People with developmental disabilities sleep less and experience higher incidence of clinical sleep disorders than the general population. Exploring the neurophysiology linking sleep with daytime performance in patients with developmental disabilities is now possible using minimally sufficient sleep and sleep-sensitive behavioral assays. Although frequent sampling represents the primary difficulty, it is required to untangle coincident effects of sleep quality amidst circadian variation. Recent evidence finds high quality sleep promotes brain plasticity, improves health measures, and enriches quality of life. Sleep treatments for apnea, insomnia, restless limbs, and conditioned sleep-aversion are available, although not readily provided, for people with developmental disabilities. This population would gain both clinical and behavioral benefits as improved sleep-monitoring, behavioral testing, and sleep-treatment technology is adapted to their needs.
One hundred years of experimental research supports the statement that impaired performance results from a night of poor or inadequate sleep (Bjerner, 1949; Harrison & Horne, 1999; Kleitman, 1923; Patrick & Gilbert, 1896; Wilkinson, 1968). Optimal sleep, on the other hand, can improve performance on a few tasks, but the benefits require several days to manifest (Siegel, 2001; Stickgold, Hobson, Fosse, & Fosse, 2001). Good sleep and poor sleep also appear to effect measurably separate performance domains. Sleepiness, the usual by-product of poor sleep, is well-known to impair sustained attention and other cognitive tasks requiring explicit, willful intention (Dinges et al., 1999; Horne, 1993). Implicit cognitive capacities, such as the application of learned rules, motor capacity, and visual pattern recognition, improve more rapidly with high quality sleep after training (Karni, Tanne, Rubenstein, Askenasy, & Sagi, 1994; Smith & Wong, 1991; Walker, Brakefield, Morgan, Hobson, & Stickgold, 2002). Sufficient total sleep time and the juxtaposition of individual sleep states (e.g., light sleep, deep sleep, and REM sleep) define sleep quality. People with developmental disabilities tend to have both insufficient sleep and disrupted individual sleep stage expression. Both the endogenous neurologic pathology responsible for an individual's disability and exogenous circumstances are known to contribute to sleep problems for people who have developmental disabilities. What remains to be discovered is whether sleep problems lead to cognitive and behavioral problems and if sleep treatments reliably improve the mood and behavioral range of people with developmental disabilities.
Experimental total sleep deprivation experienced by healthy individuals degrades mood (Pilcher & Huffcutt, 1996), impairs psychomotor performance (Dinges & Kribbs, 1991), and decreases immune responses to endotoxins (Everson, 1995). Chronic, partial sleep deprivation (defined as less than 4 hours sleep per night) has recently been shown to impair performance and decrease glycogen sensitivity, suggesting that persistent sleep restriction may accelerate some processes of aging (Spiegel, Leproult, & Cauter, 1999). In a meta-analysis, Pilcher and Huffcutt (1996) calculated mean effect size (d) from 19 total sleep deprivation and partial sleep deprivation studies in which researchers compared healthy subjects undergoing experimental sleep loss to control subjects. The calculated d values represent the number of standard deviations (SDs) the experimental groups differed from controls. Mood, d = −4.10 ± 2.88, cognitive performance, d = −3.01 ± 2.88, and motor performance, d = −0.85 ± 1.33, were all significantly impaired by total sleep deprivation or partial sleep deprivation with chronic partial sleep deprivation producing the largest effect size decrements. Mood measures taken from people undergoing sleep deprivation were about 4 SDs, d = −4.10, worse than the mood of control subjects. Cognitive capacities, d = −3.01, were also seriously effected whereas motor skills were relatively preserved during sleep loss, d = −0.85. Together, it appears that chronic partial sleep deprivation, the laboratory equivalent of an untreated sleep disorder, produces both physiologic and cognitive impairments. Furthermore, these two categories of impairment may share physiologic sources that are themselves impaired by chronic sleep loss.
In recent neuroimaging studies, researchers have measured brain glucose use before and during periods of minimal sleep deprivation (24 hours) to explore neural responses to sleep deprivation. Sleep-deprived wakefulness (subjects quietly resting awake after one night awake) show neuronal glucose patterns that resemble those recorded during deep sleep (Thomas et al., 2000). Sleep-deprived subjects who were given a verbal learning task showed reduced blood flow in nearly all brain areas except for areas of the cortex known to be involved in verbal learning. These areas actually had an increase in blood flow during sleep-deprived testing compared to doing the same task when not sleep deprived (Drummond et al., 2000). Such an increase in blood flow, in the task-specific brain area, suggests that sleepy performance engages local compensatory neural responses against a background of reduced global brain activity. Neuropsychological tests of “divergent” thinking and executive planning, impaired during total sleep deprivation, suggest that the prefrontal cortex (part of the executive attentional network) is preferentially susceptible to sleep loss (Harrison & Horne, 2000). Because prefrontal cortex-related functions show decline with aging and psychiatric disease, Harrison, Horne, and Rothwell (2000) hypothesized that chronic sleep loss contributes to cognitive decline previously attributed only to aging.
To explore how chronic sleep insufficiency might interfere with human physiology, Spiegel et al. (1999) limited sleep to 4 hours per night for 6 days while measuring hypothalamic–pituitary–adrenal functioning and carbohydrate metabolism. Sleep-restricted subjects experienced increased sympathetic central nervous system (CNS) tone and decreased carbohydrate tolerance, both risk factors for type II diabetes. They also found that partial sleep deprivation increased peak circadian levels of cortisol and delayed them 1 hour after just the first night of partial sleep deprivation (Leproult, Copinschi, Buxton, & Cauter, 1997). Increased glucose intolerance and altered hypothalamic activity suggest that sleep loss increases effects of stress by producing excess cerebral glucocorticoids and increased insulin resistance, a clinical assay known to be associated with the development of type II diabetes. At this point, sleep disruption is known to result in homeostatic dysregulation within the CNS, along with behavioral problems.
New biobehavioral evidence continues to sharpen our understanding of the link between sleep, brain plasticity, and subsequent behavioral capacity. Long-term studies in humans are rare; typically short-term sleep deprivation studies report declines in both cognition and physiology sufficient to increase the risk of accidents and health problems. Most healthy people can offset transient sleep loss through increased cognitive effort and natural homeostatic physiologic responses. However, compensation becomes harder to recruit and maintain as chronic sleep loss accumulates or total time awake increases. People with developmental disabilities experience more fragmented sleep and a higher incidence of sleep disorders than do individuals in the general population, which limits their ability to recruit compensatory responses. Furthermore, neurologic limitations likely reduce the resources available for compensatory (or even homeostatic) responses to sleep loss or sleep disruption. For these reasons a reasonable and conservative hypothesis is that sleep problems in patients with developmental disabilities are beyond their ability to compensate for, which then exacerbates resident behavioral and/or health problems.
Sleep treatment improves subjective quality of life and objective behavioral capacity in most adults. Sleep treatments properly adapted to accommodate people with developmental disabilities and sleep disorders remains an unmet medical need. Chief among the challenges to measuring the success of clinical sleep evaluation, sleep treatment, and post-treatment behavioral testing is variability between individuals diagnosed with developmental disabilities. Within-subject testing to explore the functional relationship between sleep and daytime behavior requires high temporal sampling to account for circadian variation. Statistically, using intraclass correlations in this scenario, investigators can evaluate the matrix of performance on physiologic dependent variables accounting for between-subject variability. Methodologically, measuring how developmental disabilities alter the relationship between behavior and sleep can now be accomplished using unobtrusive sleep-recording technology and behavioral assessment in the patient's natural or usual environment. In this paper we review appropriate strategies for detecting the precise relationship between sleep and behavior for each diagnostic subtype of developmental disabilities using frequent, within-subject measurements testing sleep quality and cognitive capacity in a patient's natural environment.
Present information regarding the sleep of individuals with developmental disabilities is based primarily on reports by caregivers or observers using spontaneous or scheduled charting of observed behavioral states. Large observer-based samples have been obtained, and they appear to somewhat underreport sleep disruptions and overestimate total sleep time compared to studies using physiologic sleep state monitoring (Espie & Tweedie, 1991). Researchers using observer-based assessment of sleep for individuals with developmental disabilities tend to find less sleep and more arousals per night than that for age-matched controls without disabilities. A British survey of parents of 200 adolescents with severe mental retardation found sleep disruption to be very common compared to published norms (Quine, 1991). Half of the patients had trouble settling into sleep, 67% had observable nocturnal arousals, and 33% did not get sufficient sleep. Furthermore, two thirds of the children still had sleep problems 3 years later, and those with the most persistent sleep problems were most likely to have problems with academics, communication, aberrant behavior, and epilepsy. Survey-based data have documented increased prevalence and temporal durability of sleep problems for people with developmental disabilities who require caregiver support (Durand, 1998).
Disordered sleep is a common and often life-long problem among people with neurologic disabilities and their caregivers (cf. Quine, 1991). The incidence of sleep disorders and/or sleep disruptions in individuals with developmental disabilities ranges from 13% to 88%, depending on the study and its definition of disordered sleep (Bartlett, Rooney, & Spedding, 1985; Didden, Curfs, Van Driel, & De Moor, 2002; Harvey, Baker, Horner, & Blackford, 2003; Piazza, Fisher, & Kahng, 1996). By comparison, the National Commission on Sleep Disorders Research team, William Dement, Chair (National Commission, 1993) reported chronic sleep disorders in 12% to 25% of the general population. For people with developmental disabilities, the most common sleep problem is insomnia. Poindexter and Bihm (1994) reported that nearly 40% of 103 individuals with severe mental retardation slept less than 5 hours per night, which qualifies as chronic partial sleep deprivation compared to the general population average of 7.5 hours of sleep per night (Leger, 1994). Insufficient sleep in people who have developmental disabilities has been associated with increases in aggression, self-injury, and property destruction (Durand, Gernert-Dott, & Mapstone, 1996; Horner, Vaughn, Day, & Ard, 1996; Poindexter & Bihm, 1994). The importance of improving sleep is slowly gaining visibility in the general population, but sleep problems in people with developmental disabilities remain severely underreported and understudied (Stores, 2002).
Espie et al. (1998) recorded nocturnal EEGs for 28 individuals with profound mental retardation and seizure disorders and found that less than half of the subjects had any REM sleep, whereas those who did have REM sleep averaged less than 30 minutes of REM sleep per night (normal ≥ 110 minutes). Several other researchers have reported REM sleep in adults with developmental disabilities to be below average and density of eye movements during REM above average compared to typical subjects (Diomedi et al., 1999; Feinberg, 1968; Krynicki, 1976). Human performance on a visual learning task has been shown to require an abundance of deep sleep early in the night and an abundance of REM sleep late in the night after training (Walker et al., 2002). Co-measured sleep and performance data for individuals across developmental disabilities are not available in the current literature, largely due to prior technological limitations. Modest adaptations to new technologies should allow accurate measures of sleep stage patterns in people with developmental disabilities (Harvey & Kennedy, 2002). Such investments allow testing of several hypotheses, including the idea that increasing an individual's REM sleep will improve their capacity to learn.
Persistent sampling is required to measure sleep, moment-to-moment wakeful arousal, and behavioral performance in people with developmental disabilities. Modern recording technology is able to meet that need. Portable, multichannel, digital physiologic recorders combined with patterned behavioral acceptance is possible in most environments. Each person is best evaluated as their own control using a repeated measures testing design. Patients who can adapt to the recording equipment across all phases of the night and day will provide the most sensitive measure, allowing researchers to group data according to diagnostic category, thereby providing important clinical results. Long-term EEG monitoring of arousal state, such as time and duration of wake periods, sleep time (including daily naps), REM amounts, daily drowsiness, sleep efficiency, and sleep stage distribution can be assessed as multiple measures to explore how sleep interacts with subsequent wakeful behavior.
Portable polysomnography can be used to record moment-to-moment changes in the electroencephalogram—EEG (brain electrical voltage as a function of time) to quantify temporal patterns of circadian phase, sleep/wake and behavioral arousal, and structured cognitive–behavioral ability. Minute-to-minute coding of EEG, heart rate, and body movement during wake state can reveal time and duration measures of behavioral engagement or readiness to respond. Modern polysomongraphic equipment is now portable and computerized and can allow co-measurement of EEG and behavior for several continuous days. Sleep time, sleep efficiency, time in REM sleep, deep non-REM sleep, and the incidence of body movements provide the basic sleep research measures, which can also assist in clinical sleep-disorder diagnosis.
Current technologies have the potential to record EEG using unobtrusive electrodes and wires placed into protective head gear or preferential clothing worn by patients with developmental disabilities. Digitized data can also be recorded telemetrically, without the use of wires, which extends the testing of natural behavior immensely. Similarly, a number of behavioral measures (described below) are known to be sleep sensitive and adaptable to patients with developmental disabilities. Present computer programs are easy to learn, analyze collected data very rapidly, and provide organized data and feedback to scientists and caregivers. Sleep-stage expression depends on both external factors (e.g., a quiet and safe environment) and internal factors (e.g., good health and an absence of medical problems). Because the link between sleep and behavior is reciprocal, it is important to first establish that patients are not suffering from any treatable co-morbid sleep disorders.
A single night of polysomnography can be used to rule out several sleep disorders (apnea, narcolepsy, and movement disorders), although it is a clinical procedure not often employed for individuals with developmental disabilities. Results of studies in which researchers, to date, used nocturnal polysomnograms to record sleep in people with developmental disabilities are likely to underdemonstrate existent sleep disorders because clinicians and researchers reject recording artifacts (such as those caused by muscle movements) from sleep data before analysis (Clausen, Sersen, & Lidsky, 1977). Procedural and data integrity problems with sleep recordings have reduced the body of published data leading to a severe underestimation of clinical sleep problems in individuals with developmental disabilities.
High quality sleep is defined as an abundance of deep sleep (Stages III and IV) at the beginning of the night and abundant rapid eye-movement (REM) sleep at the end of the night. In two recent studies, researchers found that visual target detection had a high correlation, r = .89, with the amount of early evening delta (deep) sleep multiplied by the amount of late evening REM sleep (Stickgold, James, & Hobson, 2000; Walker et al., 2002). Sleep quality is degraded when these states are inadequately expressed or disturbed by frequent arousals out of sleep. Basic researcher focusing on EEG-defined sleep state and simple movement measures found that they are not sufficient for clinical interpretation per se. Heart rate, respiration, videotape, and several other measures are required for an accurate clinical assessment. Fortunately, individuals participating in long-term monitoring may be more amenable to clinical sleep testing and treatment if needed.
Perhaps 39% of adults with severe intellectual disabilities have sleep disruptions that meet clinical criteria for a diagnosis of insomnia (Poindexter & Bihm, 1994). The insomnia rate in otherwise healthy adults is estimated to be 11% (Bearpark, 1994). There are very few clinical studies designed to reveal the true numbers of people with mental retardation who have polysomonographically assessed insomnia. Of the estimated 39% of adults with severe intellectual disabilities and insomnia, very few receive the labor-intensive behavioral therapy for chronic insomnia (Hauri & Linde, 1996). Patients with developmental disabilities require modified, disability-specific versions of simple insomnia treatments (e.g., stimulus control therapy, bright light exposure) and new measures of treatment success.
There are limited reliable estimates for the incidence of sleep-related breathing disorders in adults with developmental disabilities. As many as 50% of children with mental retardation are estimated to have sleep-related breathing disorders and estimates rise to nearly 65% for patients with Down syndrome (Marcus, Keens, Bautista, Pechmann, & Ward, 1991). Developmental delay diagnoses that include respiratory problems and/or muscle atonia increase the likelihood of sleep apnea (Seddon & Khan, 2003). Sleep-related breathing disorders, like most other sleep disorders, are chronically underdiagnosed even in the general population according to a 1993 Sleep Commission Report to Congress. Underdiagnosis is even more severe in adults with developmental disabilities because they require modifications to clinical diagnostic procedures and treatments for sleep-related breathing disorders. Furthermore, daytime sleepiness, the primary symptomatic complaint of the patient, is assumed to be higher in people with developmental disabilities for reasons other than a sleep disorder such as sleep apnea.
Sleep apnea causes patients to completely stop breathing for 20 to 200 seconds as a result of upper airway collapse during the reduced muscle-tone periods that occur during sleep, most severe during complete muscular atonia of REM sleep (Oksenberg, Silverberg, Arons, & Radwan, 1999). Sleep apnea and sleep-related hypoventilation can cause repetitive blood desaturation and repetitive arousals from sleep. Repetitive blood oxygen desaturation forces the cardiovascular system into periods of hypertension to compensate for reduced oxygen availability. Sequelae include mild to severe cardiac arrhythmia and contribute to cardiac insufficiency problems found in hypertensive cardiopulmonary diseases (Koskenvuo, 1987).
Cognitive problems reported by sleep-related breathing disorders patients were once thought to occur secondary to reduced blood oxygenation during apneic events. Electroencephalographic arousal frequency was the recorded variable most correlated with reduced daytime cognitive functioning (Roth, Roehrs, & Rosenthal, 1995). Many people with sleep apnea, particularly early stage sleep apnea, have sleep arousals that do not include oxygen desaturation. Even mild snoring can disrupt sleep and predisposes a person to a diagnosis of sleep apnea, a disorder producing cognitive deficits that track disease severity (Engleman, Kingshott, Martin, & Douglas, 2000).
Periodic Arousal Disorders
Sleep arousals, like the fragmentation due to sleep apnea, can also occur secondary to periodic limb or body movements during sleep. Repetitive leg-kicking and arm-flailing are common examples of periodic limb movement disorder during sleep. Nearly all motor movements during sleep result in an arousal to wakefulness. Cumulative movement-related sleep disruption increases potential wakeful dysfunction. Bonnet and colleagues (1987) experimentally controlled sleep arousal to determine that the absolute number of sleep EEG arousals correlated with cognitive impairment and daytime sleepiness (Downey & Bonnet, 1987). Periodic limb-movement disorder typically fragments sleep once every 20 seconds, resulting in significant daytime sleepiness and cognitive impairments. Incidence of this disorder in the general population increases with age; about 5% of 30-year-olds and 44% of 65-year-olds report symptoms of periodic limb movement disorder during sleep (Montplaisir, Nicolas, Godbout, & Walters, 2000). Asthma, pain, and many other medical problems can also cause repetitive sleep disruptions (Kryger, Roth, & Dement, 2000).
Because arousal thresholds are higher in non-REM sleep, this state is better conserved during times of poor or limited sleep compared to REM sleep (Daan, Beersma, & Borbely, 1984); REM sleep is the dominant sleep state during the last quarter of a night. Circadian rhythms promote waking up late in the sleep period as does the elimination of sleep need that is (theoretically) being fulfilled during the night. Combined with any sleep arousal disorder, the amount and percentage of REM sleep is expected to be the most sensitive to disruption. Base rates of movement disorders during sleep have not been estimated for people with developmental disabilities, even though they have a high incidence of repetitive, stereotyped motor behaviors while awake. Chronic wakeful motor disturbances are thought to carry over into sleep for individuals with some types of developmental disorders. For example, nocturnal head-banging persists into adulthood for some people with autism (Chisholm & Morehouse, 1996).
Circadian Sleep Problems
Wing (1996) was one of the first researchers to report a summary of the sleep records of hospitalized infants with developmental disabilities and to find that children with autism develop irregular sleep patterns very early in life. Ferber (1996) later measured sleep-related brain-wave patterns in infants with similar disabilities and found that total sleep time was dramatically reduced from normal levels of 12 hours/day down to 7 hours/day in 3-year-olds who suffered from developmental delays, spasticity, sensory deficits, and autism. Sleep schedule problems so early in life suggest that poor sleep may be a consequence of developmental delays in the circadian timing system(s) of these patients. Accordingly, developmental delays in circadian systems might occur because these children receive inadequate environmental information due to poorly developed sensory and/or sensory relay systems (Okawa & Sasaki, 1987). Alternately, the fact that individuals with developmental disabilities have reduced sleep, polydipsia, hyperphagia, and dysthermia suggests hypothalamic dysfunction (Chaney, Olmstead, & Givens, 1994). Adults with developmental disabilities tend to suffer from insomnia dominated by polyphasia, the many-sleeps-a-day sleep pattern found in otherwise healthy newborns (Espie & Tweedie, 1991). In developmentally typical infants, polyphasia resolves naturally into a monophasic sleep pattern around the age of one year (Williams, Gokcebay, Hirshkowitz, & Moore, 1994). Developmental plateaus in sleep and circadian rhythm control may occur very early in these patients and reflect the developmental achievements in sensory, motor, and/or hypothalamic systems attained by each individual.
Today, 2% to 12% of all adults with intellectual disabilities are prescribed sleeping pills for their chronic insomnia that are intended to increase total sleep time (Singh & Winton, 1989). Adults without disabilities are typically prescribed sleeping pills when experiencing a rapid onset, stressor-induced, transient form of insomnia (Hauri, 1998). Persistent use of sedative drugs eventually worsens chronic insomnia due to the addictive nature of these drugs and the insomnia rebound that is typical following withdrawal (Kales & Kales, 1984). Furthermore, drug interactions and sedative side effects of sedative–hypnotics can increase the incidence of sleep-related breathing disorder for some patients (Robinson & Zwillich, 1994). Such complications may occur more often in patients who have developmental disabilities and take a large number of medications. This is a very serious concern because patients with developmental disabilities appear to be more susceptible to nocturnal breathing disorders than are otherwise healthy people of the same age.
A broad survey of families who have a child with mental retardation (641 two- to twenty-year olds and 222 adults) found that 5.3% and 7.9%, respectively, were receiving major tranquilizers; 28.5% and 24.5%, respectively, drugs with a sedative function; and 53.4% and 52.7%, anticonvulsants (Hogg, 1992). Parental reports claimed major tranquilizer use produced more behavior problems, whereas sedative and anticonvulsant use increased reports of sleep problems and the occurrence of epileptic seizures. Some form of pharmacological treatment will always be warranted for a number of patients, but Morin and colleagues (1999) found that behavioral therapy was better than pharmacological treatments (specifically sedatives) for treating insomnia in patients without disabilities. The use of sedatives to combat the high rates of insomnia decreases REM and delta sleep and ultimately causes rebound insomnia after drug discontinuation (Kales, Bixler, Tan, & Scharf, 1977). Behavioral methods for treating insomnia may prove difficult to adapt, but graded application of stimulus control therapy and improved sleep hygiene can be modified for people with developmental disabilities. Both strategies limit internal and external stimulation in order to facilitate natural sleep (Durand, 2002; Durand & Christodulu, 2004; Piazza, Hagopian, Hughes, & Fisher, 1998).
Treatment of sleep apnea for people with developmental disabilities requires sustained individual care during the adaptation period to either durable medical equipment or surgical recovery as indicated for each patient. Tirosh, Tal, and Jaffe (1995) applied continuous positive airway pressure to young patients with mild mental retardation, which resulted in improved sleep, increased wakeful arousal, and a decrease in seizure episodes. Continuous positive airway pressure requires the patient to wear a nasal mask attached to a pressurized tube to maintain airway patency throughout the inspiration/expiration cycle, even during periods of reduced muscle tone in sleep. Sleep apnea is also treated in pediatric patients with developmental disabilities who cannot comply with continuous positive airway pressure via surgical alterations of the upper airway (Powell, Guilleminault, & Riley, 1994). This surgery involves removing excessive tissue in the upper airway and/or removing unnecessary glands in the throat. Oral appliances alter the position of the mandible during sleep mechanically to enlarge the upper airway and thereby reduce obstructive sleep apnea in some patients who cannot handle continuous positive airway pressure (Lowe, 2000). Surgical solutions to sleep apnea may be more difficult to implement because they impose additional medical risks to the patient's life, and recovery can be very painful and requires delicate healing (Riley, Powell, Li, & Guilleminault, 2000).
Otherwise healthy people, able to resolve their sleep-related breathing disorders, typically report enormous increases in physical and mental capacity and endurance (Engleman, Martin, Deary, & Douglas, 1994). Tirosh et al. (1995) reported on 4 individuals with developmental disabilities whose sleep apnea was resolved with continuous positive airway pressure. They had improved sleep, higher alertness, and improved health. Degraded health status appears to accompany undiagnosed sleep related breathing disorders. Ronald, Delaive, Roos, Manfreda, and Kryger (1998) found that during the 10 years prior to diagnosis, people with sleep-related breathing disorders had more hospital visits and medical expenses than did age-matched controls. Overweight individuals are at a higher risk for sleep apnea, and weight loss usually reduces its severity (Schafer et al., 2002; Siegfried, Siegfried, Rabenbauer, & Hebebrand, 1999).
Optimal Sleep Scheduling
Treatment of circadian rhythm disorders is often successful, with individual success typically measured in a matter of weeks. Based on the disparity between typical treatments and the specific needs of most people with developmental disabilities, one approach is to adapt the successful sleep treatment strategies used for people with physical disabilities. Uehara and Kobayashi (1986) found a reduction in polyphasic sleep patterns following vestibular and muscular sensory stimulation in people who used wheelchairs and did not have cognitive deficits. Espie and Wilson (1993) offered a widely applicable program called “optimal scheduling treatment,” which is designed to achieve 90% sleep efficiency through the maintenance of a constant waking time each morning and monitoring the subject for optimal sleep-readiness each night. Espie and Wilson relied on cues of individual “sleep readiness” (expected around 16 hours after daily wakeup time) to help structure an optimal sleep and wake schedule. Phototherapy is a noninvasive method of resetting the major circadian pacemakers in individuals of all ages. They passively receive high doses of light (10,000 to 50,000 lux) for a few hours a day, which may also prove an effective intervention for people with developmental disabilities who experience disordered sleep. Adults with mental retardation who live in Finland have been observed to have less fragmented sleep during seasons with longer photoperiods (Lindblom, Heiskala, Kaski, Leinonen, & Laakso, 2002).
Good sleep promotes learning and adaptive behavior, apparently by facilitating neural plasticity. Conversely, sleep disruption and/or deprivation contributes to accidents, impairs learning, and/or probably limits synaptic plasticity. Sleep problems in people with developmental disabilities are seriously underreported and, thereby, undertreated (Durand, 1998, Stores, 2002; Wiggs & Stores, 1996b). Treating sleep disorders in otherwise healthy adults improves mood, performance, and eradicates sleepiness. Sleep assessment and sleep disorders treatment improve health and well-being. Few guidelines are available to facilitate clinical and research strategies to assess and intervene with the physiologic sleep of people with developmental disabilities.
During sleep, the nervous system undergoes a number of state changes that are as different from each other as each is from wakefulness. Furthermore, these experientially distinct states of neuronal activity each offer unique functional advantages to sleepers. Sleep expression is a malleable biobehavioral process responsive to changes in environment, medical condition, pharmacological agents, and personal psychological constructs (Kryger et al., 2000). Behavioral capacity degrades when sleep is insufficient or disrupted. Sleep quality for people with developmental disabilities can now be improved using modern sleep assessment and treatment strategies. Caregivers, clinicians, researchers, and patients with developmental disabilities using new tools for assessment and treatment can improve sleep sufficiently to improve mood, health, and cognition. Persistent good sleep interpolated with training may well improve the capacity for learning and, thereby, improve the quality of life of people with developmental disabilities.
Sleep-Sensitive Behavioral Assessment
An ability to sustain attention, become instrumentally engaged with others, complete tasks, and produce an elevated mood should all improve following several nights of high quality sleep. Behavioral covariation with sleep quality in healthy subjects has been seen in measures of attention, memory, social skills, verbal skills, motor procedures, and mood. Behavioral methods to identify and quantify periods of focused attention in people with mental retardation have been an essential component of behavioral therapy for decades. Ault, Guy, Guess, and Bashinski (1995) used an observer-based intervention model to detect and reduce periods of inattentiveness or sleepiness. Their “ABLE” method schedules teaching during periods of maximal receptivity and uses a form of stimulus control to reduce external interference with focused attention.
Frequent sampling is required to accurately account for daily peaks and valleys of mood and performance within a day. For example, changes in sleep might only produce measurable changes in performance during naturally occurring peaks or troughs of daily alertness and behavioral engagement. Significant circadian variation in arousal, mood, and behavioral capacity have been reported, with people at their worst in the morning improving slowly throughout the day (Dijk, Duffy, & Czeisler, 1992). Similarly, sleep state changes are very dynamic and misplaced sleep (naps) can interfere with optimal nocturnal sleep. Sustained attention and alertness gains follow a short nap, so napping is sometimes of benefit to people who need extra sleep. Persistent sampling is also critical because some sleep changes due to treatment or training can require several days to fully impact sleep measures.
Human, as well as animal studies, have established that some forms of learning depend on high quality sleep immediately after training (Smith, 1995; Smith, Young, & Young, 1980). High quality sleep is defined as an abundance of deep sleep (Stages III and IV) at the beginning of the night and abundant rapid eye-movement (REM) sleep at the end of the night. In a classic study, Karni, Tanne, Rubenstein, Askenasy, and Sagi (1994) found that human subjects demonstrate perceptual learning not only after adequate posttraining sleep but after quiet rest or disturbed sleep of the same duration. In two recent studies, Stickgold et al. (2000) and Walker et al. (2002) found that nonfoveal visual target detection improvement has a high correlation, r = .89, with the amount of early evening delta (deep) sleep coupled with abundant early morning REM sleep late in the same sleep period.
Subjects undergoing sleep deprivation experiments typically provide self-assessed mood levels using numeric or visual ratings. Very simple rating scales can be used as long as they have enough meaning to encourage people to accurately discriminate how to quantify mood and avoid rating repetitions near the middle of the scale. Recently, measuring the mood of people with developmental disabilities has been a research focus, and results document associations between poor sleep and poor mood as well as between poor mood and problem behavior (Carr, Magito-McLaughlin, Giacobbe-Grieco, & Smith, 2003). Somatic complaints provide another measure of subjective mood and the intensity of somatic complaints is known to be an introspective assessment that varies as a function of time awake (Dinges et al., 1997). A wide variety of rehabilitation-based measures of self-motivation, happiness, or responsiveness to stimuli can be used to index mood. Koegel and Koegel (1988) used “responsivity” to index both the willingness of the patient to interact with external stimuli and to highlight opportunities for learning to overcome prior learned helplessness due to behavioral failure. Their measure of responsivity shows good interobserver agreement and that a measure of mood is useful when self-report mood cannot be measured directly.
Simple tasks of sustained attention are the most sensitive measure of sleep loss, and testing this skill began in the earliest days of sleep research (Kleitman, 1923; Lee & Kleitman, 1923). Repeated testing with high temporal frequency (once every 2 to 4 hours) is necessary to assess behavior because it fluctuates as a function of time of day (Dijk et al., 1992). Simple vigilance tasks require almost no learning, making them readily adaptable while allowing performance comparisons across time with minimal retesting confounds. In clinical and research testing of sustained attention, investigators use a computer to present stimuli while recording response speed and accuracy. Any task demanding attention to a singular task and quick responses offers sustained attention data. Patterned responding to a target stimulus in a stable environment can be used to measure sustained attention as long as the motor response itself is not fatiguing and the target stimuli can be easily identified. Response slowing, error rates, and larger vigilance decrement (or distraction away from the task) can provide sleep-sensitive measures of behavioral response. Improving sleep should result in faster reaction times, improved accuracy, and the ability to stay on task longer.
Task engagement is expected to improve immediately with increased total sleep time and sleep efficiency. Although measurable benefits should be seen immediately, some individuals may need to experience an increase in total sleep time for several nights in order to remedy any backlog of sleep debt. People are known to “get behind” on their sleep, and sleep debt is a convenient way to describe cumulative sleep insufficiency across time. The cumulative duration of sleep less than 7 hours a night for any given week provides a rough estimate of the sleep debt they are carrying. People with sleep debt allowed unrestricted time to sleep will produce measurable sleep rebound, transient increases in total sleep time, until the sleep debt is moderated. Good sleep maintained after “paying off” the sleep debt would be expected to decrease errors and reaction times, especially during morning testing. Adequate baseline performance is needed to help moderate learning and transient arousal effects due to testing and serve to fully explain individual variance before exploring the effects of sleep change.
One hypotheses to explain the neurology of sleep loss-induced performance deficits suggests that sleep-impaired behavior results from prefrontal cortex dysfunction (Harrison et al., 2000). Cells in the prefrontal cortex activate selectively when working memory and/or executive processes are required to coordinate simultaneous or serial cognitive demands (Miller & Cohen, 2001). Sleep deprivation selectively reduces glucose use in several distinct brain regions (e.g., prefrontal cortex) supporting cognition (Thomas et al., 2000). Neuropsychological subtests provide the best measure of executive cognitive ability and provide the most sensitive measure for tracking problems encountered due to sleep disruption. Unfortunately, very few validated neuropsychological tests have been adapted for use with people who have developmental disabilities.
One validated test, the Wechsler Preschool and Primary Scales of Intelligence, Revised (WPPSI-R), has been used to track high-level cognitive functioning in young people with developmental disabilities (Gerkin & Hodapp, 1992). Executive functioning in people with developmental disabilities or low IQ can also be measured using daily tasks, such as making a sandwich or wrapping a gift. Multistep tasks that require planning, memory recall, ordering of elements, and following rules appear to require intact prefrontal cortex functioning (Schwartz, 1995). Ordered tasks are expected to be accomplished faster with fewer errors following good sleep, and perseveration errors and alternate solutions are expected to follow poor sleep (Harrison & Horne, 1998). The ability to maintain strategic context is altered during drowsy wakefulness by interfering with the ability to incorporate new information into strategies. Other researchers have found that drowsy behavior during a simple tone-detection task revealed the abandonment of strategy in favor of exclusively automatic cognitive-processing (Doran, 1999).
Motor control, task mastery, and errors in hand–eye coordination or gross-motor body skill have been examined for effects of sleep on training and/or performance. In their meta-analysis, Pilcher and Huffcut (1996) found a trend towards impairment of motor skills during periods of reduced sleep. Mirror drawing and sequenced key-pressing in motor tasks has been shown to improve if subsequent REM sleep is elevated compared to baseline (Walker et al., 2002). Sequence learning can be used to evaluate acquisition of implicit capacities, such as the ability to use motor sequences that cannot be explicitly described or when a practiced skill becomes automatic (Posner & Snyder, 1975). Sequence training finds a stepwise improvement in both speed and accuracy with passing time that increases with the amount of REM sleep in both the young and in adults (Smith, 2001). Human neuroimaging during the sleep immediately after sequence training shows brain activity patterns during REM sleep significantly simulating patterns recorded during training (Laureys et al., 2001; Maquet et al., 2000). Motor control and the eventual mastery of new skills appear to require abundant REM sleep after training, so minimizing sleep disorders should first facilitate implicit learning, a behavioral prediction that still requires direct testing.
Concurrent changes in sleep and learning have not been directly tested (or at least reported) in people with developmental disabilities. Developmentally typically children exhibit an abundance of REM sleep (about 35% at 3 years), and REM sleep is consistently found to be reduced in people with developmental disabilities when compared to age-matched control individuals (Diomedi et al., 1999). Deficits in deep REM sleep are also known to characterize the sleep of children with developmental disabilities (Shibagaki & Kiyono, 1983; Shibagaki, Kiyono, & Matsuno, 1985), suggesting that REM sleep is not selectively impaired but may be part of the tendency towards insomnia found in people with developmental disabilities. Insufficiency of REM sleep correlates with some level of retardation in some subjects (Espie et al., 1998; Horita, Kumagai, Endo, & Niwa, 1983). Latency of REM, the time it takes to enter REM sleep, appears normal in individuals with mental retardation (Shibagaki, Kiyono, & Takeuchi, 1987), suggesting that the neural mechanisms driving sleep stage organization are intact while those controlling REM sleep are impaired.
Children with developmental disabilities who have sleep disorders may be limiting REM sleep expression during a critical developmental period for basic cognitive capacities. Preterm infants with fewer rapid eye movements during REM sleep showed lower scores on developmental scales at both 12 and 24 months of age (Scher, Steppe, Dahl, Asthana, & Guthrie, 1992). Individual capacity for REM sleep expression might more or less depend on nascent neurology or manageable sleep disorders. Established links between REM sleep and learning will require increases in REM sleep during training compared to baseline and after task mastery. As noted above, increased REM expression is predicted to occur concurrent with improved implicit cognitive capacities, such as rule learning.
Thompson, Hackenberg, Cerutti, and Baker (1994) used observations every half-hour to measure the sleep of 8 people with severe to profound mental retardation and self-injurious behavior (SIB) living in an institution as part of a pharmacological study. During baseline, matched-patient controls from the same institution slept an average of 6.72 hours per night, while those with SIB slept an average of 5.34 hours a night. This significant difference supports other observer-based findings that both problem behaviors and physical health of patients with SIB correlates with sleep time (McDermott et al., 1997; Wiggs & Stores, 1996a). Sleep deprivation has been identified as an establishing operation, increasing the likelihood of problem behaviors exhibited by individuals with developmental disabilities on subsequent days (Horner et al., 1996; Kennedy & Meyer, 1996; O'Reilly, 1995), especially negatively reinforced aggression and self-injury (Kennedy & Meyer, 1996; Lancioni, O'Reilly, & Basili, 1999; Symons, Davis, & Thompson, 2000).
Research and Clinical Applications
Although current behavioral treatments of sleep problems have not reached a critical mass and, thus, have not been established as efficacious (Schreck, 2001), it is important to establish clear a priori assessment and treatment guidelines. Future treatments should be based on multimeasure objective data rather than solely on secondary reports. Stores (2002) recently summarized the incidence of sleep problems in people with developmental disabilities in order to help provide differential sleep diagnosis when evaluating patients who have developmental disabilities. The experimental goals and research strategies we discussed in this paper support the trend for wider clinical sleep diagnosis and improved sleep care. Concurrent sleep and cognitive testing can simultaneously extend basic and clinical knowledge while treating a newly discovered sleep patient. Mood, learning, and skill performance are expected to respond to treatment of disordered sleep in people who suffer both sleep problems and have developmental disabilities. Unique behavioral limitations presented by each person with developmental disabilities are likely met by focusing on the few essential physiologic (EEG, EMG, EOG, body movement) and performance (speed and accuracy) variables. Electroencephalographically measured brain activity and motor measures are sufficient to assess sleep and wake states (i.e., active, quiescent, focused, deep REM sleep, REM sleep). Direct measurements of respiratory, heart rate, and limb movements are essential to extend basic sleep recording to individuals with developmental disabilities and assess clinical sleep disorders, such as sleep-related breathing disorders, REM sleep disorders, nocturnal seizures, movement disorders during sleep, and general insomnia.
Authors: Scott M. Doran, PhD (firstname.lastname@example.org), Senior Research Biologist, Merck & Co. Wp4-262, West Point, PA 19486 (address correspondence to 1707 Williams Way, Jeffersonville, PA 19403). Mark T. Harvey, PhD, Research Assistant Professor, Department of Special Education, Box 40, Vanderbilt University, Nashville, TN 37203. Robert H. Horner, PhD, Professor, Special Education, 1761 Alder St., Rm. 107, 5271 University of Oregon, Eugene, OR 97403-5271