1

Insomnia and adolescence

Insomnia and Associated Risk Factors in Late Adolescence 1

Linda Li, Ivan W. Kelly & Bonnie .L.Janzen

University of Saskatchewan

Abstract.----Sleep is an essential component of health and well-being. This study

examines insomnia in a sub-sample of 15 to 19 year old participants (n=2,866) using data from
the Canadian Community Health Survey (CCHS): Mental Health and Well-being (Cycle 1.2). A
multivariate analysis showed insomnia to be significantly associated with the presence of a
chronic condition, selected mood disorders (12 months), in adolescents who are experiencing
“quite a bit” to “extreme” life stress, and in adolescents who were living in households other
than with both parents. Insomnia was not found to be significantly associated with sex, selected
anxiety disorder (12 months), heaving drinking, heavy cannabis use, nor in adolescents who were
only experiencing “some life stress”.

Sleep is a necessary component of an individual health and wellbeing, without sleep one’s
quality of life is, more often than not, severely compromised. For most individuals, sleep comes
easily and without disruption, and is often taken for granted, but for others, the simple task of
“falling” and staying asleep is difficult and labourious. Such is the case for individuals who
suffer from insomnia, where the desires for restful and restorative sleep, or sleep in general, often
eludes them. Insomnia taxes its’ sufferers, mentally and physically, and can cause a lot of stress
for individuals, and their ability to be productive members of society.
According to the data from a 2002 survey done by the Canadian Community Health Survey
(CCHS): Mental Health and Well-being (Cycle 1.2), about 13% of Canadians, aged 15 and over,
were found to have met the criteria for insomnia; that is, more than an estimated three million
Canadians have difficulty going to sleep or staying asleep most of or all of the time (Tjepkema,
2005).
Insomnia effects all sufferers negatively, and there is growing evidence the burden of
insomnia among adolescents is comparable to that of major psychiatric disorders such as mood
disorders, anxiety disorders, disruptive disorders, and substance abuse (Roberts et al., 2008a;
Roberts, Roberts, & Duong, 2008b). This is especially concerning considering rates of insomnia
have been found to be quite substantial among adolescents, with nearly one-fourth of adolescents

1
Linda Li and Ivan W. Kelly, Dept of Educational Psychology & Special Education, College of Education,
Bonnie L. Janzen, Dept of Community Health & Epidemiology, College of Medicine, University of Saskatchewan

1
 2

meeting the definition of clinical insomnia in Roberts et al.’s (2008b) study. Additionally,
adolescents have been shown to carry a significant chronic sleep debt throughout the school
week, and are eight times or more, likely than children to sleep 6 hours or less per night
(Ohayon, Roberts, Zulley, Smirne, & Priest, 2000). This inability to obtain adequate amount of
sleep for optimum functioning, especially during the week, have been associated with poor
concentration, irritability, decrease cognitive functioning, and other negative outcomes in
adolescents over time (Roberts et al., 2008a). Overall, the implications of insomnia and
disrupted sleep on an adolescent’s wellbeing and future functioning are profound.
The most common insomnia symptoms experienced by adolescents varied between studies,
but difficulty initiating sleep (Johnson et al., 2006) and non-restorative sleep (Roberts et al.,
2008a), were some of the most common symptoms noted among adolescents with insomnia. In
particular, chronic non-restorative sleep, followed by difficulty initiating sleep, pose the greatest
risks for insomnia reoccurrence with adolescents in Roberts et al.’s (2008b) study.
The association with age in adolescences’ sleep problems has been inconsistent though,
with most studies reporting no age effect (Ohayon et al., 2000; Patten et al., 2000; Roberts et al.,
2008a), while others reported some age effects (Johnson et al., 2006; Roberts et al., 2008a).

2
 3

The other most identified demographic risk factor for insomnia is sex. Survey data have
consistently demonstrated that complaints of insomnia and sleep dissatisfaction are more
prevalent among women than men (APA, 2000; LeBlanc et al., 2007; Ohayon, 2002). However,
not all studies support the findings that adolescents of the female sex are more likely than males
to have insomnia (Johnson et al. 2006; Patten et al. 2000).
Adolescents who have psychiatric issues also commonly experience sleep disorders, and
sleep disorders may even, in many cases, contribute significantly to daytime symptoms and
functioning ( Johnson et al. (2006)).
Similar to the results found in adult insomniacs, the prevalence of insomnia symptoms in
adolescents is quite high in individuals with an anxiety and/or mood disorder (Ohayon et al.
2000; Meltzer & Mindell, 2006). So although sleep disturbances, such as insomnia, can be
symptoms of either an anxiety and/or mood disorder, the consequences of disrupted or
insufficient sleep often intensify these disorders as well.
Sleep deprivation, over time, has also been found to increase risks for depression and lower
self-esteem in adolescents, but was not consistently found to increase the risk of poor academic
functioning (Fredriksen et al., 2004; Gregory & O’Connor’s, 2002), although most studies do
suggest sleep deprivation can adversely affect academic performance (e.g., Boschloo, etal 2013,
Carskadon, 1999; Meltzer & Mindell, 2006; O’Brien & Mindell, 2005).
Not surprisingly, adolescents who reported trouble sleeping were significantly correlated
with aggression, attention problems, anxiety/depression, and withdrawal behaviours (Coulombe
et al., 2011). However, when psychological co-morbidity was accounted for in Coulombe et
al.’s (2011) study, associations among sleep variables and psychological symptoms appear to
accompany specific sleep patterns (e.g., nightmares, sleeping more, etc.) and the associations
were of a relatively smaller magnitude. This is not surprising as psychological co-morbidity
often increase stress in an individual, and will, therefore, magnify problem behaviours or
psychological symptoms.
Trouble sleeping has also been associated with concurrent social problems at school,
problems with peers and parents, psychiatric disorders (e.g., depression, anxiety, conduct
problems, suicidal thoughts and attempts, etc.), somatic problems (e.g., excessive fatigue and
pain), and substance use (Johnson et al., 2006; Roberts et al., 2008a). Inadequate sleep and
increased sleep problems can also leads to increased daytime sleepiness, increased risk-taking
(Carskadon 1999; Johnson et al., 2006; O’Brien & Mindell, 2005), tardiness, missed school
days, moodiness, and irritability, with overall negative effect on growth, behaviour, emotional
regulation, mood, attention, learning, and memory (Meltzer & Mindell, 2006). Generally,
adolescents with sleep disturbances perceive themselves as having poorer health than their peers
who do not have sleep problems (Roberts et al., 2008a).
This study aims to: 1) assess the prevalence of insomnia among Canadian adolescents, 15 to
19 years of age; and 2) identify variables associated with insomnia in this population. Based on
the academic literature to date, it is hypothesized that insomnia will be significantly more
prevalent among adolescents of the female sex and among those reporting psychological and/or
physical health conditions.

DISCUSSION
Most adolescents in the present study were attending school and living with their parents.
Generally, the majority of the participants rated their mental and physical health as “excellent”,
“very good”, or “good”, were experiencing little to no stress, and found their sleep refreshing.

3
 4

The prevalence rate of insomnia was 9.5%, with an estimated 5% of participants reported having
taken medication to aid their sleep at least once in the past 12 months. Of the estimated 5% of
adolescents who took medication to aid their sleep, 28.3% of them were taking sleep medication
under professional supervision. Interestingly, when the use of sleep medication under
professional supervision was analyzed with sex (analysis is not shown), adolescent males were
less likely than their female counterparts to use sleep medication under professional supervision
(16.9% and 36.1% respectively).
As anticipated, a substantial proportion of adolescents who uses medication to help with
sleep had insomnia (28.7%), and only a small percentage of adolescent with insomnia felt
refreshed after awakening (7.4%). Adolescents who reported having insomnia were also
significantly more likely to describe their days as being “quite a bit” or “extremely” stressful
(19.7%). Overall, the odds of experiencing insomnia were found to be significantly greater
among adolescents with a chronic condition, selected mood disorder (12 month), in those
experiencing “quite a bit” to “extreme” stress, and in adolescents living in households other than
with both parents.
Contrary to some studies that found adolescents of the female sex to have significantly
higher prevalence rates and increased incidences of insomnia (Ohayon et al., 2000; Roberts et al.,
2008a), the sex of the adolescent was not significantly associated with insomnia in any of this
study’s analyses, which is congruent with findings from previous studies (e.g., Johnson et al.,
2006; Patten et al., 2000; Roberts et al., 2008a) that also found no significant sex interactions
with insomnia.
Given the well-documented co-morbidity between insomnia, psychiatric disorders, and
physical illnesses, the results of the current study lend further support to the existing literature.
Similar to most adult insomniacs who report having depression and/or anxiety disorders,
adolescents who reported having insomnia in this study were also significantly more likely to
report having an anxiety and/or mood disorders. The current study’s findings also support
evidence of mood disorders (e.g., depression, manic, dysthymia, etc.) being one of the strongest
risk factor for insomnia. Unlike selected anxiety disorders (12 month), which were not
significantly associated with insomnia in the multivariate analysis, selected mood disorder (12
month) had the strongest association with insomnia (OR = 3.16). This finding compliments
earlier research with adults that show mental health variables to be strongly associated with
insomnia.
Links between poor physical health and insomnia have also been repeatedly demonstrated to
be strongly associated, as many diseases and illnesses involve pain and/or distress that can
interfere with sleep. For instance, over 20% of adults with asthma, arthritis/rheumatism, back
problems, or diabetes have reported experiencing insomnia compared to around 12% of adults
who did not have these conditions (Tjepkema, 2005). Such findings are also supported by the
current study’s analysis, where adolescents with insomnia were found to be significantly more
likely to report having a chronic condition than adolescents without insomnia.

4
 5

Findings for life stress were also consistent with earlier research (Tjepkema, 2005) where
adolescents, in this study, who reported “quite a bit” or “extreme” stress were significantly
more likely to report having insomnia than those who report having “none” to “some” stress.
Such associations persisted even when physical and mental health, along with demographic,
socio-economic, and lifestyle factors were taken into consideration. Unfortunately, due to
amount of missing data (651 adolescents did not state their family income on the survey) family
income could not be examined in this study to assess whether lower income was associated with
insomnia in this population group, as too much data was missing from the data set to provide an
accurate analysis. Past studies, however, have had varying findings on the association of income
and education levels on insomnia, with some studies noting significant associations (Leblanc et
al., 2007; Roth, 2007) and others finding no significant associations between insomnia and
education level and/or income (Roberts et al., 2008a; Patten et al., 2000; Johnson et al., 2006).
The trend of insomniacs engaging in more alcohol use (usually as a mean to aid sleep)
(Daley, 2009) was also observed in this study. Though the association was not significant,
adolescents with insomnia were more likely to report engaging in more alcohol use than
adolescents who did not have insomnia. A less examined factor associated with insomnia –
cannabis use, was also found to be significantly associated with insomnia when all other
variables were not held constant. This finding may be related to the literatures examining
insomnia in individuals who were cigarette smokers. Cigarette smokers were significantly more
likely than non-smokers to report problems initiating sleep, problems staying asleep, daytime
sleepiness, depression, being in minor accidents, consuming higher daily intake of caffeine, and
were more likely to sleep for less than 6 hours per night than non-smokers (Ohayon, 2002).
There are several limitations in this study on the contribution of various variables to the
prevalence of insomnia. Particularly, the Canadian Community Health Survey (CCHS): Mental
Health and Well-being survey often employs single questions to covered a wide range of socio-
demographic, behavioural, health and mental health variables. This is particularly problematic
when it is attempting to examine variables like insomnia and mental health conditions (e.g.,
depression, anxiety, etc.), which are considered to have multiple criteria and lists of individual
symptoms. So, although the WMH-CIDI questions and algorithms were operationalized to meet
the needs of CCHS 1.2, the use of this measure to assess various DSM disorders (e.g., anxiety,
depression, etc.) could not be used, nor assumed, to measure all aspects associated with the
DSM-IV classification and definitions of mental disorders. In particular, the distinction between
chronic insomnia and occasional transient/acute insomnia (e.g., insomnia symptoms due to
current, but short-lived, stressors) cannot be made because the questions asked in the CCHS do
not precisely measure the duration or intensity of the insomnia episodes. This meant primary,
secondary (e.g., mental disorders, medical condition, breathing disorders during sleep, other
sleep disorders, etc.), and self-induced (e.g., life style; use, abuse or withdrawal of substances;
etc.) insomnia could not be distinguished nor was it possible to separate between incident and
prevalent cases. Therefore, important differences between these types of insomnia, which have
been shown to have different risks and outcomes, could not be ascertained.
The inability to distinguish between those who suffer from insomnia – a sleep disorder,
from those who suffer from insomnia as a secondary symptom of a mental or health disorder has
been a concern other researchers have raised in the past in regards to epidemiological studies.
Due to the lack of differentiation between primary insomnia and secondary insomnia,
interpretations of the results (i.e., deductions of prevalence rates for primary insomnia) should be
done with caution. In short, the CCHS measure used in this study primarily examines the

5
 6

presence of insomnia symptoms and their frequency, but whether significant physical and mental
health problems are contributing confounding factors in sleep status cannot be determined. This
lack of differentiation between the two types of insomnia can be problematic as psychological
distress and quality of life impairments are often associated with increases in insomnia frequency
and severity.
Another, and quite common limitation, to studies done with data from large surveys
(including this study) is that they are often cross-sectional in nature, which mean the data set, and
subsequent analysis, represents only a “snapshot” of a given population at the time of data
collection. Given this limitation, we cannot determine which variables are predisposing factors,
precipitating factors, or consequences, making it difficult to trace causal pathways between
insomnia and other associated variables of interest.
The CCHS data used in this study were also based on self-reports and answers were not
validated by an independent source. For example, self-reported weight and height are known to
underestimate the prevalence of overweight and obesity, and measures of physical activities are
usually underestimated, as respondents do not account for activity at work or while doing
household chores (Mawani & Gilmour, 2010).
Despite these limitations, this study offers a number of strengths and contributions to the
literature on insomnia. This is the first nationally representative study to examine the
associations between insomnia and select self-rated health measures in adolescent Canadians,
using the latest measures available from Statistics Canada. It supports previous research in
demonstrating that, unlike their adult counterparts, gender in insomnia is not a significant factor
in adolescent insomnia – which is congruent with a number of existing studies (e.g., Johnson et
al., 2006; Patten et al., 2000). The large national population based survey and the multivariate
approach utilized in this study also adds strength to this analysis, by being representative of the
adolescents group in Canada, with strong and consistent associations across a wide range of
health variables. Such findings may provide potentially useful indicators for monitoring
adolescents sleep patterns and general health (physical and mental).
Even with the noted limitations, a number of significant associations did emerge from the
study’s analysis. This study provided a quantified measure of the most current prevalence rate
of adolescent insomnia in the Canadian population with the most recent data set from CCHS.
This study also examined whether variables (e.g., sex, mental health, and chronic health) that
were often associated with adult insomniacs were also as pronounced in adolescents. This study
analysis supported some well-established variables that are associated with adult insomniacs –
such as chronic and mental health conditions – are also associated in adolescents with insomnia.
The study also supported some previous findings that there are no gender differences in
adolescent males and females in insomnia prevalence, while concurrently at variance with other
findings that have found a significant difference in the sex of the adolescent and insomnia.