|Home | About | Journals | Submit | Contact Us | Français|
This was a pilot, cross-sectional study. Its site was West China Hospital in Chengdu, Sichuan Province.
An objective was to explore whether, in China, LBP characteristically is one symptom among co-occurring subjective symptoms. More basic objectives were to test a supplemented list of symptoms and to reconfigure findings from the literature on co-occurrence of symptoms so that they pertain specifically to cLBP. The governing metaphor was a constellation of symptoms in which cLBP is located.
With the exception of small, isolated societies, previous studies of co-occurrence of symptoms were conducted in the affluent West. Although China's population is larger than the combined populations of affluent countries of the West, research on co-occurring symptoms has been neglected in China. Unknown is whether results from studies of co-occurring symptoms in the affluent West may be extended to China.
A survey with the supplemented symptom list was cross-culturally adapted into Chinese and administered to cLBP patients (N=72) and normal controls (N=102). Multiple regression analysis was used to determine the effects of covariates (age, gender, education) on symptom reporting.
cLBP patients reported higher median numbers of symptoms than normal controls, including total symptoms (9 vs. 3), musculoskeletal symptoms (4 vs. 1), and non-musculoskeletal symptoms (6 vs. 2.5); differences between cLBP and normal controls were highly significant (p's <0.001). Covariates had little effect on symptom reporting.
cLBP characteristically was one symptom in a constellation of symptoms. This finding came from one research site in China. Nevertheless, consistency between studies is notable, with findings from the affluent West supporting the finding from the Chinese site. The tendency to concentrate on a discrete clinical entity, cLBP itself, may obscure the constellation of symptoms. The more expansive view of cLBP has implications for clinical practice and research.
Especially in a short clinic visit, patients may present chronic low back pain (cLBP) as though it were a discrete clinical entity and physicians may diagnosis it as such. Increasingly in the literature, however, cLBP is seen in the context of other subjective symptoms with which it co-occurs. The title of an article expresses co-occurrence of this nature, “Chronic pain syndromes: You can't have one without another”.1 However, with the exception of small, isolated societies,2,3 studies of co-occurrence of cLBP and other symptoms have been conducted in affluent countries of the West, notably Norway.4-8 The population of China is 1.4 billion, which is larger than the combined populations of affluent (“high income”) countries of the West.9
In view of the lack of studies of co-occurring symptoms in China, the question is whether findings from the studies conducted in the affluent West may be extended to China. An editorial in Nature sums up decades of research in the behavioral sciences and cautions against such an extension of findings.10 In the words of the editorial, research articles from the affluent West “routinely assume that their results are broadly representative…[but] a growing body of evidence suggests that this is not the case.”
Why might subjective symptoms in China and the affluent West differ? China has a distinct socio-cultural context, which exerts an influence on experiences of subjective symptoms. It has had a tumultuous recent history, from the Cultural Revolution (to go back no further) to the vast rural-to-urban migration now under way, and the attendant effects on individuals have been profound. As scholars of China have documented, these effects may lead not only to depression but also may be somaticized and manifested as such symptoms as dizziness, sleep disturbances, and chronic pain.11-13 China has also seen a rapid economic rise, which has resulted in income disparities, and, paradoxically, an increase in the symptom of sadness.14
More generally, there are well-documented discrepancies in cross-national rates of subjective symptoms.15-17 For example, among nurses, the prevalence of disabling low back pain is four times higher in Costa Rica and Nicaragua than in Pakistan and Japan.18 Rates of “cough” and “”wheezing” symptoms similarly vary multifold in cross-national comparisons.19-21 In short, distinct socio-cultural circumstances of China in conjunction with pervasive cross-national disparities in rates suggest that patterns of subjective symptoms seen in the affluent West do not necessarily also occur in China. An objective of this study, then, was to initiate research on cLBP and co-occurring symptoms specifically in China.
A more basic objective was to reconfigure findings from the literature on subjective symptoms so that they pertain to cLBP. Overlapping symptoms is a common finding in the literature.22-25 In contrast with a Venn diagram with multiple overlapping symptoms, the governing metaphor here is a constellation of symptoms in which cLBP is located. To draw out the metaphor further, single points in the constellation may variously appear at any one moment in time (in other words, cLBP patients may variously report other symptoms besides cLBP), but, aside from cLBP, no one point in the constellation is essential to discern its overall contours. Seen as one symptom within a constellation, critical issues in the treatment of cLBP come into view.
Although the term “subjective health symptoms” is used here, the underlying meaning of a number of other terms is similar, among them “medically unexplained symptoms”,22 “functional somatic syndromes”,23 and “chronic multi-symptom syndromes”.24 These terms have common core features. Each of them refers to subjective experiences obviously real to those who undergo them, often to the point of debility; on the other hand, researchers have striven to substantiate the biomedical substrate of such experiences but so far have largely come up empty-handed.26 So as not be detained further by a terminological issue, a simple, operational definition will be provided for this paper: subjective symptoms comprise those listed in the Table 1.
Of the 32 symptoms in the Table 1, 28 come from the Subjective Health Complaints Inventory.27 The four symptoms added to the list are asterisked in Table 1, and the rationales for adding them are in Table 2.
The questionnaire with the supplemented symptom list went through a process of cross-cultural adaptation, with two forward translations (English to Chinese), two backward translations (Chinese to English), and then two more forward translations (Chinese to English).28 Each translator, involved in only one step in the process, worked independently of the others. After each step, a committee resolved discrepancies between translations. For the Chinese version of subjective health symptoms used for this study, see Supplementary Data.
The research took place at West China Hospital, which is located in Chengdu, a city in Sichuan Province. West China Hospital, one of the largest hospitals in the world,29 is a referral hospital that draws patients from throughout southwest China. The Institutional Review Board of West China Hospital approved of the study.
Patients from the hospital's western-style Pain Clinic were recruited for the study. The study enrolled consecutive patients that fit selection criteria listed in Table 3. Excluded from the study were patients with an identifiable pathophysiology, such as back fracture or sciatica. Included were patients with non-specific low back pain, which is prominently classified as a “functional somatic syndrome”23,25 due to its unknown or uncertain etiology.30
Normal controls came from the eyeglasses clinic of West China Hospital. Those seeking eyeglasses themselves were not recruited for the study, since eyestrain may be a subjective symptom. Rather, people between the ages of 20 and 64 who accompanied those seeking eyeglasses, usually a family member, were recruited as normal controls.
The constellation of symptoms in which cLBP is located came into view through two contrasts:
The distribution of symptoms listed in Table 1 was skewed in the normal control population, and for this reason the Wilcoxon rank sum test was used to compare numbers of symptoms in the normal control population with numbers of symptoms in the cLBP population. Ensuing analyses in this paper pertained to the cLBP population only. Pearson correlation was used to test the association between musculoskeletal symptoms and non-musculoskeletal symptoms. In order to determine the effect of age, gender, and education on symptom reporting, multiple regression analysis was performed.
The study recruited 72 cLBP patients and 102 normal controls. Prevalence of their subjective symptoms in the past month is shown in Table 1. For each symptom, the questionnaire in this study (as in the Subjective Health Complaints Inventory27) provided four response categories (“not at all”, “a little”, “some” and “serious”). The report of a “serious” symptom was rare. Thus, in Table 1, this study followed the precedent of previous studies,4,5 and collapsed the categories of “serious” and “some” into the single category of “substantial”. Aside from data in Table 1, data presented in this paper consist of reports of “any” of the designated symptoms, which aggregates all three categories of symptom intensity (“a little”, “some” and “substantial”). Thus, Tables 4 and and55 as well as Figure 1 present data on any experience of the designated symptoms in the past month, regardless of the intensity of the experience.
As shown in Table 4, cLBP patients reported a median of 9 total symptoms vs. 3 for normal controls; a median of 4 musculoskeletal symptoms vs. 1 for normal controls; and a median of 6 non-musculoskeletal symptoms vs. 2.5 for normal controls (all ps<0.001).
Table 4 also shows demographic characteristics of the two populations. Education was significantly higher in the normal control population than the cLBP population. The percentage of women was higher in the cLBP population, although this difference was only marginally significant.
Figure 1 depicts the relationship between musculoskeletal and non-musculoskeletal symptoms in the cLBP population. These subcategories of symptoms were moderately related (Pearson r=0.38, p=0.001), with one subcategory accounting for little of the variation in the other (R2 = 0.14).
In regression analyses, the effects of age, gender, and education were tested, successively, on total symptoms and the two subcategories that compose it, i.e., musculoskeletal symptoms and non-musculoskeletal symptoms. Results are shown in Table 5. None of the covariates had a significant effect on musculoskeletal symptoms. One covariate, gender, had a marginally significant effect on, respectively, total symptoms and non-musculoskeletal symptoms; thus, men had about 2½ fewer total symptoms than women (p=0.07) and about 2 fewer non-musculoskeletal symptoms than women (p=0.08). Covariates accounted for only a minor amount of variation in all three analyses (R2 adj <10%)
In China, a population of cLBP patients endorsed significantly higher median numbers of subjective symptoms than a population of normal controls, including total symptoms (9 vs. 3), musculoskeletal symptoms (4 vs. 1), and non-musculoskeletal symptoms (6 vs. 2.5). The finding of cLBP as one symptom in a constellation of symptoms may thus be extended from smaller countries of the affluent West to a site in the most populous country in the world.
In delineating the constellation, the question was whether it consisted of a single strand composed of the two subcategories of symptoms, i.e., musculoskeletal symptoms and non-musculoskeletal symptoms, with the number of one subcategory of symptom increasing linearly with the number of the other. There was a moderately significant relationship between the subcategories of symptoms, although one subcategory accounted for little variation in the other (Figure 1). In other words, musculoskeletal symptoms may not be used s a proxy for non-musculoskeletal symptoms. Rather, in order to view the constellation in its entirety, data on symptoms in both subcategories are necessary.
Four symptoms were added to the Subjective Health Complaints Inventory, i.e., temporomandibular joint disorder, pain in the hips or thighs, and knee pain (Table 2). The added symptoms led to noteworthy results (Table 1). For example, more than one-third of the Chinese cLBP patients (36%) reported “any” pain in the hips or thighs, and for almost one-fourth of them (24%) this pain was “substantial”. All added symptoms were at least two-and-a-half times more prevalent in cLBP patients than in normal controls (chi-square test ps <0.01). The supplemented symptom list presented here was not intended to be definitive; rather, its purpose was to collect symptoms commonly occurring with cLBP in one place. In this way, symptoms may be simultaneously inspected and the list may be further revised and supplemented.
This paper concentrated on the delineation of the constellation of co-occurring symptoms, which is to be distinguished from an underlying explanation of the constellation. The multivariate analysis tested the effect of covariates (age, gender, and education), but this only constituted an initial exploration and did little to explain the emergence of the constellation. There is a substantial literature on central sensitization, a form of neuroplasticity, which has been advanced as the explanation of co-occurring symptoms.32,33 The consistency of findings cross-culturally, i.e., in the affluent West and now also in China, suggests central sensitization may be implicated in the co-occurrence of symptoms. Research in this paper, however, was not designed as a direct test of such a mechanism.
Aside from the explanation of why co-occurring symptoms emerge in the first place, a separate question is what is the prognosis of patients once they have already developed cLBP and co-occurring symptoms. Fragmentary evidence from previous studies indicates that the presence of co-occurring symptoms adversely affects the outcomes of back pain treatment.7,34 This raises the further question of what, more precisely, is the relationship between the number of subjective symptoms and the outcomes of back pain treatment. Is this relationship monotonic, i.e., as the number of symptoms increases, do treatment outcomes systematically worsen? Is there a threshold effect, i.e., is there a number of subjective symptoms beyond which treatment outcomes do not improve? A comprehensive list of subjective symptoms is necessary to test the relationship between subjective symptoms and treatment outcomes. Furthermore, the treatment of cLBP patients with multiple symptoms may result in the “balloon effect”; at follow-ups, low back pain itself may subside, but other symptoms may emerge or increase in intensity. A test for this effect again presupposes a comprehensive list of commonly co-occurring symptoms.
A study conducted in Norway also found that cLBP patients endorsed significantly more subjective symptoms than normal controls.4 The Chinese population of cLBP patients studied here, however, endorsed almost twice as many symptoms as the Norwegian cLBP patients (a median of 9 vs. 5 for Norwegian cLBP patients). This suggests another question that may be investigated in ensuing research: In the general population as well as cLBP population, is the burden of subjective symptoms greater in China than in the affluent West? If so, still another matter to investigate is whether socio-cultural circumstances in China, discussed earlier in the paper, are implicated.
This was a pilot study. Data for it came from one research site, West China Hospital, which may not be representative of other sites in China. On the other hand, regardless of where research on co-occurring symptoms has been conducted, consistency between studies is notable, with findings from the affluent West supporting findings from the Chinese research site studied here. Additionally, the effects of covariates (age, sex, education) on symptom reporting were only marginally significant or insignificant in this study, which may be due to relatively small population sizes (N's were 72 and 102 for, respectively, cLBP patients and normal controls). Studies with larger population sizes are required to clarify the effects of covariates. Larger population sizes, however, would have been unlikely to affect main results of this study. In other words, differences between numbers of symptoms reported by normal controls and numbers reported by cLBP patients were all highly significant (p values <0.001).
To conclude, cLBP of the nature discussed here is far from “uncomplicated”. Its complexities, however, may be overlooked in the clinic. Physicians, including those at West China Hospital, are trained in the process of “differential diagnosis”, which entails successive differentiations among diagnoses until one or a few discrete ones are reached.35 As opposed to casting a wide net that may pick up a multiplicity of subjective symptoms, differential diagnosis consists of a process of narrowing down. Treatment guidelines for low back pain hardly raise the matter of co-occurrence of symptoms let alone a systematic survey of them,36 and so they do little to offset this process. The ensuing tendency to concentrate on a discrete clinical entity, cLBP itself, may obscure the constellation of co-occurring symptoms. On the other hand, once cLBP is seen in a constellation, largely unexamined questions emerge, such as:
Clearly, the more expansive view of cLBP has implications that are consequential not only in research but ultimately also in the clinic.
The manuscript submitted does not contain information about medical device(s)/drug(s). Support for this research came from a Fulbright grant awarded to Ernest Volinn (Grant ID 48408578) and from a China Medical Board grant awarded to West China Hospital of Sichuan University. Two co-authors, Drs. Xiaoming Sheng and Jing Ying, are affiliated with the University of Utah, Study Design and Biostatistics Center, with funding in part from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant 8UL1TR000105 (formerly UL1RR025764).
No relevant financial activities outside the submitted work.
Level of Evidence: N/A
Ernest Volinn, University of Utah, Departments of Anesthesiology and Sociology.
Bangxiang Yang, Pain Management Department, West China Hospital, Sichuan University, Chengdu, PRC.
Nan Chen, Department of Medical Statistics, West China School of Public Health, Sichuan University, Chengdu, PRC.
Jian Ying, Dept. of Internal Medicine, University of Utah, Salt Lake City, UT, USA.
Jing Lin, Dept. of Anesthesiology, West China Hospital, Sichuan University, Chengdu, PRC.
Xiaoming Sheng, Dept. of Pediatrics, University of Utah, Salt Lake City, UT, USA.
Yunxia Zuo, Dept. of Anesthesiology, West China Hospital, Sichuan University, Chengdu, PRC.