Startle Response In Individuals With PTSD
By C. A. Morgan III, M.D., M.A.
NCP Clinical Quarterly 7(4): Fall 1997
Post traumatic Stress Disorder (PTSD), marked by symptoms of reexperiencing, avoidance and arousal, was officially delineated in 1980 as a clinical diagnosis within the category of anxiety disorders. Although initial delineation and characterization of PTSD represents a major advance, the diagnostic criteria continue to emphasize factors mainly dependent on patient self-reporting. The DSM-III-R includes physiologic reactivity and exaggerated startle as diagnostic features of PTSD. The presence of psychophysiological alterations accompanying a mental disorder has provided the opportunity to obtain data that is more objective and more readily quantifiable than self-report data. While this article will focus primarily on investigations of the startle response, a few words are in order about psychophysiologic investigations. �
Psychophysiology studies have attempted to assess the symptoms of physiologic reactivity in individuals with PTSD by measuring the heart rate (HR), electrodermal, and blood pressure responses of veterans with PTSD and control subjects when said subjects are exposed to recorded sights and sounds of combat in the laboratory. In most of these studies, the veterans with PTSD have shown greater autonomic arousal when exposed to the combat stimuli than the comparison subjects. The ability to discriminate PTSD from control subjects using these physiologic measures has been thought impressive and has ranged from 80%-95% (1-3). While these studies have contained design flaws (such as group discrepancies in age, educational level, and combat exposure; comparisons of medicated and unmedicated subjects; differences in baseline physiologic arousal), taken together, they have been perceived by biological investigators as providing convincing evidence of physiologic hyperreactivity to combat-related stimuli in veteran subjects with PTSD. Yet some investigators felt that the presentation of identical stimuli to all subjects did not permit an accurate assessment of what was uniquely stressful about a particular individual traumatic experience. It was thought that without the possibility of eliciting an individual’s unique response to a stressful event, the sensitivity and specificity of psychophysiological testing could not improve.
As a result, several researchers have used script-driven mental imagery as a means of provoking a more personalized physiologic response in the laboratory. The scripts are recorded narratives taken from a detailed description of the most traumatic event experienced by the subject, then played back to the subject in the laboratory. Each subject is asked to listen to the recording and to imagine the events portrayed. While the subject does this, heart rate, skin conductance and electomyogram activity of the left frontalis are recorded. This technique has successfully demonstrated significantly larger physiological responses during personal combat imagery compared with mentally healthy Vietnam veterans (4), and Vietnam veterans with non-PTSD anxiety disorders (5). Discriminant analyses of the physiologic measures indicated that skin conductance successfully classified 73%, EMG 67%, and HR 64% of the PTSD subjects (4). The convergent elevation of these measures in the PTSD group is noteworthy considering that concordance between physiologic measures in research of this nature has generally been found to be low.
However promising, the above described psychophysiology studies examined peripheral autonomic reactivity and behavioral changes in order to make inferences about brain function in PTSD. A number of investigators felt that a clearer understanding of central nervous system reactivity and functioning might be afforded by using a more direct probe of brain function that is sensitive to noradrenergic neuronal reactivity, fear and alarm states. It was not without reason that many believed that the acoustic startle reflex could fulfill such a need.
Historical and contemporary records provide evidence that an important symptom seen in combat veterans diagnosed with Shell Shock, Combat Fatigue or Post Traumatic Stress Disorder has been, and continues to be, an exaggerated startle reflex. (6-9) Clinical observations of exaggerated startle in distressed combat veterans were so common by mid-century, some psychiatric authorities argued that increased startle was cardinal symptom of combat fatigue (10). While not considered the cardinal symptom of PTSD today, exaggerated startle remains tightly linked to trauma related psychological illness. In fact, according to DSM-IV, PTSD is now the only anxiety disorder in which hyperstartle is listed as a core symptom.
Investigators have had various motivations for studying the acoustic startle reflex in humans and especially in those suffering from PTSD. Some have been interested in finding out whether or not exaggerated startle is a marker (or sign) indicating, or helping to provide a reliable diagnosis of PTSD. The idea of an objective test for PTSD remains extraordinarily appealing to many clinicians and forensic specialists. It is felt that such a test would enhance discrimination between individuals who do and who do not have PTSD.
For other investigators, startle has been less interesting as test for PTSD, and more interesting as a probe in examining central nervous system reactivity in individuals with PTSD. Because so much is known about the neuroanatomical pathways of, and neurotransmitters involved in the startle reflex, several studies have used startle to gain an understanding of neurohormonal functioning in PTSD. Finally, several investigators have used startle as an objective measure of the emotional states of anxiety and fear and have used startle as a tool to elucidate the neural mechanisms involved in the learning and extinction of fear and anxiety.
The startle reflex is one that is shared by very nearly all animals. In basic terms, it is the rapid motor twitch or jump that occurs when an animal or human is exposed to a sudden stimulus (such as a touch, a noise, or a visual image or light). The term acoustic startle reflex refers to the startle response to loud or sudden sounds. In humans, the most consistent and easy way to measure the acoustic startle reflex is to record the speed and intensity of the eye-blink that occurs after someone hears the noise. Technically this is done by recording the electromyographic (EMG) activity of the orbicularis oculi muscle. This method provides information about the intensity (amplitude and the speed (latency) of the eye-blink in response to the sound stimulus. Because the components of the basic startle response are relatively small in number, the time it takes for a sound stimulus to enter the ear and produce a corresponding eye-blink is rather short (21-75 milliseconds). This means that the optimal time to measure the blink component of the startle response is after 21 milliseconds and before 75 milliseconds. Startle responses recorded before 21 milliseconds are considered to be unrelated to the sound stimulus, while those occurring after 75 are considered as potentially random or consciously produced. Thus, it is possible to minimize the likelihood of random and voluntary blink responses. While many investigators have employed the methodology described above (assessment of EMG activity of the orbicularis oculi muscle in response to a sudden burst of white noise) they have not used identical sound stimuli. Some investigations have used sound stimuli of many different intensities and of very short duration (30-40 milliseconds), whereas others have used a single sound stimulus of a much longer duration (500 milliseconds). These distinctions are important to notice when reading the literature about startle for two main reasons: First, it makes it difficult to directly compare studies conducted in one laboratory to those from another; second, there is a great deal of evidence that startle varies according to stimulus intensity, frequency and duration. That the startle response may be directly enhanced by the type and characteristics of the acoustic stimulus is important and relevant to the issue of whether or not exaggerated startle can be used as a marker for PTSD.
One final comment is order regarding methodology: Some authors have used different definitions than those described here when referring to the startle response (such as the heart rate response to a startling sound) while others have used a tactile stimulus (a puff of air). These distinctions are not minor in that the neuroanatomical pathways invoked in the responses are not entirely overlapping
Theoretically, there are a number of ways that exaggerated startle and PTSD might be associated. Investigators interested in whether or not hyperstartle might be a marker in individuals with PTSD know that in healthy subjects, startle shows large variability across individuals, but high consistency within subjects over time. Therefore, it is conceivable, that people who eventually develop PTSD might be those who had high levels of startle prior to the development the disorder. Rather than being caused by exposure to trauma (and/or the development of PTSD) exaggerated startle might be a reflection of a stable trait. At this present time, no studies have measured the startle response in individuals prior to, and after exposure to intense trauma to test out this possibility.
A second possibility is that exaggerated startle in PTSD reflects a persistent sensitization (or heightened responding) of the startle reflex caused by exposure to trauma induced psychological stress. While there is some evidence from animal studies that do support the idea that intense stress may increase the overall startle response, the increased responding tends to be fairly brief in duration and is not thought to be directly related to the reports of chronic and exaggerated startle in humans with PTSD. Nevertheless, there are a number of startle studies in humans with PTSD documenting heightened startle which have invoked the stress sensitization hypothesis (or increased unconditioned responding) to explain the finding. (11-15). Importantly, there are a number of studies which have not found exaggerated, but normal (16-18) or reduced startle in individuals with PTSD (19). �The marked differences in the findings of these studies should dissuade anyone from thinking that startle might prove to be a simple or easy test for PTSD!
Some authors reviewing the startle literature have wondered whether or not exaggerated startle in individuals with PTSD might be reflective of a classically conditioned response. It is reasoned that the sudden bursts of noise might be reminiscent of the sounds of the battlefield or of gun fire and thereby produce an exaggerated (fear-conditioned) startle response. This idea however, has lost considerable explanatory power given the evidence for exaggerated startle in victims of non-combat, non-firearm related trauma who suffer from PTSD.
An alternative way of viewing exaggerated startle in PTSD is to consider the increased startle as an acute state of conditioned fear or anxiety rather than as a conditioned response unique to a specific sound or object. It is possible that exaggerated startle is seen in individuals with PTSD when they are in a state of heightened emotional arousal produced by stress or reminders of their traumatic experiences. Support for this idea comes from several sources.
First, it is well known that startle can be elevated under conditions that are emotionally salient (20). Second, we have examined startle in Vietnam veterans with PTSD in three studies (17, 21-22). In one study, startle was investigated during periods of shock anticipation (threat condition) and during periods when no shocks were administered (safe condition) (21). In the second investigation, subjects were administered placebo and yohimbine on alternative days (22). Startle was elevated in the PTSD patients throughout both experiments. Because startle was elevated in the safe condition of the shock experiment and in the placebo condition of the pharmacological challenge study, we argued that the stress of the experimental context was responsible for the elevated startle. Additional support for this idea came from our third investigation in Vietnam veteran subjects who, when tested under non-stressful conditions did not show exaggerated startle compared to healthy and combat control subjects (17). What these data suggest is that the laboratory conditions affect the emotional state of the subjects participating in startle testing and that anxious or fearful states produce an increase in the startle response.
Support for the idea that exaggerated startle in PTSD is associated with a particular emotional state, rather than a trait (or marker) can also be found in studies which have used the startle response as a measure of central nervous system neuronal reactivity. Because the startle reflex is sensitive to fear/alarm responses, and modulated by CNN noradrenergic neurotransmission, it is an ideal means of assessing these elements in PTSD. In a pharmacological challenge studies involving the administration of yohimbine and placebo on alternative days, Vietnam veterans with PTSD demonstrated significantly greater startle responses to yohimbine compared to placebo and compared to combat controls (22) Startle frequency and startle threshold were respectively increased and reduced by yohimbine. Taken together, the data from this investigation provided evidence that the exaggerated startle seen in individuals with PTSD is, in part, a manifestation of heightened noradrenergic responding.
Finally, startle has been used as a tool to study the learning and extinction of conditioned fear in healthy subjects and in individuals with PTSD. Using a laboratory procedure that involves exposing subjects to stressful electric shocks while they view a series of lights, we have recently completed a 2-day study of conditioned fear in Gulf War Veterans with PTSD (23). On the first day startle was measured before, during and after subjects were exposed to light/shock training trials. During the training trials, only one of the two lights was paired with electric shocks to the subject wrists. This procedure permitted an assessment of whether, and to what degree, startle would increase in the presence of the light that was paired with the electric shocks, an indication that the subjects had learned that the bluelight was linked with an unpleasant shock. At the conclusion of the 1st day, we disconnected the shock electrodes and measured startle in the presence of each light. We repeated the same procedures on the 2nd test day one week later. Several things happened: First, on the initial day of testing, the PTSD patients showed exaggerated startle to the threat light and the safe light, but not the dark; Second, overall startle responses were noted to be increased prior when the subjects came into the lab on the second day. These data suggest that Gulf War veterans with PTSD learn what to be fearful of, but cannot make use of their intellectual knowledge about the safe light to inhibit their fear. Second they exhibit increased fear to the context, or room where the testing occurred.
These findings important because animal studies suggest that different brain systems mediate fear to explicit cues --such as a light indicating shock-- and contextual fear --fear of place where the shocks were experienced (24-27). Lesions of a areas of the brain called the hippocampus (25-26) or the BNST (24) block context conditioning, but not explicit cue conditioning, whereas lesions of the amygdala block both. In addition, inactivation of the BNST, but not the central nucleus of the amygdala, blocks the potentiation of startle by another type of contextual stimulus (such as sustained bright lights (28). These data suggest that the hippocampus and the BNST may be especially important in contextual fear or anxiety, compared to explicit cue fear which is dependent upon the amygdala. The results of the above described study point to a dysfunction of the hippocampus and/or the BNST
So, what have we actually learned about the startle response in PTSD? Well, within recent years, a growing number of studies have reported on the startle response in individuals with PTSD. The majority of these studies have been conducted on male combat veterans, however, there are a few studies involving children and civilian women with PTSD. The startle studies that have been conducted under relatively non-stressful conditions have provided contradictory evidence as to whether startle is increased, decreased or normal in persons with PTSD. By contrast, the studies that have included threat of shock or an I.V. have consistently demonstrated exaggerated startle. Also, it is clear that a person’s age, the length of time since trauma and testing conditions all affect the startle response.
Startle provides a means of studying fear and anxiety states, as well as the learning of fear and anxiety states in individuals with and without PTSD. Since the neural pathways of the startle reflex are well understood, the findings of such studies provide �valuable information about the areas of the brain that may be dysfunctional in individuals with PTSD.
Studies of the human startle response in individuals with PTSD have provided information that may be of benefit to both patients and clinicians. First, the evidence supports the idea that exaggerated startle can be a long-lasting symptom in some individuals with PTSD. Clinicians might educate patients about this possibility so that both parties in the treatment setting might have realistic appraisal of the severity of the condition. Second, it may be helpful for clinicians to remember that increased startle may occur even in the absence of reminders of the traumatic events. For example, studies show quite clearly that startle may be increased by background noise. This may provide a clue as to why some patients find that their startle is increased in mall, or a grocery store, even when they are unable to find anything that them off. Related to this, patients will often complain that in addition to increased startle they feel rattled, agitated and restless. Third, threatening situations or fearful experiences have a powerful augmenting influence on the startle response. The research has clearly shown that the very best way to detect exaggerated startle is to test for it under stressful test conditions. Within the clinical realm this stress may take the form of anxiously anticipating an unwanted event ) such as a bill, compensation & pension examination, or after discussing PTSD-related traumatic events. Clinicians and patients can explore together ways of reducing exposure (when possible) to environments that are likely to produce increases in startle. While there are no well controlled studies examining the effects of medication on the startle response in individuals with PTSD, the animal literature suggests that medications such as buspirone or clonidine may provide some relief.
References
1. Blanchard, E.B., Kolb, L.C., Pallmeyer, T.P., & Gerardi, R.J. (1982). A psychophysiological study of posttraumatic stress disorder in Vietnam veterans. Psychiatry Quarterly, 54, 220-229.
2. Malloy, P.F., Fairbank, J.A., & Keane, T.M, (1983). Validation of a multimethod assessment of posttraumatic stress disorder in Vietnam veterans. Journal of Consulting & Clinical Psychology, 51, 488-494.
3. Pallmeyer, T.P., Blanchard, E.B., & Kolb, L.C. (1986). The psychophysiology of combat induced PTSD in Vietnam veterans. Behavioral Research Therapy, 24, 645-652.
4. Pitman, R.K., Orr, S.P., Forgue, D.F., deJong, J.B., & Claiborn, J.M. (1987). Psychophysiological assessment of post traumatic stress disorder in Vietnam combat veterans. Archives General Psychiatry, 44, 940-975.
5. Orr, S.P. (1990). Psychophysiologic studies of posttraumatic stress disorder. In E.L. Giller (Ed)., Biological assessment and treatment of posttraumatic stress disorder (pp. 135-157). Washington, DC: American Psychiatric Press.
6. Grinker, R.R., & Spiegle, J.P. (1945). War neuroses. Philidelphia: Blakiston.
7. Southard, E.E. (1919). Shell shock and other neuropsychiatric problems presented in 589 case histories from the war literature, 1914-1918. Boston: W.M Leonard.
8. Southwick, S.M., Morgan, C.A., Nagy, L.M., Bremner D., Nicolaou, A., Johnson, D.R., Rosenheck, R., & Charney, D.S. (1993). Trauma-related symptoms in veterans of Operation Desert Storm: A preliminary report. American Journal of Psychiatry, 150, 1524-1528.
9. Southwick, S.M., Morgan, C.A., Nicholau, A., Darnell, A., & Charney, D.S. Trauma-related symptomatology in Desert Storm Veterans: Two year follow-up. American Journal of Psychiatry, 152, 1150-1155.
10. Kardiner A. (1941). The traumatic neurosis of war. New York: Paul Hoeber.
11. Butler, R.W., Braff, D.L., Rausch, J.L., Jenkins, M.A., Sprock, J., & Geyer, M.A. (1990). Physiological evidence of exaggerated startle response in a subgroup of Vietnam veterans with combat related PTSD. American Journal of Psychiatry, 147, 1308-1312.
12. Morgan, C.A., Grillon, C., Southwick, S.M., Davis, M., & Charney, D.S. (1996). Exaggerated acoustic startle in Gulf War Veterans with PTSD. American Journal of Psychiatry, 153, 64-68.
13. Morgan, C.A., Grillon, C., & Southwick, S.M. (1997). Startle abnormalities in women with sexual assault related PTSD. American Journal of Psychiatry, 154, 1076-1080.
14. Orr, S.P., Pitman, R.K., & Shalev, A.Y. (1995). Physiologic responses to loud tones�in Vietnam veterans with posttraumatic stress disorder. Journal Abnormality Psychology, 104, 75-82.
15. Shalev, A.Y., Orr, S.P., Peri, T., Schreiber, S., & Pitman, R.K. (1992). Physiologic response to loud tones in Israeli patients with posttraumatic stress disorder. Archives General Psychiatry, 49, 870-875.
16. Cuthbert, B.N., Drobes, D.J., Patrick, C.J. and Lang, P. J. (1994). Autonomic and startle responding during affective imagery among anxious patients. Psychophysiology, 31, S35.
17. Grillon, C., Morgan, C.A., Southwick, S.M., Davis, M., & Charney, D.S. (1996). Baseline Startle amplitude and prepulse inhibition in Vietnam veterans with posttraumatic�stress disorder. Psychiatry Research, 64, 169-178.
18. Orr, S.P., Solomon, Z., Peri, T., Pitman, R. K. & Shalev, A.Y. (1996). Physiologic responses to loud tones in Israeli veterans of the 1973 Yom Kippur War. Biological Psychiatry, 41, 319-326.
19. Ornitz, E.M., & Pynoos, R.S. (1989). Startle modulation in children with posttraumatic stress disorder. American Journal Psychiatry, 146, 866-870.
20. Lang, P.J., Bradley, M.M., & Cuthbert, B.N. (1990). Emotion, attention, and the startle reflex. Psychology Review, 97, 1-19.
21. Morgan, C.A., Grillon, C., Southwick, S.M., Davis, M., & Charney, D.S. (1995). Fear potentiated startle in PTSD. Biological Psychiatry, 38, 378-385.
22. Morgan, C.A., Grillon, C., Southwick, S.M., Nagy, L.M., Davis, M., & Charney, D.S. (1995). Yohimbine facilitated acoustic startle in combat veterans with PTSD. Psychopharmacology, 117, 466-471.
23. Grillon C., & Morgan, C.A. (In review). Fear-potentiated startle conditioning to explicit and contextual cues in Gulf War veterans with posttraumatic stress disorder.
24. Davis, M., Gewirtz, J. C., McNish, K. A., & Kim, M. (1995). The roles of the amygdala and bed nucleus of the stria terminalis (BNST) in the acquisition of fear potentiated startle using both explicit and contextual cues. Society for Neuroscience Abstracts, 21, 1224.
25. Kim, J.J., & Fanselow, M.S. (1992). Modality-specific retrograde amnesia off ear. Science, 256, 675-677.
26. Phillips, R.G., & LeDoux, J.E. (1992). Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning. Behavioral Neuroscience, 106, 274-285.
27. Wilkinson, L. S., Humby, T., Robbins, T. W. & Everitt, B. J. (1995). Differential effects of forebrain 5-hydroxytryptamine depletions on Pavlovian aversive conditioning to discrete and contextual stimuli in the rat. European Journal of Neuroscience, 7, 2042-2052.
28. Walker, D.L. (1996). AMPA receptor blockade in the bed nucleus of the stria terminalis (BNST) but not the central nucleus of the amygdala (CNA) disrupts light-enhanced startle: A novel paradigm for the assessment of anxiety in rats. Society for Neuroscience Abstract, 22, 1117.
C. A. Morgan III is Associate Professor of Psychiatry and Research Affiliate, History of Medicine Yale University School of Medicine and Associate Director, Post Traumatic Stress Disorder Program, VA Connecticut Healthcare System, Neuroscience Division of the National Center for PTSD.
|
