HEALTH PROMOTION INTERNATIONAL
2
© Oxford University Press 1999
Perceived sleep quality and its precursors
in adolescents
JORMA TYNJÄLÄ, LASSE KANNAS, ESKO LEVÄLAHTI and
RAILI VÄLIMAA
University of Jyväskylä, Department of Health Sciences, P.O. Box 35, FIN-40351 Jyväskylä, Finland
SUMMARY
This study investigated perceived sleep quality among 11-,
13- and 15-year-old Finnish adolescents (n = 4187).
Additionally, associations of selected behavioural, social
and psychological factors with subjective sleep quality were
examined among 15-year-olds. This study is part of an in-
ternational, WHO-co-ordinated survey of school children’s
health and lifestyle (the HBSC Study). In Finland, research
data represented the whole country. The data were collected
during March–May 1994. Pupils responded anonymously
to a standardized questionnaire during a class period.
About every 10th adolescent felt that their sleep quality was
at most satisfactory, about 30% of pupils had had dif-
ficulties in falling asleep and almost every fifth adolescent
reported nocturnal awakenings every week. Thus, a large
proportion of pupils in every classroom has a weakened
ability to concentrate on school work or other activities.
Among 15-year-old boys, a good home atmosphere was the
most important contributing factor to good perceived sleep
Key words: adolescents; health behaviour; health promotion; perceived sleep quality
INTRODUCTION
Sleep quality is an important indicator of chil-
dren’s and adolescents’ health and well-being.
Poor perceived sleep quality is associated with
poor reported health, poor physical fitness and
numerous psychosomatic symptoms. Sleep dif-
ficulties and poor sleep quality may also be signs
of stress factors in living environment or indicators
of a health-compromising lifestyle. (Price et al.,
1978; Rimpelä and Rimpelä, 1983; Weydahl,
1991a; Mahon, 1994, 1995; King et al., 1996;
Pilcher et al., 1997; Vignau et al., 1997).
Vol. 14, No.
Printed in Great Britain
quality. A health-promotive lifestyle (good sleep hygiene
and infrequent use of addictive substances) and good self-
perception also had significant correlation with good
perceived sleep quality. In 15-year-old girls, a good home
atmosphere, good self-perception and health-promotive
habits played an equally important role in associations with
subjective sleep quality. Physical activity in leisure time had
minor but significant correlation with perceived sleep quality
among girls. In both sexes, perceived home atmosphere
had a significant association with all the factors that had
correlated significantly with subjective sleep quality. Results
indicate that a good home atmosphere, a health-promotive
lifestyle and good self-perception are important constituents
of good sleep among 15-year-olds. Good and refreshing sleep
is one of the constituents for general well-being among
adolescents. That is why these issues provide a challenge for
health promotion and health education.
Few studies have sleep quality as the focus of
major interest in normal adolescent populations
(Price et al., 1978; Kirmil-Gray et al., 1984; Levy
et al., 1986; Hawkins and Shaw, 1992; Andrade
et al., 1993; Manni et al., 1997; Vignau et al.,
1997). A need for a large-scale study on this issue
has been expressed in earlier studies (Price et al.,
1978; Kirmil-Gray et al., 1984; Levy et al., 1986;
Zammit, 1988; Lexcen and Hicks, 1993; Tynjälä
and Kannas, 1993; see also Hyyppä, 1991). Study-
ing perceived sleep quality in normal populations
155
156 J. Tynjälä et al.
will help us to determine if we need to be con-
cerned about sleep quality in younger age groups
(Price et al., 1978).
Perceived sleep quality is described in many
studies as difficulty in falling asleep, sleep onset
latency, difficulty in maintaining sleep, awaken-
ings at night, calmness of sleep, premature final
morning awakening, how refreshed a person
feels after sleep and a person’s own view of sleep
quality (Monroe, 1967; Price et al., 1978; Healey
et al., 1981; Lugaresi et al., 1983; Kirmil-Gray
et al., 1984; Levy et al., 1986; Strauch and Meier,
1988; Andrade et al., 1993; Lexcen and Hicks,
1993; Tynjälä et al., 1993; Tynjälä and Kannas,
1993; Åkerstedt et al., 1994; Saarenpää-Heikkilä
et al., 1995). On the basis of these indicators,
persons are often grouped as, e.g. poor sleepers,
occasional poor sleepers and good sleepers. In
these studies, the proportion of adolescents with
no sleep problems has varied between 40% and
70%, whereas about 10% have been categorized
as poor sleepers with chronic and severe sleep
disturbances.
Investigations of the above-mentioned in-
dicators of sleep quality among adolescents have
revealed that the majority of adolescents sleep
well or very well (Andrade et al., 1993; Tynjälä,
1993; Tynjälä and Kannas, 1993). Under 10% of
adolescents have reported difficulties in falling
asleep or nocturnal awakenings every or almost
every night (Morrison et al., 1992; Andrade et al.,
1993; Tynjälä, 1993; Saarenpää-Heikkilä et al.,
1995; see also Aro et al., 1987). Differences in the
indicators of subjective sleep quality between
age groups have been for the most part not sig-
nificant (Tynjälä and Kannas, 1993; see also
Strauch and Meier, 1988; Saarenpää-Heikkilä
et al., 1995). Moreover, sex differences have been
rare or have not existed (Fischer and Wilson,
1987; Strauch and Meier, 1988; Hawkins and
Shaw, 1992; Tynjälä, 1993; Tynjälä and Kannas,
1993; Yarcheski and Mahon, 1994; Saarenpää-
Heikkilä et al., 1995). However, in some studies
girls have reported more sleep difficulties than
boys (e.g. Rimpelä and Rimpelä, 1983; Kirmil-
Gray et al., 1984; Aro et al., 1987; King et al.,
1996; Vignau et al., 1997).
There are many studies that provide evidence
that poor sleep quality is associated, among other
things, with problems in social relationships,
problems with handling troubles, problems at
school, disturbances in sleeping environment and
psychological states, e.g. anxiety, depression,
tension and worry (Price et al., 1978; Kahn et al.,
1989; Weydahl, 1991a; Mahon, 1994, 1995;
Pilcher et al., 1997; Vignau et al., 1997). Positive
self-esteem or self-concept is more typical of
persons with good sleep quality (e.g. Price et al.,
1978; Healey et al., 1981; Kirmil-Gray et al., 1984;
see also Hyyppä et al., 1991).
Unhealthy behaviours tend to cluster (Lexcen
and Hicks, 1993; Aarø et al., 1995), and therefore
it is important to also study sleep quality from
the viewpoint of adolescents’ lifestyle. Studies
have indicated that frequent smoking, heavy
use of alcohol and frequent use of coffee or caf-
feined beverages are associated with poor
perceived sleep quality (e.g. Price et al., 1978;
Rimpelä and Rimpelä, 1983; Hallman et al., 1992;
Rahkonen et al., 1992; Lexcen and Hicks,
1993; Phillips and Danner, 1995; Vignau et al.,
1997). Physical activity has been found to have
positive effects and lack of physical activity
negative effects on subjective sleep quality (Price
et al., 1978; Shapiro et al., 1984; see also Weydahl,
1991b). Sleeping habits, e.g. late and irregular
bedtime, short sleeping time, large sleeping time
differences between the week and weekend, and
long daytime naps are associated with poor
perceived sleep quality (Kirmil-Gray et al., 1984;
Carskadon, 1990; Ferber, 1990; see also Rimpelä
and Rimpelä, 1983).
The first research task in the present study
was to describe the self-reports of 11-, 13- and
15-year-old Finnish boys and girls on how well
they sleep (how often they have difficulties in
falling asleep, how often they wake up at night
and how well in general they sleep). The second
research task was to investigate what factors,
e.g. self-perception, social relationships, addictive
behaviour, sleep hygiene and physical activity,
are associated with perceived sleep quality in
15-year-old Finns.
METHODS
This study is part of an international, WHO-
co-ordinated survey of school children’s health
behaviour—the HBSC Study—which describes
the perceived health and lifestyle of 11-, 13- and
15-year-old adolescents from a health promotion
viewpoint.
In Finland, the Institute for Educational
Research in Jyväskylä was responsible for the
sampling procedure. Samples were chosen from
the Finnish school register by using a special
sampling program. The sample frame was the

number of Finnish-speaking pupils in each class
level. This computer program picked the schools
using cluster sampling in such a way that the size
of schools was taken into consideration (prob-
ability proportional to size of school—PPS).
Within the school, the class was randomly
selected. In only a few cases was more than one
class per school selected in the sample. The strata
of schools were based on province division. The
other basis of stratum was the division between
urban and rural communes. In summary, the
research data represented the Finnish-speaking
part of the country well (94% of the whole popu-
lation) in the 11-, 13- and 15-year age groups
(Tynjälä and Törmäkangas, 1995).
In Finland the age groups represent pupils in
the 5th, 7th and 9th grades of the compulsory
comprehensive school. Pupils responded anony-
mously to a standardized questionnaire during a
class period with a teacher overseeing the pro-
cess. Data were collected in March–May 1994.
The final sample of the cleaned (Aarø et al.,
1994) data consisted of 4187 pupils, 869 of which
were 11-year-old boys (mean age was 11 years
and 9 months), 631 were 13-year-old boys (mean
age 13 years and 10 months) and 576 were 15-
year-old boys (mean age 15 years and 9 months);
respective figures for girls were 845 (mean age 11
years and 9 months), 648 (mean age 13 years and
9 months) and 618 (mean age 15 years and 9
months). All 206 schools included in the sample
participated in the survey. Of all the pupils in the
sample, 93% responded to the questionnaire.
Non-responses, 7% of the sample, consisted of
pupils absent from school on the data collection
day. The data have been described in more detail
elsewhere (Tynjälä and Törmäkangas, 1995; King
et al., 1996).
Perceived sleep quality was measured by
inquiring how well adolescents sleep in general,
how often they have difficulties in falling asleep
and how often they wake up during the night.
The internal consistency of the sum index of these
variables was tested using Cronbach’s Alpha,
and it was 0.66. Frequency distributions of these
variables were analysed by pupil’s sex and age
using cross-tabulations and Chi-square test. A
more detailed description of these questions and
variables is presented in Appendix 1.
The associations between perceived sleep
quality and 28 variables were investigated among
15-year-olds. These 28 variables were selected on
the basis of earlier studies. After investigating
the correlation structure and using information
Perceived sleep quality 157
from earlier studies (e.g. Tynjälä et al., 1997),
these variables were thought to formulate the
following factors: sleep hygiene (bedtime during
school week, regularity of bedtime, sleep time
during school week); addictive behaviour (cigar-
ettes per week, times being drunk, frequency of
beer drinking); leisure time physical activity
(frequency and time spent on physical activity
per week, activity in sport club); sedentary behav-
iour (hours of watching TV, hours of watching
VCR, number of evenings a week spent with
friends); self-perception (index of self-esteem
[Cronbach’s Alpha was 0.83; the index was a modi-
fication of Rosenberg, 1985; see also Lintunen
et al., 1995], feeling of self-confidence, feeling of
helplessness); index of problems at school (six
items, Cronbach’s Alpha was 0.68); social rela-
tionships (feelings of loneliness, perceived home
atmosphere, number of close friends, index of
the perceived importance to family members and
close friends—six items with Cronbach’s Alpha
= 0.46); class and school atmosphere (encour-
agement by teachers, how interested teachers
were in pupils, did pupils enjoy being together
in the class, how friendly and helpful pupils felt
the other pupils were in the class, frequency of
being bullied, how safe pupils felt at school);
attitude toward school and school work (opinion
about going to school, opinion of how good
he/she is in school, opinion of how well he/she
knows what kind of things he/she is good at in
school).
Among 15-year-old boys and girls, factors’ cor-
relation structure was analysed using structural
equation models within the framework of LISREL
(version 8.12a). The polychoric correlations
needed in LISREL were calculated using
PRELIS [version 2.12a] (Jöreskog and Sörbom,
1993). These analyses were carried out separately
for boys and girls. The LISREL analyses were
performed using the generally weighted least
squares (WLS) estimation method because the
analyses were based on polychoric correlations.
In addition, the variables were not multinormally
distributed and the data comprised mostly variables
with an ordinal scale. The analyses simultaneously
included both the confirmatory factor analysis
and structural equation modelling. The equivalence
of the regression coefficients was statistically tested
with nested models that are based on sequential
Chi-square tests. If the Chi-square test was not
significant, the regression coefficients were equal.
In the reverse case, the coefficients were not
equal.
158 J. Tynjälä et al.
Table 1: General sleep quality by age and sex

How well do you usually sleep? Boys Girls

11 years 13 years 15 years 11 years 13 years 15 years

% % % % % %

Very well 59 54 60 62 56 57
Quite well 33 35 33 32 37 34
Satisfactorily 7 10 5 6 5 6
Quite or very badly 1 1 2 1 2 3
Total 100 100 100 100 100 100
(n) (866) (630) (559) (841) (646) (609)
Between age groups, p value 0.049 NS
Between sex, p value NS 0.013 NS

NS = No statistical significant difference.

Table 2: Awakenings at night during the 6 months preceding the enquiry by age and sex

Frequency of nocturnal awakenings Boys Girls

11 years 13 years 15 years 11 years 13 years 15 years

About every night 4 3 2 3 2 3

More than once a week 7 6 4 5 6 5
About every week 10 8 8 11 10 10
About every month 19 22 18 18 19 20
Rarely or never 60 61 69 62 63 63
Total 100 100 100 100 100 100
(n) (860) (627) (573) (842) (646) (614)
Between age groups, p value 0.007 NS
Between sex, p value NS NS NS

NS = No statistical significant difference.

RESULTS week. There were no age group differences among

Perceived sleep quality in 11-, 13- and
15-year-olds
Most adolescents slept well. About 60% felt
that they slept very well and every third rather
well. There were only slight sex or age group
differences in general sleep quality (Table 1).
Every other adolescent had rarely or never
experienced difficulties in falling asleep during
the 6 months preceding the survey. However,
about 30% had had difficulties in initiating sleep
at least once a week and 5% almost every night.
There were no age group differences and only
a small (p = 0.027) sex group difference in the
youngest age group where the problem was
slightly more typical among boys. Over 60% of
the pupils did not report awakenings at night
during the 6 months preceding the survey, but
almost 20% had had nocturnal awakenings every
girls in nightly awakenings; among boys they
seemed to decrease with age (Table 2).
Perceived sleep quality and its correlates
among 15-year-olds
Structural equation models were used to describe
associations of selected behavioural, social and
personal factors with perceived sleep quality.
The models include only those variables and
factors that were accepted for the final statistical
models, which may be characterized more as of
an inter-relationship type than explanatory type.
The analyses were executed separately for boys
and girls. The factor structure is described in
Appendix 1. The polychoric correlation coefficients
of the variables that were the basis for the final
models are presented in Appendices 2 and 3.
Factors’ correlation structure used in the final

models is presented in Appendix 4. Only graph-
ically simplified versions of the structural equation
models are presented in this report: comprehen-
sive graphical presentations of the models are
available from the author.
A model for boys
Among boys, home atmosphere, self-perception
and health habits (second order factor composed
of sleep hygiene factor and addictive behaviour
factor) explained 22% of the variance in per-
ceived sleep quality. Subjective sleep quality
together with home atmosphere explained 11%
of the variance of self-perception and 14% of the
variance of health habits. The second order factor,
health habits, was formulated in the structural
equation model, because the first order factors—
sleep hygiene and addictive behaviour—correlated
strongly with each other (r = 0.53). This cor-
relation was higher than their correlation with
the sleep quality factor (r = 0.24 and 0.22). Home
atmosphere, self-perception and health habits
were not equally important in associations with
subjective sleep quality (χ2 = 6.760, d.f. = 2,
p = 0.034), but self-perception and health habits
were equally important (χ2 = 0.090, d.f. = 1, NS).
It follows that home atmosphere was the most
significant element concerning subjective sleep
quality. For the factor model with health habits
factor, but without structural equations, the χ2
value was 59.096 (d.f. = 49, p = 0.153). Tests
for the equivalence of sleep hygiene factor and
addictive behaviour factor in this second order
factor were done against this model: these two
factors were not equally important (χ2 = 14.162,
d.f. = 1, p , 0.001—test of equivalence of
residuals of the factors). Based on the previous
Chi-square test, it follows that the sleep hygiene
factor was more important than the addictive
behaviour factor in this second order factor.
Reliability of the variables (R2) used in the
model varied from 0.51 to 0.82. According to the
goodness-of-fit statistics of the model, the fit to
the data was good. The goodness of the statistical
model is presented in Figure 1.
A model for girls
Among girls, home atmosphere, self-perception,
health habits (second order factor composed
of sleep hygiene factor and addictive behaviour
factor) and leisure time physical activity ex-
plained 30% of the variance of perceived
Perceived sleep quality 159
sleep quality. Subjective sleep quality together
with home atmosphere and leisure time physical
activity explained 27% of the variance of self-
perception. Perceived sleep quality, home atmos-
phere and self-perception explained 8% of the
variance of leisure time physical activity. In add-
ition, subjective sleep quality and home atmos-
phere explained 29% of the variance of health
habits. Self-perception and leisure time physical
activity also had a significant inter-relationship
with each other. Regarding the association of
home atmosphere, self-perception, health habits
and leisure time physical activity with perceived
sleep quality, the investigation of the equality
of the regression coefficients revealed that the
coefficients were not of equal importance (χ2 =
9.290, d.f. = 3, p = 0.002). However, the first three
were equally important (χ2 = 0.140, d.f. = 2, NS).
It follows that leisure time physical activity
proved to be the least important in associations
with subjective sleep quality (Figure 2).
The second order factor, health habits, was
formulated in the structural equation model,
because the first order factors—sleep hygiene
and addictive behaviour—correlated with each
other (r = 0.33). This correlation was higher than
the correlation between addictive behaviour
factor and sleep quality factor (r = 0.25). The cor-
relation between sleep hygiene factor and sleep
quality factor was 0.36. For the factor model with
health habits factor, but without structural
equations, the χ2 value was 109.421 (d.f. = 77,
p = 0.009). Tests for the equivalence of the sleep
hygiene factor and addictive behaviour factor were
performed against this model, and they revealed
that these two factors were equally important
in this second order factor (χ2 = 0.801, d.f. = 1,
NS—test of equivalence of residuals of the
factors, Figure 2).
The reliability of the variables (R2) used in the
model varied from 0.50 to 0.96. In a few cases,
residuals of variables correlated positively with
each other. This indicates that the correlation
structure of the variables was quite complicated
(see Appendix 3). According to the goodness-
of-fit statistics of the model, the fit to the data
was not as good as in the boys’ model. The Chi-
square value was relatively high and its p value
significant. However, the normed fit index (NFI)
was very high (0.97) and the number of cases (n)
was greater than the critical number of cases
(CN). These indicated that the large sample size
(n = 551) had had an effect on the Chi-square
value and that conclusions about the goodness
160 J. Tynjälä et al.

Fig. 1: Factors associated with perceived sleep quality among 15-year-old boys: structural equation model.
Variables are presented in boxes, factors in ellipses; the R2 value above each factor describes how much of the
variance of that factor could be explained by the other factors (or variables); small arrows outside boxes or
ellipses describe residuals (actual numbers were intentionally dropped from the figure); the rest of the
abbreviations are described below the figure.

of the statistical model should not be made on
the basis of the Chi-square test. All the other
parameters of goodness of the model were
statistically adequate (Figure 2).
DISCUSSION
Åkerstedt et al. (1994) concluded in their study
that the index of sleep quality should include
the following measures: sleep quality; calmness
of sleep; easiness of falling asleep; sleeping
throughout the night; and refreshness after sleep.
They also concluded that sleep quality is mainly
a question of sleep continuity. The questions
about sleep quality used in this study covered
most of these criteria. Adolescents’ own percep-
tions indicated that about every tenth person felt
that their sleep quality was at most satisfactory.
About 30% of pupils had also had difficulties in
falling asleep and almost every fifth adolescent
reported nocturnal awakenings every week.
These prevalences were relatively high. Accord-
ing to other studies (e.g. Tynjälä et al., 1997), at
least every third pupil was tired on almost every
school morning or felt tired more often than
once a week. It may be concluded that a large
proportion of pupils in every classroom has a
weakened ability to concentrate on school work
or other activities. Results on prevalences gained
in this study are in line with earlier studies (e.g.
Andrade et al., 1993; Tynjälä, 1993; Saarenpää-
Heikkilä et al., 1995), but prevalences cannot,
however, be directly compared with each other
because of, e.g. differences in operationalizations
of sleep variables. Among 15-year-old boys, home
atmosphere seemed to be a very important
element in relation to perceived sleep quality.

Fig. 2: Factors associated with perceived sleep quality among 15-year-old girls: structural equation model.
Explanations: see caption for Figure 1.
Among girls of the same age, self-perception,
health habits and home atmosphere were equally
important in associations with subjective sleep
quality.
Interpretation of models
The association between self-esteem or self-
image and sleep quality indicators has also been
reported in many earlier studies (e.g. Price et al.,
1978; Kirmil-Gray et al., 1984; Turtle et al., 1996).
In this study, the associations between home
atmosphere, pupils’ self-perception and perceived
sleep quality indicated that good relationships at
home form an important ground for positive self-
perception and health-promotive habits which
all correlated with good subjective sleep quality.
The results support the idea that family support in
general has salutary effects on adolescents’ psycho-
logical well-being (Reicher, 1993; Vilhjalmsson,
Perceived sleep quality 161
1994). The results in this study are in line with
the finding that poor sleepers are less satisfied
in their relationships with their parents (Healey
et al., 1981). Adolescents’ complaints of poor sleep
may even be regarded as a possibly meaningful
and sensitive sign of severe family or personal
disruption. Not only family relations but also
parents’ health status reflects perceived sleep
quality. If adolescents, especially girls, are worried
about their parents’ health and well-being, it may
manifest itself as poor perceived sleep quality
(Vignau et al., 1997).
Physical activity was a significant factor only
among girls. This finding is not in concordance
with the results obtained by Weydahl (1991a).
This may result from differences in questions
concerning subjective sleep quality and physical
activity. Among boys, health habits—composed
of sleep hygiene and addictive behaviour—and
self-perception were equally important, but not
162 J. Tynjälä et al.
as important as home atmosphere in associations
with subjective sleep quality. Sleep hygiene and
addictive behaviour correlated strongly with
each other among boys, but this correlation was
also significant in girls. This association has been
found in earlier studies in Finland and elsewhere
among adolescents (e.g. Rimpelä and Rimpelä,
1983; Tynjälä et al., 1993, 1997).
One purpose of this study was to investigate
how latent behavioural and other factors are
associated with subjective sleep quality. In many
earlier studies, the focus has been on the re-
lationships between single manifest (observable)
variables and perceived sleep quality, therefore
direct comparisons with earlier studies cannot
be made. However, it can be concluded that the
results of this study support earlier studies, e.g.
frequent use of addictive substances (drinking
and smoking in this study) had a negative effect
on perceived sleep quality (e.g. Lexcen and
Hicks, 1993; Phillips and Danner, 1995; Vignau
et al., 1997). This seems to be typical in adolescent
years, but these associations have been weak or
did not exist in the adult population (Palmer
et al., 1980; Urponen et al., 1988; Vuori et al., 1988).
This study focused on the associations of
selected behavioural, social and psychological
factors with adolescents’ perceived sleep quality.
Exclusion of many relevant indicators of subject-
ive health was done intentionally to emphasize
the main interest, subjective sleep quality. Initially,
the variables in the models were selected on the
basis of earlier studies. During the formulation
of structural equation models, many originally
significant variables—on the basis of initial cor-
relations with the indicators of subjective sleep
quality variables—were dropped from the models.
Exclusions of these variables were made on a
statistical basis: if the variable was no longer
significant in the model it was dropped out. As a
result, the models describe the associations
rather efficiently. The home atmosphere variable
in the statistical models could be described as
a natural ‘background’ factor because it is an
important ground for all later developments in
children and adolescents. Home atmosphere also
seemed to have a great impact on adolescents’
self-perception and health habits in general.
Manni et al. (1997) found in their study that
gender was an important factor in associations
with chronic poor sleep. There were more
females than males among poor sleepers in their
study. However, in this study models were
compiled separately for both sexes. The models
resemble each other but the weights of different
elements are not the same.
The question of missing data arises when using
many variables in the analyses. This was the case
in the structural equation models in 15-year-olds.
A comparison of respondents who were included
in the models (no missing data in any of the
variables in the model) against those respondents
who had missing data in one or more variables
(they were automatically excluded from the
models) revealed only one difference between
the groups: the ‘excluded’ group—17% of boys
and 10% of girls—went to bed later than the
group included in the models. No other significant
differences were found. It may be concluded that
the results of the structural equation models
represent the whole 15-year-old study group
adequately well.
Future research challenges
In future research, the role of family relation-
ships, social support in the family, the health status
of parents and siblings and their associations
with adolescents’ sleep and subjective health in
general should be covered in more detail, not
only in cross-sectional studies but also in follow-
up studies (e.g. Välimaa et al., 1993; Vilhjalmsson,
1994; Vignau et al., 1997). In this study, only one
question measured family relationships, but even
that proved to be of great importance concerning
adolescents’ subjective sleep quality. Good
family relationships probably have a ‘protective’
effect in relation to adolescents’ health and in a
broader sense to their well-being (see Nestmann
and Hurrelmann, 1994). Aro (1988) indicated
in her study that children from both divorced
and discordant families reported more distress
symptoms, more abundant use and in boys more
delinquent acts than children from intact families.
Also, severe illness of a family member was highly
associated with a higher number of symptoms,
and with the greatest increase in symptoms in
both sexes. The effect of life events involving
illness or loss was greater in boys than girls (Aro,
1988; see also Rutter, 1985). The results gained in
this study emphasize the importance of family
relationships to adolescents’ health and well-
being. An interesting future topic would be to
investigate which matters constitute a good night’s
sleep from a wider perspective (Zammit, 1988).
Few studies have investigated the association
between sleep variables and perceived health
status in adolescence. The results of these studies

have been somewhat contradictory. On one
hand, studies have suggested that sleep quality
correlates more strongly than sleep quantity with
health (Pilcher et al., 1997). On the other hand, in
some studies sleep disturbance is associated with
perceived health status only in middle adolescence
(15–17-year-olds), but not in the early (12–14-
year-olds) or late phase of adolescence (18–21-
year-olds) [Yarcheski and Mahon, 1994; Mahon,
1995]. Thus, further research is required to clarify
the association between sleep and perceived
health in adolescence.
According to Mahon (1995), better subjective
measures of sleep for adolescents are needed:
an effort should be made to develop reliable and
valid measures of the experience of sleep during
adolescent development. Nurses and other health
care professionals should routinely collect data
on adolescents’ alcohol and drug use because
they have been known or thought to influence
sleep (e.g. Lexcen and Hicks, 1993; Phillips and
Danner, 1995; Tynjälä et al., 1997). Health care
professionals should create a knowledge base
regarding the extent to which sleep contributes
to the health and well-being of adolescents
(Mahon, 1995).
Some practical considerations
What kind of practical considerations should be
made in school health education, school health
care staff training and parent counselling?
Mahon (1995) suggests the following issues to be
discussed to promote good sleep in adolescence:
pupils should sleep in quiet, comfortable and
well-ventilated rooms. They should avoid exces-
sive expenditure of energy in the evening hours
immediately preceding sleep, and if needed, pupils
should be instructed in the use of relaxation
techniques.
Poor sleep quality, poor sleep hygiene, fre-
quent use of alcohol, regular smoking and poor
self-perception identified in this study provide an
important challenge to health promotion among
school children. Although adolescents’ sleep
quality is clearly related to home atmosphere,
the role of the school is also important. In order
to deal with these sleep-related problems that
affect children’s health, issues about sleep and
rest need to be included in the health education
curriculum in schools. Intensive cooperation
and a good functioning network between school
and home will also be needed to foster dialogue
between teachers, school nurses and parents.
Perceived sleep quality 163
ACKNOWLEDGEMENTS
This study was financially supported by the Yrjö
Jahnsson foundation and the Juho Vainio
foundation. The authors wish to thank Phyllis
Wacker, PhD and Ms Kirsi Talvitie for correcting
the language.
Address for correspondence:
Jorma Tynjälä
University of Jyväskylä
Department of Health Sciences
P.O. Box 35
FIN-40351 Jyväskylä
Finland
REFERENCES
Aarø, L. E., Wold, B. and Smith, C. (1994) Health Behaviour
in School-aged Children. WHO Cross-National Survey
(HBSC). Research Protocol for the 1993–94 Study.
Research Center for Health Promotion, University of
Bergen.
Aarø, L. E., Laberg, J. C. and Wold, B. (1995) Health
behaviours among adolescents: towards a hypothesis of
two dimensions. Health Education Research, 10, 83–93.
Åkerstedt, T., Hume, K., Minors, D. and Waterhouse, J.
(1994) The subjective meaning of good sleep, an intra-
individual approach using the Karolinska Sleep Diary.
Perceptual and Motor Skills, 79, 287–296.
Andrade, M. M. M., Benedito-Silva, A. A., Domenice, S.,
Arnhold, I. J. P. and Menna-Barreto, L. (1993) Sleep
characteristics of adolescents: a longitudinal study.
Journal of Adolescent Health, 14, 401–406.
Aro, H., Paronen, O. and Aro, S. (1987) Psychosomatic
symptoms among 14–16-year-old Finnish adolescents.
Social Psychiatry, 22, 171–176.
Aro, H. (1988) Stress, development and psychosomatic
symptoms in adolescence. A study of 14 to 16-year-old
school children. Acta Universitatis Tamperensis, Ser. A,
Vol. 242. University of Tampere, Tampere (Academic
dissertation).
Carskadon, M. A. (1990) Patterns of sleep and sleepiness in
adolescents. Pediatrician, 17, 5–12.
Ferber, R. (1990) Sleep schedule-dependent causes of
insomnia and sleepiness in middle childhood and
adolescents. Pediatrician, 17, 13–20.
Fischer, B. E. and Wilson, A. E. (1987) Selected sleep
disturbances in school children reported by parents:
prevalence, interrelationships, behavioral correlates and
parental attributions. Perceptual and Motor Skills, 64,
1147–1157.
Hallman, M., Kannas, L. and Tynjälä, J. (1992) Smoking,
drinking and healthiness of living. A study of smoking
and drinking among 11–16-year-old Finns. Reports of the
National Board of Health, Finland, 61/1992, Helsinki,
Finland (in Finnish).
Hawkins, J. and Shaw, P. (1992) Self-reported sleep quality
in college students: a repeated measures approach. Sleep,
15, 545–549.
Healey, E. S., Kales, A., Monroe, L. J., Bixler, E. O.,
Chamberlin, K. and Soldatos, C. R. (1981) Onset of
164 J. Tynjälä et al.
insomnia: role of life-stress events. Psychosomatic
Medicine, 43, 439–451.
Hyyppä, M. T. (1991) Promoting good sleep. Health
Promotion International, 6, 103–110.
Hyyppä, M. T., Kronholm, E. and Matlar, C.-E. (1991)
Mental well-being of good sleepers in a random population
sample. British Journal of Medical Psychology, 64, 25–34.
Jöreskog, K. G. and Sörbom, D. (1993) LISREL 8 and
PRELIS 2 Documentation. Scientific Software Inter-
national, Chicago.
Kahn, A., Merckt, C., Rebuffat, E., Mozin, M. J., Sottiaux,
M., Blum, D. and Hennart, P. (1989) Sleep problems in
healthy preadolescents. Pediatrics, 84, 542–546.
King, A., Wold, B., Tudor-Smith, C. and Harel, Y. (1996)
The Health of Youth: A Cross-National Survey. WHO
Regional Publications, European Series No. 69.
Copenhagen, Report 69, Canada.
Kirmil-Gray, K., Eagleston, J. R., Gibson, E. and Thoresen,
C. E. (1984) Sleep disturbance in adolescents: sleep quality,
sleep habits, beliefs about sleep, and daytime functioning.
Journal of Youth and Adolescence, 13, 375–384.
Levy, D., Gray-Donald, K., Leech, J., Zvagulis, I. and Pless,
B. (1986) Sleep patterns and problems in adolescents.
Journal of Adolescent Health Care, 7, 386–389.
Lexcen, F. J. and Hicks, R. A. (1993) Does cigarette smoking
increase sleep problems? Perceptual and Motor Skills, 77,
16–18.
Lintunen, T., Leskinen, E., Oinonen, M., Salinto, M. and
Rahkila, P. (1995) Change, reliability, and stability in self-
perceptions in early adolescence: a four-year follow-up
study. International Journal of Behavioral Development,
18, 351–364.
Lugaresi, E., Cirignotta, F., Zucconi, M., Mondini, S., Lenzi,
P. L. and Coggagna, G. (1983) Good and poor sleepers:
an epidemiological survey of the San Marino Population.
In Guilleminault, C. and Lugaresi, E. (eds) Sleep/Wake
Disorders: Natural History, Epidemiology and Long-
Term Evolution. Raven Press, New York, pp. 1–12.
Mahon, N. E. (1994) Loneliness and sleep during
adolescence. Perceptual and Motor Skills, 81, 227–231.
Mahon, N. E. (1995) The contributions of sleep to perceived
health status during adolescence. Public Health Nursing,
12, 127–133.
Manni, R., Ratti, M. T., Marchioni, E., Castelnovo, G.,
Murelli, R., Sartori, I., Galimberti, C. A. and Tartara, A.
(1997) Poor sleep in adolescents: a study of 869 17-year-
old Italian secondary school students. Journal of Sleep
Research, 6, 44–49.
Monroe, L. J. (1967) Psychological and physiological
differences between good and poor sleepers. Journal of
Abnormal Psychology, 72, 255–264.
Morrison, D. N., McGee, R. and Stanton, W. R. (1992) Sleep
problems in adolescence. Journal of American Academy
of Child and Adolescent Psychiatry, 31, 94–99.
Nestmann, F. and Hurrelmann, K. (1994) Child and
adolescent research as a challenge and opportunity for
social support theory, measurement, and intervention:
and vice versa. In Nestmann, F. and Hurrelman, K. (eds)
Social Networks and Social Support in Childhood and
Adolescence. Walter de Gruyter, New York, pp. 1–22.
Palmer, C. D., Harrison, G. A. and Hiorns, R. W. (1980)
Association between smoking and drinking and sleep
duration. Annals of Human Biology, 7, 103–107.
Phillips, B. A. and Danner, F. J. (1995) Cigarette smoking
and sleep disturbance. Archives of International
Medicine, 7, 734–737.
Pilcher, J. J., Ginter, D. R. and Sadowsky, B. (1997) Sleep
quality versus sleep quantity: relationships between sleep
and measures of health, well-being and sleepiness in
college students. Journal of Psychosomatic Research, 42,
583–596.
Price, T., Coates, T. J., Thoresen, C. E. and Grinstead, O. A.
(1978) Prevalence and correlates of poor sleep among
adolescents. American Journal of Diseases of Children,
132, 583–586.
Rahkonen, O., Ahlström, S. and Karvonen, S. (1992)
Drinking, health and a strenuous lifestyle among young
people. Alkoholipolitiikka, 57, 169–178 (in Finnish with
an English abstract, p. 178).
Reicher, H. (1993) Family and peer relations and social-
emotional problems in adolescence. Studia Psychologica,
35, 403–408.
Rimpelä, M. and Rimpelä, A. (1983) Sleeping habits and
disturbances. In Rimpelä, M., Rimpelä, A., Ahlström, S.,
Honkala, E., Kannas, L., Laakso, L., Paronen, O., Rajala,
M. and Telama, R. (eds) Health Habits among Finnish
Youth. The Juvenile Health Habits Study 1977–1979.
Publications of the National Board of Health, Helsinki,
No. 4, pp. 71–81 (in Finnish with an English summary,
p. 213).
Rosenberg, M. (1985) Self-concept and psychological well-
being in adolescence. In Leathy, R. L. (ed.) The Develop-
ment of the Self. Academic Press, Orlando, pp. 205–246.
Rutter, M. (1985) Resilience in the face of adversity.
Protective factors and resistance to psychiatric disorders.
British Journal of Psychiatry, 147, 598–611.
Saarenpää-Heikkilä, O. A., Rintahaka, P. J., Laippala, P. J.
and Koivikko, M. J. (1995) Sleep habits and disorders
in Finnish schoolchildren. Journal of Sleep Research, 4,
173–182.
Shapiro, C. M., Warren, P. M., Trinder, J., Paxton, S. J.,
Oswald, I., Flenley, D. C. and Catterall, J. R. (1984) Fitness
facilitates sleep. European Journal of Applied Physiology,
53, 1–4.
Strauch, I. and Meier, B. (1988) Sleep need in adolescents:
a longitudinal approach. Sleep, 11, 378–386.
Turtle, J., Jones, A. and Hickman, M. (1996) Young people
and health: the health behaviour of school-aged children.
A report on the 1995 survey. August, BMRP International,
London.
Tynjälä, J. (1993) Sleeping habits, sleep difficulties and
educational expectations among Finnish school children.
In Shemeikka, S. and Nissinen, A. (eds) The Yearbook of
the Health Education Research 1992. Publications of the
National Board of Health, Helsinki, No. 1, pp. 99–113
(in Finnish with an English summary, p. 109).
Tynjälä, J. and Kannas, L. (1993) Sleeping habits of Finnish
school children by sociodemographic background. Health
Promotion International, 8, 281–289.
Tynjälä, J., Kannas, L. and Välimaa, R. (1993) How young
Europeans sleep. Health Education Research, 7, 69–80.
Tynjälä, J. and Törmäkangas, K. (1995) The 1994 data of a
WHO-co-ordinated survey of school children’s health
behaviour. In Kannas, L. (ed.) Perceived Health, Well-being
and School Satisfaction Among School Children. National
Board of Education, Helsinki, pp. 193–197 (in Finnish).
Tynjälä, J., Kannas, L. and Levälahti, E. (1997) Perceived
tiredness among adolescents and its association with
sleep habits and use of psychoactive substances. Journal
of Sleep Research, 6, 189–198.
Urponen, H., Vuori, I., Hasan, J. and Partinen, M.
(1988) Sleeping habits, sleep quality and sleep disorders

in urban working-aged people. Sosiaalilääketieteellinen
Aikakauslehti, 25, 338–351 (in Finnish with an English
abstract, p. 351).
Välimaa, R., Kannas, L. and Pötsönen, R. (1993) Social
support, adolescents and perceived health. In Shemeikka,
S. and Nissinen, A. (eds) The Yearbook of the Health
Education Research 1992. Publications of the National
Board of Health, Helsinki, No. 1, pp. 63–80 (in Finnish
with an English summary, pp. 79–80).
Vignau, J., Bailly, D., Duhamel, A., Vervaecke, P., Beuscart,
R. and Collinet, C. (1997) Epidemiologic study of sleep
quality and troubles in French secondary school
adolescents. Journal of Adolescent Health, 21, 343–350.
Vilhjalmsson, R. (1994) Effects of social support on self-
assessed health in adolescence. Journal of Youth and
Adolescence, 23, 437–452.
Perceived sleep quality 165
Vuori, I., Urponen, H., Hasan, J. and Partinen, M. (1988)
Epidemiology of exercise effects on sleep. Acta Physio-
logica Scandinavica, 133 (Supplement 574), 3–7.
Weydahl, A. (1991a) Sex differences in the effect of
exercise upon sleep. Perceptual and Motor Skills, 72,
455–465.
Weydahl, A. (1991b) Sleep-quality among girls with differ-
ent involvement in competitive sports during fall
in the arctic circle. Perceptual and Motor Skills, 73,
883–892.
Yarcheski, A. and Mahon, N. E. (1994) A study of sleep
during adolescence. Journal of Pediatric Nursing, 9,
357–367.
Zammit, G. K. (1988) Subjective ratings of the character-
istics and sequelae of good and poor sleep in normals.
Journal of Clinical Psychology, 44, 123–130.

APPENDIX 1. FACTOR STRUCTURES USED IN THE FINAL MODELS AMONG 15-YEAR-OLDS

Factor (F), variable (V) Questions and their alternatives in parentheses

Perceived sleep quality (F)
Home atmosphere (V, F)a
Sleep hygiene (F)
Self-perception (F)
Addictive behaviour (F)
Leisure time physical activity (F)
How well do you usually sleep? (very well, quite well, satisfactorily, quite badly,
very badly).
How often do you have difficulties in falling asleep? (almost daily, more than once a
week, about once a week, about once a month, rarely or never).
How often do you wake up during the night? (almost daily, more than once a week,
about once a week, about once a month, rarely or never).
What kind of atmosphere do you have in your home? (very good, quite good, neither
good nor bad, quite bad, very bad).
Do you usually go to bed at a certain time? (very regularly, quite regularly, quite
irregularly, very irregularly).
When do you usually go to bed if you have to go to school the next morning?
(alternatives ranged from no later than 9 pm to 2 am or later in half hour intervals).
Do you feel helpless? (always, often, sometimes, rarely, never).
Do you feel confident in yourself? (always, often, sometimes, rarely, never).
Index of self-esteem was formulated from the following statements: I feel that I am as
capable and skilful as any other person; I think that I have plenty of good qualities;
I often feel that I am a failure; I can manage things as well as any other person;
I feel that I am all right; I am satisfied with myself; I sometimes feel totally useless;
I sometimes think I am no good at all (strongly agree, partly agree, partly disagree,
strongly disagree).
How many cigarettes do you usually smoke in a week?
Have you ever had so much alcohol that you were really drunk? (no, never; yes, once;
yes, two–three times; yes, four–10 times; yes, more than 10 times).
How often do you drink beer? (every day, every week, every month, less than once a
month, never).
Outside school hours: How often do you usually exercise in your free time so much
that you get out of breath or sweat? (every day, four–six times a week, two–three
times a week, once a week, once a month, less than once a month, never).
Outside school hours: How many hours do you usually exercise in your free time
so much that you get out of breath or sweat? (None, about 30 min, about 1 h, about
2–3 h, about 4–6 h, 7 h or more).
Are you member of a sports club? (no; yes, but I don’t participate in exercises;
yes, and I am training in a sports club). This variable was dichotomized afterwards:
the first two categories were combined.

a
Home atmosphere factor was composed of home atmosphere variable.
166 J. Tynjälä et al.
APPENDIX 2. POLYCHORIC CORRELATION COEFFICIENTS AMONG 15-YEAR-OLD BOYS
(n = 484)

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

(1) Wake up during night → seldom 1.00

(2) Difficulty in falling asleep → seldom 0.55 1.00
(3) How well do you sleep in general → well 0.59 0.53 1.00
(4) Bedtime during school week → early 0.02 0.18 0.00 1.00
(5) Regularity of bedtime → regular 0.12 0.07 0.19 0.46 1.00
(6) Cigarettes a week → none 0.06 0.17 0.05 0.38 0.24 1.00
(7) Times being drunk → never 0.08 0.19 0.06 0.33 0.29 0.71 1.00
(8) Drinking beer → never 0.08 0.17 0.06 0.35 0.26 0.66 0.78 1.00
(9) Feeling of helplessness → never 0.28 0.20 0.17 –0.07 0.01 –0.10 –0.04 –0.09 1.00
(10) Feeling of self-confidence → always 0.23 0.11 0.17 –0.08 0.11 –0.10 –0.04 –0.09 0.58 1.00
(11) Self-esteem → good 0.12 0.17 0.17 0.06 0.16 0.12 0.10 0.04 0.49 0.54 1.00
(12) Home atmosphere → good 0.37 0.25 0.28 0.08 0.27 0.11 0.21 0.15 0.22 0.27 0.24 1.00

APPENDIX 3. POLYCHORIC CORRELATION COEFFICIENTS AMONG 15-YEAR-OLD GIRLS

(n = 551)

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)

(1) Wake up during night → seldom 1.00

(2) Difficulty in falling asleep → seldom 0.53 1.00
(3) How well do you sleep in
general → well 0.62 0.65 1.00
(4) Bedtime during school week → early 0.04 0.18 0.06 1.00
(5) Regularity of bedtime → regular 0.16 0.37 0.30 0.59 1.00
(6) Cigarettes a week → none 0.17 0.20 0.18 0.21 0.28 1.00
(7) Times being drunk → never 0.09 0.15 0.06 0.19 0.19 0.66 1.00
(8) Drinking beer → never –0.03 0.13 0.05 0.09 0.22 0.59 0.75 1.00
(9) Physical activity (times) →
almost daily 0.03 0.09 0.16 0.02 –0.02 0.08 0.04 0.05 1.00
(10) Physical activity (h) →
many hours a week 0.18 0.10 0.16 0.04 0.06 0.21 0.05 0.05 0.81 1.00
(11) Activity in sport club → active 0.25 0.05 0.19 –0.04 –0.07 0.17 –0.04 0.00 0.59 0.71 1.00
(12) Feeling of helplessness → never 0.27 0.32 0.31 –0.05 0.06 0.01 –0.04 –0.06 0.05 0.09 0.05 1.00
(13) Feeling of self-confidence → always 0.29 0.29 0.30 –0.02 0.02 0.05 –0.02 –0.05 0.12 0.16 0.01 0.53 1.00
(14) Self-esteem → good 0.30 0.28 0.29 0.05 0.13 0.21 0.05 0.00 0.14 0.24 0.17 0.57 0.64 1.00
(15) Home atmosphere → good 0.26 0.29 0.43 0.10 0.19 0.21 0.24 0.22 0.19 0.17 0.06 0.17 0.29 0.30 1.00

APPENDIX 4. CORRELATIONS OF FACTORS AND HOME ATMOSPHERE VARIABLE IN THE

FINAL MODELS. BOYS’ CORRELATIONS IN THE UPPER TRIANGLE AND GIRLS’
CORRELATIONS IN THE LOWER TRIANGLE

(1) (2) (3) (4) (5)

(1) Perceived sleep quality → good 1.000 — 0.316 0.404 0.308

(2) Physical activity → active 0.252 1.000 — — —
(3) Self-perception → good 0.472 0.258 1.000 0.262 0.062
(4) Home atmosphere → good 0.384 0.208 0.420 1.000 0.337
(5) Health habits → healthy 0.519 0.139 0.119 0.435 1.000