Neuroimaging of Hysteria

Gereon R. Fink

Peter W. Halligan

John C. Marshall

ABSTRACT

The potential of imaging the functional neuroanatomy of hysteria was first recognized in the closing decades of the 19th century. Charcot argued that patients without apparent structural injury to the brain who nonetheless manifested clinical symptoms that mimicked organic paralysis, sensory loss, or aphasia should have “dynamic” or “functional” lesions in the regions in which structural damage could give rise to the same symptomatology. The advent of modern functional neuroimaging methods such as single photon emission computed tomography (SPECT), positron emission tomography (PET), or functional magnetic resonance imaging (fMRI) have now made it possible to detect in vivo regionally specific changes in cerebral blood flow and task-related changes in neural activity associated with sensorimotor or cognitive processes. Charcot’s hypothesis can thus be tested. Here we review the recent, but still comparatively few, attempts to elucidate the functional neuroanatomy underlying hysterical symptoms. We suggest that the data available are difficult to reconcile with a single neural mechanism, but do reliably implicate anterior cingulate cortex and dorsolateral prefrontal cortex in many cases of sensorimotor hysteria reported to date. The specific contribution of these areas to the neuropsychology underlying hysteria, however, remains to be elucidated. We conclude that hysteria as clinically diagnosed is a protean disorder. Furthermore, future studies of the neural mechanisms associated with hysteria should also take into account the high probability that functional specialization may not be a fixed property of brain regions as previously supposed, but may rather depend on the neural context. Accordingly, functional neuroimaging studies of hysteria should analyze the functional interaction (“dynamics”) between the brain regions involved in normal and defective task performance.

INTRODUCTION

The two international systems of diagnostic classification for psychiatric disorders, namely, the International Classification of Diseases (ICD-10) and the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), summarize a variety of syndromes as “conversion disorders” (DSM-IV) or “dissociative disorders” (ICD-10). These disorders are unified by the assumption that they have in common the presence of positive physical symptoms (e.g., tremor, tonic-clonic convulsion), or negative symptoms (e.g., organic paralysis, sensory loss, or aphasia) without evidence of any discernable “organic” correlate. Since these diagnoses refer to a “psychological” conflict preceding the onset of physical symptoms, conversion or dissociative disorders are considered to be related to the older concept of “hysteria” (1). Although earlier references to hysteria (Hippocrates, 460?-377 BCE) restricted the diagnostic label to a gynecologic context, the etiology of this protean disorder has “migrated over time from the uterine to the cerebral to the psychodynamic” (2) and remains to be fully elucidated. Unsurprisingly, the medical concept of hysteria (conversion disorder, dissociative disorders, etc.) has always been controversial (3), and despite its recognition in current psychiatric taxonomies, many physicians still regard
hysterical disorders as either feigned or as a failure to find the responsible organic cause for the patient’s symptoms.

The concept of hysteria evokes considerable interest because, as a putative disorder of willed action or intention (4), the underlying mechanisms are supposed to be of a “psychological” nature and, if not feigned, the product of unconscious processes. A major problem with the current concept of hysteria is that the diagnosis can only be put forward if the physician’s best guess, or, in fact, belief, is that the patient is not consciously/intentionally producing or feigning dysfunction (5). Thus, a key issue when distinguishing conversion disorder from malingering is the clinician’s ability to make inferences about the patient’s intentions when much of the evidence relies on the patient’s subjective report. Accordingly, the issue of an impairment to the volitional system becomes the focus of interest once any explanatory organic disorder has been excluded.

This chapter summarizes the findings obtained using functional imaging [single photon emission computed tomography (SPECT), positron emission tomography (PET), or functional magnetic resonance imaging (fMRI)] in cases of hysteria and related disorders. Unfortunately, functional neuroimaging studies of hysteria are still very rare and the few data available are difficult to reconcile with a single neural mechanism. The results promise, however, new insights into the neural mechanisms associated with hysteria, as they allow for the testing of hypotheses about the pathophysiology underlying hysterical conditions.

HYSTERICAL PARESTHESIA OR ANESTHESIA

One of the first studies to investigate conversion disorders using functional neuroimaging was concerned with altered cerebral blood flow in a single case of hysterical paresthesia (6). The patient was a nurse who had no history of psychiatric or neurologic treatment, but developed symptoms of depressed mood and a panic disorder when she was in a state of extreme stress due to her current marital and domestic situation. Eleven months after the first attack, she was hospitalized for left-sided paresis and paresthesia associated with mild apathic symptoms. Her somatosensory-evoked potentials (SEP) and motor-evoked potentials were normal, and her symptoms disappeared within a week. In contrast to the normal electrophysiologic examinations of her long sensory and motor tracts, SPECT during electrical stimulation of the left median nerve at the time of paresthesia showed associated alterations in cerebral blood flow. Prior to recovery from paresthesia there was increased perfusion in the right frontal lobe, but hypoperfusion in the right parietal region (compared to the equivalent left hemisphere regions).

After recovery, the perfusion in the right parietal region was greater than in the left parietal region during left median nerve stimulation, as one would expect. These results suggested that psychogenic paresthesia may have arisen from the simultaneous activation of frontal inhibitory areas and the associated inhibition of somatosensory cortex. Tiihonen et al. (6) hypothesized that “distressing psychological events may alter the neurophysiology of the human brain in a specific way and trigger symptoms such as … paresthesia through activating or inhibiting critical areas of the brain.” More importantly, this first imaging study of hysteria suggested that hysteria was indeed amenable to investigation by functional imaging.

Using fMRI, Mailis-Gagnon et al. (7) studied altered central somatosensory processing in four chronic pain patients with hysterical anesthesia. Patients with chronic pain frequently present with nondermatomal somatosensory deficits (NDSD) to various cutaneous sensory modalities (touch, pinprick, cold) ranging from mild sensory loss to complete anesthesia. These changes of somatosensory processing are often considered to be of psychogenic origin as they typically occur in the absence of substantial structural pathology. Interestingly, NDSD may occur with variable motor abnormalities including complete motor paralysis (7). Mailis-Gagnon (7) tested their hypothesis that central factors may underlie NDSD by using fMRI to measure brush and noxious stimulation-evoked brain responses. They observed altered somatosensory-evoked responses in forebrain areas: Unperceived (unreported) stimuli failed to activate areas that were activated with perceived touch and pain, including the thalamus, the posterior part of the anterior cingulate cortex (ACC), and Brodmann area 44/45. Rather, these unperceived (i.e., unreported) stimuli were associated with deactivations in primary and secondary somatosensory areas (SI, SII), posterior parietal cortex, prefrontal cortex, and rostral parts of the ACC. Application of the same stimuli to the unaffected side of the four patients resulted in the normal patterns of activations and awareness commonly associated with somatosensory stimulation. In contrast to these fMRI results, tests using cortical evoked potentials found no changes in function (8).

The results of Tiihonen et al. (6) and Mailis-Gagnon et al. (7) suggest that functional neuroimaging is a sensitive tool for examining afferent function in cases of hysterical anaesthesia, as both show evidence of altered cerebral activity in the neural networks known to support sensory information processing. In both studies, increases as well as decreases in neural activations in response to somatosensory stimulation were observed. Reduced processing of sensory information was associated with attenuated neural activity in parietal areas in both studies (comprising SI, SII, and posterior parietal cortex), while increased neural activity was observed in frontal areas. With respect to this latter activity, however, the results of the two studies differed substantially. While the data of Tiihonen et al. implicated
prefrontal areas and supervisory inhibitory processes (6), Mailis-Gagnon et al. (7) observed increased neural activity in a region of the anterior cingulate cortex that may be involved in aspects of divided attention, response inhibition, and possibly emotion (9).

The role of striatothalamocortical circuits involved in sensorimotor processing was stressed in another SPECT study of the functional correlates of unilateral hysterical sensorimotor loss in seven patients (10). Passive vibratory stimulation was given to each hand when the deficit was present and 2 to 4 months later when the patients had recovered. SPECT blood flow measurements consistently detected a decrease in rCBF in the thalamus and basal ganglia (contralateral to the deficit) which resolved after recovery. The data thus converge with the study by Tiihonen et al. (6) which also reported, using a similar paradigm, that abnormal blood flow patterns may be reversible upon recovery from hysterical sensory loss. The study by Vuilleumier et al. (10), however, extended these findings by showing that reduced activation of the contralateral caudate predicted poor recovery at follow-up (10). The authors speculated that the basal ganglia, and in particular the caudate nucleus, “might be well suited to modulate motor processes based on emotional and situational cues from the limbic system” (10).

HYSTERICAL PARALYSIS

Marshall and co-workers reported a woman with left-sided paralysis (and without somatosensory loss) in whom no organic disease or structural lesion that could have explained the paralysis was found after intensive investigation (11). By contrast, psychological trauma was associated with the onset and recurrent exacerbation of her hemiparalysis. Using PET, no activation of primary motor cortex was observed when the patient attempted to move her affected leg (as might be expected from the lack of any observed actual movement), though there was increased neural activity in right orbitofrontal cortex and anterior cingulate cortex (11). In contrast, preparing to move or moving her good leg, and also preparing to move her paralyzed leg, activated motor and/or premotor areas previously described with movement preparation and execution. Marshall et al. (11) suggested that the right orbitofrontal and right anterior cingulate cortex inhibited prefrontal (willed) brain areas with the resultant effect on right primary motor cortex [when the patient tried (i.e., willed) to move her affected left leg]. While orbitofrontal cortex had not been described in previous studies of motor imagery, or motor execution (or hysteria), activation of the anterior cingulate cortex (11) has been conjectured to reflect a “meeting place for interactions between cognitive and motivational processes, particularly related to the generation of motor output (12).”

Spence et al. used PET to examine the neural correlates of hysterical motor loss versus a feigned disorder of movement (13). They postulated that the pathophysiology of “genuine” hysterical motor symptoms would differ from that involved in feigning. Two men with hysterical motor symptoms that substantially affected their left arms were investigated while they performed joystick movements (moving the left hand or the right hand) and at rest. In addition, two healthy individuals were instructed to “feign” difficulty in moving their left arm. For control, six healthy individuals were asked to perform the joystick movements normally. Comparing brain activations during movement of the left hand (relative to rest) in patients with hysteria versus controls and feigners showed that the two patients exhibited relative hypoactivation of the left dorsolateral prefrontal cortex (DLPFC). In contrast, the two feigners exhibited hypofunction of the right anterior prefrontal cortex.

Subsequently, Spence et al. studied a right-handed man with hysterical weakness of the right upper limb and two healthy volunteers who feigned abnormality of their right upper limb (13). A combined data analysis showed that the observed left prefrontal hypoactivation was common to all three patients with hysteria when they moved the affected limb, irrespective of symptom-lateralization, while the feigners were characterized by right prefrontal hypofunction, again irrespective of the side for which they feigned malfunction. Spence et al. suggested that “taken together” their data support the hypothesis that hysteria involves the left DLPFC and differs from feigning with respect to its neural mechanisms (13).

ASTASIA-ABASIA

Five patients with psychogenic astasia-abasia were investigated by Yazící and Kostakoglu using SPECT (14). Somatosensory evoked potentials (SEPs) of the posterior tibial nerve were abnormal in two of the five patients. In these latter patients, SEPs returned to normal after 6 months (despite incomplete clinical remission). Two patients showed decreased perfusion of the left parietal lobe. Four patients showed temporal hypoperfusion (bilateral n = 1, unilateral left n = 3). Interestingly, although clinical symptoms were bilateral in the patient group, the observed effects of hypoperfusion were mostly unilateral and in the dominant (left) hemisphere. Only one patient showed a right-sided temporal hypoperfusion in addition to a left-sided temporal defect. The ongoing right-sided perfusion defect was thought to underlie the symptom of paresis in his left leg, which persisted during imaging.

HYPNOTIC PARALYSIS

A relationship between hysteria and hypnosis has been postulated by many in the history of psychiatry and psychology,
and hypnotic induction of motor paralysis has been used since the 19th century to mimic hysterical symptoms (15). Hypnotic phenomena and conversion symptoms, particularly in the acute stage, share many clinical, cognitive, and neurophysiologic features. Accordingly, hypnotic states are often considered as a kind of “controlled state of hysteria” (16). Unsurprisingly, experiments on hypnosis have thus developed into an experimental analogue for studying hysterical symptoms. The concept of a relationship between hysteria and hypnosis is based on the assumption that both hysteria and hypnosis are abnormal states of behavior and experience influenced by “ideas,” whether these ideas emerge from the outside (as in hypnosis) or from the inside (as in hysteria). The view that conversion symptoms can be usefully thought of as an “autosuggestive disorder” (17) gained support from a functional imaging study of a single case of hypnotic paralysis (18) where similar brain areas were activated during hypnotic paralysis [right orbitofrontal cortex, right anterior cingulate cortex (ACC)] as in the study by Marshall et al. of a patient with longstanding hysterical paralysis (see above). Halligan et al. (18) suggested that their findings were consistent with the hypothesis that hysterical and hypnotic paralysis might share common neural mechanisms involving the right (in both cases, contralateral) prefrontal region.

Only gold members can continue reading. Log In or Register to continue