Sleepless in America: Inadequate Sleep and Relationships to Health and Well-being of Our Nation's Children
Abstract
OBJECTIVE. Our goal was to identify characteristics associated with inadequate sleep for a national random sample of elementary school–aged children (6–11 years) and adolescents (12–17 years).
METHODS. Data from 68418 participants in the 2003 National Survey of Children's Health were analyzed by using weighted bivariate and multivariate regression models. The dependent variable was report of not getting enough sleep for a child of his or her age ≥1 night of the past week. Independent variables included demographic characteristics, child health, school and other activities, and family life.
RESULTS. Parents of elementary school–aged children with inadequate sleep were more likely to report that their child was having problems at school or had a father with fair or poor health. Parents of adolescents with inadequate sleep were more likely to report that their child had an atopic condition, frequent or severe headaches, a parent with less-than-excellent emotional health, or experienced frequent parental anger. Inadequate sleep in both age groups was associated with parental report that their child usually or always displayed depressive symptomatology, family disagreements involved heated arguing, or parental concern that the child was not always safe at home, at school, or in their neighborhood.
CONCLUSIONS. Approximately 15 million American children are affected by inadequate sleep. Primary care providers should routinely identify and address inadequate sleep and its associated health, school, and family factors.
Inadequate sleep during childhood is an invisible phenomenon that fails to receive attention from primary care providers until it interferes with the child's behavior, mood, or performance.1–4 Community and multi-site studies have consistently reported that up to 20% to 25% of US children and adolescents experience a range of sleep problems.5–10 Inadequate sleep takes many forms: difficulty with sleep onset, length, or circadian rhythms with resulting daytime sleepiness experienced by otherwise healthy children; disturbed sleep associated with acute and chronic illness; and primary sleep disorders. The least attention has been paid to the first group, and there is little consensus about the second.
Normative requirements for adequate sleep based on epidemiologic and laboratory studies11–13 reflect the need for progressively less sleep by developmental stage with averages of 10 hours for 5- to 13-year-olds (declining from 11.1 at 5 years to 9.0 hours at 13 years) and 8 to 9 hours for adolescents 14 to 18 years of age.11 More recent reports suggest that adolescent needs may be underestimated,6 especially during midadolescence.14 Evidence is accumulating that US children and teens typically sleep less than the required hours recommended.8–10,15,16 Primary care providers inadequately assess, diagnose, or treat sleep problems even when parents and children offer complaints.1 Strength of associations between sleep disturbances and cognition, behavior, or mood in otherwise healthy children range from robust to unknown.17
Relationships between sleep disturbance and chronic health conditions during childhood have been investigated. The association of sleep and attention-deficit/hyperactivity disorder (ADHD) is not well understood. The consequences of impaired sleep may resemble and/or exacerbate ADHD symptomatology. Conversely, ADHD symptoms and psychostimulant medication may exacerbate sleep impairment. Although as many as 50% of parents of children and adolescents with ADHD report sleep problems,18–20 studies using more objective polysomnography have failed to demonstrate differences in sleep architecture between children with and without ADHD.20
Similarly, both research and clinical experience supports the relationship between sleep problems and mood and anxiety disorders, although it is unclear whether the sleep problem or the psychiatric disorder is the primary problem.21,22 In some studies, primary psychopathology is associated with or worsened by sleep impairment.7 Consistently, sleep impairment and mood and anxiety disorders are comorbid conditions. Atopic disorders, such as asthma and allergic rhinitis, have also been associated with increased sleep impairment.2,23,24
The purpose of this study was to describe the prevalence and characteristics of inadequate sleep as perceived by parents of a random national sample of school-aged and adolescent children and reported as part of the 2003 National Survey of Children's Health (NSCH). We use a multi-domain conceptual approach incorporating child, family, and environmental factors. The assumptions are made that children require regular patterns and specific hours of sleep according to developmental stage and that inadequate sleep is undesirable and potentially deleterious to health.
METHODS
Data Source
Data for this study come from the responses of parents or caregivers of 68418 children between the ages of 6 and 17 years interviewed as part of the 2003 NSCH who responded to the question “During the past week, on how many nights did your child get enough sleep for a child his/her age?” “Enough sleep” was broadly interpreted as however the parent/caregiver defined it for a particular child. Parents of children <6 years of age were not asked this question as part of the survey. The survey design is described briefly in the article by Kogan and Newacheck25 in this issue; more in-depth information can be found elsewhere.26
Variables
We stratified children by age into 2 groups: school-aged children (6–11 years) and adolescents (12–17 years). Independent variables were organized by the following categories: demographic characteristics, child health (health status, comorbid conditions, and reported child behaviors); school and activities (problems in school, physical activity, and television viewing), and family/community life (family structure, parental health, and family stress). The dependent variable, inadequate sleep, was a response to the question on the survey indicating that the child did not sleep well on at least 1 night of the preceding week.
Some variables and/or response categories were merged into broader categories before inclusion in bivariate or multivariate models. Race and ethnicity variables were merged to encompass non-Hispanic white, non-Hispanic black, Hispanic, and other race. The category “other race” was inclusive of children identified as Asian or Native American. We dichotomized responses to questions about time spent watching television, watching videos, or playing video games at ≥2 hours and <2 hours in accord with the guidelines established by the American Academy of Pediatrics.27 We merged responses to 3 questions that asked parents whether a doctor or other health professional had told them that their child had asthma, hay fever or another respiratory allergy, or eczema or another skin allergy to create the variable “atopic condition.”
A variable “depressive symptoms” was created from collective responses to 4 questions that asked about level of parental concern regarding the following child behavioral characteristics: stubborn, sullen, or irritable; feeling worthless or inferior; unhappy, sad, or depressed; and withdrawn and does not get involved with others. We quantified level of concern for depressive symptoms as “never concerned,” indicating that the respondent did not express concern about any of the 4 component variables, and “sometimes” or “usually/always” concerned if the respondent expressed this level of concern for at least 1 of the 4 behaviors. We used a similar approach to create the variable “environmental safety” from 3 questions that asked parents how often they felt their child was safe in their community or neighborhood, school, and home. Human subjects review was not required for this study.
Analyses
The NSCH provides population weights to permit extrapolation of findings from this sample to national and state population estimates. Incorporating the appropriate weights, we used SUDAAN 9.0.0 (Research Triangle Institute, Research Triangle, NC)28 to perform all of our analyses. Bivariate analyses were conducted to examine relationships between the outcome variable, inadequate sleep, and variables included in each of the categories specified above for children 6 to 11 years of age and 12 to 17 years of age who experienced inadequate sleep with those who did not. Multivariate logistic regression analyses were performed to assess the independent associations between inadequate sleep and covariates of interest for each group of children.
RESULTS
The 68418 children identified for interviews were weighted to represent 47.4 million children nationwide. Overall, the parents of 31.9% of these children reported that their child's sleep had been inadequate on ≥1 night during the week before participation in the survey. The percentage of children affected by inadequate sleep demonstrates a strong relationship with age throughout childhood, with a more marked increase by age for children ≥12 years of age (Fig 1).
Percentage of children with reported inadequate sleep according to year of age (P < .001 for trend). Source: 2003 National Survey of Children's Health.
Characteristics of children with and without reported adequate sleep are presented by age group in Table 1. In both groups, children with reported inadequate sleep were more frequently non-Hispanic white, resided in families with education greater than high school and higher income levels, and were more frequently described as having less-than-excellent health compared with children with reported adequate sleep patterns. Disease comorbidities such as atopic conditions, attention-deficit disorder (ADD), or ADHD were more common in children with inadequate sleep. Parents of children with inadequate sleep reported less successful school experiences, with approximately one third having been contacted for a school-related problem and more reporting their child having been bullied by classmates.
Characteristics of Children With and Without Reported Inadequate Sleep According to Age Group
Table 2 presents results of the multivariate logistic regression analyses. All models controlled for demographic, child health, school and activities, and family life variables. When covariates were examined independently for their relationship to inadequate sleep, the following variables were associated with greater sleep needs in both age groups: (a) presence of 1 or more depressive symptoms, (b) less frequent days of physical activity, and living in a home where (c) parents argue heatedly or shout, (d) on rare occasions respond by hitting or throwing things during family disagreements, or (e) perceive that the environment at home, school, or in the community is not always safe. Some variables were exclusively associated with inadequate sleep in the elementary school age group: (a) having problems at school and (b) fair or poor paternal general health, whereas others were associated with inadequate sleep in the adolescent group only (a) an atopic condition, (b) frequent or severe headaches, (c) less-than-excellent maternal or paternal emotional health, (d) parental perception that the child is sometimes harder to care for than other children, and (e) parental anger with the child. Living in a family with reported income <400% of the federal poverty level (PL) was associated with significantly lower odds of inadequate sleep, even after controlling for potentially confounding variables. In addition, >2 hours of television viewing per day, diagnosis of learning disability, and being victimized by classmate bullying behavior were associated with significantly lower odds of inadequate sleep in the adolescent group.
Adjusted ORs for Inadequate Sleep for Children Aged 6 to 11 and 12 to 17 Years
DISCUSSION
The neuroscientist Robert Stickgold warns that because critical reparative and integrative processes “…occur exclusively during sleep and can't be reproduced when we are awake, the consequences of losing them look more and more terrifying….”29 The independent nonprofit National Sleep Foundation concluded from its 2004 Sleep in America poll that a remarkable number of children age 10 and younger have some kind of sleep problem,9 three quarters of parents are dissatisfied about children's sleep, and few pediatric providers inquire about or follow-up on sleep complaints; their more recently released adolescent data contain similar concerns.10 Carskadon30 refers to sleep as “the forgotten country” and calls for attention to the “sleeping half of children's lives.”
Is such alarm warranted? In our analysis, parents in a national random sample reported that 31.9% of their children experienced ≥1 night of inadequate sleep during the previous week. These children experienced health-related deficits associated with selected demographic, child, family, and environmental factors. This is consistent with growing evidence for deleterious effects of even occasional lapses in recommended amounts of sleep.
Acute sleep deprivation and chronic sleep disorders result in behavioral and performance deficits in school-aged children.2,31–34 Adolescents' erratic sleep patterns are fostered by pubertal changes in hormone secretion making it difficult to fall asleep before 11 pm7; yet early school start times compel them to be in class before 8 am.5 Weekday and weekend variations result in prolonged delayed sleep onset, insomnia, and daytime sleepiness throughout the week35 along with impaired mood, behavioral control, and academic performance.7,36
Ethnocultural and Socioeconomic Differences
Impact of sleep problems on children and families transcends all cultures.37 Contrary to earlier studies8,16,38 identifying minority and poorer children at greater risk, our data suggest that parents of non-Hispanic white children and those who have higher income and greater than high school education more frequently report the perception of inadequate sleep in their children. Roberts39 examined differences in sleep complaints among 5423 American adolescents of African, Chinese, Mexican, Central American, and Anglo descent and concluded minority status may affect risk for sleep problems. Primary school children in China averaged 9.3 vs 10.2 hours of sleep compared with US children and reported daytime sleepiness.40 Cross-sectional analysis of child journals and caregiver questionnaires for 755 children aged 8 to 11 years showed that at all ages minority boys slept significantly less compared with minority girls and all nonminority children.8 A parent-report survey of 472 children in 1 US city showed an inverse association between socioeconomic status and both parasomnias and noisy sleep.2 Exploration of sleep by ethnicity and socioeconomics remains in its infancy. Comparisons across studies are hampered by differences in definitions and data sources.
Sleep and Comorbidity
In our analysis, the relationship between inadequate sleep and mood disorders was present in both younger and older children. Children with depression subjectively reported poor and inadequate sleep when compared with controls; however, 1 study using objective electroencephalogram data failed to validate this relationship.41 When sleep architecture was studied in depressed, unmedicated children compared with healthy controls, only depressed older boys had significant sleep disturbance.42
Consistent with the literature, in this study adolescents with atopic conditions and headache were more likely to have inadequate sleep compared with children without these conditions. However, this finding did not hold true for younger children. Children in remission phase of atopic dermatitis evaluated by polysomnography demonstrate significantly disturbed sleep symptoms.43 As many as 64% of asthmatic children report 3 nocturnal awakenings per week.44 Individuals with allergic rhinitis sleep poorly and have sleep-disordered breathing and subsequent daytime sleepiness.45,46 Although children with migraine headaches subjectively experienced more sleep problems than did healthy control children,47,48 more objective actigraphy did not show differences in sleep parameters.49
Our data failed to demonstrate an association between inadequate sleep and ADHD in either age group in multivariate models. Studies relying on parental report of sleep behavior18,38,50 support an association between a diagnosis of ADHD and sleep problems in children. Parents of children with ADHD report sleep problems in 25% to 50% of these children,20 but these findings are not supported when more objective measurements such as polysomnography are used.51
Family Health and Social Factors
In both age groups, inadequate sleep was associated with family conflict; in teens it was further associated with parental emotional health, being perceived as a difficult child, and parental anger. Sleep issues are known to exist with child abuse and family violence.52,53 Our study makes the connection with inadequate sleep much earlier in the violence spectrum. It may be that perceived inadequate sleep is an alert for suboptimal family functioning and should trigger assessment, referral, and early intervention. Inadequate sleep was also associated with not feeling safe in school or neighborhood. The National Institute of Mental Health reports a significant increase in US children as witnesses to or victims of community violence54 with resulting physical and psychological effects.55 Regular use of brief sleep assessment tools7 can lead to addressing sleep issues and their underlying causes.
Limitations
These NSCH cross-sectional data limit interpretation to association. All bivariate and multivariate associations described in the text and displayed in the tables are bidirectional and reflect simultaneous relationships between independent variables and the dependent variable. Therefore, neither cause nor causal direction can be determined. In addition, sleep information is derived from parental response to 1 question concerning adequacy of a child's sleep during the past week. The survey does not provide information about perceived or diagnosed causes of inadequate sleep, whether it was a usual or unusual occurrence, or the norms by which each parent judged sleep. Nevertheless, prevalence of inadequate sleep and its effects in this large national sample are consistent with reports of epidemiologic sleep studies.9,10 Because evidence is accumulating concerning the health risks of inadequate sleep at all ages, future surveys should explore sleep in greater detail.
CONCLUSIONS
Sleep health is an important but underrecognized component of wellness in children. Sleep impairment may provide a critical alert for primary care providers to search for undiagnosed physical or psychological comorbidity, suboptimal coping, family dysfunction, or threats in school or community. Conversely, when any of these conditions and/or situations is known to be present, sleep impairment may be a comorbid symptom that needs to be addressed. Sleep health and its assessment should be a fundamental aspect of clinical prevention.
Footnotes
- Accepted September 15, 2006.
- Address correspondence to Arlene Smaldone, DNSc, CPNP, CDE, Columbia University School of Nursing, 630 W 168th St, New York, NY 10032. E-mail: ams130{at}columbia.edu
The authors have indicated they have no financial interests relevant to this study to disclose.
REFERENCES
- Copyright © 2007 by the American Academy of Pediatrics
