international study attempting to assess sleep behaviors on the same day and with the same standardized questionnaire and scales, and covering 10 countries in four continents. Only three multicountry European studies of the general population have been published [2–4], while the World Health Organization Collaborative Study that involved 14 countries in three continents assessed the burden of sleep problems in primary care patients rather than the general population [8]. Thus, the present survey is probably the largest ever conducted on a general population sample and its results should provide useful insights into world-wide differences in sleep behaviors. There are a number of weaknesses related to the methods of data collection in the different countries that may weaken the extrapolation of findings in the regional samples listed here to their respective national populations, and which limit the relevance of inter-country comparisons. Variability in sample sizes and the ways in which data were collected was considerable across countries. The method which was used for the recruitment of the participants is vulnerable to several biases, such as that individuals with sleep disturbances are more likely to accept to answer the questionnaire, and that people with physical disabilities or limitations are less likely to be included (for example, when the questionnaires were administered in the streets, railway stations or shopping centers). In addition, significant
Table 5 Number and percentage of subjects with at least one night-time symptom (rated 2 or 3) and one daytime symptom (rated 2 or 3) in the AIS Number (%)
Total
Austria
Belgium
Brazil
Yes No
4,087 (12.1) 48 (9.8) 1,023 (16.0) 633 (31.8) 952 (10.0) 102 (5.2) 1,021 (10.3) 49 (6.2) 55 (11.1) 39 (20.5) 165 (8.2) 29,661 (87.9) 442 (90.2) 5,368 (84.0) 1,359 (68.2) 8,592 (90.0) 1,876 (94.8) 8,865 (89.7) 735 (93.8) 439 (88.9) 151 (79.5) 1834 (91.8)
Missing dataZ1579; Overall c2Z1044.08, dfZ9, P!0.0001.
China
Germany
Japan
Portugal
Slovakia
South Africa
Spain
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
11
Fig. 3. Percentage of subjects by ESS score category.
differences between the countries existed with regard to gender ratios, mean ages, and proportions of people in the workforce. It is difficult to distinguish which of these differences were due to local social structures and which were due to dissimilar recruitment conditions. Moreover, the ‘collection day’ was March 21, which is at the end of winter in north hemisphere and at the end of summer in the south hemisphere. This seasonal difference might play a role in the heterogeneity of the results. However, since only about 7% of the total study population lived in the southern hemisphere, it should not have substantially influenced the overall results. Also, gender differences do not seem to have played any major role in the explanation of the variability among countries, since country differences in male:female ratio were not deviating considerably (with the exception of Spain) from the expected 49:51; besides, they were much smaller than differences in the rates of any insomnia-related characteristic among countries. Nonetheless, the main outcome of this study appears to be worthy of interest, if one takes into account that the data constitute an ‘instant picture’ of sleep behaviors around the world. 4.1. Sleep habits Overall, sleep behaviors were similar around the world. Although bedtimes and waking times differed between countries, possibly due to climatic conditions and cultural traditions, total sleep durations were similar and seemed to correspond to the same physiological need across countries.
More surprising was the frequency of napping, which was common in all countries, including the most industrial ones, in contrast to the widespread perception that napping is a ‘warm country’-specific habit. 4.2. Sleep disturbances The global rate of people thinking they do not sleep well is high (about one in four subjects) and appears comparable from one country to another. Globally, this fits into the range of insomnia prevalence rates reported in the literature (range 9–36%), considering either all 10 countries assessed here or only the six European ones [1,4,8,9]. Specific country comparisons with previously reported population surveys conducted in adults of all ages are possible only for seven countries. Rates of insomnia reported in the current study are similar to those reported previously in Portugal [10], Spain [10,11], Brazil [12,13] and Japan [14,15]. The rate reported in Austria (10%) was lower than that reported previously in the literature (25–32%) [16,17]. In contrast, the rate reported here for Germany (16%) is higher than that reported in a recent review (8.5%) [1]. However, this recent review evaluated patients with ‘insomnia symptoms plus daytime consequences’ [1], which probably represents a more severe form of insomnia than the general sleep impairments seen in our survey. Accordingly, Hajak et al. [18] reported a prevalence rate of 4% of ‘severe insomnia’ in the German population. In German primary care facilities, insomnia has been
12
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
reported in 19.3–29% of patients [19–21]. In Belgium, Hajak et al. [18] reported a 9% prevalence of ‘severe insomnia’, well below the rate of people who reported that they ‘do not sleep well’ in the present study (32%). The different rates should be attributed to the difference between the two studies in the level of severity of insomnia required for a subject to be considered a case, since the number of people in our study who were recruited from medical settings and sleep laboratory waiting rooms was 4.2% of the total number of subjects ascertained in Belgium (total nZ6832) and it might not have interfered significantly with the results. No prior data on sleep disturbances in Slovakia, China or South-Africa were available, and the data in the current study pertaining to these countries should be considered original. If one considers the association of at least one marked night-time AIS symptom (moderate or severe) with at least one daytime AIS symptom of similar intensity as a close image of the DSM-IV definition diagnosis of insomnia, then the percentages we obtained in Germany (5.2%), in Portugal (6.2%) and in Spain (8.3%) are similar to those reported in the literature (4–6%, 8.7 and 6.0%, respectively) [1,10,18]. It must be noted, however, that insomnia according to the DSM-IV can be diagnosed even without the presence of a compromised feeling of well-being or functioning and/or somnolence during the day. For the diagnosis to be established, a sleep problem and preoccupation with it is sufficient. In that sense, many more subjects can be considered to be insomniacs than those who report the daytime consequences assessed in the AIS, and those who report insomnia symptoms plus daytime consequences may represent cases suffering from a more severe form of insomnia. Comparisons between the rates of people who thought they did not sleep well and the prevalence of insomnia as diagnosed with the AIS (total score 6 or higher) provided interesting data. In most European countries (5 out of 6), Japan and China, both rates were very similar, indicating agreement between participants’ subjective perceptions of sleep and the semi-subjective assessment provided by the AIS. However, in Brazil, Austria and South Africa, rates of those who claimed they did not sleep well were much lower than prevalences of insomnia indicated by the AIS. The most striking difference was seen in Brazil, where only 19% of subjects reported that they did not sleep well, whereas the prevalence of AIS ‘insomnia’ was close to 79%. Some adaptation of the AIS might be required for certain populations. For example, the cut-off level of the AIS score might need to be higher than 6 for some countries where heat may have a heavy influence on the quality of the sleep. Nonetheless, the 32% global prevalence of AIS-defined insomnia in the current study is consistent with that reported in the literature [1,4,8,9]. Specific country rates were similar to published data for Portugal and Spain [10,11] slightly lower than those published for Austria [16,17], and higher than those published for Brazil and Japan [12–15]. If one
considers together AIS ‘sub-threshold insomnia’ and ‘insomnia’ (scores R4), the 49% rate is higher than all that previously reported. Thus, the present study supports previous suggestions that, at least for the majority of the world population, an AIS score of 6 or higher should be considered as the cut-off defining insomnia [7]. As well as AIS scores, types of sleep disturbances reported differed between countries. Austrian, Portuguese and Chinese subjects reported more ‘delayed sleep induction’, whereas subjects in Belgium, Germany, Slovakia, Spain and South Africa reported more ‘frequent night awakenings’. Brazilian and Japanese subjects most frequently reported ‘insufficient sleep duration’. Finally, our results regarding the influence of aging on sleep problems are consistent with previous findings that certain sleep components are altered in the elderly [22]. The subjects in our sample who were older than 65 years had significantly higher AIS ratings of sleep induction, awakenings during the night, early final awakening, and functioning during the day, compared with younger subjects. Overall quality of sleep and well-being during the day were similar in the two age groups. Elderly subjects complained less (although the statistical significance of this difference was not as robust) about sleep duration and daytime sleepiness, which might be due to the less demanding lifestyles of the elderly. 4.3. Daytime sleepiness One of the most important points addressed in this intercountry comparison is the contrast between the intensity of the symptoms according to the AIS and the daily consequences evaluated by the ESS. For example, Brazilians, who describe the most severe symptoms in the AIS, only rank fifth for being dangerously sleepy during the day, while Belgians, who describe moderate symptoms in the AIS, are the most likely to feel dangerously sleepy during the day. It should be noted, however, that percentages of subjects in all countries who were considered ‘very sleepy’ or ‘dangerously sleepy’ on the ESS were very similar to percentages of subjects with ‘moderate’ or ‘severe’ daytime sleepiness problems on the AIS. We may hypothesize that being sleepy during the day is more acceptable in some countries than in others because of the heat. The results may also indicate that some local adaptations of the ESS might be required concerning the criteria taken into account, some of which specifically pertain to industrial countries (e.g. sleeping in front of the television or in a car). Alternative criteria should be defined to better correspond to the daily lives of subjects in different countries. 4.4. Management of sleep problems The management of sleep disturbances also appeared to vary widely between the countries in this study. Only a third
C.R. Soldatos et al. / Sleep Medicine 6 (2005) 5–13
of people who reported that they ‘do not sleep well’ had visited a physician to discuss their sleep problem. This rate appeared to be much lower in Asia than in Europe. Of those subjects who reported sleep disturbances, about a third had taken sleep medications. More detailed information on sleep disorder management will form the subject of a publication separate from this study.
5. Conclusion Although there seem to be important global variations in the prevalence of insomnia, its symptoms and their management, this survey shows that about one in four individuals world-wide do not think they sleep well. Moreover, self-reported sleep problems could be underestimated in the general population. Sleep disorders seem to be considered as normal by many individuals, rather than as a real pathology that may affect their lives and, potentially, jeopardize their health. Overall, there is a need for increased awareness of the importance of disturbed sleep and the improved detection and management of sleep disorders. The under-recognition of sleep problems in primary care could be aided by the routine use of screening tools such as the AIS, which proved to be an effective and convenient scoring instrument in this study.
Acknowledgements The following SLE-EP survey investigators coordinated the data collection in their respective countries: Professor R. Cluydts (Belgium), Professor M. Zhang (China), Professor T. Ohta (Japan), Dr A. Atalaia (Portugal), and Professor E. Estivill (Spain). The Sanofi-Synthelabo Group provided organizational support.
References [1] Ohayon M. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 2002;6:97–111. [2] Janson C, Gislason T, De Backer W, Plaschke P, Bjornsson E, Hetta J, et al. Prevalence of sleep disturbances among young adults in three European countries. Sleep 1995;18:589–97. [3] Chevalier H, Los F, Boichut D, Bianchi M, Nutt DJ, Hajak G, et al. Evaluation of severe insomnia in the general population: results of a European multinational survey. J Psychopharmacol 1999;13(Suppl 1): S21–S4.
13
[4] Ohayon MM, Roth T. What are the contributing factors for insomnia in the general population? J Psychosom Res 2001;51:745–55. [5] Soldatos CR, Dikeos DG, Paparrigopoulos TJ. Athens insomnia scale: validation of an instrument based on ICD-10 criteria. J Psychosom Res 2000;48:555–60. [6] Johns MC. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540–5. [7] Soldatos CR, Dikeos DG, Paparrigopoulos TJ. The diagnostic validity of the Athens Insomnia Scale. J Psychosom Res 2003;55: 263–7. [8] Ustun TB, Privett M, Lecrubier Y. Form, frequency and burden of sleep problems in general health care: a report from the WHO collaborative study on psychological problems in general health care. Eur Psychiatry 1996;11(Suppl 11):5S–110. [9] Hajak G. Insomnia in primary care. Sleep 2000;23(Suppl 3):S54–S63. [10] Ohayon G, Paiva T, Sagales T. Comparison of insomnia prevalence in the general population of Portugal and Spain. Sleep 2001;24(Suppl): A345. [11] Vela-Bueno A, de Iceta M, Fernandez C. Prevalencia de los trastornos del sueno en la ciudad de Madrid. Gac Sanit 1999;13:441–8. [12] Souza CJ, Magna LA, Reimao R. Insomnia and hypnotic use in Campo Grande general population, Brazil. Arq Neuro-Psiquiatr 2002; 60:702–7. [13] Rocha FL, Guerra HL, Lima-Costa MF. Prevalence of insomnia and associated socio-demographic factors in a Brazilian community: The Bambui study. Sleep Med 2002;3:121–6. [14] Doi Y, Minowa M, Okawa M, Uchiyama M. Prevalence of sleep disturbance and hypnotic medication use in relation to sociodemographic factors in the general Japanese population. J Epidemiol 2000;10:79–86. [15] Kim K, Uchiyama M, Okawa M, Liu X, Ogihara R. An epidemiological study of insomnia among the Japanese general population. Sleep 2000;23:41–7. [16] Zeitlhofer J, Schmeiser-Rieder A, Tribl G, Rosenberger A, Bolitschek J, Kapfhammer G, et al. Sleep and quality of life in the Austrian population. Acta Neurol Scand 2000;102:249–57. [17] Tribl GG, Schmeiser-Rieder A, Rosenberger A, Saletu B, Bolitschek J, Kapfhammer G, et al. Sleeping habits in the Austrian population. Sleep Med 2002;3:21–8. [18] SINE Study Group. Hajak G. Epidemiology of severe insomnia and its consequences in Germany. Eur Arch Psychiatry Clin Neurosci 2001; 251:49–56. [19] Backaus J, Junghanns K, Mueller-Popkes K, Broocks A, Riemann D, Hajak G, et al. Short-term training increases diagnostic and treatment rate of insomnia in general practice. Eur Arch Psychiatry Clin Neurosci 2002;252:99–104. [20] Weske G, Voderholzer U, Riemann D, Berger M. Sleep disorders. Differential diagnosis and therapy in general practice. MMW Fortschr Med 2001;143(Suppl 2):5–8. [21] Wittchen HU, Krause P, Hofler M, Pittrow D, Winter S, Spiegel B. NISAS 2000: The Nationwide Insomnia Screening and Awareness Study. Prevalence and interventions in primary care. Fortschr Med Orig 2001;119:9–19. [22] Schubert CR, Cruickshanks KJ, Dalton DS, Klein BE, Klein R, Nondahl DM. Prevalence of sleep problems and quality of life in an older population. Sleep 2002;25:889–93.