Source: Psychosomatics Vol. 49, #3, pp 235-242 Date: May 2008 URL: http://psy.psychiatryonline.org/cgi/content/abstract/49/3/235
Comorbid Somatic Symptoms and Functional Status in Patients With Fibromyalgia and Chronic Fatigue Syndrome: Sensory Amplification as a Common Mechanism
Correspondence: Michael E. Geisser, Ph.D., Univ. of Michigan Health System Dept. of Anesthesiology and Intensive Care, 325 E. Eisenhower Parkway, Ann Arbor, MI 48108. e-mail: email@example.com
Background Somatic symptoms are common in conditions such as fibromyalgia (FM) and chronic fatigue syndrome (CFS).
Objective Authors investigated a potential shared pathologic mechanism: a generalized perceptual abnormality where there is heightened responsiveness to varied sensory stimulation, including pain.
Method A composite measure of sensory sensitivity was created and compared with measures of somatic symptoms, comorbid psychological disturbances, and self-reported physical functioning in 38 patients with FM and/or CFS.
Results Sensory amplification influenced physical functioning indirectly through pain intensity, and physical symptoms and fatigue also independently contributed to physical functioning.
Conclusion Sensory amplification may be an underlying pathophysiologic mechanism in these disorders that is relatively independent of depression and depressive symptoms.
The diagnoses of fibromyalgia (FM) and chronic fatigue syndrome (CFS) emphasize pain and fatigue, respectively, as the hallmark symptoms of each condition.1,2 These conditions and other chronic multi-symptom illnesses are not discrete, but rather share symptoms such as pain, fatigue, sleep disturbance, and memory problems.1,3 Although some literature suggests that the presence of numerous comorbid physical symptoms among persons with FM and CFS is suggestive of somatization disorder,4 others propose that the presence of comorbid physical symptoms suggests shared underlying pathogenic mechanisms.5,6
One potential shared pathologic mechanism is a generalized perceptual abnormality whereby persons display heightened responsiveness to varied sensory stimulation, including painful stimulation. Whereas increased pressure-pain sensation is the hallmark of FM, research has shown that persons with FM also display greater sensitivity to other stimuli, such as auditory tones,7-10 paired tactile and auditory stimuli,11 innocuous heat,12 and electrical stimulation.13 These studies suggest that FM could be characterized by a decrease in inhibition and/or an increase in facilitation of activity in neural systems controlling sensory responsiveness. This spectrum of illness might be best characterized as a condition of generalized sensory amplification of bodily experiences that includes, but is not limited to, complaints of pain.
Depression is a candidate factor that may be responsible for sensory amplification in FM. Multiple comorbid physical symptoms are common among persons with major depression, and it has been suggested that the heightened awareness of bodily sensations that accompanies depression explains some of the perceptual abnormalities associated with chronic pain.14 However, studies in patients with FM suggest that depression and somatization function independently of sensory augmentation.11 This independence of pain processing and depressive symptoms or depression has also been noted in a functional-MRI study in FM showing that the presence of depressive comorbidity did not influence the degree of pain-evoked neuronal activation in regions of the brain that code for the sensory dimension of pain.15 Sayar et al.16 reported that depression and sensory amplification both decreased in subjects with major depression after treatment. However, in patients with FM, only depression, but not sensory amplification, decreased after treatment. The authors concluded that sensory amplification has a more enduring association with FM, such that relief from depression does not alter this relationship.
Wolfe and colleagues17 suggested that all patients with rheumatic diseases have variable levels of sensory amplification.These authors examined comorbid conditions in 1,298 patients with FM, as compared with 2,396 patients with rheumatoid arthritis or osteoarthritis. FM patients reported significantly more comorbid medical conditions (e.g., depression, infections, diverticulitis, and allergies) than did patients with rheumatoid arthritis or osteoarthritis. Also, FM subjects rated these comorbid symptoms as being important or having a serious impact on their health. Although FM subjects reported the greatest number of comorbid conditions, individuals in all groups reported a high number of comorbid conditions, and this measure was continuously distributed over the entire large combined cohort. Given the large number of comorbid conditions observed in FM and CFS, it is likely that symptoms other than pain and fatigue contribute to the very low levels of physical functioning observed in these populations.18-20
The present study is the first that we are aware of to develop an experimental measure of sensory sensitivity and then compare this measure to clinical assessment of somatic symptoms, self-reported functional status, and depressive symptoms. To explore these relationships simultaneously, we constructed a path-analytic model utilizing multiple-regression analysis to calculate the path coefficients.
Participants were 38 subjects who either met the 1990 American College of Rheumatology criteria for FM2 or who fulfilled the diagnostic criteria for CFS.1 The research was approved by the Georgetown University Medical Center and University of Michigan institutional review boards, and informed consent was obtained from all participants before their participation. All patients underwent a comprehensive screening exam during which the diagnosis was confirmed and the presence of comorbid symptoms was evaluated. Exclusion criteria were severe physical impairment, medical conditions that were capable of causing patients' symptoms (e.g., morbid obesity, autoimmune/ inflammatory diseases, cardiopulmonary disorders), uncontrolled endocrine or allergic disorders (i.e., hyper-/hypothyroidism, diabetes, allergic rhinitis), malignancy, severe psychiatric illnesses (e.g., schizophrenia, addiction disorders), factors known to affect the hypothalamic-pituitary axis or autonomic functioning (e.g., cigarette smoking, daily intake of caffeine exceeding the equivalent of 2 cups of coffee), or medication usage other than as needed analgesics (excluding long-term narcotics).
Pain Pain ratings were measured with the short form of the McGill Pain Questionnaire (MPQ-SF),21 which measures the quality of pain by asking patients to rate their pain relative to the quality implied by 15 verbal descriptors of pain on a 0-to-3 rating scale. The Pain Rating Index (PRI) comprises two scores: 1) a sensory pain score; and 2) an affective pain score. In this study, the PRI is used as the measure of pain intensity.
Depression The Center for Epidemiological Studies-Depression Scale (CES-D)22 was used to assess self-report of depressive symptoms. The measure is a 20-item questionnaire in which patients rate the frequency of depressive symptoms on a 0-to-3 scale relative to how they felt during the past week. A total score is obtained by summing all responses, with higher scores reflecting greater depressive symptoms. The scale is reported to be valid among persons with physical disabilities23 and has good sensitivity and specificity in chronic-pain populations.24,25
Fatigue Fatigue was assessed with the Multidimensional Fatigue Inventory (MFI).26 The MFI consists of 20 items that can be scored to yield five scale dimensions: General Fatigue, Physical Fatigue, Mental Fatigue, Reduced Motivation, and Reduced Activity. This inventory has been validated in samples of cancer patients, medical students, army recruits, and junior physicians. For the purpose of this study, the General Fatigue scale was used.
Physical Symptoms The number of comorbid physical symptoms was obtained from a standardized symptom checklist (SxList). This self-report checklist instructs participants to indicate whether or not they experienced each of 51 symptoms for at least 3 months over the past year. A total score was obtained by summing Yes responses to all 51 symptoms. Sample symptoms include dry eyes, shortness of breath, dizziness, irregular heartbeat, tingling in the extremities, urinary urgency, and coughing spells. Three items, assessing muscle pain, persistent fatigue, and extreme fatigue, were excluded from the total score because these items overlapped with the hallmark symptoms and could not be considered to be comorbid. Reliability (internal consistency) of this scale with the three items eliminated in the present sample was found to be 0.91.
Physical Functional Status The Medical Outcomes Survey Short Form-36 (SF-36) is a self-administered patient-reported outcome (PRO) measure of health-related quality of life.27,28 The eight domains of health-related quality of life assessed by the SF-36 are the following: 1) physical functioning; 2) role limitations because of physical problems; 3) bodily pain; 4) general health perceptions; 5) energy/vitality; 6) social functioning; 7) role limitations due to emotional problems; and 8) mental health. Also, two summary scores: Physical Component (PCS) and Mental Component (MCS) are derived by combining and positively or negatively weighting all eight domains. Scores for all domain and summary measures are transformed to z-scores ranging between 0 and 100, with lower scores indicative of poorer functioning. In this study, only the PCS score was examined.
Measures of Sensory Amplification
Pressure Stimulation Previous studies have shown that "neutral" regions, such as the thumb, in individuals with FM, accurately reflect an individual's overall pressure-pain sensitivity.29 Pressure-pain sensitivity was evaluated by subjective scaling of sensations evoked by discrete 5-second pressure stimuli applied to the fixed thumbnail with a 1-cm^2 hard rubber probe, using the Multiple Random Staircase (MRS) method.30 The delivery of the MRS is driven by an interactive software system that presents stimuli in a random manner, facilitating the identification of stimulusintensity values associated with subjective ratings between 0 and 1 (no sensation-to-faint pain), between 9 and 10 (mild-to- moderate pain), and between 13 and 14 (strong-to-slightly intense pain) on the 0-20 box scale.(31) For this study, stimulus values eliciting mild-to- moderate pain were used. Ratings of Perceived Exertion Subjects underwent a sub-maximal exercise test on an electronically-braked cycle ergometer (Sensormedics; Yorba Linda, CA). The test was graded, with 3-minute stages culminating when the subjects' heart rate reached 85% of their age-predicted maximum. After 1 minute of unloaded pedaling, Stage 1 was set at either 25 or 50 watts (W). For Stage 2, depending upon the subject's response during the first stage, the workload was increased by either 25 or 50 W. Both objective and subjective responses were taken into consideration. Subsequent increases in workload came in 25-W increments every 3 minutes. Ratings of perceived exertion were recorded once per stage with the Borg 6-20 scale.32 Borg scores range from 6 (very, very light) to 20 (very, very heavy). An amplification measure for exercise was obtained by dividing the rating of perceived exertion by the workload (watts) during the first stage of exercise. Using this measure, a previous study found that subjects with FM reported consistently greater perceived exertion per workload during exercise than did healthy-control subjects.33
Index of Sensory Amplification We constructed an index of sensory amplification (ISA) by determining the z-score from the sample mean and standard deviation (SD) for each person for the pressure-intensity (kg) needed to evoke mild-to-moderate pain, and the z-score for the amplification measure for exercise (perceived exertion/W). Higher ratings on the exercise amplification measure represented greater augmentation, and augmentation of pressure sensitivity corresponded with lower scores on this measure. To facilitate the creation of a composite index, scores on the pressure-sensitivity measure were multiplied by -1, making the scaling and directionality of both measures equivalent. Finally, the z-scores were averaged to calculate a composite index. Thus, higher (more positive) scores on this measure are associated with greater sensory amplification.
Statistical analyses required three steps. First, zero-order (Pearson) correlations were calculated in order to determine the strength of association between the ISA and the hallmark symptoms of FM and CFS (i.e., pain and fatigue), comorbid physical concerns, and depression. Similarly, correlations were also calculated with the PCS score from the SF-36 to assess the relative strength of association between each of the previously-mentioned variables and physical functional status.
Second, correlations between the pressure-sensitivity measure, perceived exertion measure, and ISA were examined to determine whether the patterns of correlations between the individual and composite measures were similar. This was done to examine whether the individual measures of sensory augmentation were both related to the variables of interest, rather than one individual scale being responsible for any observed associations with the composite measure.
The third step involved the construction of a path-analytic model utilizing multiple-regression techniques to examine the proposed interrelationships between the defined variables when the other variables were considered simultaneously.34 Standardized regression coefficients were calculated to determine the path coefficients, with physical-functional status serving as the dependent outcome measure. Contrary to our hypothesis, if the zero-order correlations revealed a significant association between depression and the ISA, depression would be entered into the model first, followed by ISA. If not, assuming the ISA to be a common mechanism in the erosion of physical functioning, ISA was entered first into the model, followed by hallmark symptoms and the number of comorbidities serving as mediating variables. The path coefficient between the ISA and each of the mediators was calculated as the partial regression coefficient between ISA and each of the symptom-measures controlling for the other symptommeasures. The path coefficient between each of the mediators and physical functioning was calculated by determining the standardized regression coefficient between the mediator and PCS, simultaneously controlling for the other mediators and sensory amplification.
Finally, the direct relationship between the ISA and physical functioning was calculated by determining the standardized regression coefficient between these two variables, controlling for all of the mediators. Given the colinearity between some of the variables and the conservative nature of this analysis, the criterion for statistical significance was set at p<0.10.
Table 1 lists the demographic and descriptive statistics for each of the measures used in this study. Eight subjects fulfilled only the diagnostic criteria for FM; 7, only the criteria for CFS; and 23 met criteria for both; 28 subjects were women, and 10 were men. The average age of participants was 42.0 years (standard deviation [SD]: 8.8), and the mean duration of pain was 87.9 months (SD: 64.4). The demographic characteristics of the sample are similar to those of previously published samples in both the FM and CFS literature,17,35 although subjects in the present study were younger. Like other samples, over 90% complained of both pain and fatigue, and the mean number of items endorsed on the symptom checklist (SxList) was 19.2 (SD: 9.7).
Two subjects in the sample did not complete the sub-maximal stress test, and therefore an ISA could not be determined for them. In these two cases, imputation was facilitated by using the sample mean for the ISA and perceived exertion per watt.
The correlations between the individual and composite measures of sensory amplification are presented in Table 2. The individual amplification scales (i.e., pressure and exertion) were significantly correlated with each other (r=0.43; p<0.01), and both pressure sensitivity and perceived exertion were highly correlated with the composite score (r=0.85; p<0.01 and r=0.84; p<0.01, respectively). Both individual measures were significantly cor- related with pain. Each showed an opposite, but nonsignificant relationship to fatigue. Both measures were inversely correlated with functioning, but only the relationship with perceived exertion was statistically significant. Neither measure was significantly related to depression. Both indexes were positively correlated with comorbid symptoms (the correlation with perceived exertion approached significance, at p<0.07). The pattern of correlations suggests that the individual measures of sensory amplification each contribute to relationships observed with the composite index.
The zero-order correlations are presented in Table 3. Greater sensory sensitivity to evoked stimuli, as defined by the ISA, was significantly associated with greater pain intensity, poorer physical functioning, and a greater number of comorbid physical symptoms. Depressive symptoms were not significantly associated with the ISA or the other variables of interest, except that greater depressive symptoms were significantly associated with greater fatigue. The ISA did not show a significant relationship with fatigue. Pain was significantly associated with the number of comorbid secondary symptoms, but fatigue showed weak relationships with both pain and number of comorbidities. Self-reported physical functional status was significantly associated with pain, fatigue, and the number of comorbid somatic symptoms.
The path model is presented in Figure 1. Because depression was not significantly associated with the ISA, this variable was eliminated from the model. Together, the remaining four predictor variables accounted for 50% of the variance in physical functioning, and the multiple-regression coefficient was statistically significant (F=8.37; p<0.001). The path coefficient between the ISA and physical functioning, controlling for the mediators (specific symptoms) in the model, was not statistically significant, suggesting that this variable only weakly contributes directly to physical functional status. However, the indirect pathway, through pain intensity as a mediating variable, was significant; the ISA was significantly associated with pain intensity, which, in turn, was significantly associated with physical functional status (controlling for sensory amplification and the other mediating variables in this latter pathway). The indirect pathway through physical symptoms was significant as well, using the same method of calculation. Greater fatigue was significantly associated with poorer physical functioning, but the pathway between the ISA and fatigue was not statistically significant, suggesting that the influence of fatigue on physical functional status may be due to a different mechanism.
DISCUSSION Sensory amplification, as measured by a composite measure of pressure and perceived exertion during exercise, was significantly associated with higher clinical pain and a greater number of comorbid somatic symptoms. As hypothesized, this measure was not significantly associated with depressive symptoms. Contrary to our hypothesis, sensory amplification did not appear to have a strong association with fatigue, which suggests that fatigue may operate via a different underlying mechanism. Also, the path-analytic model suggested that sensory amplification only weakly independently affected physical functioning (when controlling for physical symptoms), but influenced outcomes through the origination or maintenance of multiple physical symptoms. These findings support the hypothesis that physical symptoms other than the hallmark symptoms of pain and fatigue significantly contribute to the poor physical functioning in FM and CFS. It is likely that a tendency toward sensory amplification would contribute to symptoms across many systems of the body where sensory perception is involved. It also makes sense that fatigue, which is not directly tied to specific sensory mechanisms, would show a poorer relationship with this index.
The findings also lend further support to the notion that sensory amplification in subjects with FM is not due to somatization associated with depression. Several other studies have reported similar findings.7,11,16,36 Instead, these data support the hypothesis that FM is characterized by altered central sensory processing mechanisms that are not limited to pain. Recent neuroimaging studies suggest that subjects with FM have heightened brain responses to evoked pain sensations.37,38 The insula is consistently activated in these pain studies. Various regions of the insula are believed to be involved in an interoceptive system that is activated in response to sensations arising from within the body; these include pain, temperature, itch, sensual touch, hunger, thirst, and muscular and visceral sensations.39,40 It is possible that dysfunction of this interceptive system, involving the insula, may be involved in the sensory amplification observed in FM. It would be interesting to determine whether subjects with FM display heightened responses in the insula in response to non-painful stimuli (such as auditory tones) as compared with healthy-control subjects.
The lack of association between sensory amplification and fatigue, the hallmark symptom of CFS, raises several possibilities. Although many of the comorbidities are shared between FM and CFS, making them appear to be similar, it is possible that the hallmark symptoms emanate from distinctly different mechanisms.41 It is also possible that the lack of relationship between sensory amplification and fatigue is in part due to the chronicity of symptoms in this sample. Subjects had an average symptom duration of approximately 7 years. Although fatigue may initially be associated with sensory amplification, much like the other comorbid symptoms, over time, fatigue may become more influenced by other factors, such as inactivity and deconditioning. Future research may need to assess the relationship between sensory amplification and fatigue with newly-diagnosed conditions in order to gain insight into this relationship.
Supporting the confidence one can have in theses findings is the observation that statistically significant and meaningful relationships emerged from these data with a relatively small sample size. A potential limitation of the sample size is that other, weaker relationships requiring greater power to detect were missed. A second potential limitation was that the path-analytic model was fairly conservative, since the effects of any particular variable were examined while controlling for the influence of all other variables in the model. Although a conservative approach is appropriate for exploratory purposes, future studies, building upon these results, may be enlightened by replicating these findings using less conservative methods.
In summary, the findings of this study support the hypothesis that many symptoms in FM and CFS are associated with perceptual abnormalities (i.e., sensory amplification), and this deficit may be a common mechanism underlying these disorders. Future studies will need to replicate these findings in a larger sample and examine other sensory modalities such as auditory, olfactory, and visual stimuli. Treatment implications of these findings suggest that management strategies may want to focus on dampening sensory amplification, in general; rather than treating each symptom in isolation.