Article Text
Abstract
People with epilepsy report that stress is their most common trigger for seizures and some believe it caused their epilepsy in the first place. The extensive preclinical, epidemiological and clinical studies examining the link between stress and epilepsy have given confusing results; the clinical studies in particular are fraught with confounders. However stress is clearly bad for health, and we now have substantial preclinical evidence suggesting that chronic stress worsens epilepsy; in selected cases it may even be a causal factor for epilepsy. Healthcare professionals working with people with epilepsy should pay more attention to stress in clinical practice. This review includes some practical advice and guidance for stress screening and management.
- seizures
- STRESS
- mechanisms
- treatment
- EPILEPSY
Statistics from Altmetric.com
Introduction
Providing good care for people with epilepsy is much more than ensuring optimal seizure control and minimising side effects of antiepileptic drugs, particularly when seizures persist despite best treatment.1 Psychological and psychiatric comorbidities are strongly associated with health-related quality of life and costs in epilepsy2 but frequently go undetected and undertreated. Stress is the most commonly reported trigger for seizures in people with epilepsy, and is sometimes put forward as the cause of their epilepsy. Stress is not, of course, a new phenomenon. We all experience it at some point (figure 1) and the media increasingly portrays it as a plague of modern life. How often do our patients report that their recent seizure was triggered by stress and responded with simple reassurance or an increase in medication? Or do we dismiss a patient's claims that their epilepsy resulted from a difficult childhood, and concentrate instead on their brain scan result? Unless you can honestly say never to both, this article is for you. We have based the report upon a comprehensive review of all articles identified from a search of the terms ‘stress’ together with ‘seizure’ and/or ‘epilepsy’ on PubMed and Web of Science databases, together with hand searches of the cited articles.
What is stress?
Stress is defined as the physiological and/or behavioural response to an event or events interpreted as threatening to the individual. The psychology and biology of stress are now well understood (figure 2 and box 1). Stress occurs when environmental demands exceed an individual's adaptive capacity, resulting in psychological and biological changes. Stress is a process that involves stressors, stress appraisal and stress responses.3 The brainstem and hypothalamus appraise physiological stressors (eg, pain), whereas complex emotional and experiential stressors are appraised in multiple limbic forebrain structures with inputs from higher order sensory processing and memory.4 Both systems activate the paraventricular nucleus of the hypothalamus and the brainstem (via direct projections) and the limbic system (indirectly). The paraventricular nucleus releases the corticotropin-releasing hormone, which activates both the sympathetic adrenergic pathway of the autonomic nervous system and the hypothalamic-pituitary-adrenal (HPA) axis. Sympathetic activation increases both circulating adrenalin (primarily from the adrenal medulla) and noradrenaline (primarily from the sympathetic nerves). The resultant behavioural changes include the largely protective fight or flight reactions: these can become maladaptive, for example by increasing eating, smoking, drug use and vigilance, leading to anxiety and worrying. In the HPA axis, both corticotropin-releasing hormone and vasopressin released into the portal system stimulate the anterior pituitary gland to produce adrenocorticotropic hormone. This in turn stimulates the adrenal cortex to release cortisol, which promotes the mobilisation of stored energy, and potentiates several sympathetically mediated effects. Thus both systems work in a complementary fashion, with additional feedback loops to terminate the response, influencing appraisal and thus adaption.
Stages and terminology of stress
Stressor: events or experiences objectively associated with threat. These can be physical (eg, pain, fatigue), psychological (eg, fear, sadness) or environmental (eg, temperature, noise). Most stress that patients report is psychosocial.
Appraisal: The assessment of threat in the context of individual coping resources/adaptive capability. This may involve conscious and subconscious mechanisms.
Response: The resultant biological and behavioural changes involving activation of autonomic and endocrine pathways. These can be adaptive (‘fight or flight’) but can become maladaptive, including for example, increased eating, smoking and hypervigilance leading to anxiety.
The time course of the stress and our response is crucial (figure 3). The acute stress response evolves as an adaptive process to allow the body to maintain a constant internal state in the face of a changing environment (allostasis); in moderation, this is beneficial. However, when stress is excessive, the cost of reinstating homoeostasis can become too high with potentially harmful consequences. This inappropriate stress response, or ‘allosteric load’5 produces a vulnerable phenotype and leaves genetically predisposed people at increased risk of chronic diseases, both psychiatric and physical. Different situations can create chronic stress through stress event sequences, for example, fired from job, chronic intermittent stressful events, conflict with neighbours or ongoing chronic stress conditions, including living with disability.
Is there clinical evidence that stress causes seizures/epilepsy?
In retrospective studies, patients consistently identified stress as a major seizure trigger in 10–83% of people with epilepsy from across the world, including UK, USA, Australia, Singapore, Brazil and Scandinavia, in both tertiary and community settings.6–8 Stressful life events can also trigger new-onset seizures,9 and patients with epilepsy who report stress as a seizure trigger are more likely to have experienced early life stress from childhood traumatic experiences, particularly emotional abuse.10 However, such studies inherently depend upon patient perceptions and memories, and often have inherent selection biases, and so are unreliable. Loss of self-control is the most disturbing psychosocial consequence of epilepsy and it is a natural likely coping mechanism to try to find connections between their seizures and external or internal events to give a feeling of predictability.11
There are also studies of seizures and epilepsy that follow traumatic events. For example, people evacuated due to flooding in the Netherlands had more seizures compared with controls (reviewed elsewhere6). Similarly, a study of children with epilepsy during and after the 1991–1992 Croatian war showed that those from directly affected areas had more seizures than before the war, compared with those from unaffected areas. However, the actual numbers were small (30–60 per group), the studies are retrospective, and stress was inferred rather than confirmed/measured. Furthermore, disasters on this scale are very prone to confounding factors such as non-adherence with medication and sleep deprivation, as in a study from the Persian Gulf War. A large Danish population-based national registry showed a 50% increased risk of developing epilepsy in bereaved parents; this was not explainable by sociodemographic factors, and is truly population based, although it was not possible to exclude misdiagnoses or the effects of potential lifestyle confounders such as alcohol and substance abuse. A retrospective analysis of medical information and duty assignments (combat, maintenance or administrative) from over 300 000 men compulsorily drafted into the Israeli Defence Forces and followed for 30 months showed a slightly higher risk of new-onset seizures in the ‘high stress’ combat group than in other units. However, there was no effect on risk of seizure recurrence in those with previous or current epilepsy.
The few prospective available studies are hampered by a lack of clear definitions and standardised assessment tools. Stressful verbal stimuli presented to healthy controls can produce electroencephalography (EEG) changes that are subtle (narrowing of the bandwidth and regional changes in frequency), but sufficiently large for a blinded reviewer to correctly identify 92% of stress stimuli on EEG alone.12 Similarly in people with various epilepsies, stressful interviews induced changes (exaggerated spiking, paroxysmal activity or epileptiform complexes) in the majority13 and there are case series of stress-inducing audio/video recordings inducing seizures in patients with temporal lobe epilepsy in particular.14 There are a few larger or longer-term truly prospective studies, some of which show an increased risk of seizures with stress15 and some where the effect of stress could not be distinguished from the effects of confounders such as sleep deprivation, alcohol and missed medication.16
Thus, although clinical studies give reasonable support to the theory that stress might lower seizure thresholds, it is difficult to separate the effects of stress per se from confounders, or to separate cause from effect. In addition, it is often difficult to disentangle the subjective perception of stress from other sensations that precede seizures or from mood disorder symptoms.17 In fact, changes in the brain before a seizure might increase a patient's perceptions of stress and anxiety rather than reduce it.
Does stress cause seizures/epilepsy? Neurobiological evidence
Interestingly, preclinical animal studies suggest that acute stress protects against seizures, although this varies in different models.18 The largest group of experiments is in rodent models exposed to swim stress that creates both physical and psychological stress. Acute stress reduced those seizures induced by a gamma-aminobutyric acid-A antagonist but exacerbated those induced by electroconvulsive shock, and did not affect various other models. The anticonvulsant effects are most likely mediated via neuroactive steroids acting at gamma-aminobutyric acid-A receptors, which change subunit expression following acute stress, increasing inhibition and decreasing seizure activity.19 Conversely, chronic stress either in early life or over time in adulthood, consistently influences brain excitability and seizure susceptibility at all stages in the development of epilepsy (figure 4) in a broad range of models; it acts as an initial insult, decreasing the latent period and increasing seizure frequency in established epilepsy.19–21 There are several mechanisms that may underpin this association (figure 5) (see below).
Early life stress
Preclinical studies clearly show that early life stress predisposes the brain to provoked seizures and to later epilepsy. For example, if pregnant female rodents are stressed (eg, by restraint, bright lights or injections), it is easier to induce seizures several weeks later in the pups in response to kindling, audiogenic stimuli and chemoconvulsants. Similarly, postnatal stress induced by separation of pups from their mother renders them more susceptible to later seizures, with kindling and chemoconvulsants.20 Possible mechanisms underlying this link are:
Changes in connectivity, function and structure of neuronal circuitry that act as a first hit and lessen the insult needed in later life to trigger epileptogenesis, for example, decreased gamma-aminobutyric acid inhibition leads to hyperexcitable neurotransmission; increased corticotropin-releasing hormone leads to changes in neural structure; there may also be damage from inflammation, decreased brain-derived neurotrophic factor and delayed white matter development.22
Impaired stress response due to a hyperactive HPA axis with abnormal negative feedback.20 The hippocampus normally shuts off the HPA axis but excessive glucocorticoid exposure can cause the hippocampus to shrink, disinhibiting the HPA axis and resulting in a vicious cycle of unopposed allostatic load. Rodent models with acutely provoked (kindled) seizures have HPA axis dysfunction. There is also evidence of such dysfunction in humans with status epilepticus, although antiepileptic drug treatment may be a confounder. Functional imaging studies in humans are also now starting to examine the link between cortical and physiological responses to stress and the relationship between seizure control in epilepsy and HPA axis function, as well as with functional MRI signal reactivity.23
Chronic stress in adulthood
Animal models consistently show increases in seizure risk21 during and following chronic stress. Chronic stress may render the brain more susceptible to unprovoked seizures, or at risk of more severe or frequent seizures in the event of an additional trigger.
There are two possible mechanisms:
Decreasing the latent period of epileptogenesis. Many of the processes underlying epileptogenesis are potentiated by chronic stress including inflammation,22 abnormal neurogenesis and synaptic plasticity.24
Increasing brain excitability via increased excitatory glutamatergic transmission, intrinsic neuronal excitability and/or shift in gamma-aminobutyric acid-ergic inhibition21 ,25 likely mediated via neurosteroids.19 This phenomenon would increase the baseline excitability and therefore would keep the brain activity closer to the seizure threshold.
If stress is important, can we treat it?
UK guidance emphasises the need for enhancing quality of life through ensuring people feel supported to manage their condition, with improved functional abilities; the guidance specifically includes stress as a topic on which to provide patients with information.26 Guidelines also recommend that psychologists and psychiatrists are part of a multidisciplinary epilepsy team at least at tertiary centres, but there is little clarity beyond that, and of note there are no psychiatric/psychological outcomes in proposed US27 or UK28 quality standards for epilepsy care.
In this context, it is important to recognise that the management of stress can be distinct from that of comorbid anxiety, depression or other psychiatric disorders, but that there are also complex multidirectional relationships between stress and these disorders. It is well known that stressful events can precipitate depression or anxiety, but attempting to dissect out the causal relationships between seizures, stress, anxiety, depression and other confounding factors such as sleep deprivation may give conflicting results. A longitudinal cohort study of over 400 community recruited patients using a range of validated scales applied retrospectively at two time points looked in detail at the interaction of anxiety, depression and perceived stress on seizure frequency and concluded that depression underpinned the relationship.29 In contrast, two cross-sectional studies, each with over 200 patients attending an epilepsy centre,7 ,30 found 50% and 80% of patients reported stress as a seizure trigger, and concluded that anxiety was the most important factor. Others have shown a high correlation between stress, sleep deprivation and fatigue.31 Yet others,32 including a more methodologically robust multivariate analysis of prospective paper diaries, collected over up to 1 year15 albeit in smaller numbers (n=71), concluded that sleep deprivation, anxiety and stress all independently predicted seizure occurrence. Clearly, clinicians should address and actively treat any comorbid psychiatric disorder, since this is so important to seizure control and quality of life.1 We now discuss the evidence base for treating stress itself.
Non-pharmacological treatments
Stress is a fact of life, and often the result of non-modifiable factors, so the treatment of stress primarily relies on psychology based treatments to restore a sense of control over emotions, and downregulation of the biological response mechanisms. Stress is universal but the source, quality of experience and acceptability of various treatments varies with cultural influences. Western culture has a mechanistic explanation of stress, reflected in relaxation, biofeedback and cognitive techniques. Asian cultures have a somatic explanation of stress giving a somatic focus with treatments such as yoga and mindfulness. There is a broad range of techniques studied in epilepsy, many with positive results (box 2).
Psychological and mind-body treatments of potential benefit for stress in epilepsy
Cognitive behavioural therapy: A range of techniques that involves teaching and reinforcing self-monitoring of thoughts and emotions, identifying the automatic thoughts accompanying distressing emotions, learning about different types of cognitive distortions, and working towards reducing intrusive thoughts, and/or separating them from the anxiety that they produce.
Biofeedback is a process that enables an individual to learn how to change measured physiological activity such as EEG or galvanic skin responses in order to reduce seizure frequency. Patients receive rapid and accurate feedback of physiological data in conjunction with identifying changes in thinking, emotions and behaviour.
Mindfulness is simply paying attention to moment-by-moment experience in an open and non-judgemental way, practising awareness of sensory and mental experiences as they happen.
Yoga is an ancient Indian, non-religious, mind-body approach with components centred upon meditation, mindfulness, breathing, and activity or postures.
Acceptance and commitment therapy is an empirically based psychological intervention that uses acceptance and mindfulness strategies mixed in different ways with commitment and behaviour-change strategies, to increase psychological flexibility.
There are numerous challenges to undertaking intervention studies in this area including in relation to blinding, sham controls and powering when the potential effect size is unknown; this is reviewed in detail elsewhere33 and there is now accumulating evidence of potential benefit. For example, a recent assessor-blinded randomised control trial (n=60) showed that 4 weeks of biweekly mindfulness compared with social support gave clinically important improvements in quality of life, and reduced seizure frequency in drug-resistant epilepsy.34 A Cochrane review of yoga35 in epilepsy found it to be better than no intervention and better than interventions other than yoga (postural exercises mimicking yoga). Acceptance and commitment therapy was as effective as yoga. However, all the included studies are open and small (18–32 participants) and so we cannot draw reliable conclusions. A meta-analysis of EEG-biofeedback studies in drug-refractory patients,36 extracting data from 10 of the 63 such studies undertaken at that time, found that 74% of patients (n=87) responded, with significant (61%) reduction in median seizure frequency overall. More recently, a small study (n=11) using skin conductance biofeedback gave promising initial results, with an almost 50% reduction in seizure frequency over 12 sessions.37 Although a 2008 systematic review38 of psychological treatments in epilepsy found no evidence of benefit, this largely reflected deficiencies in methods and the small number of studies. More recently, cognitive-behavioural therapy approaches have led to significantly (>50%) reduced seizure frequency (37%) in uncontrolled retrospective studies (n=60).39 There was also benefit from cognitive-behavioural therapy in a small (n=37) open randomised control trial of compared to relaxation.40 However, another small (n=27) single blinded randomised control trial of cognitive behavioural therapy versus counselling and wait list found no differences in seizure frequency.41 In all of these studies, the interventions are often time-intensive (one to two sessions/week over 4–8 weeks, sometimes with daily home programmes), with subsequent follow-up for only a few weeks. There is one long-term follow-up study, 10 years after a biofeedback intervention, which showed sustained benefit in the responders, albeit with only 19 of the original 41 participating in the later study.42
There have been no studies of hypnosis or stress inoculation therapy in epilepsy, although they are widely used in other contexts. Similarly, there is limited evidence in epilepsy for eye movement desensitisation and reprocessing—a psychotherapy that emphasises disturbing memories as the cause of psychopathology and often used in the treatment of post-traumatic stress disorder. There are only case reports, but notably one of reactivated seizures might suggest the need for caution.43
Pharmacological treatments
There is no pharmacological treatment of stress alone, though psychoactive agents can help stress-related psychiatric conditions. The Diagnostic and Statistical Manual of Mental Disorders, 5th edition, identifies the specific category ‘Trauma and stressor-related disorders’; major diagnostic categories include acute stress disorder, post-traumatic stress disorder, adjustment disorder and reactive attachment disorder. The treatment of these conditions is not covered in this review. However, although several antiepileptic drugs have been tested in post-traumatic stress disorder and related disorders, most such studies are uncontrolled, and so we have only a low level of evidence.44 Taking the broad chapter of anxiety disorders in general, pregabalin is the only antiepileptic with good clinical evidence and a current licence for treating generalised anxiety disorder.44
Implications for clinical practice
So what does this mean for clinical practice? Whatever the mechanisms and uncertainties about strength of association, stress clearly matters to patients, can impact on mood and quality of life, and there is biological plausibility in terms of potential influence on seizure control. At the very least we should actively acknowledge this. Several validated screening tools can detect relevant symptoms and can be easily incorporated into routine practice—for example, short questionnaires completed in the outpatient waiting room before appointments;1 this sort of proactive attention to symptoms than seizures can benefit patients.45 In our experience patients often readily volunteer having stress, yet it may be disregarded in busy clinics; this is perhaps influenced by the clinician's helplessness in the face of weak evidence, and lack of resources for psychotherapeutic or behavioural support. However, there is a wealth of self-help resources, and many people anyway seek complementary therapies privately. We suggest that, as a starting point, any service should at least compile a list of local resources and recommend reading/self-help websites on stress management to people for whom stress is a potential issue. In this context, particularly if the patient will be expending time or money on any intervention, we recommend offering support in defining outcome measures and time frames for that individual as a ‘trial of treatment’, just as is routine practice when changing antiepileptic drugs. However, patients and providers may have very different perspectives, and recognising individual goals and patient inputs when negotiating any programme requiring self-management is fundamentally important to success.46 On the other hand, chronic stress is well known to be bad for health and quality of life, and can be improved by stress reduction strategies. Thus, perhaps we should stop trying to prove benefits specific to epilepsy, and start by promoting stress reduction strategies in more general terms as part of holistic care.
What next for stress and epilepsy?
In recent years there has been an increasing emphasis on individualised treatment for epilepsy, recognising that epilepsies are diverse with multiple genetic and structural/metabolic causes. Brain connectivity and response to stress probably varies in different epilepsy types. For example, stress may be of particular importance in temporal lobe epilepsy, and the decreased functional connectivity in this group47 might reflect underlying network abnormalities resulting in epilepsy that also affects their response to stress. In support, one recent case control study (n=23 each group) in left-sided temporal lobe epilepsy showed a significant relationship between seizure control and both the HPA axis and functional MRI signal reactivity to acute psychosocial stress.48 The very nature of psychobehavioural treatments requires some personalisation to reflect individual stressors, appraisal and response patterns, and coping resources that will also vary considerably from person to person. Designing and delivering good clinical studies in the face of these challenges is a daunting prospect, even without the issues of blinding, cost-effectiveness and cost-utility. Potential new pharmacological treatments seem a more attractive prospect on this background. Neurosteroids, specifically ganaxolone and allopregnanolone—all potent allosteric modulators of gamma-aminobutyric acid-A receptors—are currently under investigation as antiepileptic drugs,19 with parallel studies underway in post-traumatic stress disorder and severe postpartum depression. Vagus nerve stimulation, an established treatment for epilepsy with known effects on the autonomic nervous system, probably also influences stress. Non-invasive transcutaneous vagus nerve stimulation in healthy people during functional MRI causes widespread decreased activation to high threshold stimulation in emotional/stress-related regions of the limbic system, associated with significantly improved well-being afterwards compared with sham stimulation.49 Perhaps we have been treating stress with vagus nerve stimulation for some time without realising it.
Key points
Stress is the most commonly reported trigger for seizures in people with epilepsy.
Clinical and especially preclinical evidence supports an association between stress and epilepsy, though clinical studies are often confounded by factors such as sleep deprivation, alcohol and missing antiepileptic medication.
Acute stress is anticonvulsant, but chronic stress in early life or adulthood promotes epileptogenesis and seizures. An imbalance of the autonomic nervous system and hypothalamic-pituitary-adrenal axis dysregulation probably causes dysfunction in limbic gamma-aminobutyric acid and glutamate pathways.
Subgroups of people with epilepsy appear sensitive to stress through a combination of genetics, early life stress exposure and syndrome-specific brain connectivity.
A range of psychological and mind-body behavioural treatments for stress have a potential role in improving seizure control and quality of life in epilepsy, though we lack large well controlled studies.
References
Footnotes
Twitter Follow Hannah Cock at @drhannahrc
Contributors MM and HRC conceived and outlined the idea; CMG undertook the primary literature search and wrote the first draft of the manuscript. All authors contributed to critical review and the final manuscript.
Competing interests MM reports personal fees from UCB, Eisai, Pfizer, Elsevier, personal fees and non-financial support from Bial, non-financial support from Special Products, outside the submitted work. HRC reports personal fees from UCB Pharma, Epilepsy Nurse Specialist Association UK, European Medicines Agency, Special Products, Eisai Europe, non-financial support from GSK, personal fees and non-financial support from Lupin Pharmaceuticals, grants from NINDS, NIH USA, outside the submitted work.
Authors’ note This paper was stimulated by a book chapter by two of the authors, Galtrey and Cock6 but has been completely rewritten with a practical and clinical emphasis for the readership of Practical Neurology.
Provenance and peer review Commissioned; externally peer reviewed. This paper was reviewed by Yvonne Hart, Newcastle-upon-Tyne, UK.
Linked Articles
- Editors' commentary
Other content recommended for you
- Stress and functional neurological disorders: mechanistic insights
- Psychological stress in IBD: new insights into pathogenic and therapeutic implications
- The neurobiology of stress and gastrointestinal disease
- Temporal lobe epilepsy surgery: different surgical strategies after a non-invasive diagnostic protocol
- EEG in the diagnosis, classification, and management of patients with epilepsy
- Epilepsy surgery
- Increased antigen and bacterial uptake in follicle associated epithelium induced by chronic psychological stress in rats
- Advances in epilepsy surgery
- Comparing neurostimulation technologies in refractory focal-onset epilepsy
- Sudden unexpected death in epilepsy: basic mechanisms and clinical implications for prevention