Background
Central sensitization is a pathophysiological phenomenon characterized by increased sensitivity of nociceptive neurons of central nervous system (CNS) to normal or subthreshold stimulation [1]. An increased excitability and reactivity of central nociceptive neurons in response to damaging stimuli can be recorded using electrophysiological methods in special studies. In clinical practice, central sensitization is observed in patients with dysfunctional pain syndromes (migraine, fibromyalgia, irritable bowel syndrome), neuropathic and chronic pain. Chronic pain is currently considered as an independent disease requiring complex treatment [2].
The pathophysiological term "central sensitization" is increasingly used in clinical studies. This emphasizes the relationship between knowledge about mechanisms of hyperexcitability of nociceptive brain structures with clinical manifestations. The last ones largely reflect increased sensitivity to various loads [3—5]. Typically, patients with chronic pain syndromes report fatigue, mood and sleep disorders, and poor quality of life. High degree of comorbidity of these symptoms with pain and difficult identification of "source of pain" complicate management of these patients.
There is no "gold standard" for clinical diagnosis of central sensitization although clinical symptoms of central sensitization (hyperalgesia, allodynia) are described. This means that clinicians have no criteria for identifying the symptoms following central sensitization in patients, for example, with chronic non-specific back pain. Therefore, standard treatment may not be effective enough. Central sensitization inventory was developed and introduced into clinical practice to improve the quality of diagnosis in patients with pain syndromes [6, 7].
Non-specific neck and back pain is a particular case of musculoskeletal pain and one of 10 major causes impairing quality of life of people in the world and in our country [8]. Pain properties (intensity, quality, etiology, and pathogenesis) in patients with non-specific musculoskeletal neck and back pain are similar. Thus, these patients may be combined into a sufficiently homogeneous group to study the psychometric properties of the Russian language version of the Central sensitization inventory [9].
The purpose of the study was to evaluate validity and reliability of the Russian language version of the Central sensitization inventory (CSI-RU) [9] in patients with chronic non-specific neck and back pain (cervicalgia, thoracalgia and lumbodynia).
Material and methods
Inclusion criteria: chronic (> 3 months) non-specific neck (cervicalgia, cervicocranialgia, cervicobrachialgia) and back (thoracalgia, lumbodynia, lumboischialgia) pain.
Exclusion criteria: neck and back pain accompanied by symptoms of radiculopathy or neuropathy; neck or back pain caused by specific processes — rheumatoid arthritis, tumor, fractures, etc.
Questionnaire for assessment of central sensitization
The questionnaire comprises 2 parts (A and B). Part A includes 25 statements devoted to violation of various general (sleep) and specific (concentration, memory, emotions, etc.) functions of various systems (nervous, musculoskeletal, urogenital, etc.).
These statements make up 25 items of the questionnaire. The patient is asked to answer each of the statements using the adverbs “never”, “rarely”, “sometimes”, “often” and “always”. Arranged in a specific order, these adverbs form a 5-point verbal rating scale (VRS). Each of these adverbs is assigned a corresponding point from 0 to 4 (“never” — 0 points, “always” — 4 points). The points for each item are summarized and the total score is calculated. The results are interpreted as follows: subclinical (0–20 points), mild (30–39), moderate (40–49), strong (50–59), extreme central sensitization (60–100) [10].
Part B is for physicians. This part is not taken into account during assessment of psychometric properties of central sensitization.
Internal consistency
To assess internal consistency, we calculated Cronbach's α and inter-item correlation for the entire questionnaire and for each factor identified in factor analysis. Cronbach's α values were interpreted as follows: 0.6 <α <0.7 — doubtful value, 0.7 <α <0.8 — sufficient, 0.8 <α <0.9 — good, α> 0.9 — very good. Optimal inter-item correlation does not exceed 0.5 (optimal — 0.35, satisfactory range — 0.2—0.5).
Test-retest reliability
To assess test-retest reliability, we analyzed intraclass correlation coefficient (ICC) with 95% confidence interval (95% CI). Two-sided variance analysis was chosen as a model, type ICC2.1 — as a type of intraclass correlation coefficient, ICC2.1 for “absolute agreement” — as a definition for ICC2.1. ICC2.1 value was interpreted as follows: moderate (0.50—0.77), good (0.75—0.90), excellent correlation (> 0.90) [11, 12]. The interval between the 1st and 2nd survey was 2 — 5 days. To assess test — retest reliability, we considered parameters of only those patients whose condition did not change throughout this interval.
Factor analysis
Factor analysis was carried out by distinguishing of the main components with eigenvalues > 1. We used varimax rotation to maximize the variance of initial variables. Numerical values of factor loadings were understood as Pearson correlation coefficients between the statements of the questionnaire and the identified factors. Factor loading was interpreted as follows: weak (0.32—0.44), moderate (0.45—0.54), good (0.55—0.62), very good (0.63—0.70) and excellent load (> 0.71). Factor load of 0.45 was accepted as cut-off value [11, 13].
Minimal detectable changes
Minimal detectable changes (MDC) were calculated using the formula MDC = SEM × √2 × 1.96, where SEM — standard error of measurement (SEM), 1.96 — coefficient for 95% confidence interval (since 95% of sample is within 1.96 standard deviations (SD) of the mean) [6]. Standard error of measurement is an index of reliability indicating variability of values obtained from measurement to measurement [7]. SEM was calculated using the formula SEM = Sd / √2, where Sd — standard deviation of CSI difference scores obtained by test-retest method.
Floor and Ceiling Effects
To assess the “floor and ceiling effects”, we analyzed the percentage of patients who indicated the minimum and maximum value for each statement of the questionnaire. In case of Central sensitization inventory, these are the adverbs "never" and "always". “Floor and ceiling effects” are present if over 15% of respondents choose the answer with minimum (“floor effect”) or maximum (“ceiling effect”) value [11].
Convergent validity
Convergent validity was determined by degree of “convergence” of sensitization grade according to CSI and measurements of disability grade according to 2 other questionnaires — the Oswestry questionnaire (ODI-RU) and the Neck Disability Index (NDI-RU). Convergent validity of the Russian-language version of the CSI was assessed using Pearson correlation coefficient r (at p-value <0.05). Pearson's r values were interpreted as follows: high correlation at r≥0.5, moderate correlation at r = 0.3—0.5, weak correlation at r <0.3 [14].
Results
The study involved 195 patients (142 women and 53 men) aged 18 — 65 years (mean 40.5 ± 11.4) with chronic nonspecific neck and/or back pain. The main diagnosis was cervicalgia (including cervicocranialgia and cervicobrachialgia) in 106 (54.4%) patients, lumbodynia (including lumboischialgia) in 75 (39.5%) patients, thoracalgia in 12 (6.1%) patients. In about 40% of patients, cervicalgia was combined with thoracalgia and/or lumbodynia.
The subgroup of cervicocranialgia included patients with migraine, tension headache and cervicogenic headache. However, percentages of patients with these types of headache were not measured.
There were 122 (62.6%) patients with higher education, 7 (4.4%) ones with incomplete higher education, 66 (33.9%) patients with secondary or specialized secondary education. The sample included 145 (74.4%) working and 50 (25.6%) non-working people. The last ones consisted of 33 (66%) housewives, 10 (20%) pensioners and 7 (14%) students.
Overall CSI score was 36 ± 14 points. Distribution of scores was normal (χ2 test = 1.00 with 4 degrees of freedom at p-value = 0.91; Shapiro-Wilk test 0.99 at p-value <0.05).
In patients with predominant neck pain, central sensitization was mild (overall CSI score 32 [21; 42]). However, moderate and strong central sensitization was found in 36% of patients (CSI score ≥ 40). In all of these patients, neck pain was accompanied by headache. In patients with lumbodynia and lumboischialgia, central sensitization was mild-to-moderate (overall CSI score 39 [31.75; 42]). However, moderate-to-severe central sensitization was observed in 45% of patients. All of these patients had lumboischialgia.
Reliability — internal consistency
Internal consistency of Central sensitization inventory was good (Cronbach's α = 0.89, inter-item correlation coefficient 0.26). Correlation of each statement with other statements of the questionnaire is presented in Table 1. This correlation was moderate-to-good.
|
CSI statements |
Correlation of the statement with other statements of the questionnaire |
Cronbach's α if statement removed |
|
1. I feel tired and unrefreshed when I wake from sleeping |
0.54 |
0.89 |
|
2. My muscles feel stiff and anchy |
0.52 |
0.89 |
|
3. Anxiety attacks |
0.63 |
0.89 |
|
4. I grind or clench my teeth |
0.42 |
0.89 |
|
5. I have problems with diarrhoea and/or constipation |
0.36 |
0.89 |
|
6. I need help in performing my daily activities |
0.37 |
0.89 |
|
7. I am sensitive to bright lights |
0.47 |
0.89 |
|
8. I get tired very easily when I am physically active |
0.47 |
0.89 |
|
9. I feel pain all over my body |
0.49 |
0.89 |
|
10. I have headaches |
0.45 |
0.89 |
|
11. I have discomfort in my bladder and/or burning when I urinate |
0.35 |
0.89 |
|
12. I do not sleep well |
0.56 |
0.89 |
|
13. I have difficulty concentrating |
0.58 |
0.89 |
|
14. I have skin problems |
0.42 |
0.89 |
|
15. Stress makes my physical symptoms get worse |
0.57 |
0.89 |
|
16. Depression |
0.64 |
0.89 |
|
17. I have low energy |
0.71 |
0.88 |
|
18. I have muscle tension in my neck and shoulders |
0.50 |
0.89 |
|
19. I have pain in my jaw |
0.35 |
0.89 |
|
20. Certain smells make me feel dizzy or nauseated |
0.51 |
0.89 |
|
21. I have to urinate frequently |
0.36 |
0.89 |
|
22. My legs feel uncomfortable |
0.44 |
0.89 |
|
23. I have difficulty remembering things |
0.52 |
0.89 |
|
24. I suffered trauma as a child |
0.30 |
0.89 |
|
25. I have pain in my pelvic area |
0.35 |
0.89 |
Analysis of competitive validity revealed high correlation of CSI with NDI-RU (rS = 0.56 at p-value <0.05) and moderate correlation with ODI-RU (rS = 0.36 at p-value <0.05).
Minimal detectable changes
Minimal detectable changes for CSI were 10 points.
"Floor and ceiling effects" for each statement of the questionnaire
Distribution of answers to each statement of the questionnaire is presented in Table 2. All response categories were obtained for all statements except 11 (urinary discomfort). A large percentage of responses with minimum value were received for many statements that do not correspond to the complaints of patients with non-specific neck and back pain. However, there were no “floor and ceiling effects” for the entire questionnaire as a whole.
|
CSI statements |
Responses (%) |
||||
|
never |
rarely |
sometimes |
often |
always |
|
|
1. I feel tired and unrefreshed when I wake from sleeping |
6.7 |
18.4 |
31.8 |
30.8 |
12.3 |
|
2. My muscles feel stiff and anchy |
6.2 |
12.3 |
32.3 |
38.4 |
10.8 |
|
3. Anxiety attacks |
12.8 |
30.8 |
34.9 |
18.4 |
3.1 |
|
4. I grind or clench my teeth |
56.9 |
15.9 |
17.5 |
8.7 |
1.0 |
|
5. I have problems with diarrhoea and/or constipation |
25.1 |
41.0 |
23.2 |
10.2 |
0.5 |
|
6. I need help in performing my daily activities |
68.2 |
14.9 |
10.8 |
5.1 |
1.0 |
|
7. I am sensitive to bright lights |
46.2 |
23.1 |
22.1 |
6.0 |
2.6 |
|
8. I get tired very easily when I am physically active |
9.7 |
20.5 |
31.3 |
27.7 |
10.8 |
|
9. I feel pain all over my body |
34.4 |
29.2 |
23.1 |
9.2 |
4.1 |
|
10. I have headaches |
6.2 |
25.1 |
27.2 |
37.9 |
3.6 |
|
11. I have discomfort in my bladder and/or burning when I urinate |
75.4 |
19.5 |
4.1 |
1.0 |
0 |
|
12. I do not sleep well |
12.8 |
23.6 |
31.3 |
20.0 |
12.3 |
|
13. I have difficulty concentrating |
19.5 |
28.7 |
34.8 |
14.4 |
2.6 |
|
14. I have skin problems |
28.7 |
32.3 |
16.9 |
15.9 |
6.2 |
|
15. Stress makes my physical symptoms get worse |
23.6 |
15.9 |
21.5 |
28.2 |
10.8 |
|
16. Depression |
15.4 |
32.3 |
32.3 |
16.4 |
3.6 |
|
17. I have low energy |
7.2 |
19.4 |
33.9 |
30.8 |
8.7 |
|
18. I have muscle tension in my neck and shoulders |
7.7 |
11.8 |
16.4 |
43.1 |
21.0 |
|
19. I have pain in my jaw |
72.9 |
12.8 |
9.7 |
4.1 |
0.5 |
|
20. Certain smells make me feel dizzy or nauseated |
45.1 |
20.0 |
20.5 |
10.8 |
3.6 |
|
21. I have to urinate frequently |
46.7 |
24.6 |
16.4 |
9.2 |
3.1 |
|
22. My legs feel uncomfortable |
37.4 |
20.0 |
21.0 |
18.0 |
3.6 |
|
23. I have difficulty remembering things |
18.0 |
34.4 |
30.7 |
15.4 |
1.5 |
|
24. I suffered trauma as a child |
54.9 |
23.0 |
18.5 |
2.6 |
1.0 |
|
25. I have pain in my pelvic area |
34.4 |
20.5 |
24.1 |
13.3 |
7.7 |
Test-retest reliability
Analysis of test — retest reliability is shown in Table 3. We obtained good test-retest reliability for the entire questionnaire (ICC2.1 = 0.89) and moderate-to-good reliability for its items.
|
CSI statements |
ICC |
95% CI |
|
1. I feel tired and unrefreshed when I wake from sleeping |
0.65 |
0.44—0.78 |
|
2. My muscles feel stiff and anchy |
0.61 |
0.43—0.73 |
|
3. Anxiety attacks |
0.74 |
0.63—0.82 |
|
4. I grind or clench my teeth |
0.80 |
0.71—0.86 |
|
5. I have problems with diarrhoea and/or constipation |
0.77 |
0.67—0.84 |
|
6. I need help in performing my daily activities |
0.77 |
0.68—0.84 |
|
7. I am sensitive to bright lights |
0.73 |
0.61—0.81 |
|
8. I get tired very easily when I am physically active |
0.67 |
0.54—0.77 |
|
9. I feel pain all over my body |
0.57 |
0.41—0.69 |
|
10. I have headaches |
0.64 |
0.51—0.75 |
|
11. I have discomfort in my bladder and/or burning when I urinate |
0.57 |
0.42—0.69 |
|
12. I do not sleep well |
0.62 |
0.47—0.73 |
|
13. I have difficulty concentrating |
0.71 |
0.59—0.80 |
|
14. I have skin problems |
0.79 |
0.69—0.86 |
|
15. Stress makes my physical symptoms get worse |
0.84 |
0.76—0.89 |
|
16. Depression |
0.71 |
0.58—0.80 |
|
17. I have low energy |
0.58 |
0.36—0.72 |
|
18. I have muscle tension in my neck and shoulders |
0.67 |
0.52—0.77 |
|
19. I have pain in my jaw |
0.74 |
0.63—0.82 |
|
20. Certain smells make me feel dizzy or nauseated |
0.77 |
0.68—0.84 |
|
21. I have to urinate frequently |
0.79 |
0.70—0.87 |
|
22. My legs feel uncomfortable |
0.74 |
0.63—0.82 |
|
23. I have difficulty remembering things |
0.70 |
0.57—0.79 |
|
24. I suffered trauma as a child |
0.75 |
0.65—0.83 |
|
25. I have pain in my pelvic area |
0.75 |
0.65—0.83 |
|
Total score |
0.89 |
0.65—0.95 |
Factor analysis
Factor analysis revealed 6 factors including 3 parameters with clear structure and good internal consistency. The results of factor analysis are presented in Table 4.
|
Factors CSI statements |
Factor 1. Emotional distress |
Factor 2. Fatigue, tension, pain |
Factor 3. Mental dysfunction and hypersensitivity |
Factor 4. Headache and facial pain |
Factor 5. Problems with urination |
Factor 6. Need for help |
|
1. I feel tired and unrefreshed when I wake from sleeping |
— |
0.670 |
— |
— |
— |
— |
|
2. My muscles feel stiff and anchy |
— |
0.661 |
— |
— |
— |
— |
|
3. Anxiety attacks |
0.744 |
— |
— |
— |
— |
— |
|
4. I grind or clench my teeth |
0.524 |
— |
— |
0.457 |
— |
— |
|
5. I have problems with diarrhoea and/or constipation |
— |
— |
0.547 |
— |
— |
— |
|
6. I need help in performing my daily activities |
— |
— |
— |
— |
— |
0.635 |
|
7. I am sensitive to bright lights |
— |
— |
0.552 |
— |
— |
— |
|
8. I get tired very easily when I am physically active |
— |
0.678 |
— |
— |
— |
— |
|
9. I feel pain all over my body |
— |
0.660 |
— |
— |
— |
— |
|
10. I have headaches |
— |
— |
— |
0.481 |
— |
— |
|
11. I have discomfort in my bladder and/or burning when I urinate |
— |
— |
— |
— |
0.539 |
— |
|
12. I do not sleep well |
— |
0.489 |
— |
— |
— |
— |
|
13. I have difficulty concentrating |
0.574 |
— |
0.546 |
— |
— |
— |
|
14. I have skin problems |
0.549 |
— |
— |
— |
— |
— |
|
15. Stress makes my physical symptoms get worse |
0.668 |
— |
— |
— |
— |
— |
|
16. Depression |
0.694 |
— |
— |
— |
— |
— |
|
17. I have low energy |
0.479 |
0.620 |
— |
— |
— |
— |
|
18. I have muscle tension in my neck and shoulders |
— |
0.509 |
— |
— |
— |
— |
|
19. I have pain in my jaw |
— |
— |
— |
0.786 |
— |
— |
|
20. Certain smells make me feel dizzy or nauseated |
— |
— |
0.576 |
— |
— |
— |
|
21. I have to urinate frequently |
— |
— |
— |
— |
0.675 |
— |
|
22. My legs feel uncomfortable |
— |
— |
— |
— |
0.567 |
— |
|
23. I have difficulty remembering things |
— |
— |
0.662 |
— |
— |
— |
|
24. I suffered trauma as a child |
— |
— |
— |
0.481 |
0.486 |
— |
|
25. I have pain in my pelvic area |
— |
— |
— |
— |
— |
0.621 |
The first factor included statements about anxiety, depression and stress-induced worsening of symptoms (3, 15, 16). These statements carried the greatest factor load and had the strongest correlation. Also, the first factor included other 4 statements with high factor loading (4, 13, 14, 17). Internal consistency of the items of the first factor was good (Cronbach's α 0.83, inter-item correlation coefficient 0.42). We named the first factor “emotional distress” as parameter with greatest factor loading.
The second factor consisted of 7 statements regarding fatigue (1, 8), low energy (17), tension (2, 18), pain (2, 9) and sleep disturbance (12). Internal consistency of the items of the second factor was also good (Cronbach's α 0.82, inter-item correlation coefficient 0.40). We named the second factor “fatigue, stress, pain” as parameter with greatest factor loading.
The third factor with good internal consistency (Cronbach's α 0.73, inter-item correlation coefficient 0.36) consisted of statements about impaired mental functions (concentration, memory), increased sensitivity to light (7), smells (20), as well as symptoms typical for irritable bowel syndrome (5). We named the third factor "impairment of mental functions and hypersensitivity".
The fourth factor was made up of statements 4, 10, 19, 24. Three out of four (4, 10, 19) items are devoted to headache and facial pain. At the same time, the statement “I have pain in my jaw” carries different and the greatest factor loading (Table 4). However, internal consistency of the items of the fourth factor turned out to be poor (Cronbach's α 0.58, inter-item correlation coefficient 0.26).
The fifth factor comprised statements 11, 21, 22, 24. Two items (11 and 21) are related to urination disorders and carry good and very good factor loading (Table 4). The same factor (with good factor loading) included a statement related to restless leg syndrome (22). Statement 24 with similar and sufficient factor loading could be included into both the fifth and the fourth factors. Internal consistency of the items of the fifth factor was poor (Cronbach's α 0.53, inter-item correlation coefficient 0.22).
The sixth factor comprised statements 6 and 25 with good factor loading and poor internal consistency (Cronbach's α 0.42, inter-item correlation coefficient 0.29).
Minimal detectable changes for CSI were 10 points.
Discussion
The purpose of our study was to assess reliability of the Russian-language version of CSI [9] in patients with non-specific neck and back pain (cervicalgia, thoracalgia and lumbodynia). The study showed that our version of CSI is reliable in this group of patients.
Internal consistency of CSI turned out to be good (Cronbach's α 0.89) that corresponds to internal consistency of both original questionnaire and its adapted versions [15, 16]. Test-retest reliability of CSI also turned out to be good (ICC2,1 0.89) that is also consistent with the results of other authors [16]. Assessment of competitive validity revealed high correlation of CSI with NDI-RU (rS = 0.56 at p-value <0.05) and moderate (rS = 0.36 at p-value <0.05) correlation with ODI-RU.
High direct correlation of CSI with NDI-RU can be explained by similar items (headache, neck pain, concentration, sleep) in both questionnaires related to various body dysfunctions or limitation of daily activity (in NDI-RU — “self-care” item, in CSI — “need for help” item). Direct strong correlation between these two questionnaires may indicate a direct positive correlation between central sensitization and disability due to neck pain.
The same fact can explain moderate direct positive correlation between central sensitization and disability due to back pain. Indeed, both questionnaires (CSI and ODI) have similar items and statements reflecting similar dysfunctions in daily activity.
We found indirect confirmation of our results in the literature. In our sample, patients with non-specific neck pain had mild central sensitization (CSI score 32 [21; 42]). According to a systematic review, central sensitization is not typical for patients with non-specific neck pain (except for cases of pain associated with whiplash) [17]. Central sensitization in patients with non-specific neck pain is associated with certain risk factors — pain intensity (CS is 5 times more common in severe pain) and comorbid dysfunctional pain syndromes such as migraine or irritable bowel syndrome [18].
Central sensitization in our patients with lumbodynia was mild-to-moderate (CSI score 39 [31.75; 42]). According to the literature, central sensitization occurs in some patients with chronic back pain [19-22]. CSI score ≥ 40 was proposed to identify such patients [23]. In our subgroup of patients with back pain, this value was 45%.
According to our data, CSI-RU reflects the symptoms typical for central sensitization. According to factor analysis, most of the statements constitute an understandable structure and 3 main factors can be distinguished: 1) emotional dysfunction (anxiety, depression) accompanied by reduced resistance to stress; 2) tension, fatigue and pain accompanied by sleep disturbance; 3) mental dysfunction (in particular, memory and concentrating) and hypersensitivity to external stimuli.
Lower internal consistency of the fourth and fifth factors in our patients with predominant neck and back pain is most likely due to the fact that statements forming these factors reflect symptoms nonspecific for neck and back pain. Perhaps, factor loadings of these statements will be higher and internal consistency will be better in other homogeneous samples, for example, in patients with chronic pelvic or facial pain. Accordingly, further studies of psychometric properties of CSI are required to confirm these data.
It was found that CSI-RU, original questionnaire and its adapted versions have a multifactorial structure. At the same time, some statements are included in the main factors; other statements have a very low factor loading and not included in any factor. This is understandable, because such statements are not typical for those groups of patients in whom psychometric properties of the questionnaire were assessed [11, 24—26]. In particular, the statement “I have discomfort in my bladder” is unusual for patients with neck pain. Accordingly, this statement will have a low factor loading in patients with neck pain and will not be included in any factor. On the contrary, this statement can have a high factor loading in patients with interstitial cystitis and enter into the main factor.
Comparison of our results and foreign data showed that “floor and ceiling effects” were analyzed only in our study.
We found that some CSI statements have a “floor effect”, i.e. they are irrelevant for patients with nonspecific neck and back pain. Patients mark these statements as “never”. In particular, 76% of patients marked the statement “I am sensitive to bright lights” as “never”. If the study had been conducted in patients with migraine, "floor effect" for the statement "I am sensitive to bright lights" could be absent. This fact should be considered in further studies devoted to assessing such a property of CSI as sensitivity to changes throughout the treatment.
Conclusion
Thus, the basic psychometric properties (internal consistency, convergent validity, test-retest reliability) characterize CSI as a reliable tool. Other properties (factor structure, sensitivity to changes) are variable and will depend on characteristics of study patients. Most likely, these properties will differ in patients with migraine, fibromyalgia and discogenic radiculopathy. Therefore, further research is needed to assess these properties of CSI.
The authors declare no conflicts of interest.