Globus Sensation: Pharyngoesophageal Function, Psychometric and Psychiatric Findings, and Follow-up in 88 Patients

Background  The globus sensation has been widely regarded as psychogenic, but organic disorders were found to be etiologically significant.

Objective  To investigate the structural, functional, psychological, and psychiatric factors possibly eliciting the globus sensation and influencing its course.

Methods  Eighty-eight patients, 67 women and 21 men (aged 22-71 years), referred to 2 tertiary care centers underwent history taking, otolaryngological examination, pharyngoesophageal videofluoroscopy and manometry, psychosocial evaluation, psychometric tests, psychiatric interview, and when indicated, esophagogastroduodenoscopy, esophageal bolus transport, gastroesophageal reflux, and gastric emptying studies. According to revealed disorders, therapy was initiated, and the outcome was studied.

Results  Only 15 patients had normal pharyngoesophageal function; of these 15, 6 had chronic tonsillitis or pharyngitis, 3 had thyroid adenomata, 4 had cervical spondylosis, and 1 each had dry oropharyngeal mucosa and chronic bronchitis. Of the other 73 patients, 2 had pharyngeal dysfunction, 24 had achalasia, 1 had diffuse esophageal spasms, 3 had "nutcracker esophagus," 30 had nonspecific esophageal motor disorders, and 13 had gastroesophageal reflux. Psychometry revealed no more anxiety and depression than in general medical outpatients. Of 58 patients interviewed, 37 met criteria for psychiatric disorders. Psychometric scores and psychiatric characteristics were unrelated to the sensation's course. Therapy was recommended, but only 26 patients were treated accordingly; 22 received nonspecific treatment. Follow-up 3 to 59 months later revealed that the sensation had vanished in 13 patients who had received specific treatment, 5 who had received nonspecific treatment, and 6 who had received no treatment; it was alleviated in 10 who had received specific treatment, 13 who had received nonspecific treatment, and 9 who had received no treatment; and it was unchanged in 3 who had received specific treatment, 5 who had received nonspecific treatment, and 23 patients who had received no treatment.

Conclusions  Pharyngoesophageal disorders may be sensed only vaguely, inducing the globus sensation. Psychological and psychiatric characteristics could be relevant to the discomfort experienced but are unlikely to be etiologically significant.

THE GLOBUS sensation, a feeling of a lump in the throat associated with dry swallowing and disappearing while eating or drinking, has been widely regarded as psychogenic. It was thought to be a "materialization of a repressed idea,"¹ "one of the most widely recognized indications of nervous illness,"² "a physical manifestation of suppressed emotion,"³ a "discriminant symptom of somatization disorder,"⁴ or a "single-symptom model for the study of conversion disorders."⁵ The last proposition was based on no more than higher neuroticism and lower extraversion scores on the Eysenck Personality Inventory⁶ in 37 women with globus sensation compared with 24 normal control subjects. The same group of authors later reported to have found significantly higher neuroticism scores on the Eysenck Personality Inventory and significantly higher ratings of anxiety, obsessionality, depression, phobia, and somatic concern in 28 women with the sensation than in 33 without it.⁷ Other authors found patients with the sensation to have no more anxiety and hysterical traits than healthy subjects^8-10 and outpatients with otolaryngological disorders.^11,12

A series of studies indicated that organic factors underlay the symptom: a collapse in the vertical dimension of the masticating apparatus,¹³ temporomandibular joint dysfunction,¹⁴ carcinoma on the base of the tongue,¹⁵ hyperplastic lingual^16,17 and pharyngeal tonsils,¹⁸ cervical lymph nodes,¹⁹ postcricoidal webs,¹⁷ goiter,¹⁷ thyroid tumors,¹⁸ cervical osteophytes^17,19 or arthrosis,²⁰ and an increased tension of the pharyngeal muscles.¹⁶ Paraesophageal masses,¹⁸ carcinoma,²¹ gastroduodenal lesions,¹⁷ and hiatal hernias^17,22,23 also have been thought to underlie the sensation. In the past decades, a disordered contractile activity of the upper esophageal sphincter (UES),^{11,12,17,24-26} esophageal body,^27-30 or the lower esophageal sphincter (LES)^17,20,28,29 and gastroesophageal reflux (GER)^{11,12,18,23,28,31-39} were suggested to elicit the globus sensation.

The present study was carried out to determine the structural or functional abnormalities in the pharynx, larynx, esophagus, stomach, neck, and vertebral column as well as psychological traits and psychiatric disorders possibly related to the globus sensation and its course after diagnostic and therapeutic approaches.

Of 168 patients with globus sensation referred within 312 years to the Psychosomatic Clinic and the Psychophysiology Unit of the University of Vienna's School of Medicine, Austria, for further investigation, 88 agreed to participate in the study. Sixty-seven were women, and 21 were men; their ages ranged from 22 to 71 years (median, 43 years). The sensation had been felt for 1 month to 36 years (median, 2 years). Before its initiation, the study had been approved by the institutional Committee on Studies Involving Human Beings.

At the initial evaluation, a detailed history accounting also for psychosocial conditions was obtained, and blood specimens were taken for hematologic and biochemical screening. To assess feelings of depression and anxiety, the patients were asked to complete the Beck Depression Inventory (BDI)⁴⁰ and the Spielberger State-Trait Anxiety Inventory (STAI).⁴¹ Appointments were made for an otolaryngological examination, videofluoroscopic and manometric evaluations of pharyngoesophageal function, and a psychiatric interview. Endoscopic examinations; quantitations of esophageal bolus transport, gastric emptying, and GER; and other investigations were carried out as deemed appropriate by findings or symptoms.

Before manometry, the patients underwent an interview and answered a questionnaire directed at symptoms of dysphagia, heartburn, retrosternal and epigastric pain, postprandial fullness, regurgitation, vomiting, bloating, and bowel habits. To determine personality traits, the patients completed the Minnesota Multiphasic Personality Inventory (MMPI) in its 221-item form.⁴² Individual raw scores were converted to T-scale scores.

To examine whether criteria for any of the conditions listed in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R)⁴³ were met, the patients were asked to undergo the Structured Clinical Interview for DSM-III-R disorders.⁴⁴

The otolaryngological examination comprised palpation and sonography of the neck and an inspection of the pharynx and larynx using a flexible endoscope.

The videofluoroscopic procedure followed was as described earlier.^45,46 In the oral phase of swallowing, attention was paid to bolus control and transport. In the pharyngeal phase, elevation of the hyoid and larynx, laryngeal closure, pharyngeal contractile activity, and the UES opening were evaluated. The presence of esophageal diverticula or webs and spinal osteophytes was noted. Achalasia was diagnosed when there were slow esophageal bolus transport, nonpropulsive contractions, and an incomplete or absent LES opening with or without esophageal dilatation. Diffuse esophageal spasms were diagnosed when there were intermittent, repetitive, and nonpropulsive contractions leading to a corkscrew appearance of the esophagus.

For the manometric investigation, patients were placed in the supine position with their head on a pillow. In 36 patients, a probe with 3 strain-gauge pressure sensors spaced at 5-cm intervals and oriented radially 120° apart (Konigsberg Instruments Inc, Pasadena, Calif) was used, the output of which was recorded (R-411 Dynograph, Sensormedics Inc, Anaheim, Calif) and digitized on-line. In 24 patients, a probe with 5 strain-gauge pressure sensors spaced at 3-cm intervals (Konigsberg Instruments Inc) was used. The most distal sensor was circular, and the remaining 4 were oriented 90° apart; their output was recorded using a computer-based system (Synectics Medical AB, Stockholm, Sweden). The initiation of swallowing was monitored by a submental surface electromyogram. The resting pressure and relaxation of the LES on swallowing 5 mL of water at body temperature were determined, with the mean end-expiratory fundic pressure as zero reference, drawing the probe at 5-mm steps and 30-second intervals. To evaluate the contractile responses of the esophageal body to 5-mL swallows, the probe was withdrawn at 1-cm steps and 30-second intervals. The UES resting pressure, the onset and duration of relaxation in response to 5-mL swallows, its coordination with pharyngeal contractions, the residual pressure at maximal relaxation, and the amplitude and duration of the postrelaxation contraction were determined by drawing the probe at 5-mm steps and 30-second intervals.

In 28 patients, manometry was performed using a pneumohydraulic system (Mui Scientific Enterprises Inc, Mississauga, Ontario) and a multilumen catheter fitted with a sensor (Dentsleeve, Pty Ltd, Bowden, South Australia) for measuring with the LES and side-hole openings 0, 3, 6, and 9 cm orad the sleeve's proximal border (Arndorfer Medical Specialties Inc, Greendale, Wis), as described earlier.⁴⁷

The criteria for achalasia were an LES relaxation on swallowing of less than 75% from a resting pressure of 18 mm Hg and predominantly synchronous swallow-initiated contractions in at least the lower two thirds of the esophageal body. Diffuse esophageal spasms were diagnosed when there were, in conjunction with retrosternal pain, contraction amplitudes of more than 190 mm Hg and durations of longer than 7.5 seconds in association with either or both synchronous and repetitive contractions following 30% or more of swallows in the lower esophagus. "Nutcracker esophagus" was diagnosed when the mean amplitude of contractions exceeded 190 mm Hg, but propagation velocity and LES resting pressure and relaxation were normal. A nonspecific esophageal motor disorder was diagnosed in the presence of abnormalities not fulfilling the criteria for the above disorders, ie, more than 20% of nontransmitted, triple-peaked, retrograde, or long-duration (>6 seconds) contractions; more than 10% of synchronous contractions; or an LES relaxation of less than 75% from a resting pressure of less than 18 mm Hg.

Esophageal bolus transport was recorded scintigraphically.⁴⁷ Achalasia was diagnosed when 95% of the bolus was transported through the esophagus in 20 seconds or more. Transit times of 10 to 19 seconds were considered to indicate disordered motility and of less than 10 seconds were considered normal.

Gastroesophageal reflux was recorded by ambulatory 24-hour pH monitoring. Antimony or glass electrodes and fitting recorders (Synectics; Ingold, Messtechnik AG, Urdorf, Switzerland) were used. Percentages of time with a pH of less than 4 for more than 6.3% in the upright position, more than 1.2% in the supine position, and more than 4.2% of the total recording time were considered abnormal.⁴⁸

The emptying of a semisolid standard meal was recorded scintigraphically.⁴⁹ The residual radioactivity in the stomach 50 minutes postprandially was taken as an overall measure of emptying. The recorded data were compared with those of 55 healthy, symptom-free subjects (aged 19-64 years; median, 25 years) (mean [±SD] residual radioactivity, 54.0% [±13.3%]).

Patients were asked to undergo esophagogastroduodenoscopy if there was an indication for pathological GER or a structural abnormality.

Patients were invited to return every 3 months for questionnaire-based interviews and, if indicated, further investigations. If patients were unable to return, interviews were carried out by telephone.

The hematologic, biochemical, and urine analyses showed no gross abnormalities in any patient. Otolaryngological, fluoroscopic, and manometric investigations were carried out in every patient; videofluoroscopy in 73 patients; esophagogastroduodenoscopy in 16 patients; and studies of bolus transport in 36 patients, 24-hour GER activity in 21 patients, and gastric emptying in 6 patients.

Twenty-seven patients described the globus sensation as more or less persistent, and 61 described it as intermittent. Twenty-two patients had felt it for less than 6 months, 13 for 7 months to 1 year, 40 for more than 1 to 5 years, and 13 for more than 5 years. Twenty-five patients said that it arose or became more intense during states of fear, tension, anger, or mental stress; and 7 said that it occurred with general strains. Twelve patients said that the sensation arose during or after eating, after lying down (10 patients), when bowing down (5 patients), with regurgitation (1 patient), with dry throat or in smoky air (2 patients), with cervical pain (1 patient), with reading or watching television (6 patients), when walking downhill (1 patient), with mucus in the throat (2 patients), and after taking an antidepressant medication (1 patient). The sensation was felt less intense during states of relaxation (19 patients), during work or distraction (13 patients), during eating or drinking (7 patients), when upright or supine (6 and 3 patients, respectively), and when free of rhinitis (1 patient). In 7 patients, the sensation vanished after burping or clearing the throat, and in 3 patients, it vanished with moving the head.

Twenty-two patients thought that conflicts in their private or professional lives (or both) underlay their globus sensation, while 26 patients believed that conflicts and organic disorders and 23 patients believed that only organic disorders underlay their globus sensation; 17 patients had no explanation.

The 82 patients who completed the BDI showed about the same depression scores as 80 consecutive general outpatients at a gastroenterology clinic (Table 1). Fifty-five patients with the globus sensation vs 49 of the control group had no signs of depression (scores <11), 19 vs 26 had signs of mild depression (scores, 11-19), 5 vs 3 had signs of moderate depression (scores, 20-25), and 3 vs 5 had signs of severe depression (scores ≥26). The 79 patients who completed the STAI scored only slightly higher for anxiety as a state and a personality trait than 83 outpatients at a gastroenterology clinic. On the MMPI, which was completed by 63 patients, 10 scored high (>70) on 1 or more of the scales, but none of the mean scores was higher than 70. The mean scores were slightly lower than in about 50000 medical outpatients.⁵⁰ The scores on the psychometric tests bore no relationship to the characteristics and duration of the globus sensation.

The patients who completed the BDI and STAI 4 to 10 months later scored markedly lower on these tests than at the initial evaluation (Table 2). There was no relationship, however, between either the score differences or the original scores and the globus sensation.

Of the 39 female and 19 male patients interviewed, 24 (62%) and 13 (68%), respectively, met criteria for at least 1 of the disorders listed in the DSM-III-R. Nineteen patients met criteria for more than 1 disorder (Table 3). Except for 3 subjects, none had been in psychiatric care previously.

Of the 6 patients meeting criteria for a current major depressive disorder and the 1 meeting those for bipolar I disorder, 2 had signs of mild depression, 1 had signs of moderate depression, and 1 had signs of severe depression on the BDI. Of the 10 patients fulfilling the criteria for a past major depressive disorder and the 1 fulfilling those for past dysthymia, 2 had signs of mild depression, 3 of moderate depression, and 1 of severe depression on the BDI. The diagnoses "undifferentiated somatoform disorder" posed in 2 patients and "hypochondriasis" posed in 1 as the sole diagnosis are questionable because in all 3 patients, somatic disorders potentially explaining the symptoms were revealed.

Psychoactive drugs were recommended for 7 patients and psychotherapy for 6 patients, but only 4 and 1, respectively, were treated. The presence and type of psychiatric disorders bore no relationship to the characteristics of the globus sensation, the physical disorders detected, and the sensation's course in the follow-up period.

Abnormalities potentially related to the globus sensation, but no pharyngoesophageal motor disorders, were found in 15 patients. Two patients were found to have chronic tonsillitis, 4 chronic pharyngitis, 1 dry oropharyngeal mucosa, 1 chronic obstructive bronchitis, 2 cervical spondylosis, 1 cervicothoracic spondylosis and disordered masticatory function, and 1 cervical spondylosis and gastroparesis. Three patients had thyroid adenomata, of whom 1 also had sinusitis and cervical spondylosis. The first-mentioned abnormalities were considered as being of primary potential importance for the globus sensation (Table 4).

At reevaluation 3 to 56 months (median, 29 months) later, the globus sensation had vanished with specific therapy in 2 patients. In 3 patients, it had decreased in intensity and frequency with specific treatment, in 2 with nonspecific treatment, and in 2 with no therapy. The sensation was unchanged in 2 patients after specific treatment and in 4 after no treatment (Table 4).

Velum palatinum insufficiency and disordered pharyngeal motility were observed videofluoroscopically in 1 patient, who also had pharyngitis and gastritis. No therapeutic steps were taken, and the globus sensation was unchanged 16 months later. In another patient, the UES opened at videofluoroscopy only 70% with swallowing; with the use of cisapride medication for epigastric fullness, the globus sensation ceased.

Twenty-four patients were revealed manometrically to have achalasia (for criteria, see the "Manometric Investigation" subsection in the "Patients and Methods" section). Scintigraphically, esophageal bolus transport was slow in 12 patients (10 to >120 seconds; median, 28.5 seconds) and in the upper range of normal (9 seconds) in 3 patients; on videofluoroscopy, barium passage was slow in 8 patients.

At the primary evaluation, none of the 24 patients had reported dysphagia. After the diagnosis was made, 21 volunteered at least 1 sign suggestive of achalasia. Of these, 1 now admitted to sometimes regurgitate saliva or food and to wake up from sleep because of aspirating saliva. Another reported that she ate only soft food and drank after every bite because this was the only way to finish a meal within a reasonable time. Two reported waking up every morning with the pillow wet from saliva, often having odynophagia, and occasionally having regurgitation. One volunteered that she woke up in the night because of aspirated saliva and that her pillow was wet from saliva in the morning; occasionally more solid food got stuck in her gullet. Another reported that she always felt a retrosternal pressure, which she ascribed to her spondylosis; because of the large amounts of saliva regurgitated postprandially, she feared that she had a serious disease. One patient from time to time had retrosternal pain and dysphagia for solids, which she attributed to her goiter. After the diagnosis of achalasia was made, 2 said they had "gastric cramps" and 1 admitted to having epigastric pain that ceased after self-induced vomiting; the regurgitated food always tasted like the food previously eaten. Another patient disclosed that she could avoid the globus sensation if she chose not to lie down after supper. One patient described regurgitating huge amounts of saliva, and 2 patients sometimes regurgitated ingesta that tasted like the food previously eaten; with these 2 patients, neither the "gastric cramps" nor the epigastric pain ceased after self-induced vomiting. Three patients sometimes had difficulties with swallowing more solid food and regurgitated saliva. A patient in whom an esophageal Candida albicans infection was revealed admitted that more solid food sometimes got stuck in his esophagus. Another reported that with swallowing, his globus sensation often became so painful that it brought tears to his eyes.

For treatment, slowly releasing forms of calcium channel blockers or nitroglycerin or pneumatic dilatation of the achalatic LES were recommended to the patients' physicians. Only 6 patients, however, received specific treatment or advice directed at changing their eating or postprandial behavior (Table 4).

Twenty-three of the 24 patients could be reinvestigated 18 to 59 months (median, 38 months) later (Table 4). With the intake of a calcium channel blocker,the globus sensation had disappeared in 2 patients, decreased in intensity in 3 patients (in 1 of them the medication had to be discontinued because of adverse effects), and remained unchanged in 1 patient. In 3 patients, the sensation ceased without therapy. In 1 patient, the sensation had become less intense after she had started to go for a walk after each meal; in 1 patient, it had become less intense with supportive psychotherapy, benzodiazepine medication, and relaxation training; in 1 patient it had decreased with slow eating and the avoidance of carbonated beverages; in 1 patient with slow eating and relaxation training; in 1 patient with the use of a tricyclic antidepressant; in 1 patient with an antacid and relaxation training; in 1 patient with thyrostatic therapy and physiotherapy for cervical spondylosis; and in 4 patients without therapy. One of the last patients had ended a stressful personal relationship and found a new partner; another patient, who at the primary assessment had suffered physical abuse by her husband, had attained a better marital relationship and a position in public life. The sensation was unchanged in 4 untreated patients.

Diffuse esophageal spasms were revealed manometrically in 1 patient, who on 24-hour pH monitoring had an abnormal GER. After these investigations, the patient admitted to having retrosternal pain when eating more solid food and to having heartburn postprandially. A histamine-2 (H₂) receptor antagonist and a calcium channel blocker were recommended. Because of adverse effects, the latter was taken for only 2 days; relaxation training yielded no alleviation. Twelve months later, the symptoms were unchanged.

Nutcracker esophagus was found manometrically in 3 patients. Thereafter, 1 patient admitted to having dysphagia for more solid food and salivary regurgitation. The use of a calcium channel blocker was recommended for 2 patients, but none was treated; 28 to 52 months later, their globus sensation was unchanged (Table 4). The third patient was treated with an antidepressant medication, and the sensation was unchanged at 28 months.

Nonspecific esophageal motor disorders were found manometrically in 30 patients; videofluoroscopy carried out in 27 patients showed slow transport or proximal escape (or both) of barium in 18 patients and only partial UES opening on swallowing in 1 patient. Bolus transport was delayed (10-93 seconds; median, 16 seconds) in 13 of 19 patients studied. In 4 patients, whose symptoms were suggestive of GER and of whom 1 had grade I esophagitis,⁵¹ 24-hour pH monitoring was carried out, but no pathological reflux was found. After these investigations, 1 patient admitted to occasionally having the impression that food did not "go down" properly and to having heartburn and saliva and food regurgitation; her contraction amplitudes reached 277 mm Hg with durations of 15 seconds. A patient with high-amplitude contractions reported having dysphagia; 3 patients with partly synchronous double-peaked contractions had no dysphagia, but 1 of them regurgitated saliva. One patient, who had a midesophageal high-pressure zone not relaxing on swallowing, reported postprandial epigastric fullness and heartburn. According to the findings, therapeutic measures, including a calcium channel blocker for 1 patient and a trial of antireflux therapy for 4 patients, were recommended, but only 6 patients were treated accordingly.

At reevaluation 5 to 58 months (median, 22 months) later, the globus sensation had vanished in 9 patients. This had occurred with the use of cisapride (1 patient), cisapride and a tricyclic antidepressant medication (1 patient), an antidepressant medication (1 patient), a calcium channel blocker (1 patient), and a thyrostatic medication (1 patient); 1 patient had stopped to "clean" her throat with cotton swabs, a procedure recommended by a general practitioner as a remedy against her chronic tonsillitis, and 1 patient had ceased to work with detergents; in 1 patient, the sensation had vanished with manometry, and in another patient, it had vanished after chronic bronchitis resolved. The sensation had become less frequent in 8 patients with the use of cisapride (1 patient), infiltration therapy of the cervical muscles (1 patient), cisapride and an H₂-receptor antagonist (1 patient), an H₂ antagonist (1 patient), a tricyclic antidepressant and relaxation training (1 patient), abstaining from coffee and carbonated beverages (1 patient), a cure for chronic bronchitis and relaxation training (1 patient), physical therapy of the neck (1 patient), and no treatment (1 patient). The sensation was unchanged in 13 patients, 8 of whom had received no treatment; 1 had taken cisapride, sucralfate, and a tricyclic antidepressant; 1 had practiced relaxation training; and 1 had received acupuncture treatment. Two patients who had received no treatment felt the sensation was unchanged; however, they were less worried because they felt assured that there was no underlying serious condition.

Pathological GER was revealed in 14 patients; all had heartburn, 8 had acid regurgitation, 3 had dysphagia, and 6 had odynophagia. One also had diffuse esophageal spasms (this patient is described earlier under that category). Esophagitis was found in 6 patients and gastritis in 4 patients, 1 of whom also had cervical spondylosis. Videofluoroscopy revealed a distal esophageal stenosis in 4 patients, and scintigraphy showed a massively delayed gastric emptying in 2 patients.

Patients were seen for 4 to 42 months (median, 26 months). Three patients became symptom-free with the use of an H₂ antagonist and cisapride; 1 patient with an H₂ antagonist, a tricyclic antidepressant medication, and sucralfate; 1 patient with cisapride, an antacid, and abstaining from coffee; 1 patient with psychotherapy; and 1 patient without treatment. The sensation was alleviated by an H₂ antagonist plus cisapride in 1 patient, by cisapride and domperidone in another patient, and after no treatment in 1 patient. It remained unchanged in 1 patient who took cisapride, 1 patient who took an antidepressant and received relaxation training, and 1 with no treatment (Table 4).

The results of the present investigation show that motor disorders, inflammatory conditions, and structural abnormalities of the pharynx and esophagus were associated with the globus sensation in most patients, which suggests that they were of significance for the occurrence of the sensation. That such relationships were not revealed to the same extent or did not receive attention in previous studies may have been due to an absence, or the patients' disregard, of dysphagia, odynophagia, regurgitation, and heartburn. Even in the present study, such symptoms were noted at renewed questioning in only 47 of the 73 patients in whom it had become obvious during the diagnostic workup that disorders of the above type were present. The gathered evidence, however, seems insufficient to infer a causal relationship between the organic disorders detected and the globus sensation, in particular because only some of the patients received the recommended specific treatment and responded favorably to it.

An explanation of why, at the primary assessment, the patients complained of globus sensation but not of dysphagia, retrosternal pain, or heartburn may be provided by studies showing that one third of patients with an obstruction at the lower end of the esophagus pointed to the throat when asked where they felt the obstruction.^52,53 In 1924, Jacobson²⁷ reported that a patient had the globus sensation while barium paste was held in his esophagus and "moderate spasm" was observed fluoroscopically. More recently, it was found that the sensation could be elicited by balloon distension in the middle and proximal esophagus.⁵⁴ This suggests that, in many patients, the globus sensation can yield no more than misleading information on the location of an underlying disorder.

A possible etiologic importance of esophageal motor disorders for the globus sensation also has been suggested earlier. In a study²⁸ of 18 patients with the sensation, 2 were found to have diffuse spasms and 2 to have nonspecific esophageal motor disorders. An impaired esophageal clearance in the horizontal position was found on radiography in a third of more than 300 patients with globus sensation and dysphagia.⁵⁵ No increased incidence of abnormal esophageal motility was found by others.^56,57

Most striking in the present study was the high number of patients in whom achalasia was revealed manometrically. This may be due more to the lack of dysphagia and esophageal dilatation in most patients than to the fact that this disorder is still widely underdiagnosed. It has been reported, however, that early stages of achalasia may not, or only intermittently,⁵⁸ cause dysphagia and that even patients with megaesophagus may remain symptom-free.⁵⁹ Food and saliva residing in the esophagus could explain the occurrence of the globus sensation mainly between meals and not during the distracting act of eating. Furthermore, during eating, food may reach the stomach unhampered because of gravity and the pharyngeal pump. The fact that in 59 patients in this study focal transducers were used to evaluate LES function instead of the sleeve technique⁶⁰ does not mean that the diagnoses, or nondiagnoses, made of achalasia are questionable. Whereas a swallow-related axial movement of the esophagus over the probe may suggest sphincter relaxation in instances in which the sensor in fact records from the gastric or esophageal lumen, an incomplete or lack of relaxation cannot be simulated by axial movement.

The low incidence of structural and functional abnormalities of the pharynx and UES encountered is consonant with earlier reports of no^11,12,28,61 or only doubtful⁵⁷ abnormal features in these locations. Our finding that 14 of the 88 patients had pathological GER is in accordance with reports of others^56,62 who also noted GER in patients with the globus sensation. Thus, an increased esophageal acid exposure has to be considered as underlying the sensation in some of the patients. Further studies have to be done to see whether a more effective therapeutic approach than one used by us, ie, the prescription of proton-pump inhibitors, would render a greater proportion of patients with GER free of the globus sensation.

Patients who accept a referral to a psychosomatically oriented institution may show more abnormal personality traits than general medical patients, but no such tendency was observed in these patients. Their BDI and STAI scores were no more indicative of depression and anxiety than were the scores of patients attending a general gastroenterologic clinic, and their scores on the MMPI were in the same range as those of a large group of medical outpatients.⁵⁰ These findings are at variance with reports that patients with the globus sensation had high levels of depression⁶³ and, on the Crown-Crisp Experiential Index, higher scores for anxiety, obsessionality, depression, phobia, and somatic concern⁷ than symptom-free persons. Earlier it had been reported¹⁰ that women with the sensation had high scores for anxiety, depression, and somatic concern⁷ and did not differ on the Crown-Crisp Experiential Index from psychiatric outpatients, whereas men had scores similar to those of subjects with no psychiatric symptoms. In one article,⁶⁴ it was stated that patients with the globus sensation had "markedly" higher-than-average scores for hypochondriasis and depression on the MMPI and in another⁶⁵ that 7 patients with the sensation had more state and trait anxiety on the STAI and more depression on the BDI than 13 healthy students.

Another striking finding of the present study was the high incidence of psychiatric disorders, either active or in remission. Again, this may be related to the fact that our patients were referred to psychosomatically oriented institutions, which may reflect the suspicion of the referring physicians that psychological factors played a part in the patients' symptom. There were no indications for relationships between the diagnosed psychiatric disorders and the development of the globus sensation or the physical disorders that were revealed. Similarly, Clouse and Lustman⁶⁶ were unable to differentiate, using a multidimensional psychometric inventory, subjects with esophageal contraction abnormalities from those without. A high lifetime prevalence of psychiatric disorders, ie, 71%, was found in 114 diabetic patients with a mean illness duration of 12.4 years.⁶⁷ High prevalence rates found also in other patients with long-standing diabetes mellitus^68,69 and chronic pain^70,71 suggest that psychiatric disorders are related to chronic suffering and incertitude rather than to specific symptoms. One explanation for the high number of psychiatric disorders diagnosed may be the item "Have you ever had trouble swallowing?" that is included in the somatization section of the Structured Clinical Interview for the DSM-III-R and the items "Did you feel as if you were choking?" and "Do you often have trouble swallowing or get a lump in your throat?" in the panic disorder and generalized anxiety disorder sections, respectively. In a study⁷² of 24 patients with globus sensation, however, it was found that 8 patients had had panic attacks in the past and that 6 met DSM-III-R criteria for panic disorder. Panic attacks could, as suggested earlier,⁷³ result from a basic disorder combined with a fearful attitude to visceral sensations. By contrast, there seems to be no basis for the suggestion that "globus hystericus reflects anxiety with a choking sensation or difficulty in swallowing."⁷⁴ Although patients may have the sensation in association with emotional upheaval, the presumption that the sensation is always attributable to emotional or mental origins is unjustified.

The high incidence of esophageal motor abnormalities in the studied patients suggests, together with results of previous investigations,^52,53 that such disorders may, instead of eliciting dysphagia, be sensed only vaguely and elicit the globus sensation. Psychological characteristics, stress, and psychiatric disorders may be relevant for the discomfort experienced but not the sensation's origin and course. A diagnostic workup of patients with the globus sensation should start with a thorough history taking and a physical and otolaryngological examination. If these are unrevealing, a preferably videocinematographic, radiological, and a manometric investigation of pharyngoesophageal motility as well as a pH-monitoring study of gastroesophageal reflux activity should be carried out, even if no symptoms of dysphagia or reflux are volunteered.

Accepted for publication November 4, 1997.

This study was supported by the Austrian Science Foundation, Vienna (project P7359-MED).

Corresponding author: Georg Stacher, MD, Psychophysiology Unit, Währinger Gürtel 18-20, A-1090 Wien, Austria.