• children;
  • mental health;
  • parental rules;
  • sleep quality;
  • time in bed


  1. Top of page
  2. Abstract
  7. References

A health-sleep model concerning the relationship of mental health and parental rules with time in bed and sleep quality has been developed on the basis of survey data collected from 448 children in the first trimester of the regular school year. The children had a mean age of about 11 years and were attending last two grades of primary school. The relations between mental health characteristics, parental rules concerning sleep, sleep environment, sleep quality and time in bed are analysed using a structural equation model. In addition to a latent variable for ‘mental health’, two uncorrelated latent variables had to be introduced for sleep to achieve a satisfactory fit. One latent variable is related to sleep quality (restorative sleep), and the other relates to lying awake in bed prior to sleep (awake in bed). Restorative sleep shows a strong relationship with mental health, and awake in bed is related to having an own bedroom.


  1. Top of page
  2. Abstract
  7. References

Children’s sleep is usually considered to be influenced by child rearing activities of the parents and the sleep environment. Studies on this subject are scarce, however, and mostly deal with sleep and sleep problems of infants and young children. Research on sleep environment suggests that not sharing a bedroom, but sharing a bed is associated with a higher incidence of sleep problems (Blader et al. 1997). Parental behavior that plays a role in sleep problems are the presence of parents at sleep-onset (Ottaviano et al. 1996), parenting laxness (Owens-Stively et al. 1997), settling disorders and misconceptions about normal children’s sleep behavior (Armstrong et al. 1994), and putting children to bed too early, or according to a rigid schedule (Ophir-Cohen et al. 1993).

As the child grows older, parental influence is reduced and the child takes over more control of his or her own schedule. This may lead to symptoms of sleep-onset insomnia, difficulty in awakening, and daytime sleepiness as a consequence of inadequate sleep duration (Ferber 1990). Sleep restriction of 5 h during a single night in a sleep laboratory showed shorter sleep onset latencies and sleepiness with children aged between 10 and 14 years (Randazzo et al. 1998). Comparing extreme groups of sleepers, Wolfson and Carskadon (1998) report that adolescents sleeping 6 h and 15 min or less have a more depressive mood than adolescents sleeping 8 h and 45 min or more. As a consequence of chronic sleep deprivation, daytime functioning of teenagers may be affected by sleepiness, tiredness, difficulty getting up, moodiness, shortened attention span and concentration difficulties in school (Carskadon et al. 1998; Epstein et al. 1998; Gau and Soong 1995; Wolfson and Carskadon 1998).

Convincing relationships between sleep of children and mental health have been found when assessing sleep quality. Subjective sleep quality is usually defined as difficulty falling asleep and/or maintaining sleep, the subjective feeling of poor or good sleep, or a combination of these variables (Äkerstedt et al. 1997). Having difficulty in falling asleep shows a relationship with psychosomatic symptoms in children 11–12-year-old (Tynjäläet al. 1993). Totterdell et al. (1994) indicate a relationship with mood disturbances and physical symptoms (such as backache, nausea, etc.) for adolescents. The feeling of not sleeping well was also related to emotional factors and poor sleep hygiene in 17-year-old Italian youngsters (Manni et al. 1997). According to Kirmil-Gray et al. (1984) occasional as well as chronic poor-sleeping adolescents report being significantly more depressed, moody, tense, irritable, lacking energy, feeling less rested and less alert than good sleepers. Adolescents suffering from insomnia are more anxious, depressed, inattentive and exhibit more behavioral disorders than those who do not (Morrison et al. 1992). Anxiety (Blader et al. 1997; Lindberg et al. 1997) and increased family stress (Sadeh et al. 2000) are also associated with poorer sleep quality. Based on a wide range of clinical and observational data, Dahl (1996) concludes that inadequate sleep of children may result in tiredness, irritability, quick frustration and difficult modulation of emotions.

The aim of this paper is to develop a structural equation model (Bollen 1989) in which we attempt to correlate sleep environment, child rearing aspects as well as mental health aspects with sleep duration, sleep quality, getting up in the morning and feeling rested at school. Sleep quality has to be interpreted here as the combined effect of difficulty in falling asleep and maintaining sleep, and the subjective feeling of sleeping well or not. The mental health aspects of the children consist of the presence or absence of psychoneurotic and neurosomatic complaints and the degree of self-acceptance. Because the study concerns pre-adolescents, we focus on child rearing aspects that are relevant for this age group. These are parental rules concerning sleep, such as whether the child has been brought to bed or can choose his or her own bedtime.


  1. Top of page
  2. Abstract
  7. References


Data were collected through questionnaires that were completed by the children in the first trimester of the school year. The students filled out these questionnaires in their classroom. The questionnaires were administered in the morning. To avoid weekend and holiday effects, data were not collected on Mondays or shortly after the midterm break. The researchers administered the questionnaires personally. Each child responded anonymously.


A group of 448 children of the two last grades of primary schools in Amsterdam participated in the study. Their age varied from 9 years and 5 months to 13 years and 5 months, with a mean age of 11 years and 3 months (SD=8 months). The total group consisted of 228 boys and 218 girls (missing=2). In total 18 classes from 7 schools participated in the research. To get a homogenous group of children and to prevent cultural differences in sleep patterns, schools were selected in middle class neighborhoods with a mostly white student body. All schools started at about half past eight in the morning.

In 278 cases both parents were employed, in 144 cases only one parent worked and 25 children had parents who were both unemployed (missing=1). In 98 cases the child’s parents were divorced.


The time the children generally went to bed during school nights until the moment of getting up in the morning was used to estimate sleep duration. Because this does not correspond with actual sleeping time, we call this ‘time in bed’. Sleep quality was measured on the basis of a scale consisting of four closed questions with three ordinal response categories ranked from 1 to 3 (Meijer et al. 2000). The items refer to a subjective feeling of sleep latency (fall asleep at once, stay awake for awhile, and takes a long time to fall asleep), sleeping through the night, ease of falling asleep after awakening in the night, and a question about their subjective sleep perception. The reliability of this quality-of-sleep scale was α=0.72. The minimum score is 4 (sleeping badly) and the maximum score 12 (sleeping well). Questions about difficulty in getting up the next morning and feeling rested at school could only be answered by ‘Yes’ or ‘No’ to force the child into an explicit position.

Parental rules with respect to sleep referred to whether the child was brought to bed, if the parents allowed the child to decide the bedtime itself, if the child slept alone, if the light had to be turned off immediately on school nights, and if someone checked the child’s room after the child had gone to bed. The response to these questions, which only could be answered by ‘Yes’ or ‘No’, showed no bias with age. As could be expected, older children reported more often that they might decide their own bedtimes and when to turn off the light. The correlations between children’s and parents’ answers on questions about bedtimes and sleep habits show to be high (Epstein et al. 1998).

The mental health aspects of the children were measured by the Amsterdam Biographic Questionnaire for Children (ABV-K) (van Dijl and Wilde 1982). This instrument is a forced choice (yes/no) personality questionnaire measuring neuroticism (occurrence of psychoneurotic complaints) (N), neurosomaticism (occurrence of functional somatic complaints) (NS), and self acceptance (T), i.e. a self-critical vs. a self-flattering attitude. Examples of items of the respective scales are: I worry a lot, I often have a headache and I lie sometimes. There are separate standardized scores for boys and girls. The reported reliability of the subscales runs from α=0.74 to α=0.80 (Van Dijl and Wilde 1982). Cronbach’s standardized α values in the current study are 0.80 (boys) and 0.82 (girls) for neuroticism 0.73 (boys) and 0.76 (girls) for neurosomaticism, and 0.70 (boys) and 0.74 (girls) for self-acceptance.

Statistical methods

Two hierarchical multiple regression analyses are used to examine whether background variables, parental rules and mental health characteristics adequately predict sleep quality and time in bed during school nights. Subsequently, a ‘structural equation model’ will be proposed for the variables that show a significant relationship with sleep quality and time in bed. A structural equation model consists of a measurement model (to be seen as confirmative factor analyses for the latent (unobserved) variables) and a path model in which relations between latent variables and observed variables can be specified. Although a model, in fact, consists of many regression equations, it is mostly being represented by a graph in which arrows (paths) designate the direction of the relationship. An outgoing arrow from a latent or an observed variable indicates an independent variable in a regression equation. An incoming arrow points at the dependent variable of a regression equation. If a model fits well (e.g. if the χ2 statistic is not significant), a significant path coefficient means that the regression coefficient in the corresponding regression equation differs significantly from zero.


  1. Top of page
  2. Abstract
  7. References

Sample characteristics

Time in bed during school nights varied from 502 to 720 min with a mean duration of 616 min (SD=40 min). The sleep quality of the children had a mean score of 8.28 (SD=1.79) with a minimum score of 4 (sleeps bad) and a maximum score of 12 (sleeps well). From the children, 15.1% had a score of 6 or less (roughly 1 SD below the mean). This suggests that 15.1% of the children did not sleep well. Furthermore, 192 (44%) children had difficulty in getting up in the morning and 109 (25%) of the children did not feel rested when at school.

From the children 372 (83%) had their own bedroom, but only 39 (9%) are permitted to decide for themselves when to go to bed. What’s more, 145 children (33%) were brought to bed, and 291 children (65%) were checked on by the parents to see if they were in bed. Parents were quite tolerant about putting out the light immediately on school nights: 332 children (77%) are left to decide for themselves. Cumulative percentages for this group show that 43% turned off the light by 9.00 p.m., 75% by 9.30 p.m., 90% by 10.00 p.m. and 99% by 11.00 p.m.

Hierarchical regression analyses for time in bed and sleep quality

Hierarchical regression analyses are performed for time in bed during school nights and sleep quality (Table 1). A statistically significant relationship of ‘own bedroom’ and ‘someone checks’ with ‘time in bed’ can be noted. The negative influence of age and deciding the bedtime oneself is understandable as well, because the bedtime will be later and sleep duration will be shorter in both cases. Turning off the light immediately has the opposite effect.

Table 1.   Hierarchical regressions for time in bed during school nights and sleep quality with background variables of the child, mental health characteristics and variables concerning sleep Thumbnail image of

Also noticeable is the completely different pattern of R2 changes and β coefficients for sleep quality. Here the main contribution takes place at the introduction of the mental health characteristics as measured by the ABV-K. A higher score in the direction of psychoneurotic and neurosomatic complaints corresponds to sleeping more badly. A more self-critical attitude is related to a lower sleep quality as well. None of the parental rules that parents generally think of as important for a good rest shows the slightest relation with sleep quality.

Health–sleep model

Using the results above, we will examine how significant mental health characteristics, own room and parental rules concerning sleep relate to possible consequences of sleep, such as having difficulty in getting up in the morning (44% of those polled) and not feeling rested at school (25% of those polled). The health–sleep model in Figure 1 performs well in this respect. The structural equation model consists of all remaining observed variables when the model is urged to fit the data well. The model is plausible and the fit is satisfactory: χ2=22.14 with d.f.=16 (P=0.14); RMSEA=0.05 and CFI=0.99. The fit does not differ significantly from the best fitting saturated model.


Figure 1.  Structural equation model (Lisrel model) for health and sleep standardized coefficients.

Download figure to PowerPoint

The core of the model consists of three related latent variables; ‘mental health’, ‘restorative sleep’ and ‘awake in bed’. Mental health points to emotional stability (steadiness of mood). Restorative sleep reflects that part of the sleep that corresponds to a good night’s rest, and awake in bed pertains to the time that children spend in bed or in their bedroom before falling asleep (for instance reading a book, playing a game, etc.). The structural relationship between the latent variables in the model implies that both these sleep related concepts are needed to get the model to fit: one that is being influenced by mental health conditions which is called ‘restorative sleep’, and one that is not influenced by mental health conditions which we call ‘awake in bed’. The relationship between mental health and restorative sleep (as shown in the model) means that better mental health leads to a better night’s rest. As is clear from the standardized path coefficients in Fig. 1, the negative contribution of mental health to the neurotic aspects is −0.69 and −0.62, respectively, and the positive contribution to self-acceptance is 0.45. The correlated error (e2–e3) of psychoneurotic complaints (e2) and neurosomatic complaints (e3) is plausible as well. The contributions are highly significant. This can easily be concluded from the t-values of the unstandardized path coefficients in Table 2.

Table 2.   Unstandardized parameter estimates for the mental health–sleep model and their standard normal distributed t-value Thumbnail image of

Being influenced strongly by the mental health conditions, the latent variable ‘restorative sleep’ gives rise to different values of the observed variables ‘feeling rested’, ‘sleep quality’ and ‘having difficulty in getting up’. The standardized path coefficients make clear that the latent variable ‘restorative sleep’ has a positive relation with feeling rested and sleep quality, and a negative relation with difficulty in getting up.

Completely independent from ‘restorative sleep’ and ‘mental health’, the latent variable ‘awake in bed’ is affected by ‘having his (or her) own bedroom’ as the only aspect that leads to a model which fits acceptably. ‘Awake in bed’ relates furthermore to time in bed during school days in general, difficulty in getting up and not feeling rested at school. Although the relationships concerning ‘awake in bed’ are smaller than with ‘restorative sleep’, the standardized regression coefficients are significant for time in bed, not feeling rested at school and difficulty in getting up.

‘Awake in bed’ refers to the time that children spend in bed or in their bedroom before falling asleep (such as reading a book, playing a game, etc.). We think this to take place especially with children who are sent to bed too early. If these children have their own bedroom, they will be inclined to amuse themselves before falling asleep. So the relation with time in bed, which actually is hours in bed or in the bedroom, will be positive. However, these children may stay awake longer than is beneficial to them. Consequently, they will tend to show difficulty in getting up (a positive relation) and do not feel rested at school (a negative relationship). Addition of other variables does not improve the model.


  1. Top of page
  2. Abstract
  7. References

This study indicates a strong relationship between sleep quality and mental health characteristics. The health–sleep model predicts no relationship between time in bed and sleep quality; it is therefore doubtful whether a shortened time in bed in normal situations creates a risk of inadequate sleep quality for pre-adolescents. Being an indicator of restorative sleep, sleep quality seems mainly to be related to mental characteristics (Dahl 1996; Kirmil-Gray et al. 1984; Lindberg et al. 1997; Manni et al. 1997; Morrison et al. 1992; Sadeh et al. 1995; Totterdell et al. 1994; Tynjäläet al. 1993).

A distinct conclusion is obtained, however, via the latent variable ‘awake in bed’, because this variable is not related to mental characteristics. ‘Awake in bed’ resembles the practice of ‘staying in bed while awake’ of Perlis et al. (1997) who studied insomnia. In their opinion, the bed and the bedroom become conditioned with arousal instead of sleepiness because of activities (reading, listening to music, playing games) that the child performs if he/she goes to bed too early. In this way, sleep problems may develop without reference to mental health conditions. Given this relationship it is quite conceivable that inconsistent changes of sleep patterns may develop that can lead to erratic or otherwise unsatisfactory sleep–wake cycles and so decrease sleep time (Stores 1996).

Although the direction of influence in the health–sleep model runs from mental health to restorative sleep, this does not mean that mental health ‘causes’ restorative sleep. The relationship must be seen as a process in which restorative sleep and mental health have reciprocal effects. If sleep quality is bad for a long period of time and the child does not feel rested anymore, the consequence may be that the child develops emotional problems or psychiatric disorders (Navelet 1996).

A possible limitation of the current study is that the data rely exclusively on child report. In a further study information about sleep of the child should also be provided by the parents. To corroborate children’s reports sleep diaries and actometers could be used. Only longitudinal research, however, can provide decisive answers about the specific relation of sleep, sleep habits, and mental health over time. Testing the model on different age groups as well as comparison with clinical groups seems necessary to determine the generalizability and validity of the model.

  1. The data were analyzed with the computer program Amos (Arbucle 1995). Because of two dichotomeous variables, a bootstrap of 1000 samples was used to evaluate the solution. Since bootstrapping in Amos cannot be used together with missing value, likewise deletion had to be used before, thereby reducing the sample size from 448 to 388. The parameter estimates are in the same order of magnitude.


  1. Top of page
  2. Abstract
  7. References
  • 1
    Äkerstedt, T., Hume, K., Minors, D., Waterhouse, J. Good sleep – its timing and physiological sleep characteristics. J. Sleep Res., 1997, 6: 221229.
  • 2
    Arbucle, J. L. Amos Users Manual. Chicago, Smallwaters, 1995.
  • 3
    Armstrong, K. L., Quinn, R. A., Dadds, M. R. The sleep patterns of normal children. The Med. J. Australia, 1994, 16: 202205.
  • 4
    Blader, J. C., Koplewicz, H. S., Abikoff, H., Foley, C. Sleep problems of elementary school children. A community survey. Arch. Pediatr Adolesc Med., 1997, 151: 473480.
  • 5
    Bollen, K. Structural Equations with Latent Variables. Wiley, New York, 1989.
  • 6
    Carskadon, M. A., Wolfson, A. R., Acebo, C., Tzischinsky, O., Seifer, R. Adolescent sleep patterns, circadian timing, and sleepiness at a transition to early school days. Sleep, 1998, 21: 871881.
  • 7
    Dahl, R. E. The impact of inadequate sleep on children’s daytime cognitive function. Seminars Pediatric Neurol., 1996, 3: 4450.
  • 8
    Van Dijl, H. & Wilde, G. J. S. Handleiding bij de Amsterdamse Biografische Vragenlijst voor Kinderen (ABV-K). (Manual Amsterdam Biographical Questionnaire for children). Amsterdam, F. van Rossen, 1982.
  • 9
    Epstein, R., Chillag, N., Lavie, P. Starting times of school, effects on daytime functioning of fifth grade children in Israel. Sleep, 1998, 21: 250256.
  • 10
    Ferber, R. Sleep schedule-dependent causes of insomnia and sleepiness in middle childhood and adolescence. Pediatrician, 1990, 17: 1320.
  • 11
    Gau, S. & Soong, W. Sleep problems of junior high school students in Taipei. Sleep, 1995, 18: 667673.
  • 12
    Kirmil-Gray, K., Eagleston, J. R., Gibson, E., Thoresen, C. E. Sleep disturbance in adolescents, sleep quality, sleep habits, beliefs about sleep, and daytime functioning. J. Youth Adolescence, 1984, 13: 375384.
  • 13
    Lindberg, E., Janson, C., Gislason, T., Björnsson, E., Hetta, J., Boman, G. Sleep disturbances in a young adult population, can gender differences be explained by differences in psychological status? Sleep, 1997, 20: 381387.
  • 14
    Manni, R., Ratti, M. T., Marchioni, E., Castelnovo, G., Murelli, R., Sartori, I., Galimberti, C. A., Tartara, A. Poor sleep in adolescents, a study of 869 17-year old Italian secondary school students. J. Sleep Res., 1997, 6: 4449.
  • 15
    Meijer, A. M., Habekothé, H. T., Van Den Wittenboer, G. L. H. Time in bed, quality of sleep and school functioning of children. J. Sleep Res., 2000, 9: 145153.
  • 16
    Morrison, D. N., Mc. Gee R, Stanton W. R. Sleep problems in adolescence. J. Am. Acad. Child Adolescence Psychiatry, 1992, 31: 9499.
  • 17
    Navelet, Y. Insomnia in the child and adolescent. Sleep, 1996, 19: S23S28.
  • 18
    Ophir-Cohen, M., Epstein, R., Tzischinsky, O., Tirosh, E., Lavie, P. Sleep patterns of children sleeping in residential care, in kibbutz dormitories and at home. A comparative study. Sleep, 1993, 16: 428432.
  • 19
    Ottaviano, S., Giannotti, F., Cortesi, F., Bruni, O., Ottaviano, C. Sleep characteristics in healthy children from birth to 6 years of age in the urban area of Rome. Sleep, 1996, 19: 13.
  • 20
    Owens-Stively, J., Frank, N., Smith, A., Hagino, O., Spirito, A., Arrigan, M., Alario, A. J. Child temperament, parenting discipline style, and daytime behavior in childhood sleep disorders. Dev Behav Pediatrics, 1997, 18: 314321.
  • 21
    Perlis, M. L., Giles, D. E., Mendelson, W. B., Bootzin, R. R., Wyatt, J. K. Psychophysiological insomnia, the behavioural model and a neurocognitive perspective. J. Sleep Res., 1997, 6: 179188.
  • 22
    Randazzo, A. C., Muehlbach, M. J., Schweitzer, P. K., Walsh, J. K. Cognitive function following acute sleep restricton in children ages 10–14. Sleep, 1998, 21: 861867.
  • 23
    Sadeh, A., McGuire, J., Sachs, H., Seifer, R., Tremblay, A., Civita, R., Hayden, R. M. Sleep and psychological characteristics of children on a psychiatric inpatient unit. J. Am. Acad. Child Adolescent Psychiatry, 1995, 34: 813818.
  • 24
    Sadeh, A., Raviv, A., Gruber, R. Sleep patterns and sleep disruptions in school-age children. Dev Psychol, 2000, 36: 291301.
  • 25
    Stores, G. Practitioner review, assessment and treatment of sleep disorders in children and adoloscents. J. Child Psychol Psychiatry, 1996, 37: 907925.
  • 26
    Totterdell, P., Reynolds, S., Parkinson, B., Briner, R. B. Associations of sleep with everyday mood, minor symptoms and social interaction experience. Sleep, 1994, 17: 466475.
  • 27
    Tynjälä, J., Kannas, L., Välimaa, R. How young Europeans sleep. Health Education Res., 1993, 8: 6980.
  • 28
    Wolfson, A. R. & Carskadon, M. A. Sleep schedules and daytime functioning in adolescents. Child Dev, 1998, 69: 875877.