Journal of Orthopaedic Science

, Volume 19, Issue 1, pp 15–21 | Cite as

Prevalence of neuropathic pain in cases with chronic pain related to spinal disorders

  • Toshihiko Yamashita
  • Kazuhisa Takahashi
  • Kazuo Yonenobu
  • Shin-ichi Kikuchi
Original Article



The incidence and characteristics of neuropathic pain associated with spinal disorders have not yet been fully clarified. The purpose of this study was to investigate the prevalence of neuropathic pain and the degree of deterioration of quality of life (QOL) in patients with chronic pain associated with spinal disorders who visited orthopedic outpatient clinics.


This cross-sectional study was conducted in 1,857 patients recruited from 137 medical institutions nationwide. Participants were men and women aged 20–79 years with a history of spine-related pain for at least 3 months and a visual analog scale (VAS) score of at least 30 in the previous week. Patients were screened using a neuropathic pain screening questionnaire. The degree of QOL deterioration and its correlation with the presence of neuropathic pain were assessed using the Short Form Health Survey with 36 questions (SF-36).


Overall prevalence of neuropathic pain was 53.3 %. It was relatively high in patients with cervical spondylotic myelopathy (77.3 %) and ligament ossification (75.7 %) and relatively low in those with low back pain (29.4 %) and spondylolysis (40.4 %). Only 56.9 % of patients with radiculopathy were diagnosed with neuropathic pain. Logistic regression analysis identified several risk factors, including advanced age, severe pain, disease duration of at least 6 months, and cervical lesions. In QOL assessment, physical functioning, role-physical, role-emotional, and social functioning were severely affected, and this trend was more pronounced in patients who were more likely to have neuropathic pain.


The frequency of neuropathic pain tended to be higher in patients with diseases associated with spinal cord damage and lower in patients with diseases that primarily manifested as somatic pain. A bias toward allodynia symptoms in the screening questionnaire may have resulted in the failure to diagnose neuropathic pain in some patients with radiculopathy. Poor QOL, primarily from the aspect of physical functioning, was demonstrated in patients with neuropathic pain associated with spinal disorders.


A significant number of people in the general population appear to suffer from chronic pain arising from the musculoskeletal system. According to a survey conducted by Nakamura et al. [1], 15.4 % of Japanese adults have chronic pain arising from the musculoskeletal system. The most commonly affected site (65 %) is the lumbar region, followed by neck and shoulder regions (55 % each), indicating a high incidence of chronic pain related to disorders of the spine.

Based on the mechanism of pain onset, chronic musculoskeletal pain has been classified into two types: (1) chronic nociceptive pain arising from deformation and inflammation of the bone(s), joint(s) and/or other spinal tissues; and (2) neuropathic pain caused by damage to nerve tissues per se. Spinal disorders often involve damage to nerve tissues, such as the spinal cord, nerve roots, and cauda equina, which gives rise to neuropathic pain. Neuropathic pain may also occur concurrently with nociceptive pain arising from involvement of intervertebral discs and paraspinal muscles. However, incidence and clinical features of neuropathic pain associated with spinal disorders have not yet been clearly elucidated.

Under the initiative of the Japanese Society for Spine Surgery and Related Research (JSSR), a study was conducted to determine the prevalence of neuropathic pain and the degree of quality of life (QOL) deterioration in patients with chronic pain associated with spinal disorders who visited orthopedic outpatient clinics in Japan. Results are reported herein.

Patients and methods

This was a cross-sectional study conducted on outpatients recruited from 137 Japanese medical institutions. Participant institutions were required to employ at least one board-certified spine surgeon approved by JSSR and to be an orthopedic hospital with at least 20 beds, a general hospital, or a university hospital. Participating institutions were selected from regions throughout the country: ten from Hokkaido, nine from Tohoku, ten from North Kanto, 35 from South Kanto, 11 from Tokai, seven from Hokuriku, 24 from Kansai, 13 from Chugoku-Shikoku, and 18 from Kyushu. The number of institutions was allocated according to the number of approved surgeons in each geographical area, with 10–20 patients recruited from each institution. The study was initiated in March 2010 and completed in November of the same year. The study protocol was approved by the institutional ethics committee of each participating institution.


The study participants were patients judged by spine surgeons to meet all of the following criteria:

1. Chronic pain persisting for at least 3 months

2. Spine-related pain (including in those who have undergone surgery)

3. Visual analogue scale (VAS) score of at least 30 during the previous 1 week

4. Age 20–79 years

5. Capable of filling out the questionnaire in Japanese without assistance

Patients meeting any of the following criteria were excluded from the study:

1. History of receiving nerve-block therapy in the previous 6 months

2. Pain arising from the spine, as well as from other tissues/organs

3. Being an inpatient

4. Severe paralysis

5. Incapable of giving consent due to the presence of complications

A total of 2,025 patients were recruited from 137 institutions. Of these, 168 were excluded from the study for the following reasons: not willing to provide informed consent for participation (13); history of pain for ≤3 months (36); ≥80 years (68); VAS data not available or scores ≤30 (29); data from the neuropathic pain screening questionnaire not available (22). The remaining 1,857 patients were entered into the study. Table 1 summarizes the characteristics of study participants. The patient population consisted of an almost equal number of men and women. Elderly patients predominated, particularly patients in their 70s; mean age was 63.4 years. Mean duration of illness was 47.4 months, with 203 patients having chronic pain for at least 120 months. Mean VAS pain score was 54 mm. The most commonly affected region was lumbar (1,537 cases), followed by cervical (324 cases), thoracic (108 cases), and sacral (7 cases) regions.
Table 1

Patient background characteristics

Patient variables

Corresponding statistics

Male gender (%)


Age (years)

63.4 ± 12.6

Body weight (kg)

60.4 ± 11.8

Height (cm)

159.7 ± 9.1

Body mass index (kg/m2)

23.6 ± 3.6

Current pain (VAS; mm)

54.0 ± 25.0

Duration of pain (months)

47.4 ± 62.1

 Median (min–max)

26 (3–600)

Level of the spinal disorder


324 (17.4 %)


108 (5.8 %)


1,537 (82.8 %)


7 (0.4 %)

Medical therapy



1,442 (77.7 %)


70 (3.8 %)


174 (9.4 %)


167 (9.0 %)


42 (2.3 %)


148 (8.0 %)


200 (10.8 %)

 Nerve block

553 (29.8 %)

 Physical therapy

325 (17.5 %)

 Psychological therapy

14 (0.8 %)

VAS visual analog scale, min minimum, max maximum, NSAIDs nonsteroidal anti-inflammatory drugs

Spinal disorder details are shown in Table 2. Degenerative disorders accounted for the majority (1,586) of cases, and the most common diagnosis was lumbar spinal stenosis (742 cases). Iatrogenic spinal disorder refers to residual pain present after spine surgery, such as in cases of multiple operated back and failed back surgery syndrome.
Table 2

Spinal disorders


Number (%)

Degenerative disorders


 Lumbar spinal stenosis

742 (40.0)

 Intervertebral disc disorders

358 (19.0)

 Degenerative spondylolysis

197 (10.6)


170 (9.2)

 Cervical spondylotic myelopathy

110 (5.9)

 Nerve root damage

65 (3.5)


50 (2.7)

 Low back pain

17 (0.9)

Ligament ossification

91 (4.9)

Spine/spinal cord injury

77 (4.1)

Iatrogenic spinal disorder

75 (4.0)

Spine/spinal cord tumor

38 (2.0)


15 (0.8)

Infectious spondylitis/intervertebral discitis

9 (0.5)

Other types of spondylitis

15 (0.8)

Other types of neurological disorders

11 (0.6)


13 (0.7)



Some patients were diagnosed as having more than one of these conditions; therefore, the sum of the number of patients with each diagnosis is larger than the study population


Each patient signed an informed consent form, then filled out the questionnaires regarding pain and health status. The neuropathic pain screening questionnaire, developed by Ogawa et al. [2], was used for the pain survey. When the study began, this questionnaire was the only screening tool for neuropathic pain available in Japanese. Patients’ answers to the seven-question domains were weighted and scored (Table 3). The likelihood of neuropathic pain was determined based on total score, as follows: ≥5 = highly likely to have neuropathic pain (++); 4 = likely to have neuropathic pain (+); 3 = possibility of neuropathic pain (±); ≤2 or lower = unlikely to have neuropathic pain (−). A total score ≥4 (+ and ++) was judged as representing neuropathic pain. According to an analysis conducted by Ogawa et al. [2], ratings of + and ++ have an 87.7 % sensitivity and 71.7 % specificity for the presence of neuropathic pain.
Table 3

Questions on the nature of pain






Very severe


There is a pinprick-like pain







There is an electric shock-like pain







There is a tingling burning pain







There is a pain with strong numbness







Only a light touch with clothing or cold wind causes a pain







Site of pain has decreased or increased sensation







Site of pain shows skin swelling and/or discoloration to red or purple






Logistic regression analysis was performed using factors with the potential to affect the result of judgment on the presence/absence of neuropathic pain in order to identify the risk factors for a ++ result. Variables were gender, age, severity of the current pain (VAS), duration of pain, and level of the spinal disorder.

The Short Form Health Survey with 36 items (SF-36) was used for health status. For each SF-36 subscale—physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, mental health—raw scores were calculated and converted to the 0–100 scoring system and then to the norm-based scoring [mean of the Japanese national standard is 50, with a standard deviation (SD) of 10] [3, 4, 5]. Possible correlations between assessment results for neuropathic pain and scores for individual subscales of tge SF-36 were assessed. Statistical analysis was performed using the software R 2.13.0 (R Foundation for Statistical Computing, Department of Statistics and Mathematics Wirtschafts Universitat, Wien, Austria).


Prevalence of neuropathic pain

Results of neuropathic pain assessment was + in 20.7 %, and ++ in 32.6 % of the 1,857 participants. Overall, the prevalence of neuropathic pain was 53.3 % (Fig. 1).
Fig. 1

Prevalence of neuropathic pain: ++ highly likely; + likely; ± possibility; − unlikely

Correlation between assessment of neuropathic pain and patient background factors

The incidence of neuropathic pain tended to increase with advancing age in both men and women. In particular, the incidence was lower in patients in their 20s and 30s and higher in those aged 70–74 years (Fig. 2).
Fig. 2

Correlation between neuropathic pain and age

The incidence of neuropathic pain also increased with increasing severity of ongoing pain and duration of illness. Neuropathic pain was most frequently associated with cervical spinal disorder, followed by disorders of the thoracic, lumbar, and sacral regions (Fig. 3). We found no clear correlation between assessment results and body weight or body mass index (BMI).
Fig. 3

Correlations between neuropathic pain and pain severity (a), duration (b), and level of the spinal disorder (c)

When we examined the correlation between diagnosis (type of spinal disease) and assessment results, the incidence of neuropathic pain was the highest in patients diagnosed as having cervical spondylotic myelopathy (77.3 %), followed by patients with ligament ossification (75.7 %), iatrogenic spinal disorder (68.0 %), and spine/spinal cord injury (65.0 %). On the other hand, the incidence was low in patients with low back pain (29.4 %) and spondylolysis/spondylolisthesis (40.4 %) (Fig. 4).
Fig. 4

Correlation between neuropathic pain and diagnosis

Correlation between assessment of neuropathic pain and type of medical treatment received

Nonsteroidal anti-inflammatory drugs (NSAIDs) were the most commonly used pain medications in all patients with chronic pain. Anticonvulsants and antidepressants were used at a higher frequency in the ++ group (Fig. 5). The frequency of nerve-block or physical therapy was not correlated with results of assessment of neuropathic pain.
Fig. 5

Correlation between neuropathic pain and pain relief medication used: judgment − (a), judgment ± (b), judgment + (c), judgment ++ (d). NSAIDs nonsteroidal anti-inflammatory drugs, Con anticonvulsants, Dep antidepressants, PG prostaglandins, Opi opioids, Str steroids, Oth others

Factors affecting the of judgment on presence/absence of neuropathic pain

Table 4 presents the odds ratios (OR) calculated by logistic regression analysis of factors for the presence of neuropathic pain rated as ++. As no marked differences were observed between ORs based on the simple correlation (Not adjusted, in Table 4) and those based on multiple regression analysis (Adjusted, in Table 4), these factors were considered to be independent.
Table 4

Risk factors for the presence of neuropathic pain identified by logistic regression


No. (++)

Not adjusted OR (95 % CI)

P value

Adjusted OR (95 % CI)

P value


 ≤40 years

128 (16)





 41–69 years

1,060 (338)

3.23 (1.88–5.54)


2.72 (1.56–4.75)


 70–79 years

669 (251)

4.19 (2.43–7.22)


3.75 (2.14–6.58)




920 (305)






934 (299)

0.95 (0.78–1.15)


0.90 (0.73–1.11)


Current pain score (VAS)a

 <10 mm

111 (11)





 10 to <50 mm

657 (151)

2.71 (1.42–5.19)


2.33 (1.21–4.51)


 50 to <80 mm

783 (290)

5.35 (2.82–10.13)


4.62 (2.41–8.84)


 80 to <100 mm

303 (152)

9.15 (4.72–17.74)


8.35 (4.26–16.36)


Pain duration

 <6 months

249 (52)





 ≥6 months

1,608 (553)

1.99 (1.44–2.74)


1.73 (1.23–2.43)


Level of spinal disordera,b


324 (174)

3.03 (2.37–3.88)


3.20 (2.45–4.16)



87 (35)

1.52 (0.84–2.74)


1.65 (0.89–3.05)



1,439 (392)





P value



Pseudo R2



OR odds ratio, CI confidence interval, VAS visual alanog scale, ++; highly likely to have neuropathic pain, Not adjusted OR calculated by each category, Adjusted OR calculated using all the listed all covariates, pseudo-R2 Cragg and Uhler’s pseudo-R2

aData on gender, current pain score, and level of spinal disorder were lacking in 3, 3, and 7 patients, respectively

bCervical level, patients with cervical-level disorder with/without disorder at other levels; thoracic level, patients with thoracic-level disorder with/without lumbar/sacral level disorder but no cervical-level disorder; lumbar/sacral level, patients with only lumbar/sacral-level disorder

Gender had no effect on assessment results of neuropathic pain. Multiple regression analysis by age showed that the OR was 2.72 for the age group 41–69 years and 3.75 for those aged ≥70 years vs the age group ≤40 years; thus, the ratio increased with age. As for pain severity, OR for VAS scores 10 to <50 vs <10 was 2.33; furthermore, OR doubled for VAS scores 50 to <80 and doubled again for VAS scores ≥80. OR was 1.73 for pain duration ≥6 months vs <6 months. OR for cervical-level involvement vs lumbar/sacral-level involvement was 3.20.

Correlation between degree of QOL deterioration and neuropathic pain severity

All participants in had scores below the Japanese standard value of 50 on all SF-36 subscales. By subscale, physical functioning, role-physical and role-emotional), and social functioning were more severely affected, a trend that was more pronounced in patients who were more likely to have neuropathic pain (Fig. 6).
Fig. 6

Correlation between neuropathic pain and degree of quality of life (QOL) deterioration. PF physical functioning, RP role-physical, BP bodily pain, GH general health, VT vitality, SF social functioning, RE role-emotional, MH mental health


According to the International Association for the Study of Pain, neuropathic pain is defined as pain caused by a lesion or disease of the somatosensory nervous system. A representative example of neuropathic pain associated with spinal disorders includes pain caused by compression or damage to the spinal cord or the nerve roots.

In their review, Sadosky et al. [6] reported the incidences of neuropathic pain in various diseases as follows: herpes zoster (postherpetic neuralgia) 7–27 %; diabetes (diabetic neuropathy) 9–22 %; cerebral stroke (poststroke pain); 8–11 %; spinal cord injury (postspinal cord injury pain) 10–80 %. Compared with their data, the incidence of neuropathic pain of 53.3 % in our survey of patients with spinal disorders is relatively high. However, our participants were limited to those who had chronic pain for at least 3 months, and the incidence may have been lower if we had also included spinal disorder patients who did not have a history of pain for such a long period of time.

Logistic regression analysis was carried out in this study to identify risk factors for the presence of neuropathic pain rated as ++. Specificity is important for a logistic analysis in order to identify factors predictive of neuropathic pain. Ogawa et al. [2] reported that when the screening score was ≥4 (i.e., corresponding to ratings of + or ++), specificity and sensitivity were 71.7 % and 87.7 %, respectively, which is appropriate for screening but insufficient for a logistic regression analysis. In contrast, the specificity of 89.1 % for a score ≥5 (i.e., corresponding to a rating of ++) is appropriate to conduct a logistic regression analysis. In our study, variables used in the logistic regression analysis were gender, age, severity of current pain (VAS), duration of pain, and level of the spinal disorder. The basic analysis showed no strong correlations among these variables. Logistic regression analysis identified the following risk factors for the presence of neuropathic pain: advanced age, severe pain, disease duration of at least 6 months, and a cervical-level lesion. These results indicate that the more severe the damage to the nerve tissue, the higher the incidence of neuropathic pain.

In the survey of medical treatments, NSAIDs were by far the most frequently used medications overall (~77 %). NSAIDs are effective for relieving nociceptive pain but are considered to be ineffective, in principle, for treating neuropathic pain. Anticonvulsants, which are effective in treating neuropathic pain, were used slightly more often in the ++ group than in the other groups, but the proportion was still only about 16 %. In October 2010, when our study was about to be completed, the anticonvulsant drug pregabalin was approved in Japan for insurance-covered treatment of peripheral neuropathic pain. Therefore, its may have increased after this study in the patients who were evaluated as having neuropathic pain.

When data were summarized by disease, the incidence of neuropathic pain was low in conditions such as low back pain and spondylolysis, which, as the names themselves indicate, represent spinal-tissue-related pain, i.e., somatic pain. The incidence was high in diseases such as cervical spondylotic myelopathy and spine/spinal cord injury, the names of which suggest neuropathic conditions. However, only 57 % of patients with nerve root damage were diagnosed as having neuropathic pain. According to the above-mentioned definition by the International Association for the Study of Pain, patients with nerve root damage (radiculopathy) should always be diagnosed as having neuropathic pain, because the name of the disease itself suggests it. This discrepancy between disease name and diagnosis results can be attributed to the characteristics of the neuropathic pain screening questionnaire we used in the survey. The questionnaire comprised seven components that were biased toward allodynia symptoms, such as “Only a light touch with clothing or cold wind causes a pain,” and “There is a tingling, burning pain.” However, in actual clinical situations, allodynia symptoms are less frequently seen in cases of nerve root damage associated with intervertebral disc herniation, etc. In the future, we anticipate the development of a diagnostic tool that can be applied for neuropathic pain both with and without allodynia.

On the other hand, approximately 30 % of patients with low back pain were diagnosed as having neuropathic pain, although low back pain manifests primarily as somatic pain via nociceptive mechanisms. Therefore, this result also indicates a limitation of this screening tool in that patients with nociceptive pain may receive a diagnosis of neuropathic pain if allodynia-like symptoms are present. However, low back pain may be associated with neuropathic pain in some cases when some kind of damage to the peripheral nervous tissues in the lumbar spine and surrounding areas is present or neuroplastic changes have developed in the synapses of the dorsal horn of the spinal cord as a result of prolonged afferent nociceptive signals (central sensitization). Freynhagen et al. [7] reported that 37 % of patients with chronic low back pain had some factor suggestive of neuropathic pain. Because our study assessed only 17 patients with low back pain, additional studies with a larger sample size of participants with low back pain are warranted.



The study was conducted with the collaboration of the following executive committee members in regional blocks nationwide. Manabu Ito (Hokkaido), Yushin Ishii, Shin-ichi Konno (Tohoku), Atsushi Seichi (North Kanto), Yasuaki Tokuhashi, Kazuhiro Chiba (South Kanto), Yukihiro Matsuyama, Hideo Hosoe (Tokai), Norio Kawahara, Yoshiharu Kawaguchi (Hokuriku), Hiroaki Nakamura, Motoki Iwasaki (Kinki), Toshihiko Taguchi, Toru Hasegawa, Shinichiro Taniguchi (Chugoku-Shikoku), Hiroaki Konishi (North Kyushu), and Kazunori Yone (South Kyushu). This study was sponsored and funded by the Pfizer Japan Inc.


  1. 1.
    Nakamura M, Nishiwaki Y, Ushida T, Toyama Y. Prevalence and characteristics of chronic musculoskeletal pain in Japan. J Orthop Sci. 2011;16:424–32.PubMedCentralPubMedCrossRefGoogle Scholar
  2. 2.
    Ogawa S. Development of new screening questionnaire to identify neuropathic components in Japanese patients with chronic pain. Pain Clin. 2010;31:1187–94 (in Japanese).Google Scholar
  3. 3.
    Fukushima S, Bito S, Green J, Hsiao A, Kurokawa K. Translation, adaptation, and validation of the sf-36 health survey for use in Japan. J Clin Epidemiol. 1998;51:1037–44.CrossRefGoogle Scholar
  4. 4.
    Fukuhara S, Ware JE, Kosinski M, Wada S, Gandek B. Psychometric and clinical tests of validity of the Japanese SF-36 health survey. J Clin Epidemiol. 1998;51:1045–53.PubMedCrossRefGoogle Scholar
  5. 5.
    Fukuhara S, Suzukamo Y. Manual of SF-36v2 Japanese version: Institute for Health Outcomes and Process Evaluation Research, Kyoto. 2004. (in Japanese).Google Scholar
  6. 6.
    Sadosky A, McDermott AM, Brandenburg NA, Strauss M. A review of the epidemiology of painful diabetic peripheral neuropathy, postherpetic neuralgia, and less commonly studied neuropathic pain conditions. Pain Pract. 2008;8:45–56.PubMedCrossRefGoogle Scholar
  7. 7.
    Freynhagen R, Baron R, Gockel U, Tolle TR. Pain DETECT: a new screening questionnaire to identify neuropathic components with back pain. Curr Med Res Opin. 2006;22:1911–20.PubMedCrossRefGoogle Scholar

Copyright information

© The Japanese Orthopaedic Association 2013

Authors and Affiliations

  • Toshihiko Yamashita
    • 1
  • Kazuhisa Takahashi
    • 2
  • Kazuo Yonenobu
    • 3
  • Shin-ichi Kikuchi
    • 4
  1. 1.Department of Orthopaedic SurgerySapporo Medical University School of MedicineSapporoJapan
  2. 2.Department of Orthopaedic Surgery, Graduate School of MedicineChiba UniversityChibaJapan
  3. 3.Graduate School of Health Care SciencesJikei InstituteOsakaJapan
  4. 4.Fukushima Medical UniversityFukushimaJapan

Personalised recommendations