Psychosomatic Medicine
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published online before print February 8, 2007, 10.1097/PSY.0b013e3180312cac
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kanbara, K.
Right arrow Articles by Nakai, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kanbara, K.
Right arrow Articles by Nakai, Y.
Psychosomatic Medicine 69:158-165 (2007)
© 2007 American Psychosomatic Society

ORIGINAL ARTICLES

An Exploratory Study of Subgrouping of Patients With Functional Somatic Syndrome Based on the Psychophysiological Stress Response: Its Relationship With Moods and Subjective Variables

Kenji Kanbara, MD, PhD, Mikihiko Fukunaga, MD, PhD, Hiromi Mutsuura, MD, Hiroharu Takeuchi, MD, Kana Kitamura, MD and Yoshihide Nakai, MD, PhD

From the Department of Psychosomatic Medicine, Kansai Medical University, Osaka, Japan.

Address correspondence and reprint requests to Kenji Kanbara, Department of Psychosomatic Medicine, Kansai Medical University, 10-15, Fumizono-cho, Moriguchi-shi, Osaka, #570-8507, Japan. E-mail: kanbara{at}body-thinking.com


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Objective: To study the characteristics and subjective estimations of subgroups of patients with functional somatic syndrome (FSS). A characteristic in patients with FSS was reportedly hyporeactivity in the psychophysiological stress response (PSR).

Methods: The PSR was measured in 59 FSS patients and 41 healthy controls. Autonomic lability scores (ALSs) of six psychophysiological measurements on PSR were calculated. Cluster analysis using the ALSs was performed in the FSS group. A discriminant analysis was also performed to identify the criterion of the subgrouping. Factor analysis scores of the six ALSs, and moods and subjective variables were compared between the subgroups.

Results: Cluster analysis divided the FSS patients into two clusters. Three groups (low-lability, high-lability, and control groups) were compared. All factor scores of autonomic lability significantly differed between the low- and high-lability groups, and between the low-lability and control groups. The mood scores were higher in the high-lability group than in the low-lability group. The duration of suffering was significantly longer in the high-lability group than in the low-lability group. The distributions of symptoms and diagnosis did not significantly differ between the subgroups.

Conclusions: We have tentatively verified that there are two subgroups based on the autonomic lability among FSS patients, which were independent of the type of symptoms and diagnostic category. Autonomic lability is an important axis in the multiaxial diagnosis of FSS.

Key Words: functional somatic syndrome • psychophysiological stress response • autonomic lability • subgrouping • cluster analysis • mood and subjective variables

Abbreviations: FSS = functional somatic syndrome; IBS = irritable bowel syndrome; FMS = fibromyalgia syndrome; FGID = functional gastrointestinal disorder; ALS = autonomic lability score; PSR = psychophysiological stress response; ANS = autonomic nervous system; VAS = visual analogue scale; SEMG = surface-electromyography; TEMP = skin temperature; SCL = skin conductance level; NSSCR = nonspecific skin conductance response; BVPAmp = blood volume pulse amplitude; STS = subjective tension score.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Somaticsymptoms of medically unknown origin are highly prevalent in not only the primary care setting but also the secondary care setting (1–3) and are clinically important (4). Functional somatic syndrome (FSS) describes the groups of syndromes of medically unknown origin (4–6). In the present study, the following definition of FSS is used: "functional somatic syndrome refers to several related syndromes that are characterized more by symptoms, suffering, and disability than by disease-specific, demonstrable abnormalities of structure or function" (6). FSSs include various diseases in many medical specialties such as irritable bowel syndrome (IBS), functional dyspepsia, fibromyalgia syndrome (FMS), and chronic fatigue syndrome. Although FSS is poorly understood and heterogeneous, the similarities seemingly outweigh the differences (4). Therefore, they are expected to share a common pathophysiology (6).

The following two aspects should be considered in the pathophysiology of FSS: a) dysregulation of the stress response and b) psychological co-factors that modulate the expression of symptoms. Dysregulation of the stress response may be caused by dysregulation of the autonomic nervous system (ANS) and/or the hypothalamic-pituitary axis. In the present study, we focused on dysregulation of the ANS and evaluated the response of ANS-related psychophysiological parameters to stress. To evaluate the autonomic dysregulation, assessment of the psychophysiological stress response (PSR), which is a method of estimating the response to a stress by simultaneously measuring multiple psychophysiological parameters based on the concept of "autonomic response specificity" (7,8), has been used.

Various psychophysiological parameters have been used to assess the PSR: surface-electromyography (SEMG), skin temperature (TEMP), skin conductance, and blood volume pulse (BVP). Skin conductance reflects changes in emotional palmer sweating. BVP is a phasic measure of the pulsatile change in blood flow. The stress loads have been categorized into mental work and sensory intake (9). The mental arithmetic task is one of the mental works, easy to administer and noninvasive, and is similar to the stress in the workplace.

The discrepancy between medical assessment and the subjective experience of the patient presents difficulties in FSS. Considering the second aspect, psychological co-factors, the subjective feelings and symptoms of patients were also investigated.

There have been many studies on the PSR of patients with physical or mental disorders. Psychophysiological hyperreactivity has been thought to be one of the pathways in the link between stress and illness (10). SEMG showed a hyperreactive response to stress in patients with migraine (11) and tension-type headache (12). Stress-related increases in the SEMG have been observed in the paraspinal muscles of patients with chronic back pain (13). Patients with somatization syndrome reportedly have increased psychophysiological arousal and morning cortisol concentration (14). In patients with FMS, the baseline level of skin conductance is elevated and there is less vasoconstriction on stimulation (15), or impaired ANS was observed (16–18). Okifuji and Turk published a review on FMS and stated that hypofunctional stress systems play an important role in the pathophysiology of FMS (19). They also mentioned that no studies have attempted to test the psychophysiologic-heterogeneity hypothesis of FMS. Regarding IBS, several studies have suggested that these patients have abnormalities in both sympathetic and parasympathetic influences (20). Thus, although several studies on each disorder of FSS have been reported, to our knowledge, there have been no studies on psychophysiological characteristics in FSS as a whole.

Our previous study suggested that the tension-related psychophysiological parameters were significantly hyporeactive to stress and the subjective feelings of tension were higher in the patients with FSS compared with the controls (21). Another study suggested that the patients with psychosomatic diseases tended to have hyporeactive PSRs compared with the control group (22). Possible explanations for these findings were as follows: a) the psychophysiological tension is consistently hyporeactive in patients with FSS, or b) there are both hyporeactive and hyperreactive patients with FSS, although the number of hyporeactive patients is greater. In the present study, we hypothesized that there are at least two subgroups of FSS patients with different PSRs independent of orthodox classification of diseases.

To verify the hypothesis, we measured psychophysiological parameters and subjective parameters during a stress task. Subsequently, cluster analysis using autonomic lability to the stress was performed to investigate the existence of subgroups of FSS patients. The characteristics of the subgroups, their relationship with subjective estimation, and the independency of the subgrouping were investigated.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Subjects
The subjects were 59 patients (29 males and 30 females; age range 22 to 60 years (35.9 ± 9.5 SD)) who were diagnosed with FSS according to the steps described below.

Patients who visited the Department of Psychosomatic Medicine at Kansai Medical University Hospital in Osaka, Japan, between April 2002 and July 2004, were assessed for inclusion in the study; those who were subsequently referred to other departments were excluded from the study. Of the remaining patients, 91 patients whom the attending physicians thought would benefit from psychophysiological assessment were tested for their PSR. Twelve patients who were either <21 years or >61 years, and five patients on whom we had inadequate data were excluded from the study.

In the present study, a patient as diagnosed as having FSS if he/she met the following three conditions, which were based on the definition of FSS and were partially brought from proposed criteria for somatization (23): a) the chief complaints were somatic symptoms that could not be explained medically or by a psychiatric disorder; b) the subjective symptom rating score based on a visual analogue scale (VAS) was ≥3, and the duration of the symptoms was ≥6 months; and c) the patient had disabilities in social or daily activity due to the symptoms. Based on these conditions, 12 patients were excluded from the study. Of the remaining 62 patients, three patients had inadequate data at the time of analysis and were excluded. Finally, the remaining 59 patients were entered into the analyses of this study.

The patients were classified into the following four categories according to the patient’s predominant symptoms for the analyses: a) digestive symptoms (n = 14); b) pain or headache (n = 16); c) musculoskeletal symptoms (n = 15); and d) other (including general fatigue) (n = 14). They were also classified according to the clinical diagnosis: a) functional gastrointestinal disorder (FGID) (n = 13); b) chronic pain and tension-type headache (n = 16); c) dystonia (writer’s cramp, torticollis, and other dystonia) (n = 13); and d) other/systematic (n = 17).1

Forty-one healthy subjects (18 males and 23 females; age range 23 to 59 years, (33.1 ± 9.4 SD)) served as controls. They were recruited through public announcement; individuals who were regularly receiving medical care or had somatic symptoms were excluded. They were paid 5000 yen each for their participation.

Table 1 shows the characteristics of the subjects. There were no significant differences between the FSS and control groups in age (Mann-Whitney U test; p = .151), body mass index (BMI) (p = .760), and the male/female ratio ({chi}2 test, {chi}2 = 0.268, p = .685). The study was approved by the ethical committee of Kansai Medical University.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Characteristics and POMS Scores of the FSS Patients and the Control Subjects

 

Psychophysiological Measurements and Data Processing
The PSR was examined using a biofeedback system (MultiTrace, Stens Corp., Oakland, CA)/ ProComp+ (Thought Technology Ltd., Montreal, Canada), which was connected to a computer. Psychophysiological measurements were made before any therapies were begun. The following indices were measured using ProComp+standard sensors: SEMG, skin conductance level (SCL), TEMP, and BVP. SEMG was recorded from the frontal muscle just above both eyebrows (27), with a bandpass of 20 to 500 Hz. The SCL was recorded from the middle phalanx of the second and third fingers of the nondominant hand. The electrodes used were pregelled 1 cm2 circular Ag-AgCl single standards. TEMP was recorded from the distal phalanx of the index finger of the nondominant hand. BVP was recorded from the distal phalanx of the thumb of the nondominant hand. Data were sampled at 32 Hz.

The data that had been recorded during the initial 1 minute of each 5-minute period were deleted, and the data during the remaining 4 minutes were obtained. For SEMG (unit: µV), SCL (µS), and TEMP (°C), we calculated the mean values during 4 minutes in each period. The pulse rate (PR) (beats/minute) and BVPAmplitude (BVPAmp) (%) were automatically calculated from the BVP using the MultiTrace software. BVPAmp reflects the degree of vasodilation or constriction of peripheral vessels. SCL is the tonic level and skin conductance response is the phasic increase in conductance. The number of nonspecific skin conductance response (NSSCR) (number/minute) was calculated from the SCL by peak analysis with a width of 1 second and size of 0.05 µS (28). Thus, the six variables (PR, SEMG, SCL, NSSCR, TEMP, and BVPAmp) were obtained in each of the three periods. The data processing was performed using the waveform data analysis program (DADiSP/PRO v.4.1, DSP Development Corp., Newton, MA).

Lacey’s autonomic lability score (ALS) (29) was used to assess the responses to the stress, excluding the effect of the baseline values. The ALS was calculated for each psychophysiological variable using the standardized value measured during the baseline and the stress period.2 For TEMP, standardized values measured during the baseline and poststress periods were used because the TEMP value changes more slowly than the other parameters and the period-group interaction was found to have an effect at baseline and during the poststress period, but not during the stress period. Standardization was performed on the values of the FSS subjects when cluster analysis on the FSS subjects was performed. Standardization was performed across all samples when calculating the ALSs for factor analysis, and it became possible to perform comparisons between subgroups of the FSS group and the control group.

Subjective and Mood Measurements
To evaluate the temporary mood states, the Profile of Mood States (POMS) test (30) was administered. The POMS, a 65-item scale, assesses six affective mood dimensions. The subscales consist of tension-anxiety (TA), depression-dejection (D; depressive mood or discouragement), anger-hostility (A; irritation, fury, and aggression), vigor-activity (V; activity, liveliness, and happiness), fatigue-inertia (F; remaining inactive), and confusion-bewilderment (C). The POMS scales were demonstrated to show high internal consistency (31).

To evaluate the subjective severity of symptoms in the FSS patients, the subjective symptom score was obtained using a VAS from 0 (absent) to 10 (most severe). The duration of suffering, i.e., the length of time after the onset of the subjective symptoms, was recorded on a monthly basis. To evaluate the subjective feelings of tension, the subjective tension score (STS) (21) was obtained using a VAS from 0 to 10 during the three periods. These scores were determined to 1 decimal point.

Procedure
Psychophysiological and subjective measurements were made in an examination room of the hospital. The room temperature was kept at 23° to 26°C. Before applying the electrodes to the subjects with FSS, these persons were asked to rate the subjective symptom score and the duration of suffering. Then, the POMS test was administered. Subsequently, the PSR was examined as follows.

The subject was seated in a chair with eyes closed during the measurements. After a 5-minute adaptation period and 2-minute preparation period, psychophysiological measurements were made during the following three periods (5 minutes each, total of 15 minutes): a) baseline resting period: the subject was instructed to relax and make himself/herself comfortable; b) stress period (mental arithmetic task): the subject was instructed to subtract 7 serially from 1000 as accurately and as quickly as possible; and c) poststress period: the subject was instructed to relax. After all measurements were made, the subject was asked to estimate the STS.

Statistical Methods
To compare the changes in each parameter in the three periods in the FSS and control groups, two-way repeated measures analysis of variance (ANOVA) with "period" (three levels: baseline resting, stress, and poststress periods) as the within-subjects factor and "group" (two levels: control and FSS groups) as the between-subjects factor was performed. Greenhouse-Geisser adjustments for degree of freedom were used. To evaluate the validity of the stress task, the effect of period on the STS was analyzed by one-way repeated measures ANOVA.

Hierarchical cluster analysis with the 6 ALSs (standardized on subjects in the FSS group) in the FSS group was carried out. Ward method was used for clustering and cases only were clustered. A discriminant analysis by Mahalanobis distance was performed to investigate the criterion of subgrouping and evaluate which variables discriminate between the subgroups. Multivariate ANOVA using Wilks’ {lambda} and univariate ANOVAs were performed to confirm the difference among the clusters.

For data reduction and evaluation of the latent factor of variables, factor analysis with promax rotation using the 6 ALSs (standardized on all subjects) of psychophysiological variables in all subjects was performed. Factors with Eigen values >1.0 were extracted using a principal factor method. Based on the factor analysis, factor scores for each factor in each subject were estimated by the regression method. One-way ANOVA with Tukey’s "Honestly Significantly Different" (HSD) post hoc test was performed to investigate the significance of differences among subgroups and the control group in the factor score and subjective variables. The {chi}2 test was performed to confirm the deviation of the distribution of symptom categories or diagnosis categories.

Statistical analyses were performed (SPSS 11.5J for Windows, SPSS Inc., Chicago, IL). The {alpha} level was fixed at 0.05.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 NOTES
 REFERENCES
 
Comparison of the PSR Between the FSS and Control Groups
The POMS scores for TA, D, F, and C were significantly higher and the score for V was significantly lower in the FSS group than in the control group (Mann-Whitney U test; p < .001 each) (Table 1).

Mean and SD values of each parameter at each period in the FSS group and the control group are expressed in Table 2. Two-way multivariate ANOVA showed that period (Wilks’ {lambda} (12, 382) = 0.245, p < .001) and the period-group interaction (Wilks’ {lambda} (12, 382) = 0.827, p < .001) had significant effects on the PSR. Univariate analyses revealed that the PR was significantly higher (F(1, 98) = 9.12, p < .01); the frontal SEMG was significantly lower (F(1, 98) = 11.82, p < .001); and the PR, NSSCR, TEMP, and BVPAmp were significantly hyporeactive to the stress task (F(1.43, 139.90) = 5.31, p < .05; F(1.20, 117.87) = 4.38, p < .05; F(1.59, 155.74) = 3.92, p < .05; F(1.87, 183.60) = 5.89, p < .01; respectively) in the FSS group than in the control group.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Mean Values and Standard Deviations of Each Parameter at Each Period

 

Clustering and Discriminant Analysis
Cluster analysis divided the 59 FSS patients into two clusters: Cluster 1 (n = 39 patients) and Cluster 2 (n = 20 patients). The dendrogram was presented in Figure 1. In the present study, we determined the number of clusters was two because it is meaningful based on the following ANOVA and the number of samples. The ALSs for each parameter in the two clusters are shown in Table 3. Multivariate ANOVA showed that six parameters were differentiated by the cluster solution (Wilks’ {lambda} (6, 52) = 0.390, p < .001). On univariate ANOVAs, the ALSs for PR, SCL, NSSCR and BVPAmp in Cluster 1 were significantly smaller than in Cluster 2 (Table 3). Therefore, we tentatively called Cluster 1 the "low-ALS group" and Cluster 2 the "high-ALS group".


Figure 17
View larger version (7K):
[in this window]
[in a new window]

  		Figure 1. Dendrogram by a cluster analysis using Ward method in the patients with functional somatic syndrome. Cases are listed along the left vertical axis. The horizontal axis shows the mean Euclidean distance between clusters.

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Autonomic Lability Scores of Each Parameter in the Two Clusters of the Group With Functional Somatic Syndrome

 

Discriminant analysis identified the weight of the variables to separate the clusters. The standardized coefficients are shown in Table 4. The discriminant function was significant (Wilks’ {lambda} = 0.390, p < .001). The correct classification was 93.2% for the original cases and 89.8% for the cross-validated cases in the patient group. Control cases were classified by the discriminant criteria of the patient group clustering into 16 cases (39.0%) of low-ALS group and 25 cases (61.0%) of high-ALS group.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Standardized Coefficients of Discriminant Analysis in the Group With Functional Somatic Syndrome

 

Factor Analysis
Factor analysis with the ALSs of six psychophysiological variables in all subjects demonstrated three factors. The rotated component matrix is shown in Table 5. The cumulative sum of squared loadings of the three factors was 72.72%. The three factors were assumed to be the following from the factor matrix: a) mental tension lability-related factor, b) peripheral vascular response-related factor, and c) heart rate lability-related factor.


View this table:
[in this window]
[in a new window]

  		TABLE 5. Rotated Factor Matrix by Factor Analysis with Promax Rotation

 

Significant positive correlations between the first factor score and POMS TA, D, A, and F (Spearman r = 0.391, p < .005; r = 0.299, p < .05; r = 0.345, p < .01; r = 0.410, p < .001, respectively) and between the third factor score and POMS TA (Spearman r = 0.261, p < .05) were observed in the FSS group. No significant correlations were observed between factor score and POMS score in the control group except for between first factor score and POMS C (r = 0.328, p < .05).

Characteristics and Differences Between the Subgroups
The factor scores and subjective variables in the low-ALS, high-ALS, and control groups are shown in Table 6. Significant differences were not observed among the three groups in age (F(2, 97) = 0.20, p = .14), BMI (F(2, 97) = 0.65, p = .52), and male/female ratio ({chi}2 = 3.32, p = .190).


View this table:
[in this window]
[in a new window]

  		TABLE 6. Characteristics and Factor Scores of the Two Subgroups of Patients With Functional Somatic Syndrome and Control Group

 

ANOVA showed significant differences in all three factor scores among the three groups (F(2, 97) = 20.39, p < .001; F(2, 97) = 9.44, p < .001; F(2, 97) = 15.54, p < .001; 1st to 3rd factor, respectively). The post hoc test showed that all three factor scores were significantly lower in the low-ALS group than in the control group (p < .001, p < .001, p < .001; 1st to 3rd factor, respectively), and lower in the low-ALS group than in the high-ALS group (p < .001, p < .01, p < .001; 1st to 3rd factor, respectively). Although the differences in the three factor scores between the high-ALS group and control group did not reach statistical significance, the 1st and 3rd factor scores tended to be higher in the high-ALS group than in the control group.

With respect to the subjective and mood variables, significant differences among the three groups were found in all POMS scores (TA, F(2, 95) = 16.90, p < .001; D, F(2, 95) = 14.81, p < .001; A, F(2, 95) = 7.92, p < .001; V, F(2, 95) = 12.55, p < .001; F, F(2, 95) = 14.75, p < .001; C, F(2, 95) = 8.48, p < .001). The profiles of the POMS scores of the patient subgroups and control group are shown in Figure 2. The POMS scores were high (low in POMS V) in the high-ALS group, lower (higher in POMS V) in the low-ALS group, and lowest (highest in POMS V) in the control group. The post hoc tests revealed significant differences between each two subgroups with a few exceptions (Table 6).


Figure 27
View larger version (16K):
[in this window]
[in a new window]

  		Figure 2. Profile of the POMS scores of the two subgroups of FSS patients and control group. The control group (n = 41; thin solid line with black circles) displayed low POMS TA, D, A, F, and C scores and a high POMS V score. The high-ALS group (n = 20, thick solid line with triangles) displayed high POMS TA, D, A, F, and C scores and a low POMS V score. The low-ALS group (n = 39, broken line with squares) displayed moderate POMS scores. POMS = Profile of Mood States; TA = tension-anxiety; D = depressive mood or discouragement; A = anger-hostility; F = fatigue-inertia; C = confusion-bewilderment; V = vigor-activity; ALS = autonomic lability score.

 

The duration of suffering was significantly longer in the high-ALS group than in the low-ALS group (F(1, 57) = 5.98, p < .05). There was no significant difference in the subjective symptom score between the two subgroups. The STS among the three groups did not significantly differ.

Distributions by Symptom and Diagnosis Categories
Table 7 shows the distributions of the FSS patients according to categories of their predominant symptoms or categories of diagnosis in each subgroup. The distributions in the two subgroups did not significantly differ (symptoms, {chi}2 = 2.197, p = .53; diagnosis, {chi}2 = 0.566, p = .90).


View this table:
[in this window]
[in a new window]

  		TABLE 7. Distribution of Predominant Symptoms and Diagnoses in the Low- and High-ALS Subgroups

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
NOTES
 REFERENCES
 
The FSS group was significantly hyporeactive to the stress task in most of the parameters compared with the control group. The baseline value of the FSS group was higher in PR. Since hyporeactivity might be based on the high baseline value, we used not the difference score but ALS to exclude the influences. The observation that the STS significantly changed indicates that the stress task caused alterations in the feeling of tension. These tendencies were compatible with previous work (21,22) and confirm the characteristics of FSS patients. Our hypothesis was that there are several subtypes of dysfunction in FSS.

Factor analysis showed three factors (Table 5). The first factor was closely correlated with SCL and NSSCR. Skin conductance reflects palmer sweating, which originates in the central nervous system rather than peripheral perspiration, and indicates the general state of arousal and alertness (28). Thus, it seems reasonable to designate the first factor as the "mental tension lability-related factor." The second factor is closely correlated with TEMP and BVPAmp. Because these two parameters are associated with constriction and dilation of peripheral arteries (32), it seems reasonable to designate the second factor as the "peripheral vascular response-related factor." The third factor is closely correlated with PR and SEMG. The PR is usually equivalent to the heart rate and designation of the third factor as the "heart rate lability-related factor" seems reasonable, but this factor may include partially somatic tension.

Cluster analysis divided the FSS patients into two clusters by autonomic lability. All three factor scores were significantly lower in the low-ALS group (Cluster 1) than in the high-ALS group (Cluster 2) (Table 6), indicating the validity of the naming of these clusters. Differences between the two clusters were particularly large for PR, SCL, and NSSCR (Table 3). Discriminant analysis also showed these parameters mainly discriminate the subgroups (Table 4). Based on these findings and factor scores in Table 6, the two subgroups were found to differ especially with regard to mental tension and heart rate labilities.

Because a larger proportion was in the low-ALS group than in the high-ALS group, the features of the FSS group as a whole reflected those of the low-ALS group. Although the lability in the low-ALS group was smaller than that in the control group, there were no significant differences between the high-ALS group and control group. This finding may suggest that the low-ALS group has a more critical pathology of autonomic lability than the high-ALS group, considering that control cases were classified into a larger proportion in the high-ALS group by the discriminant criteria. Furthermore, the samples have possibilities to be clustered into three groups: low-ALS, normal, and high-ALS group, if the sample size is larger. Then, the pathologic condition of high-ALS group (and normal lability group) may be clarified distinctly.

The duration of suffering was significantly longer in the high-ALS group (Table 6). One possibility is that because the disorders in the high-ALS group are refractory, the duration of suffering of this group becomes longer. Another possibility is that when a disorder is prolonged, the psychophysiological lability increases. Steptoe stated that even if exaggerated reactivity may not be directly involved in etiology, it may be involved in maintenance of the preexisting health problem (10). The assumption that a primarily vulnerable system is maintained by autonomic hyperreactivity is compatible with the present result. On the other hand, the subjective symptom score did not significantly differ among subgroups. This indicates that the axis of subgrouping, psychophysiological lability, related to not the severity but the sustainment of symptoms.

The results of POMS scores indicate that moods of tension, anxiety, depression, and anger are high and vigor was low in the high-ALS group than in the low-ALS group (Figure 1). In healthy subjects, negative emotions positively correlate with physiological measures (33,34). To our knowledge, the literature on patients with FSS is sparse. In the present study, significant correlations were observed between the first factor and most of the POMS score in FSS patients. This result supports our decision to designate the first factor as the "mental tension lability-related factor" and indicates that abnormal moods were heightened along with the first factor of ALSs in this subgroup.

Thus, the group with FSS could be divided into two subgroups by autonomic lability and consequently the two subgroups differed in mood. Autonomic functions are involved in emotional processing. Several studies reported the relationship between alexithymia (a trait characterized by difficulties in identifying, communicating, or expressing own feelings) (35) and attenuated autonomic reactivity to a stress or hypoarousal autonomic response (36,37). Therefore, low-ALS group in the present study is likely to include certain amount of alexithymics. This assumption is consistent with the results that mood scores in the low-ALS group were lower than those in the high-ALS group. Although an etiology of FSS was still unknown, our results suggested at least two pathologies: a) one subset has low autonomic lability and the alexithymic process and b) another subset has high autonomic lability (but rather close to the control group) and heightened mood states.

Such subgrouping would be meaningless if the subgroups reflected a difference in symptoms or diagnostic category. There were no significant differences between the two subgroups with regard to symptoms and diagnostic categories (Table 7). Therefore, the subgrouping was based on a criterion independent of symptom or diagnostic category.

Patients with FMS have been divided into subgroups based on psychosocial and behavioral aspects (38), or based on mood, cognitive, and neurobiologic aspects (39). However, no previous study has subgrouped patients based on autonomic function. Some studies reported that the autonomic function on diseases related to FSS is hyporeactive, whereas other studies reported it is hyperreactive. We have tentatively verified the hypothesis that there are hyporeactive and hyperreactive subgroups of FSS patients. This proposal is not contradictory to the results of various previous studies, but it seems to be more essential.

We consider that the subgrouping of FSS patients is useful for the diagnosis, treatment, and prediction of the prognosis of FSS patients. First, for the diagnosis of FSS, multiaxial diagnosis has been reportedly more useful than orthodox classification (4). The results of our study indicate the importance of examining patients with FSS on the axis of autonomic lability, especially when psychophysiological interventions are being considered. Second, with regard to therapeutic approaches for patients with FSS, antidepressant (40), psychological, or cognitive behavioral therapy (4) have been recommended. Psychophysiological interventions are considered to be effective in reverting the hypofunctional stress systems to normal. Based on the findings of the present study, the method of approach should be altered according to the subgroup to which patients belong by using the discriminant criteria. Patients in the high-ALS-and-high-POMS group should be treated in such a manner that their mood is improved and their lability is lowered by relaxation. Although the patients in this group have rather normal autonomic labilities and abnormal mood states, they must be treated with a focus on improvement of mood states. On the other hand, biofeedback or behavioral therapy, which improves the flexibility of patients, may be more suitable for those in the low-ALS-and-low-POMS group. Third, with regard to prognosis, we have come across no description in the literature regarding the prognosis of patients with FSS. Because we found a relationship between subgrouping and the duration of suffering, however, there is a possibility that subgrouping is a prognostic factor. A prospective follow-up study on the matter should be conducted in the future.

Limitations of This Study
The minimum sample size for clustering and factor analysis is upheld. But the number of clusters is limited in this sample size. The results about differences among subgroups (Table 6) suggested that, for comparison between the high-ALS group and the control group, clustering is expected into three groups: low-ALS, normal, and high-ALS. Such investigation needs a larger sample size.

We cannot exclude the possibility that our group with FSS was a biased sample of all patients with FSS. But the bias might not be so large based on the distributions of the diagnostic or symptomatic category. Also, the psychophysiological indices in this study were not complete. Indices representing parasympathetic activities were particularly deficient. We are pursuing our study with a complete set of indices.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
REFERENCES
 
1FGID was diagnosed by the ROME II criteria (24). Chronic pain was diagnosed according to the guidelines of the Chronic Pain Section, American Society of Anesthesiologists Task Force on Pain Management (25). Dystonia was diagnosed according to the concept and classification of dystonia (26). Back

2ALS was calculated by the following formula: ALS = 50 + 10(Yi – Xirxy)/(1 – r2xy)0.5, where Xi is standardized baseline value, Yi is standardized value during stress period, rxy is correlation coefficient for the sample between baseline and values during stress period (29). Back

Received for publication April 21, 2005; revision received September 26, 2006.

DOI:10.1097/PSY.0b013e3180312cac


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 

  1. Kellner R. Functional somatic symptoms and hypochondriasis. A survey of empirical studies. Arch Gen Psychiatry 1985;42:821–33.[Abstract/Free Full Text]
  2. Peveler R, Kilkenny L, Kinmonth AL. Medically unexplained physical symptoms in primary care: a comparison of self-report screening questionnaires and clinical opinion. J Psychosom Res 1997;42:245–52.[Medline]
  3. Reid S, Wessely S, Crayford T, Hotopf M. Frequent attenders with medically unexplained symptoms: service use and costs in secondary care. Br J Psychiatry 2002;180:248–53.[Abstract/Free Full Text]
  4. Wessely S, Nimnuan C, Sharpe M. Functional somatic syndromes: one or many? Lancet 1999;354:936–9.[CrossRef][Medline]
  5. Kirmayer LJ, Robbins JM. Functional somatic syndromes. In: Kirmayer LJ, Robbins JM, editors. Current concepts of somatization: research and clinical perspectives. Washington, DC: American Psychiatric Press; 1991.
  6. Barsky AJ, Borus JF. Functional somatic syndromes. Ann Intern Med 1999;130:910–21.[Abstract/Free Full Text]
  7. Lacey JI, Bateman DE, VanLehn R. Autonomic response specificity. An experimental study. Psychosom Med 1953;15:8–21.[Abstract/Free Full Text]
  8. Wenger MA, Clemens TL, Coleman MA, Cullen TD, Engel BT. Autonomic response specificity. Psyshosom Med 1961;23:185–93.
  9. Williams RB Jr, Lane JD, Kuhn CM, Melosh W, White AD, Schanberg SM. Type A behavior and elevated physiological and neuroendocrine responses to cognitive tasks. Science 1982;218:483–5.[Abstract/Free Full Text]
  10. Steptoe A. The links between stress and illness. J Psychosom Res 1991;35:633–44.[CrossRef][Medline]
  11. Cohen MJ, Rickles WH, McArthur DL. Evidence for physiological response stereotypy in migraine headache. Psychosom Med 1978;40:344–54.[Abstract/Free Full Text]
  12. Hatch JP, Moore PJ, Borcherding S, Cyr-Provost M, Boutros NN, Seleshi E. Electromyographic and affective responses of episodic tension-type headache patients and headache-free controls during stressful task performance. J Behav Med 1992;15:89–112.[CrossRef][Medline]
  13. Flor H, Turk DC. Psychophysiology of chronic pain: do chronic pain patients exhibit symptom-specific psychophysiological responses? Psychol Bull 1989;105:215–59.[CrossRef][Medline]
  14. Rief W, Shaw R, Fichter MM. Elevated levels of psychophysiological arousal and cortisol in patients with somatization syndrome. Psychosom Med 1998;60:198–203.[Abstract/Free Full Text]
  15. Qiao ZG, Vaeroy H, Morkrid L. Electrodermal and microcirculatory activity in patients with fibromyalgia during baseline, acoustic stimulation and cold pressor tests. J Rheumatol 1991;18:1383–9.[Medline]
  16. van Denderen JC, Boersma JW, Zeinstra P, Hollander AP, van Neerbos BR. Physiological effects of exhaustive physical exercise in primary fibromyalgia syndrome (PFS): is PFS a disorder of neuroendocrine reactivity? Scand J Rheumatol 1992;21:35–7.[Medline]
  17. Cohen H, Neumann L, Shore M, Amir M, Cassuto Y, Buskila D. Autonomic dysfunction in patients with fibromyalgia: application of power spectral analysis of heart rate variability. Semin Arthritis Rheum 2000;29:217–27.[CrossRef][Medline]
  18. Martinez-Lavin M, Hermosillo AG, Rosas M, Soto ME. Circadian studies of autonomic nervous balance in patients with fibromyalgia: a heart rate variability analysis. Arthritis Rheum 1998;41:1966–71.[CrossRef][Medline]
  19. Okifuji A, Turk DC. Stress and psychophysiological dysregulation in patients with fibromyalgia syndrome. Appl Psychophysiol Biofeedback 2002;27:129–41.[CrossRef][Medline]
  20. Aggarwal A, Cutts TF, Abell TL, Cardoso S, Familoni B, Bremer J, Karas J. Predominant symptoms in irritable bowel syndrome correlate with specific autonomic nervous system abnormalities. Gastroenterology 1994;106:945–50.[Medline]
  21. Kanbara K, Mitani Y, Fukunaga M, Ishino S, Takebayashi N, Nakai Y. Paradoxical results of psychophysiological stress profile in functional somatic syndrome: correlation between subjective tension score and objective stress response. Appl Psychophysiol Biofeedback 2004;29:255–68.[CrossRef][Medline]
  22. Kanbara K, Mitani Y, Fukunaga M, Ishino S, Takebayashi N, Nakai Y. Characteristics of psychophysiological stress responses in patients with psychosomatic disorders. Jpn J Psychosom Med 2005;45:685–95.
  23. Rief W, Hiller W. Toward empirically based criteria for the classification of somatoform disorders. J Psychosom Res 1999;46:507–18.[CrossRef][Medline]
  24. Drossman DA, Corazziari E, Talley NJ, Thompson WG, Whitehead WE. Rome II: The functional gastrointestinal disorders. 2nd ed. McLean: Degnon Associates Inc.; 2000.
  25. Practice guidelines for chronic pain management. A report by the American Society of Anesthesiologists Task Force on Pain Management, Chronic Pain Section. Anesthesiology 1997;86:995–1004.[CrossRef][Medline]
  26. Fahn S. Concept and classification of dystonia. Adv Neurol 1988;50:1–8.[Medline]
  27. Lippold OCJ. Electromyography. In: Venables PH, Martin I, editors. A manual of psychophysiological methods. Amsterdam: North-Holland; 1967.
  28. Dawson ME, Schell AM, Filion DL. The electrodermal system. In: Cacioppo JT, Tassinary LG, Berntson GG, editors. Handbook of psychophysiology. 2nd ed. Cambridge: Cambridge University Press; 2000.
  29. Lacey JI. The evaluation of autonomic responses: toward a general solution. Ann NY Acad Sci 1956;67:125–63.[CrossRef][Medline]
  30. McNair DM, Lorr M, DroppLemn LF. EITS manual for the profile of mood states (POMS). San Diego: Educational and Industrial Testing Service; 1971.
  31. McNair DM, Lorr M, Droppleman LF. Profile of mood states manual (rev.). San Diego: Educational and Industrial Testing Service; 1992.
  32. Cohen MJ, Schandler SL. A modular system for detecting and displaying plethysmographic blood volume pulse amplitude. Psychophysiology 1979;16:80–3.[CrossRef][Medline]
  33. Pollock V, Cho DW, Reker D, Volavka J. Profile of mood states: the factors and their physiological correlates. J Nerv Ment Dis 1979;167:612–4.[CrossRef][Medline]
  34. Carrillo E, Moya-Albiol L, Gonzalez-Bono E, Salvador A, Ricarte J, Gomez-Amor J. Gender differences in cardiovascular and electrodermal responses to public speaking task: the role of anxiety and mood states. Int J Psychophysiol 2001;42:253–64.[CrossRef][Medline]
  35. Sifneos PE. The prevalence of ‘alexithymic' characteristics in psychosomatic patients. Psychother Psychosom 1973;22:255–62.[Medline]
  36. Neumann SA, Sollers JJ III, Thayer JF, Waldstein SR. Alexithymia predicts attenuated autonomic reactivity, but prolonged recovery to anger recall in young women. Int J Psychophysiol 2004;53:183–95.[CrossRef][Medline]
  37. Linden W, Lenz JW, Stossel C. Alexithymia, defensiveness and cardiovascular reactivity to stress. J Psychosom Res 1996;41:575–83.[CrossRef][Medline]
  38. Turk DC, Okifuji A, Sinclair JD, Starz TW. Pain, disability, and physical functioning in subgroups of patients with fibromyalgia. J Rheumatol 1996;23:1255–62.[Medline]
  39. Giesecke T, Williams DA, Harris RE, Cupps TR, Tian X, Tian TX, Gracely RH, Clauw DJ. Subgrouping of fibromyalgia patients on the basis of pressure-pain thresholds and psychological factors. Arthritis Rheum 2003;48:2916–22.[CrossRef][Medline]
  40. O'Malley PG, Jackson JL, Santoro J, Tomkins G, Balden E, Kroenke K. Antidepressant therapy for unexplained symptoms and symptom syndromes. J Fam Pract 1999;48:980–90.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kanbara, K.
Right arrow Articles by Nakai, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kanbara, K.
Right arrow Articles by Nakai, Y.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS