Research Review By Dr. Demetry Assimakopoulos©


Download MP3

Date Posted:

January 2016

Study Title:

Somatosensory nociceptive characteristics differentiate subgroups in people with chronic low back pain: a cluster analysis


Rabey M, Slater H, O’Sullivan, Beales D & Smith A

Author's Affiliations:

School of Physiotherapy and Exercise Science, Curtin University, Perth, Australia.

Publication Information:

Pain 2015; 156: 1874–1884.

Background Information:

We as manual therapists are frequently tested both clinically and emotionally by patients suffering from chronic low back pain (CLBP). Unfortunately, the evidence-informed treatment outcomes for CLBP are unpredictable, likely due to the underlying heterogeneity of CLBP patients/populations. Subgrouping low back pain populations to individually tailor treatment has proven to improve treatment outcomes for both acute and subacute LBP (1). However, such subgrouping has not been performed comprehensively or entirely successfully in a CLBP population.

Bedside sensory testing is often a component of a comprehensive CLBP assessment. Several factors, some of which fall in the biopsychosocial category, are capable of influencing sensory testing, as well as general pain sensitivity factors – these include age, gender, mood, catastrophizing, fear, activity level and sleep (2-8). In light of this, the authors sought to identify if a large CLBP cohort could be sub-grouped based on clinical bedside sensory findings discovered during physical examination. They subsequently attempted to profile the subgroups according to other biopsychosocial variables in hopes of offering further insight into the complexity of CLBP.

Pertinent Results:

The chronic LBP study subjects were separated into 3 clusters, according to their somatosensory nociceptive characteristics, age and gender:
  1. Cluster 1: characterized by average-to-high pressure and thermal pain sensitivity both in the local lumbar spinal region and peripherally. These patients had superficial and deep tissue, multisensory nociceptive hypersensitivity. A number of the individuals assigned to this category also showed temporal summation (see explanation below), suggesting central sensitization as a likely mechanism.
  2. Cluster 2: characterized by average-to-high pressure pain sensitivity and normal/slightly below-average thermal pain sensitivity. The authors were unclear as to why these individuals had high pressure pain sensitivity, but hypothesize mechanisms such as peripheral sensitization of deep tissue afferents, and acidosis.
  3. Cluster 3: characterized by decreased pressure and thermal sensitivity, similar to a control population.
Depression and sleep disturbance were more prevalent in clusters 1 and 2. Interestingly, some experimental data has shown that thermal and pressure hypersensitivity can be induced in otherwise healthy participants who are sleep deprived. Additionally, the individuals in cluster 1 were less physically active, which has also been linked to pressure sensitivity.

The analysis failed to reveal any relationship between nociceptive sensitivity and pain threshold, pain intensity or disability. Hypersensitivity has been linked to poor outcomes/prognosis in other datasets. However, prognosis was not integrated into this analysis.

Clinical Application & Conclusions:

This unique analysis revealed that a large chronic LBP cohort can be subgrouped based on clinical sensory findings. Cluster 1 featured subjects that were hypersensitive to all cutaneous modalities and, due to the presence of temporal summation, were likely centrally sensitized to pain. Cluster 2 featured subjects who were only hypersensitive to pressure. Cluster 3 had subjects that had a normal sensory examination, similar to controls. The individuals in clusters 1 and 2 were also sleep deprived, depressed and inactive, pointing to a relationship among these variables and some manifestation of CLBP.

This study’s analysis is exciting, because the results can be taken directly into the clinic in the following ways:
  1. Performing a comprehensive sensory examination is important when faced with a chronic pain patient (including those with LBP) – it should not be overlooked or forgotten.
  2. Screen your patients for other psychosocial issues, such as depression, anxiety, poor sleep and inactivity. Individuals who fall into Clusters 1 and 2 likely require more than just manual therapy and condition specific rehabilitation. Specifically, they should be counselled on how to improve their sleep quality. Additionally, you and your patient should set physical activity goals to improve their fitness level and day-to-day function. Setting goals such as walking down the hall for 2 minutes multiple times per day, taking their dog for an extra walk, or walking to the grocery store when they would normally drive, are all ways a person can improve their physical activity levels. Get creative!
  3. Screen your chronic pain patients for depression and refer accordingly. Working with a psychiatrist, psychologist or social worker can help your patient confront barriers to recovery that you and, more importantly, they were unaware of.
In closing, the assessment of CLBP should include a sensory examination, as it can reveal important information about a patient’s central and peripheral nervous systems. Secondly, CLBP management should be patient-specific, and often requires an integrarted, biopsychosocial approach.

Study Methods:

The authors conducted a cross-sectional study. The inclusion criteria were as follows:
  • Male or female, aged 19-70
  • Axial, back dominant, low back pain > 3 months duration
  • Score of ≥ 2/10 on the numeric pain rating scale (NRS, 0-10)
  • ≥5 on the Roland-Morris Disability Questionnaire (RMDQ)
The exclusion criteria were:
  • Previous extensive spinal surgery (greater than a single-level fusion, instrumentation or discectomy)
  • Spinal surgery of any kind within the last 6 months
  • Serious spinal pathology, such as cancer, inflammatory arthropathy, acute vertebral fracture, etc.
  • Diagnosed neurological disease
  • Bilateral pain at the dorsum of the wrist or hand
  • Pregnancy
  • Inability to understand English
Sensory Testing Procedures:
  • A combination of bedside and laboratory quantitative sensory testing (QST) were performed on every patient in the prone position.
  • Sensation was tested first on the dorsal wrist, and then over the most painful area of the low back.
  • Two point discrimination (TPD): performed only over the region of maximal lumbar spinal pain, with a plastic calliper ruler. Each time the subject was touched, they were asked to indicate if they perceived one or two calliper points. The distance between the two calliper points narrowed with repeated stimulation. Testing continued until the patient could only perceive one point touching their back, rather than two. Trick stimuli were used to decrease the chances of guessing, by randomly applying the callipers at a distance that was out of sequence or by contacting the patient with only one point.
  • The Standardized Evaluation of Pain:
    1. Patients were stimulated with a 1 cm wide make-up brush over the wrist and then the most painful part of the low back, and asked if the two sensations felt the same. If brushing in the low back elicited pain, they were asked to rate the pain on the NPS scale.
    2. Detection of vibration: patients were tested using a 128 Hz tuning fork, placed in both the area of maximal lumbar spine pain and at the dorsal wrist. Subjects were asked to compare the vibration sensation in both regions.
    3. Detection of pinprick sensation: conducted over both the low back and wrist. Enough pressure was applied onto a toothpick to elicit pain, but not enough to indent the skin. If they denied pain on 3-4 locations, they were deemed hypo-algesic. If the sensation in the low back was more painful than the wrist, they were deemed hyper-algesic, and asked to rate the pain on the NPS.
    4. Detection of temporal summation: performed using monofilament stimulation. A positive response was defined as an initial perception of non-noxious stimulation, which became noxious with repeated testing.
  • Mechanical Detection Threshold: This was the lowest mechanical force, applied to skin that the subject could detect, using monofilaments. If the subject could detect the light touch, gradually smaller filaments were used until no response was elicited. The test was terminated if they could not feel the light touch at the starting point.
  • Pressure Pain Threshold (PPT): The point at which the sensation of pressure changes to a sensation of both pressure and pain. This was performed by gradually increasing the skin pressure contact of a pressure algometer.
  • Heat and Cold Pain Threshold (HPT/CPT): HPT was the temperature at which a sensation of warmth becomes the sensation of heat and pain. Testing began at 32°C, and increased by 1°C/sec until the patient reached their pain threshold, or until the upper limit of 50°C was reached. CPT was the temperature at which the sensation of cold became a sensation of cold and pain. Testing started at 32°C, at decreased by 1°C/sec until they reached their threshold, or until the lower limit of 4°C.
  • Conditioned Pain Modulation: The testing stimulus (TS) was applied to the most painful region in the low back. Pressure using an algometer was applied, and gradually increased until it reached pain intensity of 6/10 on the NRS. The conditioning stimulus (CS) was noxious heat applied to the dorsal hand, starting at 40°C and increased by 1°C until the stimulus was too great to tolerate. The TS over the low back was then re-measured, after the CS was applied to the hand. The authors were specifically looking for whether a pain intensity of 6/10 was elicited more quickly, after the CS.
  • Demographic information, psychosocial history, pain intensity, back pain disability (RMDQ), pain duration, opioid analgesic usage, depression and anxiety (Depression Anxiety Stress Scale 21), stress levels, kinesiophobia (FABQ), catastrophization (PAS), back awareness (Fremantle Back Awareness Questionnaire), health comorbidities, pain sites, sleep quality (Pittsburgh Sleep Quality Index), and physical activity levels were elicited through history taking, or measured using reliable questionnaires.
  • Latent Class Analysis (LCA) was performed to examine inherent relationships between variables within the dataset, and to determine if the dataset could be clustered based on individual responses to multimodal sensory testing.
  • Between-cluster differences in quantitative sensory testing variables were ascertained using linear regression analysis for normally distributed variables, ANOVA for variables with skewed data and X2 for dichotomous data.
  • Biospsychosocial and demographic variables were profiled using univariate multimodal logistic regression.

Study Strengths / Weaknesses:

  • Use of the latent cluster analysis (LCA) allowed for clustering of the study subjects, depending on how they reacted to cutaneous stimulation. Clustering like this has never been done before in this type of patient population to our knowledge.
  • The researchers used a combination of bedside and laboratory sensory testing, which allowed for subgroup derivation. The inclusion of the bedside testing allows for translation of the results into daily practice.
  • The study included individuals suffering from predominant back pain, and not leg pain.
  • The LCA requires all included variables (i.e. sensory tests) to have clear, definable maximum likelihood ratios. There was difficulty obtaining clear maximum likelihood ratios for the bedside and laboratory sensory testing, which might have affected results.
  • There is risk of type 1 error (detecting an effect that is not present), due to the overwhelming number of comparisons.

Additional References:

  1. Vibe Fersum K, O’Sullivan P, Skouen J, et al. Efficacy of classification-based cognitive functional therapy in patients with nonspecific chronic low back pain: a randomized controlled trial. Eur J Pain 2013; 17: 916–928.
  2. Neziri AY, Scaramozzino P, Andersen OK, et al. Reference values of mechanical and thermal pain tests in a pain-free population. Eur J Pain 2011; 15: 376–383.
  3. Pfau DB, Krumova EK, Treede RD, et al. Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): reference data for the trunk and application in patients with chronic postherpetic neuralgia. Pain 2014; 155: 1002–1015.
  4. Klauenberg S, Maier C, Assion HJ, et al. Depression and changed pain perception: hints for a central disinhibition mechanism. Pain 2008; 140: 332–343.
  5. Campbell CM, Kronfli T, Buenaver LF, et al. Situational versus dispositional measurement of catastrophizing: associations with pain responses in multiple samples. J Pain 2010; 11: 443–453.e442.
  6. Kelley NJ, Schmeichel BJ. The effects of negative emotions on sensory perception: fear but not anger decreases tactile sensitivity. Front Psychol 2014; 5: 942.
  7. Ellingson L, Colbert L, Cook D. Physical activity is related to pain sensitivity in healthy women. Med Sci Sports Exerc 2012; 44: 1401–1406.
  8. Kundermann B, Spernal J, Huber MT, et al. Sleep deprivation affects thermal pain thresholds but not somatosensory thresholds in healthy volunteers. Psychosom Med 2004; 66: 932–937.