- Top of page
- Data analysis
The extensive biological, personal, and psychosocial changes that characterize adolescence affect sleep-wake regulation and timing, making them vulnerable to insufficient sleep, irregular sleep-wake schedules, and sleep disturbances (Carskadon, Acebo & Jenni, 2004). Cross-sectional studies in adolescents found that insomnia is common and links with emotional, behavioral, and substance-related problems (Liu & Zhou, 2002; Roane & Taylor, 2008). Prospective studies indicated an association between insomnia and impaired psychosocial functioning at follow-up in 1–2 years and the onset of depression and substance abuse in young adulthood (Kaneita et al., 2009; Roane & Taylor, 2008; Roberts, Roberts & Duong, 2008).
Because of the significant health risk, many studies have been conducted on the epidemiology of insomnia during adolescence. Although Asian adolescents tend to go to bed later and obtain less sleep than their counterparts in North America and Europe, there is insufficient data to compare the prevalence of insomnia between adolescents from different cultures (Gradisar, Gardner & Dohnt, 2011). Due to different definition and methodology, the prevalence of insomnia derived from the studies varied widely, with estimates ranging from 10% to 30% (Chung & Cheung, 2008; Kaneita et al., 2006; Roberts, Lee, Hernandez & Solari, 2004; Roberts, Roberts & Chan, 2006). These epidemiological studies were limited by a lack of in-person clinical assessment and that recognized diagnostic criteria were not often used. Using structured interviews and the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition (DSM-IV) criteria (American Psychiatric Association, 2000), Ohayon and colleagues (Ohayon, Roberts, Zulley, Smirne & Priest, 2000) found that 4.1% of 15- to 18-year-olds in general population samples from France, Great Britain, Germany, and Italy had a diagnosis of insomnia in the past 30 days. Another interview-based study in the US showed that the prevalence of DSM-IV-defined current insomnia among 13- to 16-year-olds was 9.4% (Johnson, Roth, Schultz & Breslau, 2006). The prevalence of primary insomnia was estimated at 2.2% in Ohayon's study and 5.8% in the US study. Additional studies utilizing face-to-face assessment and standardized diagnostic criteria are warranted to accurately determine the prevalence of insomnia among adolescents and whether cross-cultural difference is present.
Apart from prevalence, other dimensions of insomnia epidemiology, including help-seeking behaviors and types of interventions, were poorly documented in the adolescent population. It was recognized that the utilization of health services in adolescents with mental health problems was quite low. A population survey in Norway, a country where mental health services are readily available and free of charge, found that among adolescents within the 10th percentile on anxiety and depression, only 24% sought help from health care professionals in the last 12 months (Zachrisson, Rodje & Mykletun, 2006). Available data is also limited regarding the interventions used by adolescents to alleviate their sleep difficulties. Only the use of hypnotics has been reported in previous studies, whereas the prevalence of other types of interventions is unknown. For example, in a survey of 1629 secondary school students in Hong Kong, 1.3% of 12- to 19-year-olds reported using hypnotics occasionally and 1.0% took hypnotics at least once per week (Chung & Cheung, 2008). Another study among 557 seventh-graders and 509 tenth-graders in China showed that 2.5% of the sample used hypnotics at least once per week (Liu, Zhao, Jin & Buysse, 2008). The reluctance in seeking treatment among adolescents with mental health problems suggested that a majority of adolescents with insomnia might not seek pharmacological or psychological treatments. The use of alcohol and over-the-counter medication for insomnia might have harmful side effects; whereas the benefits of other self-help treatments remained unclear. Given the negative consequences of untreated insomnia, it is important to gather more information on the types of interventions commonly used by adolescents to promote sleep.
The aim of the present study was to estimate the prevalence of insomnia symptoms and DSM-IV-defined insomnia diagnoses in a Chinese population by clinical interview and to describe the help-seeking behaviors and types of interventions among adolescents with insomnia. The findings would contribute to our understanding of insomnia in adolescents.
- Top of page
- Data analysis
The Insomnia Interview Schedule (Morin, 1993), a semistructured interview, was used to assess the subjects' sleep-wake problems, current sleep-wake schedule, use of sleeping aids, sleep problem history, bedroom environment, eating, exercise, and substance use habits, functional impairment, symptoms of other sleep disorders, and history of psychopathology and psychiatric treatment. Additional open-ended questions ‘Have you ever tried anything to help yourself sleep better?’ and ‘Have you ever told someone about your sleep problem?’ were asked. All types of self-help strategies and help-seeking behaviors that the participants have tried were included. For help-seeking behaviors, the students were specially asked if they have ever told a friend, a family member, a social worker, or consulted a healthcare provider for insomnia. For those who had no insomnia symptoms, we would skip the questions on functional analysis, psychopathology, psychiatric treatment, self-help strategies, and help-seeking behaviors. The interview took around 10–40 min and was administered by the author (KC).
Insomnia symptoms, including difficulty initiating sleep, difficulty maintaining sleep, and early morning awakening were rated using a 4-point scale (0 = no, 1 = mild, 2 = moderate or 3 = severe) based on both severity and frequency in the past month (Table 1). Sleep onset latency (SOL) was defined as the time usually taken to fall asleep after light off and no activity present. Wake after sleep onset (WASO) and waking up earlier than planned were based on self-report. Although a previous study showed that SOL or WASO ≥20 min every night for 2 weeks was the best cutoff criteria for insomnia (Lineberger, Carney, Edinger & Means, 2006), the study was conducted in middle-aged adults. We have adopted less severe quantitative criteria for insomnia in adolescents because younger people have stronger homeostatic sleep drive and less complaints of insomnia (Klerman & Dijk, 2008; Stewart et al., 2006); hence more likely to fall asleep quicker and have deeper sleep than middle-aged adults.
Table 1. Quantitative criteria for insomnia symptoms and DSM-IV insomnia in the past month
| ||Severitya||Frequencyb||Subjective distress/Functional impairment|
|Insomnia symptoms|| || ||Not required|
|Difficulty initiating sleep|
|Difficulty maintaining sleep|
|Early morning awakening|
Among the subjects with insomnia symptoms, we assessed the associated level of distress and functional impairment to determine whether they were having DSM-IV insomnia (American Psychiatric Association, 2000), of which difficulty initiating and maintaining sleep and early morning awakening have lasted for at least 1 month and the sleep disturbance or associated daytime fatigue has caused clinically significant distress or impairment in social, occupational, or other important areas of functioning. Although previous studies have used either three times per week or four times per week as a cutoff for insomnia (Johnson et al., 2006; Ohayon, 2002), we followed the DSM-IV and did not set frequency criteria for the diagnosis. To determine the causes of insomnia, we screened for symptoms of restless leg syndrome, periodic limb movement disorder, obstructive sleep apnea, narcolepsy, gastro-esophageal reflux, parasomnia, sleep-wake schedule disorder, and other medical and psychiatric disorders.
Other data included in the current analysis were derived from the questionnaire survey. The severity of daytime sleepiness was reported using the Chinese version of Epworth Sleepiness Scale (ESS; Chung, 2000; Johns, 1991), which is an 8-item scale assessing the likelihood of dozing in daily life situations, with higher scores representing more severe daytime sleepiness. A recent study found that the test-retest reliability of ESS over two weeks was high in adolescents (Gibson et al., 2006). The 12-item General Health Questionnaire (GHQ-12) was used to assess the current level of psychological distress (Goldberg, 1972). The students were asked to rate six positive and six negative statements using a 4-point Likert scale (less than usual, no more than usual, rather more than usual, or much more than usual). Each statement was scored using a binary code (0, 0, 1, 1) with total score ranging from 0 to 12. Higher scores indicate greater psychological distress. The Chinese version of GHQ-12 has been widely used in different populations (Li, Chan, Chung & Chui, 2010; Pan & Goldberg, 1990). On the basis of the interviewed sample, the Chinese versions of ESS and GHQ-12 had high internal consistency, with Cronbach's alpha .73 and .86 and moderate test-retest reliability over 2 weeks, using Pearson correlation coefficient .70 and .68, respectively. The students reported their habit of current tobacco use and the number of cigarette smoked per day. Alcohol use was reported using a 5-point scale (0: never; 1: rarely; 2: sometimes; 3: almost always; 4: often). Academic performance was assessed using a standardized scale as excellent, good or marginal (Chung & Cheung, 2008). The students also reported their parents' marital status, occupation, and years of education. We derived a 7-point index of family occupational status as the higher value of the father's and mother's occupational status (1: professional; 2: manager and administrator; 3: associate professional; 4: skilled nonmanual worker; 5: skilled or semiskilled manual worker; 6: unskilled manual worker; 7: unemployed/homemaker).
- Top of page
- Data analysis
Table 2 presents the students' sociodemographic, school, lifestyle, and clinical variables. There was a slightly higher proportion of girls in our sample (56.2%) and the proportion of students aged 12–13, 14, 15, 16, and 17–18 years was 27.2%, 25.9%, 25.9%, 14.5%, and 6.5%, respectively. Very few participants reported to have drinking and smoking habits and most of them enjoyed good physical health.
Table 2. Sample characteristics
|Variablesa||Total sample (n = 290)||No insomnia (n = 174)||Insomnia symptoms only (n = 89)||DSM-IV insomnia (n = 27)||Chi-square/F valueb||p-value|
|Female gender||163 (56.2)||103 (59.2)||53 (59.6)||15 (55.6)||0.15||.93|
|Age in year||14.4 ± 1.4 (12–19)||14.4 ± 1.3||14.4 ± 1.5||14.7 ± 1.7||0.80||.45|
|Parents martial statusc|
|Single||8 (2.8)||4 (2.3)||4 (4.5)||0 (0.0)||2.02||.73|
|Married/cohabiting||246 (85.1)||147 (85.0)||75 (84.3)||24 (88.9)|
|Divorced/widowed||35 (12.1)||22 (12.7)||10 (11.2)||3 (11.1)|
|Average parental education in year||8.7 ± 2.9 (0–16)||8.8 ± 3.0||8.8 ± 2.9||8.3 ± 2.8||0.41||.66|
|Professional, associate professionals and nonmanual workers||104 (36.2)||62 (36.0)||34 (38.6)||8 (29.6)||4.67||.59|
|Skilled and semiskilled manual workers||105 (36.6)||61 (35.5)||30 (34.1)||14 (51.9)|
|Unskilled manual workers||39 (13.6)||26 (15.1)||12 (13.6)||1 (3.7)|
|Unemployed/homemakers||39 (13.6)||23 (13.2)||12 (13.6)||4 (14.8)|
|Occasional or regular use of alcohol||31 (10.7)||13 (7.5)||15 (16.9)||3 (11.1)||5.44||.066|
|Current tobacco use||7 (2.4)||5 (2.9)||2 (5.6)||0 (0.0)||0.83||.66|
|Chronic medical problemsd||23 (7.9)||10 (5.7)||11 (12.4)||2 (7.4)||3.54||.17|
|Regular medical treatments||9 (3.1)||3 (1.7)||5 (5.6)||1 (3.7)||3.01||.22|
|Self-reported academic performance|
|Excellent||82 (28.3)||51 (29.3)||23 (29.3)||8 (29.3)||0.93||.92|
|Good||128 (44.1)||76 (29.3)||39 (29.3)||13 (29.3)|
|Marginal||80 (27.6)||47 (29.3)||27 (29.3)||6 (29.3)|
|Self-reported snoring||59 (20.3)||43 (24.7)||12 (13.5)||4 (14.8)||5.14||.076|
|History of sleepwalking||27 (9.3)||18 (10.3)||7 (7.9)||2 (7.4)||0.56||.76|
|History of sleep paralysis||27 (9.3)||11 (6.3)||6 (6.7)||10 (37.0)||27.12||.0001|
|Epworth sleepiness scale score||7.7 ± 4.3 (0–21)||7.4 ± 1.0 e||7.9 ± 4.5||9.6 ± 4.5e||3.38||.036|
|12-item general health questionnaire score||2.2 ± 2.9 (0–12)||1.8 ± 2.5e||2.3 ± 2.9e||4.5 ± 3.8e||11.32||.0001|
Prevalence of insomnia symptoms
The prevalence of the different types and severities of insomnia symptoms was presented in Table 3. One-hundred and sixteen of the 290 adolescents were assessed to have insomnia symptoms by our quantitative criteria, which was equivalent to an overall prevalence of 40.0%. Problem with difficulty initiating sleep was the most frequent (35.9%), followed by difficulty maintaining sleep (9.3%), and early morning awakening (6.6%). Ninety subjects of the total sample (31.0%) had one insomnia symptom and 26 (9.0%) had two or three insomnia symptoms simultaneously. Fifty-eight of the 290 subjects (20.0%) presented with only mild symptoms, 30 (10.3%) had at least one symptom that was moderate in severity, and 28 (9.7%) had at least one symptom that was severe.
Table 3. Prevalence, help-seeking behaviors, and self-help strategies for insomnia in adolescents
|Variablesa||Total sample (n = 290)||All insomnia (n = 116)||Insomnia symptoms only (n = 89)||DSM-IV insomnia (n = 27)||Chi-square/t valueb||p-value|
|Type of insomnia symptoms|
|Difficulty initiating sleep||104 (35.9)||104 (90.0)||77 (86.5)||27 (100.0)||4.06||.044|
|Difficulty maintaining sleep||27 (9.3)||27 (23.3)||16 (18.0)||11 (40.7)||6.01||.014|
|Early morning awakening||19 (6.6)||19 (16.4)||16 (18.0)||3 (11.1)|| ||.56|
|Number of insomnia symptoms|| || || || ||17.45||.0001|
|1||90 (31.0)||90 (77.6)||76 (85.4)||14 (51.9)|| || |
|2||18 (6.2)||18 (15.5)||7 (7.9)||11 (40.7)|
|3||8 (2.8)||8 (6.9)||6 (6.7)||2 (7.4)|
|Severity of insomnia symptoms|| || || || ||30.87||.0001|
|All mild||58 (20.0)||58 (50.0)||54 (60.7)||4 (14.8)|| || |
|At least 1 moderate||30 (10.3)||30 (25.9)||24 (27.0)||6 (22.2)|| |
|At least 1 severe||28 (9.7)||28 (24.1)||11 (12.4)||17 (63.0)|| |
|Frequency of insomnia per week|| ||5.7 ± 2.1 (1–7)||5.5 ± 2.2 (1–7)||6.6 ± 1.0 (1–7)||2.54||.012|
|Friends|| ||12 (10.3)||7 (7.9)||5 (18.5)||2.54||.11|
|Parents|| ||16 (13.8)||4 (4.5)||12 (44.4)|| ||.0001|
|Social workers|| ||1 (0.9)||1 (1.1)||0 (0.0)|| ||.15|
|Healthcare providers|| ||1 (0.9)||0 (0.0)||1 (3.7)|| ||.15|
|Any type|| ||26 (22.4)||12 (13.5)||14 (51.9)||17.54||.0001|
|Listening to music|| ||11 (9.5)||7 (7.9)||4 (14.8)|| ||.28|
|Relaxation|| ||7 (6.0)||2 (2.2)||5 (18.5)|| ||.007|
|Reading|| ||5 (4.3)||3 (3.4)||2 (7.4)|| ||.33|
|Earlier bedtime|| ||5 (4.3)||3 (3.4)||2 (7.4)|| ||.33|
|Others|| ||12 (10.3)||8 (9.0)||4 (14.8)|| ||.47|
|Any type|| ||37 (31.9)||22 (24.7)||15 (55.6)||9.07||.003|
Based on clinical assessment, 27 of the 290 adolescents or 9.3% were diagnosed to have DSM-IV current insomnia. Primary insomnia, defined as insomnia unaccounted by medical, psychiatric, circadian rhythm or other sleep disorders, was the most common insomnia diagnosis (7.9%). The other insomnia diagnoses included insomnia associated with major depressive disorder (0.7%), insomnia associated with anxiety disorder (0.3%), and insomnia associated with comorbid depressive and anxiety disorder (0.3%). None of the subjects was diagnosed to have circadian rhythm sleep disorder, breathing-related sleep disorder, substance-related sleep disorder or sleep disorder due to general medical condition. The prevalence of DSM-IV insomnia and primary insomnia in boys was 9.4% and 7.9%, respectively; among girls, it was 9.2% and 8.0%, respectively. There was no significant difference in the prevalence of DSM-IV insomnia in subjects aged <15 years and those aged ≥15 years (7.1% vs. 11.8%, χ2 = 1.83, p = .18). The mean age at onset of DSM-IV insomnia was 13.1 years (SD = 2.5). Female students had significantly earlier age at onset than male students (mean ± SD: 12.2 ± 2.8 years vs. 14.2 ± 1.7 years, t = 2.16, p = .040). The average duration of DSM-IV insomnia was 19.5 months (SD = 20.7 months) and there was no significant difference in insomnia duration between girls and boys (mean ± SD: 24.5 ± 26.0 months vs. 13.4 ± 8.8 months, t = 1.42, p = .17).
DSM-IV insomnia vs. insomnia symptoms alone
Compared with subjects who met the quantitative criteria of insomnia symptoms, but not up to the level of DSM-IV insomnia, the DSM-IV insomnia group was significantly more likely to have two or more insomnia symptoms (48.1% vs. 14.6%, χ2 = 17.45, p = .0001), insomnia of at least moderate in severity (85.2% vs. 39.3%, χ2 = 30.87, p = .0001) and greater frequency of insomnia per week (6.6 ± 1.0 vs. 5.5 ± 2.2, t = 2.54, p = .012; Table 3). Adolescents with DSM-IV insomnia also had higher GHQ-12 score (4.5 ± 3.8 vs. 2.3 ± 2.9, F = 11.32, p = .0001) and were more likely to have a history of sleep paralysis (37.0% vs. 6.7%, χ2 = 27.12, p = .0001) than those with insomnia symptoms alone, but there was no significant difference in sociodemographic, school, lifestyle, and other clinical variables (Table 2).
Help-seeking behaviors and self-help strategies
Twenty-six of the 116 subjects with insomnia symptoms (22.4%) sought help from at least one source (Table 3). Only one (0.9%) consulted healthcare providers for insomnia, one (0.9%) talked to social workers, 16 (13.8%) to their parents, and 12 (10.3%) to their friends. The likelihood of any type of help-seeking was almost four times greater among subjects with DSM-IV insomnia than those with insomnia symptoms alone (51.9% vs. 13.5%, χ2 = 17.54, p = .0001) and was nearly three times greater in adolescents with moderate or severe insomnia than those with mild insomnia (32.8% vs. 12.1%, χ2 = 7.14, p = .008). Help-seeking behavior for insomnia was not significantly associated with age, gender, alcohol use, chronic medical problem, regular medical treatment, academic performance, severity of daytime sleepiness, GHQ-12 score, number, frequency, and duration of insomnia symptoms, and parental marital status, occupation, and educational attainment (all p > .05).
Of the 116 respondents with insomnia symptoms, 37 (31.9%) used at least one strategy to promote sleep (Table 3). Listening to music was utilized with the highest frequency (9.5%), followed by relaxation (6.0%), reading (4.3%), earlier bedtime (4.3%), drinking milk (3.4%), counting (2.6%), improving bedroom environment (1.7%), hot bath (0.9%), physical exercise (0.9%), and reduced daytime napping (0.9%). Only one subject used prescribed medicine (0.9%), one tried over-the-counter medication (0.9%), but none used alcohol as sleep aids. A higher proportion of subjects with DSM-IV insomnia tried to improve sleep by any kind of self-help than those with insomnia symptoms alone (55.6% vs. 24.7%, χ2 = 9.07, p = .003) and a higher likelihood for adolescents with moderate or severe insomnia symptoms trying to improve their sleep than subjects with mild insomnia symptoms (48.3% vs. 15.5%, χ2 = 14.33, p = .0001). No other demographic and clinical variables were significantly associated with the use of interventions for insomnia (all p > .05).
- Top of page
- Data analysis
This study extended our understanding of the epidemiology of insomnia among adolescents and was the first to examine their help-seeking behaviors and self-help strategies for insomnia. On the basis of clinician-administered interviews of 290 Chinese adolescents, we found that the prevalence of DSM-IV insomnia was similar to a study in the United States using similar procedure and diagnostic criteria (Johnson et al., 2006), with almost one in 10 adolescents having insomnia symptoms that were associated with significant distress or functional impairment. Our estimate of the prevalence of insomnia symptoms was 40%, which was higher than the range of 10%–30% reported in previous studies (Gradisar et al., 2011; Ohayon, 2002). One of the possible explanations was that our severity and frequency criteria of insomnia symptoms were less stringent than in previous studies. For example, adolescents with SOL ≥10 min every night or SOL ≥30 min twice a week were considered to have mild problem of difficulty initiating sleep. In addition, our definition of insomnia symptoms was based on quantitative criteria irrespective of subjective distress and daytime consequences. We found that only 27 of the 116 subjects with insomnia symptoms were diagnosed with DSM-IV insomnia; whereas the rest had no report of distress or impairment in daytime functioning. Our findings suggested that some participants who fulfilled our quantitative criteria might not regard themselves having sleep difficulties if they were asked in a questionnaire survey. The review by Ohayon (Ohayon, 2002) points out that the prevalence of insomnia will be lower when daytime consequences and dissatisfaction with sleep quality or quantity are included in the definition. It would be worthwhile to further examine the severity and frequency of insomnia that are associated with daytime consequences and sleep dissatisfaction across different age groups.
The prevalence of DSM-IV insomnia with and without comorbid psychiatric disorders was 1.4% and 7.9%, respectively. Thus, only 14.8% of DSM-IV insomnia had comorbid psychiatric disorders. The percentage was much lower when compared with the 52.8% reported in Johnson's study (Johnson et al., 2006). In view of the limited data available, additional studies are needed to determine the psychiatric comorbidity of insomnia in adolescents. We could not detect any age or gender differences in the rate of DSM-IV insomnia, but the onset was 2 years earlier in girls than in boys. Two previous studies using the DSM-IV diagnostic criteria could not detect any age difference in the prevalence of insomnia (Johnson et al., 2006; Roberts et al., 2006) and gender difference was partially supported in one study, of which higher prevalence of insomnia was observed in girls after menarche when compared with girls before menarche and boys (Johnson et al., 2006). Our other finding of an earlier onset of insomnia in girls was in contrary to Johnson's study, which found the opposite (Johnson et al., 2006). As girls have greater and faster bodily and hormonal changes in puberty than boys (Patton & Viner, 2007), it is expected that girls may have earlier onset of insomnia. Additional studies are needed to further examine the age and gender effect on the onset and prevalence of insomnia in adolescents.
We found that adolescents with DSM-IV insomnia had more severe and frequent sleep disturbances, higher GHQ-12 scores, and a history of sleep paralysis than those with insomnia symptoms alone. It was understandable that subjects with more severe and frequent insomnia symptoms would experience greater distress or impairment. The finding that adolescents with DSM-IV insomnia had higher GHQ-12 scores was consistent with the diagnosis of insomnia and that they were in greater psychological distress. Although cross-sectional studies have shown that sleep paralysis is associated with sleep disruption, anxiety, and depression (Otto et al., 2006; Szklo-Coxe, Young, Finn & Mignot, 2007), limited research have been conducted regarding the relevance of a history of sleep paralysis. Our finding that a history of sleep paralysis was associated with insomnia suggested that additional studies might be needed to examine whether childhood sleep paralysis was an early indicator of psychopathology during adolescence and adulthood.
In this study, about 22% of adolescents with insomnia sought help from other people about their sleep problem and they most often talked to their parents; followed up by friends, but formal help-seeking behaviors remained uncommon. About 32% of the subjects with insomnia symptoms tried various methods to improve sleep. Prescribed or over-the-counter medications have been used in less than 2% of the participants; whereas psychological and behavioral treatments have never been used. We were uncertain whether listening to music, reading, and other self-help strategies would help to improve sleep or actually delay proper treatments. Sleep-health education at schools might increase the students' awareness of sleep problems and knowledge about effective treatments. Apart from the severity of insomnia and diagnosis of DSM-IV insomnia, we could not detect any sociodemographic and personal factors that were related to help-seeking and self-management of insomnia in adolescents.
This study has several limitations that should be considered when interpreting the findings. First, the sample was not randomly selected from the general population and was small in size. There may be sampling error in our estimates of the prevalence rate. Analysis of the gender difference in the duration and age of onset of DSM-IV insomnia is also limited by the small sample size. The duration of insomnia and history of parasomnia was assessed retrospectively, with a potential of recall bias. The lack of information on pubertal development disallowed us to test whether there was an increased risk of insomnia after the onset of menses. Objective sleep measurement was not performed, although insomnia was regarded as subjective and the objective and subjective sleep measures did not usually correlate. Finally, we have not assessed nonrestorative sleep in the subjects because no consensus regarding its definition and measurement is currently available (Vernon, Dugar, Revicki, Treglia & Buysse, 2010).
Despite these limitations, our findings add to the database on the epidemiology of insomnia among adolescents. We found that many adolescents were troubled by sleep difficulties, but they seldom sought help or tried to improve their sleep. Given the negative consequences of untreated insomnia, education, and intervention programs on sleep disturbances should be specially designed and targeted at adolescents. Additional studies on the phenomenology of insomnia in adolescents are also important.