Sleep complaints are among the most frequently reported symptoms by patients diagnosed with a paychiatric disorder (Benca, Obermeyer, Thisted & Gillin,1992). As a result a considerable amount of research has focused on identifying sleep patterns which are distinctly associated with certain psychiatric conditions in an effort to devise more reliable diagnostic criteria (Bence, Obermeyer, Thisted & Gillin,1992). In addition, the observation of sleep events can help define mechanisms which are abnormal or dysfunctional in particualr psychiatric conditions because measurements are being collected at a time when cognitive and emotional influences are assumed to be at particularly low levels and as such the effects of these variables should be minimized (Koenigsberg, Pollak, Fine & Kakuma,1994).
Posttraumatic stress disorder (PTSD) and panic disorder (PD) are both classified as anxiety disorders and appear to exhibit many similarities, especially those involving sleep disturbances. In particular, both disorders demonstrate evidence of involving a dysregulation of arousal, which is especially apparent during sleep. Research investigating these conditions has yet to establish conclusively where the source of this malfunction is located or how it functions to control arousal. There is however accumulating evidence to suggest that, even though PTSD and PD appear to be caused by a common dysfunction in the ability to maintain diminished levels of arousal, this final outcome is precipitated by distinctly different sequences of neural events.
SOME CHARACTERISTICS OF NORMAL SLEEP
Before discussing the sleep abnormalities evident in PTSD and PD it is important to have a general understanding of the characteristics of normal sleep. First, in the healthy adult, there is a cyclic progression through several sleep stages (stages 1 through 4 and REM sleep). When measured, each stage shows a distinct pattern of physiological variables, such as EEG, EMG and EOG activity (Ross, Ball, Sullivan & Caroff,
1989), Generally sleep begins with a decrease in arousal and drowsiness and evolves through a cyclic pattern which is repeated several times through the night, moving through the non-REM (NREM) stages and into REM sleep (which on average represents 20 - 25% of total sleep time (Ross, Ball, Sullivan & Caroff,1989). Finally, the REM control mechanism is hypothesized to consist of two circuits of neurons: 1) the excitatory component, involving cholinergic neurons in the brainstem which trigger REM onset; and 2) the inhibitory component, comprising noradrenergic and serotonergic neurons in the brainstem (Ross, Ball, Sullivan & Caroff,1989).
POSTTRAUMATIC STRESS DISORDER
One of the characteristic features of PTSD is chronic sleep disturbance (Mellman, Kulick-Bell, Ashlock & Nolan,1995). Some of the most commonly reported abnormal sleep events in PTSD are: nightmares, difficulties with sleep initiation and maintenance, nocturnal panic and excessive body and limb movements (Mellman, Kulick-Bell,
Ashlock & Nolan,1995). It has been suggested that these symptoms of PTSD are the consequence of an inability to sustain a diminished level of arousal compatible with sleep maintenance and efficiency (Mellman, Kulick-Bell, Ashlock & Nolan,1995). However, these symptoms also characterize various other sleep and psychiatric disorders, for example, REM sleep behaviour disorder, night terrors and PD (Mellman, Kulick-Bell, Ashlock & Nolan,1995). Specifically, PTSD and PD are associated with increased nightime awakenings, increased levels of physiological arousal in general and sensitivity to particular biological challenges (Mellman, Kulick-Bell, Ashlock & Nolan,1995). Consequently, precise delineations of the abnormal sleep events associated with these different disorders are required so that the underlying biological mechanisms can be identified and possible relationships between the disorders can be verified or rejected.
In one study, (Mellman, Kulick-Bell, Ashlock & Nolan,1995) three types of data were collected in an effort to distinguish the sleep characteristics of PTSD. Specifically, information was collected in the form of sleep surveys, sleep diaries and sleep laboratory assessments from groups of male Vietnam veterans with PTSD. The purpose of this study was to gather physiological sleep data characteristic of combat-related PTSD and through this effort to expand on the current knowledge of the pathophysiology of PTSD (mellman, Kulick-Bell, Ashlock & Nolan,1995). The results from the data collected through the sleep survey are outlined below.
An outline of sleep syptoms gathered by survey
Compared with veterans without PTSD, veterans with PTSD reported:
Data accumulated through the observation of physiological measures in the sleep laboratory reveal that the veterans with PTSD experienced significantly more microawakenings (lasting 15-45 seconds), more time awake, more body movement, more subclinical respiratory events and decreased sleep efficiency compared with healthy subjects (Mellman, Kulick-Bell, Ashlock * Nolan,1995). A portion of awakenings (83%) experienced by veterans with PTSD were not associated with dream recall even though they reported a sensation of fear or physiological arousal (Mellman, Kulick-Bell, Ashlock & Nolan,1995). Half of these awakenings occured following an episode of REM sleep and EMG readings were not indicative of REM behaviour disorder (Mellman, Kulick-Bell, Ashlock & Nolan,1995). The authors suggest that these findings, especially the data revealing subthreshol awakenings and subclinical sleep apneas, correspond with the subjects' self-reports confirming that sleep is less restorative in individuals with PTSD (Mellman, Kulick-Bell, Ashlock & Nolan, 1995).
The authors consider the data from this study to favour the hypothesis that PTSD is associated with arousability from sleep (Mellman, Kulick-Bell, Aslock & Nolan,1995) and contemplate two possible explanations for this phenomenon. One hypothesis proposes that REM awakenings are a learned response to dream activity (Mellman, Kulick-Bell, Ashlock & Nolan,1995). This conditioning is presumed to take place during the acute phase of PTSD when nightmares are especially intense. Later, during the chronic phase, even though nightmares occur less frequently, the psychophysiological response to REM sleep activity may persist (Mellman,Kulick-Bell, Ashlock & Nolan,1995). This interpretation might account for an interesting difference revealed between PTSD sufferers in this study. It was found that veterans who received medication for their symptoms were more likely to report awakenings due to nightmares; whereas, veterans not receiving medication reported more awakenings due to panic not associated with dream recall (Mellman, Kulick-Bell, Ashlock & Nolan,1995). This explanation seems to concur with evidence gathered on other related phenomena, which Greenberg refers to as "dream interruption insomnia" (Mellman, Kulick-Bell, Ashlock & Nolan,1995,114), which has been successfully treated with medications such as diazepoxide, resulting in fewer awakenings and more dream recall (Mellman, Kulick-Bell, Ashlock & Nolan,1995).
The second explanation for increased arousability argues that neurobiological dysfunctions can account for the increased arousal response in PTSD. This belief is supported by research findings that PTSD patients are predisposed to experiencing panic symptoms or flashbacks when administered certain drugs (Mellman, Kulick-Bell, Ashlock & Nolan,1995). One chemical found to exert this type of influence is yohimbine; which effects locus coeruleus activity through the stimulation of noradrenergic neurons (Mellman, Kulick-Bell, Ashlock & Nolan,1995). The authors note that this interpretation coincides with Hobson et a's model of sleep regulation proposing that an increase in locus coeruleus activity is the "primary stimulus for terminating REM activity" (Mellman, Kulick-Bell, Ashlock & Nolan,1995,114). Moreover, the data from the present study supports this hypothesis as most REM awakenings were found to occur after a period of sustained REM sleep (Mellman, Kulick-Bell, Ashlock & Nolan,1995).
In conclusion, the authors note that their findings extend upon earlier research demonstrating the presence of "a conditioned heightened arousal response to trauma-related stimuli during wake states" (Mellman, Kulick-Bell, Ashlock & Nolan,1995,115) and specifically propose that PTSD may represent a more general dysfunction in the ability to maintain states of diminished arousal (Mellman, Kulick-Bell, Ashlock & Nolan,1995).
Another study, (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995) evaluating the effects of trauma on sleep, attempts to address a common limitation on research in this area by obtaining data within a short time after the trauma rather than years later. In this study, victims of hurricane Andrew were evaluated between 6 to 12 months after the incident for psychiatric and sleep symptoms. Victims of the hurricane without the presence of a psychiatric disorder during the 6 months prior to the trauma were compared with victims with prior psychiatric morbidity and individuals who were unaffected by the event. The aim of the study was to assess the relationships between sleep complaints and the prevalence of PTSD symptomatology. This objective was inspired by previous research findings which have revealed a role for sleep disturbance in the development and maintenance of PTSD. For example, the authors report evidence that certain mood and cognitive symptoms associated with PTSD, such as irritability and problems with concentration, have been found in healthy subjects during sleep deprivation (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995). They also refer to data in support of an association between insomnia and the onset of mood and anxiety disorders and a relationship between nightmares, sleep disturbance and the predisposition to additional psychological deterioration from stressors subsequent to the precipitating trauma (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995). Psychological and sleep data was collected through several measures: the Structured Clinical Interview for DSM-III-R-Non-Patient Version, the Pittsburgh Sleep Quality Index and sleep laboratory assessment. It was found that sleep complaints are a frequently reported phenomenon after a traumatic event and that those subjects who had psychiatric and sleep disturbances before the trauma were especially vulnerable to sleep difficulties after the hurricane (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995). Moreover, the present study supports previous findings that PTSD sufferers tend to experience "brief shifts toward higher levels of arousal" (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995,1662), for example, a greater number of arousals and more entries into stage 1 sleep (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995; Mellman, Kulick-Bell, Ashlock & Nolan,1995). This data strengthens the hypothesis that PTSD involves an inability to maintain diminished levels of arousal, and that this may function as a predisposing factor in the development of PTSD, or as a factor in maintaining PTSD symptomatology (Mellman, David, Kulick-Bell, Hebdig & Nolan,1995).
PANIC DISORDER
The investigation of PD and its relationship with sleep disturbances is another method through which a greater understanding of anxiety disorders, in general, can be accomplished. Sleep is a particularly pertinent and revealing condition in which to monitor the incidence of panic. As PD is defined by the occurence of spontaneous and unexpected panic attacks (Mellman & Uhde,1989), studying PD subjects during sleep allows one the opportunity to observe the occurence of a relatively pure form of panic phenomenon, the sleep panic attack (Mellman & Uhde,1989).
Additionally, studying physiological activity of PD patients during sleep, diminishes the possibility of confound with the effects of anxious cognition and expectancy set (Koenigsberg, Pollak, Fine & Kakuma,1994).
Similar to subjects with PTSD, subjects with PD manifest sleep disturbances, such as, increased sleep latency and decreased sleep time and efficiency (Mellman & Uhde,1989). In contrast, however, the awakenings experienced by individuals with PD appear in NREM as opposed to REM sleep (Mellman & Uhde,1989). Some of the EEG characteristics of PD subjects are shown below.
EEG characteristics of subjects with panic disorder
The EEG findings for PD patients during sleep raise several possible theoretical explanations for PD. Some researchers have suggested a link between night terrors and sleep panic; however, important differences between these disorders do exist. Night terrors tend to occur most often in children and following sustained delta activity; whereas, sleep panic attacks occur more frequently in adults and after less than 2 minutes of delta sleep (Mellman & Uhde,1989). Regardless, there still remains the possibility that these disorders are linked developmentally and that night terrors may represent a predisposing factor for sleep panic or PD (Mellman & Uhde,1989).
Other possible explanations involve physiological sensitivities to bodily sensations, such as, increased heart or respiratory rate; however, as a significant number of sleep panic attacks seem to occur following progression into a deeper sleep stage, this explanation seems unlikely (Mellman & Uhde,1989). Nevertheless, the authors note that these findings could be related to a carbon dioxide (CO2) sensitivity which is present in some PD patients (Mellman & Uhde,1989).
The hypothesis regarding CO@ sensitivity has been studied in awake PD patients and is supported by evidence demonstrating abnormalities of the respiratory system, especially in regard to a hypersensitivity to increasing levels of CO2 (Stein, Millar, Larsen & Kryger,1995; Papp et al, 1997; Pine et al,1994; Koenigsberg, Pollack, Fine & Kakuma,1994), which is an alternate method to induce the sensation of suffocation. The relevance of the effects of CO2 biological challenge is that the neural regions implicated in CO2 monitoring are located near the locus coeruleus (LC) and the raphe nuclei, and increased levels of CO2 are associated with increased firing rates in LC neurons (Pine et al,1994). In addition, these brainstem regions are believed to be implicated in anxiety disorders (Pine et al,1994) and the regulation of sleep features, such as, the inhibitory control of acetylcholinergic neurons which promote the onset of REM sleep (Hobson,1995).
In an attempt to extend on the findings related to CO2 sensitivity in PD patients during the waking state, a study (Stein, Millar, Larsen & Kryger,1995) was conducted to observe whether these abnormalities were also present in PD patients during sleep. Data was collected by recording the respiratory rate, tidal volume and the occurence of microapneas, microhypopneas, and sighs in subjects with PD and control subjects. The data reveal that there was no difference in respiratory rate, but that there was significantly more irregular tidal volumes and a greater number of respiratory pauses in PD subjects. Subjects with PD were also found to have a larger quantity of microapneas and microhypopneas, although this difference was not significant due to the large variance in these measures between subjects. The authors do note however that this trend could be indicative of irregular respiratory events during sleep occuring in a distinct subgroup of PD patients not yet identified (Stein, Millar, Larsen & Kryger,1995). These results offer support for a physiological interpretation of panic as the events were recorded during sleep, when anxiety symptoms are assumed to be at a lower level of intensity (Stein, Millar, Larsen & Kryger,1995). In sum, the results do support a conclusion that patients with PD have a heightened sensitivity to CO2, which also coincides with Klein's suffocation false alarm theory (Stein, Millar, Larsen & Kryger,1995). It is believed that this heightened sensitivity causes individuals to engage in repeated sighing to decrease levels of arterial CO2, which in turn results in pauses in respiration because there is no CO2 stimulus to trigger breathing (Stein, Millar, Larsen & Kryger,1995). This explanation is not entirely satisfactory however as there was no significant increase in the number of sighs observed in subjects with PD compared with subjects in the control group. The authors suggest that the finding of increased minor apnea episodes could be "associated with brief arousals sufficient to "alarm" sensitive patients" (Stein, Millar, Larsen & Kryger,1995,1172) and trigger sleep panic attacks.
One study (Pine et al,1994), which serves to clarify the effects of CO2 sensitivity in the pathophysiology of PD, compared subjects with a central nervous system disorder which prevents the perception of increased CO2 levels (congenital central hypoventilation syndrome) with subjects that had a psychological disturbance and intact CO2 perception. The researchers predicted that if individuals with this disorder were protected from the development of PD through an inability to detect increasing levels of CO2 they would manifest fewer symptoms of anxiety than individuals with other disorders. Conversely, if conditional learning, rather than physiological factors, is involved in the etiology of PD, then the subject group with congenital central hypoventilation syndrome would be predicted to show more symptoms of anxiety, as this group is subjected to many frightening experiences, such as medical procedures and life-threatening occurences of suffocation (Pine et al,1994). The results showed that subjects with congenital central hypoventilation syndrome had significantly fewer anxiety symptoms, lower rates of anxiety disorders and lower rates of childhood disorders linked with the development of PD in adulthood. In sum, the data supports the hypothesis that congenital central hypoventilation syndrome protects rather than predisposes individuals to anxious symptomatology and that biological sensitivity plays a greater role in the development and maintenance of anxiety than fear conditioning (Pine et al,1994).
continue to the conclusion of this essay
return to the physiological psychology page
return to the main psychology page
