1. Introduction
Stress is a reaction to a stressor. A stressor can be defined as any external threat through which an organism cannot sustain optimal homeostasis. Homeostasis is the mechanism by which biological systems try to maintain stability so that internal conditions remain optimal, which is vital for survival. A stressor is, roughly said, anything that causes stress to an organism. It’s important to note that a stressor is subjective, meaning that for one person the situation causes stress, while for someone else the situation does not trigger a response. The nature of the stressor can be acute, for example an exam or be confronted with a spider, or chronic, for example high pressure at work or a sick family member that needs constant care. The theory of Selye is an old vision on stress, but has been the basis for stress related research later on. The theory of Selye is also known as the general adaptation syndrome (GAS). The starting point of is theory is this: if a person perceives a situation as stressful, a stress response will occur. This model consists of three stages. The first stage is what’s called the alarm reaction. We can understand this stage as the first reaction to a stressor. This is more commonly known as the ‘fight-or-flight-response’. It prepares the body to make an adequate stress response by producing adrenaline, loss of body fluids, sharpened vision, higher heart rate and less blood flow to the body ends like fingers and toes. The second stage is called the stage of resistance. Here, the person is still faced with the stressor and needs to adapt in order to face the stressor. The final stage is called the stage of exhaustion. If this stage occurs, the person faces problems in his daily life as a result of prolonged stress. The person was not able to reach new homeostasis, nor did he adapt well to the stressor (Van der Feltz-Cornelis, 2015). According to Hunter, Minnis and Wilson (2011) a normal response to an acute stressor comprises rapid activation of the sympathetic nervous system (SNS), which is part of the autonomic nervous system (ANS). The SNS can be seen as the accelerator (Swinnen & Coeckelbergh, 2017).This is followed by the activation of the hypothalamic-pituitary-adrenal (HPA) axis. This axis accounts for the production of the glucocorticoid hormones, including cortisol as most important hormone. In a normal situation, there is a negative feedback system to stop the stress response when the time is right. The activation and deactivation of this stress response system helps an organism to handle a threat and return to a normal level of functioning (Hunter et al., 2011). The other part of the ANS is the parasympathetic nervous system. This system can be seen as the brake of the stress response (Swinnen & Coeckelbergh, 2017). Contrary to the sympathetic nervous system, responses coordinated by the parasympathetic nervous system (PNS) are slow and targeted to specific organs. While the sympathetic nervous system prepares the body for a ‘fight-or-flight-response’, the actions of the parasympathetic nervous system can be described as the ‘rest-and-digest-response’. These actions can be understood as being complementary to that of the sympathetic nervous system. Stimulation of the PNS leads to calming adjustments. These include vasodilation, lower heartrates, and/or decreased contractility (Laizzo, 2015). The PNS counterbalances both inflammatory responses and sympathetic activation (Shonkoff & Garner, 2012).
Mostly, the psychophysiological response to a stressor is that the body gets prepared for physical activity. When stress is used in this way, one can see the benefits of this mechanism. However, in modern times stress as a result of psychosocial stimulation can be seen as unsuitable. The organism will be aroused for physical activity, yet rarely it Is allowed. The result is that the stress response hardly follows (Everly, Jr & Lating, 2013). Or imagine a man walking in the woods and suddenly faces a snake. In this situation, it is good to have a quick ‘fight-or-flight-response’. It’s a life threatening situation. Now, in modern times stressors are from a different order. Situations causing stress are (most of the time) not life threating. Yet our ‘fight-flight-response’ will be activated (Swinnen & Coeckelbergh, 2017). In sum, the body is prepared for a ‘fight-or-flight-response’ ( ex. elevated heartbeat, losing excessive weight), yet there will be no cooling down.
If all goes well, the body will return to normal levels of functioning. This process is called allostasis. We can see allostasis as the process by which a person can maintain his or her homeostasis ( Van der Feltz-Cornelis, 2015). Allostatic load refers to the burden once body has to fulfill the allostasis (Van der Feltz-Cornelis). It can be seen as the ‘ wear and tear on the body’. Allostatic overload can influence the working of the HPA-axis, metabolic pathways and the immune system (Murphy, Cohn & Loria, 2016). When an individual experiences stress without recovery and relaxation, chronic stress occurs (Swinnen & Coeckelbergh, 2017). Chronic stress can lead to a broad range of complaints (Van der Feltz-Cornelis, 2015; Swinnen & Coeckelberg, 2017). It can manifest itself in physical complaints (pain, fatigue, nausea, hyperventilation) as in psychological complaints (depersonalization, panic attacks, pessimism, problems with concentration) (Van der Feltz-Cornelis, 2015; Swinnen & Coeckelberg, 2017).
Stress in early life has been found to exert profound short-and long-term effects on human physiology. These effects can be found in the nervous system as well as peripherally (Syed & Nemeroff, 2017). It is clearly demonstrated that early life stress is associated with many psychiatric disorders (Syed & Nemeroff, 2017). Prenatal stress and maltreatment as a child are correlated with an abnormal HPA axis (re)activity (Frodl & O’Keane, 2013). Exposure to chronic stress considerably raises vulnerability to negative medical outcomes. Chronic stress is correlated with day-to-day output of cortisol and abnormal HPA activity (Miller, Chen & Zhou, 2007). As mentioned above, the HPA-axis normally has a negative feedback system to stop the stress response. This activation and deactivation helps the individual to successfully cope with a stressor (Hunter, Minnis & Wilson, 2010). If there is an abnormal stress response, the HPA axis may not be able to response well to a stressor, which in turn can lead to long-term consequences for memory, mental illness, behavior and physical complaints like pain and fatigue (Hunter, Minnis & Wilson, 2010). In a review of Hunter et al. (2010), the authors mention that there is evidence from both animal and human studies that misery in earlier life affects the pathological stress responses in adulthood. A meta-analysis of Miller et al. (2007), showed that chronic stress indeed alters the activity of the HPA axis, but that this modification is dependent of nature of threat and the person facing it and how he or she respond. People who were sexually abused show abnormal levels of corticotropin-releasing factor and adrenocorticotropic hormone (Hunter et al., 2010). However, although chronic pain syndromes are related to sexual abuse and/or childhood trauma, results are conflicting. Some studies support the idea, while others fail to show a connection (Lawson & Wallace, 2015).
Many research has shown that different psychiatric and somatic disorders are related to dysfunctions in stress reactivity. Functional somatic syndromes (like fibromyalgia, chronic fatigue syndrome and irritable bowel syndrome) (FSS) are syndromes that are characterized more by disability, adversity and symptoms without the presence of a clear, verifiable abnormality (Kanbara, Fukunaga, Mutsuura, Kitamura & Nakai, 2007). A potential mechanism that connects psychosocial stress with FSS is ANS dysfunction (Tak, Riese, de Bock, Manoharan, Kok & Rosmalen, 2009). A meta-analysis showed a lower activity of the PNS in patients with FSS in comparison with controls (Tak et al., 2009). A study of Boneva et al. (Boneva, Decker, Maloney, Lin, Jones, Helgason, Heim et al., 2007) found a higher heart rate and reduced heart rate variability in patients with CFS. Several studies found a lower heart rate variability in people with clinical burnout (Sjörs & Jonsdottir, 2015; Lennartsson, Jonsdottir & Sjörs,2016). One study (Juster & Durand, 2013) showed that global burnout and emotional exhaustion demonstrated the most tough and steady negative association to cortisol in the afternoon and evening. Lower cortisol levels were found as soon as 30 minutes after waking up till sleep time. A study of Han (Han, Stegen, Schepers, Van den Bergh & Van De Woestijne, 1998) showed that patients with panic varied with other anxiety disorders in terms of a drop in FETCO2 at rest. Data from Nardi et al. (Nardi, Valença, Nascimento & Zin, 2001) proposes that there is a link between panic attacks and hyperreactivity to an acute hyperventilation test. Reduced heart rate variability is found in patients with panic disorder (Klein, Cnaani, Harel, Braun & Ben-Haim, 1995; Garanki, Martinez, Aaronson, Voustianiouk, Kaufmann & Gorman, 2008). A study of McTeague & Lang (2012) compared the whole spectrum of anxiety disorders with healthy subjects on the startle reflex. This was used as a physiological measure of fear. The brain’s fear and defense circuit mediates the startle reflex. Their results showed that fear circuitry can be altered in chronic, pervasive anxiety. Finally, Smets et al. (Smets, Schiavone, Velazquez, De Raedt, Bogaerts, Van Diest & Van Hoof, in progress) demonstrated that physiological signals during a stress tasks can be used to differentiate healthy patients from patients with stress related complaints. In sum, we can see that there are modifications of the stress response system in people with FFS, panic disorder and burnout. However, there is variability (sometimes stronger response is found, sometimes decreased response).
When making a diagnosis, health practitioners tend to rely on symptom reporting of the patient. In the medical model, it is assumed that patients tend to have a high level of interoceptive accuracy (Bogaerts, Van Diest & Van den Bergh, 2014). Interoceptive accuracy can be defined as the within-subject correspondence between the corresponding physiological referent and a perceived sensation (Bogaerts, Van Eylen, Li, Bresseleers et al., 2010). People having panic complaints perform better on the heartbeat detection task. This suggests that in children and adults higher levels of panic symptoms are associated with increased ability to perceive internal physiological cues (Eley, Stirling, Ehlers, Gregory & Clark, 2002). Ehlers & Breuer (1996) found in their study a remarkable heartbeat perception in the subgroup of panic disorder. In contrary to people with panic symptoms, studies show that in people with FSS perceived symptoms are less strongly linked to relevant physiological parameters (Bogaerts et al., 2010). Van den Bergh et al. (Van den Bergh, Witthöft, Petersen, & Brown, 2017) showed that this relationship is highly variable and depends on personal characteristics, contextual factors and interaction of both. This does not mean, however, that they imagine these complaints. The brain proceeds and regulates bodily signals in an atypical way. This leads to identify these signals as painful physical symptoms (Bogaerts & Van Oudenhove, 2013; Bogaerts, Van Diest & Van den Bergh, 2014). Trait negative affect (NA, Watson & Clark) seems to play a role. Trait NA refers to susceptibility to interpret situations as more negative and threatening (Watson & Clark, 1984,1992). A steady correlation (.40) exist between trait NA and symptom reporting (Bogaerts et al., 2014). Another study showed that in a healthy population high habitual symptom reporters differ from low habitual symptom reporters in interoceptive accuracy (Bogaerts, Millen, Li, De Peuter, Van Diest, Vlemincx, Fannes & Van den Bergh, 2008). Di Lernia et al. (Di Lernia, Serino & Riva, 2016) found in their review that patients with chronic pain conditions showed a lower interoceptive accuracy. To our knowledge, there are no studies regarding interoception in patients with burnout. In sum, we can say that in patients with FSS there is a low relationship between interoception and physiological parameters, whereas in patients with panic symptoms a high interoceptive accuracy is found.
If we look at all these results, we can hypothesize that complaints in patients with panic disorder are more driven by autonomic dysfunction (stress reactivity), while complaints in patients with FSS show more deficiencies in symptom perception and interoceptive accuracy. Burnout can be placed between these 2 extremes. So if we place these conditions on a continuum of stress reactivity, panic disorder is more disturbed followed by burnout and FSS.
On the other side, it is possible that there is variability within each patient group. We hypothesize that chronicity plays a role in such way that at the beginning of a disorder, complaints are more driven by a dysregulated stress reactivity while after some time problems with symptom perception begin.
In our study, we test these hypothesis by comparing healthy subjects with 4 patient groups and by comparing these patient groups to each other on stress reactivity. We do not, however, test symptom perception. We use the following physiological parameters: (1) CO2-levels at rest and after hyperventilation provocation test and (2) autonomic parameters (skin, temperature and EMG) at baseline level and after 3 stress tasks. We also compare these groups on NA.
Next, we will discuss the patient groups.
Fibromyalgia/ Chronic fatigue syndrome (CFS)
Fibromyalgia (FM) is functional somatic syndrome characterized by musculoskeletal pain, fatigue, depression, cognitive dysfunction and sleep disturbance (Hassett et al., 2007). The diagnosis of fibromyalgia needs to be done by a rheumatologist. The criteria are based on the ACR-criteria (American College of Rheumathology) This protocol (2010) states that (1) a score of 7 or more on the Widespread pain index (WPI) and symptom severity (SS) of 5 and more, or WPI-score between 3-6 and SS score over 9 (2) symptoms are present for 3 months at a similar level and (3) no other disorder is present that can explain the pain. Psychological stress seems to be part of possible factors that attribute to the development of this disease. (Lawson & Wallace, 2015). However, most rheumatologists still use criteria from 1990. Lawson & Wallace (22015) refer to these criteria in their book (widespread pain for minimum 3 months, 18 bodily parts are determined on forehand of which at least 11 must hurt). A meta-analysis of Afari et al. (Afari, Ahumada, Wright, Mostoufi, Golnari, Reis & Cuneo, 2013) showed that people exposed to a trauma had almost 3 times (2.7) more chance to have a FSS. A significant relation between physical and sexual abuse with FM was found (Häuser, Kosseva, Üceyler, Klose & Sommer,2011). Jones (2014) found a connection between emotional and physical trauma with FM. A shortcoming in the central nervous system has been found in patients with FM. More particular, a failure in the descending inhibitory pathway has been documented (Lawson & Wallace,2015). This could be an explanation of why they tend to focus on pain. Also, emotional trauma and early life adversity are risk factors for the progress of FM in adulthood. Patients with FM have been found with diminished HRV over short and long (24 hours) periods (Hassett et al. 2007). According to Hassett et al. HRV biofeedback is a potential helpful treatment for people with FM. However, in a critical review of Wheat and Larkin (2010) the authors state that the presence of both short and long-term effects of HRV is less clear.
Diagnostic criteria for CFS were formulated in 1994. The criteria states that there is physically unexplained fatigue, for at least 6 months with at least 4 of the next symptoms: muscle- and/or joint pain, concentration- and/or memory problems, still tired after getting sleep, headaches, sore throat and painful lymph nodes, and feeling unwell after labor (Van der Filtz-Cornelis, 2015). In Belgium, the diagnosis of CFS can only be made by a recognized CFS-referencecentre.
Hyperventilation is considered as a concomitant and possible maintaining factor that contributes to the symptom pattern of CFS. Research suggests that people who expect a distressing and demanding activity in which high arousal is expected, tend to hyperventilate more. This means that it is not a stable characteristic of the patient, as was previously considered. In this view, hyperventilation is more likely an important perpetuating factor than it is a causal factor (Bogaerts et al. 2007). Patients suffering from CFS show modifications of both PNS as SNS function (Freeman & Komaroff, 1997) and central nervous system (Siemionow, Fang, Calabrese, Sahgal & Yue, 2004). Comorbidity between FM and CFS has been shown (Aaron, Burke & Buchwald, 2000; Friedberg & Jason, 2001; Hassett et al., 2007; Geisser, Donnell, Petzke, Gracely, Clauw & Williams, 2008). Marcus and Deodhar (2011) stated that for 81% of patients with FM, CFS is reported.
Nervous exhaustion (Neurasthenia)
Nervous exhaustion, also called neurasthenia, is described as a syndrome with mental as well as physical fatigue. It is followed by withdrawal and depression (http://academic.eb.com.kuleuven.ezproxy.kuleuven.be/levels/collegiate/article/neurasthenia/101820). It is important to note that brief fatigue in a healthy, solid person caused by a demanding activity, which can be resolved by eating and sleeping, is not viewed as neurasthenia. Neurasthenia should be seen as the consequence of continued overstrain (mentally and/or physically), inadequate food and rest for a long period resulting in loss of nerve power (Jewell, 1878). Temporary fatigue is a normal process, neurasthenia should be seen as the poorly grade. The nervous system is not destroyed, however it is extremely exhausted. It is not fixed with adequate rest (Jewell, 1878). People have trouble to maintain control over their expectations. Most, not all, stop working. There are significant limitations in social life and work. Work pressure contributes to nervous exhaustion. The period in which the complaints are started, is not more than 12 weeks (Hoeymans, Eysink &de Hollander, RIVM rapport 270052001/2005). When there is a longer period (at least 12 months), fatigue and exhaustion are the main complaints and they are related to work stress, diagnosis of burnout is made (DSM-IV). One could see burnout as the end phase of nervous exhaustion. In depression, there mostly will be an increased activity of the HPA axis, while in pain disorders (like neurasthenia), hypocorticolism is found (Van der Feltz-Cornelis, 2015). In this paper, we use the word ‘overstrain’. Overstrain is often referred to as nervous exhaustion/neurasthenia in literature.
Burnout
The term burnout stems from the personal and professional observation of Freudenberger. He described numerous behavioural, emotional and cognitive exhausting factors among overcommitted community workers (Marchand, Juster, Durand & Lupien, 2013). We can describe burnout as the protracted depletion of an individual’s energies (see above, GAS theory). Burnout is characterized by emotional exhaustion, reduced personal accomplishment and feelings of insufficiency and depersonalization. One hypothesis is that burnout can be seen as the response of an individual to persistent stress. It’s a subjective experience. The response consists of emotional core elements and the individual is displaying psychic and somatic symptoms. Burnout is often mentioned in the context of work and specific occupation groups, but is not restricted to this. It often has an association with stress. As until today, burnout hasn’t been demonstrated to be a predictor of a mental disorder, but some studies show it to be related to mental disorders. There is growing evidence that burnout could be associated with distinct changes in diurnal production of the stress hormone cortisol (Marchand, A., Juster, R-P., Durand, P. et al, 2013). It is important to make a differentiation between burnout and some other psychiatric diagnosis, because burnout symptoms are similar to some of the symptoms of depression. A critical difference between burnout and chronic or persistent fatigue, is that chronic fatigue from job stress is a commonly seen process in the general population. People often see burnout and fatigue as the same phenomenon, but there is a difference. People with chronic fatigue (not to be confused with CFS) will return to a normal level of energy after an adequate time of rest. If the exhaustion is prolonged nevertheless, a burnout or mental disorder can occur (Bährer-Kohler, 2013). The distinction between CFS and burnout is that in people with CFS physical discomfort with pain and fatigue are most prominent, while in burnout people suffer from exhaustion and tend to withdraw themselves (Van der Feltz-Cornelis, 2015). The difference between burnout and depression is that burnout can be a risk factor for developing a depression (Bährer-Kohler, 2013). The symptoms of a burnout can be found in people suffering from a depression. If this is the case, than the diagnosis of depression is made and treatment must be started for depression (Bährer-Kohler, 2013). The emphasis lies on pessimism and hopelessness (Van der Feltz-Cornelis, 2015). Despite the common believe that women are more likely to suffer from a burnout than men, a study by Purvanova and Muros (2010) failed to find such results. Men and women differ in the way they report about burnout. Meta-analysis of gender differences showed that men and women experience burnout in a different way.
Panic Disorder (PD)
Panic disorder is a common mental disorder, yet it has received little historical attention (Nardi & Freire, 2016). A panic disorder is a disorder that is characterized by episodic attacks of acute anxiety (Carlson, 2010). It consists of periods of acute and ongoing terror. These periods can grip people for some time, lasting from seconds to a few hours (Carlson, 2010). It reaches a peak in intensity in a timeframe of 10 minutes (Parks & Marek, 2007). When someone experiences a panic attack, they report many physical symptoms like shortness of breath, dizziness, clammy sweat and irregular heartbeat. Victims also often report the feeling that he or she is going to die during an attack (Carlson,2010). A theory that is widely accepted considering panic disorder, is the so called ‘ False Suffocation Alarm Theory’ of Donald Klein (Nardi & Freire, 2016). This theory suggest that a panic attack actually is a disorder of the physiological suffocation alarm. Here the central nervous system would be informed of an imminent suffocation situation by the regulating monitor when this wasn’t actually happening (Nardi & Freire, 2016). It is believed that panic attacks stems from a network of fear with adjusted sensibility. This network include: the prefrontal cortex, the thalamus, amygdala, insula and amygdala extensions into the hypothalamus and brain stem (Nardi & Freire, 2016).
2. Methods
Participants
Data collected from 120 patients in Tumi therapeutics was retrospectively analysed. Tumi therapeutics is an expertise center that is specialised in the diagnosis and treatment of hyperventilation, medically unexplained symptoms and stress-related diseases. We analysed data from 30 individuals diagnosed with FM/CFS, 30 individuals with panic disorder, 30 overstrained individuals and 30 individuals with burnout. The control group consisted of 30 healthy individuals who were relatively matched for age (ranging from 18-65 years) and gender. Healthy participants were selected on an inclusion criteria (cutoff point ≤75) based on a physical symptom checklist questionnaire. An adaptation of the Checklist for Symptoms in Daily Life (CSDL, Wientjes & Grossman, 1994) was used ( see Bogaerts, Millen, Li, De Peuter, Van Diest, Vlemincx, Fannes & Van den Bergh, 2007). This questionnaire consisted of 39 questions targeting physical complaints (ex. headaches, nausea, vomiting,…) on a 5-point Likert scale Healthy participants that met the inclusion criteria were invited to the lab for a test session. Before coming to the lab, at home, healthy participants filled out the battery of questionnaires that is also used at Tumi therapeutics. The study was approved by the Social and Societal Ethics Committee of the University of Leuven. Healthy controls gave written informed consent before participating in the study. Patients at Tumi therapeutics gave informed consent to use their data for scientific purpose during the intake. Healthy participants were rewarded with a gift card of 15 euros.
Materials and apparatuses
Self-report measures
The Positive and Negative Affect Schedule (PANAS) consists of two scales, one measuring positive affect, while the other measures negative affect (Watson & Clark, 2016). Participants answer 20 items by using a 5-point scale. This scale ranges from 1 (=not at all) to 5 (=extremely) (Watson & Clark, 2016). The items consist of feelings, and the respondent answers to what rate he or she feels this way in everyday situations. The reliability of the PANAS has been shown by Crawford & Henry (2004) as well as the validity.
Physiological measures
The NeXus 10 Mark II (Mind Media, Herten, The Netherlands) was used to measure psychophysiological signs. With the help of NeXus 10, we could measure multiple physiological reactions simultaneously. Skin conductance was measured with two Ag/AgCL sensors attached to the middle finger and the ring finger and was recorded at 32 Hz. A thermistor was used for measuring skin temperature, placed at the little finger and was also recorded at 32 Hz. All sensors where placed on the non-dominant hand. Finally, muscle tension was measured at the trapezius muscle by four electrodes, two at each side of the neck and recorded at 32 Hz. The Bio-Trace 10 software was used to record these physiological reactions and to present the standardised stress tests included in the BioTrace package. During the test, the participants were asked to keep their hand steady. The physiological signs were visually inspected and processed offline.
A handheld capnography (Oridion) was used to measure fractional end-tidal CO2 (FetCO2). It was measured through a nasal cannula. FetCo2 and respiration rate (RR) were noted every 15 seconds by the researcher.
Procedure
This examination consisted of three components: a capnography test, a long stress test and a short stress test. For both the stress tests, we use the BioTrace+ software.
Capnography test
The first component was a capnography test. In this test, the feedback of end-tidal carbon dioxide (FetCO2) was examined. This was measured during the baseline and while doing two exercises. First, FetCO2 and RR of the subject was measured for five minutes while he or she was in rest. After that, the subject was asked to exhale deeply for five times. After a five minutes recovery phase, the subject was asked to do the hyperventilation provocation test, in which hyperventilation is induced voluntarily by following a certain breathing rhythm indicated by the experimenter. The hyperventilation provocation test ended when the subjects claimed to experience physical complaints. After this, they could breathe normally again.
Long stress test
The second part was the long stress test. This test lasted for fifteen minutes. In this test, the physiological reactions of the subject were measured when the person was at rest (baseline) when confronted with three stressors, and in recovery from these three stressors. The first stressor was the stroop colour word task. The aim of this task is to say the colour the word is written in aloud, and suppress the automatic impulse to read the word itself. The words appeared in groups of 5 and switched every 5 seconds. The words itself are also a colour ( ex. the word red is written in blue). The participants were motivated to respond faster and faster, as extra stressor. The second stressor was a math task in which the subject needed to subtract seven, starting from 1081. Like with the stroop colour word task, the participant was asked to respond faster each time. The final stressor was called the stress talk. The purpose of this stressor was for the subject to talk about a stressful memory to the experimenter, focusing on the emotions and feelings he or she felt during this event. All these stressors lasted for two minutes, and after each stressor, the subject had a recovery period of two minutes. During this recovery phase, the participant was shown a relaxing photo and relaxing music in the background and they had to close their eyes after the first stressor until the end of the experiment. It’s important to know that the interest of this study did not lie in the accuracy of the answers or how fast they reacted. The main focus was the physiological reaction of the subject and how he or she recovered.
Short stress test
The final part was a short stress test. This test lasted for six minutes. Again, the test started with determining a baseline for three minutes. After that, an announcement, stating that a stressor would start soon, was presented on the screen (lasting for 20 seconds) and finally, the subject was confronted with negatively loaded pictures and loud noises for thirty seconds. When this was over, the subject had a two minute recovery period, again with their eyes closed.