Prevalence and Risk Factors of Excessive Daytime Sleepiness in a Community Sample of Young Children: The Role of Obesity, Asthma, Anxiety/Depression, and Sleep

Abstract

Study Objectives:

We investigated the prevalence and association of excessive daytime sleepiness (EDS) with a wide range of factors (e.g., medical complaints, obesity, objective sleep [including sleep disordered breathing], and parent-reported anxiety/depression and sleep difficulties) in a large general population sample of children. Few studies have researched the prevalence and predictors of EDS in young children, none in a general population sample of children, and the results are inconsistent.

Design:

Cross-sectional

Setting:

Population -based.

Participants:

508 school-aged children from the general population.

Interventions:

N/A

Measurements and Results:

Children underwent a 9-hour polysomnogram (PSG), physical exam, and parent completed health, sleep and psychological questionnaires. Children were divided into 2 groups: those with and without parent reported EDS. The prevalence of subjective EDS was approximately 15%. Significant univariate relationships were found between children with EDS and BMI percentile, waist circumference, heartburn, asthma, and parent reported anxiety/depression, and sleep difficulties. The strongest predictors of EDS were waist circumference, asthma, and parent-reported symptoms of anxiety/depression and trouble falling asleep. All PSG sleep variables including apnea/hypopnea index, caffeine consumption, and allergies were not significantly related to EDS.

Conclusions:

It appears that the presence of EDS is more strongly associated with obesity, asthma, parent reported anxiety/depression, and trouble falling asleep than with sleep disordered breathing (SDB) or objective sleep disruption per se. Our findings suggest that children with EDS should be thoroughly assessed for anxiety/depression, nocturnal sleep difficulties, asthma, obesity, and other metabolic factors, whereas objective sleep findings may not be as clinically useful.

Citation:

Calhoun SL; Vgontzas AN; Fernandez-Mendoza J; Mayes SD; Tsaoussoglou M; Basta M; Bixler EO. Prevalence and risk factors of excessive daytime sleepiness in a community sample of young children: the role of obesity, asthma, anxiety/depression, and sleep. SLEEP 2011;34(4):503-507.

INTRODUCTION

Although excessive daytime sleepiness (EDS) in adults has been the focus of extensive research, studies on the risk factors associated with EDS in children have been limited. The prevalence of EDS in young children with sleep disordered breathing (SDB) varies greatly, from just 7% to as high as 49%.^1,2 This variation may be explained by the different methods used for determining EDS, sample size, and referral source, as well as confounding factors that have not been examined, such as obesity.

Although EDS in children is commonly assumed by physicians and lay persons alike to be the result of disturbed or inadequate sleep, which in turn may interfere with daytime functioning (e.g., academic performance, behavioral and psychological problems), it remains unclear whether EDS is a frequent manifestation of SDB or disturbed sleep in young children. One study reported a weak association with EDS and SDB in children,³ while 2 other studies found a strong association of SDB with EDS in obese children.^4,5 One population-based study on subjective report of sleep disturbance and behavioral problems in children found no association between EDS and emotional or disruptive behaviors in school.⁶ In children scheduled for adenotonsillectomy, Chervin et al. reported more objectively assessed (multiple sleep latency test) and subjectively reported sleepiness in children with moderate SDB than controls.^7,8

There have been no published general population studies of EDS in children, as defined by parent and/or teacher report of sleepiness during the day, with objective sleep data. Therefore, our study is the first to report on the association between EDS and objective measures of sleep, demographic factors, health, and parent-reported sleep difficulties and emotional problems in a general population of young children. The purposes of this study were to (1) establish the prevalence of EDS, and (2) identify associations between demographic, emotional, and medical factors and the quantity and quality of sleep—measured objectively and by parent report—in young children with EDS.

METHODS

This study was designed in 2 phases. In Phase I, general information from parents about their child's sleep and behavioral patterns was collected. A screening questionnaire based on the survey published by Ali et al.,⁹ validated to identify children at high risk for SDB, was sent home to parents of every student (K-5th grade) in 4 local school districts (n = 7,312), with a 78.5% response rate. The procedure for Phase II of this study was initiated each year for 5 years by randomly selecting 200 children based on stratification for grade, gender, and risk for SDB from the current year's returned questionnaires. We studied 704 children in this phase. Four children did not complete the polysomnographic (PSG) recordings; thus 700 children out of 1000 children were included in Phase II, for a response rate of 70%. All children from Phase II who completed the Pediatric Sleep Questionnaire (PSQ) were included in this study. Children diagnosed with medical problems (37.8% allergies, 13.3% asthma, 1.2% juvenile diabetes), mental health disorders (11.1 % ADHD, 1.7% depression/anxiety, 0.8% autism), or a learning disability (9.1%) were not excluded from the study, so that the sample is representative of the general population. Thus, our final sample for this study consisted of 508 children from the Penn State Child Cohort. We contrasted the subjects who completed the PSQ and PSG with those who did not complete Phase II. There were no significant differences between the 2 groups on grade, sex, and risk for SDB. This study was approved by the Institutional Review Board of Penn State College of Medicine. Informed consent was obtained from parents of all participants, and assent was obtained from all children prior to participation.

Sleep Laboratory

During their visit in the laboratory, all subjects underwent a series of subjective and objective measurements. Height and weight were recorded for each child, and body mass index (BMI) was calculated. Waist circumference was measured.

All subjects were then evaluated for one night in sound-attenuated and temperature controlled rooms. During this time, the child's sleep was continuously monitored for 9 hours (24 analog channel and 10 dc channel TS amplifier using Gamma software, Grass-Telefactor Inc). A 4-channel electroencephalogram (EEG), 2-channel electrooculogram (EOG), and single-channel chin and anterior bilateral tibial electromyogram (EMG) were recorded. Throughout the night, respiration was monitored by thermocouples at the nose and mouth (model TCT1R, Grass Instrument Co., Quincey, MA), nasal pressure (Validyne Engineering Corp) and thoracic and abdominal strain gauges (model 1312 Sleepmate Technologies, Midlothian, VA). All-night recordings of hemoglobin oxygen saturation (SpO₂) were obtained using a cardiorespiratory oximeter (model 8800, Nonin Medical, Inc., Plymouth, MN) attached to the finger. Snoring sounds were monitored by a sensor attached to the throat (Sleepmate model, 1250). Our records were screened for sleep apnea using criteria that are currently used clinically.^10,11 An obstructive apnea was defined as a cessation of airflow of ≥ 5 sec and an out-of-phase strain gauge movement. A hypopnea was defined as a reduction of airflow of approximately 50% with an associated decrease in oxygen saturation (SpO₂) ≥ 3% or an associated arousal. Based on these data an apnea/hypopnea index (A/HI) was calculated [(apnea+hypopnea)/hours of sleep]. A long awakening was defined as ≥ 10 min.

Parent Rating Scales

For the purposes of this study, the Pediatric Sleep Questionnaire developed by Chervin¹² was completed by a parent in order to assess EDS in our study population. Children were classified as having EDS when the parent reported “yes” for “Does your child have a problem with sleepiness during the day?” and/or “Has a teacher or other supervisor commented that your child appears sleepy during the day?” The same definition of EDS was used in a recently published study by Tsaoussoglou et al.⁵ A parent also completed the Child Behavior Checklist (CBCL),¹³ which is a widely used tool for the assessment of childhood behavioral abnormalities. One of the 8 syndrome scales (anxious/ depressed) was used. In addition, a parent completed the Pediatric Behavior Scale, ¹⁴ which has norm-referenced T scores for several subscales including problems with sleep. Each item is scored on a 0 to 3 point scale, with 0 indicating no problems and 3 indicating that a behavior is very much or very often a problem. The PBS has been used in several studies by our group to assess sleep problems in children with autism and ADHD.^15–17

Statistical Analyses

The primary objective of the analysis was to evaluate the prevalence of EDS and associations with various risk factors, including SDB in a general population of young children. BMI was expressed as BMI percentiles (BMI %) adjusted for age and gender using the formula and data of the NHANES CDC growth charts.¹⁸ AHI was analyzed as a continuous variable. Univariate analyses of these data were initially conducted to compare those with and without a complaint of EDS with respect to various outcomes using t-test or χ² tests. Effect size (Cohen's d), P values, and odds ratios (ORs) ± 95% confidence intervals (CIs) based on the difference between the 2 groups are reported. Binary logistic regression was used for the multivariate analysis. The statistical confidence level selected for all analyses was P < 0.05. All analyses were performed using Predictive Analytics Software (PASW, Inc, Chicago, IL) Version 17.0.

RESULTS

The final sample of 508 children consisted of 431 children without EDS and 77 children with EDS. The age range was 5-12 years, with an average age of 102.0 ± 0.08 months. Approximately one-quarter of our sample was minority(African American, Asian and Hispanic ) as defined by a parent; (51.8% were boys, and 45% were from a professional family. The average AHI was 0.8 ± 0.06, with only 6 children with an AHI ≥ 5. The prevalence of EDS was 15.0% (Table 1).

Table 1.

Sample characteristics

	No EDS	EDS
	n = 431	n = 77	P
Gender (% male)	51	53	0.80
Age (y)	8.5	8.7	0.36
Race (%minority)	24	25	0.05
Professional status (%)	45	41	0.53
Waist (cm)	64.7	69.2	0.001
BMI %ile	62	70	0.02
AHI	0.76	0.87	0.54
Full Scale IQ	108	105	0.11

The distribution of demographic factors and potential risk factors for EDS is described in Table 2. Waist circumference, positive history of asthma, use of asthma medication, heart-burn, and parent reported symptoms of anxiety/depression were significantly associated with EDS. In addition, parent-reported symptoms of sleep difficulties (i.e., trouble falling asleep, restless sleep, and wakes often during the night) were also significantly associated with EDS. The parent-reported sleep difficulties remained significant even when controlling for waist circumference, asthma, and anxiety/depression. Caffeine intake (weekly) and history of allergies were not significantly associated with EDS. Objective sleep factors (Table 3) included AHI, minimum SpO_2, sleep latency, REM latency, total sleep time, number of long awakenings, sleep efficiency, number of arousals, and percent of REM, stage 1, 2, and slow wave sleep. None of the objective sleep factors were significantly associated with EDS.

Table 2.

Univariate associations between children with and without EDS

Risk Factors	Univariate	ES	P	Ors	CI
Health	Heartburn	0.35	0.008	3.1	1.4, 7.2
	Asthma	0.35	0.006	2.4	1.3, 4.3
	Asthma medication	0.41	0.002	2.9	1.5, 5.7
	Allergies	0.23	0.07	0.63	0.39, 1.03
	Caffeine consumption	0.05	0.58	1.2	0.66, 2.1
	Waist (cm)	0.43	0.001	1.04	1.01, 1.06
Objective Sleep	Total sleep time	0.03	0.82	1.00	0.99, 1.01
	Number of long awakenings	0.08	0.48	0.91	0.69, 1.2
	Sleep latency	0.03	0.82	1.00	0.99, 1.01
	REM latency	0.12	0.34	1.00	0.99, 1.00
	%Stage 1	0.02	0.90	0.99	0.93, 1.07
	% Stage 2	0.15	0.23	1.01	0.99, 1.04
	%SW	0.15	0.23	0.99	0.96, 1.01
	% REM	0.01	0.99	1.00	0.96, 1.04
	Arousal index	0.06	0.62	0.97	0.88, 1.09
	Min SpO2	0.10	0.45	0.98	0.91, 1.04
	AHI	0.08	0.55	1.04	0.91, 1.2
	Sleep efficiency	0.05	0.74	1.00	0.98, 1.04
Subjective Sleep	Trouble falling asleep	0.59	< 0.001	1.7	1.4, 2.3
	Restless sleep	0.46	< 0.001	1.6	1.3, 2.0
	Wakes often during the night	0.56	< 0.001	1.8	1.4, 2.3
Psychological	Depression and/or anxiety	0.48	< 0.001	2.9	1.6, 5.1

ES, Effect size Cohen's d

Table 3.

Objective sleep variables: Means for the children with and without EDS

	No EDS	EDS	P value
Sleep latency (min)	28.5 ± 1.2	29.2 ± 2.7	0.67
Total sleep time (min)	456.9 ± 2.4	458.3 ± 4.9	0.33
Sleep efficiency (%)	85.8 ± 0.41	86.1 ± 0.85	0.45
REM latency (min)	160.1 ± 3.2	152.4 ± 6.6	0.36
Stage 1 (%)	3.6 ± 0.16	3.5 ± 0.35	0.76
Stage 2 (%)	45.6 ± 0.56	47.4 ± 1.3	0.92
Slow wave (%)	31.2 ± 0.54	29.5 ± 1.3	0.79
REM (%)	19.7 ± 0.27	19.7 ± 5.8	0.92
Arousal index	3.1 ± 0.12	3.0 ± 0.26	0.42
SpO2 low* (%)	94.1 ± 0.18	93.8 ± 0.29	0.18

Mean and standard error.

^*Mean percentage of oxygen saturation during respiratory events.

In order to establish the relative independent contribution of these risk factors we further analyzed the data from a multivariate perspective using binary logistic regression. Four models were created. The most plausible theoretical predictors of EDS were tested beginning with metabolic factors and objective sleep difficulties, then subjective history of medical and psychological factors, ending with parent-reported sleep difficulties. The initial model included all objective sleep variables and waist circumference. The second model included all of Model 1 variables plus asthma and heartburn. Model 3 included all variables from Model 2 plus parent-reported anxiety/depression. Model 4 included all variables from Model 3 plus parent-reported sleep difficulties. Waist circumference and anxiety/depression remained independent predictors of EDS in Model 4, while asthma was eliminated from the model (Table 4). None of the objective measures of sleep, including sleep stages and AHI, were independently associated with EDS in any of the four models.

Table 4.

Risk factors for EDS based on multiple logistic regression

	Model 1		Model 2		Model 3		Model 4
	P	OR	P	OR	P	OR	P	OR
Waist circumference	0.003	1.04	0.01	1.04	0.003	1.04	0.004	1.03
Asthma			0.03	2.10	0.040	2.10	0.160	1.60
Depression/Anxiety					0.010	2.50	0.050	1.90
Trouble falling asleep							0.030	1.40

Waist circumference is a continuous variable; asthma, depression/anxiety, and trouble falling asleep are binary variables. Model 1: Waist circumference and objective sleep variables. Model 2: Waist circumference, objective sleep, and medical variables. Model 3: Waist circumference, objective sleep, medical, and depression/anxiety variable. Model 4: Waist circumference, objective sleep, medical, and subjective reported depression/anxiety and sleep related variables.

DISCUSSION

In our general population sample of 508 children, we observed a prevalence of 15% for EDS. Our study indicates that EDS is highly prevalent in children, a symptom that may adversely affect daytime functioning. Interestingly, independent predictors of EDS were waist circumference, parent report of anxiety/depressive symptoms and trouble falling asleep, as well as a history of asthma.

This study is the first to evaluate simultaneously a wide range of potential risk factors that included demographic, medical, psychological, objective and parent-reported sleep variables associated with EDS in a general population of young children (Penn State Child Cohort). This study suggests an association between childhood EDS and medical factors (i.e., heartburn, asthma), medication for asthma, waist circumference, and parent-reported anxiety/depression and sleep difficulties (i.e., trouble falling asleep, restless sleep, and wakes often during the night). Parent report of allergies, and objective sleep factors (AHI, minimum SpO_2, sleep latency, REM latency, total sleep time, number of long awakenings, sleep efficiency, number of arousals, and percent of REM, stage 1, 2, and slow wave sleep) were not significantly associated with EDS in our study. However, these objective sleep findings should be considered within the context of limitations that include the possible impact of first-night effect and a relatively low prevalence of children with moderate to severe SDB in a population sample (AHI ≥ 5) affecting our power.

To assess the relative contribution of various factors for the presence of EDS, we evaluated our data from a multivariate perspective. Waist circumference was the most strongly associated with EDS. This finding is consistent with previous studies,^2,4,5 demonstrating that obesity in children is independently associated with an increased risk for EDS, even in children with SDB. Our finding that waist circumference contributes to the independent prediction of EDS suggests that metabolic factors may play a contributing role in the mechanism of EDS, as others have reported in children and adults with SDB.^19–21 One study⁴ found that in children matched for SDB, EDS was linked to increased levels of inflammatory mediators (e.g., Interleukin-6, high sensitivity C Reactive Protein, Tumor Necrosis Factor 1), suggesting that pro-inflammatory cytokines are mediators of EDS in children similar to adults.²⁰ Most recently a study⁵ comparing obese children with mild to moderate SDB to obese children without SDB and lean controls, suggested that obesity and SDB were independently associated with EDS; and that inflammatory markers and leptin increased and adiponectin decreased in these obese children with SDB. EDS frequency increased progressively and significantly in the groups, with the lowest frequency in the lean group and the highest in the group with an AHI ≥ 5 (11%), in contrast to the results of our study, which only had 5 children with an AHI ≥ 5 (1%). This difference in the samples’ composition may help explain why SDB was not an independent risk factor in our study. An alternative but not mutually exclusive explanation is the potential influence of obesity on lung volume, which may have an impact on daytime sleepiness.

The second and third strongest independent risk factors in our multivariate analysis were parent-reported anxiety/depression and trouble falling asleep. The anxiety/depression finding is consistent with a recent study by Mayes et al.²² that suggests children with a clinical diagnosis of anxiety/depression had more daytime sleepiness than children with ADHD, autism, brain injury, and controls. Similarly, depression was independently associated with EDS in two studies of adults.^21,23 Our data suggest that EDS in this population of young children may be the result of anxiety/depression that should be appropriately evaluated and managed. The effect of anxiety and depression on EDS could be mediated, through the known effects of these conditions on the quality and quantity of sleep, or/and through the activation of physiological systems such as the stress system, that may result fatigue. Finally, the association of parent-reported trouble falling asleep with EDS is consistent with the fact that in adults nighttime sleep difficulties are associated with fatigue and sleepiness.^24,25

Parent report of wheezing/nocturnal asthma was the fourth strongest risk factor for EDS. When trouble falling asleep was added to the final model, asthma was eliminated. This suggests that parent-reported trouble falling asleep mediates the association between asthma and EDS. Thus, in children with asthma, trouble falling asleep may partially explain their symptoms of EDS. This finding is supported by a study that reported an increase in daytime sleepiness in children who wheeze or have asthma, with an association between a complaint of EDS and parent reported sleep disturbance.²⁶ We found no association, however, with any objective markers of sleep between those with and without asthma (data not shown). Our finding is compromised by the fact that one night in the lab may not be representative of the child's habitual sleep patterns or seasonal exacerbation of asthma symptoms. An alternative explanation is that the inflammatory process associated with a chronic respiratory disease (as already reported in children with obesity) or the side effect of asthma medications is the link to EDS.

Although EDS is commonly assumed to be the result of disturbed or inadequate sleep (quantity), it appears that in a large general population of children representing the typical range of SDB (i.e., mild), objective sleep was not related to EDS. Instead, the presence of EDS is more strongly associated with obesity, parent-reported depression/anxiety and trouble falling asleep, and asthma. Thus, from a clinical standpoint, professionals who evaluate and treat children with EDS should be cognizant of comorbid risk factors associated with daytime sleepiness. Although PSG is extremely useful in screening children for a number of sleep disorders (i.e., narcolepsy, seizures, parasomnias, SDB), in children with a parent complaint of EDS, parent information regarding sleep difficulties, particularly trouble falling asleep, may be more relevant. Primary lines of treatment might include weight loss if the child is overweight, treatment for underlying depressive and anxious symptoms, and implementation of nocturnal asthma prevention methods (e.g., making your bedroom free of allergens, such as dust mites and cigarette smoke, and using a humidifier in the house to keep the air warm and moist) if the child is diagnosed with asthma. Future research needs to determine if children with moderate to severe SDB (AHI ≥5) are at greater risk for EDS than children without SDB.

DISCLOSURE STATEMENT

This was not an industry supported study. The authors have indicated no financial conflicts of interest.