Theme : Dreams

Sleep Spindles and Arousal: The Effects of Age and Sensory Stimulation

R. T. Pivik1, S. Joncas2 and K. A. Busby1
1Department of Psychiatry, University of Ottawa and The Ottawa Hospital, General Campus, Ottawa, Ontario, Canada, K1H 8L6 and 2Centre d’etude du Sommeil, Hopital du Sacre-Coeur, Montreal, Quebec, Canada, H4J 1C5
Abstract
This study assessed a proposed sleep-preserving role for sleep spindles by evaluating variations in this activity as a function of factors, both naturally occurring and experimentally induced, known to affect and effect arousal from sleep. These factors included age, auditory stimulation, and experimentally induced arousal from sleep. Analyses were based on data from 84 males (5-49 yrs. old) from normal and clinical (hyperactive, enuretic, and chronic pain) populations who had participated in sleep auditory arousal threshold studies involving adaptation and 1-2 experimental nights. Spindles on experimental nights were visually analyzed and incidence determined for the two minutes preceding and throughout all Stage 2 arousal attempts. Prestimulation spindle occurrence in 39 preadolescent subjects with two experimental nights did not vary significantly from night-to-night, and prestimulation period comparisons between clinical groups and their respective controls were also non-significant. Anticipated relationships between spindle activity and indices of arousal-either inverse with respect to known variations in arousal threshold, i.e., decreases with age and across the night, or direct with respect to stimulus intensity particularly on trials when arousal did not occur-were not observed. Instead, all age groups showed significant decreases in spindle density as an increasing negative function of stimulus intensity. These findings suggest that to the extent to which sleep spindles can be considered to play a role in sleep preservation by inhibiting or attenuating potentially arousing stimuli, these effects appear to be restricted to endogenously generated stimuli and are passive rather than reactive in nature.

Current Claim: To the extent to which sleep spindles play a role in sleep preservation by inhibiting or attenuating potentially arousing stimuli, these effects appear to be restricted to endogenously generated stimuli and are passive rather than reactive in nature.

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
Sleep spindles, first described in the human electroencephalogram (EEG) by Loomis et al. (1935), are 12-14 Hz waxing and waning oscillations lasting 0.5 seconds or more (Rechtschaffen and Kales, 1968; Silverstein and Levy, 1976) which recur every 3 to 10 seconds (Steriade et al., 1993) during non-rapid-eye-movement (NREM) Stages 2-4. This activity, thought to be generated by interacting thalamocortical processes (Steriade, 1993a), is most readily recorded from central (McCormick et al., 1997) or centro-parietal (Zeitlhofer et al., 1997) cortical placements and in humans is most prominent during Stage 2 NREM sleep (Rechtschaffen and Kales, 1968; Dijk et al., 1993; Bové et al., 1994). Spindle incidence is characterized by wide inter-subject variability-evident even at the earliest stages of development (Tanguay et al., 1975)-but night-to-night consistency (Silverstein and Levy, 1976; Di Perri et al., 1977b, c; Shirakawa et al., 1978; Zeitlhofer et al., 1997). However, investigations of within-night variations disagree as to whether spindle activity remains constant (Di Perri et al., 1977c; Gaillard and Blois, 1981; Dijk et al., 1989; Bové et al., 1994), increases (Goetz et al., 1983; Guazzelli et al., 1986; Uchida et al., 1992), or decreases (Keane et al., 1977; Bové et al., 1994; Zeitlhofer et al., 1997) across the night. Since delta activity is concentrated early in the night (Williams et al., 1964, 1966), a reported reciprocal relationship between delta and sigma (which includes spindle activity) frequencies (Uchida et al., 1992, 1994; Dijk et al., 1993) would be consistent with an increase in spindle activity across the night.
Investigations of gender-related variations in spindle incidence have produced inconsistent results with respect to male-female comparisons, with findings of no differences (Goetz et al., 1983; Bové et al., 1994) as well as reports that females exhibit more spindles than males (Wu et al., 1980; Gaillard and Blois, 1981). Spindle incidence has been shown to be quite variable across the menstrual cycle (Driver et al., 1996).

Spindle activity has been studied in a variety of neurologic, behavioral, and sleep-related clinical conditions. A review of findings in various neurologic conditions noted spindle increases in association with motor disorders, but decreases, often localized, were noted in cases of cerebral pathology involving tumors or following selective cortical surgical procedures (Jankel and Niedermeyer, 1985). In children diagnosed with hyperactivity, a behavioral disorder with suspected maturational and motor control deficiencies, Khan and Rechtschaffen (1978) observed that spindle incidence was reduced in unmedicated subjects, but increased when subjects were treated with methylphenidate. Poitras et al. (1981), however, found more spindles in unmedicated hyperactive males than normal controls. Spindle density, evaluated in patients with affective disorders to determine if there is a relationship between this variable and sleep complaints in these subjects has been reported to be at control levels in prepubertal depressives (Goetz et al., 1983), but reduced in adult patients with bipolar or unipolar depression (de Maertelaer et al., 1987).

Studies charting the developmental time course of sleep spindles have indicated that rudimentary spindles can be detected in the EEG of full-term infants at 4 weeks of age, become fully developed within 3 months, attain maximal density and duration values between 4 and 6 months (Metcalf, 1969; Tanguay et al., 1975; Louis et al., 1992), and then decrease in incidence until the fifth year of life (Tanguay et al., 1975). Gestational age may be an important determinant of this activity since those born prematurely have reduced spindle activity relative to full-term infants (Wu et al., 1980). These observations and findings of abnormal spindle activity in children afflicted with conditions such as mental retardation (Gibbs and Gibbs, 1962; Shibagaki et al., 1982), perinatal injury (Monod et al., 1977), congenital hypothyroidism (Lenard and Schulte, 1974), PKU (Poley and Dummermuth, 1968), and autism (Ornitz, 1972), have led to the suggestion that sleep spindles may serve as an index of neural maturation (Tanguay et al., 1975; Shibagaki et al., 1982).

Studies comparing spindle incidence in children and adolescents with that in adults have reported both fewer (Smith et al., 1979; Principe and Smith, 1982) and more (Nicolas et al., 1998) spindles in younger subjects. Spindle density comparisons among young and middle-aged adults and the elderly have generally (Guazzelli et al., 1986; Dijk et al., 1989; Wauquier, 1993; Landolt et al., 1996; Nicolas et al., 1997,1998), but not consistently (Keane et al., 1977; Bové et al., 1994), indicated a decrease in this measure with age. One investigation (Di Perri et al., 1977a) examining various spindle-related parameters in young and aged adults reported similar spindle density values for both groups when this measure was derived from right frontal EEG recordings, but a significant relative decrease in the aged subjects was reported when the measure was based on left frontal recordings.

There is evidence to suggest species-specific variations in the occurrence and correlates of spindle activity. For example, there are multiple reports indicating that relative to their occurrence in humans, spindles are not only less prominent in several types of non-human primates, but are largely absent in the adults of these primates (Caveness, 1962; Hughes and Mazurowski, 1964; Bert et al., 1970a, b; Balzamo, 1980). Spindles are prominent during sleep in cats, but unlike the general decrease in incidence with age reported in the human, Bowersox et al. (1985) found that spindle occurrence was enhanced in older compared to younger animals.

Speculation regarding the functional significance of sleep spindles has focused on inhibitory correlates of this activity observed in studies conducted at both cellular and behavioral levels. Early investigations detailing the neuronal basis for sleep spindles emphasized the contribution of inhibitory postsynaptic potentials to the generation of this activity (Andersen and Sears, 1964; Andersen et al., 1967), and these observations have been confirmed and extended by Steriade and his colleagues (Steriade and McCarley, 1990; Steriade, 1993a; Steriade et al., 1993). Consistent with these inhibitory features of spindle activity are reports of diminished auditory evoked responses (Yamadori, 1971; Elton et al., 1997) and reduced motor activity and excitability (Hongo et al., 1963; Ehrhart et al., 1981; Sterman and Bowersox, 1981; Pivik and Bylsma, 1982) in association with sleep spindles. These findings have been interpreted as indicating a sleep preserving or protective role for sleep spindles (Johnson et al., 1976; Naitoh et al., 1982; Bowersox et al., 1985; Steriade and Amzica, 1998).

If spindles do serve a sleep preserving function, then an inverse relationship between spindle activity and arousal during sleep might be expected. The findings of reduced sensorimotor activity and excitability in association with spindle activity are suggestive of such a relationship, as are reports of greater spindle density in conditions of hypersomnia (Bové et al., 1994) and following the administration of hypnotics (Johnson et al., 1976). However, there are also reports apparently inconsistent with the arousal-suppressing effect of spindles. For example, Church et al. (1978) found that evoked K-complexes were potentiated during spindles, and studies in the elderly have failed to observe a negative correlation between spindle density and the amount of waking in all-night sleep recordings (Feinberg et al., 1967; Guazzelli et al., 1986). Based on the spindle-facilitating effect of hypnotics Johnson et al. (1976) hypothesized that "... sleep spindles serve to raise the arousal threshold which, in turn, helps to maintain sleep by reducing the probability of being awakened" (p.74). Subsequent studies by this group and others indicated that hypnotics did elevate both arousal threshold (Bonnet et al., 1979; Johnson et al., 1979a) and spindle density (Johnson et al., 1979b). However, referring to data from an unpublished study by their group in which these measures were examined in the same subjects, Church et al. (1978) reported the absence of a significant positive relationship between spindle rate and average arousal threshold in subjects receiving placebo or a hypnotic. From these data and their finding of K-complex potentiation during spindles, these investigators argued that spindles were not inhibitory, but instead "reflect a phasic reduction in inhibitory action" (p.451). With the exception of this reference in the Church et al. (1978) publication, the relationship between spindle incidence and experimentally induced arousal from sleep has not been directly addressed.

The literature detailing the correlates of sleep depth has established several subject and sleep parameters relevant to the spindle-arousal relationship. This literature is based primarily on studies similar to those of Johnson et al. (1979a) in which the susceptibility to arousal from sleep has been evaluated by determining the intensity of auditory stimulation required to awaken subjects during different sleep stages and at various times during the night. These auditory arousal threshold (AAT) studies have reported higher thresholds in Stage 4 relative to Stage 2 or REM sleep (Goodenough et al., 1965; Bonnet and Moore, 1982; Busby et al., 1994), and decreases in threshold across the night (Rechtschaffen et al., 1966; Keefe et al., 1971; Busby and Pivik, 1985; Busby et al., 1994) and as a function of increasing age (Busby and Pivik, 1983; Zepelin et al., 1984; Busby et al., 1994). The extent to which spindle incidence covaries with the results of these studies in a way consistent with the postulated sleep-preserving function of spindles has not been determined, and the previously noted inconsistencies in the literature regarding variations in spindle activity as a function of ontogenesis and time-of-night further obfuscate the nature of this relationship. The purpose of the present study was, therefore, to address issues related to this relationship by examining variations in the incidence of sleep spindles as a function of age, sensory (auditory) stimulation, and experimentally induced arousal from sleep. The investigation was based on an extensive dataset derived from four arousal threshold studies (Busby and Pivik, 1985; Busby et al., 1994; Wolfish et al., 1997; Harman et al., unpublished observations) using similar methodologies and involving developmentally diverse subjects (children, preadolescents, adolescents, young adults and adults), both normal and presenting with clinical symptoms (hyperactivity, enuresis, chronic pain syndrome). Since there are no persuasive data indicating variations in spindle activity unique to any of these clinical populations, hypotheses were based on age and arousal-related parameters only. Consistent with the arousal threshold data in humans and the hypothesized sleep preserving function of spindle activity, it was expected that spindle incidence would: be higher in younger subjects; decrease with age and across the night; increase in response to sleep-disturbing auditory stimulation; and be intensified on non-arousal trials relative to those on which subjects awakened.

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.

Figure 1

Subjects
Eighty-four males ranging in age from 5-49 years were studied, including 48 normal control (6 children, 18 preadolescents, 10 adolescents, 10 young adults, and 4 adults), 15 preadolescent enuretic, 8 medicated and 8 non-medicated preadolescent hyperactive, and 5 adult chronic low back pain subjects. None of the subjects had medical, psychiatric, or sleep disorders other than those which defined the clinical groups, and all subjects had an audiometric evaluation to ensure normal hearing. Adult subjects signed a consent form. Parental consent and approval were obtained for younger subjects. Table 1 presents more detailed group- and age-related information relevant to study variables.

Procedure
Subjects reported to the sleep laboratory one hour prior to their normal bedtime for application of electrodes for electrographic recording of monopolar electroencephal-ographic (EEG;C3/M2), and bipolar horizontal and vertical electrooculographic and electromyographic (orbicularis oris) activities. Recordings were taken using standard procedures (Rechtschaffen and Kales, 1968) while subjects slept in a sound-attenuated, radio-frequency shielded, and temperature-humidity controlled room. Total bedtime allowed for sleep was 10 hours for children (5-7 yrs.) and preadolescents (8-12 yrs.), 9 hours for adolescents (13-16 yrs.), and 8.5 hours for adults (20-49 yrs.). All subjects had two adaptation nights immediately preceding the experimental night or nights (two for enuretic and hyperactive subjects and their age-matched controls) on which arousal threshold determinations were made.

On experimental nights 1500 Hz pure tones (3 seconds on, 3 seconds off) were delivered through an earphone insert during Stages 2,4, and REM sleep across the night (Figure 1). Stimuli were delivered in an ascending series of increasing intensity (5 dB increments beginning at 30 dB SPL, the threshold for tone detection during wakefulness), twice at each intensity until the subject awoke, or until 10 repetitions were presented at the maximum intensity of 120 dB. Subjects were instructed to press a hand-held button and say "I am awake" as soon as they were aware of the stimulation. At least 5 minutes of uninterrupted sleep in each stage was required before stimulation was initiated. A minimum of 30 minutes of sleep was allowed after a trial effecting arousal before another trial was attempted. Following an unsuccessful arousal attempt, another trial could be initiated after 10 minutes of sleep.

Analyses
Sleep recordings were independently stage scored by two individuals with high reliability (>.90) using standardized criteria (Rechtschaffen and Kales, 1968). Amplitude criteria for spindle determination are not provided in this manual which is commonly utilized in human (non-infant) sleep investigations. Consequently, studies of spindle activity across age groups have used various amplitude criteria for spindles, generally in the 5µV-25µV range. The values selected in the present investigation reflect those previously used in younger, non-adult (Goetz et al., 1983), adolescent and adult populations (de Maetalaer et al., 1987; Zeitlhofer et al., 1997). Spindles were visually scored on experimental nights only during Stage 2 awakening attempts since, among NREM sleep stages containing spindles, this activity is most prevalent during this sleep stage and spindle determination by visual analysis is most reliable for this stage. This stage also has the advantage of being distributed across the night, thereby permitting time-of-night analyses. The following criteria were used for spindle determination: a) the presence of 6-7 cycles within a 0.5 second interval; and b) within this interval, at least 3 cycles had to be =20µV for subjects =12 yrs. old and =25µV for subjects =13 yrs. old. Spindle activity was scored for the 2 minutes immediately preceding stimulus onset and throughout the stimulus period. The pre-trial period provided a baseline, non-stimulation-associated measure of spindle activity. Prior to analyses of the study database, two individuals established reliability for scoring sleep spindles (.88) using recordings other than those in that database. Interrater reliability, rechecked using random samples of dataset recordings scored near the end of these analyses, remained acceptably high (.85).

To assess more immediate variations in spindle activity in response to stimulation, spindles beginning within 200 milliseconds after the onset or termination of each 3-second tone were tabulated. This time interval was selected since it allows sufficient time for auditory stimulation to evoke responses in cortical brain regions from which spindles were recorded in the present investigation (Picton et al., 1974). Portions of recordings containing physiologic or electronic artifact were excluded from analyses. For time-of-night analyses, experimental night recordings were divided into thirds, and arousal attempts and associated spindle rates determined for each third.

Subjects were classified into the following five age groups for analyses: children (5-7 yrs.); preadolescents (8-12 yrs.); adolescents (13-16 yrs.); young adults (20-24 yrs.); and, adults (25-49 yrs.). Data were analyzed using univariate and multivariate analyses of variance procedures with post-hoct-tests and trend analyses (SYSTAT, 1994).

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Table 1 details the distribution of recording times sampled across groups, and indicates the relative contributions of each subject category to the analyses. Prestimulation sampling time was constant (2 minutes) but time intervals during which stimuli were delivered varied depending on when or whether subjects awoke, i.e., variable times for successful trials and 4.6 minutes for non-arousal trials. Over three-quarters (77.4%) of the subjects were non-adults, and this fact, together with the greater number of unsuccessful awakening attempts in younger subjects, resulted in substantially more stimulus-associated sample time for these subjects. To permit cross-group comparisons not confounded by this imbalance in sampling time, spindle activity was standardized (expressed as spindles/minute) for analyses.
A total of 18,826 spindles were scored, 7,525 in prestimulation periods and 11,301 during stimulation. Analyses were initially conducted to determine if data from subject groups with two experimental nights, i.e., hyperactive, enuretic and their age-matched controls, showed between-night consistency. The absence of significant variations in spindle activity across nights in these subjects would simplify group analyses by permitting averaging across nights and provide, comparable to other groups, a single experimental condition spindle-rate value, as well as confirm in younger subjects the presence of night-to-night consistency in spindle activity reported in adults. Between-night comparisons of prestimulation spindle rates by trials were non-significant (F[1,343]=1.908; p=.168), and data from the two experimental nights in these subjects were therefore averaged for subsequent analyses.

Comparisons between prestimulation spindle rates in clinical subjects and their respective age-matched controls were non-significant (Table 2). These data were further examined for group differences in spindle distribution across thirds of the night. Only subjects with at least one trial in each third were included in these analyses, and this restriction resulted in the exclusion of 3 subjects from these analyses (1 preadolescent and 2 adults) and an additional two others (1 adolescent and 1 young adult) from subsequent similar group analyses. Clinical subjects did not differ significantly from controls in any of these analyses (Table 3), and consequently data from these groups were combined within age categories in later analyses.

Analyses of prestimulation spindle activity conducted to determine if spontaneous spindle rate varied as a function of age did not reveal significant group differences (F[4,79]=0.961, p=.434). However, time-of-night analyses indicated an interaction between this variable and age (F[8,148]=2.141, GG=.035) resulting from an across-night variation which was quadratic in form (F[4, 74]= 3.154, p= .019; Figure 2). This interaction reflected increases in spindle rate in children between the 1st and 2nd (t[5]= -7.854, p<.001) and the 1st and 3rd thirds (t[5]=-2.573, p=.05) of the night. Significant time-of-night differences were not present for any other group.

Having determined that spindle rates preceding stimulation were similar across age groups, prestimulation and stimulation-associated spindle rates were then compared within age categories to assess the influence of stimulation on spindle incidence. These analyses indicated a decrease in spindle density during stimulation for all groups. This effect was significant for all but the adults [age x condition interaction: F (4, 79)= 2.588, p=.043; within-group comparisons significant at p= .01], and for these subjects the decrease approached significance (p=.07; Figure 3). Although the relative mean percent decrease varied across groups, ranging from 13.5% in preadolescents to 34.7% in young adults and averaging 21.02% overall, spindle rates during stimulation did not vary significantly across groups, whether stimulation periods were considered on a whole night (F[4,79]=2.043, p=.096) or thirds of the night basis (F[4, 74]=1.846, p=.129).

To determine whether spindle activity varied with changes in stimulus intensity, analyses were performed on nonarousal trials since these trials were equal in duration and the most extended in time, making possible assessments across the full range of intensities, i.e., 30 dB-120 dB. These considerations excluded both adult groups since nonarousals were rare among these subjects. For these analyses average spindle/minute values for each 10 dB stimulus increment were calculated and comparisons made between prestimulation spindle rates and those associated with the various stimulation intensities. A significant prestimulation-stimulation difference was present (F[10,520]=2.2, GG=.035), which did not vary significantly across age groups (F[2,52]=0.643, p=.53). There was a negative quadratic relationship between spindle rate and stimulus intensity (F[1,52]=4.25, p=.044), indicating a general decrease in spindle density with increasing intensity (Figure 4). Post-hoc analyses indicated the greatest decreases relative to prestimulation levels at the following intensities: 50-60 dB, t(54)=1.899, p=.06; 60-70 dB, t(54)=2.262, p=.03; 70-80 dB, t(54)=2.913, p=.005; 80-90 dB, t= 1.965, p= .06; 90-100 dB, t(54)=4.499, p=.001; 100-110 dB, t(54)= 3.090, p=.003; 110-120 dB, t(54)=2.679, p=.01, and, 120 dB, t(54)=3.602, p<001.

The evaluation of variations in spindle rate in relation to whether subjects awoke in response to stimulation was limited to the three youngest age groups since only these groups had a sufficient number of trials in both arousal and nonarousal conditions to support these analyses (see Table 1). Within-group, between-condition (arousal-nonarousal) comparisons of spindle rates indicated group by condition interactions during both prestimulation (F[2,233]= 6.09, p=.003) and stimulation (F[2,233]= 5.127, p=.007) periods. In both instances, spindle rates were higher in association with nonarousal relative to arousal for children and preadolescents-significantly so for children during prestimulation (t[49]= 3.113, p=.003) and for both children (t[49]= 3.512, p=.001) and preadolescents (t[155], p=.053) during stimulation. Spindle rates in adolescents did not vary significantly across these conditions. The magnitude of the difference in degree to which spindle rate was decreased relative to prestimulation values on these conditions across groups was slight, i.e., the overall mean difference between conditions was less than 2%.

Time-of-night analyses of these data showed that for both conditions spindle rate was significantly reduced relative to prestimulation values within each third of the night (Table 4). Prestimulation values on arousal and nonarousal trials did not vary significantly from one another across thirds (all Fs<1) and similar condition (arousal/nonarousal) by time-of-night comparisons of spindle rates during stimulation differed significantly only during the last third (F[1,80]=6.585, p=.012; Figure 5), with a higher mean rate on nonarousal relative to arousal trials.

A final set of analyses evaluated variations in spindle activity as a function of trial outcome by correlating prestimulation spindle rates with the stimulus intensity required to effect arousal and by comparing the effects of 10 dB increments in stimulus intensity on spindle rates across arousal-nonarousal conditions. The correlational analyses indicated the absence of a predictive relationship between prestimulation rates and arousal threshold values, either when subjects were considered as a single group (r= -.06, p=.40) or dichotomized into non-adult (r= -.02, p=.82) and adult (r=.03, p=.82) groupings. The incremental stimulus intensity comparisons revealed no significant differences between conditions when data were collapsed across the night (Figure 6) or considered within thirds of the night.

Of the more than 11,000 spindles that occurred during stimulation periods, only 562 (~5%) fell within 200 milliseconds of stimulus onset or offset. These were nearly equally distributed between these periods (onset: n=283; offset: n=279) and were not differentially distributed across age groups (F[4,79]=0.95, p=.442). This incidence is significantly less than that expected by chance ( 2= 5,746.29, df=1, p=.001; odds ratio=.05; 95% CI: 0.05-0.06), indicating that spindle activity was not immediately influenced by either the occurrence or termination of the auditory stimuli.

Activate the ShortNotes by clicking on this link. Your notes will be stored in this area and automatically retrieved upon your next visit.
The proposed sleep-protective, sleep-promoting role of sleep spindles (Johnson et al., 1976; Naitoh et al.,1982; Bowersox et al.,1985; Steriade and McCarley, 1990; Steriade and Amzica, 1998) provided the basis for the general expectation of an inverse relationship between spindle activity and arousal, and directed the specific predictions-based on variations in sleep depth derived from arousal threshold studies-of decreases in spindle activity with age, decreases across the night and increases with sensory stimulation. The results failed to confirm any of these expected relationships. However, the anticipated absence of significant variations in spindle rates in clinical groups relative to their controls was observed. These latter findings, together with observations of consistent night-to-night rates in both adult (Silverstein and Levy, 1976; Di Perri et al.,1977b, c; Shirakawa et al.,1978) and, as observed in this study, preadolescent subjects, suggest that spindle activity is a stable feature of sleep not readily influenced by normal daily activities or behavioral disorders.
The failure to observe significant age-related differences in baseline spindle rates corroborates the findings of Keane et al. (1977) and Bové et al. (1994), but not those of others who have variously reported increases (Smith et al., 1979; Bowersox et al., 1985) or decreases (Principe and Smith, 1982; Guazzelli et al., 1986; Wauquier, 1993; Landolt et al., 1996; Nicolas et al., 1997,1998) in spindle activity with advancing age. These across-study inconsistencies may derive from a variety of factors, including between-subject variability and variations in aspects of methodology, such as differences in sampling procedures (whole night vs. selected intervals; distribution within individual sleep cycles and across the night), methods of analysis (automated vs. visual scoring), as well as in amplitude and duration criteria. Despite such subject, methodologic and analytic variations, spindle rates reported for Stage 2 have remained within a fairly restricted range of values across studies (2-10/minute), and those observed in the present study fall within this range. The methodology of the present investigation (i.e., scoring criteria and reliability, distribution of samples during sleep, sample size, and analytic procedures) does not offer any obvious basis for considering either that the data were unrepresentative or the evaluation of spindle activity unreliable.

Significant time-of-night variations in the form of decreased spindle incidence early in the night were present only in children. Preadolescents showed an across-night spindle density profile similar to that observed in children, but attenuated in degree of between-thirds differences. These findings are generally consistent with those of Goetz et al. (1983) who reported a significant increase in spindle density in the second half of the night in subjects within the age range of the child and preadolescent groups in the present study.

The within-night distribution of spindles has not been previously studied in adolescents, and the results of these analyses in this investigation did not differentiate this group from any other than the children. Findings regarding this measure in adult subjects have been mixed. Similar to the reports of Di Perri et al. (1977c) and Gaillard and Blois (1981), but unlike others who have observed increases (Guazzelli et al., 1986) or decreases (Keane et al., 1977; Bové et al., 1994) in this measure, no significant across-night variation in spindle density was found in adult subjects in the present study. In view of the reported within-night inverse relationship between delta and sigma activities (Dijk et al., 1993; Uchida et al., 1992, 1994) and the generally acknowledged concentration of delta activity early in the night, the absence of consensus among studies in reporting increases in spindle density across the night appears perplexing. However, the lack of perfect identity between bursts of oscillatory 12-14 Hz activity which serve as the basic data in studies of spindle density, and activity in the sigma frequency band-of which coherent spindle bursts are only a component-may account in part for the apparent discrepancies noted above.

The most striking effect observed in this investigation was the decrease in spindle activity during auditory stimulation. This effect was present in all groups and the extent of the decrease was generally comparable across groups. Since adult subjects awakened on virtually all arousal attempts, more detailed analyses of this effect were conducted in younger subjects who provided adequate numbers of arousal and nonarousal trials for comparison. These analyses showed that the decrement in spindle rate increased with increasing stimulus intensity, and was even evident, although to a non-significant degree, at the lowest stimulus intensity (Figure 5). This effect occurred regardless of trial outcome, i.e., arousal or sleep maintenance, but was somewhat attenuated on nonarousal trials relative to those on which subjects awakened. Attempts to relate the relatively enhanced spindle activity on nonarousal trials to other variables, e.g., prestimulation levels, time-of-night, and stimulus intensity, which would provide convergent indications that spindle density at these times-even if decreased-might index a sleep-preserving mechanism, were largely unsuccessful. In general, these findings support those of Church et al. (1978) who reported no significant relationship between spindle rate and arousal threshold.

The literature dealing with the effects of sensory stimulation on spindle incidence in humans is limited and largely descriptive. In 6 subjects ranging in age from 4-68 months, Tanguay et al. (1975) reported that the number of spindles occurring within 400 milliseconds of auditory stimuli (1 millisecond clicks, 60 dB above waking hearing threshold) was neither decreased nor facilitated relative to those occurring during a comparable interval without stimulation. Other studies using human subjects in which sensory stimuli, generally auditory, have been delivered in the presence and absence of spindles, have had as their focus variations in associated evoked potentials rather than the effects of such stimulation on spindle activity. Nevertheless, in two such studies-one in 12 infants using auditory and somatosensory stimuli (Lenard and Ohlsen, 1972) and the other using auditory stimuli in 3 young male subjects of unspecified age (Yamadori, 1971)-brief statements, without supporting analyses, are made that stimulation did not produce (Lenard and Ohlsen, 1972) or alter the distribution of sleep spindles (Yamadori, 1971).

Prominent among methodologic parameters distinguishing this study from others that have not observed an effect of stimulation on sleep spindles are differences in stimulus duration and intensity. Relative to other studies, stimulus duration in this study was much longer, i.e., 3-second tones every 6 seconds throughout the trials, compared with 1 millisecond clicks every 2 seconds (Tanguay et al., 1975), 10 millisecond pips at variable rates (Yamadori, 1971), and trains of 70-80,1 second tones or 0.5 second square wave pulses/sleep cycle (Lenard and Ohlsen, 1972). Furthermore, unlike the fixed intensities used in previous studies, stimulus intensity in this study varied within trials (30-120 dB) and exceeded levels used in those studies, i.e., 60 dB above hearing threshold (Tanguay et al., 1975), 30 or 40 dB above hearing threshold (Yamadori, 1971), and 85 dB (Lenard and Ohlsen, 1972). Since ascending brainstem and basal forebrain activity can disrupt thalamocortical processes underlying spindle generation (Steriade et al., 1990), it is possible that the stimuli used in the present investigation were of sufficient intensity and duration to activate such pathways and effect the observed attenuation of spindle activity. The negative relationship between stimulus intensity and spindle density is consistent with this interpretation.

The analyses of spindle incidence immediately after stimulus onset and termination allowed for the determination of the extent to which spindles might be evoked in response to stimulus change. In view of the observed suppressive effect of auditory stimulation on sleep spindles, the failure of such stimulation to evoke spindles is not surprising. Although there are no previous reports in humans of systematically attempting to elicit spindles using external stimulation, or documenting spindle incidence relative to such stimulation, anecdotal reports (Lenard and Ohlsen, 1972; Tanguay et al., 1975) suggest that spindles are not readily evoked by auditory or somatosensory stimuli. Consistent with these impressions is the observation by Jankel and Niedermeyer (1985) that "Personal observation in the EEG laboratory clearly shows that a spindle train may be produced by an auditory stimulus (instead of a K-complex). But this is the exception rather than the rule; most spindle trains appear to be unrelated to arousing stimuli" (pp. 25-26). These reports and the results of the present investigation indicate a consensus among investigations in human subjects regarding the inability to evoke sleep spindles with external auditory stimuli. A similar, but qualified, conclusion can be drawn from a report by Contreras et al. (1997) who studied cats under barbiturate anesthesia--a condition in which spindles are produced with predictable regularity. These authors were unable to evoke spindles in these animals with low intensity direct cortical stimulation except near (within .4 seconds) the time of the expected occurrence of a barbiturate spindle, but with high intensity, stimuli spindles could be evoked at any time in this preparation.

The results of the present investigation are remarkable for their general similarity across subject groups varying widely in age and behavioral composition as reflected by the general absence of group differences for either baseline or stimulus-related measures, as well as by the within-subject night-to-night consistency of spindle rate observed for preadolescents. These observations reinforce existing literature indicating that the spindle is a robust correlate of NREM sleep generated in a regulated, and within-subjects, somewhat predictable manner, and is generally resistant to influence except by invasive procedures or events or pathology.

Equally remarkable is the uniform inconsistency of the results with a view of sleep spindles as proactive sleep promoting or protective events. Sleep spindles were shown not to covary meaningfully with known fluctuations in sleep depth as a function of either age, time of night, or arousal threshold, and were significantly-suppressed rather than provoked-by potentially sleep disturbing auditory stimuli, regardless of whether subjects awakened or not in response to the stimulation. These findings do not obviate or preclude spindle-synchronous effects which may serve to modify associated sensory and/or motor activities, but discrepancies in the literature regarding the nature of these effects and their relationship to momentary variations in arousal remain to be resolved. However, the results indicate that spindle generation is not enhanced in response to potentially sleep disturbing events and may be suppressed if such stimuli are of sufficient intensity, and therefore the functional consequences associated with sleep spindles must be applied passively, not reactively, to events that happen to coincide with spontaneous spindle occurrence.

1. Andersen P, Andersson SA, Lomo T. Nature of thalamocortical relations during spontaneous barbiturate spindle activity. J Physiol Lond 1967; 192: 283-307.
2. Andersen P, Sears TA. The role of inhibition in the phasing of spontaneous thalamic discharge. J Physiol Lond 1964; 173: 459-80.

3. Balzamo E. Papio Unibus: un primate palmi les primates. Etas de vigilance et activités ponto-géniculocorticales (PGO). Electroencephalogr Clin Neurophysiol 1980; 48: 694-705.

4. Bert J, Kripke DF, Rhodes J. Electroencephalogram of the mature chimpanzee: twenty-four hour recordings. Electroencephalogr Clin Neurophysiol 1970a; 28: 368-73.

5. Bert J, Pegram V, Rhodes JM, Balzano EM, Naquwt R. A comparative study of two cercopthecinae. Electroencephalogr Clin Neurophysiol 1970b; 28: 32-40.

6. Bonnet MH, Moore SE. The threshold of sleep: Perception of sleep as a function of time asleep and auditory threshold. Sleep 1982; 5: 267-76.

7. Bonnet MH, Webb WB, Barnard G. Effect of flurazepam, pentobarbital, and caffeine on arousal threshold. Sleep 1979; 1(3): 271-9.

8. Bové A, Culebras A, Moore JT, Westlake RE. Relationship between sleep spindles and hypersomnia. Sleep 1994; 17(5): 449-55.

9. Bowersox SS, Kaitin K I, Dement WC. EEG spindle activity as a function of age: Relationship to sleep continuity. Brain Res 1985; 334: 303-8.

10. Busby K, Mercier L, Pivik RT. Ontogenetic variation in auditory arousal threshold during sleep. Psychophysiology 1994; 31: 182-8.

11. Busby K, Pivik RT. Auditory arousal thresholds during sleep in hyperkinetic children. Sleep 1985; 8(4): 332-41.

12. Busby K, Pivik RT. Failure of high intensity auditory stimuli to affect behavioral arousal in children during the first cycle. Ped Res 1983; 17(10): 802-5.

13. Caveness WF. Atlas of EEG in the developing monkey (Macaca mulatta). Reading, MA: Addison-Wesley Press, 1962.

14. Church MW, Johnson LC, Seales DM. Evoked K-complexes and cardiovascular responses to spindle-synchronous and spindle-asynchronous click stimuli during NREM sleep. Electroencephalogr Clin Neurophysiol 1978; 45: 443-53.

15. Contreras D, Destexhe A, Sejnowski TJ, Steriade M. Spatiotemporal patterns of spindle oscillation in cortex and thalamus. J Neurosci 1997; 17(3): 1179-96.

16. de Maertelaer V, Hoffman G, Lemaire M, Mendlewicz J. Sleep spindle activity changes in patients with affective disorders. Sleep 1987; 10(5): 443-51.

17. Dijk DJ, Beersma DMG, Hoofdakker RH. All night spectral analysis of EEG sleep in young adult and middle-aged male subjects. Neurobiol Aging 1989; 10: 677-82.

18. Dijk DJ, Hayes B, Czeisler CA. Dynamics of electroencephalographic sleep spindles and slow wave activity in men: effect of sleep deprivation. Brain Res 1993; 626: 190-9.

19. Di Perri R, Meduri M, Di Rosa AE, Musolino R, Simone F. Sleep spindles in healthy people: 3) An automated, quantitative and topographical analysis in aged subjects (preliminary report). Boll Soc It Biol Sper 1977a; 53: 997-1003.

20. Di Perri R, Meduri M, Di Rosa AE, Simone F. Sleep spindles in healthy people: 1) a quantitative, automatic analysis in young adults subjects. Boll Soc Ital Bio Sper 1977b; 53: 983-9.

21. Di Perri R, Meduri M, Di Rosa AE, Simone F. Sleep spindles in healthy people: 2) automatic, topographical analysis in young-adult subjects. Boll Soc It Biol Sper 1977c; 53: 990-6.

22. Driver HS, Dijk DJ, Werth E, Biedermann K, Borbély AA. Sleep and the sleep electroencephalogram across the menstrual cycle in young healthy women. J Clin Endocrinol Metabol 1996; 81: 728-35.

23. Ehrhart J, Ehrhart A, Muzet A, Schieber JP, Naitoh P. K-complexes and sleep spindles before transient activation during sleep. Sleep 1981; 4(4): 400-7.

24. Elton M, Winter O, Heslenfeld D, Loewy D, Campbell K, Kok A. Event-related potentials to tones in the absence and presence of sleep spindles. J Sleep Res 1997; 6: 78-83.

25. Feinberg I, Koresko R, Heller N. EEG sleep patterns as a function of normal and pathological aging in man. J Psychiat Res 1967; 5: 107-44.

26. Gaillard JM, Blois R. Spindle density in sleep of normal subjects. Sleep 1981; 4(4): 385-91.

27. Gibbs EL, Gibbs FA. Extreme spindles: connection of electroencephalographic sleep pattern with mental retardation. Science 1962; 138: 1106.

28. Goetz RR, Goetz DM, Hanlon C, Davies M, Weitzman ED, Puig-Antich J. Spindle characteristics in prepubertal major depressives during an episode and after sustained recovery: a controlled study. Sleep 1983; 6(4): 369-75.

29. Goodenough DR, Lewis HB, Shapiro A, Jaret L, Sleser I. Dream reporting following abrupt and gradual awakenings. J Personality Soc Psychol 1965; 2: 170-9.

30. Guazzelli M, Feinberg I, Aminoff M, Fein G, Floyd TC, Maggini C. Sleep spindles in normal elderly: comparison with young adult patterns and relation to nocturnal awakening, cognitive function, and brain atrophy. Electroencephalogr Clin Neurophysiol 1986; 63: 526-39.

31. Hongo TK, Kubota K, Shimazu H. EEG spindles and depression of gamma motor activity. J Neurophysiol 1963; 26: 568-80.

32. Hughes JR, Mazurowski JA. Studies on the supracallosal mesial cortex of unanesthetized conscious mammals. II. Monkey. D. Vertex waves and epileptiform activity. Electroencephalogr Clin Neurophysiol 1964; 16: 561-74.

33. Jankel WR, Niedermeyer E. Sleep spindles. J Clin Neurophysiol 1985; 2(1): 1-35.

34. Johnson LC, Church MW, Seales DM, Rossiter VS. Auditory arousal thresholds of good sleepers and poor sleepers with and without flurazepam. Sleep 1979a; 1(3): 259-70.

35. Johnson LC, Hanson K, Bickford RG. Effect of flurazepam on sleep spindles and K-complexes. Electroencephalogr Clin Neurophysiol 1976; 40: 67-77.

36. Johnson LC, Seales DM, Naitoh P, Church MW, Sinclair M. The effects of flurazepam hydrochloride on brain electrical activity during sleep. Electroencephalogr Clin Neurophysiol 1979b; 47: 309-21.

37. Johnson LC, Spinweber CL, Seidel WF, Dement WC. Sleep spindle and delta changes during chronic use of a short-acting and a long-acting benzodiazepine hypnotic. Electroencephalogr Clin Neurophysiol 1983; 55: 662-7.

38. Keane B, Smith J, Webb W. Temporal distribution and ontogenetic development of EEG activity during sleep. Psychophysiology 1977; 14: 315-21.

39. Keefe FB, Johnson LC, Hunter EJ. EEG and autonomic response pattern during waking and sleep stages. Psychophysiology 1971; 8: 198-212.

40. Khan A, Rechtschaffen A. Sleep pattern and sleep spindles in hyperkinetic children. Sleep Res 1978; 7: 137.

41. Landolt HP, Dijk DJ, Achermann P, Borbély AA. Effect of age on the sleep EEG: slow-wave activity and spindle frequency activity in young and middle-aged men. Brain Res 1996; 738: 205-12.

42. Lenard HG, Ohlsen I. Cortical responsivity during sleep spindles in young children. Neuropädiatrie 1972; 3(3): 258-67.

43. Lenard HG, Schulte FJ. Sleep spindles in hormonal and metabolic diseases of infancy and childhood. In: Petre-Quadens O, Schlag JD eds. Basic Sleep Mechanisms. New York: Academic Press, 1974, pp. 380-403.

44. Loomis A L, Harvey EN, Hobart G. Potential rhythms of the cerebral cortex during sleep. Science 1935; 81: 597-8.

45. Louis J, Zhang JX, Revol M, Debilly G, Challamel MJ. Ontogenesis of nocturnal organization of sleep spindles: a longitudinal study during the first 6 months of life. Electroencephalogr Clin Neurophysiol 1992; 83: 289-96.

46. McCormick DA, Bal T. Sleep and arousal: thalamocortical mechanisms. Ann Rev Neurosci 1997; 20: 185-215.

47. McCormick L, Nielsen T, Nicolas A, Ptito M, Montplaisir J. Topographical distribution of spindles and K-complexes in normal subjects. Sleep 1997; 20(11): 939-41.

48. Metcalf DR. EEG sleep spindles ontogenesis. Neuropädiatrie 1969; 1(4): 428-33.

49. Monod N, Eliet-Flesher J, Dreyfus-Brisac C. Le sommeil du nouveau-né et du prémature. III. Les troubles de l'organization du sommeil chez le nouveau-né pathologique: analyse des études polygraphiques. Biol Neonatorum 1977; 11: 216-47.

50. Naitoh P, Antony-Bass V, Muzet A, Ehrhart J. Dynamic relation of sleep spindles and K-complexes to spontaneous phasic arousal in sleeping human subjects. Sleep 1982; 5(1): 58-72.

51. Nicolas A, Rompré S, Dumont M, Laberge L, Montplaisir J. Four to five seconds periodicity of sleep spindles in different age groups. Sleep Res 1997; 26: 31.

52. Nicolas A, Rompré S, Paquet J, Montplaisir J. Influence of age on sleep spindles characteristics. Sleep 1998; 21: 54.

53. Ornitz EM. Development of sleep patterns in autistic children. In: Clement C, Purpura DP, Mayer FE, eds. Sleep and the Maturing Nervous System. New York: Academic Press, 1972, pp. 321-63.

54. Picton TW, Hillyard SA, Krausz HI, Galambos R. Human auditory evoked potentials. I: Evaluation of components. Electroencephalogr Clin Neurophysiol 1974; 36: 179-90.

55. Pivik RT, Bylsma FW. Spinal reflex modulation during sleep spindles in man. Sleep Res 1982; 11: 37.

56. Poitras L, Bylsma FW, Simeon J, Pivik RT. Cortical sleep spindle activity in hyperkinetic children. Sleep Res 1981; 10: 117.

57. Poley D, Dummermuth G. EEG finding in patients with phenylketonuria before and during treatment with a low-phenylalanine diet in patients with some other inborn errors of amino-acid metabolism. In: Holt KS, Coffey O, eds. Some Recent Advances in Inborn Errors of Metabolism. Edinburgh: Livingstone, 1968, pp. 61-79.

58. Principe JC, Smith JR. Sleep spindle characteristic as a function of age. Sleep 1982; 5(1): 73-84.

59. Rechtschaffen A, Hauri P, Zeitlin M. Auditory awakening thresholds in REM and NREM sleep stages. Percept Motor Skills 1966; 22: 927-42.

60. Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages in Human Subjects. Los Angeles: Brain Information Service/Brain Research Institute, 1968.

61. Shibagaki M, Kiyono S, Watanabe K. Spindle evolution in normal and mentally retarded children: a review. Sleep 1982; 5(1): 47-57.

62. Shirakawa S, Sumizono T, Azumi K. Characteristics of sleep spindle activity during seven consecutive nights. Sleep Res 1978; 7: 48.

63. Shouse MN, Sterman MB. Change in seizure susceptibility, sleep time and sleep spindles following thalamic and cerebellar lesions. Electroencephalogr Clin Neurophysiol 1979; 46: 1-12.

64. Silverstein LD, Levy CM. The stability of the sigma sleep spindle. Electroencephalogr Clin Neurophysiol 1976; 40: 666-70.

65. Smith JR, Karacan I, Yang M. Automated measurement of alpha, beta, sigma, and theta burst characteristics. Sleep 1979; 1(4): 435-43.

66. Steriade M. Cellular substrates of brain rhythms. In: Niedermeyer E, Lopes da Silva F, eds. Electroencephalography: Basic Principles, Clinical Applications and Related Fields. Baltimore: Williams & Wilkins, 1993a, pp. 27-62.

67. Steriade M. Sleep oscillations and their blockage by activating systems. J Psychiatry Neurosci 1993b; 19(5): 354-9.

68. Steriade M, Amzica F. Coalescence of sleep rhythms and their chronology in corticothalamic networks. Sleep Res Online 1998; 1(1): 1-10.

69. Steriade M, Deschêne M, Domich L, Mulle C. Abolition of spindle oscillation in thalamic neurons disconnected from nucleus reticularis thalami. J Neurophysiol 1985; 54(6): 1473-97.

70. Steriade M, Domich L, Oakson G, Deschêne M. The deafferented reticular thalamic nucleus generates spindle rhythmicity. J Neurophysiol 1987; 57(6): 260-73.

71. Steriade M, Gloor P, Llinàs RR, Lopes da Silva FH, Mesulam MM. Basic mechanisms of cerebral rhythmic activities. Electroencephalogr Clin Neurophysiol 1990; 76: 481-508.

72. Steriade M, McCarley RW. Brainstem Control of Wakefulness and Sleep. New York: Plenum Press, 1990.

73. Steriade M, McCormick DA, Sejnowski TJ. Thalamocortical oscillations in the sleeping and aroused brain. Science 1993; 262: 679-85.

74. Sterman MB, Bowersox SS. Sensorimotor electroencephalogram rhythmic activity: a functional gate mechanism. Sleep 1981; 4(4): 408-22.

75. Tanguay PE, Ornitz EM, Kaplan A, Bozzo ES. Evolution of sleep spindles in childhood. Electroencephalogr Clin Neurophysiol 1975; 38: 175-81.

76. Uchida S, Atasumi Y, Kojima T. Dynamic relationship between sleep spindles and delta waves during a NREM period. Brain Res Bull 1994; 33: 351-5.

77. Uchida S, Maloney T, Feinberg I. Beta (20-28 Hz) and delta (0.3-3 Hz) EEGs oscillate reciprocally across NREM and REM sleep. Sleep 1992; 15(4): 352-8.

78. Wauquier A. Aging and changes in phasic events during sleep. Physiol Behav 1993; 54: 803-6.

79. Williams HL, Hammack JT, Daly RL, Dement WC, Lubin A. Responses to auditory stimulation, sleep loss and the EEG stages of sleep. Electroencephalogr Clin Neurophysiol 1964; 16: 269-79.

80. Williams RL, Agnew HW, Webb WB. Sleep patterns in the young adult female: an EEG study. Electroencephalogr Clin Neurophysiol 1966; 20: 260-6.

81. Wolfish NM, Pivik RT, Busby KA. Elevated sleep arousal thresholds in enuretic boys: clinical implications. Acta Paediatr 1997; 86: 381-4.

82. Wu HS, Gould JB, Lee AFS, Fineberg N. Factors affecting sleep spindle activity during infancy. Dev Med Child Neurol 1980; 22: 344-51.

83. Yamadori A. Role of the spindles in the onset of sleep. Kobe J Med Sci 1971; 17: 97-111.

84. Zeitlhofer J, Gruber G, Anderer P, Asenbaum S, Schimicek P, Saletu B. Topographic distribution of sleep spindles in young healthy subjects. J Sleep Res 1997; 6: 149-55.

85. Zepelin H, McDonald CS, Zammit GK. Effects of age on auditory awakening thresholds. J Gerontol 1984; 39(3): 294-300.

The authors gratefully acknowledge the secretarial and technical assistance of Jane Buttrum in the preparation of this manuscript.

Original address of this text :

http://www.sro.org/bin/article.dll?Paper&1614&0&0

Please copy this address to the address bar of your internet browser and press the "enter" key.
(We prefer not to put actual links because often page locations change and then our log files get cluttered with error messages
- if the address does not work try to find it from the homepage of the website in question).